ML20212N445

Seismic Qualification of Equipment in Operating Nuclear Power Plants.Unresolved Safety Issue A-46

ML20212N445
Person / Time
Issue date:	02/28/1987
From:	Chang T Office of Nuclear Reactor Regulation
To:
References
REF-GTECI-A-46, REF-GTECI-SC, TASK-A-46, TASK-OR NUREG-1030, NUDOCS 8703130028
Download: ML20212N445 (183)

Text

.

. . . s NUREG-1030 xg Seismic Qualification of Equipment in Operating Nuclear Power Plants l.

t Unresolved Safety issue A-46 4 .

s ..

,/<

U.S. Nuclear Regulatory Commission Office of Nuclear Reactor Regulation T. Y.'Cnang i

! l' s** "'%q E

l f',.,

: 20

??R *R8Hil8 1030 R PDR

~ . _ _ _ . . . . _ . . _ _ _ _ . . . _ _ . . . . . _ ._-._.-___,_,_______,...,z. , . _ . . . , - . , - - _ _ . _ _ , _ _ _ - . . , , . . . . , - . . _ _ . . _ _ . . . , - _ _ . _ , _ . . _ _ _

. . _. - _ . =

. n -. _-

.e _- - - - . - . __

3 g ,. ..

NOTICE y Availability of Reference Materials Cited in NRC Publications

Most documents cited in NRC pub!! cations will be available from one of the following sources

e

. 9 l 1. The NRC Public Document Room,1717 H Street, N.W.

Washington, DC 20655 ,

s

! 4 2. - The Superintendent of Documents, U.S. Government Printing Office, Post Office Box 37082,

<. Washington, DC 20013 7082

.(. 4 3.9 The National Technical information Service, Springfield, VA 22161 Althou'gh the listing that follows represents the majority of documents cited in NRC publications, it is not intended to be exhaustive.

) Referenced documents available for inspection a' nd copying for a fee from the NRC Public Docu-ment Room include NRC correspondence and internal NRC memoranda; N RC Office of Inspection and Enforcement bulletins, circulars, information notices, inspection and investigation notices:

Licensee Event Reports; vendor reports and correspondence: Commission papers; and applicant and i

licensee documents and correspondence.

The following documents in the NUREG series are available for purchase from the GPO Sales Program: formal NRC staff and contractor reports, NRC-sponsored conference proceedings, and NRC booklets and brochures. Also available are Regulatory Guides, NRC regulations in the Code of Federal Reputations, and Nuclear Regulatory Comminion Issuances.

Documents available from the National Technical Information Service include NUREG series j '

reports and technical reports prepared by other federal agencies and reports prepared by the Atomic Energy Commission, forerunner agency to the Nuclear Regulatory Commission. t t

Documents available from public and special technical libraries include all open literature items, such as books, joumal and periodical articles, and transactions. Federal Register notices, federal and

state legislatiori, and congressional reports can usually be obtained from these libraries.

Documents such as theses, dissertations, foreign reports and translations, and non NRC conference ,

proceedings are available for purchase from the organization sponsoring the publication cited. l Single copies of NRC draft reports are available free, to the extent of supply, t:pon written request to the Division of Technical Information and Document Control, U.S. Nuclear Regulatory Com-mission, Washington, DC 70555. ,

Copies of industry codes and standards used in a substantive manner in the NRC regulatory process are maintained at the NRC Library, 7920 Norfolk Avenue, Bethesda, Maryland, and are available <

there for reference use by the public. Codes and standards are usually copyrighted and may be i purchased from the originating organization or, if they are American National Standards, from the American National Standards Institute,1430 Broadway, New York, NY 10018, 1

i e

h

l l NUREG-1030 l

Seismic Qualification of Equipment in 0 aerating Nuclear Power 'lants Unresolved Safety issue A-46 M:nuscript Completed: February 1987 DIte Published: February 1987 T. Y. Chang Division of Safety Review and Oversight Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, DC 206Ei6 l

l

ABSTRACT The margin of sdfety provided in existing nuclear power plant equipment to resist seismically induced loads and perfurm their intended safety functions may vary considerably, because of signifiant changes in design criteria and methods for the seismic qualification of equipment over the years. Therefore, the seismic qualification of equipment in operating plants must be reassessed to determine whether requdlification is necessary.

The objective of technical studies performed under the Task Action Plan A-46 was to establish on explicit set of guidelines and acceptance _ criteria to judge the ddequacy of equipment under seismic loading at all operating plants, in lieu of requiring qualification to the current criteria that are applied to new plants.

This report summarizes the work accomplished on USI A-46 by the Nuclear Regulatory Commission staff and its contrdctors, Idaho National Engineering Labordtory, Southwest Research Institute, Brookhaven National Laboratory, and Lawrence Livermore National Laboratory. In addition, the collection and review of seismic experience ddta and existing seismic test oata by the Seismic Qualification Utility Group and the Electric Power Research Institute, respectively, and the review and recommendations of a group of seismic consultants. .the Senior Seismic Review Advisory Panel, are presented. Staff dssessment of Work occomplished under USI A-46 leads to the conclusion that the use of seismic experience data provides the most reasonable alternative to current qualification criteria. Consideration of seismic qualification by use of experience data was a specific task in USI A-46. Several other A-46 tasks serve to support the use of an. experience data base.

The proposed resolution of USI A-46 which included NUREG 1030 draft for public comment and the attached Regulatory Analysis were issued for public coment in September 1985. The public comment period ended on November 15, 1985. All public comments were subsequently addressed dnd both the public comments and NRC staff resolution of comments are presented as Appendix D to this NUREG-1030. The Regulatory Analysis for proposed resolution of USI A-46 has also been revised to incorporate the resolution of r@lic comments, and is reissued as NUREG-1211.

l l

l The principal technical finding of USI A-46 is that seismic experience data,

! supplemented by existing seismic test data, applied in accordance with the guidelines developed, can be used to verify the seismic adequacy of mechanical and electrical equipment in operating nuclear plants. Explicit seismic qualification should be reoutred only if seismic experience data or existing test data on similar components cannot be shown to apply.

NUREG-1030 111

TABLE OF CONTENTS P6ge ABSTRACT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii ABBREVIATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x 1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1 1.1 Background ......................... 1-1 1.2 Description of A-46 Tasks . . . . . . . . . . . . . . . . . . 1-2 1.3 A-46 Technical Findings . . . . . . . . . . . . . . . . . . . 1-4 1.3.1 General Conclusions ................. 1-4 1.3.2 Scope of Seismic Adequacy Review . . . . . . . . . . . 1-5 1.3.3 Equipment Outside Applicability of Seismic Experience Dato Base . . . . . . . . . . . . . . . . . 1-5 2

SUMMARY

OF TECHNICAL WORK WHICH SUPPORTS USI A-46 TECHNICAL RESOLUTION . . . . . . . . . . . . . . . . . . . . . . . 2-1 2.1 Seismic Qualification of Equipment Using Seismic Experience Data Base .................... 2-1 2.1.1 Background . . . . . . . . . . . . . . . . . . . . . . 2-1 2.1.2 Summary of LLNL Report, " Correlation of Seismic Experience Data in Non-Nuclear Facilities With Seismic Equipment Qualification in Nuclear Plonts (A-46)". . . 2-3 2.1.3 Summary of EQE Report, " Pilot Program Report -

Program for the Development of an Alternative Approoch to Seismic Equipment Qualification" . . . . . 2-20 2.1.3.1 Methods used in the Pilot Program . . . . . . 2-21 2.1.3.2 Conclusion end NRC Stoff Comments . . . . . . 2-25 2.1.4 Summary of EQE Reports, " Seismic Experience Doto Bose--Data Base Tobles for Seven Types of Equipment,"

" Seismic Experience Doto Bose--Average Horizontal Data Bose Site Response Spectra," and "Investigotion of Equipment Performance in Foreign Earthquakes and the 1964 Alaska Earthquake" ............. 2-28 2.1.4.1 EQE Report, " Seismic Experience Data Bose--

Data Bose Tables for Seven Types of Equipment" ................. 2-30 2.1.4.2 EQE Report, " Seismic Experience Data Base--

Average Horizontal Data Base Site Response Spectra". . . . . . . . . . . . . . . . . . . 2-33 2.1.4.3 EQE Report, " Investigation of Equipment Performance in Foreign Earthquokes and the 1964 Alaska Eorthquake" . . . . . . . . . . . 2-42 2.1.4.3.1 Survey of U.S. Experts . . . . . . 2-54 NUREG-1030 v

TABLE OF CONTENTS (Continued)

P_ age 2.1.4.3.2 Literature _ Survey of Equipment Performance in the 1964 Alaska Earthquake . . . . . . . . . . . . 2-54 2.1.4.3.3 Literature Survey of Equipment Performance in Foreign Earthq0akes . . . . . . . . . . . . . 2-57 2.1.4.3.4 Conclusions and Staff Connents on -

Alaskan and Foreign Earthquakes. . 2-58 2.1.5 Summary of SSRAP Report, "Use of Past Earthquake i

Experience Data to Show Seismic Ruggedness of Certain Classes of. Equipment in Nuclear Power Plants". . . . . 2-58 2.1.5.1 Seismic Motion Bounds . . . . . . . . . . . . 2-60 2.1.5.2 Motor Control Centers . . . . . . . . . . . . 2-60 2.1.5.3 Low-Vol tage Switchgea r. . . . . . . . . . . . 2-62 2.1.5.4 Metal-Clad Switchgear _ . . . . . . . . . . . . 2-63 2.1.5.5 Unit Substation Transformers. . . . . . . . . 2-63 2.1.5.6 Motor-0perated Valves . . . . . . . . . . . . 2-63 1

2.1.5.7 Ai r-0perated Val ves . . . . . . . . . . . . . 2-64 2.1.5.8 Horizontal and Vertical Pumps . . . . . . . . 2-64 2.1.5.9 Conclusion and NRC Staff Comments . . ... . . 2-66 2.2 Development and Assessment of In-Situ Testing Methods to Assist in Qualification of Equipment . . . . . . . . . . . 2-69 2.2.1 Background . . . . . . . . . . . . . . . . . . . . . . 2-69 2.2.2 Summary of INEL Report, "The Use of In-Situ Procedures for Seismic Qualification of Equipment in Currently Operating Plants" .................. 2-70 2.2.2.1 Summary of Part A and Part B, " Preliminary Study of the Use of In-Situ Procedures for Seismic Equipment Qualification in Currently Operating Plants" and " Improved In-Situ Procedures and Analysis Methods". . ... . . . 2-70 1

2.2.2.2 Summary of Part C, " Guidance and Acceptance Criteria for Application of Combined In-Situ and Analysis Procedures" .......... 2-76 2.2.2.3 Summary of Part D, " Seismic Qualification

, Cost Estimating Task" . . . . . . . . . . . . 2-83 2.2.3 Staff Conclusions .................. 2-86 2.3 Development of Methods To Generate Generic Floor Res Spectra . . . . . . . . . . . . . . . . . . . . . . ponse ..... 2-86 4

NUREG-1030 vi

TABLE OF CONTENTS (Continued)

Page 2.3.1 Background . . ._. . . . . . . . . . . . . . . . . . . 2-86 2.3.2 Sumory of BNL Report, " Seismic and Dynamic

^

Qualification.of Safety-Related Electrical and Mechanical Equipment in Operating Nuclear Power Plants" ....................... 2-87 2.3.3 Staff Conclusions . . . . . . . . . . . . . . . . . .- 2-91 2.4 Seismic Qualification of Equipment Using Existing Test Data ... 2-93 2.4.1 Background . . . . . . . . . . . . . . . .' . . . . . . 2-93 2.4.2 Sumery of EPRI Report " Seismic Equipment Qualification Using Existing Test Data" .............. 2-93 2.4.3 Conclusion and NRC Staff Coments . . . . . . . . .. 2-96 3- REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1 APPENDIX A Summary of Technical Work Completed That Is Not Implemented in USI A-46 Resolution . . . . . . . . . . . A-1 APPENDIX B Perfomance of Power Facilities During the 1964 Alaska Earthquake ................... B-1 APPENDIX C Performance of Power and Industrial Facilities During Some Foreign Earthquakes . . . . . . . . . . . . . . . . C-1 APPENDIX D Consideration of Public Comments on USI A-46

-Proposed Resolution . . . . . . . . . . . . . . . . . . . D-1 LIST OF TABLES 2.1-1 Seismic Qualification Utility Group members . . . . . . . . . 2-2 2.1-2 Categories of possible seismic equipment qualification (EQ) requirements . . . . . . . . . . . . . . . . . . . . . . 2-4 2.1-3 Documents most important for seismic equipment qualification . . . . . . . . . . . . . . . . . . . . . . . . 2-7 2.1-4 Sumary of feasibility evoluation . . . . . . . . . . . . . . 2-8 2.1-5 Selected major earthquakes that have affected 2-21

~

power and industrial facilities . . . . . . . . . . . . . . .

2.1-6 Equipment selection for SQUG pilot progrom. . . . . . . . . . 2-23 2.1-7 Comparison of equipment data. . . . . . . . . . . . . . . . . 2-24 2.1-8 Sumary of data base plants and earthquakes . . . . . . . . . 2-25 2.1-9 Major conclusions of SQUG . . . . . . . . . . . . . . . . . . 2-26 2.1-10 Sumary: Motor control centers . . . . . . . . . . . . . . . 2-34 2.1-11 Motor control centers at the Sylmar Converter Station . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-35 2.1-12 Sumary: Motor-operated valves . . . . . . . . . . . . . . . 2-36 2.1-13 Motor-operated volves at near-field sites nea r Coal i nga . . . . . . . . . . . . . . . . . . . . . . . . 2-37 2.1-14 Motor-operated valves et far-field sites near Coalinga. . . . . . . . . . . . . . . . . . . . . . . . . . . 2-39 NUREG-1030 vii

. LIST OF TABLES (Continued)

Page.

2.1-15. Vertical pumps . . . . . . . . . . . . . . . . . . . . . . . 2-40  !

2.1 Vertical pumps at near-field sites near Coalinga.' . . . . . . 2-41

'2.1-17 Procedure to estimate data base site response spectra . . . . 2-50 2.' 2-1 C.1-1 Cost estimates. . . . . . . . . . . . . . . . . . . . . . . . 2-85 Damage to Enaluf Steam Plant. . . . . . . . . . . . . . . . . C ,

LIST OF FIGURES

~1.3-1 USI A-46 screening procedure . . . . . . . . . . . . . . . . 1-7 2.1-1

' Methods used in pilot study. . . . . . . . . . . . . . . . . 2-22 2.1-2 Distribution of motor control centers as a function of vintage, manufacturer, acceleration, and number of assemblies ....................... 2-42 2.1-3 Motor control centers surviving PGA > 0.18 g, data base of motor control centers plotte3 as a function 2.1-4 of width in sections-. . . . . . . . . . . . . . . . . . . . 2-43 Motor control centers surviving PGA 1 0.28 g . . . . . . . . 2-44 4 ~

2.1-5 Motor ~ control centers surviving PGA 10.45 g . . . . . . . . 2-45 2.1 Motor-operated valves surviving PGA 1 0.18 g, data base of motor-operated values plotted as a function of supporting

pipe diameter and operator height ............. 2-46

! '2.1-7 Motor-operated valves surviving PGA t 0.18 g, data base of motor-operated valves plotted as a function of supporting l pipe diameter and operator weight ............. 2-47

2.1-8 ' Vertical pumps surviving PGA > 0.18 g, data base of vertical

, pumps plotted as a function oY pump horsepower . . . . . . . 2-48 2.1-9 Vertical turbine pumps surviving PGA 1 0.18 g, data base of vertical turbine pumps plotted as a function of shaft l 1ength . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-49 2.1-10 Location of the San Fernando Valley data base sites and the ground motion records which are the basis for the estimated average peak horizontal ground accelerations . . . . . . . . 2-51 2.1-11 Response spectra for the ground motion record at Pacoima Dam, 5% damping, two horizontal components and their average. . . 2-52 2.1-12 Average horizontal response spectrum, Sylmar Converter Station, S% damping based on the ground motion record at Pacoima Dam, scaled by a factor of 0.50/1.25. . . . . . . . . 2-53 2.1-13 Questionnaire. . . . . . . . . . . . . . . . . . . . . . . . 2-55 2.1-14 Seismic motion bounding spectra. . . . . . . . . . . . . . . 2-62 2.1-15 Motor-operated valves for which type A spectrum is to be used .......................... 2-65 2.1-16 Motor-operated valves for which type C spectrum is to be used .......................... 2-66 2.1-17 Air-operated valves for which type A spectrum is 2.2-1 to be used . . . . . . . . . . . . . . . . . . . . . . . . . 2-67 Line graph definition of Region 1, Region 2 separation AD. . . . . . . . . . . . . . . ., and frequency

........ 2-78 2.2-2 Best estimate in structure response spectra and broadened response spectra . . . . . . . . . . . . . . . . . . . . . . 2-79 i

[ NUREG-1030 viii

LIST OF FIGURES (continued)

Page 2.2-3 Coupled building and support structure natural frequencies . 2-80 Comparison of envelopment ................. 2-81 2.2-4 .

