ML20099D625

Analysis of the Lasalle Unit 2 Nuclear Power Plant: Risk Methods Integration and Evaluation Program (Rmiep).External Event Scoping Quantification

ML20099D625
Person / Time
Site:	LaSalle
Issue date:	07/31/1992
From:	Banon H, Ravindra M NTS/SMA, INC., SANDIA NATIONAL LABORATORIES
To:	NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
References
CON-FIN-A-1386 NUREG-CR-4832, NUREG-CR-4832-V07, NUREG-CR-4832-V7, SAND92-0537, SAND92-537, NUDOCS 9208060219
Download: ML20099D625 (161)

Text

{{#Wiki_filter:- -.. - - NUREG/CR-4832 SAND 92-0537 Vol. 7 Ana:ysis of tae LaSalle Unit 2 Nuclear Power Plant: Risk Methoc.s Integration anc Evaluation Program (RMIEP) External Event Scoping Quantificatian Prepared by ' M. K. Ravindra, H Banon NTS/ Structural Mechanics Associates Sandia National Laboratories i l Prepared for U.S. Nuclear Regulatory Commission

1. 88 288M 8!888174 P

PDR

.m._.. AVAILABluTY NOTICE Availabety of Reference Matonals Csted in NRC Pubiscations I Most documents cited in NRC publications will be aval:able from one cf the following sources: 1. The NRC Public Document Room,2t20 L Street, NW, Lower Level, Washington, DC 20555 l 2. The Superintendent of Documents, U.S. Government Printing Office, P.O. Box 37082. Washington, DC 20013-7082 3. The National Techntcal information Service, Springheld, VA 22161 Atthough the listing that follows represents the majority of documents cited in NRC publications, it is not Intended to be exhaustive. ' Referer.ced documents available for inspection and copying for a fee from the NRC Pubuc Document Roorn include NRC correspondence and internal NRC memoranda: NRC buitetins, circularsc information notices, [ ' inspection and Investigation notices; licenses event reports; vendor reports and corf espondence; Commis-sion papers; and applicant and licenses documents and correspondence. The following documents in the NUREG series are available for purchase from the GPO Sales Progra' l. formal NRC staff and contractor reports NRC-sponsored conference proceedings, international agreement reports, grant publications and NRC booklets and brochures, Also avahable are regulatory guides, NRC regutations in the Code of Federal Reguladons, and Nuclear Regulatory Comrnission Issuances, i-Documen*,3 available from the National Technical information Service include NUREG-series reports and technical reports prepared by other Federal agencies and reports prepared by the Atomic Energy Commis-sion, forerunner agency to the Nuclear Regulatory Commission. Documents availabla from public and special technical hbraries include a!! open literature items, st.A as l books, journal articles..and transactions. Federal Register notices, Federal and State legislation, and con. grossional reports can usualty be obtained from these libraries, l Documents such as theses, dissertations, foreign reports and translations, and non-NRC conference pro-ceedings are available for purchase from the organization sponsoring the pubiscation cited. Single copies of NRC draft reports are available free, to the extent of suppt6 upon written toquest to the Office of Administration, Distributton and Mal Services Section, U.S. Nuclear Reguiatory Commission, Washington, DC 20555.- I. Copies of Industry codes and standards used in a substantive manner in the NRC regulatory process are mantained at the NRC Lfbrary; 7920 Norfolk Avenue Dethesda, Marylano, for use by the pub!Ic. Codes and standards are t.sually copyrighted and may be purchased from the originat;ng organization or, if they are American National Standards, from the American National Standards Institute,1430 Broadway, New York, NY.10018. I DISCLAIMER NOTICE This report was prepared as an account of work sponsored by an agency of the United States Govemment. Noither the United States Government nor any agency thereof, or any of their employees, makes any warranty, expressmi or implied, or assumes any legal liabmty of responsibility for any third party's use, or the results of [j such use, of any information, apparatus, product or process disclosed in t*ts report, or represents that its use by such third party would not ir,fringo privately owrmd rights. l 1-I I: j.

NUREG/CR-4832 4 SAND 92-0537 - Vol. 7= RX Analysis of the LaSalle Unit 2 l Nuclear Power Plant:- 1 Risk Methods Integration and Evaluation Program (RMIEP) LExternal Event Scoping Quantification - Manuscript Completed: February 1985 Date Published: July 1992 Prepared by = ' M, K. Ravindrai, H. Banon2 NTS/ Structural Mechanics Associates - - 5160 Birch Strect: Newport Beach, CA 92660 . Under Contract to: Sandia National 1;tboratories -

Alouquerque, NM 87185 -

Prepared for-Division of Safety Issue Reso_lution - Office.of Nuclear Regulatory Research ' U.S. Nuclear Regulatory Commission l Washington, DC 20555 ' NRC FIN A1386 1 Currently with EQE, Inc., Costa Mesa, CA 2 Currently _with Exxon Production'Research. Houston, TX l-L a .-maw--- m- --y g+ y a-- wr y e w-

ABSTRACT This report is a description of the scoping quantification study which selected the external events to be included in the -Level III-PRA of the LaSalle County Nuclear Generating Station Unit II. The study was. performed by NTS/ Structural Mechanics R Associates (SMA) for Sandia National Laboratories as part of the Level I analysis being performed - by the Risk Methods Integration and Evaluation Program (RMIEP). The methodology used is described in detail in a companion report, NUREG/CR-4839. In this report, we describe the process for selecting the external events, the screening analysis, and the detailed bounding calculations for those events not eliminated in the screening analysis. As a result of this analysis, it was concluded that only internal flooding, internal fire, and seismic events were potentially significant at LaSalle. Detailed analyses were performed for each of these and are reports in NUREG/CR-4832, Volumes 10, 9, and 8, respectively. l' lii/iv

CONTENTS l Pace ABSTRACT iii/iv FOREWORD xiii x ACKNOWLEDGEMENT xvii

1. 0 - INTRODUCTION 1-1 1.1 Background ~

1-1 1.2 Objective 1-2 1.3 Outline.and Contents of Report 1-2 2.0 EXTERNAL EVENT METHODOLOGY 2-1 2.1 Review of General Techniques and Mathematical Models 2-1 2.2 Identification of Potential External Events 2-1 2.3 Initial Screening of Events 2-2 2.4 Bounding Analysis. 2-3 2.5 Detailed Analysis 2-5 2.6 Information 2-5 2.7 Technical Quality 2-6 2.8 Uncertainty Analysis 2-6

3. 0-SCOPING QUANTIFICATION STUDY 3-1

-3.1 Plant Description 3-1 3.1.1 Site, Terrain, Meteorology 3-2 3.1.2 Site Visit 3-3 3.2 Initial Screening of External' Events 3-13 3.3 Screening of External Events Based on FSAR Information 3-24 3.'3.1 Accidents in Industrial and Military Facilities 3-24

3.3.2 Pipeline Accidents 3-26 3.4

' Bounding-Analysis 3-30 3.4.1 Model, Uncertainty and Acceptance / Rejection Criterion 3-30 3.4.2 Aircraft Impact 3-31 3.4.2.1 FSAR Information 3-32 ~ 3.4.2.2 Update on FSAR Information 3-33 t-V

CONTENTS (Continued)__ Pace 3.4.2.3 Aircraft Impact Bounding Analysis 3-33 a 3.4.2.4 Aircraft Impact Uncertainty Analysis 3-36 3.4.3 Winds and Tornadoes 3-38 3.4.3.1 Plant Design Criteria 3-38 3.4.3.2 Seismic Category I Structures 3-39 3.4.3.2.1 Tornado Loads 3-39 3.4.3.2.1.1 Characteristics of Tornadoes 3-40 3.4.3.2.1.2 Tornado Occur-rence Rate 3-41 3.4.3.2.1.3 Tornado Hazard Model 3-42 3.4.3.2.2 Tornado-Gener-ated, Missiles 3-48 3.4.3.3 Nonseismic Category I Structures 3-52 3.4.3.3.1 Design Capacity 3-52 3.4.3.3.2 Exceedence Probability 3-53 3.4.3.4 Uncertainty Analysis for Winds and Tornadoes 3-54 3.4.3.5 Conclusions 3-57 3.4.4 Transportation-Accidents 3-57 3.4.4.1 Chemical Explosions 3-53 3.4.4.2 Toxic Chemicals 3-62 3.4.5 Turbine Missiles 3-63 3.4.5.1 Historical Background 3-63 3.4.5.2 Probabilistic Methodology 3-6L 3.4.5.2.1 Probability of Turbine Failure P1 3-66 vi

h 3 i T' CONTENTS (Concluded) 2A9.fi 3.4.5.2.2 Probability of Missile Strike P2 3-67 3.4.5.2.3 Probability of Barrier Damage P3 3-69 i 3'4.5.3-FSAR Analysis 3-71 3.4.5.4 Recent Turbine Missile Issues 3-72 3.4.5.4.1 Stress Corrosion Cracking Issues 3-72 3.4.5.4.2 Refinements-in Turbine Missile -Risk Analysis 73 3.4.5.5 Conclusion-3-73 3.4.6 External Flooding 3-74 3.4.6.1 Illinois River. 3-74 3.4.6.2 Cooling Lake 3-75 3.4.6.3 Local Preci'pitation 3-75 3.5 Events Requiring Detailed-PRA 3-118 4.0-

SUMMARY

AND-RECOMMENDATIONS ' 4-1 -4.1-Summary 4-1 4.2 ' Recommendations 4-2 .- REFERENCES ' R, e vii 3 ,e -r=- T >=re u- ---t~ = e -~ --ww-

.-~ _ l LIST OF TABLES Table Easa 3.1-1 Code Requirements for Components and Systeas Ordered after July 1, 1974 3-6 3.2-1 Preliminary Screening of External Events for LaSalle County Station. 3-20 3.3-1 Industries with Hazardous Materials within 10 Miles of the Site 3-27 3.4-1 Commercial Airports within 20 Miles of the Site 3-77 3.4-2 Private Airstrips within 20 Miles of the Site 3-78 3.4-3 Aircraic Traffic Statistics Near the LaSalle Site for June 7,-1984 3-79 3.4-4 Annual In-Flight Crash Rates (1 Mile) 3-80 3.4-5 Annual Frequencies of Aircraft Impact for LaSalle Structures 3-81 3.4-6 Intensity, Length, Width, and Area Scales 3-82 3.4-7 Regional Tornado Occurrence - Intensity Relationships Corrected for Direct Classification Errors and Random Encounter Errors 3-83 3.4-8 Intensity-Area Relationship Including Correction and Random Encounter Errors (AIM Matrix) 3-84 3.4-9 Variation of Tornado Intensity Along Path Length and Across Path Width (VWL Matrix) 3-85 3.4-10 Intensity-Length Re73tionship Including Corrections for Direct Observation and Random Encounter Errors (LIM Matrix) 3-86 3.4-11 Variation of Intensity Along Length Based on Percentage of Length Per Tornado (VL Matrix)- 3-87 3.4-12. NRC.SRP Tornado Missiles (Standard Review 3-88 3.4-13 Plan Minimum Reinforced Concrete Thicknesses (Inches) Required to Prevent Scrubbing (NDRC and Chang's Formulas) 3-89 3.4-14 Estimates of Annual-Probability of Turbine Missile Generation 3-90 3.4-15 30-Inch Last Stage Bucket, 1200 RPM Low-Pressure-Turbine - Hypothetical Missile Data 3-91 3.4-16 Maximum 24-Hour Precipitation for Chicago 3-93 l l ix l

1 LIST OF FIGURES ,E30ure Pace 3.1 General Arrangement-Roof Plan 3-9 3.1-2 General PlantJArrangement 3-10 3.1-3 Location'of the-Site within the State of Illinois - 3-11 3.1-4 General Site Arrangement 3-12 3.4-1 Airports and Flight Patterns within 20 Miles of the Site 3-94 3.4-2 Geometry for Aircraft Impact Probabilistic Model. 3-95 3.4-3 -Tornado Risk Regionalization Scheme Proposed by WASH-1300, Markee, et al. (1975) 3-96 3.4-4 Tornado Risk Regionalization Scheme Proposed by Twisdale and Dunn (1983) 3-97 3.4-5 _ Tornado Parameters and Damage Origin Area Definition 3-98 3.4-6 Sketch of Hypothetical F4 Tornado Illustrating Variation of Intensity 3-99 3.4 ~ Tornado Hazard Curves for LaSalle Site 3-100 3.4-8 Station Locations 3-101 3.4-5 Family at Tornado Hazard Curves for the LaSalle Site with Corresponding Subjective Probabilities-3-102 3.4-10 Tornado Fragility Curves for LaSalle Including Uncertainty in Median Capacity 3-103 3.4-11 Distribution of Annual Frequency of Core Melt in LaSalle Due to Tornadoes 3-104 3 -4.12 Transportation Routes Near LaSalle County Station: 3-105 3.4-13 Radius to Peak' Incident Pressure of 1 psi 3-106 3.4-14 Probability of Flammable-Plume Ignition .Versus Plume Area at Time of Ignition 3-107 '3.4-15 Pressure Pulses from TNT 3-108 3.4-1b Free-Field Blast Wave Parameters Versus Scaled Distances for-TNT Surface Bursts (Hemispherical Charges) 3-109 3.4-17 Dynamic Load Factors. Maximum response of one-degree clastic systems-(undamped) subjected to rectangular and triangular load pules having zero rise time 3-110 3.4 Variables and Terminology Used in Calcu-lating Missile Strike Probabilities 3-111

3.4-19

-Histogram of Maxirum Daily Precipitation i for~ Chicago 3-112 l Normal Distribution Fit for Maximum Daily 3.4 20 Precipitation 3-113 x1

LIST OF FIGURES (Cont.)- Ficure Page ' 3.4-21 Lognormal Distribution Fit for Maximum Daily Precipitation 3-114 3.4. Gamma Distribution Fit for Maximum Daily Precipitation 3-115 3.4 Extreme Value Type I Distribution Fit for Maximum Daily Precipitation 3-116 3.4-24 Log-Pearson Type III Distribution Fit for Maximum Daily Precipitation 3-117 L l. l l I I l l I xii

.. - ~, FOREWORD ~LaSalle Unit 2 Level III Probabilistic Risk Assessment In recent years, applications of Probabilistic Risk Assessment (PRA)-.to - -nuclear power plants have experienced increasing acceptance. -and' use, - particularly in-addressing regulatory issues. Although ' progress on the - PRA front has been impressive, the usage of PRA methods and insights to address increasingly broader regulatory issues has resulted in the need for continued improvement in and expansion of PRA methods to. support-the needs of the Nuclear Regulatory Commission (NRC). -Before any new PRA - - methods can be considered suitable for routine use in the regulatory arena, =they need to be integrated into-the overall framework of a PRA, appropriate interfaces defined, and the utility of the methods evaluated. The LaSalle Unit 2 Level III PRA, described in this and asso'clated

reports, integrates new methods-and new applications of previous methods into a PRA framework-that provides for this integration and evaluation.

It helps lay the bases for both the routine use of the methods and the > preparation' of procedures that will provide - guidance for future PRAs used in addressing regulatory issues. These-new

methods, once _ integrated into the framework of a. PRA and evaluated, lead to a more complete PRA analysis, a better understanding of the-uncertainties in PRA results, and broader i~nsights into the importance of plant design and operational

. characteristics to public risk. In order to satisfy the needs described above, the LaSalle Unit 2, Level III PRA addresses the following broad objectives: 1) To develop ' and apply methods -to integrate internal, L

external, and dependent failure risk methods to achieve -greater efficiency, consisten-4,

and completeness in the-conduct of risk assessments; 2) To evaluate PRA technology developments and formulate I improved PRA procedures;- 3) To identify,

evaluate, and effectively display the uncertainties in PRA' risk predictions-that stem from limitations in plant modeling, PRA methods, data, or L

physical processes that occur during the evolution of l a' severe accident;- xiii

4) To conduct a. PRA on a BWR 5, Mark II nuclear power -plant, ascertain the plant's d.ominant accident sequences, evaluate the core and containment response-to accidents, calculate the consequences of the accidents, and assess overall risk; and finally 5)- To formulate the results in such a manner as.to allow the PRA to be easily updated and to allow testing of future improvements in methodology,

data, and the treatment of phenomena.

The - LaSalle Unit 2 PRA was performed for the NRC by Sandia National Laboratories- (SNL). with substantial help from Commonwealth' Edison (CECO) and its contractors. Because of -the size and scope of the PRA, various related programs were set up-to conduct different aspects of the analysis. Additionally, existing programs had tasks added to perform some analyses for the LaSalle PRA. The responsibility for overall direction of the PRA was assigned to the Risk Methods Integration and Evaluation Program (RMIEP). RMIEP was specifically responsible for all aspects of the Level I-analysis (i.e., the core damage-analysis). The Phenomenology-and Risk Uncertainty Evaluation Program (PRUEP) was responsible for the Level II/III analysis (i.e., accident progression, source term, consequence analyses, and risk integration). Other programs provided support in various areas or performed some of the subanalyses. These prcgrams include the Seismic Safety Margins Research Program - (SSMRP) at __ Lawrence Livermore National Laboratory (LLNL), which performed the seismic. analysis; the Integrated Dependent Failure Analysis Program, which developed methods and analyzed data for dependent failure modeling; the MELCOR Program, which ~ modified the MELCOR code in response to the PRA's modeling needs;- the Fire Research Program, which performed the fire analysis; the PRA Methods-Development Program, which developed some of the new' methods used in the PRA; and the Data -Programs, which provided new and updated data for BWR plants similar to LaSalle. CECO provided plant design and operational information and reviewed many of the analysis results. The LaSalle PRA was begun before the NUREG-1150 analysis and the LaSalle program. has supplied the NUREG-1150 program with simplified location analysis methods ' for integrated analysis of external events, insights on possible subtle interactions that come from the very detailed - system models used in.the LaSalle PRA, core vulnerable sequence resolution methods, methods for handling and propagating statistical uncertainties in-an integrated way through the entire analysis, and BWR thermal-hydraulic models which were adapted for the Peach Bottom and Grand Gulf analyses. i xiv 1 l -I

~ The Level _-I _- results of the LaSalle Unit 2 PRA are presented in:- " Analysis of the LaSalle Unit 2 Nuclear Power Plant: - Risk - Methods Integration and Evaluation Program (RMIEP), " NUREG/CR-4832, SAND 92-0537,- ten volumes. The reports are organized as follows: NUREG/CR-4832 - Volume 1: Summary Report. NUREG/CR-4832 - Volume 2: Integrated Quantification and Uncertainty Analysis. NUREG/CR-4832 - Volume 3: Internal Events Accident Sequence Quantification. NUREG/CR-4832 - Volume 4: Initiating Events and Accident Sequence Delineation. NUREG/CR-4832 - Volume 5: Parameter Estimation Analysis and Human Reliability Screening Analysis. s tlUREG/CR-4832 - Volume 6: System Descriptions and Fault Tree Definition. NUREG/CR-4832 - Volume 7: External Event Scoping =Quantification. NUREG/CR-4832 - Volume 8: Seismic Analysis. NUREG/CR-4832 - Volume 9: Internal Fire Analysis. NUREG/CR-4832 - Volume 10: Internal Flood Analysis. The Level II/III results. of the LaSalle Unit 2 PRA are presented in: " Integrated Risk Assessment For the LaSalle Unit 2-Nuclear Power Plant: Phenomenology and Risk Uncertainty Evaluation Program .( PRUEP), "- NUREG/CR-5305, SAND 90-2765, 3 volumes. The reports are organized as follows: + NUREG/CR-5305 --Volume 1: Main Report NUREG/CR-5305 - Volume 2: Appendices A-G NUREG/CR-5305 - Volume 3: MELCOR Code Calculations Important associated. reports have been issued by the RMIEP Methods Development Program in: NUREG/CR-4834, Recovery ActionsLin PRA for the Risk' Methods Integration'and Evaluation Program (RMIEP) ; NUREG/CR-4835, Comparison and Application of Quantitative Human-Reliability Analysis Methods for the Risk xv

Methods Integration and Evaluation Program (RMIEP) ; NUREG/CR-4836, Approaches to Uncertainty Analysis in Probabilistic Risk Assessment; NUREG/CR-4838, Microcomputer Applications and Modifications to the Modular Fault Trees; and NUREG/CR-4840, Procedures for the External Event Core Damage Frequency Analysis for NUREG-1150. Some of the computer codes, expert judgement elicitations, and other supporting information used in this analysis are documented in associated reports, including: NUREG/CR-4586, User's Guide for a Personal-Computer-Based Nuclear Power Plant Fire Data Base; NUREG/CR-4598, A User's Guide for the Top Event Matrix Analysis Code (TEMAC) ; NUREG/CR-5032, Modeling Time to Recovery and Initiating Event Frequency for Loss of Off-Site Power Incidents at Nuclear Power Plants; NUREG/CR-5088, Fire Risk Scoping Study: Investigation of Nuclear Power Flant Fire Risk, Including Previously Unaddressed Issues; NUREG/CR-5174, A Reference Manual for the Event Progression Analysis Code (EVNTRE) ; NUREG/CR-5253, PARTITION: A Program for Defining the Source Term / Consequence Analysis Interface in the NUREG-ll50 Probabilistic Risk Assessments, User's Guide; NUREG/CR-5262, PRAMIS: Probabilistic Risk Assessment Model Integration System, User's Guide; NUREG/CR-5331, MELCOR Anelysis for Accident Progression Issues; UUREC/CR-53<6, Assessment of the XXSOR Codes; and NUREG/CR-5380, A User's Manual for the Postprocessing Ptogram PSTEVNT. In addition the reader is directed to the NUREG-1150 technical support reports in NUREG/CR-4550 and 4551. Arthur C. Payne, Jr. Principal Investigator Phenomenology and Risk Uncertainty Evaluation Program and Risk Methods Integration and Evaluation Program Division 6412, Reactor Systems Safety Analysis 3 Sandia National Laboratories Albuquerque, New Mexico 87185 xvi

l i ' ACKNOWLEDGMENT 1 The~ authorra w3sh to acknowledge the guidance and technical - support of Gregory J. Kolb and Kathleen Diegert. 4 r I-l- xvii

l.0 INTRODUCTION - A full-scope Probabilistic Risk Assessment (PRA) 'of a nuclear - power plant should consider all internal and external events - that-- may ' pose a potential threat to the plant safety and contribute to the public risk. The detail to which the risk analysis is performed for each event depends on its frequency of occurrence'and its effect on plant systems. In recent PRA studies, some external events .e.g., seismic, fire, internal ( flood, and extreme winds) have been treated in detail; other external events (e.g., turbine missiles,-aircraft impact, and external flooding) have been dismissed as insignificant based on available datu and judgment. Since PRA is a logical ari formal procedure-tor examining all potential accidents, - a logical and formal approach is needed for selection of important external events. The aim is to ensure that all potential external eventa are considered and that the significant ones are selected for more detailed studies. In fact, such a formal procedure has been developed in the PRA Procedures Guide, hUREG/CR-2300 (USNRC, 1983). This procedure also facilitates a complete documente. tion of the basis for selecting the external hazards which deserve further detailed attention. Because the PRA Procedtres Guide only described detailed methods for seismic, .;1 ci od, and fire events, a separate' analysis was performed to develop scoping . quantification methods for other e.<tet nal events (Ravindra and 'Ba non - 1992). This report is a description of the scoping quantification study which-selected the external events to be included in the detailed PRA of'the LaSalle County Nuclear Generating Station. The study was performed by NTS/ Structural Mechanics Associates (SMA) for Sandia National Laboratories as part of.the Risk Methods Integration.and Evaluation Program (RMIEP). _The study generally followed the procedures outlined in the PRA Proce-dures Guide (USNRC, 1983) as to methodology, presentation, and technical quality assurance, but was supplemented by scoping - quantification methods developed and described in the report by'Ravindra and Banon mentioned above. l'.3 Backcround The Risk Methods. Integration and Evaluation Yrogram (RMIEP) l performed by Sandia National Laboratories for the NRC selected j the LaSalle county Station for application of the new - methodologies developed as part of the full scope PRA. One . task of-the _ RMIEP plan was defined as an external event scoping -quantification study which would select the external ' events to be included in a-detailed external events analysis. l F u r. this purpose, NTS/ Structural Mechanics Associates was l 1-1 ~, ,-m-, e e w ,,w- ,,+-

__ _. _ _ _ - _. _.. ~ retained ; by : Sandia National Laboratories to perform the scoping quantification study for the LaSalle County Station. Although'a_-general' external event scoping study would consider all the1possible events at-the site; seismic, internal flood, and fire events were excluded from the present study. Based on the'results of recent PRA Studies, they.were considered to be potential contributors to the plant ' risk and thus were included for'a detailed study in the other tasks of. RMIEP. . The - LaSalle County Station has been derigned against-the ef fects of extreme _ winds, tornadoes and tornado-generated missiles, and chlorine release. Exampler, of other external events 'which were considered in the LaSa'. le FSAR but were 'not - specifically included in the design basis loads are external flooding, turbine missiles, and aircraft impact. The FSAR - analysis was' based -on meeting the Regulatory Guide requirements rather than_ quantifyiag the plant risk from external events from a PRA' standpoint. The methods for performing an external event scoping quantification have been outlined in the PRA Procedures Guide (USNRC, 1983).

However, the methods are described in a general 1' f ashion and the spcific mathematical models and analytical techniques to be used are' not described.

The general methods described in=the-PRA Procedures. Guide form the basis for the scoping. procedures to be used in this study. - In addition to the PRA Procedures Guide, a review of the 7 techniquessand~the mathematical models used to scope external events in other.NRC and industry-sponsored' studies was carried out.. These models and techniques were examined for their applicability to the LaSalle scoping quantification study, including -_ detailed bounding analyses, and the results were used to_ develop more_ detailed scoping quantification methods for use in this study (Ravindra and Banon,-1992). ~ 1.2 Obiective The' objective of-this= ' report was to - perform a scoping-quantification in order ~ to define the additional external-events, if any, that the LaSalle PRA should analyze in detail. -As reported previously, the PRA ~ analyzed seismic, fire, and - internal _ flooding. events in detail (see-volumes 8, 9, and 10 respectively of this report). . l.3 Outline and Contents of Report U

This report describas the external events scoping lJ

--quantification performed for the LaSalle County Station l ( L3CS ) ~. This report is divided into four chapters. Chapter 1 L -is an overview of t h e' study including background.and 1-2 L l ___-____________,_,._,___,,____,_______m .m.- p-3- .y ---4_ y m ,<ya 9 g

objectives, Chapter 2: describes the selection of methods for the external events risk analysis, identification of potential external events, and the general-methodology for an external event bounding' analysis. Also, Sections 2.6, 2.7, and 2.8 in Chapter 2 are general descriptions of the sources of information, technical quality assurance requirements, and the uncertainty analysis for external ~ events. Chapter 3 describes the initial screening of = the external events, aad the more detailed bounding analysis performed for - the events which could not be eliminated through the initial screening process. For each bounding analysis, a mathematical model is presented and sources of the data for estimation of parameters.of the model are reported. The bounding analysis in Chapter 3 shows the significance of each external event to the plant risk. Tnerefore, events which require further detailed analysis are identified in Chapter 3. Chapter 4 summarizes the results of initial screening and bounding analyses. Also a set of recommendations based on these results is presented in Chapter 4. l l-l' 1-3

7 2.0 EXTERNAL EVENT METHODOLOGY An external event analysis in a PRA has three important goals. The first goal is that no significant events should be overlooked. The second goal is an optimal allocation of limited resources to the study of significant events, and the third goal is that the differences between external events and internal events (i.e., common-cause and fragility related failures) should be recognized and explicitly treated. Based on these goals, four tasks were identified for the present study. 1. Review of external event scoping quantification general techniques and mathcuatical models. 2. Identify pote.itial external events. 3. Initial screening of external events. 4. Approximate bounding analysis to calculate risks from external events. a eneral description of each task is given in the rollowing u tions. 2.1 Review of General Techniaues and Mathematical Models During the last four years, several Probabilistic Risk Assessments for nuclear power plants have been published. Aside from seismic, fire, and internal floods, other external events have not been treated in-depth in these PRAs.

However, the general techniques and models for quantification of risk from external events have experienced much modification as more PRA studies were completed.

Therefore, there is a need to study and compare these models and techniques before performing the LaSalle external event scoping quantification. It may be noted that not all of these m dels are applicable to the LaSalle site. For example, the Lin.srick PRA (PECO, 1983) which was performed by NUS Corporation studies the hazard from a chlorine explosion on site in great detail.

However, information about chlorine stored at the LaSalle site indicates that only a small amount of liquid chlorine is stored on site.

Therefore, it was judged that there is no T possible risk from chlorine to the LaSalle County Station. On the other hand, reviews of the models and information which were carried out in this task would be used in developing the external event scoping quantification methods document. i 2.2 Ldentification of Potential External Events The PRA Procedures Guide (USNRC, 1983) was used as a guide for identification of potential ex t. crna l events at the LaSalle 2-1 l

eite. Table 10-1 of the PRA Procedures Guide lists most of the possible external events for a plant site. This information was reviewed in the present study.

Also, 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 Final Safety Analysis Report (FSAR) regarding the geologic, seismologic, hydrologic, and meteorological characteristics of the site region as well as present and projected industrial activities (i.e., increases in the number of flignts, construction of new industrial facilities) in the vicinity of the plant were reviewed for this purpose. A description of external events considered for the LaSalle site appears in Section 3.2. 2.3 Initial Screeninc of Events At this stage, the external events identified as described above were screened in order to select the events for either approximate or detailed risk quantification. A set of screening criteria was formulated that should minimize the possibility of omitting significant risk contributors while reducing the amount of detailed analyses to manageable proportions. The set of screening criteria givcn by the PRA Procedures Guide used in this study is as 'follows. An external event is excluded if: 1. The events for which the plant has been designed. This screening criterion is not applicable to events like earthquakes, floods, and extreme winds since their hazard intensities could conceivably exceed the plant design basis. An evaluation of plant design basis is made in order to estimate the resistance of plant structures and systems to a particular external event. For example, it is shown by Kennedy,

Blejwas, and Bennett (1982) that safety-related structures designed for earthquake and tornado loadings in Zone 1 can safely withstand a 3.0 psi static pressure from explosions.

Hence, if the PRA analyst demonstrates that the overpressure resulting from explosions at a source (e.g.,

railroad, highway, or industrial facility) can not exceed 3

psi, these postulated explosions need not be considered.