2-84 2.2-5 USI A-46 screening procedure . . . . . . . . . . . . . . . .

2-88 2.3-1 Model 3 ........................... 2-89 2.3-2 Model 4 ..........................

2.3-3 Generic floor response spectra . . . . . . . . . . . . . . . 2-90 Generic peak responses at top, middle, and bottom levels . . 2-92 2.3-4 2.4-1 GenericEquipmentRuggednessSpectrum(GERS)for operability of motor valve operators . . . . . . . . . . . . 2-95 Conceptual approach to vibration correlation . . . . . . . . A-3 A.2-1 Comparison of actual with acceptable fragility surface . . . A-5 A.2-2 A-5 A.2-3 Basis for damage fragility ratio . . . . . . . . . . . . . .

A.2-4 Possible combinations of fragility function and qualification parameters . . . . . . . . . . . . . . . . . . A-6 A.3-1 Effect of aging on seismic capacity ............ A-9 l

l NUREG-1030 ix

ABBREVI4TIONS ACRS Advisory Committee on Reactor S.ifety ANSI American National Standards Institute BNL Brookhaven National Laboratory BWR ~ boiling water reactor CFR.  ; Code of Federal Rcgulations CQC complete quadratic combination DC .U.S. Department of Commerce EERI Earthquake Engineering Research Institute EPRI Electric-Power Research Institute EQ environmental qualification-EQE EQE Incorporated ERS experience response spectra FRF frequency response function GDC. General Design Criterion HVAC heating, ventilating,-and air conditioning IEEE Institute of Electrical and Electronics Engineers INEL Idaho National Engineering Laboratory (EG&G Idaho, Inc.)

JAERI Japan Atomic Energy Research Institute LLNL Lawrence Livermore National Laboratory LOCA' loss-of-coolant accident MCC- motor control center MMI modified Mercalli intensity MPF modal participation factor NAS National Academy of Sciences NCEE U.S. National Conference on Earthquake Engineering NCEL U.S. ' Naval Civil Engineering Laboratory NEMA National Electrical Manufacturers Association NRC Nuclear Regulatory Commission OBE operating basis earthquake PRA probabilistic risk assessment PSD power spectra density PWR pressurized water reactor RES Richardson Engineering Services, Inc.

RG Regulatory Guide RRS required response spectra SEP Systematic Evaluation Program SQUG Seismic Qualification Utility Group SRP Standard Review Plan SRSS square root of the sum of the squares i SSE safe shutdown earthquake '

SSRAP Senior Seismic Review Advisory Panel SWRI Southwest Research Institute TAP task action plan UBC Uniform Building Code USI unresolved safety issue 1

NUREG-1030 x l

i 1 INTRODUCTION

1.1 Background

General Design Criterion (GDC) 2 in Appendix A to Title 10 of the Code of Federal Regulations (CFR) Part 50 (10 CFR 50) states that structures, systems, and components important to safety in nuclear power plants shall be designed to withstand the effects of natural phenomena, such as earthquakes, without a loss of capability to perform their safety functions. Section III of Appendix B

-to 10 CFR 50 states that design control measures shall provide for verifying or checking the adequacy of design by the performance of a suitable testing program. It also requires that this program include suitable qualification testing under the most adverse design conditions. These requirements point to the need for seismic qualification of safety-related electrical and mechanical equipment to ensure structural integrity and functional capability during and after a seismic event. Current criteria and methods of compliance are in the Nuclear Regulatory Commission (NRC) Revision 2 to Standard Review Plan (SRP)

Section 3.10, " Seismic and Dynamic Qualification of Mechanical and Electrical Equipment" (NUREG-0800) (NRC, July 1981)* and NRC's Regulatory Guide (RG) 1.100,

" Seismic Qualification of Electric Equipment for Nuclear Power Plants." With some exceptions, RG 1.100 basically endorses the Insitute of Electric and Electronics Engineers (IEEE) Standard 344-1975, "IEEE Recommended Practices for Seismic Qualification of Class 1E Equipment for Nuclear Power Generating Stations."

On the basis of the requirements and recommendations of these criteria and methods, equipment is seismically qualified today by analysis and/or laboratory test. Analyses alone are acceptable only if the necessary functional capability of the equipment is ensured by its structural integrity. Otherwise, some testing is required using the required response spectra or required time histories for the seismic input motion to equipment. When equipment is tested, it is mounted on a shake table and subjected to certain types of excitation corresponding to a test response spectrum that envelopes the required response spectra. The equipment is tested in the operating condition. For equipment too large to fit on a shake table, a combined analysis and test procedure is used.

Since commercial nuclear power plants were first introduced, seismic qualifi-cation criteria have been changed to a significant degree. The analytical and experimental methods used to qualify equipment have also changed. Because of these changes the margins of safety provided in existing nuclear power plant equipment to resist seismically induced loads and perform their intended safety functions may vary considerably, and may not meet current seismic quali-fication criteria. Therefore, to ensure performance during and after a seismic event, seismic capability of equipment in operating plants must be reassessed.

• References in this report are cited parenthetically by author. See Section 3,

" References," for a complete citation.

NUREG-1030 1-1

It was also recognized that it may not be practical to qualify operating plant equipment using current seismic qualification criteria and methods because of (1) excessive plant down time, (2) difficulties in shipping irradiated equipment to a test laboratory, and (3) difficulties in acquiring identical vintage equipment for laboratory testing. In December 1980, the Nuclear Regulatory Commission designated " Seismic Qualification of Equipment in Operating Plants" as an unresolved safety issue (USI). The objective of USI A-46 is to develop alternative methods and acceptance criteria t% t can be used to assess the capability of mechanical and electrical equipment in operating nuclear power plants to perform the intended safety functions.

1.2 Description of A-46 Tasks A task action plan (TAP) was developed for USI A-46 in the spring of 1981.

Tasks for study were selected on the basis of their potential for providing reasonable alternatives to current requirements for seismic qualification. It was recognized that a utility always has the option to requalify equipment using procedures required for new plants. Only alternative procedures which provide some advantage over current requirements are likely to be used. In addition, any alternative procedure must be sufficiently rigorous to provide a level of safety comparable to that achieved by current requirements. A key element of the approach was to take advantage of experience gained by previous qualification tests and analysis, and experience with actual seismic events.

Tasks selected for study were:

(1) Identification of seismic-sensitive systems and equipment; (2) Assessment of adequacy of existing seismic qualification; (3) Development and assessment of in-situ testing methods to assist in quali-fication of equipment; (4) Seismic qualification of equipment using seismic experience data; (5) Development of methods to generate generic floor response spectra.

As work progressed it became increasingly apparent that Task 4, " Seismic Qualification of Equipment Using Seismic Experience Data" was the most likely alternative for assessing seismic capability, which is both economically attractive to the plant owners and acceptable from a public safety viewpoint.

Lawrence Livermore National Laboratory (LLNL), under contract to NRC, conducted a feasibility study (NRC, August 1983) which concluded that use of seismic experience data is feasible and can be as effective as current qualification i methods. This study is discussed in more detail later in Section 2.1.2. In addition, a utilities group, Seismic Qualification Utility Group (SQUG), in conjunction with its consultants EQE Incorporated, conducted a pilot program to independently demonstrate the feasibility of using seismic experience data.

Their report was issued by EQE in September 1982. A more detailed discussion of this effort is presented in Section 2.1.3.

In March 1983, SQUG proposed to NRC management the formation of a Senior Seismic Review Advisory Panel (SSRAP) to provide consulting services and expert opinion on the use of experience data. This idea was endorsed by NRC management and SSRAP was subsequently formed in June 1983.

NUREG-1030 1-2

4 In January 1985, SSRAP released its report which describes the SSRAP findings

-dnd recommenddtions for using seismic expelenCt data for non-nuclear plants to evaluate seismic adequacy of equipment in operating nuclear plants. Conclusions on the use of seismic experience data including caveats and exclusions were presented in the SSRAP report. These technical findings dre presented in i Section-1,3 of this report. More detailed description of this study and its conclusions can be found in Sections 2.1.4 and 2.1.5 of this report.

.In addition to earthqueke experience data, another type of experience data is

the large amount of information generated by the nuclear industry during seismic qualification-testing of equipment in the p4st several decades. In 1984, the Electric Power Research Institute (EPRI) initiated a project to -

collect and evaluate seismic test data. SQUG, SSRAP and the NRC staff worked

. very closely with EPRI since the beginning of this project to make sure that results could be considered in resolution of USl A-46. The specific goals of i

the EPRI test experience project are to establish 1) the classes of equipment for which sufficient qualification test data exist, 2) the generic seismic ruggedness level for each equipment class, and 3) the functionality of equipment required to. operate during selsmic motion (" operability" or

" functionality") and after seismic motion _("survivsbility"). The anticipated program accomplishments are 1) both operobility and survivability ruggedness

! levels for each identified equipment class, 2) inclusion rules and cautions j for each equipment class, and 3) field checklists for screening of equipment for class applicability. Final results from this EPRI generic qualification

. program are scheduled to be available for review by SQUG/SSRAP and the NRC l staff by December 1986. A more detailed aescription of this stuoy and its conclusions can be found in Sections 2.4.2 and 2.4.3 of this report.

i Tasks 3 and 5, " Development and Assessment of In-situ Testing Metnods To Assist

}

in Qualification of Equipment" and " Development of Methods To Generate Generic i.

Floor Response Spectra," play a supporting role. The emphasis on both tasks was focused to support use of an experience date base. Such emphasis is described in Sections 2.2 and 2.3 in this report.

Task 2, " Assessment of the Adequacy of Existing Seismic Qualification," was an effort to develop methods to evaluate the. acceptability of qualification by procedures used before current requirements were instituted. For instance, a method was developed to assess results of a single axis test in tems of expected multiple axis response. Although Southwest Research Institute devel-oped a proceoure for such assessment, it is of limited immediate value in its present form because of the need to either know the fragility level or estimate the fragility of the equipment and know the required response spectra. It may be useful in special cases. Task 2 is described in more detail in Appendix A.

Task 1, " Identification of Seismic Risk Sensitive Systems and Equipment," was an attempt to develop, on a generic basis, a minimum equipment list. The study, performed by Brookhaven National Laboratur model of 4 pressurized water reactor plant (PWR)y and a hybrid (BNL),

model was uf aconducted boiling on water reactor (BWR) plant using a seismic probabilistic risk assessment (PRA) model. The contribution to risk of major systems and components was calculated and ordered by risk importance. Although this study did provide some insight NUREG-1030 1-3 i - - _ .--_ ___- - - _-._. . . - - . _ - - - - . - - _ - - -

i

.into the risk importance of. systems and components _ and demonstrated the effect of varying equipment f ragility.on overall risk, it.is of limited usefulness

.in defining a generic equipment. list. ..The major conclusion of the BNL study was that BNL had demonstrated a methodology that could be applied on a plant-L

= specific' basis to develop a risk-based minimum equipment list. For plants for.

1 which an existing. seismic PRA model is'available, it may be' feasible to evaluate

~ the necessity to. qualify specific systems or components on the basis of- rit,k contribution. This task is described in more detail in Appendix A. i

'1.3 A-46 Technical Findings

! The principal technical finding of A-46'is -that seismic experience dato

, supplementea by existing seismic test data, applied in accordance with the

guidelines developed, can be-used to verify the seismic adequacy of mechanical and electrical equipment in operating nuclear plants. Explicit seismic 1 qualification should be required only if seism 1c experience data or existing

, test dota on similar components con not be shown to apply.

This finaing 1s based primarily on the steff's review of the work accomplished by SQUG, EPRI and SSRAP to develop a seismic experience data base and seismic test data base, and to develop guidance for its application. In' addition to

' endorsement of'the SSRAP conclusions, the staff has developed general guidance for extending the applicobility of seismic experience data to other classes of 4

components.

i 1.3.1-' General. Conclusions l The study completed by SQUG and SSRAP (SSRAP, January 1985) was limited to eight classes of equipment: motor control centers, low-voltage (480-V) switch-

! . gear, metal-clad -(2.4 to 4-kV) switchgear, unit substation tronsformers, motor-4 operated valves, air-operated valves, horizontal pumps, and vertical pumps.

General conclusions of the study on these eight classes of equipment were sum-marized by SSRAP as follows:

l (1) Equipment installed in nuclear power plants is generally similar to and i

at least as rugged as that installed in conventional power. plants.

(2) This equipment, when properly anchored, and with some reservations, has L an inherent seismic ruggedness and a demonstrated capability to withstand l significant seismic motion without structurol damage.

(3) For' this equipment, functionality after the strong shaking has ended has also been demonstrated, but the absence of relay chatter during strong t

-shaking has not been demonstrated.

(4) With several important caveats and exclusions, it is SSRAP's judgment that for excitations.below certain seismic motion bounds, it is unnecessary to perform. explicit seismic qualification of existing equipment in these eight classes -for operating nuclear power plants to demonstrate functionality after the strong shaking has ended.

(5) The existing data base reasonobly demonstrates the seismic ruggedness of this equipment up to these seismic motion bounds.

NUREG-1030 1-4

Specific caveats and exclusions for each equipment type are delineated .in 2.1.5.2 to 2.1.5.8. Additional observations on using the seismic experience data base are presented in 2.1.5.9. SSRAP believes that similar conclusions might apply to other classes of equipment. SQUG has initiated work to document the seismic adequacy of all equipment required for hot shutdown. In addition, EPRI started to collect and evaluate existing seismic test data on nuclear power plant equipment in 1984. These results will be reviewed end approved by SSRAP and the NRC staff before A-46 implementation.

1.3.2 Scope of Seismic Adequocy Review The staff concluded that it is unnecessary to verify the seismic adequacy of all plant equipment defined as.selsmic Cldss I in RG I.29 (HRC, September 1978).

This implies that only those systems, subsystems, and components requireo to bring the plant to a safe hot shutdown condition and to maintain it in that :

condition are important to assure safety during and after a Safe Shutdown Earthquake (SSE) event. The scope of the seismic verification, therefore, can be limited to the minimum equipment necessary to perform-the functions related to plant safe shutdown. This opproach is consistent with seismic reviews conducted by the Systematic Evaluation Program (SEP) anc with current staff thinking on simultaneous occurrence of a loss-of-coolant accident (LOCA) with a seismic event.

The initial intent of Task 1 of the A-46 TAP, " Identification of Seismic Risk Sensitive Systems and Equipment" was to develop a generic risk-ordered list of equipment. This effort however did not result in on equipment list that could be generally applied to operating plants. This effort is described in Appendix A.

The staff developed assumptions related to defining the equipment scope anc guidance on required plant functions. The assumptions which dictate the systems and equipment that will be needed ore *

(1) The seismic event does not cause a LOCA, a Steam-Line-Break Accident (SLBA), or a High-Energy-Line-Break (HELB).

(2) The LOCA, SLBA or HELB will not be postulated to occur simultaneously with or curing a seismic event. However the effects of transients that may result from ground shaking should be considered.

(3) Offsite power may be lost during and/or after a seismic event.

Given these basic assumptions, the scope of systems and equipment needed is less than currently required for new licenses.

1.3.3 Equipment Outside Applicability of Seismic Experience Data Base Not all equipment required following a seismic event was within the original scope of the experience data. An extension of the data base to cover additional classes of equipment and to extend the limits for the original eight equipment classes has been undertaken. In addition, other procedures are being developed. Steps taken for equipment not covered by the original data base are listed below ano shown graphically on Figure 1.3-1.

NUREG-1030 1-5

1) Extend experience data to include additional classes.

2) Find test data which are applicable to equipment.

Develop other evidence of seismic ruggedness.

Test prototype.

Perform enalysis and/or in-situ test to show seismic ruggedness or similarity with data base or test data (see Section 2.2).

(6) Perform simple modification to provide similarity with data base.

(7) Replace equipment with qualified equipment.

(8) Qualify to current requirements.

EPRI is in the process of collecting and evaluating seismic test data (see Section 2.4). In addition, SQUG is currently documentin of all equipment needed for hot shutdown (using (1), (2)g , (3)the seismic and adequacy possibly (5) dbove). The results of these studies will be reviewed by SSRAP and the NRC staff before A-46 implementation.

1-6

5 A

9 8

o Equipment Screened Out by Data Base (Satisfies 8 Classes Recommended by SSRAP)

Equipment Outside Limits of Data Base (Caveats and Bounding Spectra) or Not Belonging to the 8 1f Licensee Develops Classes Recommended by Equipment Plant-specific Compare List SSRAP in Data Bas Screened Out Seismic Adequacy

emin of Equipment Equipment List With Experience Means.. L From Functional Data Base Assured Requirement

• T u

7

• From Section 1.3.2.

• • 1. Extend experience data which are comparable to SSRAP guidance and caveats.

2. Find test data which are applicable to equipment.

3. Develop other evidence of seismic ruggedness.

4. Test prototype.

5. Perform analysis and/or in-situ test to show seismic ruggedness or similarity with data base or test data.

6. Simple modification to provide similarity with data base.

7. Replacement by qualified equipment.

3. Qualify to current requirement.

Figure 1.3-1 USI A-46 screening procedure

2

SUMMARY

OF TECHNICAL WORK WHICH SUPPORTS USI A-46 TECHNICAL RESOLUTION As mentioned in Section 1.2, of the five tasks selected for study in USI A-46, the most practical proved to be Task 4 " Seismic Qualification of Equipment Using Seismic Experience Data." Tasks 3 and 5 " Development end Assessment of In-Situ Testing Methods To Assist in Qualification of Equipment" and " Develop-ment of Methods To Generate Generic Floor Response Spectra," play a supporting rol e. _The emphesis on.these two tasks was focused to support use of on experi-ence data base. In the following paragraphs'these three tdsks are described.

The other two tdsks included in the task dction plan did not directly contribute to. resolution of USI A-46. They are discussea in Appendix A. In addition, the EPRI work on the seismic qualification of equipment using existing seismic test data, which supplements the seismic experience data in the resolution of USI A-46, is also described.

2.1 Seismic Qualification of Equipment Using Seismic Experience Data Base 2.1.1 Background It is well known that many non-nuclear power plants and industrial facilities containing equipment similar to that in nuclear power pidnts experienced major earthqudkes. It is also recognized that during the course of qualifying safety-related equipment for licensing nuclear plants in the last decade or so, numerous equipment items were tested on shake tables in laborotories for seismic capability. Therefore, there is a wealth of information regarding seismic experience that potentially con be utilized as an alternative to formal qualification of equipment in operdting plants. To use this information the data must be collected and organized, and guidelines and criteria must be developed. Two independent efforts to develop a seismic experience data base were initiateo. The SQUG (T4ble 2.1-1) conducted a pilot progrem, " Program for Development of an Alternative Approach to Seismic Equipment Qualification."

The pilot program was completed by the SQUG contractor, EQE Incorporated.

Results of this pilot program were recorded in a two-volume report issued in September 1982. A second effort was initiated by the NRC staff, with LLNL as the contractor. NRC published " Correlation of Seismic Experience Ddta in Non-Nuclear Facilities With Seismic Equipment Qualification in Nuclear Plants" i

in August 1983.

l-The results of both studies confirmed the feasibility of utilizing non-nuclear seismic experience data to verify seismic adequacy of equipment in operating nuclear power plants.

A group of seismic consultants, the Senior Seismic Review Advisory Panel (SSRAP) was formed by the SQUG in June 1983 to provide consulting services and expert opinion on the use of experience data. The stoff worked closely with SQUG and SSRAP to develop an acceptable approach to using seismic experience l data.

In Janudry 1985, SSRAP released its report which describes the SSRAP findings dnd recommendations (SSRAP, Jdnuary 1985). Conclusions on the use of seismic experience data including caveats and exclusions were presented in the report.

The study included motor control centers, low-voltage (480-V) switchgear, metal-clad (2.4 to 4-kV) switchgear, unit substotion transforiners, motor-operated NUREG-1030 2-1

i i

. Table 2.1-l' Seismic Qualification Utility Group members American Electric Power Co. Nebraska Public Power District

- Arkansas Power & Light Co. New York Power Authority Baltimore Gas & Electric Co. - Niagara Mohawk Power Corp.

Boston Edison Co. Northeast Utilities Service Co.-

. Carolina Power & Light Co. Northern States Power Co.

Central Electricity Generating Board Omaha Public Power District i Commonwealth. Edison Co. Philadelphia Electric Co.

Consolidated Edison Co. Public Service Electric & Gas Co.

Consumers Power Co. Rochester Gas & Electric Co. -

Detroit Edison Co. Sacramento Municipal Utility District Duke Power Co. Southern California Edison Co.

! ENEL ctn/NIRA' Tennessee Valley Authority Florida Power Corp. Toledo Edison Co.

Georgia Power Co.- Vermont Yankee Nuclear Power Corp.

GPU Nuclear Corp. Virginia Power Co.

? -INTERCOM /Electrobel Wisconsin Electric Power Co. '

Iowa Electric Light & Power Co. Wisconsin Public Service Corp.

Maine Yankee Atomic Power Co'. Yankee Atomic Electric Co.

valves, air-operated valves, horizontal pumps and vertical pumps.

General conclusions of SSRAP on these eight classes of equipment can be

summarized as follows

(1) Equipment installed in nuclear power plants is generally similar and at least as rugged as that installed in conventional power plants.

(2) This equipment, when properly anchored and with some reservations, has an t-inherent seismic ruggedness and has a demonstrated capability to withstand substantial seismic motion without structural damage.

! (3) Functionality after the strong shaking has ended has also been demonstrated, but the absence of relay chatter during _ strong shaking has not been demonstrated.

(4) With several important caveats and exclusions, it is the SSRAP judgment s

that below certain seismic motion bounds it is unnecessary to perform explicit seismic qualification of existing equipment in these eight classes for operating nuclear power plants to demonstrate functionality after the strong shaking has ended.

l (5) The existing data base reasonably demonstrates the seismic ruggedness of this equipment up to these seismic motion bounds.

j_ NUREG-1030 2-2

i l

Furthennore, SSRAP stated in their January 1985 report that their conclusions can be extended to other classes of equipment, but only with further study on experience data and test data on a class-by-class basis. Candidate classes for extension of equipment types are: heat exchangers, diesel generators, electrical motors, air compressors, fans, HVAC (heating, ventilating, and air conditioning) ducts, piping, and cable trays.

2.1.2 Sumary of LLNL Report, " Correlation of Seismic Experience Data in Non-Nuclear Facilities With Seismic Equipment Qualification in Nuclear Plants (A-46)"

The study was completed by LLNL and NRC issued a report (NUREG/CR-3017) (NRC, August 1983). This study was intended to answer the question: Is it feasible to use experience data on the performance of equipment in non-nuclear facili-ties during earthquakes in addressing issues concerning the seismic qualifica-tion of equipment in operating nuclear power plants located in the eastern United States?

The study shows that the answer to this question is affirmative. LLNL's general dpproach to the feasibility determination is based on the assumption that if experience data can be shown to be equivalent to current seismic equipment qualification requirements, then it is f easible to use experience data. The basic approach was to develop an overall summary statement evaluating seismic experience data and current requirements, as embodied in 12 different NRC Standard Review Plan sections, regulatory guides, and national standards. A .

conparison of the two summary statements provides the basis for the feasibility determination.

In LLNL's approach, 30 categories (issues) of possible seismic equipment quali-fication requirements are identified. That is, seismic equipment qualification standards might be (but presently are not) formulated in terms of requirements and criteria that address each of the 30 issues. Each of the 30 issues was ranked and a minimum set was identified. Table 2.1-2 lists the 30 issues and j

briefly describes each issue.

The 12 " current requirements" documents which are considered most important in terms of seismic equipment qualification for new plants are listed in Table 2.1-3.

LLNL's evaluation was performed by first reviewing the 12 current requirements in each of the 30 categories in Table 2.1-2, and then providing a comprehensive evaluation of these requirements. The evaluation was performed by ranking the current requirements in the 30 categories using the following numerical weights:

l

' Adequate - 3: This is the highest ranking. It is used to show that the current requirements are judged to adequately address the particular issue.

" Adequately" means that "the issue is addressed as well as is needed." It l

should not be interpreted as " ideally" or " perfectly" addressed or that it

" addresses the issue as perfectly as can be conceived."

• Moderately Acequate - 2: This is the next highest ranking.

Marginally Adequate - 1: This is a poor ranking. The issue is addressed, j

but not very satisfactorily.

l NUREG-1030 2-3

Table 2.1-2 Categories of possible seismic equipment qualification (EQ) requirements Category of Possible Seismic EQ Requirement Brief Description of Category Physical attributes

1. Sampling For equipment items qualified by testing, only a limited number of the items installed in a plant are tested.

2. Similarity The EQ for one item of equipment is sometimes extended to similar but differant items.

3. Mounting simulation The mounting and orientation used in the qualification of equipment may be different from those of installed equipment.

4. Peripheral attachments Peripheral items, such as electrical cables, small control piping, large pipe, and so forth, are often attached to the major item of equipment.

5. Dummy components Equipment is sometimes qualified by testing with a dummy item substituted for the actual item. For example, an electrical cabinet might be qualified with a dummy component substituted for a relay.

Seismic loads

6. Generic loads Generic loads (loads that envelop *all the required design loads for a parti-cular category of equipment) are sometimes defined.

7. Enveloping load assumption It is often assumed that if an item of equipment is qualified for load L , i then it is also qualified for load L 2.

where L 3 is greater than L 2-

8. Required design load Do the required design load and parameters adequately reflect EQ issues and concerns?

9. Margin Is there sufficient margin in the capacity of the equipment?

10. Tolerances Are tolerances specified for the required qualification load?

NUREG-1030 2-4

Table 2.1-2 Categories of possible seismic equipment qualification (EQ) requirements (continued)

Category of Possible Seismic EQ Requirement Brief Description of Category Seismic loads (continued)

11. Single vs. multiaxis testing How many independent test excitation axes are required?

12. Wave form A number of issues are related to the waveform of the test motion imparted to equipment.

13. Fatigue The fatigue requirements are considered here. An example is 5 OBE plus 1 SSE.

Strength / capacity

14. Fragility Do the EQ requirements address the strength of equipment, and if they do, how do they address it?

15. Failures Addresses failures that occur during qualification testing.

16. Functional requirements Addresses the functional performance of the equipment before, during, and after qualification testing.

17. Critical parameters Addresses the parameters that are most important to the survivability or func-tionality of equipment.

18. Degradation under test Has the qualification testing has been so severe that the capacity of the equipment to perform as required in the future can be questioned?

i,

19. Response Addresses the observed response of the

equipment during qualification testing.

20. Unexpected results Includes failures at unexpectedly low levels, unusual response patterns, and behavior that is inconsistent with predictions.

I Seismic and other loads

21. Load combination Relates to appropriate combinations of loads such as seismic, thermal, and pressure.

f L

l 22. Load sequencing A variant of load combination.

l l NUREG-1030 2-5 l

,~

4 r

> .' ^

LTable'2.1-2 Categories of possible-seismic equipment qualification (EQ) requirements (continued)

- Category of Possible

-Seismic.EQ Requirement

. Brief ~ Description of: Category Miscellaneous

- 23.

Errors ' Includes design qualification, ,

~ construction, mounting, and maintenance errors.

124. Maintenance' Includes consideration of how normal' .

(rather than erroneous) maintenance might affect the qualification. status of equipment.

25. . Mounting adequacy Addresses the adequacy of the equipment.

mounting.

26. Post earthquake Addresses the issue of assessing EQ subsequent to an earthquake.

. 27. Value/ impact. Addresses the benefit' of seismic EQ l in risk reduction (value) versus the cost of such requirements (impact).

l 28. EQ by analysis Addresses the issue of performing EQ by analysis rather than testing.

29. EQ by testing and analysis Addresses the issue of performing EQ "

by a combination of testing and j- analysis, f

30. In-situ. testing Addresses the. issue of the possible role of in-situ testing in EQ.

t

• Inadequate - 0: This is the worst ranking. The issue is either not addressed j at all or, if it is addressed, it is addressed poorly.

l

!

• Ranking not required: This ranking usually occurs when an issue that does I j not have to be addressed is included for completeness.

Next, the use of experience data was also evaluated for each of the 30 categories.

l The same ranking as above was used. These rankings were then weighted according l to importance, and the two sums (current requirements and experience data) were

! compared to arrive at a feasibility judgment. The result of the evaluation is

- summarized in Table 2.1-4. Table 2.1-4 shows that when the current requirements j in existing NRC and national standards were evaluated against the common set of l 30 issues, they were estimated to score 91 out of 156 overall, or about 60%.

Experience data were estimated to score 97 out of 156 overall, also about 60%.

l The fact that the current requirements and experience data score about the same NUREG-1030 2-6 l

1

, , , , , , + ~ - , e-- ,, --,,n..-,,,,w- -,---,m--,-,,a-,.,n,,,,w w,---_,,,-,-,,,,---. _ - - , -.,,-n-,,,,

Table 2.1-3 Documents most important for seismic equipment qualification U.S. Nuclear Regulatory Commission, Standard Review Plan, Section 3.10,

" Seismic and Dynamic Qualification of Mechanical and Electrical Equipment,"

NUREG-0800, Rev. 2, July 1981.

U.S. Nuclear Regulatory Commission, Regulatory Guides:

1.40 " Qualification Tests of Continuous-Duty Motors Installed Inside the Containment of Water-Cooled Nuclear Power Plants," March 16, 1973.

1.73 " Qualification Tests of Electric Valve Operators Installed Inside the Containment of Nuclear Power Plans," January 1974.

1.100 " Seismic Qualification of Electric Equipment for Nuclear Power Plants," Rev. 1, August 1977.

1.148 " Functional Specification for Active Valve Assemblies in Systems Important to Safety in Nuclear Power Plants," March 1981.

- IEEE Standard for Type Tests of Continuous Duty Class 1E Motors for Nuclear Power Generating Stations, ANSI N41.9-1976, IEEE Std. 334-1974.

IEEE Recommended Practices for Seismic Qualification of Class 1E Equipment for Nuclear Power Generating Stations, ANSI /IEEE Std. 344-1975.

- IEEE Standard for Qualification of Safety-Related Valve Actuators, IEEE Std. 382-1980.

- IEEE Standard Seismic Testing of Relays, IEEE Std. 501-1978.

I

- IEEE Standard for Qualifying Class 1E Motor Control Centers for Nuclear Power Generating Stations, IEEE Std. 649-1980.

- Self-Operated and Power-Operated Safety-Related Valves Functional i Specification Standard, ANSI N278.1-1975.

- Functional Qualification Requirements for Power-0perated Active Valve Assemblies for Nuclear Power Plants, ANSI B16.41, Draft 3, Rev. II, l

June 1981.

l (60%), led to LLNL's conclusion that it was feasible to use experience data on seismic equipment qualification issues.

Besides the feasibility study, LLNL's report also addressed recommended guide-

! lines for the use of experience data.

For all the categories considered to be the most important (those given an im-portance ranking of 3), guidelines were developed. Categories considered are:

NUREG-1030 2-7

3C C:

23 Table 2.1-4a Summary of feasibility evaluation: Sources 1-7 (as Ifsted in Table 2.1-3)*<

C1 C$ Category SRP 3.10 RG 1.40 RG 1.73 RG 1.100 RG 1.148 IEEE Std. IEEE Std.

cd C3 334-1974 344-1975

1. Sampling Sampling is A " prototype A

• prototype A " prototype ImpIfcit acceptance acceptable. unit" to be unft" to be unit" to be . of. sampling, at Sample size is tested under tested under tested under - least for cases not defined. most adverse most adverse most adverse where fragility _

design design design testing is conditions, conditions. conditions. performed.

2. Similarity Extension of EQ by test to sfullar.

equipment is allowed using a' combination of .

test and analysis.