2. 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 events. In this case, the uncertainty in the frequency estimate for the excluded 2-2

~ _ _ _... _ -. - _ _ -. _ _. _ _ _ -_._ _ L event 11s judged by the - PRA analyst as not significantly influencing the total risk. l 3._ The event cannot occur close enough to the plant to affect I it. _ This is also a function of the magnitude of the

event.

-Examples =of-such events are landslides, volcanic eruptions, and earthquake fault ruptures. 4. The event is included in the definition of another event. For example, storm surges and solches are included in external flooding; the release of toxic gases from sources external to the plant is included in the effects of either pipeline accidents, industrial or military facility accidents, or transportation accidents. J By this process of initial screening, a smaller set of external events is identified for risk assessment. A bounding analysis is'then performed for these external events. 2.4 Boundina Analysis Although the-screening process has identified a set of external events for further risk analysis, it ir still possible to perform simplified analyses to show that some of the. events are not significant contributors to the risk._ The bounding risk analysis is an essential step in the external event PRA.as-it minimizes the effort that is required for a detailed external - events analysis. The key elements of a complete bounding risk analysis for an external event are: o Hazard analysis o Plant system and structure response analysis o Evaluation of the fragility and vulnerability of-plant structures and equipment o-Plant system and accident sequence analysis o Consequence analysis A -hazard analysis estimates the frequency of occurrence of 'different-intensities of an external event. These are called " hazard intensities." Typically, the output of hazard analysis is a hazard curve of exceedence frequency versus hazard. intensity. Since there is normally a great deal of uncertai nty 1 in the parameter values and-in the mathematical model-of the hazard, the effects of uncertainty are represented through a family of hazard curves, and a - probability value-4.s assigned to each curve. The purpose of structural response analysis is to translate the hazard input into responses of structures, piping systems, 2-3

and equipment. The fragility or vulnerability of a structure I or equipment is the conditional frequency of its failure given a value of the response parameter. In some external event analyses, the response and fragility evaluation are combined and the fragili&y is expressed in terms of a global parameter of the hazard (e.g., tornado wind speed). The analysis of plant systems and accident sequences consists of developing event trees and fault trees in which the initiating event can be the external hazard itself or a transient or LOCA initiating event induced by the external event. Various failure sequences that lead to core damage, containment failure, and a specific release category are identified and their conditional frequencies of occurrence are calculated. The unconditional frequency of core damage or of radionuclide release for a given release category is obtained by integrating over the entire range of hazard intensities. If the consequence analysis is carried out separately for the external event, the output would be curves of frequencies of damage (i.e., early fatalities, latent cancer deaths, or property damage). After a bounding analysis is performed, an external event can be excluded from further risk assessment based on the same considerations as in the initial screening analysis. For example, calculation of the core damage frequency may be done using different bounding assumptions explained by the following example. Typically, nuclear power plants are sited such that the accidental impact of plant structures by aircraft is highly unlikely. For the purposes of an external event PRA, the risk from aircraft accidents may be assessed at different levels. The mean annual frequency of aircraft impacting the plant during take-off or landing, or in flight may be determined. If this hazard frequency is very low (e.g., 510-7 per year) then the aircraft impact as an external event may be eliminated from further study. This approach assumes that the aircraft impact results in damage of the structures leading to core damage or serious release. This assumption may or may not be highly conservative. The assessment of the conditional probability of core damage will determine the actual cutoff level used here. If the frequency of aircraft impacting the plant structures is estimated to be larger, the fragility of the structures may be evaluated to make a refined estimate of the frequency of core damage. Further refinements could include (1) elimination of certain structural failures as not resulting in core damage (e.g., damage to the diesel generator building may not result in core damage if offsite electrical power is available), and (2) performing a plant system and accident analysis to calculate the core damage frequency. This example shows that for some external events, it may be sufficient to perform only the 2-4

hazard analysis; for some others the hazard analysis and a simple fragility evaluation may be needed; only in rare cases, a plant-systems and accident sequence analysis may be necessary. The procedure of screening out the external events in this stage consists of: (1) establishing an acceptably low mean frequency of core damage based upon simplifying conservative assumptions (i.e., $10-7 per year), (2) performing bounding calculations of the mean core damage frequency for each external event, and (3) eliminating from further consideration those events which have mean core damage frequencies less than the acceptable value (i.e., 10-7 per year). As part of the licensing evaluation of nuclear power plants, probabilistic analyses are performed for a few external events, and the frequencies of unacceptable damage (i.e., exceedence of 10 CFR Part 100 guideline exposures) caused by these external events are shown to be very small. The information contained in the plant safety analysis reports and the analyses performed at the design stage in support of FSAR are reviewed and new information is gathered as part of this effort. Since the PRA attempts a realistic risk evaluation, the conservative bias introduced by the assumptions made in the licensing analysis are appropriately removed. 2,3 Detailed Analysis For the external events that are not screened out by the initial screening process and the bounding analysis, a detailed risk analysis is necessii. Such an analysis is typically done for seismic eventu, internal flooding, and fire. The risk analysis methods for these events are described in Chapter 11 of the PRA Procedures Guide. Any other external events identified to be potentially significant contributors to the risk based on the results of this study would need to be studied in detail. However, such detailed PRA analysis is outside the scope of this report. 2.6 Information Plant specific information for the present study was obtained from the LaSalle FSAR (CECO), and engineering drawings of the plant. This information was augmented by other information regarding the plant design basis provided to SMA by the Commonwealth Edison Company and Sargent and Lundy Engineers. Some of the generic data which were used in the external event bounding analysis were reported in previous PRA studies, e.g., the Limerick Severe Accident Risk Assessment (PECO, 1983) and the Midland PRA (CPCO, 1984).

Also, a site visit was conducted by the SMA personnel.

The objectives of the site 2-5 l

- - _ ~. - _ -. _ x, visit were -to verify the information which was given _ in the FSAR and to - gather new information concerning the effect of potential external events on the plant. 2.7 Technical Ouality This-study conforms to the requirements of the assurance of technical quality as outlined in the PRA Procedures Guide, Chapters 2 and 10. The study was performed at the Newport Beach offices of NTS/ Structural Mechanics Associates by the authors. The methods used, whether previously developed in a published PRA or developed as part of this study, were documented and internally reviewed. The results were internally reviewed by Dr. D. A. Wesley who is a senior consultant to the_ project. An external quality assurance audit of the project was also performed. 2.8 Uncertainty Analysis Uncertainties exist in the hazard analysis and the fragility evaluation of plant structures and equipment. These arise from lack of data (i.e., parameter uncertainty) and in the use I of analytical models to predict fai4ure (i.e., model uncertainty). The uncertainty in frequency of the plant damage due to an-external event is particularly important if the - event is a potential contributor to the plant risk. Therefore, for these events, an attempt was made to address the question of model and parameter uncertainties, i.e., an integrated - assessment of both parameter and model uncertainties. was made to calculate the high confidence (95 percent) value.of the annual frequency of plant damage. As will be_ described in Chapter 3, uncertainty analyses performed for these external events were in accordance with the methods and models used by SMA in previous Probabilistic Risk Assessment studies. An effort is currently underway at Sandia to develop new methods of uncertainty assessment as part of the RMIEP. Therefore, detailed information regarding the data which were used to estimate the parameters and choice of the models were provided to Sandia personnel to be used in an uncertainty assessment which is consistent with the RMIEP uncertainty methodology. 2-6

__.__.___..m.- s 3.0L SCOPINGJQUANTIFICATION STUDY This chapter describes the initial screening of external -events and the bounding analyses which were performed as part of the.LaSalle scoping quantification study. Section 3.1 is a general-description of-the plant structures, site characteristics,. and transportation routes near the site. S e c t i o n _ _ 3 '. 2 - lists all the external events which were identified for the LaSalle site. Also, the initial screening of these external events has been described in Section 3.2. Some of the events which required a more detailed screening analysis based on the LaSalle FSAR information are listed in Section _3. _3. The external events which required a bounding analysis appear in Section 3.4, and those events which may require a detailed PRA analysis are identified in Section 3.5. -3.1' Plant Descrintion The LaSalle Nuclear Power Generating Station was designed in the early 1970's in accordance with criteria and codes in effect at that time ( LaSalle FSAR). The station consists of two Boiling Water Reactors (BWR), each rated at 3323 Mwt and 1100 Mwe. The plant, with the exception of the Nuclear Steam Supply System (NSSS), was designed by Sargent & Lundy (S&L) Engineers. The NSSS was designed by the Nuclear Energy Division of che General Electric tompany. The BWR Mark Il containment design is used. The primary containment is a steel-lined, post-tensione'd concrete structure enclosed in the reinforced concrete reactor building. The primary structure consists of a' combined building which houses both NSSS units, the turbine buildings, an auxiliary building, the diesel generator buildings, a radwaste building, the service ' building, and the of f-gas building. A lake screen house is -locatedLon'the inlet flume but does not contain any critical equipment. Seismic Category I structures and-equipment were designed to withstand both a Safe Shutdown Ea rthqua k_e - (SSE) and an Operating Basis Earthquake (OBE). The maximum horizontal ground _ design accelerations at the foundation _ level were 20 percent'of gravity for the SSE and 10 percent of grav_ity for the-OBE. The corresponding maximum vertical design acceleration was two-thirds of horizontal for both the SSE and .OBE.. Plant structures and equipment important to safety ware classed as Seismic Category I in the original design. Codes and standards used in the design and qualification of- . structures and. equipment - for the LaSalle Plant are listed in Table 3.1-1 (LaSalle FSAR). Figure 3.1-1 (LaSalle FSAR) shows the general arrangement of the LaSalle structures. It may be -noted that the-outside walls of LaSalle structures do not have the ' same - thickness, e.g., the diesel generato-alls are 12" 3-1 l l: u.-

thick whereas the reactor building walls are 2'0" thick. Thickhoss of the outside walls is important in the analysis of structures for winds and tornadoos, tornado missiles, and turbine missiles.- Figure 3.1-2 (LaSallo PSAR) shows a section of the plant structures including the reactor building, the auxiliary building, and the turbine building. Although the reactor building is enclosed by 2'0" walls below the refueling floor at Elevation 843#6", it is sh.elded by only metal siding above the refueling floor. The refueling floor of the reactor building in LaSa110 does not contain any ongincor9d safoty features (ESP) equipment. 3.1.1 Site, Terrain, Motoorology f The LaSallo County Station Units 1 and 2 are located in north-castern Illinois. The Illinois River is approximately 5 miles north of the plant. Figure 3.1-3 (Laballe PSAR) shows the general location of the site within the state of_ Illinois. The LSCS site occupies approximately 3060 acres, of which 2058 acros comprise the cooling lake. There are no industries or residences _ on the sito. Thoro is a state fish hatchery associated with the plant. The general layout of the plant is shown in Figuro 3.1-4 (LaSalles FSAR). The major transportation routes near the site include the Illinois River, approximately 3.5 miles north of the northorn boundary; Illinois State Highway 170,_ 0.5 mile east of the eastern boundary; and Interstate Highway 80, 8 miles north of the northern boundary of the site. The Chicago, Rock Island, - and Pacific. Railroad; approximately 3.25 miles north of the northern site boundary is the closest opera _ing railroad lino. The LaSalle FSAR includes a descriptior, of tJisting and projected population centers near the siv. 4 The population within 10 miles of the site vas - 15,600 :as ut' 1970 and it was relatively projected to grow to 24,300 by P020. Tno most heavily populated areas near the sito lie la the northeaC [ direction towards the city of Chicago. There are no storage facilities, mining and quarry operations, t ransportation facilities, tank farms, or oil and gas pipolines within 5' miles of the plant.. There are no military bases, missile sites, military firing or bombing ranges, refineries, or underground gas storage facilities within 10 miles. There are no products or materials regularly manufactured, l stored, used, or traNported within 5 miles of the sito. Tho nearest industrioO 'aro located in Seneca,

Illinois, approximately 5.6 miles northeast of the site.-

There are no 3-2 - = -....

A commercial airports within lo miles of tne site, and there are no privatu airstrips within 5 mileu. At the present time, there are two airport site investigations in progress in the vicinity of the LSCS site. The LaSalle-Peru area approximately 23 miles west-northwest of the plant site is being studied as one possible site. The second airport study is being conducted in the area betwe 1 the towns of Pontiac, Streator, and Dwight, approximately la miles south of the LSCS site. Both of those airports will be designed to handle commercial planes in addition to the single-engine and twin-engine planes common to the area. Also, the Continental Grain Company is developing a river terminal to handle both barge cargo and truck cargo, but there are no plans to handle hnzardous or explosive materials. The LSCS site experiences a high variability and a wide range i e' temperature extremes. For example, extreme temperatures recorded at nearby Ottawa, Illinois, range from 112' to -26*F. Temperature data recorded at Peoria Airport and Argonne National Laboratory as well as data from the LSCS 4 meteorological tower were used in the plar

• design.

Precipitation in the LSCS site area averages about 34 inches annually with monthly averages ranging from about 1.8 inches in January to 5.0 inches in July. Precipitation is not monitored at the LSCS site. Long term data from Peoria airport and Argonne National Laboratory were used in the plant design. Sleet or freezing rain can occur during the colder months of the year. Glazo storms with ice thickness of 0.75 inch or greater are expected to occur once every three years. The LSCS site, located in mid-Illinois, experiences a wide spectrum. of extreme winds. In addition, tornadoes have been historically observed in the State of Illinois. For-the period 1916 through 1969, there were a total of 43 tornadoes in the ten county areas surrounding and-including-the LSCS site. The terrain around the plant site is gently rolling, with ground surface elevations varying from 700 feet to 724 foot mean sea level (MSL) which is 217 feet above the norme.1 pool elevation in the Illinois River. The river screen house and the outfall structure, both nonsafety-related structures, are the - only plant facilities that are potentially af fected - by floods in the Illinois River. 3.1.2 Sito Visit A site visit was conducted in April 1984 by Drs. M. 1(. Ravindra and 11. Banon (Structural Mechanics Associates) and 3-3 .u-~. -___,u - _, ~. _ - -

K. Campe (NRC, Site Analysis Branch). The purpose of the site visit was twofold:- firut t o confirm the information in TSAR which is being used in the LaSalle scoping quantification

study, and second to collect new information and look for posaible changes in the plant and site conditions which could affect the risk-from external hazards to the site.

Therefore, -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. 1.- Ho -major changes or deviatione from the information in LaSalle FSAR were observed in the plant or its surroundings. Since this study is concerned with the external events, the effort was concentrated on those factors which could affect-the risk from these events. 2. A survey of the structures in LaSalle revealed that all the doors which open to the outside of the plant are leak-tight. Also, the~ ground floor in every structure has an adequate drainage system in case of flooding. This information was used for the external flooding analysis. i 3. It was confirmed that the refueling floor of the reactor building as well as the top floor in the auxiliary building - do not contain any ESF equipment. This information is needed in the analysis for wind and tornadoes. 4. During the site visit, a survey of the objects in the . plant boundary which could potentially become tornado-The site visit generated missiles was carried out.. missiles at the confirmed that the potential number of LaSalle site is less than the number which-has been used in a tornado missilo simulation study by Twisdale and Dunn --- ( 19 81). - -Tornado missiles are discussed in Section 3.4.3. 5. It was-observed that collapse of the stack under winds or tornado loads could affect the safety of category I structures in LaSalle. Further information from the Commonwealth Edison Company showed that the stack has been cusigned for the effects of the Design Basis Tornadoes. Thocefore, the stack does not add =to-the risk from winds and tornadoes. This - is - described in more detail-in Section 3.4.3. 6. The site _ visit confirmed that there are no industrien, airportra, pipelines, or major highways-in the vicinity of the site. Il o w e v e r,. no attempt was. made to -find-information regarding future ' construction of such-3-4 . -. = - ~...

. -. ~.. ~ -. -. -. - _ -. -.. _. - facilities near the cite, i.e., this study would rely on the FSAR information for this purpose. In addition to the site visit, the SMA personnel also visited the officos of Sargent and Lundy in Chicago,. the Architect-Engineer for the LaSa110 Plant to gather information for the scoping quantification study. 1 1 6 F M F 3-5

a....

-.,=.a.-.-.-_---...--..._,-=-.,.

ta>Je 3,1 1 Code Requireeente for Cowtonente W1 lyettas 0.dered Af ter July 1.1974 QUALITY f WP CIA 55151 CAT 10N 3 A 2 'ressure Vesselse ASMt toller and A5MI 6eller er. ASMt teller and A5MS Beller and Pressure vessel Code, Preceute vessel Code. Fressute Vessel Code. Pressure vessel Code, Section 111

1976, Section ill 1974 Section 111

1974, Seetten V!!!, niv. 1 Class 1.

Class 2. Cla66 3.

1976, ritirig

_ ASMI better and ASMt &ctler and AtMt teller and AN$1 831.1 1973, tiessure Vessel Code. Pressure vessel Code, Fretoure vessel Code, Code fet treasure . taction til. 1974, Section 171 197=, $ection ill

1974, tiping.

a Claes 1. Class 2. Clase

• 1 tumps and Valves A&Mt Beller and ASRI boiler and A$ME boiler and ANi! $)1 d 19?),

Pressute Vessel Code, Pressure vessel Code, Pressute Vessel Code, C6de for Pressure Section 111 + 1974, lection 111'. 1976, settien 111 4 1974, tiping.** Clast 1. Class 2. Clast 3. lev Preseure Tar.ks American Petroleur American Fettoteum Institute, Retos. It's titute, Recom. mended Rules for sended Rules for a Design and Construe. Design and Construe. tion of lange Welded tion of 1arge Welded inw. treasure Storage law Pressure Storage tanks, API 670 1963 fanks, API 620 196) edition. American Waterwerbs Ameriesn Waterwerke American Waterworks Atsusphette $terage p fanks Assoelation. Standard Aeonciation, Standard Aseociation, Standard ter steel Tanks, for $ teel tanks, for steel tanks. l Standpipes, ke tra $tendpipes. Reser. Standpipes, peset. votre and Elevat ed votre med tierated - veits and llevated Tanks for Water Stot. Tanks fon Water $ter. Tanke fe,r Water Ster. ogo,' AWA.D1001967 age, AWA.D1001967 ego, AWo.01001967 editioni er Weided edition; or Welded edtrion: er Welded - l Steel Tanks for 011 Steel Tanks for 011 Stool Tanks for 011 Storage, Att 650 1964 storege, AF16SO 1964 5torage. AP16501967 edition)

edition, edition.

Heat Enchangers .ARMf. Soller and tree. ASME Soller and Pres. ASME Soller and Pres..ASME better and Pres. aute Veseel Code, sure Vessel Code, sure Vessel Code, -su.se Vessel Code, lection 111

1974, Section 111

1976, Section til.1974, Section V111 Div. 1 Class 3, 1974, and tubutor Class 1, Clase 2.

i Enchanger Manufacturere Associatten (TIAA) Class C. eftfV and Contairment Vessel eteluded, t ' **For pumpe eperating ebeve 150 pet and 211*F A$MR 8eetsen V111. Divlelen 1 shall be used as a guide for [ esteulating thicknene of presoura retaining parts and in sising cover betting; below 160 pel-and 112'T manufacturer's standards for servisa intended w!!! be used. ' 1 upplementary WDE.' 1009 vclumetrie esamination of the side wall for plates 3/16.tnr.h thtch and 1004 sur f ace k 5 .esamination of welds for plates 3/16. inch thich or lose. Aloe, 100 persent surfsee enemination for olde to hotton

welds, r

s k T + 3-6 .w ..m - .s

i 4 i l' e Teble 3 1 1 l. Code 5egattements f&r Component s and Systees s' Os dened Priet to Jult 1,1971 VJA1.1Tf CRWF CLAlstr!C.AT!pH s'* I C Feessure Vesselei Asnt beller and ALMt boiler and ASMF 5eller and AsME totter and treasure Vessel Cede, Pressure Vessel Cooe. Pressure vessel Codes. Fiessure Vessel Codes., section Ill. Class A. $setten lit. Class C. Section V111, Div. 1 8*etten Yllt. Div.1 I 1968 Addenda 1960 Addenda through 1968 Addenda thrpugh 1968 Addehde thttugh through summer 1970. Summer 1970. Summer 1970. Suseer 1970, tipingee AN81 831.1 Nuclear AN5! $11.7 pueleat ANs! $13,7 Nuclear AN81 211.1,0 Code for Power Flping, Class Fevet tipitig, Paping, Clase !!! - Fiessote Piping 196 7. 1e 1969. Class II + 1969.

1969, Addendus. 1969, Pumpe and Valvesti ASMF Code for Pumps ASME Cenas for Pumpe ASMt Code for Pumps Anal $11.3 0 Code for and Valves for nucleat and Valves for Nuclear and Velves for Nuclear Pressure tiptr;g*

Power, Class 1 1946 Power, Class !! 1968 Power, Class 111 1968 1967. I Draf t addenda March Draft Addenda March Draf t Addenda March 1970. 1970. 1970. Low tressure Tanks American Petrolean American tetroleum

• • =

Institute, Recommended leetitute, Recomeended Rules for bestgn and Rules for benign and Construction of Large. Construction of targe Welded Low treasure Welded Luw.Fressure $terage Tanks. API 610 storage Tanks. Af1 620 1961 editten. 1961 edition. Atmospherte-Ametteen Waterworks American Vetowworks American Vaterworks Storage Tanks Association 8tandard Asso41stle s, Standard Association, 8tendat d I f or Steel Tanks, land-for steel Tanks, Sand

• for Steel Ta?As, sand-pipes, keservoirs and pipes, Reservelts and pipes, Rese rvoirs and Elevated Tanks for llevated Tanks for tievated Tanks for Vatet Steroge, AWWA.

Water storage. AVWA-Water Storage, AWWA. D100 1967 edition; er D100 1967 editten; er D100 1967 edition; er Weided Steel Tanks for Welded Steel Tanks for Welded Steel Tanks for 011 Storage. AP1 650-011 Stossgo, AP1,650 Oil $torese. API 6$0 1964 edition," le64 edition. 1964 editten, keat Enchangers ASKI Soller and Pres + ASME noiter and Pres. ASRt Satter and Fres+ A$ME toller and Pres. sure Vessel Code, sute Vessel Code, sure. Vessel Code, sure, Vessel Code, tection lit. Class A.. Section lit Class C, Section V111. Div, 1, Section Div. 1. 1968 1968 Addenda through ^1968 Addende through 1964 Addenda through Addenda through Summer Summer 1970. Summer 1970, and Summer 1970 and 1970 and Tubular Tubular Enchanger Manu. Tubular tachanger Manu. Enchanger Manufacturers facturera Assestation facturere Association Association (TEMA7 (TEMA) Class C. (TEMA) Class Cr Class C. ePumpe operating above 150 pet and 211*F ASMS Section V111. Divisten 1 of the Beller and Pressure Vessel Code shall be used se a guide for saleulating the thickness of pressure rotatning parts and in etsing cover botting; below 150 Pet and 111'T nanufacturer's standarde for service itaended w!!1 be used. ' ** Croup A nuclest piping, pumpe ann volves will meet the provisions of ASME Setter and Pressure Vessel Code, Section L - 111. Summer Addenda 1969.' fatagraph H+151. 'RW and Containment Vessel escluded. $upplementory NDE e 10b4 welumetric emaatnattom of the side voll for plates.ever 3/16 Alech thttk and 1004 U surf ace esmalnatten of welds for pistes 3/16. inch thick or less. ~ welds. Also, 1004 sutract omaatnetten of elde to-bottom 3-7 ,~-e 5,,, vrid-ey-v,,www-.- wr n-,-vm-ww--w+.veew,,,-,,-a,-~wv5,,m.- .-,,-~,%ww-w.3m--+.vam- ' m e r w y mir-v----r-

• • -=r-'

Tabit 3,1 1 rada kequiremente for Contaments and Systes;s Ordered # f ter July 1,1971 OUALITY Ch'r CLA$$171 CAT!0te t 4 1 tressure veeselse ASMt st uer and AsMt s.11er and asMr teller and AsMt seller and cressur i vessel code, treasure vessel Code, riessure vase.1 code, Pressure veeste Code, Section 111-1971, Section 111

1971, Section !!!. 1971.

Section Vllt. Div. 1 Claea 1, Class 2, Clase L 1968. Addenda through wintet 1970. r$ ping ASMt Settet and ARMt totter and ASME boiler and Akl1411.1.0

1967, Freteure vessel Ceda, Presouro Vessel Ceda, Pressure Yessel Cede, Ctde for tressure i

section ill. 1971, Section !!!

1971, Seetten 111

1971, tiping. Addendum Class 1.

Class 2. Cisse L 811.1 De. 1969. Pumpt and Velves H Mt botter and A&Mt be 1er and ASME Batist and ANi! $117,0

1967, treasure vesse' Coue, treasure Vessel Code, Pressure Yeesel Code, Code for Pressure seet tec !!!.1971, lettien 111

1971, Section 111

1971, Piping. Addendum i

f Class I, Clasa 2. Class 3. $11,1,0s - 1969.ee American letteleus American totteleum i low Pressure tanks institute Recce. Institute, tecos. mended pulos for sended Rules for Destge, and Constrve. Design and Construc. tion of large velded tien of large Welded Imw tressure Storage tav. Pressure storage Tanks, AP16101961 Tanks, Att 620 1963

edition, edition.

Atopopheric Storage American Waterverks American Votorworks American Waterworks Tanks Association, Standard. Association, $tandard Association, Standard 5 for Steel tanks, for Steel Tanks, for Steel Tanks, = i Standpipes, kene(* Standpipes, Reser. Standpipes, koser. ~ veits and tievated votra and flevated votre and Elevated Tanks for Water Ster. Tanks for Water Ator. Tanks for Water Stor. age, AWA. DIDO 1967 age AWA t:100 1967 age, AWA t1001967 settion; of Welded. edition; er Welded edition; or Welded Steel Tanks fer Oil Steel Tanks for Oil Steel tanks for 011 r storage, AF1 4 0 1964 Storage. (?!.HO 1964 Storage, At!.450 19% edition,'

edition, edition, 5

llest tachangers A5ME Seller and Pres. ASME better and Pres. alm 2 Setter and Pres. AAME Soller and Pres. eure Vessel Code, eute Vessel Cede, sure Vessel Code, sure Vessel Code, Sac. - Section 111

1971, Section 111

1971, Soetten 118

1971, tion Vill, Div. 1 1964, Class 1,

' Class 2. Class 3. Addenda through Vinter 1970; and Tutsular En. changes Manufacturere Association (TEMA) Class C, I

• sry and Centsinnent Vessel sueluded,

~

• ef er pumps opersting above 150 pst and 212*r ASME Section Vill Division 1, shall be used as a guide for toteulating thickness of pressure retaining parte and in etsing cover bolting; below 150 pet and 212'r manuf acturer's standarde for service intended will be used.

' Supplementary NDt.1004 volumetric esseinstion of the ~ side wall for plates 3/16. inch thick and 1004 suiface.maninatt.n of welde for et.tes 3/16. inch thlet er le.e. A1... loce eurface essaination for side.to. detto.

1ds,

- Reproduced from the LaSalle-FSAR. i 3-8 6-r W ^1--r e-v=wr rw-r w-.- w+" w m-v ---"-as wee -i -we r*r-m"m'"e-ame.ir v

rei,

-,e wow -w,==i w we v-wi-- w.- ---w*--=-----+----a

---

ww

==-m

E -M I }'> ,I Q 3.y'1 -@) 9* ..;e h,,f C I

I j

g; ; v a__ _ _. g , p,;,} g q;i l, y;} 3_ 2._.. t i e* q I l i I a Sq .. _t -- 1 o. gg I, n ,Il ;j

g lh i

I g .P h s - l1111[! i eo ,i it t. u wj s a e mwe=r ~ e a, + ,s 'h 0c -.~u l

• et ner-- e.re 7;

D l c i _.__si. cu e o a, .pt. _ __es O i, l8j ! )i o g. ),,it.,>Oj.g=;iafti? i u i t = i ii:

• d c

0-Q. b L.Ag L e w.= =4 )I l !b e .,._;-===) 19 g 0 I 4(h j;. ' [t U NI E' I i l c i d e , m f' I l s = 11 ;[w[ hI .i l i [ f 0 0 ![1 L r u y i n ::!!,s l p h p y li '. i j .us g !h;lg'O #n ! =i 1 il !! !!q !.! 'e:t)s0 l-13

1 B

0 d e. E 4% g = l m-e. r;'g;rg _.b- ^l:L'l(%

• m w

{ ll 'I l e ct; a n .c 4 y !! Mp 1 v 1

• bir g

m 3, Cg': y-),< M M @w.c G i ~

t "* 9{_O j d; h:-

T L itt. 0 'I .-am, { I t, ; _.. u - l 1 }.;. lil D-e

1 g'

[ e' I H ~ h* 1l1 .i 4, i ~~~~'% l ' L. 4 i .i J' I % li,lj4 a. 'n !!I e -3; E E } m g { [> !! l j bllij,I! [ )g? G O a u a a t l ,I'! '.3 j o i o x a,,,,, l - u p.#_a_ __jg, g i, m o l l 0 Q hb Y- .l. t s"f, y ^A 'l ilj 1 plo [d, n p; q :8sd, 8 i it i 1.' 4 g ji l c i t !c I ,iIil e k- .n. i i i . O l t 2 .us-I 1 h I 9 I { O l G l0 v.) '.) d k W M} j o E' ~ CL 1 t i

• p d

.e d ) 5 e 4 a -- - e e 3-9

ir ir.r ir ir ' e. ( k 1 ,h M .i 3 i i !! il 'J i

ik at ss wi er 1 i

.I F,I; SI;f7g_ +.d r? ~' y, n c d 4.;7J;eg - g, {p ,e E)!L gh '.l, p _. htenga6 $>1 e-c !! n ! i n Jy)jjkg!dhilik@j7 i l 3%FMI hu: - I l L !My I i! ltr.~,) '"e@hr4[Y,1 a t c i') L.'L9 r *,i. a. ~ il S,,iti 3;w% 3.,I 5 u, n,, C.,F,t n * *l - ~ i )F@%4l$ Q tr$* Eh e e e __s(h.F.,_. owe"EF35iML 7w ai .d !j swWIg"k [go-gr I' />,$ r U ), j di a3 473' bfrWh,,p?$,.3: _.i 3 AB@gi.Ct% "2 05: Ls n!l!!! Ilill!! !! li[tiu ! ! fl f,if!lf 'I If f, l 8 it L :L :L itit L it l 8 4 e ?x m 1 3-10 ~~~_

Reproduced from the LaSalle FSAR. _w- .a m-1 l ( i fm n-i ,l ~~ SITE S O I %l x =- i IL LilNOIS ' Q) f J

• f

,f-b x,---- l ..,1 y usv- "-f me 49ei \\ u --- h g ~* r e mm-m .e n.- ~a. L % r s r-N._ J' ~~ SCALE e 0 40 Mdes t"5EiiiPJ'EEiiid - [.,.2 l-o so xitometm - a. 25&d'Eiit$ } a, p 2 is tv:, w ,\\ l Floure 3,1-3. Location of the Site within the State of Illinois 3-11 e ? -g. ai.es-w. eg

---

y?r-

-eu--we'- u' wa- ---m-sw<*e-w-3-- a2---- fr - e a--wr-e--* w--m-r -w--4

__mm._ m _.__m, ..m t 't ,],k

• p
• y-4 4*s j ',

' 7 4 ' _i, r, e -1*.e

• 1

, ~ ~ seso ur*' l' ,y, !!w

• t. 1 ;

p -)*?.: '1 .a j '.,, nse f n.* *f,. n' ng

p "* b.V, '., m--. _ _ _. _.,

, ') ' D. I ' .k i

• . '. -S' *
• g-.,10 4 -a U.s5 e

~ ns. ,y g ,p ~ ~- * - "p ?.'*EITy* ***"d N l" ! 7:4.*

, IW ;pi~~.;s W hy.