, 3. Mounting The fixture design The equfpment shall simulation should simulate be mounted in a.

the actual service manner that simu-

, mounting. lates the intended '

service mounting.

[o 4. Peripheral Major peripheral The effects of i attachments attachments are . peripheral attach-addressed. ments must be considered.

5. Dummy Dummy specimens Use of dummy spect-j components are allowed to mens is allowed.

! simulate the mass effects and dynamic coupling to the supports.

6. Generic loads

7. Enveloping Not clear whether- The assumption The assumption is load . the assumption is is made. made, assumption made.

8. Required design load

9. Margin Margins are Margins are . 105 margins are required but required but' specified for the not specified. not specified. response spectrum

, at the mounting

, point of the equipment.

"A blank indicates no requirement was found.

.- a

Table 2.1-42 . Summary of fsasibility evaluation: Sources 1-7 (as'1fsted in Table 2.1-3) ' (continued) f$ Category. SRP 3.10 RG 1.40 RG 1.73 RG 1.100 RG 1.148 IEEE Std. IEEE Std.

Cf 334-1974 344-1975 ha ca 10. Tolerances LJ 3

11. Single vs. Two simultaneous Multfaxis testing is multfaxis axes of input are suggested. Single-

, testing 9enera11y required, axis testing is General procedures allowed if conser-are specified. wative, or if the responses in the axes are Independent..

12. Wave form The characteristics Requirements for of the required simulating earth-input should be quake are given.

j specified by Specific require-i response spectrum ments for proof

! or time history testing are methods. speelffed.

i 13. Fatigue Structural Performance must The requirement is 4

integrity and must be assured five OBEs plus an operability must during and after SSE.

be assumed under an SSE preceded Sf an SSE preceded by by several 08Es.

uD several 08Es.

14. Fragility Fragf11ty testing recommended, but not required, for equipment to be used in a number

of applications.

15. Failures

16. Functional Operationality General, indf- Reference is made Seismic input is For devices (relays, require- should be veriffed rect references to ANSI N278.1- assumed to occur motors, sensors),

ments during and/or to functionality 1975. with motor stand- ft is assumed that after testing. are given. still, starting, the seismic input running, or . can be imposed coasting down. whfie simulating normal operative and sensing per-formance.

17. Critical Some parameters are parameters suggested as pos-

)

sibly critical and are recommended for

] (dentification.

1

)

l 1

I i

i

' !! Table 2.1-4a Summary of feasibility evaluation: Sources 1-7 (as listed in Table 2.1-3) (continued) -

30 rn G4

/a Category SRP 3.10 RG 1.40 RG 1.73 RG 1.100 RG 1.148 IEEE Std. IEEE Std.

c) 334-1974 344-1975 00

18. Degradation upon completion under test of the test, the motor shall be dismantled and inspected.

19. Response Monitoring is

. required, but specific require-ments are not given.

20. Unexpected Analysis might be results used to explain unexpected behavior..

during a test.

21. Load It is not clear Normal operating combina- what combinations -loads which adversely-l tion are acceptable. af fect function must be combined with

$f seismic loads.

He ca 22. Load Load sequencing Load sequencing .

sequencing is to follow is indirectly IEEE Std. 323-1974. addressed.

23. Errors

24. Maintenance

25. Mounting Requirements The mounting adequacy on mounting ade- method shall be quacy are given the same as that with respect to recommended for testing and/or active service, analysis assess-ments.

26. Post earthquake

27. Value/

impact

. ~ ,_.- - . . . _

Table 2.1-4a Summary of feasibility evaluation: Sources 1-7 (as listed in Table 2.1-3) (continued)'

2 C

=

@ Category SRP 3.10 RG 1.40 RG 1.73 RC 1.100 RG 1.148 IEEE Std. IEEE Std.'

s 334-1974 344-1975 w

i O O

28. EQ by EQ by testing EQ by other EQ by analysis'is analysis is preferred. testing is not generally implicitly recosaended with-accepted by IEEE out test except s Std. 344-1975. where structural integrity alone can

, ensure equipan,nt function.

29. EQ by ',.

EQ by combined ~ -

2

- testing and '

testing and'enalysis Analysis is acceptable, but ,

s only vaguely defined.

30. In-situ In-situ. testing In-sito testits can testing i~s not required. .. be a part of EQ by but it is allowed. combined testing

- and analysis, n ~v

I .

H W t .

.. z >

l g .

v- s

~

, 'S ,

b 1

e P

N

• '~ [

9 y (

1  % , 4, s

J -

' , n_

6, s

• 9

~...

, .w-

n. ~

1 -k:.  ; ,h ci N a y .c g:.s

~

~

g '

2 Table 2.1-4b Summary of feasibility evaluation:.' dadeces 8-12 (as listed in Table 2.1-3) and other' data' C

c

+

Score on

$ e <

' current Score on

' ANSI B16.41- experience' y Category

- IEEE 5td..

382-1980 IEEE Std.

501-1978 IEEE Std.

6 0-1980 ANSI N278.1-1975 1981

. r%ulre ;

ment ~

~Fxperience data.. data w

O .,

y

' N-e '

1. Samp1fng .A procedure sug- A minimum of 3 At least one de- Testing of at 3 , .Several units are,-6 gested for se- specimens is vice must be most one sample commonly excited '

lecting the test required. -tested, but not is acceptable. at once by an, ~ g-units is given one motor control earthquake. .

4*

in App. A. center. . Therefore experf ; . s ence data are, ,. -i potentially rich. , L, in sampling. -

2. Stellarity Siellarity is ad- Extension of General guide-. Guidelines are .- 3 ." Equipment among , 6 dressed in terms qualified relays lines are given given to extend non-nuclear fact .. .'

of generic groups to relays not '.to extead the qualification of lities is usually, ,

of valve actua- tested is qualification of- .value assemb11es quite similar.

tors from which allowed. motor control to siellar units. A casual compart-son also indi ,

test units are centers to other I drawn. units. cates that the equipment is also quite siellar to that BJ in nuclear facilities, w'

to 4 Experience data 6

3. Nounting The valve actua- The relay must The motor control simulation tor is required be mounted as it center must be. reflect the true to be mounted to normally would mounted as it . mounting condi-would be in a tions. Therefore, the shaker table be in service.

as it would be plant, mounting is not an .

attached to the issue for'such valve. data.

Anticipated addi- Electrical, 4- The credibility 6

4. Peripheral Electrical, hy-attachments draulic, or tional weight and hydraulic, or :cf effects from pneumatic connec- external connec- pneumatic con- peripheral tions must be tions shall be. nections shall attachments is attached. simulatei. be required. not an issue for : I experience data. l

5. Dummy 4 Dummy specimens do 6-components not represent an .

issue for experi-ence data.

6. Generic Generic loads Fragility test- Generic load EQ Not .Not loads for valve actu- Ing is required techniques are required.. required.-

ators are estab- for reliys; allowed for 11shed for most therefore, groups of equip-plants. generic loads ment.

are essentially required.

_..T:,1

~

4

~.a k

'c Table 2.1-4b Summary of f"easibility evaluation: , . .-

2-Sources 8-12 (as listed in Table 2.1-3) and other data (continued)

E' r'1

!\ " Score on current' . Score on IEEE Std. 'IEEE Std.'~wS. IEEE Std. ANSI N278.1- l ANSI 816.41- require *

.p Category . 382-1980 501-1978.

, experience 649-1980 1975 -1981 ment Experience data. ' data

j. o

$ 7. Enveloping load Enveloping is probably estab-2 Experience data 2 could provide en-assumption lished through . Indication of generic loads. equipment per-fstmance'at loads-

-..that envelope -

required loads for EQ.

8. Required The required -

'6' Although loads ex- 3 design load design load may X perienced are real- a, be deficient. istic' , the ade- ,. " 51 quate reflection ";i -w

?vt of such loads to areas of concern.

.in EQ of nuclear plant equipment .

may be laciting.

9. Margin Margins are Fragjlltytest- Margins are  :._- 3 Some evaluations - 6  %

ro included in the ing 1(LCludes 'specified in indicate that

'A, 8

generic loads. 'the concept of Table 1.. some non-nuclear - , '%.i:

U margins. facilities have experienced

~

seismic loadings-in excess of design' loadings

.in nuclear facil-ities.

10. Tolerances Tolerances are Not Not specifled for required. required.

instrumentation.

11. Single vs. 81 axial testing Triaxial testing 6 ~ Experience data .6 multtaxis required. is desired, but generally consists blanfal testing of three-is acceptable. dimensional.

excitation.

12. Wave form Requirements are Two multi- 9 Inputs in expert-- 6 consistent with frequency, stan- ence data can be.

IEEE Std.344- dard response either narrow ban-1974. spectra are ded if the equip-specified for ment.is mounted qualification ,

on a structure of relays.' or piping sys-tem, or broad' banded if mounted y on the foundation.

~

t y -J .

. - - . _ _m__- _

z hh Table 2.1-4b Summary of feasibility evaluation: Sources 8-12 (as listed in Table 2.1-3) and other data (continued) rn C4 y, Score on c3 current Score on oo IEEE Std. IEEE Std. IEEE Std. ANSI N278.1- ANSI B16.41- require- ,

experience L C3 Category 382-1980 501-1978 649-1980 1975 -1981 ment Experience data . data

13. Fatigue CBE and SSE Five OBE plus an 3 Low-cycle fatigue 3' testing are SSE testing are .may be revealed required. Each required. Mini- by experience test must be mum duration is data.

15-s minimum. 15-s per test

14. Fragility Fragility test- 3 Present indications 3 ing is required from a limited re-for relays. . view of experience -

data suggest that.

few or no failures of equipment will be observed.

15. Failures Determination of 3 Failure informa- 3 what constitutes tion may be failure for re- limited.

ha lays is given.

t bd 45 16. Functional valve actuators Relays must be Motor control Valve assemblies Functional re- 9 Experience data on 6 require- must be func- tested in the center opera- must be operable quirements are the functionality ments tional before, transition from tional capa- during and after given for valve of equipment may during, and nonoperating . must be demon- the test. assemblies, be relatively after testing. to operating strated. scarce, i condition.

17. Critical 1 Since few or no- 0 I

parameters failures have been observed, it is' unilkely thag ex-perience data will reveal critical parameters. The most important fattures observed -l have been fail-ures of mountings

or attachments, l

18. Degradation Inspection of 0 Degradation is 3 i under test valve assen- generally not an bites shall be issue for exper*

Performed be- tence data.

fore and after testing.

..- , s _~_

3D Table 2.1-4b Summary of feasibility evaluation: Sources 8-12 (as . listed in Table 2.1-3) and other data (continued) rn Cf Score on pa ' current ' Score on c3 IEEE Std. IEEE Std. IEEE Std. ANSI N278.1- ANSI 816.41- require- emperience . -

. [j Category 382-1980 501-1978 649-1980 1975 1981 ment Experience data data

19. Response Not Not required. required.

20. Unexpected Not Not results required. required. .

21. Load Seismic tcsting 6 Normal operating 4 combination of relays can be loads are expected performed under to be present.

prevailing ambi- already when an ent conditions of earthquake occurs.

the test labor-atory.

22 toad A standard load Sequencing of A sequence of 6 ~ Equipment in 4 sequencing sequence is preaging and - testing is spe- operating plants required. seismic testing cified for valve can be expected is specified. assemblies. to have normal p3 environments, e

transients, and bn in-situ vibration.

23. Errors O Equipment in 2 plants presumably has been in-stalled with a more or less typical set of errors.

24. Maintenance Maintenance to Maintenance can Modifications If maintenance O Experience data 2 be performed be performed during testing or adjustments should be valuable-during the test after a given shall be evalu- are required in assessing if, must be fragility test. ated to deter- during testing, and how maintenace specified. mine their acceptance of affects seismic effect on the the test must performance.

EQ. be evaluated.

25. Mounting The valve actu- Recommended Mounting must The valve assen- 9 Failure of mount- 6 adequacy ator must be mounting hard- be by welding bly must be sup- ings appears to mounted to the ware must be or bolting for ported as re- be the single shake table as used. seismic. testing. quired to permit most important it would be testing-in accor- failure: there-mounted to a dance with the fore, experience valve. standard. data can be ex-pected to pro-

! vide useful in--

formation re-garding mount-ing adequacy.

9

3D Table 2.1-4b Summary of feasibility evaluation: Sources 8-12 (as listed in Table 2.1-3) and other data' (continued) rn C1 Score on ph current Score on c) IEEE Std. IEEE Std. IEEE Std. ANSI N278.1- ANSI 816.41- require- experience

[j Category 382-1980 501-1978 649-198F 1975 1981 ment Experience data data

26. Post 2 Equipment exposed -1 earthquake to an earthquake is subsequently subjected to nor .- j mal operation, .I transients, etc.

l Therefore experi-l ence data should-be useful for j

assessing post-earthquake be .

havior, but only partially.

27. Value/ impact Not Not required. ' required.

28. EQ by EQ by analysis EQ by analysis EQ by analysis EQ by analysis 1 Experience data 1 analysis is allowed to is assumed to is allowed. ,is allowed. are at least as Sf extend qualifi- be possible, amenable to .

Ha cation of a ge- analysis as EQ Ch neric group to is through d. I a specific appli- ordinary means. 1 ication, j

29. EQ by EQ by combina- General renuire- bQbycombina- EQ by combina- 1 The use of com- 1 testing and tion of test ments are tion of test and tion of test bined test and analysis and analysis is given. analysis is and analysis analysis in allowed to ex- allowed. is allowed. experience data tend EQ of a would have to generic group be defined in to specific more detail to applications. make a good evaluation of' its value.

30. In-situ 1 It should be- 3 testing possible to develop accept-able in-situ techniques for nonnuclear .

facilities and -

nuclear facil-ities alike.

Total 91 1 97

-(1) Sampling

-(2) Similarity (3) Required design load (4) Margin

.(5) Single vs. multi-axis testing

.(6) Wave form.

(7)' Fragility (8) Failures (9) Functional requirements (10) Mounting adequacy The guidelines, as taken directly from the LLNL report, are co,ibined under the five headings as follows:

Sampling (1) Experience data should be gathered on all non-nuclear facilities that have experienced (a) a significant earthquake, or (b) failures of any kind or either temporary or permanent loss of functional capability. LLNL antici-pates that 10 to 50 facilities will fall into this class. If fewer than ten facilities, three significant earthquakes, or all facilities that have experienced some kind of mechanical, structural, or functional failure are included in the data base, LLNL does not recommend that the NRC accept experience data as fully as it has otherwise recommended.

(2) The numbers of each type and size of affected equipment should be obtained for each facility in (1). If fewer than three items of each type and size of interest are found, then a justification must be provided to extend the experience data.

I Similarity (3) The issue of the similarity of equipment in non-nuclear facilities to equipment in nuclear facilities must be addressed. However, exact simi-larity need not be established.

Rather, what is required is reasonable assurance that the equipment in non-nuclear facilities (a) is of the same type and basic design, and (b) was manufactured by the same manufacturers in the same period as the j

equipment of interest in nuclear facilities.

Required Design Load, Wave Form, and Dimensionality (4) The approximate location of each item of equipment in non-nuclear facil-ities must be established in order to obtain a " rough" idea of the type of earthquake motion it experienced. " Rough" means that dynamic modeling or analysis is not required. Two categories are suggested:

l

! (a) Dimensionality. Was the earthquake motion affecting the equipment j predominantly one , two , or three-dimensional in nature?

l (b) Wave form. Was the earthquake motion affecting the equipment:

NUREG-1030 2-17 l

w --. -e e+w- ,- --y _ _ .u

random like an earthquake (as for equipment in the foundation or free-field) random because of superposition of a number of narrow-band pass motions, each with a different center frequency (as for hori-zontal motions on equipment in the lower elevations of a structure) sinusoidally random, that is, essentially a single-band pass motion (as for horizontal motions on equipment in the higher elevations of a structure).

Criteria are difficult to establish in this area, because in some respects they are dependent on the motions expected for the equipment of interest in nuclear facilities. However, if the experience data indicate signifi-cant two- or three-dimensionality of motion and sinusoidally random motion with a mix of center frequencies, then the experience data are acceptable.

Margin (5) The facilities in (1) should be selected in order of decreasing severity (for example, peak acceleration) of earthquake, that is, the most severe earthquake first. A reasonable assurance of margin for plants in the eastern U.S. is provided if the experience data are obtained from earth-quakes with a peak acceleration greater than the SSE peak acceleration for the nuclear plants of interest and the duration is greater than 10 seconds.

However, inevitably questions will arise about the most detailed aspects of the motion affecting the equipment in non-nuclear facilities (for example, in-structure response spectra) and the relation to similar motions in nuclear facilities.

The staff believes that the above requirement for acceleration and dura-tion provides reasonable assurance on the issue of margin, and nothing further is recommended. If, however, the NRC decides that more needs to be done on the margin issue, three steps are recommended:

(a) As a first step, realistic analyses can be performed on the non-nuclear facilities. For example, a comparison of realistic non-nuclear and nuclear design in-structure spectra, as in the EQ report of September 1982 may establish the required confidence in margin.

(b) If (a) is not chosen or if it does not indicate margin is present, then the following may be an acceptable alternative. Realistic, best-estimate analyses, with uncertainties explicitly characterized, as in the LLNL report of July 1981, should be performed on both the non-nuclear (for the earthquake that occurred) and nuclear (for design earthquakes) facilities. The median of the two results should be used as a measure of whether or not adequate margin exists. For example, median in-structure spectra from the two analyses can be compared.

(c) As part of either (a) or (b) above, margin is assured if, for example, margin exists at the frequencies of interest but not at some other frequencies in the spectra.

NUREG-1030 2-18

Fragility, Failures, Functional Requirements, and Mounting Adequacy (6) A vigorous effort to seek out failures or incipient failures in experience data is required. In addition to mechanical or structural distress or failure, incipient or actual functional failures should also be sought.

This effort includes examination of plant system logs and interviews with plant operators or other personnel present during the earthquake.

The six guidelines above are concerned with experience data obtained from non-nuclear facilities.

The next three guidelines are concerned with actions recommended by LLNL for nuclear facilities.

Functional or Other Failures (7) Nuclear plant equipment should be examined very closely for any and all failures revealed in (6). For example, experience data suggest that mounting failure is the single most important cause of failure of equipment.

All nuclear equipment of interest should be examined for adequacy of mounting or attachment.

(8) The NRC should develop a detailed and definitive check list to aid in a

" walk-down" of equipment of interest in nuclear plants. Such a walk-down should then be performed in each operating nuclear power plant where there is concern about the seismic adequacy of equipment. The items and proce-dures in the checklist should be drawn from three sources:

(a) Information gathered from the collection of experience data; (b) Information gathered from laboratories experienced in seismic equip-ment qualification testing; (c) Recognized experts who have performed walk-downs in the past.

(9) A limited amount of shake table testing should be performed on equipment obtained from operating nuclear power plants to confirm the perceived strength of equipment. This testing should satisfy the following:

(a) The test objective is to obtain the " capacity" of each equipment item tested. Capacity includes:

incipient or actual " structural" failure degradation of or loss of function

• identification of failure modes and key parameters related to failure or capacity anomalous behavior An example of such testing can be found in the JAERI report of i August 1979.

i t

NUREG-1030 2-19

(b) The equipment should be tested while functioning or in such a manner that capability.of-function is assured.

(c) The equipment need not be artificially aged or subjected to loads or environments other than seismic.

(d) The equipment should be tested as is. That is, it should not-be modified, adjusted, disassembled and tested separately,.etc., after it is selected for removal or removed from the plant.

(e) The testing should be limited in the number of categories of equip-ment tested, but comprehensive in addressing each operating plant and category of equipment. For example, one item of each category of equipment should be obtained from each category of equipment, and the same test program executed for each.

(f) The number of categories of equipment should be limited. The selec-tion of the category of equipment to be tested should be based on importance, estimated vulnerability, (that is, choose a category that is believed to be relatively weak rather than strong) and diversity of equipment type. For example, these objectives may be satisfied if the testing is limited to:

125-V vital bus (electrical equipment)

• motor-operated valves (mechanical equipment)

(g) The above requirements may lead to testing on the order of 100 items of equipment, depending on the number of plants involved. As an al-ternative to 100 tests on only 2 categories of equipment, as outlined above, a minimum of 5 tests on 20 or so categories would be acceptable.

2.1. 3 Summary of EQE Report, " Pilot Program Report - Program for the Develop-ment of an Alternative Approach to Seismic Equipment Qualification" Many non-nuclear power plants and industrial facilities containing equipment similar to that found in nuclear power plants have experienced major earth-quakes. A sample of this experience is shown in Table 2.1-5. The SQUG with help from EQE, initiated a pilot program to evaluate the potential for using experience data as the basis for qualification. The results of this pilot pro-gram were documented in this EQE report (EQE, September 1982). Stated goals of the pilot program were:

(1) To develop a historical data base on the performance of equipment in power plants during and after strong earthquakes.

(2) To show that much of the equipment in those plants is similar to equip-ment found in nuclear power plants.

(3) To determine whether data from actual earthquakes are sufficient to  ;

conclude that seismic qualification by conventional methods is not necessary for certain classes of equipment.

(4) To develop a methodology for using earthquake data to evaluate the necessity for seismic qualification of specific items of equipment by conventional methods.

NUREG-1030 2-20

-Table 2.1-5 Selected major earthquakes that have affected power and industrial facilities Recorded Estimated Peak Number of Power Approximate Ground Ground Plant Richter Accelera- Motion Units Earthquake Location Year Magnitude tion (g) Records Affected

1. Eureka, Ca. 1980 7.0 0.15+ 8 '3

2. Imperial Valley, CA 1979 6.6 0.81+ 50 4

3. Miyagi-Ken-Oki, Japan 1978 7.4 0.40 100+ 10+

4. Friuli, Italy 1976 6.5 0.30+ 30+  ?

5. Eureka, CA 1975 5.5 0.35 Several* 3

6. Point Mugu, CA 1973- 5.9 0.09 10+ 4

7. Managua, Nicaragua 1972/3 6.2 0.60 4+ 3

8. San Fernando, CA 1971 6.5 1.25 60+ 20+

9. Caracas, Venezuela 1967 6.5 --

Several*

10. Seattle, WA 1965 6.5 0.08 3 Several*

11. Alaska 1964 8.4 --

7

12. Niigata, Japan 1964 7.5 0.18+ Several* Several*

13. Chile 1960 8.5 None Several*

14. Kern County, CA 1952 7.7 0.13 5+ 1

15. Long Beach, CA 1933 6.3 0.15+ Several* 5 Source: EERI,'1981.

+ Indicates equal to or greater than the number shown.

• Actual number not determined.

2.1.3.1 Methods Used in the Pilot Program Two types of facilities were addressed in the pilot program: nuclear power plants and non-nuclear power facilities that have experienced strong earth-quakes (also referred to as data base plants by SQUG).

The steps involved in collecting data from the data base plants and the nuclear power plants and in comparing the data are shown in Figure 2.1-1. Before walk-downs of the data base plants were conducted, available records of the seismic event at each site were collected. These data included ground motion traces recorded near the plant sites. Facilities that had experienced significant ground motion and that also appeared to contain equipment appropriate to the investigation were selected for visits and walkdowns.

Preliminary and final walkdowns were conducted at both the nuclear power plants and the non-nuclear facilities. Preliminary walkdowns at the nuclear power plants were used to identify types of cunmonly encountered safety-related equip-ment. Preliminary walkdowns at the non-nuclear facilities were used to record the locations of types of equipment that are similar to nuclear power plant equipment. Following the walkdowns, particular classes of equipment were selected to be the focus for the remainder of the pilot program. Final walk-downs were used for collection of detailed data, including conducting in-situ j dynamic testing.

NUREG-1030 2-21

[

NUCLEAR POWER PLANTS DATA BASE PLANTS Review Records on Facilities Review Type of Equipment Which Have Experienced Earthquakes 1 f 1 f Select Representative Plants and Select Representative Plants Equipment and Perform Waikdowns and Perform Walkdowns 1 f 1 f Select Plants and Equipment Sefect Plants and Equipment for Detailed Sampling for Detailed Sampling 1 f 1 I Collect Equipment Data and Floor Collect Equipment Data and Response Spectra Floor Response Spectra 1 P Compare Equipment Data and Response Spectra 1 P Determine if Equipment Requires Detailed Qualification Figure 2.1-1 Methods used in pilot study Low-excitation-level in-situ testing was conducted on approximately 200 pieces of equipment in the data base and nuclear power plants to determine approximate primary response frequencies and mode shapes. This permitted estimates to be made of equipment response to floor motion.

Seven classes of equipment

• were selected for detailed study (see Table 2.1-6).

Each class was reviewed to determine similarities between equipment in the two types of power plants. The following characteristics were examined to establish similarity: primary structural and functional characteristics; dimensions and

• An eighth equipment class was later added.

I NUREG-1030 2-22

. Table 2.1-6' Equipment selection for SQUG pilot program-Eq'uipment selected:

Motor control centers 480-V switchgear 2.4 to 4kV switchgear Motor-operated valves Air-operated valves-

' Horizontal pumps Vertical pumps Of seven nuclear power plants visited, three.were selected for equipment data collection:

Plant Design-basis SSE Dresden 3 0.21 g Calvert Cliffs 1 0.15 g Pilgrim 0.15 g name plate data; and ranges of dynamic-response frequency. The response frequencies found during the in-situ testing were compared to determine whether the equipment in the. data base plants and the nuclear plants could be expected to have similar dynamic response properties.

It was noted by SQUG that most of the equipment of interest in the data base plants is located at grade, in basements, or in the first two floors of the structure (up to the turbine decks). In addition, most of the data base struc-tures are relatively stiff, many are either light concrete structures with shear walls or braced steel-frame structures. Therefore, SQUG concluded that no large amplification of ground motion by the structure was expected for the locations of most of the equipment of interest. Free-field ground spectra were used as conservative estimates of the floor response spectra for the data base structures that were not analyzed. Thus, amplification of the data base floor response spectra was excluded.

The floor response spectra required for the nuclear power plants were obtained from the operating utility. Wherever spectra were unavailable for a specific item, amplified floor spectra were assumed on the basis of nearby spectra.

The data base floor response spectra and the nuclear equipment required response spectra obtained as above are then compared to assure that floor response spectra of the data base envelope those of the nuclear equipment.

The performance of data base equipment during past earthquakes was evaluated and conclusions regarding the seismic resistance capability of similar nuclear equipment were reached. A typical comparison is shown in Table 2.1-7.

For the purpose of the pilot program, non-nuclear power plants and other facil-ities in southern California where significant earthquakes have occurred were chosen for the study. Table 2.1-8 shows the four earthquakes in southern NUREG-1030 2-23

Table 2.1-7 Comparison of equipment data

! Variable Data Base Equipment Nuclear Equipment 8 ITEM: 480-V motor control center cabinets i

i 480-V motor control center 39-3

(( IVA-6VA, P3A & P4A (Eight Units) o PLANT: Sylmar Converter Station Dresden Nuclear Plant, Unit 3 MANUFACTURER: General Electric 7700 Line Series, 1970 General Electric 7700 Line Series,1971 l LOCATON: Sylmar Converter Station basement, facing Reactor bulding elevation 570 ft, facing

] northeast and southwest east (grade is at elevation 517.5 ft)

FUNCTION / SYSTEM: Control of pumps and valves for rectifier Control of various Class I mechanical systems cooling systems

CABINET

Each cabinet is four cubicles wide; the Cabinet is six cubicles wide. The cabinet

! specific arrangement of starter units varies contains starter units in cubicles of from cabinet to cabinet; they are otherwise various sizes.

very similar.

i' COMPONENTS: A typical starter unit consists of a General A typical starter unit consists of a General

$2 Electric CR-106 magnetic contractor, a circuit Electric CR-106 or CR-105 magnetic contractor, breaker switch, a control transformer, on-off a circuit breaker switch, a control trans-pushbuttons and a terminal block. former, on-off pushbuttons, and a terminal l

block.

ANCHORAGE: The bottom channel of the cabinet is tack The bottom channel is tack welded to an welded to a baseplate embedded in the concrete embedded baseplate, two welds at the base floor. At least one cabinet was inadequately of each stack of cubicles, front and back, anchored at the time of the earthquake and slid a few inches.

APPLICABLE The records taken at Pacolma Dam are shown The calculated floor spectra for the reactor RESPONSE scaled to 40% of the measured amplitudes building, elevation 589 ft are shown.

SPECTRA: as a conservative estimate of the ground Spectra at elevation 570 ft were not motion at Sylmar. generated.

EQUIPMENT The MCCs were in operation at the time of the STATUS DURING AND earthquake. No damage to either cabinet or FOLLOWING THE components was reported. One cabinet slid EARTHQUAKE: a few inches due to lack of floor anchorage.

Toble 2.1-8 Summary of data bose plants ond earthquakes Earthquake & Dete Facility Estimoted PGA San Fernando 1. Sylmer Converter Station 0.50 - 0.75*

1971 2. Valley Steam Plant 0.40*

3. Burbank Power Plent 0.35*

4. Glendale Power Pldnt 0.30*

5. Pasadena Power Plant 0.20*

6. Rinoldt Receiving 0.50*

7. Vincent Substotion 0.20*

8. Saugus Substation 0.39**

Point Mogu 9. Ormond Beach Plent 0.20*

1973 10. Santa Clora Substation 0.10*

Sonta Barbara 11. Golete Substation 0.28**

1978 12. Ellwood Peaker Plant 0.30 - 0.40*

Imperial Valley 13. El Centro Steam Plant 0.51**

1979 14. Magmamox Geothermal Plant 0.20 - 0.30*

• Locdted near strong motion records.

• • Recorded peek ground occeleration - at plant site.

California that were reviewed in detail in this progrorri. The facilities thot contoined the largest number of equipment items of interest and were reviewed in detail dre the Sylmar converter station, Valley steem plont, Burbank power plant, Glendole power plant, Pasadena power plant, and El Centro steam plant.

Seven nuclear power plants were visited, and three were selected for equipment data collection, they ore Dresden Unit 3, Calvert Cliffs Unit 1, end Pilgrim.

These plents were selected so that the equipment reviewed for the project would form o representative sample of a variety of nuclear plant characteristics, including reactor type and vintage. Only equipment required for safe shutdown Wds Considered.

2.1.3.2 Conclusion and NRC Staff Comments

! The goals of this pilot program were evoluoted by SQUG against the results

! obtdined from the study. Table 2.1-9 lists the gools, findings, and conclu-sion as seen by SQUG. Finally, SQUG reached the following two conclusions:

The structural integrity of onchored power plant equipment and component is not compromised in strong edrthquakes of up to 0.50 g peak ground

[

dCCeleration.

l

• Typically, operability of power plant equipment is not compromised in strong earthquakes with peak ground acceleration of obout 0.20 g to 0.30 g.

Although the staff is in general agreement with SQUG on the first overall point, it has some reservation on the second point, particularly with respect to electrical relays, i

l NUREG-1030 2-25 l

Table 2.1-9 Major conclusions of SQUG GOAL 1: Develop a historical data base on the performance of equipment in conventional power plants during and after strong earth-quakes.

FINDINGS: Several power plants and other industrial facilities have experienced strong earthquakes exceeding the free-field safe-shutdown earthquakes required for the design of most U.S. nuclear power plants.

The plants responded well to the earthquakes and usually continued to operate or were back on line shortly after the earthquakes.

Many of the facilities were in operation at the time of the earthquakes; thus their equipment was subjected to normal operating loads in addition to the seismic loads from the earthquakes.

With a few minor exceptions, the equipment contained in the power facilities was undamaged and was functional after the earthquakes. The equipment was not known to be modified because of the earthquakes.

Sufficient data exist to estimate the spectra experienced by the plants and their equipment.

There is a large, available data base, only a portion of which was sampled in this study, of power plant equipment that has been subjected to strong earthquakes.

CONCLUSION: There is a large body of available data on the performance of power plant equipment in strong earthquakes, including both mechanical and electrical equipment. Many conventional power plants and industrial facilities have experienced earthquakes that subjected their equipment to seismic environments equal to or exceeding seismic loads associated with safe shutdown earth-quakes required for the design of most nuclear power plants.

GOAL 2: Show that much of the equipment investigated, which has experienced strong earthquakes, is similar to equipment found in nuclear power plants.

FINDINGS: A few major equipment manufacturers supply much of the equipment for both conventional and nuclear power plants.

There is little observable difference between the measured dynamic response frequencies of equipment in nuclear power plants and those in conventional plants.