Eh [* I ',).. 'F N Y( M JM- . s.yh, W. w'.e 3 c-i' @e " n%;--*s. _,f, q .3* c MD c:,7 _ =.. - J.o. gJ v.,e. - %., g... ;b ; +.ws! r.r v ,. ;, a < yr y, s *m, ,9 - ~A',, _. t 1 - m q, (,,s am, ~ - e,g, ..[ , [ ,M. ' ai.'., ' ab d ' / (h'M '.]h"_ q ~*', ' (N # ""',.WP' r ye4 :eY ~1 'y

• E

~~ ' 4 N W>,n.m. e.n 'i g %. s! K - 4,$;,.d A,.,.m...:., j.1, J... t' .gs.ml Q s. w.

n. r.s u s.

a.--.- r '. '.h.,

• r g

4. t v 'fn' ' ' '. h+,', k%' l'"9 f .7 4 lg, .r., V.., s, m, - s ss e )j. %,u.

f PI R

n l 3 9,

• r

'\\ g y 7.s l '. ',,:. ! * %. * ,e. ', o..+*{ ,'Qx v .s 5. t cc. ,, %. u,. 2,

• V.J,

/.i f. w,- tp j : - ;,4(, ' t, - s d; d.#--l s. @ - Q'*** ',. iQ , A.9 <f c" m )p..,. A ' ,. t!.. ..., ~ c , ';'w y, e,g$ M p,m A

jg M

jefJ [ 4, } o h, " e.. s u. g m %.I!Ib.j (,,,'i ls 7.yq. g...< t N )k .',' b6 s .h. 9-r,% m e m%, f w s. c.,,. 4.,. v, p,...-,9 a. p.. - s,. y .a s, )',... p f. ).D,.*

• t -{ #

' y,d,.L.) s.[ i It r. .,f U. F 8 1 4,J '{ a w w

1. a 9

,4,' \\ -.-m t< i 3 q i e ;,,,,, 4 . k. =%:, ~l.. v, + ,1 u, g.a. y) + x ,~ t,

- >.s i

m.~. fR l- ) v' s;s'. n h - A-n ~, 1% M % j, -.n 1 \\ ,t e.< . yg * - --~~~,.c.. g ,f.*- em l_ ,./ $4 a,. 'g '. f,,,,N _ + e .A A e..,' nl s -g. y ' ,,',4, ,J _y, ,,q ,4 %. ' d* I ,7,,. j G Id f-m "*,. g ,\\ 0 c ..:T.l p 'i y 7 , A,, G ; g/, (.,.

t..*.

d 'n n (3 .,s. w. i ff,Al4 1 e C '6. v,,. u, a, !! -lb._, !g, j .p~ }, j j-s .y k,, w.g... s p% .s s ,. Q '.. s - ' y. , f ; \\,. i. g* - . J, ,..4 ~K. .. < y-- 4 7 gI '1 s. 4 s t s , y Q ? ~. y' ',]_ f. I, 8

' g p ',*."$. 7

-y [ (f5 y t m '.,, d.7 j v. .'.2.P j

i. '

N /.s. A,tb ' ~ p*, *x* _,_-.f.e., __, m - ,.s r-+- - r. v. s*g. g ~.., r* ..e. , s -v a**"4J..W e % w 3, ,.,t, -e'

• • f

/ e in kg .,4 1rK dg g p [*g's i p.4 g, i,. ~. ,,,,,,.7 m fy *[7..

• ""l* w., 7 %,1 r'%$"k

-t,..; b ~ ~ .'t/O' 4"" ** *" Y N^**** f3 3, .u hY,3., ic r. p-..,f :.,, '; ) f Y .f. ,s. w u) "A '?

w. h, wli

w$ p$d @% D m

w M+Wi.k.. w<M~w.2~,

;w

!%MN N. g T v.; b,u.$ n.. n iw v g m,. - v!:,. s K M C,\\ 4-'r-m*- n.h"~ 1L <' : W. L u t g. e.m u.o , I :.3 e., s?

2.. h

. NA, 4.). i, m;y.

• eMg, t t,a; u.<

.4 . -4 y. ,.-. L,z.y mi. 3~ p r - a, G - 1 :,wa .ni, ;;3..t .c 5.. #, y%, s,., s s ,o G.ww. e o, lt ,, ~, n ,~ g, y' y s,c, 3..... ,ts. ~.4. m ., ~.,,. -- o y, g i f m'*l ~. N l '. } t .,,1 ,"p f 3 v. o, < >

• y.,,g :..s. 7 e-y-;.

,i u, M..- ~% ${ o ., L;,: k?f, c y1v.5,J. Y -a a O y Y W. Y ' '~ ' _u'

,,_'i.g _, _. ;_ fg ~~ "

7 O. -*y.; y'7'"~l ~Tl'"-'~~P"' Igg',,,I -s-. e T- - I l 3-12

~ _ .. -. - ~ _. 3.2 Initial Screening 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 LaSalle Final Safety Analysis Report as well as other data obtained from the utility and the r information gathered in the site virtt were reviewed for this purpose. A general guide for this task is the PRA Proceduren Guide (1983) which lists the poc,sible external events for a nuclear power plant. Table 3.2-1 is a listing of external hazards for the LaSalle County station. This table is similar to Table 10-1 of the PRA Procedures Guide. A set of screening criteria was daveloped which should minimize the possibility of omitting significant risk contributors while reducing the amount of analysis to manageable proportions. These screening criturla were described in Section 2.3 and are also listed at the end of Table 3.2-1. For each external

event, the applicable screening critoria and a brief remark are included in the table.

In _- the following paragraphs, the external _ events in Table 3.2-1 are discussed in more detail. Also, the reasons for screening some of the events are presented. Aircraft Imoact i A bounding analysis is performed for this event. Avalancha LaSalle County Station is built on a gently rolling terrain where there ' are no mountains. - Therefore, avalanches cannot . occur near the site. Bioloaical Events The only biological event which may affect safety of the plant is aquatic life in the cooling lake, i. e '., fish may block flow of water from the lake to the plant. This event is not-considered further because there would be adequate warning, and therefore remedial action can be taken. Coastal Ero4120 LaSalle County _ Station is located inland and therefore this event is not applicable to the site. Drouabt LSCS has been designed for the possible offects of droughts or -low flow rates in the Illinois River. The total capacity of 3-13 -,, a -. -, - -.. - -

the makeup pumps at the river screen house is 200 cfs which is much less than a 100-year low flow level of 1592 cfs in the Illinois River. In additlon, loss of watet from the Illinois River or from the cooling lake does not affect the ability of safety-related facilition to function adequately. The Ultimato 11 eat Sink (UllS) for the LaSalle is an excavated pond which is located under the southeast corner of the cooling lako area. In the unlikely event of unevailability of water from the cooling lake, emergnney water supply u.11 d be obtained from the U!!S. The Ulls has a 30 day supply of water based on the worst period of recorded weather conditions at the site. Therefore, in case of a worst possible drought there vould be enough time for remedial action to be taken. External Flooding A bounding analysis is performed for this event. Extreme H1D510 and Tornadoog A bounding analysis is }. eforned for this eve.t. f.QII 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. The effect of fog on aircraft landing of takeoff accident rates may be neglected because there are no airports within 5 miles of the site, i.e., only in-flight accidents contribute to aircraft hazard at-the site.- Forest Fire 3 There are no forests in the vicinity of the LaSalle site, -i.e., the. site has been cleared. Therefore, this event is not applicable to the site. 1 Frost Loads ' induced ' on - LaSalle 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. liail -Hail was considered as one of the meteorological conditions in the design of LaSalle structures (LaSalle FSAR).

Ilowever, 3-14

-x-.

- -...~ - l l ' hail is less damaging than other missiles which are generated ) [ outside of the plant such as tornado missiles and turbine i missiles. Therefore, hail is not considered further in the p scoping study. j Iliah' Tide, Hlah Lg};e Level or lilah River Staac High tido is not applicable to the site because the plant is located inland. High lake level and high river stage are considered in the bounding analysis under external flooding. I high! Summer TejlipsrAtnrg As mentioned under drought, - the Ulls is designed to provide a t minimum of 30 days water supply for cooling taking into account evaporation,

drift, seepage, and other water-loss mechanisms.

Therefore, high record temperatures were indirectly included in the design of LaSalle under drought conditions, llurricane LaSalle site is inland and thus is not affected by hurricanos. Ice Cover Ice loading is considered in the plant design along with snow loads. For this study, ice loads and snow loads are considered to act together (see snow loads). Indup_ trial or Military Facility Accident This c' vent is included in the scoping study. Internal Flooding _ This event is included in the detailed internal events analysis. Landslides The LaSalle plant is-built on-flat land where landslides are not possible. Lichtnina The plant ' structures and electrical systems are protected-by lightning ' conductors agEinst a current of 200. kilo-ampere (kA). In a study by the Electric Power Research Institute '(NSAC, 1981),. the range of predicted number of cloud-to-ground i 3-15

lightning strikes of 25 kA or larger is estimated to be from 1.8 to 11.6 strikes per square kilometer per year. Of these strikes, only one percent have current amplitudes in excess of 200 kA. -If the plant area is taken as 2,000' x 3,000', the annual _ frequency of lightning strikes damaging the plant systems is calculated to be from 10-3 to 6.4 x 10-2 Therefore, lightning events cannot be screened out on the basis of their frequency of occurrence alone. Studies performed by Sandia National Laboratories under the NRC research program TAP A-45 have estimated the frequency of severe core damage may be as high as 1.7 x 10-6 per year due to lightning strikes for a plant in the vicinity of LaSalle with a minimum AC/DC system (i.e., two electrical divisiono). The relevant scenario " station blackout" is the lightning i strike results in the loss of offsite power and the onsite electric power is unavailable due to random causes. Since LaSalle has three electrical divisions, additional damaging lightning strikes or random electrical failures must occur in order for this scenario to happen. Inclusion of -these additional events for LaSalle is judged to lower the scenario frequency below 104 por year. Since the lightning conductors are expected to sustain currents _in excess of 200 kA, the above estimate of damage frequency is expected to be overly conservative. - Also, the reactor building has metal siding permitting grounding of lightning strikes. Since the calculated frequency of damage is low, lightning is not expected to contribute to the plant core damage frequency and it will not be considered further in the current scoping study. The ef fects of lightning in inducing LOSP are included in the internal event quantification of LOSP and its time recovery curve. Low Lake or River Water Level This event is included under drought. Lov Winter Temnerature Low temperatures can affect the plant structures as well as the cooling lake or the Illinois River. Thermal stresses and embrittlement which _are induced _ by low temperatures are insignificant compared to other design loads. In addition,- these effects are covered by design codes and standards for plant design. Ice cover on the cooling lake or on the Illinois River does not affect the plant safety becanae of the availability of the ultimate heat sink. In case of an ice cover on the. ultimate heat sink,_there is adequate warning so that remedia' action can be taken. (provision..for ice melting in lake screenhouse forebay). 1 3-16 l l

t t ligtaorite This event has a very low probability of occurrence. A study by Solomon _(1974) showed that the probability of a meteorite impacting a nuclear power plant is negligible, and therefore ,1aeteorites will not be considered in the scoping study. P l Eingline Accide.Itt L This_ event is included in the scoping study. Intense Prenipitation This-event is ~. included under ex';ernal flooding. Release of Chemicale in Onsite storatte This event-'is included in the scoping study. River Diversion The-Illinois River in 5 miles away from the plant and the site -. i n approximately 180 foot above the river elevation. Therefore, ~ any river ~ diversion could not become a hazard to the. plant. Sandstorr This event is not relevant for the LaSalle site. D.iLiSh.Q This event is included under external flooding. i Seismic Activity This' event is included -in the detailed external events analysis. Snow ' Snow and-ice-loads were considered in the design of category-I structures._ The following statistics were calculated for the design of structures due - to local probable maximum precipitation (PMP) at the LaSalle site - (FSAR) : o ._100-year recurrence interval ground snow load = 24.0 psf o 48-hour probable maximum Winter precipitation = 15.9 inches-3-17

Prom those data, it was found that the corresponding vator load of snow and ico loads due to a winter PMP with a 100-year recurrence interval antecedent snow pack is less than the design load (83.2 psf) for the roofs of safety-related structures. The roof drains are designed for a precipitation intensity of 4 in/hr. Conservatively assuming that the roof drains are clogged at the timo of the PMP, the maximum accumulation of water on the roofs of safety-related structures is limited by the height of parapet walls, viz. 16 inches. The corresponding water load is thorofore 83.2 2 lb/ft. The roofs of safety-rolated structures at LaSalle can . withstand this load. Thoroforo, snow and ice loads are excluded from further study, apj1 5hrio bSyn R _Qgneel1dqtipn Plant structures are designed for the effects of differential cottlement due to consolidation. In addition, such offects occur over a long period and they do not poco a hazard during the plant oporation, i.e., the plant can be safely shutdown if

nooded, t

B orm1Surga This event is included under external flooding. Trnngoortation Ag_g.idnata A bounding analysis is performed for this event. Tsunami. LaSalle site is inland and therefore this event is not applicable to the site. Ipxic Gag 'This event is_ included under transportation accidents, onsito chemical - rolease, and industry and military facilities -accidents. Inrbino qsn m tsd Missiles 'A-bounding analysis is performed for this event. Y21 m jpg Activity The sito.is.not_close to any_ active volcanoes. 3-18

Eaves [ The LaSalle site is inland and therefore ocean waves can be excluded. Waves in the Illinois River or the cooling lake are included under external flooding. In summary, the findings of the preliminary screening are as follows: Aside from seinmic, fire and flood which have already been included in the detailed external hazards analysis, the following e/ents were identified for a more detailed study. 1. Aircraft Impact 2. External Flooding 3. Military and Industrial Facilities Accidents 4. Pipeline Accidents 5. Transportation Accidents 6. Turbine Missiles 7. Winds and Tornadoes 8. Release of Chemicals in Onsite Storage The above events are discussed in Sections 3.3 and 3.4. 3-19 .,.....u - -

Table 3.2-1 Preliminary Screening of External Events for laSalle County Station Applicable

• Event Screening Remarks criteria Aircraft Impact Included in scoping study Avalanche 3

Topography is such that no avalanche is possible Diological Events 1 There would be adequate warning for these events Coastal Erosion 3 LaSalle Site is inland Drought 1 LaSalle is designed for probable maximum drought. There would be adequate warning so that remedial action can be taken. External Flooding Included in scoping study -~~ ? Extreme Winds and Included in scoping Tornadoes

stud, Fog 1

It effects frequency of occurrence-of other

hazards, e.g.,

highway accidents,. aircraft landing and take-off a Forest Fire 1 -There are no forests in the vicinity of the site; site is cleared Frost 1 Snow and ice loads govern Itail 1 Tornado and turbine generated missiles govern High Tide, Ifigh Lake 4 Included under external Level-- or.'liigh River. flooding Stage

• See notes-3-20

~. ., -.. - - - - _ ~ ~ - - l Table 3.2-1 l Preliminary Scrooning of External Events for LaSalle County Station (Continued) l-Applicablo* l Event Scrooning itemarks i criteria liigh Summer 1 Ultimate heat sink is l Temperaturo designed for at loar' 30 days of operation, taking into account ovaporation,

drift, scopago, and other i

water-loss mechanioms; gives adequate warning. l llurricano 3 LaSallo sito -is inland and is not affected by hurricanos Ico Cover 1,4 Plant structures and systems are designed for the ico offects Industrial or Included in scoping Military; Facility study Accident Internal Flooding Included in external-events analysis Landslido 3 Topography is such that no landslides are possible Lightning. 1 Plant is designed for lightning. All buildings havo lightning conductors. Low Lake or River -1 The plant is designed Water Lovel for this condition. .Also, there-will be adoquate L warning so that remedial action can be taken. L . Low Winter 1 Thormal stresses and l Temperature ombrittlement aro insignificant or. covered by-l. -* Soc. notes l 3-21 F ~

Table 3.2-1 Preliminary Screening of External Events for LaSalle County Station (Continued) Applicabic* Event Screening Remarks Criteria design codes and standards for plant

design, generally, there.is adequate warning of icing on the ultimate heat sink so that remedial action can be taken.

~ Noteorite 2 This event has a very low frequency of occurrence for all-sites. Pipeline Accident Included in scoping study Intenso= Precipitation 4 Included under internal and external flooding Release of chemicals Included in scoping in Onsite Storage study River Diversion 3 Illinois river is 5 miles away from the plant. at a much lower elevation, i.e., river diversion could not become a hazard. Sandstorm 3 This is not ' relevant for this region Soiche' ~ 4 Included under external 4 flooding Seismic Activity Included in external events analysis . Snow Plant is designed - for snow . load ponding effects and-combinations of snow with other loads.

• See-notes 3-22

.s:

Table 3.2-1 Preliminary Screening of External Events for LaSalle County Station (concluded) Applicable * [ Event Screening Remarks criteria Soil Shrink-Swell 1 Plant structures are all consolidation designed for the effects of consolidation. Storm Surge 4 Included under external flooding Included in scoping Transportation Accidents study Tsunami 3 LaSalle site is inland . Toxic Gas 4 Included in transportation

accident, onsite chemical release and industry and military facilities accident.

Included in scoping Turbine Generated Missiles study Volcan'ic Activity' 3 The site is not close to any active volcanoes- [ Waves 3 LaSalle is inland ~

• llOTES :

l: 1.- The event.is of equal or lesser damage potential than the. l events for which the plant has been designed. 2. The event has a significantly lower mean frequency of Neurysrce than other events with similar uncertainties and ' auld:not result in worse consequences than those events. .he event cannot occur close.enough to the plant to affect it. -- 4. The event is: included in the definition of another event. 3-23

3.3 Screenina Qf Txternal Events Bqged on TEAR Information This section describes the external events which could be screened based on the FSAR information supplemented with new data. Section 3.3.1 discusses the military and industrial facilities accidents and Section 3.3.2 describes the pipeline accidents. It is shown that these accidents are unlikely to contribute the plant risk. a An accident scenario which is usually considered for a BWR plant like LaSalle is an explosion caused by the chlorine which in stored on site. However, the information which was provided by the Commonwealth Edison Company indicated that only a small amount of liquid chlorine is stored on the LaSalle site. Therefore, a. chlorine accident is not significant for the LaSalle County Station. 3.3.1. Accidents in Industrial and Military raci]ities According to the LaSalle

FSAR, there are no storage facilities, mining and quarry operations, industrial plants, or military facilities within 5 miles of the plant site.

The nearest industrial facility which stores hazardous materials i is'E.'I. DuPont de : Nemours and Company which is located in

Seneca, Ill.nois, approximately 5.6 miles northwest of the site.

.There are two other industrial plants within 10 miles of the site which store hazardous materials, namely Beker Industries and Borg-Warner - Chemical Corporation. Both of f these ' plants are located in Marseilles, Illinois, which is approximately - 6.8 miles north-northwest of the site. Table 3.3-1, which is duplicated from-LaSalle FSAR, lists all the hazardous materials, quantities - - stored, and mode of transportation - for the above mentioned industries. In addition to the fccilitier listed in Table 3.3-1, Tri-State Motor Transit, which.is a trucking firm approximately 5 miles northeast of the-sito,- has a holding area for trailers with explosive and/or sensitive loads. Since there has been no activity in this holding area and also there are no plans to increase the use of this area, Tri-State Motor Transit was not included in Table 3.3-1. There are'three possible-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 control room which ' could : incapacitate the operators, and (3)' flammable vapor. clouds-leading to heat hazard at the site.

Industrial accidents at distances farther than 5 miles'to the- -site are-not expected-to cause significant overpressure loads on the plant structures.

Also, the plant Category I structures are designed for Zone I torna 5 wind loads, i.e.,

~ the Category I structures have a minimum capacity of-3 psi 3-24 _., _ - _ _ _._ _ _ _ _ _ _.. _ _.. ~.

against blant loads. A detailed description of the Category I structural capacities is given in Section 3.4.4 under Transportation Accidents. Since an industrial accident at a distance of 5 miles or more would result in overpressures on wall panels which are less than 1 psi, an overpressure hazard due to in'*ustrial accidents could be screened for the LaSallna site. Flammable vapor clouds at a distance of 5 miles or more would not generate much heat at the site.

Also, the probability of a flammable cloud travelling a distance of 5 miles or more to the site is negligible.

Thus, flammable vapor clouds due to inductrial accidents will not be considered further in the LaSalle external events neoping study.

Release of toxic cheuicals near nuclear power plants can potentially result in the control room being uninhabitable. This condition can happer if: (1) large quantities of toxic chemicals are

released, (2) there are favorable wind conditions and insufficient dilution of chemicals such that these chemicals reach the control room air intakes, and (3) there are no detection systems and air isolation systems in the control room.

According to Regulatory Guide 1.78, chemicals stored or situated at distances greater than 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, atmospherin dispersion will dilute and disperse the incoming plume to such a degree that there should be sufficient time for the control room operators to take appropriate action. The control room !!VAC in LaSalle has redundant equipment and provides chlorine and anhydrous ammonia detectors with appropriate alarms and interlocks. Provision has been made for the control room air to be recirculated through charcoal filters and also provision has been made to pass outdoor makeup air through impregnated charcoal filters before introduction to the control room system. From the foregoing discussion, the following conclusions are made: 1. The only toxic chemicals which are stored in large quantities near the site are chlorine and anhydrous ammonia. The control room is equipped with detectors for chlorine and anhydrous ammonia and therefore they would not pose a hazard to the plant. The only other hazardous chemical which is stored in large quantities is Butadiene, ll o w e v e r, the maximum quantity of Dutadiene stored at the Borg-Warner chemical plant (Table 3.3-1) is well within the allowable limit which is calculated based on the Regulatory Guide 1.78 criteria. 3-25 l .h

2. Even if there is an accident at the Dupont chemical plant in Sonoca,

Illinois, the hazardous chemicals have to travel a distance of more than 5 miles and an olevation of more than 180 feet before they reach the control room air intakes.

Therefore, it is concluded that the probability of coro damage due to an industrial. accident is negligible. 3. Overpressure and heat load due to industrial accidents at a distanco of more than 5 miles would not affect the LaSalle plant. 3.3.2 pipelino Accidents j The LaSalla FSAn information is used to show that the i probability of damage to LaSallo structures due to a pipelino accident is negligibly small. According to the FSAR, thoro are no gas pipelines or oil pipelines within 5 miles of the sito. However, there are two natural gas pipelines betwoon 5 l to 7 miles of the sito which are operated by Northern Illinois Gas company. T1.ese-pipelines are 6" and 8" pipes and operato at 230 psi pressuro._ Both of the pipelines are buried approximately 30 inches below ground. These two pipelines are not used for storage and are not likely to be used to transport ' store any product other than natural gas. An accident in a gas' pipeline would load to either a fire or an explosion. In any of these ovents, the distance from existing pipelines to the LaSalle sito is such that there t<ould-be no damaging offect on the plant structures. i I + r 3-26

m._._ . _ _. -. _. _. _........ -. _ _ m._. .__..__._--_m_m, 1 Table 3.3*1 Industries with Hazardous Materials Within 10 Miles of the Site Pacility (Location) Has tenen Hade of Quentitles Transportation Bakar Industrleal I (Haastilloo) I r l AnhyJsous anruar.ta 10,000 t,on bange i i Sulfuric atid 3,000 ton truck I Dynamite 100 lb. I Wet procene 7,500 ton tell & truck 3 hoe % orte ocid l Illinola Nitrogen Corp.I Dieteellies) l Anhydrous anmonia 42,000 ton barge +111&nois River all-Chiceso-l Rock !aland & Pacific Truck

• e U.S. Hwy, 6 Boti pallied berge*

rail-tauck Liquid blended-ba r s e-- i ~ rail-L.P.O., gasoline, (small goantittee f1 fuel oil. chlorine for plant use only) i E.I. DuPont de Nemours -A Co,8 (Seneca) . Anhydrous armonia 150,000 lb, barge 30,000,000 lb, at reti-Chicago Rock !aland & Pacific Seneca Port Operating Authorit.y Storage Penn Central i Honomethiamine 2'0,000 lb. rail-Chicago Rock ' Island & Pacific Fenn Central Honomethiuminenitiate'. 7,000 lb. used in high i explosives manu-facture-not Shipped N dric' acid 56 801 3.250,000 lb. rall* Chicago Rock Island & Pacific truck .Mit:Ac acid 95 460,0D0 lb, rati-Chicago Rock Island & Pacific truck 3-27

-.2.

_.._......-. a. _, .u -... u.. 2. .u.

~..~., ~.-.. ~-.-.-. .-_---..n .~n_~.--..w..c Table 3.3 1 Industries with Basandous Hatoriale Within 10 Miles of t.he Site (Continued) t I Facility (Location) Hamimum Mode of Quantities Transgertation-l tilzed acid (Natric) 450,000 lb, rait-Chicago Rock Island & Facific truck Aramntwo nitrate 7,600,000 lb. rail-Chicago Rock I prills !aland & Facific j truck Dynamite 80,000 lb. truck initiating erplosives 40,000 teach) truck (caps) [ Initiating explosives-7,500 teach) truck -(primers) Jet tappets 3,000 (each) truck (orplosives) Nitrocellulos. 300,000 lb. truck p (alcohc,1 wet) ~ Chlorine (H P. 7,275 lb, truce cylinders) Anstentum 650,000 lb, ra' .ago Rock - nitrate liquor (80 ad & Facific equeou6 solution) Watog gel 2,1t>0',000 lb. - truck ._ high' explosives) ( -1 Aluminura powder 200,000 lb. truck 011sonite" 100,000 lb. trut.k Vinyl acetate 480,000 lb. , tail Chicago Aock Isler.d & Facific . Liquid ethylene 100,000 truck-State Highway 47 or U.C. Highway 6 -{ Nitrogen (lieruid) 4.000 lb. truck State' Highway -i 47 or U,$. t Highway 6 Nitrogen (gas) 45.000 ft trucu State Highway: 47 or U.S. Highway 6 - - Hethanol 40,000 1ho, la 55-truch Stat.e Highway gal, drums 47 or U.S. Highway 6 Ip 3-28 L a- ___._._,u.._. . _,, _ _.. -,,..-_, a,

Table 3.3 1 Industries with Hasardous Materials Within 10 Miles of the Site (Concluded) Facility (Locationi Marinas hade et Quantittee fransportation j^ Tormaldehyde $0.000 lb. truck State Highway 67 or U.8, Bishway 6 l l Bots Wernet Chemical, i ' Fora Warner Corp.' ) (Hatseilles) Acrylonitrile 500,000 gal tell Butadiene 1,006,000 set targe, rail Nitrogen '$50,bu0 ft truck Buifuric acid 30,000 gal, 9$ acid truck ) Fuel ett 1,200,000 set truck j i 1 Source: Mr. W. H.~ Fresor, Plant'Hanager, Beker Industries, letter to J. C. Frey, Cultural Resource Analyst, Sargent & Lundy, August 13, 1973. l A Cource: Mr. R. P. Feser, Manager, Illinois Nitrogen Corporation, letter to J. C.

Prey,

Cultural Resource Analyst, Bergent & Lundy, July 7, 1975.

3 Sou cet Mr. J. D, Graham, E. 3. DuPont de Nemout s & Cornpany, letter to J. C. Frey, Cultural l Resource Aanlyst, Sersert & Lundy, August 6, 1975 ' Sourcot ~ Hr. K. T. Bruns, Project f.netneerles Hanaser, Bors* Warner Chemicals. Bors-Watner Comporation. letter to J. C, Prey, Cultural Resource Analyst, Sargent & Lundy, September 9,

1975,

' Reproduced from the LaSalle FSAR [ 3-29

6.,e y

-...,..v.,.- m...,,,.w ..,.. ~,., ,.y.,.y, p.n,. re.,. .,,w.,

3.4 Doundino Analysis The external events which may be expected to contribute to the plant risk are-included in this section. A bounding analysis is: performed for each external event to find the annual frequency of core damage due to -the event. Section 3.4.1 describes the general methodology of a bounding analysis, and Sections 3.4.2 through 3.4.6 describe the analysis for each individual external event. The events which are inclr id in 'this - section are aircraft impact, winds and torri .oes, transportation accidents, turbine miselles, and external flooding. 3.4.1 Model, Uncertainty, and Acceptunce/ Rejection Criterion The pro}. bilistic models used in bounding analyses should integrate the randomness anl uncertainty associated with loads, response analysis, and capacities to predict the annual frequency'of the plant damage. Thn aim sf the present study ' is '. to use conservativo models for calculating the annual frequency of core damage. Obviously, if both the median frequency <and the high confidence'(e.g., 95 percent) value of frequency according to the conservative r.odel are predicted to be low (e.g., slo-7/ year), the external event may be eliminated from further consideration. The bounding analyses would therefore identify those external events which need to be studied in more detail as part of the PRA external events analysis. Elements of a complete bounding analysis are described in Section 2.4. For some external events,-it is possible to perform a bou'nding analysis without a _ structural response _ analysis. In effecte one could show that the frequency of exceeding design loads is very small. Since the design capacities wnich are based on the design loads are also conservatively defined, the external ev6nt would ra?. contribute significantly to the plant risk. This approach is used in analyses for transportation accidents and external floodi nt In - a _ complete bour?ag analysis, one needs the probability. distribution of leaa as well as the conditional probability distributions (fragilit en) of those components which: appear -in the plant syster 1d,icident sequence analysis. _The-londs-are usually defined q terms of a hazard curve which shows the annual frequency' ceedence for different load levels. The wertainty in the nazard analysis.can be represented by ' eloping a family of hazard curvec where each hazard curve s assigned a: subjective _ probability. An example of this plot i -can:be found in the bounding analysis for winds and_ tornadoes (Section 3.4.3) where the hazard curves are plots of ? 30

the probability of ec eedence versus maximum tornadic wind spoM o. The component. fragilities are also developed as a fam y of fragility curves which represent the median fragility curve and the uncertainty in the median fragility. The probability of core damage (CD) can be expressed as: P[CD)=[P if n C gg < Ryg - g(x)dx (3.4-1) L= f i x 1 j where Ci$ is the capacity of component i in cut set j, R j "is i the resiMance of component i in cut set j, and ft (x) is t.he probability density function of input load. The first term in the above integral represents the component fragilities appearing in the plant sequence and system analysis and the second term is the slope of the hazard curve. For the present study, some simplifications to the above equation were introduc.d. One simplification was to represent each cut set by only one component. As an example, back-face scabbing of the auxiliary building walls in case of an aircraft impact was assumed to lead to core damage even though a sequence of failures is necessary to lead to this damage state. In addition to calculating a point estimate (median) frequency o. mre damage, the uncertainties in hazard and component .iAties may be used to find the high confidence (95 e rc., ) frequency of damage. An uncertainty analysis is c

• r. 1 on11 if the external event leads to a best estimate a

fregun. icy which is close to the rejection frequency Jv ear). For this reason, uncertainty analyses were i pn n.ed for winds and tornadoes and aircraft impact. An unca tainty analysis was not performed for transportation accidents and external flooding because these events were shown to contribute insignificantly to the plant risk. For turbine missiles, the results include a best estimate frequency as well as confidence bounds based on the-FS AR analysis and other recent information. 3.4.2 Aircraft Impact An arsessment of the risk from aircraft crashes into the LaSalle structures is presented in this section. For this

purpose, information in the LaSalle FSAR as well as more recent date concerning airports, air corridors, and aircraft activity nt. c r the site were used.

An attempt was made to correct the data for anticipated changes in aircraft activity near the site. It was concluded that the frequency of plant 3-31 ] 4 --_____.s-___ _. ____ . _ _ _ _. _ _ - -. _ - _ ~ -. _ _ _ _. _ _ _ _ - - _ _ _. _ - _ _. _ _ - _ _ - - _ _ - _ _ - _ _ _ _ _ _ - - _ - _. - - - - - -

~, -- -~ - damage states initiated by aircraft crashes is on the order of 10 7 year. Section 3.4.2.1 describes the information in / 5 x ~ FSAR and Section 3.4.2.2 describes tho'present aircraft hazard analysis. 3.4.2.1-FSAR Information The LaSalle FSAR includes a description of airports and aircraft activity near the site. According to the LaSalle FCAR,-there are no commercial airports within 10 miles of the site and there are no private airstrips within 5 miles. Tables 3.4-1 ~ (LaSalle FSAR) and 3.4-2 (LaSalle FSAR) list all conimercial airports and private airstrips within 20 miles of tha site. As indicated in Table 3.4-1, these commercial airports can handle both single-engine and twin-engine aircraft. The annual number of operations for commercial aircraft is also given in Table 3.4-1. The aircraft using the private airfields are very small single-engine aircraft. The number of operations for private airfields near the site is expected to be low and, in addition, the random path of i these aircraft would make the potential risk to the plant negligible. There are three airway corridors within 10 miles of the site. These airway corridors are approximately 8 miles wide, and most aircraft fly within two miles of their centerline (Figure 3.4-1 (LaSalle FSAR)). All the traffic on these airways are expected - to conform to the ' FAA regulations concerning.the minimum low altitudes, i.e., all aircraft must fly at least 1000 feet.above the tallest object in the corridor. According to the FSAR, aircraft hazards can be excluded from the external events analysis because of the following reasons: 1. There are no federal airways or airport approaches passing'within 2 iniles of the station. The closest airway-corridor is 3 miles away from the station. 2. There are no commercial airports existing within 10 miles of the site and there are no private airstrips within 5 miles. -3. The projected landing and take-of f operatixns out of those airports located wlthin 10 miles : of the site are. far less than 500*dZ per year, where d is the distance in miles. The projected operations per year for airports located outside-of 10 miles is less than 1000.d2 per year. 4. There are no military installations or any airspace usage for military purposes within 20 miles of the station. 3-32 { L l av,- e wf+m-- -e L.-r -e

-3.4.2.2 -Update on FSAR Inf'ormation In order to perform a bounding analysis for aircraft impact at the LaSalle site, the-information in the FSAR as well as new information on - aircraf t activity near the site wac used. Recent traffic data was provided by the FAA to Sargent and Lundy Engineers in the June 15, 1984 letter to S. IIallaron. Table 3 4-3 summarizes the FAA data which was gathered for June 7, 1984. Among the air corridors in this table, routes V156 and V9 are approximately within 3 miles of the site, whereas routes Vll6 and V69 are approximately 7 miles away from the plant. Other airway corridors in Table 3.4-3 are far enough from the site such that they would not contribute to the. _ aircraf t hazard as discussed in the next paragraphs. According to the FAA letter, aircraft listed.as flying at 9000 feet and below (96 percent) are single and twin-engine light aircraft. Also, aircraft listed as flying at 10,000 feet and above (92 percent) are three and four engine heavy. jet aircraft. Although - the information which is presented in Table 3.4-3 is for one day traffic only, the data was provided for a peak traffic day and it is felt that it could be used to conservatively estimate the annual traffic volumes. In addition, the data in Table 3.4-3 were increased by-50 percent and then used in the bounding analysis to account for future increases in aircraft activity during lifetime of the plant. 3.4.2.3 Aircraft Impact Bounding Analysis The methodology that is used to calculate the frequency of aircraft impact has been described in the Midland Probabilistic Risk-Assessment. The probability of an aircraft j impact on the plant structures may be written as: l J f -EEN A d p p-(3.4-2) l-ij PJ where N j = Number of 'aircraf t operations of type j along i airway 1, Aj = Crash rate of aircraft type j, dj = Distance traveled by aircraf t type j where. the site is within striking distance, AkJ " Crar area of the structures, Apj = Area where the aircraft may crash. 3-33

,m. _ _ . ___~ _ - _. _. The -term A j/ Apj in Equation ~ (3.4-2) represents the k probability of an impact'given-a crash in the vicinity of the site.

This: probability and also the distance dj are determined -- geometrically.

The other-variables in' the above D equation are assigned-distributions representing our state of h knowledge about their valuese - Figure 3.4-2 shows the geometry of an aircraft accident. Assuming that the aircraft is disabled at an elevation h, the distance that it would travel before the crash is gh where g is - the glide distance per unit of altitude lost. For the present. study,- it is assumed that there is an equal probability of crash termination anywhere in the sector of radial-length gh and angle 4 = 180' in front of the aircraft. Aj is the half circle defined by radius gh where g Therefore, p - was assumed to be the maximum glide ratio, equal to 17. Aj k is.the impact area of structures which is minimum when the aircraft crash is vertical and Jt is maximum when the glide ratio -'g is maximum. An average value of the two areas was k skid used - for _ A j in the present study. In addition, a -which distance of 100- feet was assumed for the aircraft - increases the structure impact area (A j)- k The - aircraft impact frequency in Equation (3.4-2) was calculated for different types of aircraft. In this study,- three types of aircraft were identified for these calculations, i.e., single-engine, twin-engine, and commercial aircraft. Also, :a fragility analysis was performed to determine whether these aircraft types are capable of inducing damage to the Category I structures in case of an impact. Capacities' of Category. I structures against aircraft impact were determined using the formulas which have been developed for impact of non-deformable missiles on reinforced concrete walls and panels. For an aircraft, it may be assured that the engine' and part of the aircraft body represents the non-deformable missile. Information regarding-the characteristics of -single-engine and twin-engine aircraft was obtained from Niyogi, et al. (1977). Also, it was conservatively assumed - that if an: aircraft impacts one of the Category I structures and causes back face scabbing, it would lead to a plant damage state. Another conservatism is that all impacto are assumed to be normal, glancing impacts would have less chance of - causing damage. The formulas which - have been developed-to predict the minimum scabbing thickness all indicate that the concrete; wall thickness = required to prevent scabbing is independent of the amount of steel reinforcement for low to moderate steel ratios. The formula used in -this study was developed by Chang (1981). Chang's formula is based on full-scale and model impact tests. According to chang, the minimum 3-34

wall-thickness (inches) which is required to prevent scabbing (ts) is given as: 0.4 0.67 9 y t = 2.47 g o.2 0.4 (3.4-3) d g where w = weight of missile (lbs), v = velocity of missile (ft/sec), 4A d = missile effective diameter (inches) =1 fc = ultimate strength of concrete (psi), Ac = contact area of missile (in2), The results indicated -that a single-engine aircraft must be traveling at s'peeds faster than 200 mph at the time of impact to cause scabring of'2'6" reactor building walls. Since this velocity is 'n the range of the maximum velocity of single-engine aircraft, it was concluded that single-engine aircraft would not 6amage the reactor building in case of an impact 'below Eleva'clon 843'. However, a single-engine aircraft could cause damage to the reactor building if it _ crashes into the building'above Elevation 843_' (which has metal siding _ walls) and penetrates the slab at this elevation. It should be noted that there is no safety-related equipment in the reactor building _ at Elevation 843', so in this analysis only twin engine and commercial aircraft will be considered. The auxiliary building at LaSalle is surrounded by the turbine

building, the diesel generator buildings, and the reactor building.

A fragility evaluatioa of the auxiliary building walls at LaSalle showed that only twin engine and commercial aircraft are capable of scabbing the auxiliary building walls. Because the _ auxiliary _ building down to Elevation 786'6" does not contain any non-redundant safety systems, a single-engine aircraft impact at the higher floors of the auxiliary building would not cause_ damage to critical equipment. Also, the lower elevation walls of the auxiliary building are thick enough to withstand a single-engine aircraft impact. The diesel generator building for Unit II at LaSalle was excluded from the aircraft impact risk calculations because of the following reasons: 1) the diesel generator building is 3-35

much-smaller -than the other buildings '(less impact area),

2) _ it is shielded on two sides by the_ reactor building and auxiliary building,_ and 3) while a crash into this building might fail two diesel generators and_also result _in loss of of f site-power to Unit II only (which enters near the building), the swing diesel is in the Unit I diesel generator building on the opposite side of the plant and AC power would still; be available.. The conditi'onal probability of getting core damage by crashing into the diesel generator building is,-

therefore, much smaller than for the other buildings. The crash rate statistics for different types of aircraft are listed in Table 3.4-4. These statistics were calculated from the 10 years of crash data involving air carriers published in the FAA Statistical Handbook of Aviation (1979) and accident rates for general aviation aircraft published in the Annual Review of Airport Accident Rates by the-National Transportation Safety Board (1980). The statistics in Table 3.4-4 were calculated assuming a lognormal distribution for aircraft crash rates. 1 Table 3.4-5 summarizes the results of LaSalle aircraft hazard ~ bounding analysis. These results were obtained assuming that single-engine aircraft fly at an average altitude of 4000 feet -and twin-engine aircraft fly at an average altitude of 5000 feet. For commercial airplanes, data for air corridors near the site was used to estimate average aircraft altitudes. As shown in this table, the point (median) estimate frequency of ~ an aircraft impact on the LaSalle structures leading to a 10-7 year. It is plant ; damage state is approximately 5 x / noted -- that mostL of-the contribution to ' the risk comes from twin-engine aircraft. These aircraft have much higher crash rates than commercial aircraft. 3.4.2.4 Aircraft Impact Uncertainty Analysis The aircraft impact bounding analysis for LaSalle showed that-the median frequency of - plant _ damage due to a crash is 5 x 10-7/ year. -In order-to evaluate the uncertainty in this frequency, -distributions of the random'. variables -in Equation (3.4-2) have to' be identified. For this purpose, the probability distribution of._ crash rate was obtained from the iFAA data. In addition, distributions of ' the other random variables' in Equation (3.4-2) were obtained from subjective l engineering judgment. It was assumed-that for each aircraft type j, the random variable representing uncertainty in crash rate (c fj ) can be'modeled as: a 2 cgj.= cN (1 Ch Eg ( 3. 4 -4 ) - l: I' ? i 3-36

where the c's are lognormal variables with median equal to unity and logarithmic standard deviation denoted by B. Therefore, for each aircraft type j, the logarithmic standard deviation-of crash rate pfj may be written as: + Pf + P +O O.W P P y g g3 g

However, since commercial aircraft do not contribute significantly to the i-isk, they could be excluded from the following ancertainty analysis.

From the FAA crash data, px for single-engine and twin-engine aircraft were found to be 0.10 and 0.15, respectively. The logarithmic standard deviation in aircraft altitude (h) was obtained by assuming that the median altitude for single-engine aircraft is 3000 feet and the 95 percent value is 4000 feet. Therefore, #h can be calculated as 4000 s- - 0,17 (3.4 6) h 1.65 This #h was used for the twin-engine aircraft. As discussed previously, a factor of 1.5 was applied on the number of aircraft operations to estimate the median value of the operation activity accounting for future increases in the aircraft activity near the site. Assuming that a factor of 2.0 represents the 95 percent value, #g was calculated to be 0.17. For glide ratio (g), it was assumed that the value of 17 which was used in the analysis is the best estimate and a glide angle of 10' represents the 99 percent value. Thus, p g was determined to be 0.47. Using Equation (3.4-5), Srj for twin-engine aircraft was calculated to be 0.55. Assuming a lognormal distribution for the annual crash frequency, the 95 percent confidence bound was found to be 10-6 Therefore, based on our model, the high confidence (95 percent) frequency of impact resulting in damage is expected to be in the same order of magnitude as the median frequency of impact. I 3-37 l

3.4.3 Winds and Tornadoes This section describes the bounding analysis of LaSalle structures for the offects of winds and tornadoes. Both seismic Category I structures and non-Category I structures I were considered for this task. Seismic Category I structures at LaSalle have been designed for both extreme wind and tornado load effects. Therefore, they are expected to have a high capacity against extreme winds and tornadoes. Non-Category I structures at LaSalle were generally designed against wind loads. However, in the design of the plant, non-Category I structures were shown to not collapse on adjacent seismic Category I structures, if any, in the event of a tornado. 3.4.3.1 Plant Design Criteria Cateaory I Structures A design wind velocity of 90 mph based on a 100-year return period was used for Seismic Category I structures (i.e., reactor building, diesel-generator building, and auxiliary building including control room) at LaSalle (LaSalle FSAR). For the purpose of structural analysis, dynamic wind pressures on the structures were converted into equivalent static forces which vary along the height of each structure. In addition, Category I structures at IaSalle were designed to withstand a Design Basis Tornado (DBT) which is defined as follows: o maximum rotational velocity of 300 mph o translational velocity of 60 mph o external pressure drop of 3 psi at the vortex within E a 3-second interval o radius of maximum wind speed of 227 feet Pressures due to both wind velocity and tornado velocity were assumed to be static in uhe design of the structures at LaSalle. Since the natural periods of buildings at LaSalle are short compared with the rise in time of applied design pressures, the above assumption is well justified. A comparison of the design wind loads and the design tornado loads along with the corresponding allowable stresses revealed that the tornado loads are more critical. Therefore, it is sufficient to limit the bounding analysis to tornado loads for the Category I structures which were designed for both winds and tornadoes. 3-38

The safety related structures at LaSalle were also designed for the effects of postulated tornado missiles. The used in the design of Category I postulated tornado missiles structures are as follows: o Wood plank, 4 in. x 12 in. x 12 in. impact velocity = 225 mph Automobile weighing 4000 lbs, 20 ft2 front area, o impact velocity = 50 mph The reactor building superstructure above Elevation 843'6" has roof. The metal siding has decking retal siding and metal at wind speeds much less than that been designed to blow off of the DBT. However, there are no ESF equipment at this elevation in the building. Non-Catecocv I Struslures The non-Cetegory I structures (i.e., turbine building, radwaste building and service building) at laSalle have been designed to withstand the effects of 90 mph wind velocity. The turbine building which adjoins the auxiliary building is designed such that it will not collapse on the auxiliary design basis tornado strike. The building as a result of a missiles produced by the tornado induced damage of non-Category I structures (i.e., girts, subgirts and parlins) are less damaging than the spectrum of missiles generally spec'ified in the Standard Review Plan. The bounding analysis of LaSalle structures for extreme winds, in tornado winds and tornado generated missiles are described the following sections. Seismic Category I Structures 3.4.3.2 controlled by the The design of Category I structures was tornado loading and tornado missiles. For this reason, the bounding analysis described herein addresses only tornado effects. The probabi]ity of straight winds exceeding the capacity of Category I structures is much smaller than the probability of tornadic winds exceeding the same capacity. 3.4.3.2.1 Tornado Loads The probability of structural failure resulting from tornado strikes on LaSalle Station structures is calculated using the tornado occurrence data and plant design features. It is shown that the probability of tornadoes striking the plant 300 mph is structures with tornadic wind speeds in excess of 3-39

of the order of 10-7 per year. Even if the plant structures are assumed to fail at this design value (of 300 mph), the contribution of the tornado events to the plant risk is negligibly small. 3.4.3.2.1.1 Characteristics of Tornadoes Tornadoes are rare events which are usually characterized by their rate of occurrence, direction, maximum intensity, ' path length and path width. The most important aspect of a tornado is its maximum wind speed. Other characteristics of a tornado such as velocity, pressure, and pressure drop can be estimated from maximum tornado wind speeds. The bounding analysis used in this study is based on the methodology described in Reinhold and Ellingwood (1982). In this approach, the tornado hazard curves at the site are developed in terms of maximum tornado wind speeds, i.e., the hazard curve is a plot of annual frequency of exceedence for a range of maximum tornado wind speeds. It will be shown later in this report that such tornado hazard curves are dependent on the geometry of structures exposed to tornadoes. Tornadoes are usually classified accordir.g to their intensity. The most common classification of tornadoes is the Fujita F-Scale and Pearson length and width scala (FPP) which is a measure of destructiveness of a tornado (Fujita and Pearson, 1973). In this scale, tornadoes are assigned a number from 0 to 6 (FO - F6) with higher numbers indicating higher intensity tornadoes. Table 3.4-6, reproduced from Fujita with permission, shows the FPP classification of tornadoes along with intensity scale, length scale, and width scale.

Also, listed in Table 3.4-6 is an area intensity scale whicn is based on total damage area.

The F-scale intensities are assigned using a qualitative assessmeic of the worst damage that occurs during a tornado. This is usually accomplished by observing the damage to residential buildings or other structures and calculating the pressure that is needed to cause the observed damage. From calculated tornado wind

pressure, one can find the maximum velocity which could generate such pressures.

Since classification of tornadoes is based on observation of damage rather than direct measure-ment of wind speed, two types of errors can be introduced in this process. Direct classification errors are due to inaccuracies in assigning intensity scales to tornadoes whereas random encounter errors are due to lack of damage observation. The uncertainty due to direct classification is expected to be unbiased, i.e., it is equally likely errors that a tornado is underscaled as it is overscaled. On the other hand, random encounter errors are due to the lack of damage medium in a tornado path which could subsequently be used for the tornado classification. Therefore, random 3-40 u

-. ~ - _. - encounter errors are always associated with underestimating the tornado characteristics. Another source of random encounter errors -is that small tornadoes are often undetected in unpopulated areas. As-an example, increased public awareness has led to a trend toward increased reporting of weaker tornudoes in recent years whereas the average number of strong-tornadoes reported is basically unchanged (Twisdale and Dunn, 1983). This error would tend to underestimate the rate of occurrence of all tornado intensities but it would overestimate the occurrence rates of higher intensity tornadoes. An--attempt was made in the study by Twisdale and Dunn (1983) to correct the reported tornado data for the above errors. The tornado hazard model in this study includes the following elements: l l o variation of tornado intensity with occurrence frequency; the frequency of tornado occurrences decrease rapidly with increased intensity o correlation of width and length of damage area; longer tornadoes are usually wider o correlation of area and intensity; stronger tornadoes are usually larger than weaker tornadoes o variation in tornado intensity along the damage path length; tornado intensity varies throughout its life cycle o variation of tornado intensity across the tornado path vidth 3.4.3.2.1.2 Tornado Occurrence Rate As a first step in the bounding analysis, the frequency of occurrence of all - tornadoes (irrespective of their inten-sities) at the site was-calculated. Based on historical data, the frequency of occurrence of all tornadoes at LaSalle County has been-reported to be 1.7 tornadoes per year for a l' x 1* square (LaSalle FSAR). Assuming a Poisson process for the occurrence of tornadoes, mean-arrival rate of tornadoes at the site is found to be 4.8 x 10 4 tornadoes / year-square mile. The calculated occurrence rate for the LaSalle site is compared to two other tornado risk regionalizations. Figure 3.4-3 shows the tornado risk regionalization scheme which was reported by WASH-1300- (Markee et al., 1974) and Figure 3.4-4 shows the regionalization scheme which was proposed by Twisdale and Dunn (1981). Regulatory Guide 1.76 (USNRC) describes the design basis tornado for nuclear power plants and has adopted the 3-41

~.. ~, - . ~

scheme in WASH-1300.

The occurrence ratec _for each region is shown in Table 3.4-7,-reproduced from Reinhold-and Ellingwood. These occurrence rates - have been corrected for - possible unreported tornadoes in sparsely populated areas. It is noted -that using either regionalization scheme, the occurrence rates of 4.12: x 10-4/ year-mi2 for Region I or 5.18 x lO-4/ year-mi2 for" Region A compare favorably with the calculated occurrence rate of 4.8 x 10 4/ year-mi2 for the LaSalle site. 3. 4. 3. 2 ~.1. 3 Tornado Hazard Model Using a Poisson process for occurrence of tornadoes, the probability of a tornado striking the structures during time T with a velocity exceeding V* may be written as: P[ strike-by tornado with V > V*] = vT E[V( A ) > V* ( A ) ) I I (3.4-7) where v is the mean arrival rate per unit area per year for the' site, V(A ) is the velocity in an area AI which will be I defined below, and E(. ) is the expectation operator taken over all tornado parameters. Figure 3.4-5, (reproduced from Garson, et al, 1974 with permission) ' shows a rectangular structure with dimensions A and B. Assume that-this structure is approached by a tornado that travels at an angle a measured from the side B.

Also, let us assume _that this tornado travels a distance equal to L and the damage is limited to width W during. lifetime of the tornado.

Knowing the above information, - one can define an area AI where any tornado initiated in this area would strike the structure. Here, the point of initiation for the. tornado .is assumed to.be the mid-point of width W,. in general the but following results' are ' not dependent on this assumption. The area. AI is shown -in the lower part of Figure 3.4-5. Using simple geometry, --it is observed that AI is made up of four distinct _ regions (Garson et al., 1974). 1. The sum of the-areas _ denoted by T1 and-T2 iu equal to the= total tornado damage area _WL. 2._ The area denoted by P is equal - to HL where H is the projection of the structure on a line which is perpendicular to the tornado path. -3.

The areas denoted by BA1 and BA2 sum to the structure

-area AB. 4. The areas denoted by E, E, E3 and E4 sum to WG where I 2 G is the projection of the structure on the tornado path. 3-42

i Therefore,-- it is observed that_ the tornado will st.rike the structure if it is initiated within an area AI given by -AI = WL + HL + WG + AB (3.4-8) The first term in. Equation (3.4-8) is the tornado damage area whereas the next two terms indicate an interaction between the-tornado and the structure. Finally, the last term in Equation (3.4-8) is the structure's area. Thus, the_ tornado hazard curves for a site are expected to depend on the structure's size. For typical - structures struck by tornadoes, the last two terms in Equation - (3. 4-8) may be neglected and AI may be written as AI = WL + HL (3.4-9) where - WL is the area for a point structure and HL is-the lifeline: term which also contributes to the probability of a tornado _ strike. -Normal)v, one would integrate the results over the-probability dist.zibution of angle a for all possible tornado strikes. For this study, angle a was conservatively chosen such that it would maximize the second term in Equation. (3.4-9), i.e., H was chosen as the maximum projection _ length o f-the. atructure. In the following paragraphs, a matrix formulation for calculating the annual frequency of tornado-strikes with-V > V* is presented which accounts for both terms in Equation (3.4-9). The probabilistic model for-calculating tornado hazard curves i l at the site may be briefly= _ described as follows. The occurrence of- - tornadoes in this model is assumed to have.a Poisson' distribution (Equation (3.4-7)),-1.e., the probability distribution of-toraado inter-arrival times is assumed to be exponential. Given that a tornado has. occurred at the site, the conditional probability of the-tornado intensity scale (FPP) is then based on historical data. Next,.for each tornado-intensity scale, one has to determine-the average or the expected value of tornado area (WL).and tornado path length (L) which is to-be used in Equation (3.4-9). Thus, one can calculate -the expected value of area AI for each tornado intensity scale (FPP). Assuming that the maximum tornado Wind velocity for each FPP intensity scale is the mid-point of the velocity scale as reported in Table. 3.4-6, the probability of a> tornado strike with maximum wind speeds exceeding a given-velocity V* is equivalent to the probability of that tornado 3-43

being initiated in the area A. As an example, an F3 tornado I in Table 3.4-6 would correspond to a maximum wind velocity of 182Lmph. .Also, one can calculate a corresponding AI area for F3 tornadoes. Therefore, the probability of exceeding 182 mph . winds at the. site is equivalent to the probability of an F3 . tornado occurring in the corresponding A1 at the site. However, the problem is complicated by the fact that an F3 tornado does not exhihit a-uniform level of damage along its . path. A detailed description of the probabilistic model is given in the next paragraphs. . Table 3.4-7 shows the variation of tornado intensity with-occurrence for the regions which are identified in Figures 3.4-3.and 3.4-4. The occurrencerintensity (OI) relationships in-this.tablo are based en historical data and they have been corrected for direct classification errors and random encounter errors. Each row of Table 3.4-7 3s a vector (OI}- which shows the conditional probability of each F-scale inter.sity tornado given that c tornado has occurred. As stated previously, each tornado FPP scale is also associated with an area scale, a length scale, and a width scale as shown in Table 3.4-6. For example, an F4 tornado is expected to'have a damage area of 1.0 mi2 to 9.999 mi2 On the other hand, it is possible. for an F4. tornado to have a smaller or a larger damage area. The same statement may be made about the length scale and width scale of tornadoes which are listed in Table 3.4-6. For the present study, one is-interested in the expected value of tornado-damage area -(WL) for each FPP intensity scale. These average areas may be calculated from historical measured damage areas of observed tornadoes, i.e., one has to obtain an area-intensity relstionship for tornadoes. Table 3.4-8 (reproduced from Reinhold and Ellingwood, 1983) shows a matrix of area-intensity relationship for all tornadoes..This area-intensity relationship is based on the area and intensity of 10,240 observed tornadoes (Schaefer et al., 1980). Each row of this table shows the percentages.of each F-scale intensity tornado which were classified.according to area classifications in Table 3.4-6. Since F6 tornadoes have not been observed in the-

past, the last row in Table 3.4-8 represents engineering judgment in: assigning area classifications.

This matrix shows -that the calculated area and wind scales are slightly skewed and that no tornados are expected to have areas in the A6 range. Representing the. average of-area scales in Table 3.4-6 by a vector =(AA) and the matrix in Table 3.4-8 by (AIM), the -vector.of expected values of areas for each F-scale intensity (AI) may be written as (AA) (3.4-10) (AI) = (AIM) e 3-44 =

. - - -. ~. ~. _ - - Thus, mean tornado area - (m12) for each F-scale intensity were j obtained as'{AI)T = (0.30,.0.72, 1.8, 4.3, 8.5, 15.7, 18.9). Another-characteristic of a tornado is that its intensity does not__ stay constant along its path.- As noted previously, an FPP ^ intensity, scale -'is assigned to a tornado based on the most severe observed _ damage. However, a tornado is usually at its highest intensity only for a fraction of the time.that it is active. F i g u r e 3. 4 - 6,. reproduced from Reinhold and Ellingwood, shows a hypothetical F4 tornado with varia. tion of intensity along its path. Table 3'. 4 - 9, reproduced from Reinhold. and Ellingwood, shows a matrix (VWL) for combined variation of tornado intensity - along its path length and across its path width. Each column of matrix (VWL) in Table 3.4-9 shows the percentage of each F-scale damage in the area - (WL) for a tornado which has been assigned an intensity scale based on the-nost severe observed damage. As an

example, F3 tornadoes are expected to inflict F3 damage on only 2.7 percent of the total damage area.

In fact, 61.5 percent of the damage that is indicated by an F3 tornado is . expected to be very light (FO). This matrix was obtained from the analysis of the damage from 149 tornadoes that occurred on April 3.and 4, 1974. For a point structure where AI WL (see Equation (3.4-9)), = the probability of wind speeds exceeding (V*) at the site _may be written as: P((V( A,wt) ) > (V*)] OI) (3.4-11) (VWL) (AI = 1 e where (V*) is taken to be the mid-point of tornado velocity scales as shown in Table 3.4-6, i.e., the left-hand side of Equation (3.4-11), which;is_the probability of exceedence for F-scale _ intensities, is also equivalent to the probability of exceedence of_tha mid-point velocities _for_F-scale intensities from Table 3.4-6.- The matrix (VWL) was described in the above -paragraph and-(AI

• UI) is a' vector where its elements are the expected _ values of tornadc areas times the. occurrence-intensity ~ rates for the same F-scale intensity.