NUREG-1030 2-26

Table 2.1-9 Major conclusions of SQUG (continued)

G0AL 2:

There are no generic differences other than age FINDINGS: between equipment found in conventional and nuclear (CONTINUED) power plants.

CONCLUSIONS: Certain types of mechanical and electrical. equipment found in nuclear power plants are very similar in configuration, function, manufacturer, and model to the types found in conventional plants. Much of the equipment in nuclear power plants and i conventional power plants is the same.  ;

GOAL 3: Determine whether actual earthquake data are sufficient to conclude that seismic qualification of certain classes of equipment by conventional methods is not necessary.

FINDINGS: Excluding some unanchored equipment and one air-operated valve, no failures were reported in any of the seven types of equipment addressed in this study.

With the possible exception of electrical relays, there is no evidence of malfunction of the reviewed equipment during the earthquakes.

The estimated ground-response spectra from several California earthquakes and the conventional power plants affected by them envelop the floor-response spectra for the safe shutdown earthquakes required for nuclear power plants in the ranges of most equipment response frequencies.

Conventional plants that were subjected to earth-quakes with peak ground acceleration of about 0.30 g or lower generally continued to operate throughout the earthquakes.

CONCLUSION: Seismic qualification of nuclear equipment by conventional methods does not appear to be necessary for the classes of equipment evaluated for most levels of safe-shutdown earthquakes.

G0AL 4: Develop a methodology for the use of actual earthquake data to determine whether seismic qualification of specific items of equipment by conventional methods is necessary.

FINDINGS:

The seismic perfonnance of the reviewed equipment appears to be independent from any of the following factors:

Age of equipment Years of service NUREG-1030 2-27

Table 2.1-9 Major conclusions of SQUG (continued)

GOAL 4: Manufacturer and model FINDINGS:

(CONTINUED) Mounting configuration Dynamic properties The methodology used in the pilot program to evaluate classes of equipment would be equally applicable to specific items of equipment.

CONCLUSION: The pilot has demonstrated the methodology. There is an abundance of data that can be used to identify specific items of equipment that do not require additional seismic qualification.

The NRC staff completed the review of the pilpt program report, and concluded that it is feasible to accept experience data as a basis for seismic qualifica-tion. Staff comments on the SQUG pilot program were generally an assessment of what further work should be done to provide an acceptable experience data base.

The comments were sent to SQUG in December 1982.

2.1.4 Summary of EQE Reports, " Seismic Experience Data Base--Data Base Tables for Seven Types of Equipment," " Seismic Experience Data Base--Average Horizontal Data Base Site Response Spectra," and " Investigation of Equip-ment Performance in Foreign Earthquakes and the 1964 Alaska Earthquake"

After reviewing the SQUG pilot program report, the staff concluded that it is feasible to accept experience data as a basis for seismic qualification, so long as some additional work is done to provide an acceptable data base. In a meeting with NRC management in March 1983, SQUG suggested the formation of a third-party Senior Seismic Review Advisory Panel (SSRAP) to provide consulting services and expert opinion for the further development of experience data.

The members of SSRAP were to be five recognized experts in the field of seismic engineering, and in the design, operation and qualification of electrical and mechanical equipment in both nuclear and fossil power plants. The functions of SSRAP were to be:

(1) To review and comment on the validity of the conclusions reached by SQUG.

(2) To provide guidance in the use of earthquake experience data as a screening method to exclude certain classes of equipment from formal seismic qualifi-cation and focus qualification efforts on the more fragile equipment.

(3) To evaluate the data collection and review process and methods used by SQUG in the screening of equipment.

NRC management endorsed formation of SSRAP and the panel was subsequently formed in June 1983 and is organized as follows:

i NUREG-1030 2-28

Chairman Robert P. Kennedy (Structural Mechanics Associates)

Vice Chairman - Water A. von Riesemann (Sandia National Laboratories)

Secretary - Paul Ibanez (ANC0 Engineers, Inc.)

Member - Anshel J. Schiff (Purdue University)

Member - Loring A. Wyllie, Jr. (H. J. Degenkolb Associates, Engineers)

On July 8,1983, SQUG presented its pilot program to the ACRS during the 279th ACRS meeting. The response from ACRS was generally favorable to the pilot program; however, the Comittee observed that "more work is required to estab-lish the operability of equipment during and after an earthquake and more data will be required to support conclusions drawn concerning the seismic resistance of the equipment investigated."

After a. review of SQUG's pilot program report and the staff's coments on the report, SSRAP compiled a list of issues and requested additional information to help the panel in its review. Briefly, the requests and observations follow.

(1) Data Deaggregation. The SSRAP recomended that the data base be deaggregated to provide the following information.

(a) average spectra for the two horizontal components for each plant, rather than the larger (or smaller) of the two; (b) a list of equipment by plant; (c) a list of equipment located more than 40 feet above grade in a structure whose first mode resonant frequency is below 3 Hz. Also, percentage of the data base, on an equipment category-by-category basis, above.40 feet. These data are needed to assess the signifi-cance of possible base isolation and spectra reduction effects of low-tuned buildings; (d) a breakdown of equipment by manufacturer /model, size, and type (e.g., gate versus butterfly valves).

(2) Data Base Extent. The SSRAP recommended that the current data base be extended to include the 1964 Alaska earthquake and 1983 Coalinga earth-quake. These earthquakes should be reviewed largely with emphasis on investigating whether failures occurred or not. The Alaska event is particularly useful because of its long duration. These data will help satisfy the issue of repeated or longer duration shaking. Also, SSRAP recomended that knowledgeable U.S. power industry people be surveyed about their experiences in selected foreign earthquakes (including, at least, Fruili, Managua, and Miyagi-Ken-Oki). The emphasis should be to document, in writing, their experience as to whether a significant number of generic equipment failures occurred.

(3) SSRAP endorsed the SQUG pilot program in general, and agreed that the SQUG activity should be limited to the seven classes of equipment (see Table 2.1-6).

(4) The goal of the SSRAP review will be to establish, if possible, a set of I screening criteria for the seven classes of equipment. The intent was to NUREG-1030 2-29

c. ;_

avoid piece-by piece comparison of equipment in the data base with equip-mentfin the operating nuclear plants. No-further seismic qualification

.of equipment'should be required if it is satisfactorily established by the screening criteria that the equipment belongs to one of the seven classes of equipment. -In order to make this approach feasible, SSRAP believed that a significant amount of data will be'needed for each of the seven classes of equipment.

-(5) Similarity and operability of equipment are the two most important issues to be resolved in developing the screening criteria. . Operability -of equip-ment must be more fully addressed. The conventional plant data do not yet indicate how phenomena such_as relay chatter and breaker trip would affect operations in a nuclear plant. More data and_ study are needed, including studies of the differences in requirements _between conventional and nuclear plants. Alternatively, specific relay qualification or re-placement may be required.

-(6) Generic qualification of the kind proposed may not be possible with struc-tures containing certain brittle materials, such as cast iron and porcelain.

(7) Walkdown of nuclear plant equipment will probably be an essential part of a_ generic qualification procecure.

(8). More explanation is needed for the data on vertical pumps (e.g., nature of shaft supports'and overall size).

(9) The' data base needs to be expanded on motor-operated valves and vertical pumps.

(10) Adequate equipment anchorage should be established before equipment is screened.

SSRAP met with the NRC staff and SQUG seven times from June 1983 to January 1984, and reviewed, exchanged ideas, and commented on the SQUG study. In addi-tion, walk-throughs of several of the non-nuclear facilities in the Los Angeles area used.in the data base were conducted, and Zion and Dresden nuclear power plants were visited. During the November 1983 meeting, EQE provided SSRAP with the information it asked for in the form of three draft reports. Following are summaries'of these reports.

2.1.4.1 EQE Report, " Seismic Experience Data Base--Data Base Tables for Seven Types of Equipment" SSRAP asked SQUG to deaggregate the data base to provide the needed information.

EQE, consultant to SQUG, prepared the report described here (EQE, November 1983c). This report not only deaggregated the data base but included the 1983 Coalinga earthquake.* Foreign earthquakes and the 1964 Alaska earthquake are surveyed in a separate EQE report, described in Section 2.1.4.3. Average hori-zontal spectra for each plant are covered in another EQE report, and are de-scribed in Section 2.1.4.2.

• The performance of equipment in the Coalinga earthquake is documented in an EQE report, dated August 1984 (see EQE, August 1984).

NUREG-1030 2-30

The tables in this EQE report include a count of equipment found within the power plants and industrial facilities studied. The count is limited to items of the seven types of equipment under study. For horizontal pumps and for air-operated valves the count is approximate and conservatively low because of the large number of these items found in the facilities surveyed. Small pumps, both vertical and horizontal, under 50 horsepower, were not included in the count. Data are included in the table entries in varying levels of detail. In general, more detail was collected on equipment which was most representative of that found in nuclear plants. All equipment listeri survived the earthquake without damage, unless otherwise noted.

For each of the seven types of equipment, data are su aarized in a series of columns. The data columns vary slightly among the di.ferent equipment types.

The headings of columns are defined below.

(1) location / Elevation - This entry locates the floor elevation of equipment with respect to grade elevation within the plant. If the equipment is located in the yard adjacent to the plant structures the location is designated as " ground level."

(2) Number of Assemblies (No. Asm.) - For electrical equipment, an assembly consists of multiple cubicles or cabinets mounted in vertical sections which are bolted together to form a single structure.

(3) Number of Units (No. Units /No. Un.) - For electrical equipment, a unit is defined as one circuit breaker cabinet or one motor controller cubicle mounted within an assembly.

(4) Estimated Peak Ground Acceleration (Est. PGA) - This is the peak horizontal ground acceleration estimated for the particular site as an average of two horizontal components (see Section 2.1.4.2).

(5) Size - For electrical equipment, size includes the width of the assembly in vertical sections. The dimensions of the assembly are also included, although for many entries these numbers are simply estimates based on standard cubicle dimensions. Motor control centers are designated as being double- or single-faced assemblies, with cubicles either mounted in both sides or in only one side of the assembly. For metal-clad switchgear, the operating voltage is noted as either 2.4 or 4.16 kV. Motor control centers and low voltage switchgear always operate at 480 V unless otherwise noted on the table. For pumps, size is designated by the motor horsepower j (hp) and by the pump flow rate (gpm) and discharge pressure (in feet of i

head). The total height of vertical pumps is also included, measuring from the base plate to the top of the motor. The size of valves is des-ignated by the pipe diameter and by the operator height measured from the pipe centerline to the top of the operator. Where accurate data are avail-able, entries for valves include an estimate of the flexibility of the supporting line. Very flexible lines are those with measured or estimated frequencies less than 4 Hz. Moderately flexible lines are those with frequencies between 4 Hz and 10 Hz. Supporting lines would be considered stiff if they had no response frequencies below 10 Hz, and rigid if they had no frequencies below 33 Hz.

l (6) Frequency - For a few sample items, measurements were made of the lowest response frequency as an indication of the typical flexibility of the NUREG-1030 2-31

u

. 4

. type of. equipment. -For electrical equipment,_the rocking or overturning frequency of the~ assembly is.noted where measured. For valve operators, the rocking or.." cantilever" frequency is noted where measured. . Valve-operator-cantilever frequencies correspond to the response.of the' operator-relative ~to the supporting piping.

(7) Form -'For electrical _ equipment, details of internal devices are'provided where available. Data on specific components are given for typical

~

cubicles or cabinets within an assembly. For example, the major compo-nents for a typical motor controller within a motor control center (MCC) may be listed-including the manufacturer and the model number if avail-

=able. 'For switchgear, the model number of a_ typical circuit breaker in the assembly may be noted._ along with the types of door-mounted-relays on the front face of the assembly. For vertical pumps, the type of pump is designated as either a turbine or a centrifugal pump. ' Vertical turbine pumps include the length of the shaft below the base plate, if known. The means of. support for the suction _line containing the shaft is also noted if known.. Most vertical turbine pump suction casings are supported only at the pump base plate. 'The suction casing thus forms an inverted canti-lever into the source of water below the pump motor. For horizontal-pumps, the drive mechanism for the pump is noted as either electric motor, steam turbine, or diesel engine. The drive train is noted as either through a gearbox or-transmission, or by a direct connection between motor and pump.

The type of pump is noted as either a centrifugal single impeller, a multistage turbine pump, or a screw. For valves, the type of valve is designated (if not covered by insulation). The orientation of the attached operator is noted with respect to the valve.

(8) . Attached Pipina - For pumps, the diameters of the suction and discharge lines are listed if this information is available.

(9) Manufacturer, Model Vintage - The manufacturer of the equipment is noted where nameplate data were collected. If-a designation of model, size, or type was include on the nameplate, this is noted. The equipment vintage is usually estimated according to the year of construction of the particu-lar unit of the plant.

(10) Internal Details - For electrical equipment a short description is provided of the units which make up the assembly, including variations in the size of cubicles, and the ratio of occupied to blank cubicles in the assembly.

An assembly is listed as full if all or nearly all of its aVailable cubicles contain motor controllers (in MCCs) or circuit breakers (in switchgear). Additional details are included, such as the pressure of door mounted components such as relays, or the inclusion within the assembly of equipment such as transformers. '

(11) Installation - The anchorage of the equipment is described where this information was collected. For some entries, the size of anchor bolts are estimates. Any additional supporting structure other than anchorage to the floor (or pipe) is noted.

(12) Photographs Available (Photo Avail.) - Photographs are rvailable for nearly all equipment listed. Exceptions exist for a portion of the hori-zontal pumps and air-operated valves which are usually found to be repeti-tions in a particular facility. Where only a portion of the individual NUREG-1030 2-32

y E

4 items counted in a table entry have available photos, the photo inventory is listed as " partial."

(13) Catalogue Available (Cat. Avail.) - For a portion of the equipment, 4- manufacturer's catalogues, equipment specifications, or drawings for the 1

particular-item have been collected.

3 A--summary table is included in this EQE report for each of the seven types of <

equipment. The summary table provides a total count of equipment, broken down i according to earthquake, data base plant, and elevation with respect to ground.

A summary is also provided of the manufacturers and vintage of the equipment, and the performance of the data plant during the earthquake.

) The tables are followed by a series of plots in which certain parameters for

- each equipment type are presented in graphic form.

Typical samples of tables and plots are presented in Tables 2.1-10 through 2.1-16 and Figures 2.1-2 through 2.1-9 for the seven types of equipment for a random selection from various data base plants.

]

L 2.1.4.2 EQE Report, " Seismic Experience Data Base--Average Horizontal Data i Base Site Response Spectra" i For some of the facilities included in the seismic experience data base, ground j motion records were not available at their specific locations. The nearest '

j. ground motion record was then used by EQE to extrapolate an estimate of the

peak ground acceleration and the shape of the ground motion response spectra

at the data base site. This EQE report (EQE, November 1983b) includes plots of

the horizontal ground motion response spectra for the various data base sites  !

, used in the SQUG studies. The two horizontal ground motion response spectra 1 are plotted as dashed lines for each record. The average response spectrum of i the two horizontal components is plotted as a solid line. This average hort-zontal spectrum is then used for the various data base sites, multiplied by a

- scaling factor to account for the location of the data base site with respect i to the causative fault or the epicenter. As an example, the development of the

! estimated data base site horizontal response spectra during the February 9 j 1971 San Fernando earthquake is described below, i Scaling factors to estimate data base site response spectra for the San Fernando j sites were developed by EQE in the following manner. Peak ground accelerations

! were obtained from the sites of actual ground motion records. These peak ground accelerations are the higher acceleration of the two horizontal components

! recorded. By comparing the location of the various data base sites with the l

locations of the records with respect to the causative fault, estimates were j made of the peak ground acceleration at the data base sites. These estimates l were based on past studies of ground motion attenuation as a function of l distance from the fault. The average ground motion response spectrum for the

, nearest ground motion record was then scaled by the ratio of estimated peak .

l ground acceleration at the data base site to the measured peak ground accelera-l tion at the record site. For the data base sites in the San Fernando Valley, j this procedure is summarized in Table 2.1-17.

Figure 2.1-10 shows a map of the San Fernando Valley included to locate data

base and ground motion record sites. Figures 2.1-11 and 2.1-12 show the j response spectra at Pacoima Dam and Sylmar Converter Station, respectively.

i NUREG-1030 2-33 l

b

s Table 2.1-10 Summary: flotor control centers fTt O

H a

Est.