As an example, for F6 tornadoes, the above equation for Region A may be written as P [ F a F 3 = P [V ( A,WL) > 349 mph) =_0.001-x 18.9 x 0.0013 A 6 A I = 2.46 x 104 (3.4-12) l-1-45

~ ..- ~...- _-.---~ As described _prevjously,.there is a second contribution to the probability -of the -tornado wind ~ _ speeds exceeding a cartain value_which arises from the lifeline term-in Equation (3.4-9). As shown in Equation (3.4-9), the lifeline _ term (HL) depends on the tornado length _and it-is. independent of tornado width. In_ fact, the effect of tornado width variations on the probability.of exceedence was ignored by neglecting the term WG~in. Equation (3.4-8). Table 3.4-10, reproduced from Reinhold and Ellingwood, shows a matrix of intensity-length relationship (LIM) where each row of the matrix is the fraction of tornadoes with a given F-scale intensity which were observed to have length scales according to Table 3.4-6. This matrix was - based on an analysis of 7953 tornadoes between 1971-1979 (Reinhold and Ellingwood, 1982). The_ expected value of tornado length for each F-scale intensity tornado (LI) may then be computed from (LL} (3.4-13) (LI) = (LIM) wher_e (LL) is the vector of mid-point length scales from Table 3.4-6. Thuc a length-intensity vector. ( LI)T (1.53, 3.01, = 4.76, 9.15,.18.8, 26.9, 30.1) was-obtained (miles). Since a tornado's intensity varies along its length, one needs to. establish _ a ' relationship between the total length for a given F-scale tornado and the percentages of total length which were observed to have different F-scale intenalties. Such a relationship -is shown in terms of the matrix of - variation of intensity along length (VL) in Table 3.4-11, - reproduced from Reinhold and Ellingwood, whero.cach column of . the matrix lists the percentages of total tornado length with - different F-scale intensities. This matrix was based on 14 9 tornadoes which occurred on April 3 and 4,

1974,

~ Thus, the contribut' ion of the lifeline term to the probability z of exceedence of a wind speed-(V*) at-the site may be written

as

{LI. OI) e H (3.4-14) P((V(A,yg) ) >-(V*)) = (VL) I = Again,-(V*) is taken to be tho'mid-point of-velocity scales for each F-scale tornado as shown in Table 3.4-6. The vector 3-46

OI) is obtained by multiplying each term of the length-(LI e intensity vector;(LI) by the occurrence-intensity vector (01). As an - example, the contribution of a structure with a characteristic longth of H 1 ft. to the probability of = exceedence of F6 tornadoes for Region A is P [F a F6) = P (V(A,Wu) > 349 EPh] = 0.160 x 30.1 x 0.0013 A A I 1 ft

• 5280 ft/mfle

= 1.19 x 10-6 (3.4-15) l Combining the point structure strike probability and the lifoline strike probability and using the Poisson arrivals for tornadoes (Equation (3.4-7)), thes annual probability of exceedence for each F-scale velocity may be written as (P[F 2 F ]) = (P[V > V *]) = v[{c1} + (c2}H] (3.4-16) i i where vectors (c1} and (c2) are obtained from Equations (3'.4-11) and (3.4-14). For the LaSalle site located in Region A, vectors {ct} and (c2) are obtained as {ct)T = (1.28, 4.76(E-1), 1.52(E-1), 3.08(E-2), 4.39(E-3), 3.66(E-4), 2.46(E-5)) (3.4-17) (c2}T = (2.15(E-4), 2.79(E-4) 2.69(E-4),

1. 31(E-4),

4. 84 (E-5), 0.31(E-5),

1.19(E-5)) -(3.4-18) Figure - 3. 4-7 shows the tornado hazard curves for the LaSalle site which were calculated for lifeline lengths of 100, 300 and 500 feet. The Category I structures-at Lasalle are built adjacent to - each other. For Unit 2, the dimensions of a rectangle which would. enclose all Category I structures are approximately 180' x - 215 '.. Assuming that a tornado approaches the plant at 45' angle to one of the sides, the maximum lifeline length of the structure is calculated to be H = 280'. From Figure 3.4-7,-the annual probability of exceedence of 300 3-47

Eph' winds for a characteristic. length of 280' is approximately 1x 10-6.. The ' Category I structures _ are designed for rotational - tornado wind peeds of 300 mph-and translational s tornado velocity of 60 mph, i.e., a total wind speed of-360 mph was used in design and therefore 300 mph may be assumed to be a lower limit on the wind load capacity of the Category 1 structures. Thus, it is concluded that structural failures due to tornado' wind pressures are not significant contributors to the overall plant risk. 3.'4.3.2.2 Tornv a-Generated Missiles Missiles generated by tornadoes may lead to a plant damage state if they impact the Category I structural walls or roof slabs with critical velocities. The tornado missile hazard is a low _ probability event because a sequence of events must occur in order for the missile to cause any damage. This sequence--includes the missile injection and transport, missile impact and barrier damage of _ Category I structures, and an accident sequence. A description of tornado missile bounding analysis for LaSalle Category I structures follows. The tornado missiles used in the present study are representative ' of construction site debris and they are the set of' missiles-which have been listed in the Standard Review Plan. . Table 3.4-12 (from the Standard Review Plan, USNRC, 1975)- gives a = description of these missiles and-their respective maximum horizontal' velocities for tornado Zone I hs defined in Figure 3.4-3. _ Missiles A, D and F in Table 3. 4--12 may be classified as deformable missiles -whereas missiles B and C are nondeformable missiles. Except for missile C, these missiles have vertical velocities'of 70 percent of postulated l horizontal velocities. -Missile C which is used to test barrier openings is_ assumed to have the same velocity in all

directions.

. Missiles A, B, C'and E _ are considered at all elevations and missiles D and F are considered at elevations up toL30-' feet.above grade. Basedion test data, several formulas-have_been suggested for nondeformable missile ~ impact on reinforced concrete walls. In all of.the studies on micsile impact which have been performed l to date, it has been concluded that the amount of l reinforcement 'is -- not an important factor in calculating the scabbing thickness or perforation thickness of a reinforced concrete. wall. The most widely used formulas for determination of minimum wall thicknesses required to prevent scabbing are Chang's formula and the modified-National Defense _Research. Committee (NDRC) formula (Chang, 1981). According to chang, the' scabbing thickness (ts) of a wall or slab may be calculated.by (Equation (3.4-3)). 3-48

0.4 0.67-y y t, - 2.47 d.2 (f ) O.4 0 e where w = weight of missile (lbs), v = velocity of missile (ft/sec), 4A d = missile effective diameter (inches) = e fe = ultimate strength of concrete (psi), Ac = contact area of missiles (in2), The modified NDRC formula gives the penetration depth x of a solid missile as -x-4Md 1000d f T 1. 8~ fr 2.0 QMO 1000df,+ d -x-M where 0 K= kc N is_an empirical constant equal to 0.72 for flat-nosed

missiles, 0.84 for blunt-nosed missiles,

1. 0 ' for average bullet nosed missiles, and 1.14 for very sharp missiles.

Scabbing thickness is - then related to penetration depth as follows: t 2 [-7.91.({}-5.06({} for { s 0.65 t [-2.12+1.36 for0.6'5<{s11.75 (3.4-20) 3-49

~ _... -. _ ~. .. _ ~ For 'the NDRC formula, best results are obtained for pipe missiles when d is the actual outside diameter of the pipe in calculating __ penetration depth and equal to an effective diameter in calculating scabbing thickness. Using the above formulas-for the missiles in Table 3.4-12, wall and slab thicknesses which are required to prevent horizontal and vertical missiles from scabbing were calculated (Table _3. 4-13). The NRC recommended minimum thicknesses of 16" for - roof _ and 20" for walls compare favorably with the results obtained by the Chang's formula. These calculated thicknesses are higher than some of the wall and roof slab -thicknesses of the LaSalle Category I structures. For example, the diesel generator structures at LaSalle have 12" walls and 12" roof slabs. Although the auxiliary building roof and the_ reactor building roof have 6" slabs on top of a metal deck, they are not considered in this study because the floors which are immediately below the roof slabs in these structures do-not contain any ESF equipment.

Also, as mentioned in the FSAR, the spent fuel pool which is located at Elevation 843'6" on the operating floor of the reactor building 'has been analyzed for postulated tornado missiles.

All other Category I buildings at LaSalle are protected by walls or slabs which are at least 18" thick. Therefore, it is concluded that the _ only-critical structure at LaSalle ti:at needs to be analyzed further for tornado missile impact 12 the diesel generator building which has 12" thick walls and 12" thick roof slab. In performing a bounding analysis for the diesel generator I building tornado missile impact, the following factors should be taken into consideration: I 1. Given that there is a tornado.at the site, the probability _of a missile injection and transport resulting in the missile impact of the diesel generator building ir very_ low. l 2. 'Even if a tornado missile impacts the diesel generator L

building, it may not - have enough energy to cause scabbing _of the_ walls or the roof slab.

Twisdale and Dunn (1981)'.have performed a simulation study _for-a typical nuclear power plant to obtain tornado missile impact probabilities and probability distributions of missile velocities. They used-a total of 65,550 potential missiles which could be injected from different zones near the plant. Since most of these^ missiles represent objects which would-be available during construction _of a plant, the total number of missiles.is expected to be. conservative for-the LaSalle sta-tion wnere.both units are operating. In-fact, the site visit 3-50 ~

LbyfSMA personnel verified.that thez potential missile popula-tion at-LaSalle is about-one-fifth to one-tenth of the number used by Twisdale and-Dann (1981). In Twisdale - and Dunn (1981), - a flat terrain similar to the LaSalle site was used. - Also, a comparison - of the plant layout and geometry of-the buildings -- between LaSalle and the example plant in Twisdale and<Dunn (1981) showed that the diesel generator building at -LaSalle.is protected on two_ sides whereas the diesel generator-building-for the example plant is protected on one side only. Therefore,- using - the results of the simulation study by Twisdale and Dunn. (1981) for-the diesel generator building at LaSalle.isiexpected to be conservative. Results of the simulation-study by Twisdale and Dunn (1981) . indicates 1that given a tornado at the site, the probability-of a -tornado missile impacting the diesel generator building.is 'approximately 10-2 Since the total number of potential L misciles -for LaSalle site was estimated to be approximately 12,000, which is lower than 65,500,- the conditional probability-of missile impact for LaSalle was estimated to be 2 x 10- 3... Also, distributions of the missile-velocities show - that t given a nondeformable missile (6" pipe or 12" pipe) impact with the diesel generator building, the probabilitv of scabbing is high, e.g., roughly 0.6 for the 6" pips and v.98 for the 12" pipe. The probability of scabbing due to a nondeformable tornado missile. impact may be written as P[S)

• P[TS] = P[MIlTS] = P[SlMI]

(3.4-21) where S~= scabbing TS =~. tornado strike MI = missile impact Assuming that. tornadoes with intensities greater than F1 can transport missiles and cause-damage to the diesel generator building, the probability.of a tornado strike was estimated to be _ l'. s x : 10 - 4/ year (see Figure 3.4-7). Since the nondeformable - 6" and 12" ' pipe missiles represent only 25 percent of the total ' potential missile population, the last I

term in Equation (3.4-21),

P[S MI), is' estimated to be 0.25. 2 x 10 the probability. of Therefore, using P(MIlTS) = scabbing-was conservatively estimated to be 5.0 x 10-8 This probability is comparable to a probability of scabbing of 2.8 x 10-7 year - for. Region A reported by Twisdale and Dunn (1981). / 3-51

The deformable tornado missiles, -namely wood plank and automobile impact, were included in the design of Category I structures._ :The-velocity used for wood plank in the design was 225. mph which is higher than the suggested velocity by the Standard Review Plan (Table 3.4-12). On the other hand, the automobile velocity used in the design was 50 mph which is lower than the value listed in Table 3.4-12. Results of the simulation study by Twisdale and Dunn (1981) show that given a tornado,_the probability of an automobile impacting any of the structures in the plant with a velocity greater than 57 mph is less than 10-3 Due to the inherent conservatisms in design, it may be concluded that the capacity of diesel generator walls for an automobile impact is at least 57_ mph. Therefore, the automobile -impact's contribution to the plant risk would be less than 10 8/ year. The only deformable tornado missile which was not specifically considered in the plant design is the utility pole. I!owever, based on the full-scale tornado missile impact tests conducted by EPRI (Stephenson, 1976), utility poles are not expected to cause any damage to t_u 12"-thick reinforced concrete walls. Based on the conservative bounding analysis performed in this study, it is concluded that nondeformable tornado missiles as wel] as deformable missiles are not significant contributors to--the plant risk. It is noted that-the HVAC air intakes and exhausts are protected from tornado missiles using adequate concrete barriers. The barriers are placed such that the tornado missiles cannot reach the fan-openings. Also, the auxiliary building roof ventilation stack which is the tallest structure in the plant is' designed to withstand the effects of the design basis tornado and therefore will not collapse on the auxiliary building. 3.4.3.3 Non-seismic Category I Structures 3.4.3.3.1 -Design' Capacity The non-seismic Category I structures at LaSalle are designed to. withstand the effects of 80 miles per hour straight winds and the approaching tornado. The siding enclosures for the j following structures are designed to blow-in and blow-out under predetermined tornado wind pressure: Reactor buildinge i-Turbine building (above Elevation 767'0") The metal roof decking for the following structures is designed to blow off under tornado conditions: Reactor buildings 3-52

i Turbine building Auxiliary building A -review of the metal siding specifications used for LaSalle indicated that the siding is designed to blow-in (o* out) at tornadic wind pressures between 52 psf and 84 ps f, i.e., the siding will start blowing in at 52 psf, and all the siding will have blown in at 84 psf leaving the bare structural frame. Therefore, the structural frame is designed to withstand the 84 psf wind pressure acting on the building with the entire siding intact. The structure is also analyzed for the design basi.s tornado of 300 mph maximum tangential velocity acting on the bare frame to ensure that it will not collapse on adjacent seismic Category I structures. The lowest wind speed at which the siding-will start blowing in (or out) is estimated as (52/C y 0.002558)1/2 136 mph = for Cp= 1.1 from Figure 4 ir. ANSI $58.1 (1982). 3.4.3.3.2 Exceedence Probability Structural failure of the siding or roof decking could occur when the wind speed exceeds 136 mph. This could happen in either a tornado or a strong wind storm. Therefore, the exceedence probability is estimated considering both tornadees and straight winds. Tornado Loads The probability of exceedence of the lowest capacity of siding by tornadic winds is obtained from Figure 3.4-7 as 1 x 10 " per year. Straight Extreme Winds. The probability of extreme wind speeds at LaSalle exceeding 136 mph is estimated by reviewing the wind speed data for the pertinent weather stations. Figure 3.4-8 from the report by Changery (1982) shows the weather etations in the vicinity of.the site. The LaSalle t reathe r station had only 9 years of wind speed data. Shorefore, data from the neighboring stations (Chicago, Moline a.id Peoria, Illinois) were utilized in estimating the wind speed probabilities at LaSalle. It was found that Moline, Illinois, station had the highest annual wind speed exceedence probabilities among these stations. Thus, for the purpose of bounding analysis, Moline,

Illinois, data was used.

The probability of exceeding 136 nph wind speed was calculated by 3-53

~ _ _,__.. _ _.. _._ _ _ _.._ _ m fitting -'an extreme-value Type I distribution' to the annual maximum Wind. speed data as recommended in ANSI A58.1 (1982). This probability of exceedence.value was-obtained as 3.8 x 10 6 per year. .The probability-of wind speeds exceeding 136 mph as a result-of tornado strikes or extreme wind storms was estimated as 1 x l 10 4 + _ 3. 8 ' x 10 1 x 10-4 per year. It is assumed that the failure _of non-seismic category I structures will not lead to core damage. However, -i f any components housed in these structures are included in the fault trees, the failure rates used in calculating _ their unavailabilities should be assumed not-less than 1 x.10 - 4 per year (lower rates might be used if -the components are protected somehow from the structural failure)._ Similarly, the exposed tanks (e.g., condensate storage-tank) which are typically designed to withstand the effects of earthquake and straight. wind loads using the Uniform Building-code (1973) requirements should be assumed to have failure rates not less than 3 x 10-4 per year. 3.4.3.'4 Uncertainty Analysis for Winds and Tornadoes A probabilistic bounding analysis for wind and tornado hazard i ar.d' tornado - missile. hazard at the LaSalle site was performed in - Sections 3.4.3.1 through _3. 4. 3. 3. Based on the results presented in these sections, it was concluded that-the -probability of potential core damage due to winds and ~ tornadoes is negligible. The bounding analysis was based on conservative assumptions regarding tornado hazard and structural ~ fragility models; however, it did not address the question _ of uncertainties in models and modeling parameters. -In _ this section, estimates of these uncertainties are

presented.

Also, these uncertainties are propagated in the -bounding analysis to obtain an estimate of the uncertainty in theLprobability of severe core damage. Since wind loads were

shown to - be of lesser importance in comparison with tornado loads for. LaSalle structures, attention will be focused on uncertainty in. tornado loads and tornado. generated missiles.

Uncertainty'in the calculated probability of core damage due to tornado loads arises from the following: 1. Uncertainty in_ tornado hazard calculations 2. Uncertainty-- in wind pressure calculations given a . tornado wind: speed-3. Uncertainty in -- structural response and fragility calculations. 3-54 u

The hazard model utilized in this study has been previously discussed in detail (Section 3.4.3.1). The model is based on historical data as well as subjective classification of tornadocs based on their maximum observed damage. Also, the data base for the model is not uniform, e.g., the area-intensity relationship is based on a sample of 10,240 observed tornadoes whereas variation of tornado intensity with path is based on an analysis of 149 tornadoes which occurred in a tornado outbreak during a two-day period. The tornado model developed by Reinhold and Ellingtood (1982) has corrections for tornado classification errors and random encounter errors.

Recently, Mcdonald (1983) completed a tornado hazard probability assessment which accounts for uncertainty in area-intensity and occurrence-intensity relationships.

The model used by Mcdonald (1983) is very similar to the model used in the present study. Hazard uncertainty reported in Mcdonald (1983) is due to dispersion in data, i.e., regression models were fitted to historical tornado data to represent arer intensity and wind speed occurrence relationships. Confidence bounds on the best estimate tornado hazard carve were established from uncertainties in regression models. Figure 3.4-9 shows the median tornado hazard curve for the LaSalle sits with 95 percent confidence bounds as estimated based on the study by Mcdonald (1983). Tne uncertainty in tornado pressure cocfficient was estimated using the reported uncertainty for straight wind pressure coefficient (Ellingwood, 1978). The pressure coefficient relates the induced pressure on wa: 3anels to maximum wind velocity. Induced pressure on a wall panel is a function of structural shape as uell as the location on a wall panel. Therefore, there is some uncertainty associated with the pressure coefficinnt. Ellingwood (1978) reports a coefficient of variation equal to 0.15 for uncertainty in the straight wind pressure coefficient ar.3 a coef ficient of variation equal to 0.05 representing the uncertainty in wind modcling. Due to lack of data for tornado wind pressures, the uncertainty in straight wind pressure coefficient was used in the present study. Since the physical phenomenon of induced presaure due to straight winds is the same for tornadoes, this assumption is judged to be realistic. As discussed in Section 3.4.3.1, the seismic Category I structures at LaSalle have an effectiva design capacity of 360 mph against tornado wind loads. Ths. are two sources of conservatism in design: (1) there is an inherent conservatism in ae nominal steel yield stresses and the nomin;l concrete strengths specified by the designer, and (2) the code allowable stresses are lower than ultimate or yield stresses. The conservatism factors in nominal yield and design code 3-55

.~. =-. allowable-stresses were estimated to be 1.2 and 1.1, respectively, for screening--purposes due to assumed variations -in _ material -behavior. . Since-the induced wind pressure un-a wall panel -is proportional to the square of applied wind velocity,=the median. wind capacity of LaSalle category I structures (V) is calculated as: 'l =[1.1 x 1.2 (360)2jl/2 = 414 mph (3.4-22) Un' certainty in the median wind capacity of the LaSalle buildings is due to uncertainties in the material behavior used in the structural model. The coefficient of variation in material yield stress was estimated to be 0.15 (Galambos and Ravindra, 1978; Mirza and MacGregor, 1979; Mirza, Hatzinikolas and'MacGregor, 1979). Also, a coefficient of variation equal to 0.15 was used for modeling uncertainty, Next, it is assumed that the variability in wind pressure (cp) can be modeled as the product'of random vat-lables representing variabilities in pressure coefficient (cpe), wind modeling (cwm) e material yield- ((my) and structural modeling (csm) - op = rpe twm (my (sm (3.4-23) Assuming that the c's are lognormally distributed, the logarithmic standard deviation for wind pressure (p ) was p calculated to be 0.26. Since the calculated wind pressure is -proportional to _ the square of wind velocity, logarithmic standard deviation of wind velocity is 1/2(0.26) 0.13. = -Thus,- the wind fragilities of reinforced concrete structures at LaSalle'are defined in terms of their median capacity (v = 414 mph)_ and a composite logarithmic standard deviation (pv = 0.13). Figure 3.4-10 shows-the tornado fragility curves for LaSalle, category I-structures. In order to develop the f ragi)ity curves, it was assumed - that the composite variability-sv can be split into two terms #v,r = 0.08 and Bv,u = 0.11 representing the randomness. and -uncertainty in the tornado wind capacity calculations. Figure :3.4-11 shows the distribution of annual frequency of severe core -damage calculated from the family of tornado ' hazard.and structural wind fragilities. From this distribution, the median frequency of = severe core damage was 10*8 year _ whereas the 95 percent confidence / found: to'be 3 x bound was calculated to be -3 x 10-7 year. Since the bounding / analysis has been-conservative and the 95 percent confidence ' bound probability is extremely low, it is concluded that 3-56

tornadoes do not contribute significantly to the probability of core damage. 3.4.3.5 Conclusions The bounding analysis described in this section has shown that the high confidence probability of failure under wind and tornado loading for Seismic Category I structures housing critical equipment is on the order c,f 10-7 per year. Even if these structural failures are conservatively assumed to lead to core damage, their contribution to the plant risk is negligible smail when compared to other events. The ncn-seismic Category I structures and exposed tanks have frequencies of failure under wind and tornado loading on the order of 10-4 per year. Their failures may not lead to core damage; if components in the structures should ar, car in the fault trees, the failure rates used to calculate their unavailabilities should not be less than 10-4 per year unless the components have additional protection. 3.4.4 Transportation Accidents This section describes the bounding analysis for transportation accidents near the LaSalle site which could contribute to the plant core damage frequency. A transportation accident near the plant may lead to core damage in one of the following ways: (1) a chemical explosion due to a transportation accident may cause damage to Category I structurec and safety-related equipment, and (2) toxic chemicals which are released in a transportation accident may drift into the control room and cause incapacitation of the operators. A bounding analysis was performed taking into consideration the frequency of occurrence of transportation accidents as well as fragility of tha plant structures against accident effects. The bounding analysis for chemical explosions is described in Section 3.4.4.1 and the analysis for toxic chemical release is described in Section 3.4.4.2. There are three modes of transportation near the site, i.e., highway, railroad, and river. Major highways near the site (Interstate 30 and U.S. High'ay 6) are farther than 5 miles from the plant and therefore will not be considered in this study. LaSalle County Road 6 is the only paved road n< r the site and passes approxiuately 2000 feet south of the plant structures. The Chicago Rock Island and Pacific railroad is farther than 3.5 miles north of the plant structures. The Illinois River is approximately 3.5 miles north of the plant at its closest point. The transportation routes near LaSalle County Station are shown in Figure 3.4-12 (LaSalle FSAR). 3-57 l 4

3.4.4.1 Chemical Explosions A chemical explosion near the plant structures may cause overpressure, dynamic pressures, blast-induced ground motion, or blast generated missiles. However from previous research in this topic, it has been deternined that overpressures would be the controlling corwideration for explosions resulting from transportation accidents (Regulatory Guide 1.91, USNRC). An accident overpressure at the site enn also occur because of vapor cloud explosions drifting towards the structures, This type of explosion. involves complex phenomena which depimd on the material involved, combustion process, and tcpographical and meteorological conditions. According to a study by Eichler and Hapadensky (1978), 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 which are used to represent tranoportation accidents. According to the Regulatory Guide 1.91 (USNRC), chemical expI Q 1on' which would result in free-field overpressures of less than 1 psi at the site do not need to be considered in the plant design. Based on experimental data on hemispherical _ Jarges _ _ of TNT, a 1 psi pressure would be safe distance R (feet) which is defined ast j -translated intt a R a KW /3 (3.4-24) l 45 and W is an equivalent weight of TNT charges. where K = The maximum probable equivalent TNT charge is 50,000 lbs for a highway truck, 132,30 lba for a single railroad box car, and 1 x 107.lbs for a river barge. A recent study be Eichler, Napadensky and Mavec (1978) shows.that accidents in an empty barge due to vaporization of liquid left in the tank would lead to a maximum TNT equivalent explosive load of 1000 lbs. Since this type of accident does not produce a more severe condition, it will not be considered further in thin analysis. 41gure 3.4-13, which is reproduced from Regulatory luide 1.91 (USNRC), shows the safe distances for a highway truck, - a railroad _ box car, and a river barge. Based on this analysis, it may_be concluded that explosions outside of LaSalle County Station - in any of the transportation routes will not pose an i overpressure hazard to the plant structures, In the study by Eichler, Napadensky, and Havoc (1978), the hazard from vapor ; cloud _ drifts _ which could be generated in [~ 3-58' -, - ~,,. ~. ~...... -, _ _.. -. -. _. ~ -. _. -. _ _. - - - - - - - -..

k barge accidents were examined. According to this study, although a vapor cloud may theoretically drift towards the site and produce higher incident overpressuren at the sito, the following reasons minimize the throat due to drifting vapor clouds. 1. Probability of. vapor cloud explosion rapidly decreases .due to the -docrease in concentration as it travols -away-from the accident site. i 2. Rango of unfavorable wind directions (i.e., wind directions that can impact the plant) rapidly l decreases as spill to site distance increases. Dased on this study, it was concluded that the equivalent TNT l oxplosive weights which aro specified by the 11RC are very conservativo. Vapor cloud explosions woro also considered in the Limerick-Severe Accident Risk Assessment. In the Limerick study, vapor cloud drifts from a railroad accident which is approximately 600 feet away ' from -the nearest Category I structure wore considered. The: equivalent TNT in the Limorick study was calculated according to: S Qp g 1 F3 au t 500 Kcal/lb of TNT (3.4-25) v c I wherot F = fraction of spill quantity involved in vapor j

cloud, 0 0#

1 . gm-molo of' combustible chemicals spilled ~, A Si = spill fraction, Q = quantity of shipment, pT = density of liquid, A = molecular weight, AHe = heat of combustion (Kcal/gm-mole), E - yield.of. explosion. 3-59

Also, based on historical data, the cumulative density function of distance from the accident uite to ignition was obtained.

This is shown in Figure 3.4-14, reproduced from Eicher 1978, where the curve may be reprecented by a line as: / log 10^ ~ 1 14 erf (3.4 26) l' y ,j$ j A - 0.175 r where r is the distance from the spill site in meters. As mentjoned before, the railroad in the LaSalle area is approximately 4 miles from the site. Therefore, if the plume travels a distance of 1 mile, the probability of not having an ignition before that distance is reached reduces to approximately 10*3 If the same CDP is assumed for a barge

accident, it is observed that vapor clouds do not pose a hazard to the plant structures, i.e.,

assuming that the vapor cloud can travel a maximum distance of 1 mile, an explosion will result in a small incident overpressure on the buildings. Althous: the liRC Fugulatory Guide is conservative in defining the eq alent T111 explosive loads, it is unconservative wich respect to structural capacities because of the following reason. The free-field pressure wave which results from a TllT explosion is reproduced from Kennedy et al. (1983) in Figure 3.4-15. This pressure consists of an instantaneous rise and a decay to zero followed by a slight negative press are. The values of peak incident overpressure (Pso), positive phase impulse (I), and positive duration (t ) which were based on d experiments are shown in Figure 3.4-16, also reproduced from Kennedy. 11ote from Figure 3.4-15 that the overpressure acting on the wall phnels of a structure also includes a reflected pressure. Therefore, the overpressure on the wall panels is approximately twice the incident overpressure. In addition, the dynamic effect of peak overprossure for a wall panel may be significant. Figure 3.4-17 shows dynamic load factors for a single-degree-of-freedom system as a function of the ratio of pulso duration (t ) to period of structure (T) for a d t :'iangular pulse and a rectangular pulse (reproduced from Biggs, 1964 with permission). It can be observed that the dynamic load f actor for a pulse can reach a maximum value of t /T ratios. As a result of pressure 2.0 for higher d reflection and dynamic effects, a free-field overpressure of 1 psi at the site could result in an offective static overpressure of up to 4 psi on the wall panels. Therefore, a 3-60

nore detailed study of overprousure due to transportation explosions was doomed necessary. An examination of the transportation accidents in the vicinity of.the LaSalle site showed that the controlling accident is a truck explosion on County Road 6 south of the plant. Assuming Regulatory Guide maximum explosivo load of 50,000 lbs, a peak free-field incident overprossuro Pso of 0.66 psi was calculated from Figure 3.4-16.- Therefore, maximum static overprossure on the wall panels could be as high as 2.64 poi. 3 Since the LaSalle category I structures have been designed for i

Zone I tornado offects, their minimum static lateral design load capacity is at least 3.0 pai.

Based on this conservative comparison, it may be concluded that the category I structures have a higher capacity than the maximum postulated overpressure due to an explosion. The above analysis for calculating overpressure capacity of l the wall panels neglected the ability of structural walls to l absorb energy under inelastic behavior. In fact, Kennedy I .ot al. (1983).suggest that a conservativo ductility value i l equal _to 3.0 should be used as the limit of inolastic behavior i for structural _ wall panels. Ductility is defined as the ratio of peak inelastic displacement to the yield displacement for an clastic-plastic structure. The maximum ductility which was assumod - by Kennedy-et al. is conservative because of the following reason. When a reinforced concreto panel is subjected to blast loads, it develops extensivo cracking which means that-the tusior: in cracked sections is resisted by the stool reinforcosent. In fact, ultimate capacity of a reinforced concret.o ' panel may _ be calculated using the yield line theory (Porguson,- 1973, - Park 'and Paulay, 1975). According to the yield line theory, ultimato capacity of a-reinforced _ concreto panel which is subjected to a uniform pressure is dependent on its geometry and ultimato moment capacity _ of the cracked sections. Since ultimato moment capacity _ of a cracked section is. dominated by t'.1e steel . ultimate strength,. well designed -reinforced concrete W '1els-are expected - to exhibit fairly high ductilities under blast loads. Using the results from Kennedy, _ ot al. (1983), freo-field incident _ overpressure capacity of wall panels in LSCS structures was calculated to be a minimum of 1.95-psi. There are two. dif ferences between the calculations for wall panel capacities - in Kennedy, et al.- (1983) and the present study. 3-61

The first difference is that blast capacities in Kennedy et al. (1983) were calculated for a barge explosion. This is a conservative assumption because barge explosions correspond to largest pulso durations and therefore result in higher dynamic load factors (see Figure 3.4-17). The second difference is that the wall panel thicknesses used in Kennedy, et al. were 18 and 24 inches. This is an unconservative f actor because the diesel generator walls are 12 inches thick. However, it is shown in Kennedy et al. (1983) that the wall thickness does not have a significant effect on the wall capacity, i.e. a maximum dif ference of 15 percent was observed between capacities of 18" Walls and 24" walls. Considering all other conservative assumptions used in Kennedy et al., (1983), 1.95 psi-may be accepted as a lower bound capacity of structural wall panels in LaSalle. A comparison of minimum wall capacity of 1 95 psi (incident overpressure) with a free-field incident overpressure of 0.66 psi reveals that there is at least a factor of 3 against an overpressure failure of structures due to the worst truck explosion. Thorofore, it is concluded that chemical explosions do not contribute to the plant risk. -3.4.4.2 Toxic Chemicals A toxic chemical spill near the LaSalle site would pose a danger to the plant if toxic chemicals penetrate into the control room through air intakes and cause the operators to be incapacitated. As discussed in Section 3.3.1, this condition can happen if (1) large quantities of toxic chemicals are l released, (2) there are favorable wind conditions which would cause a drift of chemicals towards the control room air intakes at excessive concentrations, and (3) there are no -detection systems and air isolation. systems in the control-room. Among the three transportation modes near the site, a barge

accident in the Illinois River could result in the. largest amount of chemical spill.

As reported previously, the Illinois River.is 3.5 miles away from the plant structures.at' its' closest distance.

Also, the river elevation is-approximately 180 feet below the plant-grado.

Considecing the fact that many of the toxic vapors are denser than air, the atmospheric dispersion of those chemicals towards the plant under favorable wind conditions is unlikely because of the difference in _ plant and. river elevations._

Also, for more turbulent wind conditions, it is highly unlikely that a toxic vapor would reach the control room air intaken at excessive concentrations.

An examination of ' Table

3. 3 shows. that among the _ hazardous -chemicals transported-on barge _ to the nearby _ industrial facilities, chlorine, anhydrous ammonia, and 3-62

.u_-., ..,_,_..,_,_-,,..-,__._,__,,_,_....,4 ,--,.,_-,,m...-,..,, .,_._-_..,_.._m.._

-_m h i butadiene are shipped at large quantities. Since the control room IIVAC at LaSalle is equipped with detectors for chlorine and anhydrous ammonia, these two chemicals are excluded from i further consideration. Acco.d.ng to the Regulatory Guide 1.78 l (USNRC), -butadione has a low toxicity limit. Therefore, even if the--maximum quantity of butadiene required at the Borg-Warner chemical facility was shipped on one barge it would still meet the requirements of Regulatory Guide 1.78 (USNRC) as: to the proximity of toxic chemicals to a nuclear power plant. From the foregoing discussion, it was concludes _that 1 chemical spills resulting from barge accidents do not contribute -significantly to the plant risk. Using the same i

logic, railroad accidents are also excluded from external c

events -analysis because the Chicago Rock Island and Pacific i Railroad is further from the plant than-the Illinois River and a railroad accident - would result in a much lower quantity cf-spill than a barge accident. As shown in Figure 3.4-12, majnr U.S. highways in the vicinity of LaSalle - site are en n ' miles away from the i u plant structures.

Also, st Be h@ W N is more than 3 miles from the plant structut y Ya r cast paved road to the-plant is LaSalle County Roaa a.' u n is 4.000 feet south of the plant structures.- Thorofore, tae only possible hazard to the site would come from the county ttoad 6.

Since this road is not a major highway, there is nr reason to believe that it is:used for - transportation of chemicals other than those i shipped to'the plant or to the nearby industrial facilities. On this basis, a chemical spill near the site. would be either I

detected, i.e.,

chlorine or anhydrous ammonia spill, or it 4 Lwould be of no consequence to the plant operators, i.e., butadiene-- spill. _Thus, it was - concluded that transportation ' accidents leading to toxic chemical spills are nd significant contributors to the plant risk. 3.4.5-Turbine Missiles -This section describes the bounding analysis of the LaSalle-plant-for the risks from turbine missiles. A review of the historical background, FSAR analysis and recent issues in regards to turbine missiles is given. 3.4.5.1 !!istorical Background Failures of large steam turbines in both nuclear and fossil-fueled power-plants, althouch rare, have occurred occasionally 'in the past. These failurus have occurred because of one or more of the following broad classes of reasons: (1) metal-lurgical and/or design inadequacies, (2) environmental 3-63 -_____,.a_ - z _..,.-..~--._a.

offects, (3) out-of-phase or generator field failures, and (4) failures of overspoed protection systems.

The failures have resulted in loss of blades, ditik cracking, rotor and disk

rupture, and even missiles.

Turbine missiles aro highly energotic and have. the potential to damage safety-related structures housing critical components. Therefore, protection of nuclear power plants from turbine missiles is an important safety consideration. Also, rupture of the turbino casing in a boiling water reactor plant (e.g., LaSalle) may lead to release of primary coolant steam and radioactivity to the environment.

Ilonce, the plant owners aim to minimize the frequency of turbine failures resulting in casing rupturo even if there are no significant turbino missile strikes on safety-related components.

In a total of 2,500 years of turbine operation in nuclear power plants in the free world, only four failures have occurred: Calder liall (1958), liinkley Point (1969), Shippingport (1974), and Yankee Rowe (1980). External missiles were produced in the liinkley Point-and Caldor Hall failures. Although the causativo mechanisms of these failures have boon identified and are generally corrected in the modern nuclear turbines, there - is no assurance that other types of turbino failures will not occur 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 turbino failures. Nuclear plant turbines rotate at 1800 rpm with the low-pressure -(LP) and high-pressure (IIP) sections on a contiguous shaft. The LP sections have blado hubs (called "whcols" or " disks") shrunk onto the rotor. Depending on the manufacturer and rated capacity of the turbino, there could be 10-to 14 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 - turbine._ With the turbine unit running at less-than 120 percent of the rated speed, the disks aro 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-ir.termediate speeds (i.e., 120- to 180 percent), rupture of disks may be caused by a combination of flaws and weaker material in the disks. 3-64 ., - _.... ~.. _. - -. - - - _.. - -... -. - -. ~. -. -.-. ~. -. -..

Turbine missiles are spanning, irregular fragments with weights in the range of 100 to 8,000 pounds, and velocities in the range of 30 ft/sec to 800 ft/sec. It is conventional to - discuss two types of turbine missile trajectories: low trajectory missiles (LTH) and high trajectory missiles (llTM). The low trajectory missiles are those which are ejected from the turbine casing at a low angle toward a barrier protecting an essential system. liigh trajectory missiles are ejected vertically (almost) upward.through the turbine casing and may strike critical targets by falling on them. The customary ballistic distinction between LTM and HTM is the initial elevation angle (4) of the missile (LTM is for & < 45' and itTM is for p a 4 5'). Turbine manufacturers have specified that the maximum 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 (USNRC) and recommends that the essential systems be located outside this LTM 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-the full scale and model tests, it is estimated that concrete barriers should be at least 4 feet 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 turbine missile trajectories. In the design of a nuclear . power plant, the designers have many alternative approaches for treating the potential offects of turbine failures (Sliter, Chu, and Ravindra, 1983). These approaches can be grouped as:= (1) prevention of turbine failure, (2) prevention of missiles, (3) prevention of strike on critical components, and (4) performance of probabilistic analysis to demonstrate that the probability of turbine miselle damage is acceptably 1 aw. In the LaSalle FSAR, it is shown that the probability of turbino missile damage is acceptably low. The following subsections review the FSAR Analysis from a-PRA standpoint and utilize and update the results for the bounding analysis. 3.4.5.2 Probabilistic Methodology The probability of serious damage from turbine missiles to a specific system in the plant is calculated as (Bush, 1973): P4"P1 P2 P3 (3.4-27) 3-65 l- ---~ s- .c, --a e v ,w. ~w .e -n~

whero: P1 = probability of turbino failuro leading to missile generation, P2 = probability of missiles striking a barrier which encloses the safety system given that the missile (s) have been generated, P3 " probability of unacceptable damage to the system given that' one or more missiles strike the barrier. In practico, the evaluation of P4 should include consideration of different speed conditions, distribution of missiles, and all the safety-related components and systems in the plant. 3.4.5.2.1 Probability of Turbine Failuro Pi LaSallo county Station has 38" last stago bucket 1800 rpm turbine generators manufactured by the General Electric company (GE). Typically, turbine failures under three speed conditions are considered. Failures at or near the rated speed of the turbine could occur primarily due to brittle fracture of disk material. - Overspeed failures could occur bccauso of turbine overspooding and subsequent disk rupt ure due to - brittle fracture or ultimate tensile failure of material'. Design overspoed.is defined as follows. The calculated speed attained following-the loss of full. load and the malfunctioning _of the turbine speed governing system along with _a successful. tripping of the turbine overspeed trip mechanism will not exceed overspeed which is 120 to 130 percent of_the rated _ speed. The turbino disks may rupture at this overspeed 'from brittle fracture ~ propagating from an . undetected ' flaw. Destructive overspood is the lowest calculated speed at which :any. LP rotor disk- (or wheel) will burst based on the average - tangential tensile stress being i equal - to ' the - maximum ultimate tensile strength ~ of the disk material, assuming - nol flaws or cracks in the disk. The i destructive overspeed is typically between 180 and.190 percent of the rated speed of-.the turbine. Probability of failure at_an overspeed-(e.g.,-design overspeed and destructive'overspeed) is calculated as the product of the-probability _ P11 of attaining the specified overspeed condition when the turbino generator unit at full load is unexpectedly-separated-from the system und the probability P12 that a turbino disk (s) = ruptures and disk fragments exit.the turbine casing wnon the ;overspeed condition is reached. The probability of attaining an overspeed, P11, is calculated by modeling the - overspeed event as a sequence of simple events 3-66 i

---

~~,.. -....-.-- - -- ---. -+ -~ --~,

l and performing a fault tree analysis. The analysis "tilizes the failure rates for electronic compononts, control valves, stop valves, overspeed trips, etc., and incorporates the effects of in-service inspection (CE, 1973). j General Electric (1973a). has established that the probability of missile generation at the rated speed or at the design overspeed conditions (called "the low speed burst") 'is statistically insignificant and as such no missiles are postulated at these speeds. The probability of disk failure leading to the ejection of a missile at the destructive overspeed (called'the "high speed burst") is calculated by GE as 5 x 10 9 por year. Dush ' (1973) has analyzed nuclear and relevant fossil turbine failure data with the objective of mLking a realistic estimate of the probability of turbine failure leading to mics11e generation.. . Operating history of nuclear turbines is too short to make a reliable estimate of the failure probability based on' only nuclear data. Hence, fossil turbine failures that are judged to be relevant to this analysis are also -included. The most comprehensive study to date on the historical failure data is that performed by Potton et al., (1983) for the Electric Power Research Institute. They estimate the probabilities of turbine missile generation at operating speed and overspeed as 1.20 x 10 4 per year and 0.44 .x 10*4 por year, respectively. These estimates are several ' orders of magnitude higher than those reported by GE (1973a). Recent discovery of stress corrosion incidents in the operating GE turbine-ganerato'rs-(Southwest Research Institute, 1982) -suggest that P values are not as low as what the manufacturers have estimated. Following tho' approach taken in the Seabrook PRA (Pickard, Lowe and Garrick, Inc., -1983), the estimates made by GE (1973a)'Were taken to be the lower bounds (i.e., 5 percentile) on P1 for'the two speed conditions. Similarly, the estimates made by Patton et - al. (1983) were assumed to be the upper bounds (i.e., 95 percentile). The uncertainty in the P1 _ values Lwas modeled as :lognormally distributed with ' the percontiles given above. Table 3.4-14 shows the estimates of annual : probability of turbine missile generation. Since the mean value of Pi is estimated to be.about three orders of magnitude higher than '10-7/ year, turbine missiles cannot be excluded-in the scoping.quantification solely on the basis of

the. probability of missile generation.

3.'4.5.2.2

Probability of' Missile Strike P2 When-the fragments produced in a disk rupture escape the

. turbine casing, their paths have to be determined in order to 3-67 ~

know if they intersect barriers protecting essential systems i of the nuclear power plant. For this purpose, a description of the parameters of these missiles is needed. Major turbine manufacturers have developed their own - generally proprietory - techniques for assessing whether or not disk fragments exit the turbine casing and the parameters of resulting missiles. By making a set of concorvative assumptions regarding the dich breakup mechanism and the impact between the disk fragments and casing structure, they estimate the missile exit conditions. These conditions include weight, closs-scetional areas, chape, size, number of fragments, and exit velocities at different speed conditions. Table 3.4-15, reproduced from the LaSalle FSAR, shows the properties of missiles postulated in a whool burnt of GE 38" last stage bucket 1800 rpm low pressure turbine generators installed at LaSalle County Station. The probability of missile striking a barrier is calculated as follows: low trajectory missilou are considered to travel in straight line paths. Their direction is defined in terms of two angles i.e., the ejection angle, et, Irom the horizontal plane and the deflection angle op from the plane of rotation of the ruptured disk (Figure 3.4-18). The angle el could vary from 0* to 90'. The limits on 02 are specified by the turbine manufacturer (e.g., GE specifies -5* to +5' for interior disks and O' to 25* for end disks). It is customary to assume that the angles 01 and 92 are distributed uniformly within the specified limits. The probability of a low trajectory missile strike on a structural barrier protecting an essential system is calculated as the ratio of the solid angle the barrier subtends at the missile origin to the total solid angle within which the missile can be ejected out of the turbine casing (GE, 1973a). High trajectory missile strikes are analyzed using ballistic theory (Bush, 1973; General Electric 1973a; Semanderes, 1972; Filstein and Ravindra, 1979). The missile is modeled as a point mass experiencing no drag forces. Since the initial velocity of a missile and the ejection and deflection angles are random variables, there is a finite probability that any essential system will be struck by high trajectory missiles. The strike probability density, pA per unit horizontal strike area, located at a radial distance r from the missile origin is expressed as (Filstein and Ravindra, 1979). 3 3 x .x max min p^ (3.4-28) 48 r g sinA(V2'V)I 3-68

where x ~ inin 2 r 2V sina x_ f- - if r s -- (3.4 29) I g cose) frgsin? cy,-l fsina 3 otherwise. \\ C"$"3 [ In the above equations, the missilo velocity is assumod to vary between V1 and V ; the coordinatos of the point along the 2 missilo trajectory are (x,y,z) where x = resine3 and y = recos93 03 is given in terms of 01 and 92 by Cot 03 = Cot 92 Cot 01 (3.4-30) 1 and g is the acceleration due to gravity. Twisdalo et al. (1983) have developed a Monto Carlo simulation methodology for tracking the turbino missiles. A six-degroo-of-freedom (6D) model for predicting the free-flight motion of rigid bodies has boon formulated. It considers drag, lift, and side forces and simulatos missile tumbling by periodic reorientation. A computer code called TURMIS has boon developed to integrate the coupled nonlinear ordinary differential equations of motion. Sensitivity studios performed using this sophisticated 6D model clearly support the use of no-drag ballistic model for low-trajectory turbino missile calculations. For high trajectory missiles, the ballistic model introduces prediction errors for individual trajectories,-but those errors may not be significant (due to componsating offects of reduced speed 'and increased impact probability) when statistically averaged for plant risk analysis. 3.4.5.2.3 Probability of Darrior Damago P3 When a missile impacts a structural barrior (i.e., wall or roof) protecting an essential system, one or more of the following events could. take place: ponotration, front-face spalling, perforation or back-face scabbing of the barrior, i overall responso of'the barrior, and ricochet of the missilo. All of thoso events may be important in ovaluating the damage 3-69 ~ - -

potential of turbine missiles,

llowever, local offects of turbine missiles on concrete and steel barriers normally provided in nuclear power plants are particularly important and include penetration, perforation, and scabbing.

^ Penetration into a reinforced concrete barrier that does not produce back-face scabbing may not constitute a safety-related damage-event unless front-face spalling ja of concern. perforation is the event in which the missile completely penetrates the barrier and continues its flight with a residual velocity less than the initial impact velocity. Scabbing is the - failure mode of most interest because the scabbed concrete fragments may damage the enclosed safety-related component or the piping, electrical

cable, or instrumentation attached to it.

The probability of barrier damage P3 is calculated using the random properties of the missile (i.e., weight, velocity, impact area, obliquity, and noncollinearity) and the empirical impact formulas (Chang, 1981; Berriaud et al., 1978; Twisdale et al., 1983). The dispersion in the-impact tout data about the empirical-formulas is used to develop probability density functions of perforation or scabbing. thickness. For any given missile impacting a structural barrier of known material and thickness, the probability of perforation or scabbing is calculated using these probability density' functions. Evaluation of P2 and P3 can be done numerically if the missile initial conditions are described by a limited set of parameters ano if the plant' is assumed to be damaged when the external barrier of a safety-related structure is breached - (i.e., perforated or acabbed). In general, turbine missiles i are described by'a number of random parameters and-several barriers separate the safety-related components from the missile sources. A Monte Carlo simulation procedure such as the TUT 0iIS ' computer code developed by Twisdale et al. (1983) would be needed to handle the multitude of missile trajectories and_possible impact conditions encountered in a - nuclear power _ plant. The nuclear power plant is modeled for t this analysis as follows. A component may be damaged by a missile physically impacting it, or by the missile damaging the 01cetrical cables or piping that are needed for the L component to function. Since it is impractical to-model all l

piping, electrical cables, and 11VAc ducts for the turbine missile analysis, the components may be modeled as being o

enclosed in fire zones. Each fire ~ zone's boundaries are delineated such that - the component and _ al1 its lifelines electrical cables, etc.) are within this zone. L (piping,. zones - are - independent of each--other, By Thorofore,-the fire this technique,_the safety-related structures of a plant _are divided into a small-number of fire zones (at each elevation l 3-70 lb,-_., - -,. ~ _ -,L,~. ...-.._._...,,._--..,.m._ .....,.....-._,_:__..,.-..=_____._.._.m.-..

-.~--..--- -.--.-. -.- -. In the structures and/or through dif ferent elevations). The -sequences of fire zones which if damaged by missiles in a singlo turbine failure may lead to core damage or serious release (i.e., " cut sets") are obtained by fault tree analysis. 3.4.5.3 FSAR Analysis An analysis was. performed _ during the PSAR preparation to evaluate the probability of damage from turbino missiles to LaSallo County station. The turbine placement and orientation at' LaSalle are such that thera are some safety-related components located within the low trajectory missilo zone. However, the main control room is outside this zone. Both low and high - trajectory missiles were considered in the FSAR analysis. General Electric Company _1973a) provided the P1 ( values-and missilo data as input to this analysis (Table 3.4-15). GE has established that the probability of missilo generation at or near the operating speed (i.e., low speed burst). is statistically insignificant; thorofore, no missilos 'woro postulated for this speed condition. The i probability of disk failure-leading to ojection of missiles at tho_ destructive overspeed -(i. e., high speed burst) was calculated by GE as 5 x 10 9 per year. The FSAR analysis considered the redundancy of equipment and systems and the multiple barriers that must be breached by the missilos before they could affect the equipment and systems. It concluded that the portion of auxiliary building housing the turbine-driven feedwater pump and 480V Switchgear-(betwoon column rows. R and N and between column lines 19 and 21 at Elevations 768'0" and 786'6") is the only area exposed to LTM strikes. Similarly, the reactor building.'was assessed to be the only area exposed to high trajectory missile -strikes and that has equipment that.does not_have redundant-items in other areas of the - plant.- The probability, o'f L missilo damago. to concrete barriers was calculated using modified Petry formula and by treating.the impact velocity and impact area as random-variables. _-The probability of turbino missilo damage conditional on the missilo generation was calculated as 6.06 x 10 4 for two reactor units. Using the estimates of' the probability of turbine missile generation'given in Tablo i 3.4-14,- the probability of - turbine missile damage to the plant is calculated as: P4 (5 percentile) = 3. 4 2 x 12/ year - P4 (mean) = 9.50 -x 10 8 year / P4-(95 porcentile) = 1.-12 x 10 7 year / 3-71 ~

Based on these low probability values of unacceptable turbine missile damage, it was concluded that turbine missiles were not a significant contributor to the plant risk. 3.4.5.4 Recent Turbine Missile Issues Subsequent to the preparation of the FSAR, there have been some significant activities in the area of turbine missile analysis. These may be grouped into two categories: stress corrosion cracking issues and refinements in the analytical techniques. 3.4.5.4.1 Stress Corrosion Cracking Issues Following the discovery of widespread stress-corrosion cracking in disks and rotors of operating turbines, turbine manufacturers have proposed several " hardware" fixes and changes in' the operating procedures. Until the proposed hardware fixes are accepted, the manufacturers suggest that the-turbine disks, and turbine control and overspeed protection systems be periodically inspected. Two approaches have been proposed for deriving the - frequency of volumetric inspection of the turbine disks. In the deterministic

approach, several conservative assumptions are made in the initiation and growth rate of stress-corrosion cracking and in the critical crack size.

The disks are inspected periodically such that any existing crack is detected before it reaches the critical crack size. In the probabilistic approach, a program for inspection of turbine. disk, valve, and control systems is chosen such that the probability of unacceptable. damage to the nuclear power plant systems due to turbine missiles is maintained at some acceptable level. The uncertainties in the crack initiation, crack growth rate, critical crack size, and in the success of overspeed protection systems are explicitly modeled in the evaluation of turbine failure probability. The probabilistic analysis would also - consider the particular features of the turbine (i.o., missile paramotors), - the arrangement of safety systems within the specific plant, and the effect of barriers in the path of turbine missiles. The NRC staff has established the maximum value of P1, i.e., -probability of turbine missile generation using an acceptable limit of 10-7 per year for-P. For unfavorably oriented 4 ' turbine _ generators (i.e., for plants having some safety-related -systems within the LTM zone), the NRC staff has PP23 would lie in the range of 10-3 to 10-2, concluded that Therefore, the staf f recommends that-P1 should.not be larger

than 10 5 per year (NUREG-0887, USNRC, 1983).

The value of P1 calculated using historical failure data (Patton et al., 1983) may not be appropriate in calculating the turbine missile risks; since, it is our judgment 'aat the stress corrosion 1 3-72

cracking issue would be resolved in the near future and that the probability of turbine failure leading to missile generation at LaSalle would be less than 10 5 per year. 3.4.5.4.2 Refinements in Turbine Missile Risk Analysis The FSAR analysis was utilized as a screening evaluation to show that the probability of unacceptable damage from turbine missiles to any of the ESF systems is acceptably small. The conservatisms and uncertainties in these analyses have to be assessed in light of the recent developments in the techniques for turbine missile analysis. The range of P P2 3 calculated in the FSAR has many conservatismst o the missile data provided by the turbine manufacturers l tend to overpredict the missile sizes and velocities o damage was assumed when scabbing of concrete barrier occurred; scabbing could lead to equipment damage only if there are sensitive instrumentation linec, valves, and cables in the path of scabbed pieces of concrete o damage ta any ESP equipment was deemed unacceptable; typicsily, a sequence of - equipment failures (" cut sets") < 10 to take-place in order to have core damage. The FSAR analysis used the modified Petry formula for calculating the value of P. Recent filll-scale missile impact 3 tests have shown that this formula is not a good predictor of scabbing or perforation thickness. As described in Section 3.4.5.1, a comprehensive probabilistic analysis of turbine missile damage would consider both the probabilistic characteristics of missile generation events, missile transportation, and missile impact with barriors, and the nuclear plant system characteristics wherein a sequence of components have to_ fail for the undesired event. If such an -analysis is done for LaSalle, it ir judged that the probability of turbine missile induced c re damage would be estimated as less than 1 x 10-7 per year u' ng the value of P1 l of 10*5 por year. 3 4.5.5 Conclusion Based on the bound'ng analysis, it is concluded that the turbine missiles are not a significant contributor to the plant risk. Therefore, no further detailed analysis of this event is-considered necessary. 3-73

~ 3.4.6 External Flooding The LaSalle County Station is located approximately 5 miles south of the Illinois River. The man-mado cooling lake adjacent to the plant has a surface area of 2058 acres at its normal pool elevation of 700' MSL. Make-up water for the cooling lake is pumped from the Illinois River, and part of the water in the lake is blown down to the Illinois River to provent dissolved solids in the lake from building up to excessive lovels. The ultimato heat sink for LSCS is an excavated pond which is constructed within the lake and has a surfaco aron of 83 acros at the design lovel of 690' MSL. Three modos of flooding were considered in the design of LaSallo County Station, i.e., (1) a postulated probablo maximum flood (pMF) in the Illinois River, (?) a probablo maximum precipitation (PMP) with antocodont standard project storm (SPS) on the cooling lake and its drainage area, and (3) a local' PHP at the plant site. For the present bounding analysis, modos of flooding for tho site woro also judged to be either from the river or from the lake or duo an intonne precipitation at the site. The plant design critoria as well as motoorological data for the site vero used to perform a bounding frequency analyais for external flooding. As shown below, the contribution cf flooding to the overall plant risk is-negligible. 3.4.6.1 Illinois River The structures in LaSallo Station have a floor clovation of 710.5' MSL and the plant grade is at Elevation 710' MSL. In comparison, the Illinola-River is normally at olevations under 500 MSL in the vicinity c' the site. The terrain at the sito is. gently rolling with ground surface elevations which vary-from 700' to 724' MSL, i.e., the sito elevation is much higher th a r. the Illinois River at all locations.- For the plant design, probable maximum flood elevation attthe Illinois River including coincident wayo offect:was calculated to be 522.5' MSL (FSAR). This in 188' below the plant floor elevation. Although tho' probable maximum flood lovel is not calculated on the basis of a given annual probability of exceedence, it is i thought to bc associated with-a very low oxceedence probability. In fact, -the observed maximum flood-water elevation in the-Illinois River has boon 504.7' MSL recorded in 1831. The river scroon house and the out-f all-structure .which are not safety-related structures are the only plant facilitics which could be damaged by floods in the. Illinois t -River. There are some low navigation dams in the Illinois River upstream.from the plant. However,-failure of those dams due to - floods or other events would not affect the site. p 3-74 s i

l Therefore, it may be concluded that floods at the Illinois River would not either directly or indirectly affect the plant uafety. 3.4.6.2 Cooling Lake The cooling lake at LaSalle site is at a lower elevation than the plant grado elevation, i.e., 700' MSL vs 710' MSL. There are three baffle dikes within the lake which chanrael the flow of water and increase the flow path for efficient heat dissipation. In case of an overflow due to an intense precipitation, runoff from the lake would flow away from the plant towards existing creeks and gullies. Also, in case of breaching of the peripheral dikes of the cooling lake, the impounded water woul.d discharge directly into local creeks that meet the Illinois River. Thus, it is concluded that the plant safety-related structures would not be affected by the probable maximum water level in the lake with coincident wind waves. 3.4.6.3 Local Precipitation In the LaSalle FSAR, it was concitund that the critical mode of flooding at the site is due to an intense local precipitation. The assumptions which were used in the design were as follows: a standard project storm followed by three rainless days and next followed by the probable maximum precipitation for a period of 48 hours. A hydrological analysis of the site was carried out which included the site topographic date, the cooling lake, and data for both the main spillway and the auxiliary spillway, l was shown that the water surf ace elevation near the plant buildings could reach an olevation of 710.34' MSL which is slightly lower than the floor elevation of 710.5' MSL. Ilowever, it is shown in the following paragraphs that the analysis was conservative and the calculated flood level corresponds to a very low annual probability of exceedence. An examination of the hydrologic analysis of LaSalle site showcd that conservatism in the analysis is mostly due to the definition of probable maximum pre;ipitation. In the plant design, the 24 hour local probable maximum precipitation for the site was calculated to be 32.1". Date. from the meteorological tower at the site and other weather stations near the site which were considered in the FSAR indicated maximum 24-hour and 48-hour precipitations of 4.45" and 8.62", respectively for record periods of up to 15 yeare. In the present study, r.eteorological data for weather stations near the site were obtained from the National Oceanic and Atmospheric Agency in Ashville, North Carolina. The most 3-75

complete set of precipitation records near the site is for the city of Chicaga which covers a 100 year period starting in t i 1871. 'I he s e duta show a maximum 24 hour precipitation of 6.19" which occurred in 1885. An examination of other weather station data in northern and central Illinois revealed a maximum recorded 24 hour precipitation of 7.56" which occurred in Cairo, Illinois in 1952. Therefore, the probable maximum precipitation which was calculated for LaSalle is expected to have a low negligible probability of exceedence. Table 3.4-16 shows the 100 year maximum 24 hour precipitation data for Chicago. Figure 3.4-19 shows the histogram of maxinum 24 hour rainfall for the 100 year period 1871 to 1970. Also, shown in Figure 3.4-19 is a normal distribution fit to the rainfall data. In addition to the normal distribution, four other probability distributions were also fit to the

data, i.e.,

lognormal uistribution, gamna distribution, extreme value type I distribution, and Log-Pct rson type III distribution. Figures 3.4-20 through 3.4-24 show plots of probability of exceedence of daily rainfall for frequencies of 10-10 to 10-4 based on these distributions. Also, stepicted in these figures are 90 percent confidence bounds on the probability of exceedence. From Figures 3.4-20 through 3.4-24, it may be concluded that the 24 hour PMP has a very low probability of occurrence, i.e., the 95 percent confidence value of 24 hour PMP has a probability of occurrence of less than 10-8 por year. Other conservative assumptions which were made in the site hydrological analysis are as follows: 1. It was conservatively assumed that all drains are clogged during the PMP. 2. No leakage or permeation of water into ground was assumed to occur during the storm. 3. The maximum precipitation is expected to occur for a vwy short period of time.

However, the analysis assumed a 48 hour PMP for the site.

4. An inspection of the plant during the site visit by SMA personnel revealed that the doors are leak-tight, i.e., even if water elevation rises above the plant

grade, the buildings will not be flooded.

In

addition, the structures have adequate drainage at ground elevation and they have been designed for possible flooding.