O Ito. Ito. PGA flenufacturer, flodel, Earthquake Location Elevation Asm. Units,(g) Vintage O Performance During Earthquake San Fernando Sylmar Besement 11 100 0.50 1971 General Electric. Cutler Facility it,st power for several Hammer, *1970 months; no meter controllers 12 ft. 7 109 required replacement; one assembly

$11d slightly.

43 ft. 5  : 35 Valley Ground floor 6 83 0.30 General Electric, Federal Three units were on-1 foe; two Pacific 1950s tripped off-line and lost power, 15 ft. 11 218 one remained on-line. Its damage to meter controllers.

Burbank Ground floor 5 126 -0.32 Westinghouse, Cutler Four units were en-1tne; twe Olive Mammer, +1960 tripped off-line, two remained Plant on. All shut down shortly after Electric Itachinery the earthquake as offsite power late 1960s was lost, les damage to motor controllers.

Glendale Basement 16 162 0.27 Westinghouse. *1963 Three units were on-line; all m remained on-line.

w General Electric, *1959 A

Square D. *1953 Pasadena Ground floor 1 24 0.18 General Electric, *1965 Two units were on-Ifne; both remained on-line.

Federal Pacific, *1957 17 ft. 2 20 33 ft. 1 30 Imperial El Centro Ground floor 3 30 0.42 Westinghouse *1957 Two units were on-line; one lost Valley power; one tripped off-ifne but 1979 Square D *1968 continued to operate. No damage 20 ft. 2 26 to motor controllers.

Coatinga Withf.: Ground level 7 212 0.60 Nelson Electric, *1970 All facilities lost power. Two 1983 10 km of unenchored assemblies slid; two epicenter Furnace Electric, *1980 anchored assembifes failed anchorage and slipped. leo Westinghouse, +1980 damage to motor controllers.*

ITE, *1972 and 1980 Within Ground level 4 25 0.35 Westinghouse, +1970 All pimping stations lost power.

20 km of stotor controllers were not epicenter General Electric *1970 damaged Total - --

81 1280 -- -- --

• 0ne motor controller unit at the Union oli Butane Plant was inoperable following the earthquake because of a thermal overload relay that would.

not reset. Operators at the plant thought that the controller's condition had been noticed before the earthquake, but positive confirmation could not be made, l

l

l x

C Es m

G1 e

> Table 2.1-11 Motor control centers at the Sylmar Converter Station o

ha O Est.

No. No. PGA Frequency Manufacturer, Photo. Cat.

Location Asm. Un. (g) Size (Hz) Form Model, Vintage Internal Details Installation Avail. Avati.

Basement 8 110 0.50 4 secticas Not Typical unit contains General Electric. Cubicles of 3 sizes; Tack welds to embed- Yes Yes wide; 90" m measured GE CR-106 contactor. 7700 Line Series assemblies are 2/3 ded base plate in 20" x 80*; circuit breaker, NCC, +1970 full; assembly concrete floor; cthicles on control transformer, includes a switch- about 6 per assembly.

I side. pilot lights, & push board.

buttons.

Basement 2 10 0.50 2 sections Not Typical unit contains General Electric, Cubicles of I size; Tack welds to embed- Yes Yes wide; 90" x measured CR-106 contactor, 7700 Line Series 1 section are spares. ded base plate in 20" x 40*; circuit breaker, MCC,+1970 concrete floor; cubicles on control transformer, about 4 per assembly.

I side. pliot Ilghts, & push buttons.

Basement 1 60 0.50 8 sections Not Internals not Culter-Hammer, Cubicles of 3 sizes; Anchor bolts,1/4" Yes No wide; 90* x measured inspected. Unitrol, +1970 assembifes are 3/4 diameter, at corners fu 20" x 160*; full; assembly in- of each section.

O cubicles on cludes a large trans-us both sides, former at one end.

Second 1 32 0.50 5 sections Not Internals not Cutler-Hammer, Cubicles of 2 sizes; Anchor bolts, 1/4". Yes No floor wide 90* x measured inspected. Unitrol *1970 assemblies are 3/4 diameter, at corners 12' 20" x 100"; full; I secticn of each section, above cubicles on supports instrumen-ground both sides. tation rather than motor controllers.

Second 6 77 0.50 3 sections Not Internals not General Electric, Cubicles of I size; Anchor bolts, 1/4" Yes Yes floor wide; 90" x measured inspected, but 7700 Line Series assemblies are full; diameter, across 12* 20" x 60*: probably siellar MCC, *1970 1 cubicle supports a center of assembly above cubicles on to other GE units. door-mounted relay, base.

ground I side.

Fourth 5 35 0.50 2 sections Not Internals not Cutler-Hammer, Cubicles of 2 sizes; Anchor bolts, 1/4' Yes Yes.-

floor wide; 90" x measured inspected. Unitrol *1970 assemblies are 1/2 diameter, at corners 43' above 20" x 40"; full, of each section.

grade cubicles on I side.

9

=

C

M3

$ e fable 2.1-12 Summary: Motor-operated valves N

O Est. Operator,

$ Earthquake Location Elevation /Pfping No. of Valves PGA (g)

Manufacturer, Flexibility Model, Vintage Performance During Earthquake San Fernando Valley E1. 10 ft. 14 0.30 Limitorque. Three units were on-line. Two 1971 Spring-supported *1953 trfpped off-Ifne and lost power; feedwater lines one remained on-line. No damage to valves.

E1. 20 ft. 17 0.30 McBain Torkeaster, Spring-supported *1957 feedvater ifnes Burbank Ground level 2 0.32 Limitorque, *1958 Four units were on-line. Two Rigid 24* lines tripped off-line; two remained on. All units shut down shortly El. 20 ft. 2 0.32 Limitorque,*1958 after the earthquake as offsite Very fleutble Ifnes power was lost. No damage to valves.

Glendale Basement Mezzanine 4 0.27 Limitorque, +1959 Three units wers on-line; all y Moderately flexible remained on-line. No damage to e ifnes valves, w

• El. 6 ft. 1 0.27 Limitorque, +1959 Very fleutble Ifne E1. 20 ft. 1 0.27 Limitorque, *1953 Adjacent to boiler E1. 60 ft 1 0.27 Lleitorque, *1965 Adjacent to better Imperial El Centro Ground level 2 0.42 Limitorque Two units were on-Ifne. One Valley Rigid 24" lines unit lost power; one tripped 1979 off-Ifne but continued operating.

E1. 80 ft. 3 0.42 Limitorque, No damage to motor-operated valves.

Adjacent to boiler 1953 - 1968 Coatinga Main oil Ground level 55 0.60 Limitorque, Plant lost power and all 1983 pumping Short piping runs 1967 - 1980 equipment shut down. - Some plant probably rigid damage to plastic conduit attached to valve motors.

San Luis Ground level 29 0.35 Limitorque, Stations lost power. No Canal Short piping runs 1963 - 1979 damage to valves, pumping probably rigid stations Total --

131 -- -- --

i 1

+

2 C

m Table 2.1-13 Motor-operated valves at near-field sites near Coalinga rvi C) d o Est.

Operator, Manufacturer, W Mo. of PGA Frequency Model, Photo. ' Cat.

O tocation Valves (g) Size (Hz) Form Vintage Installation Avail. Avail.

Ground level, 1 0.60 Pipe diameter = 12" Not measured Cate valve, operator Limitorque, Motor /gearbos bolted ' Yes Yes Main oft mounted above and to Type SMC-03, to yoke with four 1/2" pumping plant Operator ht. = 75" one side of the valve. *1980 bolts; shaft bolted to 1000 wt. valve with 1/2" bolts.

Short span of pipe, probably rigid Ground level, 2 0.60 Ffpe diameter = 12" Not measured Gate valves, operator Limitorque, Motor / gearbox bolted Yes Yes Main oil mounted directly Type See, to yoke with eight 1/2" pumping plant Operator ht. = 60* above. Size 1, bolts; yoke bolted to

• 1967 valve with eight 1/2" short span of pipe, 400f wt. bolts, probably rigid Ground level, 4 0.60 Pipe diameter = 12" Not measured Gate valves, operator Limitorque, Motor / gearbox bolted Yes Yes Main oil mounted directly Type SMB to yoke with eight 1/2" pumping plant Operator ht. = 40" above. Size 00 bolts; yoke bolted to 7

w Short span of pipe,

+1967 2008 wt.

valve with eight 1/2" bolts.

N probably rfgl Ground level. 7 0.60 Pipe diameter = 24" Not measured Gate valves, operator Limitorque, Motor / gearbox bolted Yes Yes Main oil mounted directly Type SMC-03 to yoke with efght 1/2" pumping plant Operator ht. = 90" above. *1980 bolts; yoke bolted to 100f wt. valve with eight 1/2" Short spans of pipe bolts.

well supported, probably rigid Ground level. 4 0.60 Pipe diameter = 8" Not measured Cate valves, operator Limitorque, Motor / gearbox bolted Yes_ Yes Main all mounted directly Type SMC-03, to yoke with four 3/8" rumping plant Operator ht. = 40" above. *1980 bolts; yoke bolted to 100f wt. valve with four 3/8" Short spans of pipe, bolts.

probably rigid Ground level, 4 0.60 Pipe diameter = 10" Not measured Globe valves, operator Limitorque, Motor / gearbox bolted Yes No Main oil mounted above and to Ident. No. to yoke with four 3/8" pumping plant Operator ht. = 20" one side of valves, 876P0576M-WF bolts; yoke bolted to

• 1%7 valve with four 3/8" Short spans of pipe, bolts.

probabl; rigid Ground level, 4 0.60 8uried pipe Not measured Gate valves, operator Limitorque, Motor / gearbox bolted Yes Yes Main oil projects out of ground Type SMC-03 to yoke with four 3/8" pumping plant Operator ht. = 20" directly above valves. *1980 bolts; yoke bolted to 100f wt. valve with four 3/8" Short spans of pipe, bolts.

probably rigid

2 C

x m

o B

(a Table 2.1-13 Motor-operated valves at near-field sites neae Coalinga (continued) o Operator.

Est. Manufacturer, No. PGA Frequency Model, Photo. Cat.

Location Valves (g) Size (Hz) Form Vintage Installation Avail. Avall.

Ground level, 2 0.60 Surfed pipe Not measured Gate valves, operator Limitorque, Motor / gearbox bolted Yes No Main oil projects out of ground Ident. No.

• to yoke with four 3/8" pumping plant Operator ht. = 30" and is then offset to 876P0576M-WF bolts; yoke bolted to one side. $1967 valve with four 3/8".

Above ground bolts.

Ground level, 2 0.60 Pipe diameter = 24" Not measured Gate valves, operator Limitorque, Motor / gearbox bolted Yes Yes Main ott . mounted directly Type See to yoke with eight 1/2" pumping plant Operator ht. = 70" above. Size 1 bolts; yoke bolted to

+1%7 valve with eight 1/2" Short spans of pipe, bolts, probably rigid Ground level, 1 0.60 Buried pipe Not measured Gate valves, operator Limitorque Motor / gearbox bolted Yes Yes N Main off mounted directly Type SMC-03 to yoke; yoke bolted to d

co pumping plant Operator ht. 50" above. *1980 100# wt.

valve with four 3/8" bolts.

Ground level, 5 0.60 Buried pipe Not measured Gate valves, operator Limitorque, Motor / gearbox bolted Yes 'Yes Main oil mounted directly Type SMC-03 to yoke; yoke bolted to pumping plant Operator ht. = 30* above. *1980 valve with four 3/8" 100# wt. bolts, Ground level, 5 0.60 Pipe diameter = 24" Not measured Gate valves, operator Limitorque Motor / gearbox bolted Yes No Main oil mounted directly (no name- to yoke; yoke bolted pumping plant Operator ht. = %" above. plate) to valve with four 1/2"

! bolts.

Short spans of pipe, probably rigid Ground level, 8 0.60 Buried pipe Not measured Gate valves, operator Limitorque Motor / gearbox bolted Yes Yes Main oli mounted directly Type SMC-03 to yoke; yoke bolted pumping plant Operator ht. = 40" above. $1980 to valve with eight 5/8" 100# wt. bolts.

Above ground Ground level, 6 0.60 Buried pipe Not measured Gate valves, operator Limitorque Anchorage not visible. Yes Yes Main oil mounted directly Type $8e pumping plant Operator ht. = 36" above. Size 0

• 1967 Above ground 300# wt.

i

1 2

C ll0 rrt Ch e

H o

LJ o .

Table 2.1-14 Motor-operated valves at far-field sites near Coalinge ,,

Mo. Est. Operator. C of PGA Frequency Manufacturer, f '- ' '

Photo. Cat.

Location "alves (g) Size (Hz) Form Model. Viretage Installation Avail, Avail.

San Luis 4 0.35 Pipe diameter = 8*-16' Not Butterfly valves; Limitarm e, Motor / gearbox sconted atop Yes Yes '

• 9 Canal Operator ht. = 16" m 'q.

measured operator mounted Type SMC-04,*'

Pumping Short spans of pipe,

. wore gear actuator; s  %)

to one side. 1979 . actuator halted to valve Station probably rigid. 100f wt. flange with four 3/4" 20-R bolts.

i San Luis 8 0.35 Pipe diameter = 8*-16* , ,Not Butterfly valves; thdtorque, Motor / gearbox mounted atop Yes Yes Canal Operator ht = 18" measured operator mounted Type H18C-Sp5-00, worm gear actuator; 4 Pumping Short spans of pipe, ' to one side. 1975 actuator bolted to valve.

g Stations probably rigid. 2008 wt. flange with four 3/4" e 21-R & bolts.

M 22-F J c Pumping 9 0.35 Pipe diameter = 8*-14' Not Butterfly valves; Limitorque, Motor / gearbox mounted atop Yes Yes Stations Operator ht. = 18" measured operator mounted Type SMC-04, worm gear actuator; 16-RC & Short spans of pipe, to one side.

1978 actuator bolted to valve ' " '7 4 14-RC probably rigid. 100f wt. flange with two 1/2"

, bolts.

Pumping 4 0.16 P$ediametereN' Not Sutterflyfialves; Limito*que, Motor /gearbos mounted atop Yes )es Itstion '

Operator ht. = 24' measured operator mounted Ty,e H, woromaar erivator; e ,

7-1 Short spans of pipe, to one side. 1963., actuator bolted to valve -s

~ * '

probably rigid. 200s wt. flange witt his 3/4" M bolts.

• Pumping 4 0.35 Pipe diameter

• 10*-20" Not Sutterfly valves; Limitorque, Motor /geart.Feounted atop Yes Yes d

Station Operator ht. = 18" measured operator mounted Type SMB-00, worn gear actuator; th-RA to one side. 1979 actuator bolted to valve 200f wt. flange with tws 3/4 #

bolts.

1 ..

~

U

.,, x - ..,

e k

3' .

-3 ki f

n (

.W '

fl<

r. , _ -

e _. q t

g

.Z<

s Table 2.1-15 Vertical pumps 2

C ,

• • ""P' Manufacturer, 'peodel, e

@ Earthquake Location Elevation (g) Vintage Performance During Earthquake

>.4 50-200 4 400 5 o Three units were on-line. Two tripped off-w San Valley Ground floor 23 4 0.30 Motors - General Electric, O Fernando 'tlllot, Westinghouse, US line and lost power; one remained on-Ifne.

l 1971 E1. 20 ft 8 0 Electric. No damage to pug s.

Pumps - Johnston, Byron-Jackson, Peerless, United.

1954-1956.

Barbank Ground flose 4 2' O.32 -Motors - Allis Chalmers, Four units were on-line. Two tripped off-General Electric, US Ifne; two remained on. All units shut down Electric. shortly after earthquake as off-site power was lost. No damage to pumps.

Pumps - Byron-Jackon, 1*60.

Glendale Basement 6 0 0.27 Motors - General Electric. Three units were on-line; all remained Allis Chalmers, on-line.

Ground level 1 2 Pumps - Syron-Jackson, Peerless, US Pump, 1941-1964.

to S Pasadena Ground level 0 4 0.18 Motors - General Electric, Two units were on-Ifne; both remained O on-1tne.

Pumps - Foster-Wheeler, 1957.

Coalinga Facilities Ground level 0 8 0.60 Motors - Westinghouse, A11 facilities lost power and shut down.

1983 within Stemans-Allis, US No damage to pumps.

10 km of Electric.

epicenter Pumps - Byron-Jackson, Union, Veriline.

1967-1980.