In view of these conservatisms and the conservatism in definition of PMP, it was concluded that external flooding does not contribute significantly to the risk of core damage in LaSalle. 3-76

f 1 Table 3.4-1 1 t Commercial Airports Within 20 Miles of the Site i 4 f Runways: Number of Distance & l a i Orientation / Type of. Operations Per Direction Airport Length (/ft) Type Aircraft Year By Type From Site i l Dwight. 90-27/2340 asphalt a) single-engine a) 9,850 16 miles SE f 4 1E-36/2000 turf b) twin-engine b) 1,100 t y Morris 16-36/3000-asphalt a) single-engine a) 6,570 17 miles ENE j j j Municipal 9-27/2500 turf b) twin-engine b) 730 [ Ottawa 5-23/2300 ' paved. a) single-engine a) 2,500 16 miles NW 4~ 18-36/2600 turf b) twin-engine b) 2,500 { 9-27/1900 turf l i Starved 10-28/3200 turf 17 miles WNW Rock I ![ Streator 9-27/2500 . asphalt a) single-engine a) 9,000 12 miles SW (B&S 18-36/1700' turf b) twin-engine b) 1,000 i l Aviation) [ I I I ReproducedEfrom the LaSalle FSAR. 1 i i t t 1-I t i

r p 4 Table 3.4-2 . Private' Airstrips Within 20 Miles of the Site . Airstrip Distance & Direction From Site r Cody Port 11 miles NW 5 Cwain' 18 miles N Fillman 14 miles ESE (( I' Gillespie 5 miles N i 1 w Holverson 6 miles N

a i

e Kenzie 16 miles NW i lentman 17 miles SW y Matteson 15 miles ESE [ i-l Mitchell 5 miles N l r r l Prairie Lake 7 miles N i l I j-Reicheing 18 miles NNW I t i Skinner 12 miles WSW i l [: Testoni 16 miles S I l Reproduced'from the-LaSalle FSAR. I f i f i

i Table 3.4-3 Aircraf t Traf fic Statistics Near the LaSalle Site for June 7, 1984 9,000 Feet 10,000 Feet and Below and Above 1. Peoria, IL direct Joliet, IL, V116 3 2

• 2.

Pontiac, IL direct Joliet, IL, V69 22 36 3. Airway J64 or direct routes which 0 61 overly the airway (24,000 and above) 4. Airway V156 or direct routes which 5 14 overly the airway (23,000 and above) b 5. Airway V38 or direct routes which 11 4 overly the airway 1 6. Pontiac, IL direct Joliet, IL 0 13 Joliet 360 Radial

• 7.

Pontiac, IL 24 92 V9 Planc, IL 0 4 8. Random routes over your facility 65 225 Totals Preferential irrival Routes. 20,000 feet descending to 10,000 feet. Summarized from June 15, 1984 Ltr. to S. Halloran.

t L l (. Table 3.4-4 -Annual In-Flight' Crash Rates (1 Mile) l Aircraft Type 5th Percentile. 50th Percentile 95th Percentile r. I w i-e Single-Engine 1.91 x 10-7 2.27 x 10-7 2.70 x 10-7 m l-o f Twin-Engine 5.54 x 10-8 7.14 x 10-8 9.20 x 10-8 I i Commercial 6.95 x 10-10 1.39 x 10-9 2.76 x 10-9 u.

t I Table 3.4-5 Annual Frequencies of Aircraft Impact For 'LaSalle Structures n i i . Building ' Aircraft Type Impact Area Airway Impact Frequency (mi2) (/yr) i i Reactor Building. . Twin-Engine 0.0115 V9,V156 2.1 x 10-7 f i V69,V116 1.7 x 10-7 t 3.8 x 10-7 f i Commercial O.0115 V9,V156 1.2 x 10-8 { t i-w i Random

7. 5 x 10-10 3-m j

H' 1.3 x 10-8 [ [ I 4 Aux Building Twin-Engine 0.0026 V9,V156 4.8 x 10-8 l V69,V116

4. 0 x 10-8 8.8 x 10-8 I

ti i Commercial O.0026 V9,V156 2.6 x 10-9 s I Random 1.7 x 10-10 2.7 x 10-9 i' i: j i-Total 5.0 x 10-7 l j I l 1 r r

l f I Table 3.4-6 I Intensity, Length, Width and Area Scales Fujita - F Pearson - P Pearson - P Area Scale Scale Intensity Scale Length Scale Width Scale No. (mph) (mi) (mi) (mi2) O 72 1.00 0.010 0. 0 0 '. Y 1.00-3.15 0.010-0.31 0.001-0.009 m 1 73-112 2 113-157 3.16-9.99 0.032-0.099 0.010-0.099 3 158-306 10.0-31.5 0.100-0.315 0.100-0.999 4 207-260 31.6-99.9 0.316-0.999 1.000-9.999 5 261-318 100-315 1.00-3.15 10.00-99.99 6 319-380 316-999 3.16-9.99 100.0-999.9 T. Fujita. Pennission to use this copywrited material was granted by T.

es .. ? l Table 3.4-7 -Regional Tornado-Occurrence - Intensity Relationsbips Corrected [ for Direct Classification Errors and Random Encouater Errors l -(EachERow in the Table is the Vector OI) I j' i ) Corrected Probability of Occurrence i at Each F-Scale Intensity F [ p Region. Scale F0 F1 F2 F3 F4 F5 F6 i i L ) Fig. 3.4.3-1 I- .2227 .3785 .2576 .1016 .0324 .0066 .0009 l II .3610 .3116 .2198- .0912 .0147 .0015 .0002. l [ - III .3044 .4421 .1730 .0681 .0112 .0012 .0001 i A .1658 .3379 .3122 1322- .0413 .0093 .0013 [ Fig. 3.4.3-2 j w B .2263 .3527 .2785 .1040 .0312 .0063 .0008 j m C .2830 .3611- .2426 .0856 .0225 .0047 .0006 L D .3034 .3799 .2436 .0622 .0096 .0011 .0001 I' l Region Regional Occurrence Rates Corrected for Unreported Tornadoes (occurrences per square mile. per year) l 1 ) l Fig. 3.4.3-1 I 4.12 x 10-6 [ i II 2.67 x 10-5 [ l III 1.35 x 10-5 ( { Fig. 3.4.3-2 A 5.18 x 10-4 L - B 6.98 x 10-4 i j C 3.37 x 10-4 f - D 3.53 x 10-5 I [ 4: Reproduced from Reinhold and Ellingwood 1983. i I ~ t

Table 3.4-8 Intensity-Area Relationship Including Corrections for Direct Observation and Randon Encounter Errors (AIM Matrix) Percentage of Tornadoes With Indicated Area Classification Actual Maximum Tornado State A0 Al A2 A3 A4 A5 i F0" .155 .421 .269 .125 .029 .0016 u m F1" .057 .255 .355 .259 .071 .003 b F2" .022 .139 .303 .368 .155 .013 F3" .009 .070 .210 .376 .289 .046 F4" .003 .033 .123 .299 .435 .107 F5" .001 .017 .068 .216 .4G1 .237 ~ F6" .001 .012 .049 .185 .458 .295 1 1983. Reproduced from Reinhold and Ellingwood

4 Table 3.4-9 Variation of Tornado Intensity Along Path Length and Across Path Width (VWL Matrix) True Maximum Tornado State Local 1._nado State F0" F1" F2" F3" F4" F5" F6" FO* 1.000 .743 .658 .615 .637 .632 .625 Fl* O .257 .248 .267 .234 .236 .23d F2* O O .094 .091 .093 .088 .089 F3* O O O .027 .028 .033 .033 F4* O O O O .008 .009 .011 FS* O O O O O .002 .003 F6* O O O O O O .001 Reproduced from Reinhold and Ellingwood 1983.

l Tab 1v6 3.4-10 4 Intensity-Length Relationship Including Corrections for Direct Observation and Random Encounter Errors (LIM Matrix) Percentage of Tornadoes With Indicated Length Classification Actual Mcximum Tornado State PLO PL1 PL2 PL3 PL4 PL5 w FO" .801 .115 .069 .014 N. 0 i F1" .590 .219 .140 .046 .005 0 F2" .436 .249 .212 .093 .010 0 F3" .272 .226 .268 .195 .C38 .001 F4" .141 .152 .272 .326 .090 .019 F5" .079 .113 .197 .444 .131 .036 F6" .058 .101 .155 .496 .147 .043 Reproduced from Reinhold and Ellingwood 1983. l l

-Table 3.4-11 Variation of Intensity Along' Length Based on. Percentage of Length Per Tornado (VL Matrix) i Recorded Tornado State Local Tornado State FO F1 F2 F3 F4 F5 F6 F0 1.000 .383 .180 .077 .130 .118 .100 ' F1-0 .617 .279 .245 .131 .125 .110 F2 0 0 .541 .310 .248 .162 .120 F3 0 0 0 .368 .234 .236 .160 F4 0 0 0 0 .257 .187 .200' F5 0 0 0 0 0 .172 .150 F6 0 0 0 0 0 0 .160 Reproduced from Reinhold and Ellingwood 1983. s t t ai

T Table 3.4-12 NRC SRP Tornado Missilcs.(Standard Review Plan)- V (Ft/Sec)' Missile Weight (Lbs) Dimensions Region I- 'A. ' Wood Planki 120 3.6" x 11.4" x 144" 270 B. 6" Sch. 40 Pipe 300 6.6".3 x 180" 170 C. 1" Steel ~ Rod 9 1" D x 36" 167 D. Utility Pole 1100 13.5" D x 420" 180 E.. 12" Sch. 40 Pipe 750 12.6" D x 180" 254 F. Automobile 4000 16.4' x 6.6' x 4.3' 194 m .. A.

Table 3.4-13 Minimum Reinforced Concrete Thicknesses (Inches) Required to Prevent Scabbing (NDRC and Chang's Formulas) Missile ts (NDRC) In ts (Chang) In Horizontal B. 6" Sch. 40 Pipe 8.2 18.8 i C. 1" Steel Rod 4.5 6.6 E. 12" Sch. 40 Pipe 10.3 22.3 Vertical B. 6" Sch. 40 Pipe 7.3 14.8 C. 1" Steel Rod 4.5 6.6 E. 12" Sch. 40 Pipe 8.3 17.5

}] ~ c l s Table 3.4 ' Estimates of Annual _ Probability of Turbine Missile Generation Failure Mode-Source Operating Overspeed Total Speed P_' P' ie O General. Electric 0 5.00x10 5.00x10-9 Patton et al. 1.20x10-4 0.44x10-4 1.64x10 This Report 1.17-x10-4 2.10x10-5 1.38x10-4

i ' Table 3.4-15 38-Inch Last Stage Bucket, 18&O RPM Low-Preseare Tazbine " Hypothetical Missile Data (1) (2) STAse CROUP 'l E - p Stage Members'in Gioup; . Number of.5.epresentative ' ~ Stage 1-4 2 4 - 6;5 (Last);7 -{ MISSIt'! DIMENSTOMS Tragment Group a b e d. a b e d a b e d '[ l Number of Fragments in droup 2 1 3 10 2 1 3 10 2 1 3 10 Sector Anale, degrees' 120 60 120 60 120 60 Fragment Weight, lbs 2000 1000 300 ~100 3000 1500 500 150 6500 3200 1800 200 LJ R3 Bore 18 18 17 17 16 16 3 Radius, in., UP R2 Rub 24 24 15 25 25 15 R3 Vane Root 45 45 45 45 45 45 Thickness, in., Hub 10 10 12 12 21 21 .Tg Web 3-3 5 5 10 10 T2 l Approximate Rectangular DLeensions, in. 19x19x3 11x11x3 19:19x5 19x10x5 tex 19x10 8x8x10 LOW SPEED B'JRST Postulated Speed: '2160 RPM (120) Lifetime Prebability: Not Statistically Signific ant 3 s.

L ? 1 t ! I T y y t t i i c 000 c 00 0 o 8 5 o 8 9 l 7 5 l 96 e e V V b d y y 5 I r m I r 035 r I e 31 e 031 n n P 8 E E U O E y y .R t t G 7 1 t i e 00 0 c 000 / E 1 2 o 0 1 2 o 1 5 G l 4 65 l 96 A e e d T V v n S a e o y y 5 c r 68 7 s e r 13 2 r 036 s e e 1 / n n t 'e E E e n e i y y f b t t r i i n .t u c 000 c 00 0 i s s T o 53 o 3 6 e o l 7 5 l 96 yg o e e e t nb r V v i a u b d crr s) I y 5 v o e 2 a ld p s I r e( r 03S r 021 ee P) 7 8 e 1 e vt r r n n a u 1 M E E E

t c

) w( D s s c d o G 7 a 4 y y d o 't L / t t nr t E 3 5 i t ue o n a o MD G c 000 e 00 0 ovt I A o 5 9 o 8 2 po c R e T l 53 t 8 6 d ( S e e td e l 0 v v oet 5 i 0 a e ot a 1 s 8 y y f ul s 1 m m b u 4 ,M i r 04 7 r 063 nit e 1 e ors 3 t l n n i t o e a E E l s p e l k li R c c y y id e b ui t t m r B i 000 i 000 ya A a t T e c 2 4 c 3 0 nl S eh o 64 e 1 8 i ms t g l l 1 re F o a I e e nos p V v ef s t y P 8 b d via e S H 1 y v i nl 8 t 0 E n a guc l s R r r e 063 e 021 s ee l a G 7 4 L / n n ib z a E 3 6 E E i h' G s os S c A y y et n .T t t l rI a u e o '0 L i id I S i 00 0 c 0 0 0 s c 8 o 7 4 o 3 6 stf1 3 l 4 3 l 96 i a e e ml nx e V v ui h 5 t a e d t y v es s 5 e t oe1 g r c pl r e e 07 3 e 04 2 e i s m 0 e c n n j es n 4 yg n E E er s a o 2 t a e aie 3 it r f mm r 3 t - l S r os f y i u e n7 3 i b - c yi e-c g gee ( l a1 ' Op r rt d T d i b S e b oe u e ex5 e R e a r c f o p nni U p b P n or u EES1 c B S) o) e G o u 1 r* l r y r ) ) ) ) D d0P( a r t e G 1 234 d E e8 7 nu i g ( ( ( ( E t 1 e-oc l a t t t t o P a( mE i c pit n mmn n s n n a r S l i tO ubS e uui. e ui p n uMt5 i oa m mmo. m a n o s H tI e df rb n g ii p g ii p e G sRf1 nog oi a nxd a nxd t e o i o r r iai r iai o P L C P I MMM F MMM N R s OV WI

b Table 3.4-16 Maximum 24 Hour Precipitation for Chicago . Year Inches Year Inches Year Inches 1871 2.57 1906 2.91' 194l'- 1.71: - 1872 2.70 1907-1.80 1942 1.98- . 1873 2.82 1908 4.34' .1943 3.93 . 1874 2.19 1909 3.52 1944 1.64 ' 1875 3.44 1910 1.81 1945 1.96 1876 1.94 1911 1.51 1946 2.46 1877 2.65 1912 1.87 1947 4.08 1878 4.14 1913 1.83 1948 2.50-1879' 3.25 1914 1.65 1949 2.73 1880 1.91 1915 2.48 1950 3.52 4 + - 1881 3.35 ~1916 2.61 1951 2.93 1882 1.92 1917 1.51-1952 1.60 1883

3.39 1918 1.92 1953 2.42-

' 1884 3.26 1919 .2. 28 ' 1954 2.20 'i y 1885. 6.19' -1920 2.28 1955 3.11 i e 1886-2.11 1921 2.60 1956 1.57 w 1887 1.39 1922 2.64 1957 6.24 1888 2.43 1923 3.70 1958 2.25 1889 4.02 1924' 3.75 1959 4.58 1890 2.60 1925 1.85 1960 2.86 1 1891 1.92 1926 3.02 1961 2.63 i 1892 3.11 1927 2.92 1962 1.82 1893 1.46 1928 2.71 1963 2.67 1894 3.35-1929 3.12 1964 2.09 1895 3.65 1930 1.48 1965 2.78 1896 2.42-1931 3.84 1966 5.39 c 1897 2.01 1932 2.03 1967 2.95 1898 2.50 1933 2.81 1968 3.83 1899 2.17 1934 1.86 1969 3.29 1900 1.48 1935 3.00 1970 2.97 1901-1.96 1936 2.69 'l 1902 2.02 1937 1.85 1903 1.54 1938 1.63 1904 1.83 1939 2.09 1905 2.78-1940-1.91

• 1957, 6.24 was 100 year maximum r

1 e - t -..m..-

F7. ",%c rg.u,,,y,).,,, _ 4e ' n* t ., y y fr A 1 -en.,,,& g',g m rrr gs

i.,,,f,,

as ' ,,$,004 [l u r..t. e. i, uwota ' i '" ' ' 3lr. n) . th ', / ('"iY/. M e i. < + c.w,o . N's p[I %e.. q q__iT y j; .LsQ.tp

1. 'l

/ 8 4 '. 2 0,. Miles '.:u. i~ p i f a ~= T p- /, s.: e .. ~ - p.. AIos ',kl A .. c.... J s # s. 4 WE. *i a=,u, i" ,d

1. J','

V /, .... 4>.:n i *".? M,' ~ f'-3@*Y 64.@,.d',, C., A ')oe a

,,, j p=.'Dr_f...o 3

A,, s f4 ~ z / n;. *. e... hy {',.", Gi A '"/A'J,h d"~KGJ " =* 7W i }". n. m g. p ,p f. :: w... p!yy,w ;.,a e,

p 1

,,,,, j

4.., g.:.~ -

ph a 9 s ,hi E S I T E

1. ' D f

m. w%ei, e'%ln.u,

.e 4m. w; $'. h s 7 N ed F 6 ~ A m w

c. J

<.. s.- e,_ ~ Yr.' ! [j% m *: ,f 1\\7 ret.. ^s /- T --/ ,, t ',f ;n,c,,,5 4 , r,,' rerm. y. n.' l2 4,s _ x.; e, .2*. ase, m,us e. ,A i w'.N I., f).orowi....e,,,, j. 4 "v/,7,'N - u.. 1 ....a s v. 'j4,c ,-m...a*J. 4 f .m ] 3 I p 8., g $ odh J / s._ 7, f &y: %~~ n t~4 f~ '7,,, ,7; ^ . k4 g.,.. [.; rf'=/ s .n/ i ]'I q4 I " }i 3.../ M ..a6 l A. d't! .~ "2 /h '""**j f1/_-- e-J d5 re jo a..., {^ j.o b,,- jo N /,,.:O" y

• ~is

\\ _. /'y"~ ' @,. r.mA'N....L' " - "ce.,,....!. 7P !~ yj77,:ci:.sy..,qcy].!> t n SCALE O 10 20 Miles ~ O 10 20 Kilomsters W LEGEND Airporis O No paved runway g No paved run'way (Restricted use) $ Herd surfoes runway Figure 3.4-1. Airports and Flight Patterns Within 20 Miles of the Site Reproduced from the LaSalle FSAR. 3-94

gh w' AIRCRAFT

• /2 TRAJECTORY a

+ B b )/ PLANT LOCATION d ALONG THIS LINE Figure 3.4-2. Geometry for Aircraft Impact Probabilistic Model

~ f %.. / \\ 7 / ___. i _p,, j/ 3o.- 7 ,/ WP'/g ( ,t*= a+- f 3, \\ / ,/, .g.9 9 },.... g, s z ... x,f, 6 -f ~ .. = " " * ' * * ' ///// s e=

9-...

/l i 4 \\ 4 l \\ 'J ). e a-' Markee Figure 3.4-3. Tornado Risk Regionalization Scheme Proposed by Wash-1300, et al. (1975) Reproduced from Reinhold and Ellingwood 1983. - - = = = = =

1 A D- / \\ ._ i I / 7 [' / B ph d \\ i Cf

\\

/,\\_Y 4r e w 1 / v C a D 7 q i A. ~ 1 I 4 Bi C l I Figure 3.4-4. Tornado Risk Regionalization Scheme Proposed by Twisdale and Dunn (1983), i ' Permission to use this copywrited material granted by W.'R. Sugnet.

+e*'- STRCTURE +-A-* 7 s s B. ,s ,y H 'J L_ \\ hW ,/ f TORNADO / \\ W* / I / '/ L / l 4 ./ ). l/ / y b / \\BA2 \\ <E 3

1. ,/

/\\ ~ l\\ ,/ , r. p ,/ .,s s' /.

g. -

,4 .l\\( n /. l' 1. / T = TORNADO ARE A = WL-t \\ I B A g(. / - P = PROJECTED AREA = HL -BA = BUILDING AREA = AB L .v l E E = EDGE AREA = WG Figure 3.4-5.- Tornado Parameters and Damage Origin Area -Definition-Permission to use this copywrited material granted by MIT. 3-98 1

' ~ ~. ~ F0 /// / fl F3 // F4 Figure 3.4-6. Sketch of Hypothetical F4 Tornado Illustrating Variation of Intensity Reproduced from Reinhold and Ellingwood 1983. 3-99

10-3 10-4 d a od 10-5 i W H=500' [ H=300' O g H=100' C E y 10~0 n.

d

'Ee 10' 10'O i t i I O 50 100 150 200 250 300 350 MAXIMUM TORNADO WIhD VELOCITY (mph) Figure 3.4-7. Tornado Hazard Curves for LaSalle Site 3-100

95* 90 '85* 80* - 75* 50' 70* 65*- 1 3

v ~.

I \\ ~ 50 .( k-45' ._.'s,..._.. . ~... i w -,..a. ~.. '. .\\ .-1.. '). .-T <' 1~1 \\. i,, w ...c .,. c. 3...-- c y e .a- - 3

s..

t. t

• i.>

45 u: i, ,j. ., 2 s,w..t.. .c. _.e.x i p w j s.-. 40* a..- /. 0 .. i

t. -- -- ;;,7..._..

~ \\ i .o* .g,,,. ,,,. 3 ./ 7~- ......... -.\\ \\

1.. -. - i,

...-p.-:;?e..- 70* ...s s.

.),

g g

y..... :,;-f

\\ 3~ \\. - r' 'u. u.i, - P ',.- 40 ---- -- - \\

• ~ ' ' ' ' *. -

'~~" ~ s. k .'...'.t- [ .m %i - r-o t.,s yyes '- i. .?.;* l _..t 95 90 85 80 75* 1 Figure 3.4-8. Station Locations Reproduced from Changery 1982.

m - 10'3 10'4 d5 10-5 hl 0.1 0.8 5 0.1 l

• b R

f

10-6 E

.i L-E4 '10-7 10-8 3 '0 50 -100 150 '200 250 300 350 L.. 1 HAXIMUM TORNADO WIND VELOCITY (mph) l l l Figure 3.4-9. Family of Tornado Hazard Curves for the LaSalle Site With. Corresponding Subjective Probabilities 3-102

1.0 i

PROB, O.100 0.8-0.200 0.400 0.200 i

1 e 0.100 3 2 w 0.6_ g U58u w N 0.4 - g p E5 v 0.2_ 0 0 200 4b0 6d0 80'0 1000 MAXIMUM TORNADO WIND VELOCITY (MPH) Figure 3.4-10. Tornado Fragility Curves for LaSalle Including Uncertainty in Median Capacity 3-103

a .~r d' z~ o.. -o F--uzo LL. :r N mz tu m a-d-t- -.s e coa C f-coo cc o_ o I l I o l10" 10~" 10 'l 0" 10~' l0~' RNNUAL FREQUENCY OF FA] LURE Figure 3.4-11. Distribution of Annual Frequency of Core Melt in LaSalle Due to Tornadoes 3-104

. _ _. = l l N MI* r"E M'. "j:M i /g-/} "&.Ni p,* l ~ ^d Immre

• j l

lw n m. ~ -

,p

f..x N'Q &l"$7. f,,,Y k ~.ff n}

'.T W 3 V.>. v +. %'sTMQd. g/- (_qm%@1c?.#.;g,g?.,'e,e ;e,:f2 a v33 L 3 _ > gM" S' " ^ j+.ji?"?, '( n. F - 570 /, ' .;p .c i*i t ta'~ -I~t ::. ;. ;g spi ? p -.;;,r'~ = ~. y R-f t

g s x4,.5 c v

%;i.Q w.b'.~t-4.-'~' N. :v.r. ~, - i t. - @ ~.;a nf u,<w ~* M. m .a s,.;. -M ' C8 ",t .V'^ c { ( rhg b fh f i .f.c.' .upr eyi.. %g,.- 4. ',.. i@1 -(- M,+. 3.4 '.. f~ v g,,,.......,, a r g *b.... s.i e. 44 3,"[htX2QGeI.* r-U. o 'j ' - V ii JJ n-q d-J.7 ' w- -e----,--- 4a: m +

._a L

3 : '. bs.1:: /. [c-4q',3,.P r.. u, -,2__~,_ _. t.- N er m !1p- < _ M m,. W{ - h 1, _ m..m.:1m;.- 4..(, r I y w,,,, 'n..i. .m, s >j! \\v zV N_

• t 1_c;,N.s.j,170.%a

',.. *,I. ,e. ,p,; o- --.c _,} ~., - i g y.. ,_i \\,,, 'p_ SITE {i ; Jl W r I . o"J, m 716_'.. .f.6,.n.I (gl 6^t_ /l..iL. i N r-'-%e]. T <' w I l'2 ' l,_S 1_. Z P *ap is - 1 g. .. SQ' M. ?M,'. y b 4 i Y' ' '., @ 2 :. - L i I C ) M., e,* l.J). {'i$l' r,4 i f 3 id i F ~ 'l

• /

44 I I I 'r g g a.,.y e,. d p o u. a "-5 ', d b [bg"r...)....,M.$.py;bE--. p, q,q q; - b ", - - rd 'I.N l -- w i y p. !, ,J i 1 .. _. -, -- 1,_ :o,q-i p 2g. L,.i .L._ 3 ._*n. i q e,4 ,,_ g, - 7 i u 9,a ;: 1<y' I _%n l.- j t y) i m r.S i s q w,**e,- :;,.c s. ",J. 4. p ...,._.i - m m' a r ~1 r if,3 .r, . c._ c -. e -4 i / g . a4 N -l M, I i F Wi _ m,, ehfvM 15 ~- f# . _i; a.d C@os :T Opp,i,

-*J (-e.-.,.

- ~.

. _ 5 e_r_r:aucr p...:.L,.C, e-L_,i 4 " -l - --.d o. 'a. s r , j _E 4 1 d l J. 'H. ' '>A. ; ; E k, > -. '3 ep I, 6'/;t, .,,. m i w,2. R o ..u ~, ..w 1 4 .~ 1 -- s , Ag ' -.. ..Q g

",p'*.--] h ,i %nIno T

- ;.c, wu,,a'b:r'[tb'upWrAto* <',H' '.r * " A I Cl11

m b u. }g. g N r

-A. '. * ~ +w CL y 6 - n,a M ~,, e n --'"(17 -. j j

• nN N, r,,.,

ym - --j'.O :.f % Q,E ~}l$l.. Q V ' V""> ' y.1?-C. - +- -{2 n c ,i> 1g,- .,.-s/ y..f.1 ' A,%s L,- t - 8,4 f.%vi $ M l / w l -l.

s >r v

0 5 to Miles O 5 10 Kilome ters LEGEND ..Q

• iriterstate Highways h U S. Highways

-O-stot, s,s.o,s G co e., s.,s.o,, Rstroods Figure 3.4-12. Transportation Routes Near LaSalle County Station Reproduced from the LaSalle PSAR. 3-105

100A00 ~ 4 U + - - - - t,,,,, 8 g ~ g,LY g Q = -3 g J w 4 *== = - l ,,oo } I m W u E li 1l .1 E a 5 2 51 al 8 -1 "l 8 a T l 1I S g I 81 'gg

• g S

Il -l N-l 103 104 105 ige ig7 TNT EQUIVALENT IN POUNDS l Radius to Peak Incident Pressure of 1 psi Figure 3.4-13. 1.91. Reproduced from Regulatory Guide

l I l l I l l l 1 0 cate Equation of hne: f LOG A -138021 RA) * % [f +,rf 1 j a 2.45318 4h 103 g 4 ll a E O I E 100 O 10 .02 0.1 0.2 0.5 08 0.9 0 98 Fraction of flammable plumes egnated. P(A) A = 0.175 r2 Figure 3.4-14. Probability of Flammable-Plume Ignition Versus Plume Area at Time of 1_ . tion Reproduced from Eichler 1978. 3-107 l

e PRESSURE a) INCIDENT PRESSURE P, P-TIME o t. PRESSURE P,r b) REFLECTED PRESSURE P,,- e N % I T TIME t n t+ e .t d Figure 3.4-15. Pressure Pulses From TNT -Reproduced from Kennedy 1983. 3-108 ~

i 10,000 g PEAM PEAK OVEjRPRESSURE-Pill OVERPREISURE ARRIVALTIME-MSEC /L8il1#8 POSITIVE SURATION-MSI:C/L881/3 . POSITIVE lMPULSE-986-M BEC/LBS1/3 SHOCK VI:LOCITY-FT/MllEC 1000 I I" ~ IMPULSE YELOCITY 10 3 y 1.0 3 4 DURATION v 0.1 7 Ani1 VAL TIME .01 .01_ .10 1.0 to 100 1000 GROUND RAM A (FT/LBS '8) I Figure 3.4-16. ' Free-Field Blast Wave-Parameters Versus Scaled Distance for TNT Surface Bursts (Hemispherical Charges) Reproduced from Kennedy 1983. 3-109

l l 2.0 y g g p p l / f .s s 1.6 ,7 l g l / / I I// / s 08 4

1. d

! / Y / Y = It f fI / I g l 0 l I 005 0.10 0.2 0.5 1.0 2.0 to 1/7 4 Figure 3.4-17. Dynamic Load Factors. Maximum response of one-degree elastic systems (undamped) subjected to rectangular and triangular load pulses having zero rise time. Reproduced from Biggs 1964 with permission. 1 3-110

b Va . j Vg e I+0 !1 iW y 1 ? V l' Vo Initial Velocity 01 Angle from Y Axis to Vyz, Ose, 590* 02 Angle from YZ Plant To Vo, -Ase2sA, A<5 (Inner Disk) ; Os0 sa~,_As25* _(Outer Disk) 2 03-Angle on the Ground, _90* 503s90* 4 Angle from Ground to Vo, 0240, Then 4401 Figure 3.4-18. Var ables and Terminology Used in Calculating i Missile Strike Probabilities 3-111

LP, SRLLE EXTERNRL-EVENT tRRIM FRLL RECORD, 1971-13701 O

• NORMRL-D151RIBU110N

,o d-E d-E o- >-uzo $ d-aw te LL. ?. d-S =i ~ o 6-N o S 1.00 2'.00 3'.00 14'. 0 0 5'.00 6.00 7.00 Maxiinum 24 Hour Precipitation Figure 3.4-19. Histogram of Maximum Daily Precipitation for Chicago l 3-112

L A 5ALLE EXTERNAL EVENT tR91N FALL REC 3RD,!B71-1970! 10-y NORMAL OlS1RIDUT10N ' x x 'N \\ \\ 10-5 A 's N \\ \\ \\ N N 10-0 ~ 'N 'N \\ o.ss \\ \\ 0 \\ \\ 10-7 M \\ \\~\\ \\ s N s E \\ \\ .1 se \\ \\ 10 8 x Ag 5 ~ 10-9 \\ \\ \\ \\ O.05 !0 10 6.00 6.4C 6'.00 7.20 ~.60 8.00 0.40 tiaximum 24 Hour Precipitation Figure 3.4-20. Normal Distribution Fit for Maximum Daily Precipitation 3-113

LR 5ALLE EX1ERNAL EVEN1 (RR)N F A' L RECSRD, 1871-19701 10-y L of_NSRMAL 0151R18U11SN w x 's N N. \\ 10-5 N x \\ N x 10-6 'm \\ 'N oo x 'M N 'N d N 's w -

• 1

.s g x x 3 N, N, g N N ! O-l 'Ua g , 'x N 10-3 x A h \\ s ,,,g !O-10 19.00 !!.00 12.00 13.00 14.00 15.00 16.00 Maximum 24 Hour Precipitation Figuro 3.4-21. Lognormal Distributiot Fit For Maximum Daily Precipitation 3-114

LR SRLLE EX1ERNAL EVENT tRRIN F ALL RECORD,1671-IS10s 10-y CAMMA DISTRIBUTION ~' ^ ~ N \\ 10-5 x N N 'N 'N N. 10-6 'm s 1 's 'x X \\ 'd \\ \\ l0-7 O r. A M i..__. w s s( N, L6 A N \\ d \\ \\ fl0-e \\ = x 8 x h h a s 10-9 's s A \\ 0,05 10-10 8.00 0.50 9.00 9.50 10.00 10.53 11.00 Maximum 24 Hour Precipitation Figure 3.4-22. Gamma Distribution Fit for Maximum Daily Precipitation 3-115

LA SALLE EXTERNAL EVENT TRAIN FALL RECORD, 1811-1970) 10~ 9 E *I"I"E # AL UE I #E I OISI IOUIIO A ,x x N \\. 10-5 N ' N. s -x N N x \\ \\ 10-6 w N 's e et N '\\ \\ d 's b10-1 \\ c + Es N-N g x x U \\ \\ 5 'N 'N 8 10~9 ataa d s N \\ il N 10~9 xx \\ 'N

o. :s 10-10 10.00 11.00 12.00 13.00 14.00 15.00 16.00 Maximum 24 Hour Precipitation Figuro 3.4-23.

Extreme Valut Type I Distribution Fit For Maximum Daily Precipitation I-116

LA SALLE EXTERNAL EVENT (PAlH FALL RECORD, 1871-1970) 30-4 LOG.PEARSON TYPE 111 DISTRIBUTION ,,.[s \\ -4,. w X

1. f

\\ \\ ,0-s 's 1 \\ \\ \\ \\ ( \\ \\ \\ \\ \\ ,~ 1 x \\ \\ \\ 0.05 w \\ ~\\ \\ M \\ \\ \\ \\ \\ S $10-7 w 1 \\ \\ \\ \\ O \\ \\ 2 \\ \\ \\, \\. 10 - at== y \\ '1 a af \\ \\ \\ \\ \\ ,~ 's \\ \\ \\ \\ o.os gg-tu d D) 12 14 lb 18 U Maxinum 24 Hour Precipitation Figure 3.4-24. Log-Pearson Type III Distribution Pit for Maximum Daily Precipitation 3-117

3.5 Eygnts R_qmi_ir_ing_Itel!Liled ELM i Dounding analyses for the events which could not be excluded based on the initial screening process were presented in Section 3.4. These events included aircraft impact, winds and tornadoes, transportation accidents, turbine generated missiles and external flooding. Among these external events, aircraft impact and tornadoes woro found to be potential contributors to the plant risk. Based on the bounding analysis for aircraft impact, the median frequency of core 10-7 year. Also, the damage was found to be equal to 5 x / uncertainty analysis showed that the 95 percent confidence bound for the frequency of damage due to aircraft inpact is 10-6 year. For tornadoes, the median frequency of core damage / was calculated to be 3 x 10-8/your whereas the 95 percent confidence bound was calculated n be 3 x 10-7/vear. As mentioned in Section 3.4, the bounding analyses did not account for the plant systems failures and accident sequences leading to a core damage and, therefore, was generally conservative. In light of the conservatism in the bounding analyses and also the low frequencies of core damage for aircraft impact and tornadoes, it was concluded that the external events considered in this scoping quantification study are not significant contributors to the plant risk. llowever, a detailed evaluation of aircraft impact risk to the LaSalle site may become necessary if the contribution of internal events to the risk is found to be less than 10 6/ year. 3-118

4.0

SUMMARY

A11D RECOMMEllDATIONS A scoping quantification study was performed for the LaSalle County Station to determine the external events which should be included in the detailed PRA study performed as part of the RMIEP program. Section 4.1 summarizes the results and Section 4.2 presents the recommendations of this study. 4.1 finlamary The scoping quantification study which was performed here considered all possible external events at the site except for internal flooding, seismic and fire events, 1.c., these three events were included in a detailed external events analysis. The PRA Procedures Guide (1983) was used as a guideline for identification of all possible external events at the LaSalle site. Next, an initial screening proccas was carried out to eliminate some of the events from the list. For this purpose, a not of screening criteria was developed and then each external event was examined for possible elimination based on these criteria. After the initial screening process was completed, the following events were found to be potential contributions to the plant risk. 1. Military and industrial facilities accidents 2. Pipeline accidents 3. Release of chemicals in onsite storage 4. Aircraft impact 5. Extreme winds and tornadoes 6. Transportation accidents 7. Turbine generated missiles 8. External flooding The top three events in this group were climinated based on the analyses and information which is presented in the LaSalle PSAR. A probabilistic bounding analysis was performed for each of the remaining five events in the above list. The degree of sophistication in the bounding analysis for each event was dependent on whether the event could be eliminated based on only a hazard analysis or a complete analysis including hazard

analysis, fragility evaluation, and response analysis.

Ilowever, the plant system and accident sequence analysis was conservatively neglected for these external events. For aircraft impact, the median frequency of core damage was calculated as 5 x 10-7/ year whereas the 95 percent confidence 10 6 year. An evaluation of the plant bound was found to be / 4-1 Y,

structures for extreme winds and tornadoes revealed that I extreme winds do not dominate the response and therefore could be eliminated from the bounding analysis. The median frequency of plant core damage due to tornadoes was calculated 10 8 year, and its 95 percent confidence bound was / to be 3 x found to be 3 >: 10-7/ year. The bounding analysis for transportation accidents including toxic chemical release and chemical explosions showed that these accidents do not significantly contribute to the plant risk. For turbine generated missilon, the FSAR onalysis was re-examined in light of new information regarding the generation of such missiles. It was concluded that the 95 percent confidenc-bound on the frequency of a plant damage state due to turbine missiles is on the order of 10 7/ year. The bounding analysis for external flooding showed that probability of occurrence of the probable maximum precipitation ( PMP) at the site for which the plant has been designed for is indeed very low. Since the only credible mode of flooding at LaSalle is due to an intense local precipitation, this event could be climinated from the detailed PRA study. 4.2 ILeg.gpmenda t ions The bounding analysis of potential external events at the LaSalle site showed that only aircraft impact and tornadoes may be potential contributors to the plant risk. For aircraft impact, the 95 percent confidence bound on the frequency of core damage was calculated to be 10-6/ year and for tornadoes, the 95 percent confidence bound was calculated to be 10 7 year. Since the bounding analysis did not consider / 3 x the plant systems failures and consequence analysis, these frequencies are generally conservative. It is our judgement that none of the external events considered in this scoping quantification study is a significant contributor to the plant risk. Ilowever, if the PRA analysis for internal events should show that contribution of the internal events to the risk is less than 10 - 6/ye a r, then there may be a need to further examine aircraft impact and tornado events. 4-2

REFERENCES

Abbey, R.

F., " Risk Probabilities Associated with Tornado Wind Speeds," Proceedings, Symposium on Tornadoes,

Lubbock, Texas, June, 1976.

" Annual Review of Airport Accident Rates, Calendar Year 1980," National Transportation Safety Board, NTSB-ARG-80-1, May 1980.

Berriaud, C.,

Sokolovsky, A.,

Gueraud, R.,

Dulac, J.,

Labrot, R.,

and R. Avet-Flancard, " Local Behavior of the Reinforced Concrete Walls Under Missile Impact", Huclear En91DR2Tino ajnLDenian, pp. 457-469, Vol. 45, 1978.

Diggs, J.

M., IntrpAu c t i o n _ t o_J t r u cj;nr31_py n a n i gff, McGraw-liill Book company, 1964.

Boboe, B.,

"The Log Pearson Type 3 Distribution and Its Application in Hydrology", liat.qE_Re qo_q.t.c e s Research, Vol. II, No. 5, October 1975.

Bobee, B.,

"The Use of the Pearson Type 3 and nog Pearson Type 3 Distributions Revisited," Water Recources Research, Vol. 13, No. 2, April 1977.

Bush, S.

I!., ' Probability of Damage to Nuclear Components Due to Turbine Failure," Nucl ear _jiaLqty, Vol. 14, No. 3, May-June 1973. Chang, W. S., " Impact of Solid Missiles on Concrete Barriers," Journal of the Structural D i v i s tp_D,

ASCE, Vol.

107, No. ST2, pp. 257-271, February 1981.

Changery, M.

J., " Historical Extreme Winds for the United Great Lakes and Adjacent Regions," National States Oceanic and Atmospheric Administration, NURE3/CR-2890, August 1982. Chelapati, C. V.,

Kennedy, R.

P. and I. B.

Wall, "Probabilistic Assessment of Aircraft llazard for Nuclear Power Plants," linglear Enaineerino and DqE13D, 1972.

Eichler, T.

V. and H. S. Napadensky, " Accidental Vapor Phase Explosions on Transportation Routes Near Nuclear Power Plants," IIT, Research Institute, Chicago, I L, NUREG/CR-0075, May 1978.

Eichler, T.,

Napadensky, H. and J. Mavec, " Evaluation of the Risks to the Marble Hill Nuclear Generating Station from Traffic on the Ohio River," Engineering Division, IIT Research Institute, Chicago, I L, October 1978. R-1

REFERENCES (Continued) Ellingwood, D., " Reliability Basis of Load and Resistance Factors for Reinforced Concrete Design," National Bureau of Standards, Washington, D.C., February 1978. "FAA Statistical llandbook of Aviation, Calendar Year 1979," U.S. Department of Transportation, Federal Aviation Administration, December 1979.

Ferguson, P.

M., IlpMQIr_pd Concrete Fun (10ntnt a l s, Third Edition, John Wiley and Sons, 1973.

Filstein, E.,

and M. K.

Ravindra,

" Probability Analysis of Turbine Missila Hazard," Proceedings of ASCE/EMD Spc 2alty Conference, Austin, Texas, September 1979.

Fujita, T.

T. and A. D.

Pearson, "Results of FPP Classification of 1971 and 1972 Tornadoes,"

Paper presented at the Eighth Conference on Severe Local Storms, October 1973.

Galambos, T.

V. and M. K. Ravindra, " Properties of Steel for Use in LRFD," Journal of the StIJmt;3 ural Divis h, ASCE, September 1978.

Garson, R.

C.,

Catalan, J.

M. and C. A.

Cornell,

" Tornado Design Winds Based on Risk", Department of Civil Engineering, MIT, August 1974. General Electric Company, " Memo Report - liypothetical Turbine Misciles - Probability of occurrence," March 14, 1973. General Electric Company, " Memo Report - Hypothetical Turbino Missile Data 38-Inch Last-Stage Bucket Units," March 16, 1973. Ho, F. P. and J. T. Riedel, " Seasonal Variation of 10-Square-United Mile Probable Maximum Precipitation Estimates States East of 105th Meridian," U. S. National Weather Service, Silver Springs, MD, NUREG/CR-1486, June 1980. Kennedy, R. P.,

Blejwas, T.

E. and D. E. Bennett, " Capacity of Nuclear Power Plant Structures to Resist Blast Loadings," Sandia National Laboratories, Albuquerque, NM, NUREG/CR-2462, SAND 83-1250, September 1983. Letter from Department of Transportation (FAA) to S.

Hallaron, Sargent and Lundy Engineers, June 15, 1984.

"LaSalle County Station, Final Safety Analysis Report," Commonwealth Edison Company. R-2

REFERENCES (Continued)

Markee, E.

J., Beckerley, J. G. and K. E. Sanders," Technical Basis for Interim Regional Tornado Criteria," WAS!! 1300, U.S. Government Printing Office, Washington, May 1974. Mcdonald, J. R., "A Methodology f or Tornado llazard Probability Assessment," Texas Tech University, Lubbock, TX, liUREG/CR-3058, October 1983.

Mirza, S.

A. and J. G. MacGregor, " Variability of Mechanical Proporties of Reinf orcing Bars," LoMEnal of the Stntetural DIY.in19D, ASCE, May 1979.

Mirza, S.

A., llatzinikolas, M. and J. G. MacGregor, " Statistical Descriptions of Strength of Concrete," Lo.llI.Dal of tJ1e Structural Diviqi9D, ASCE, June 1979.

Niyogi, P.

K. et. al., " Safety Design of 11uclear Power Plants Against Aircraft Impacts," Proceedings of Topical Meeting on Thermal Reactor Safety, Sun Valley, Idaho, July 31 August 4, 1977. NSAC (Nuclear Safety Analysis Center, Electric Power Research Institute), " Lightning Problems and Protection at 11uclear Power Plants," NSAC/41, December 1981.

Park, P.

and T. Paulay, Eg.intgrsed Concrete Sintginte.E, John Wiley and Sons, 1975.

Patton, E.

M., et al., "Probabilistic Analysis of Low-Pressure Steam Turbine-Missile Generation Events," Battelle Pacific Northwest Laboratories, EPRI 11 P-2 7 4 9, August 1983. " Midland Probabilistic Risk Assessment," Consumer Power Company, Midland, MI, 1984.

Reinhold, T.

A. and B. Ellingwood, "Tornndo Damage Risk Assessment," National Bureau of Standards, Washington, DC, NUREG/CR-2944, September 1982.

Ravindra, M.

K. and II.

Banon, "Mothods for External Event Screening Quantification:

Risk Methods Integration and Evaluation Program (RMIEP) Methods Development," prepared for Sandia National Laboratories, NUREG/CR-4839, SAND 07-7156, March 1992 4 R-3

l l REFERE!1CES (Continued) "Seabrook Probabilistic Safety Study,"

Pickard, Lowe and Garrick, Inc., 11ewport Beach, California, 1983.

Semanderes, S. N., "Mothods for Determining the Probability of a Turbine Missile liitting a Particular Plant Reglen," Westinghouse Topical Report MCAP-7861, February 1972. Philadelphia Electric Company, Severe Accident Rish 2LQJen sme nt. Limer19.k_Jlc.De ratinct Station, Philadelphia, PA, 1983. Schaefrar, J. T.,

Kelley, D.

L. and R. F. Abbey, " Tornado Track Characteristics and llazard Probabilities," Proceedings of the Fifth International Conference on Wind Engineering, Pergamon Press, New York, 1980.

Sliter, G.

E., " Assessment of Empirical Concrete Impact Formulas," Journal of the Stnig_tural DiviD1QD, ASCE, May 1980.

Sliter, G.

E.,

Chu, B.

B. and M. K. Ravindra, "EPRI Research on Turbine Missile Effects in 11uclear Power Plants," Transactions of 7th International Conference on Structural Mechanics in Reactor Technology, Paper J8/5, Chicago, IL, pp. 401-409, 1983. Stephenson, A. E., " Full Scale Tornado-Missile Impact Tests," Electric Power Research Institute, RP-399, August 1976.

Solomon, K.

A. et. al., " Estimate of the llazards to a Nuclear Reactor from the Random Impact of Meteoritos," UCLA-ENG-7426, March 1974. Southwest Research Institute, " Steam Turbine Disc Cracking Experience, Volume 2: Data Summarios and Discussion," g Prepared for the Electric Power Research Institute, EPRI NP-2429, June 1982. " Tornado-Borne Missile Speeds," NBSIR 76-1050, National Bureau of Standards, April 1976.

Twisdale, L.

A. and W. L. Dunn, " Tornado Missile Simulation and Design Methndology," Research Triangle Institute, NP-2005, Vol. 1, August 1981.

Twisdale, L.

A. and W. L.

Dunn, "Probabilistic Analysis of Tornado Wind Risks," Journal of tlLe Structural Division, ASCE, February 1983.

R-4

REFERrllCES (Continued)

Twisdale, L.

A. et al., "Probabilistic Analysis of Turbine-Missile Risks," Report prepared for the Electric Power Research Institute, EPRI llP-2749, February 1983.

Uman, M.

A., Ende.ligdilndi.p y Lichtnina, Bok Technical Publications, Inc., Caraegia, Pennsylvania, 1971. United States Water Resources Council, " Guidelines for Determining Flood frequency," US/WRC-1121, March 1976. USNRC, "PRA Procedures Guide," 14UREG/CR-2300, January, 1983.

USNRC, Regulatory Guide 1.115,

" Protection Againut Low-Trajectory Turbine Missiles."

USNRC, Regulatory Guide 1.91,

" Evaluations of Explosions Postulated to Occur on Transportation Routes Near Nuclear Power Plants." USNRC, Regulatory Guide 1.78, " Assumptions for Evaluating the liabitability of a lluclear Power Plant Control Room During a Postulated llazardous Chemical Release."

USNRC, Regulatory Guide 1.76,

" Design Basis Tornado for Nuclear Power Plants." USNRC, " Safety Evaluation Deport Related to the Operation of Perry 11uclea r Power Plant Units 1 and 2," NUREG/0887, supplement No. 3, 1983. USNRC, " Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants," NUREG-0800-75/087, 1975.

Wen, Y-K,

" Note on Analytical Modeling in Assessment of Tornado Risk," Proceedings, Symposium on Tornadoes, Lubbock, Texas, June 1976. R-5

Distribution James Abel Commonwealth Edison Co. 35 1st National Vest Chicago, IL 60600 Kiyoharu Abe Department of Riactor datety Research-Nuclear Safety Research Center Tokai Research Establishment JAERI Tokai mura Naga gun Ibaraki ken, JAPAN Bharat B. Agrawal -USNRC.RES/PRAB MS: NLS 372 J. Alman Commonwealth Edison Co. LaSalle County Station RR1, Box 220 2601 North 21st Rd. Marnielles,'IL 61341 Ceorge Apostolakis UCLA Boelter llall, Room 5532 Los Angeles, CA 90024 Vladimar Asmolov Ilead, Nuclear Safety Department 1.:V. Kurchatov Institute of Atomic.Enegry Moscow,- 123182- -U.S.S.R.

11. Banon' Exxon Production Research P.O. Box 2189 llouston, TX 77252-2189 Patrick W. Baranowsky USNRC AEOD/TPAB MS:

9112 Dist-1

Robert A. Bari Brookhaven National Laboratories Building 130 Upton, NY 11973 Richard J. Barrett USNRC-NRR/PD3 2 MS: 13 D1 William D. Beckner USNRC-NRR/PRAB MS: 10 E4 Dennis Bley Pickard, Lowe & Garrick 2260 University Drive Newport Beach, CA 92660 Gary Boyd Safety & Reliability Optirnization Services 9724 Kingston Pike, Suite 102 Knoxville, TN 37922 Robert J. Budnitz Future Resources Associates 734 Alameda Berkeley, CA 94707 Cary R. Burdick USNRC RES/RPSIB MS: NLS-314 Arthur J. Buslik USNRC-RES/PRAB MS: NLS - 372 Annick Carnino E1cetricito de France 32 Rue de Monceau 8EME = Paris, F5008 FRANCE

5. Chakraborty Radiation Protection Section

.Div. De La Securite Des Inst. Nuc. 5303 Wren 11ngen SWITV.ERLAND Dist-2

Michael Corradini University of Wisconsin 1500 Johnson Drive Madison, WI 53706 George Crane 1570 E. Ilobble Creek Dr. Springville, Utah 84663 Mark A. Cunningham USNRC RES/PRAB !!S : NiS - 372 G. Diederick Commonwealth Edison Co. LaSalle County Station RR1, Box 220 2601 North 21st Rd. Marsfelles, IL 61341 Mary T. Drouin Science Applications International Corporation 2109 Air Park Road S.E. Albuquerque, NM 87106 Adel A. El-Bassioni USNRC-NRR/lRAB MS: 10 E4 Robert Elliott USNRC NRR/PD3-2 MS: 13 D1 Farouk Eltavila USNRC-RES/AEB MS: NLN-344 John 11. Flack USNRC RES/SATB MS: NLS 324 Karl Fleming Pickard, Lowe & Garrick 2260 University Drive Newport Beach, CA 92660 James C. Glynn USNRC-RES/PRAB MS: NLS - 372 Dist-3

T. llammerich Commonwealth Edison Co. I LaSalle County Station RR1, Box 220 2601 North 21st Rd. Marsielles, IL 61341 Robert A. llasse USNRC RCN-III MS: RIII Sharif lieger UNM Chemical and Nuclear Engineering Department Farris Engineering Room 209 Albuquerque, NM 87131 P. M. lierttrich Federal Ministry for the Environment, Preservation of Nature and Reactor Safety liusacenstrasse 30 Postfach 120629 D 5300 Bonn 1 FEDERAL REPUBLIC OF GERMANY S. Ilirschberg Department of Nuclear Energy Division of Nuclear Safety International Atomic Energy Agency Wagramerstrasse 5, P,0. Box 100 A-1400 Vienna AUSTRIA ~ M. Dean llouston USNRC ACRS MS: P-315 Alejandro lluerta Bahena National Commission on Nuclear Safety and Safeguards (CNSNS) Insurgentes Sur N. 1776 C. P. 04230 Mexico, D. F. MEXICO Peter llumphreys US Atomic Energy Authority Wigshaw Lane, Culcheth Warrington, Cheshire UNITED KINGDOM, WA3 4NE Dist-4

W, lluntington Cormnonweal th Edison Co. laSalle County Station RR1, Box 220 2601 !Jorth 21st Rd. Harnielics, 11 61341 Brian Ives UNC 14uclear Industries P. O. Box 490 Richland, VA 99352 V1111am Kastenberg UCIA Boolter liall, Room 5532 Ims Angeles, CA 90024 George Klopp [10) Commonwealth Edison Company P.O. Box 767, Room 35W Chicago, IL 60690 Alan Kolaczkowski Science Applications Int. Corp. 2109 Air park Rd. SE Albuquerquo, NM 87106 Jim Kolanowski Commonwealth Edison Co. 35 1st National Vest Chicago, IL 60690 S. Kondo Department of Nuclear Engineering Facility of Engineering University of Tokyo 3-1,liongo 7 Bunkyo ku Tokyo JAPAN Jose A. Lantaron Cosejo de Suguridad Nuclear Sub. Analisis y Evaluaciones Junto Dorado, 11 28040 Madrid SPAIN Dist 5

Josette Larchier Boulanger Electricte de France Direction des Etudes Et Recherches 30, Rue de Conde 65006 Paris FRANCE Librarian NUMARC/USCEA 1776 I Street NW, Suite 400 Washington, DC 80006 Bo Livnang 1AEA A 1400 Swedish Nuclear Power Inspectorate P.O. Box 27106 S 102 52 Stockholm SWEDEN Peter Lohnberg Expresswork International, Inc. 1740 Technology Drive San Jose, CA 95110 Steven M. Long USNRC NRR/PRAB MS: 10 E4 llerbert Massin Commonwealth Edison Co. 35 1st National West Chicago, IL 60690 Andrew S. McClymont IT-Delian Corporation 1340 Saratoga Sunnyvale Rd. Suite 206 San Jose, CA 95129 Jose I. C.slvo Molins licad, Division of P.S. A. and lluman Factors Consejo Do Seguridad Nuclear Justo Dorado, 11 28040 Madrid SPAIN Joseph A. Murphy USNRC RES/DSR MS: NLS-007 Dist 6

Y.enneth G. Murphy, Jr. US Departtoent of Energy 19901 Germantown Rd. Germantown, MD 20545 Robert L. Palla, Jr. USNRC NRR/PRAB MS: 10 E4 Careth Parry NUS Corporation 910 Clopper Rd. Gaithersburg, MD 20878 G. Petrangeli ENEA Nuclear Energy ALT Disp Via V. Brancati, 48 00144 Rome ITALY Ing. Jose Antonio Bocorra Perez Comision Nacional De Seguridad Nuclear Y Salvaguardias Insurgentes Sur 1806 01030 Mexico, D. F. MEXICO William T. Pratt Brookhaven National Laboratory Building 130 Upton, NY 11973 William Raisin NUHARC 1726 M. St. NW Suite 904 Washington, DC 20036 D. M. Rasmuson 'JSNRC-RES/S AIB MS: NLS-372 M. K. Ravindra (10] EQE Inc. 2150 L istol St., Suite 350 Costa Mesa, CA 92626 John N. Ridgely USNRC-RES/SAIB MS: NLS-324 Dist 7

Richard C. Robinson Jr. USNRC RES/PRAB MS: NLS-37 2 Denwood F. Ross USNRC AEOD MS: 3701 Christopher P. Ryder (10) USNRC RES/PRAB MS: NLS 372 Takashi Sato Deputy Manager Nuclear Safety Engineering Section Reactor Design Engineering Dept. Nuclear Energy Group Toshiba Corporation Isogo Engineering Center 8, Shiusugita cho, Isogo-ku, Yokohama 235, JAPAN Martin Sattison Idaho National Engineering Lab. P. O. Box 1625 Idaho Falls, ID 83415 Louis M. Shotkin USNRC RES/RPSB MS: NLN 353 Desmond Stack Los Alamos Nationni Laboratory Croup Q-6, Mail Stop K556 Los Alamos, NM 87545 T. C. Theofanous University of California, S. B. Department of Chemical and Nuclear Engineering Santa Barbara, CA 93106 liarold VanderMolen USNRC-RES/PRAB MS: NLS-372 Dist-8

Magiel F. Versteeg Ministry of Socini Affairs and Employment P.O. Box 90804 2509 LV Den llaag Tile NETilERLANDS Edward Varman Stone & Webster Engineering Corp. P.O. Box 2325 Boston, MA 02107 Wolfgang Werner Gesellschaft Fur Reaktorsicherheit Forschungs6elande D 8046 Garching FEDERAL REPUBLIC OF GERMANY 3141 S. A. Landenberger [5] 3151 C. L. Esch 6321 T. A. Wheeler 6400 N. R. Ortiz 6410 D. A. Dahlgren 6411 D. D. Carlson 6411 D. M. Kunsman 6411 R. J. Breeding 6411 K. J. Maloney 6412 A, L. Camp 6412 S. L. Daniel 6412 S. E. Dingman 6412 B. D. Staple 6412 C. D. Wyas 6412 A. C. Payne, Jr. [25] 6412 D. W. Whitehead 6413 F. T, liarper 6413 T. D. Brown 6419 M. P. Bohn 8524 J. A. Wackerly f Dist-9

I sFC8th'/ 334 v s Nyct g ap a g GUL A106 4 CON *VittiON

• t*;**'..

'ts* i E m, 0,*,' C ?.'fd.*0,' 7,',.S*" w na BIBLIOGRAPHIC DATA SHEET NUREG/CR-4832 3 e, v.m r,,,,.,,, SAND 92-0537

sini e,; 5. p u Vol ' 7 Analysis of the LaSalle Unit 2 Nuclear Power Plant

Risk Methods Integration and Evaluation Program (RMIEP) 2 w t "I'0 '

5. ri:

e. External Event Scoping Quantification July 1992 ,e,Aouca.aseen A13A6 $ au1" Dos-e 3,,6 e.nsi M.K. Ravindra, H. Banon Technical 7 stooocoanto.

• UI,',?,"fi7h Mi.'N'DN - "'"E ** ^'
  • c*""'"'"'*"**"*'1*-""+->c'-+~~~-""-en NTS/ Structural Mechanics Asso. ates Under Contract to:

5160 Birch Street Sandia National Laboratories Newport Beach, CA 92660 Albuquerque, NM 87185

• *h*.?l2
• Q2, * * * * * ' ' "' ' * * * * * * * ' D * "' " * * ' ~
• * ~ ~ ~ ~ ~ ~ '~ ~~ * * * : b ~ o "~ v * " e ' s r ~
• nm, c e -- -.

Division of Safety Issue Resolution Office of Nuclear Regulutory Research US Nuclear Regulatory Commission Washington, DC 20555 us suecavtstaa, sons A55T m A;i w,.wa..c ne This report is a description of the scoping quantification study which selected the external events to be included in the Level III PRA of the LaSalle County Nuclear Generating Station Unit 2. The study was performed by NTS/ Structural Mechanics Associates (SMA) for Sandia National Laboratories as part of the Level I analysis being performed by the Risk Methods Integration and Evaluation Program (RMIEP). The methodology used is described in detail in a companion report, NUREG/CR-4839. In this report, we describe the process for selecting the external events, the screening analysis, and the detailed bounding calculations for thost events not eliminated in the screening analysis. As a result of this analysis, it ws concluded that only internal flooding, internal fire, and seismic events were potentially significant at LaSalle. Detailed analyses were performed for each of these and are reports in NUREG/CR-4832, Volumes 10, 9, and 8, respectively, u n 6

• w o a vs o i s c a e i u a u..,...,....,, r....

,4, Unlimited External Events PM Mehs Risk Methods Integration and Evaluation Program (RMIEP) hnlassified Screening Procedures ,,m_,.. Unclassified % ww u % s i t h !;( t hat 8 Dav m un

THIS DOCUMENT WAS PRINTED USING RECYCLED PAPER

l. UNITED STATES SPECIAL FOURrri CLASS RATE POSTAGE AND FEES PAID NUCLEAR REGULATORY COMMISSION ususc WASH!NGTON, D.C. 20555-0001 Pf RPJJT NO G 67 CFF6CIAL SUSINESS PENALTY FOR PRIVATE USE, $ 300 1 : 0 5 5' 13 G'i 31 1 1 2 " l P '" !;s n'o -enev r-UP L I"A TIri;c SVCS 'Tv FOI3 Tog _ pac _,ycrr D - 2 '. ! w a s a ! *,-3 T r P; CC '^5~= l 1 t i l l l l l l ._____ _}}