Pleasant Ground level 0 9 0.49 Motors - Toshiba Shiburu. Plant lost power and all equipment shut down.

Valley No damage to pumps.

M - Ebaru,1%9.

San Luis Cround level 29 27 0.35 Motors _ General Electric, Stations lost power and all equipment shut' Canal Westinghouse, US Electric. down. All pumps were operable following the earthquake. A few pug s displayed Pumps - Peabody Floway, excessive vibration because of worn' bearings.

1970-1979.

Pump Ground level 0 4 0.16 Motors - General Electric. Station was down at time of earthquake. No Station damage to equipment.

7-1 Ptamps - Fairbanks, Morse, 1%3.

Total 68 60

A f'

.g 6 2

C 3D -

rre O

e N

O to o

Table 2.1-16 Vertical pumps at near-field sites near Coatinga Est.

No. PGA Attached Photo. Cat.

Location Purps (g) Size Form Piping Manufacturer, Model. Vintage Installation Ave *1. Avall.

Ground level 2 0.60 Motor - 300 hp. Turbine pump; shaft 12" suction Motor - Westing *HMrse, tifeline Base of pisup Yes No Main oft length unknown. 24" discharge . anchored to con-pumping M no nameplate. ' Induction Motor.

crete pad with t

plant Puug no nameplate *1%7. twelve 1" bolts.

Total ht. = 8 ft.

Ground leve') 2 0.60 Motor - 500 hp. Turbine pump; shaft 16" suction Motor - Seimans-Allis Base of pump Ves No Main oil length unknown. 16" discharge Induction Motor. anchored to con-pu : ping M - 3500 gpe, crete pad with +

N plant 271 ft. head. M - Byron-Jackson -1980. four 1" bolts.

e M Total bt. = 9 ft.

Ground level 4 0.60 Motor - 700 hp. Turbine pump; shaft 12" discharge Motor - U.S. Electric. Base of pump , Pertfel No Water length = 20 ft.

bolted to con- - -?

filtration P g - no nameplate. M - Verfline Turbine Pump,- crete with four.

plant *1970. 1/2" bolts.

Total ht. - 10 ft.

Pleasant 9 0.49 Motor - 7000 hp. Centrifugal pump; Suction from Motor - Toshiba Shiburu The motors are Partial No Valley motor and pump on canal, 36" Type TAK. buf1t into a con-Pumping M - 225 ft.8/ different floors, discharge crete pedestal on Plant sec., 197 ft. head. connected by a 30-ft. .Ifne M - Ebaru centrifr. gal pump the ground floor; drive shaft. Type 54-39VLM,1%9. the pump is moun-ted on the base-ment floor below the canal water ifne.

Nueber of Assesblies P6A year Manufacturer 10 15 20 5

+ - -+.....---------+----..---------

Zinsco 0.30g 1952  !!IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII Square 8 0.27g 1953 IIIIIIIIIIIIIIIIIII l Federal Pacific 0.30g 1956 lIIIIIIIIIIIIIIIIIIIIIIIII l Federal Pacific 0.lBq 1957 !!!!!!!! l Westinghouse 0.42g 1957 l1111111 l 6eneral Electric 0.27 1959 lIIIIIIIIIIIIIIIIIIIIIIIII i.

Cutler Hanser 0.32 1960 +11111111111 Westinghouse 0.32 1960 lIIII l 0.27g 1963 Westinghouse j!!!!IIIIIIIIIIIIII l 6eneral Electric 0.18g 1965 jIIIIIII i Electric Machinery Corp. 0'.32 1968 IIIII j Square 3 0.42 1968 IIIIIIIIIll! .

Culer-Hanser 0.5 1970 l!IIIIIIIIIIIIIIIIIIIIIIII  ;

Delta Switchboard Co. 0.3 1970 + H31 .

Seneral Electric 0.50 1970 lIIIIIIIIIIlllI1XIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII Nelson Electric 0.60 1970 lIIIIIII i Westinghouse 0.3 1970 IIIIIIII l ITE 0.6 1972 IIIII  ;

Seirra Switchboard Co. 0.3 1973 !!!!!  ;

l l l l I

l

-Furnas 0.60g 1980 +111I I ITE-Gould 0.60g 1981 11111 l Wastinghouse 0.60g 1981 11111111 l

! ----+--------------+

. ..+. . - - . . + . --

Figure 2.1-2 Distribution of motor control centers as a function of vintage, manufacturer, acceleration, and number of assemblies 2.1.4.3 EQE Report, " Investigation of Equipment Performance in Foreign Earthquakes and the 1964 Alaska Earthquake" The equipment earthquake experience data base compiled by EQE for the SQUG project during the pilot program indicates a lack of failure for the seven types of equipment considered. The data base equipment was subjected to seismic motions comparable to the design earthquakes for the operating nuclear power plants in the eastern U.S. However, it does not include any data from earth-quakes outside the U.S. The possibility of discovering numerous equipment failures during well-known earthquakes not investigated by the project is a serious concern on the part of SSRAP.

NUREG-1030 2-42

20 +--------+--------+--------+ + - +---------+--------+ -+--------+----- -------+

i i i  :

1 10 + 1 +.

l I l

! I l

! I I i I I  :

16 + I I +

l I I  :

I I 1 I I i i I I  :

i I I 14 + 1 I +

N  !  ! I I u  ! I I e I I I i b l I I I i e  : I I I i r 12 + I I I I I +

i I I I I I i o i I I I I I i

(  ! I I I I I i I I I I I  :

A  : I I I I I  :

s 10 + I' I I I I +

s i I I I I I l e i I I I I I i e  ! I I I I I b  ! I I I I I 1 1 I I I I I  :

i 8+ I I I I I +

e  : I I I I I  :

5  : I I I I I  :

I I I I I  :

! I I I I I  :

I I- 1 I I '!

6+ 1 I I I I +

l I I I I I l l I I I I I  :

I I I I I  !

! I I I I I  :

I I I I I 4+ 1 I I I I I +

! I I I I I I  :

! I I I I I I

! 'I I I I I I I I I I I I  !

! I I I I I I  :.

2+ I I I I I I I I +

!  ! I I I I I I I  !

!  ! I I I I I I I

: I I I I I I I I I  :

I I I I I I I I I  :

1 I I I I I I I I  :

l 0+ +------+-------+------+-------+ --+---- +--------+- -+ --+- - ----+

0 1 2 3 4 5 6 7 8 9 10 11 Number of Sections in Width Figure 2.1-3 Motor control centers surviving PGA > 0.18 g, data base of motor control centers plotted as a function of width in sections i

NUREG-1030 2-43

20+--------+--------+--------+--------+--------+---------+--------+-----i------+--------+ ---+

l l i  :

l l 18 + +

t l 1  :

i i  !

16 + +

I  :

i  :

l l l 14 + +

N i i u  !  !

a  !  :

b  :

e  ! l r 12 + I I +

1 I I  !

o  ! I I f i I I 1  :

i I I I i A  : I I I  !

s 10 + I I I I I +

s-  !  ! I I I I  !

e i  ! I I I I a  ! I I I I I  :

b I I I I I  :

1 -! I I I I I l 1 8+ I. I I I I +

e- l I I I I I l

s.  !  !  ! I I I I I I I I

! I I I I I  :

i I I I I I l 6+- I I I I I +

!  !  ! I I I  :

!  ! I I I I  !

l I I I I I  :

!  !  ! I I I l l 1 I I I I 4+ 1 I I I I +

l I I I I I 1 I I I I I  ;

i  ! I I I I I I I I I I I  ;

I I I I I I  :

2+ 1 I I I I I I +

l I I I I I I I  :

1 I I I I I I  :

I I I I I I I I  :

I I I I I I I I I I I I I I I I l 0 +- -+ - ..+... ...+ .+ _+ .+ .. ..+ -+- + -+........+

0 1 2 3 4 5 6 7 8 9 10  !!

Nueber of Sections in Wicth j Figure 2.1-4 Motor control centers surviving PGA > 0.28 g _

NUREG-1030 2-44

~

i. a {,i.

20 +---- - -+ - --+--------+--------+--------+- -+--------+--------+--------+--------+--------+

l-  !

18 + +

-!- I

' 16 + +

~!  !

14 + +

11  :  !

u  !  :

e  :

b.  !  :

e  ! I r 12 + +

c  :  !

f  !  !

A 1, s 10 +  ! +

s~ -!  !  :

.e  !  !  !

e  : I b  : I.

.1 I  :

i ~B+ I +

e  : I  :

s I l

=!  ! I  ;

l 1 I  !

! I I 6+ 1 I I +

l I I I  :

! I I I  :

! I I I  :

! I I I  !

! 1 -1 I  :

4+ I I I +

! I I I I I I  :

I I I I  :

!  ! I I I.

I I I I  :

2+ I I I I I +

! I I I I I

! I I I I I  :

I I I I I I I

! I I I I I I I  !

!- I I I I I I I  !

o +-... .+.. +... ...+. + .+ . + + ..+........+__ -+ +

0 1 2 3 4 5 6 7 8 9 10 11 l Nueber of Sections in Width Figure 2.1-5 Motor control centers surviving PGA > 0.45 g I.

NIJREG-1030 2-45 t-l i

+

?

100.0 + + + -+--+------+------4-------+----+- +- +---+ +-+

5 :

90.0 + 4 '711000l* 1

. i

':  : I 80.0 + +

l l  ! i

-l  :

~:

70.0 + 8:6509) 2(4004)+

0  :  !

p  : ~!

e 1110001  :

r  :

a 2(40001  :

t 60.0 + 1(5000) +

o  !  !

r  : 1  :

H e  : 1(65081 7  :

i . 50.0 + 4(335tl 2(6500) +

2 i i t  !  !

l 2(30001  :

i l 2 i n 40.0 + 4(1000) 1 4(2006) +

C 'l l' h  : 2(20006  :

e  ! 2(2000):

s-  :

30.0 + 1 2 +

2(10001  :

! 1(10001  :

l' 20.0 + 5(1000) 4 +

!. 4(2000) 2(2000) 2(20001 4(1000) 2(2000) 2(20001  :

- 2(10001 1(1000) 1(109tl  :

10.0 + +

l l l l l l

-0.0 + +. +__ +. ..+. +. + + ..+ . ..+ . .+- +. .+...,

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 22.0 24.0 j

Pipe Diaseter (anches) i i

! Key: Number of valves (operator weight #)

Operator weight is omitted if data are not available.

Buried pipes are omitted Figure 2.1-6 Motor-operated valves surviving PGA > 0.18 g, data base of _

motor-operated valves plotted as a function of supporting pipe diameter and operator height NUREG-1030 2-46

150.0 +- -----+ ~ ~---+ -+----~-+-~~~+--- +- -+-~~ -~+- ~ ~ + +-------+----~+---+

i  :

100.0 + +

1  :

i 2(50')  :

650.0 + 1(50') 8(10') +

I I  !

600.0 + +

I  :

550.0 + +

O  ;

p 500.0 + 1(60') +

e  !  :

r i  :

a i  :

t 450.0 + +

o r  :

11 400.0 + 2(60') 2(70'l+

e 1 i  !  !

.e  !  !

E 350.0 + +

t i 4(50')  !

P  :  :

o 300.0 + 2(44') +

u  :

n d I  :

s 250.0 + +

! 2(18')

200.0 + 2(36*) 4(18') 2(18') 4(40') 2(18'l 2(18') 2(35'l+

I  :

150.0 + +

1(24') 5(18') 1(16'l  :

100.0 + 2(26') 2(16') 1(65'l 4(18'l 1(16'l 7(90'l+

l 4(40')  !

1 l 50.0 + +

l l l l l-  :

0.0 s + ....+ .. +.... .+... + + ~+ -+.. +-- .+- -+- +---+

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 22.0 24.0 Pipe 0 aseter (anchest Key: Number of valves (operator height ")

Only valves for which operator weights and pipe diameters are known are plotted.

Figure 2.1-7 Motor-operated valves surviving PGA > 0.18 g, data base of motor-operated valves plotted as a function of supporting pipe diameter and operator weight NUREG-1030 2-47

20+------------I--+--------------+--- ---------+-- --------+--------------+----+

l I .

! I

I

• I

,I I i

+

1B + I l 1 I .

l I  !

. I .

! I .

l I +

16 + 1

I l I I

{.1 l I I .

! I i 14 + 1 I N I I ' +l u I I  !

t  : I I I .

b  ! I I I e  !  !  ! I +

r 12 + I I I l I I I o l I I I .

'f l II I I II I I l.

I l P l II I u 10 + II I I I I a

p II I II I I

I I

I I

I +ll I

5  : II I I I I

! II I I I I I i II I I I I I  !

8+ IIII I I I I +

I I I  !

-l IIII I I l IIII I I I

IIII I I I I

IIII I I I I  !

! IIII I I I I l 6+ IIII I I I I +

! IIII I I I I  :

l IIII I I I I  !

! IIII I I I I  :

! IIII I I I I  :

IIII I I I I 4+ IIII I I I I I I I +

-l IIII I I I I I I I  !

IIII I I I I I I I  :

! IIII I I I I I I I I .

l IIII I I I I I I I I l

l IIII I I I I I I I I  !

l I

• l 2+ II I I I I I I I I I I I i II I I I I I I I I I I I I  :

II I I I I I I I I la I I I  !

! I

! II I I I I I I I I I I I I

! II I I I I I I I I I I I I I  !

! II I I I I I I I I I I I I I  :

-+... . -+ +- - ..+. +

0+ . +

1000 0 200 400 600 800 l

l Puep Horsepower l

Figure 2.1-8 Vertical pumps surviving PGA > 0.18 g, data base of vertical L pumps plotted as a function of pump horsepower l

NUREG-1030 2-48

.gn s + ..+........+...... +. +..._+ + +..L ...+ . ..+

t 55 + +

! I  :

I  :

. I  :

50 + I +

! I

.: I  :

i I  :

45 + I +

I .

1. I  :

N.  ! I  :

u '40 + 1 +

a i I'  :

6  : I  :

e I r  !  !  !

35 + I +

o l I  !

f  : I  :

I P  !  !

u 30 + 1 +

e  ! I  :

p  : 1 s  !  !  !

!  ! l 25 + I +

-1 I- l

! I  :

I  :

20 + I +

l I  :

I  :

I  :

i I i 15 + 1 +

l I  :

I  :

I  :

I  :

10 + I +

I  :

I  :

1 I  :

! I I I  :

5+ I I I +

!  ! I I I I  !

! I I I I I I I I I I I I I I I I I I I I I I  :

0+ +- + + -+--------+ -+--------+----- -+ ---+--------+ -+

0 2 4 6 8 10 12 14 16 18 20 22 Shaft Length (feet)

NOTE: Only pumps for which shaft length data are available are plotted.

Figure 2.1-9 Vertical turbine pumps surviving PGA > 0.18 g, data base of vertical turbine pumps plotted as a function of shaft length NUREG-1030 2-49

a 7

1 Table 2.1-17_; Procedure ~to estimate data base site response' spectra l Estimated' Estimated.

Measured Peak Peak Hori- Scaling . Average _

Ground - Horizontal zontal Ground Factor _ Horizontal:

Motion Ground Accel- -Acceleration for -Acceleration Record erations (two Data Base at Data- Response, 'at Data Site components) Site Base Site Spectrum Base Site-Pacoima 1.25g, 1.24g Sylmar. 0.50g 0.50g 0.50g Dam Converter 1.25g Station Orion 0.27g, 0.14g Valley _0.40g - 0.40g 0.30g Blvd. Generating 0.27g Plant Broadway 0.28g, 0.23g Burbank 0.35g_ 0.35g 0.32g Ave.,- Power 0.28g Glendale Plant Glendale 0.30g 0.30g ' 0'. 27g -

Power 0.28g

~ Plant Milikan 0.22g, 0.18g Pasadena 0.20g 0.20g 0.18g

. Library,. Power 0.22g Cal. Tech. Plant Because of this concern EQE studied selected foreign earthquakes and the 1964-

' Alaska earthquake (magnitude 8.4). The findings and conclusions of this study are covered in an EQE report titled " Investigation of Equipment Performance in Foreign Earthquakes and the 1964 Alaska Earthquake" (EQE, November 1983a). The-Alaska earthquake is of interest mainly because of its long strong motion dura-tion, which may be a characteristic of the larger eastern U.S. earthquakes.

This study was not performed just to collect more detailed dr.ta similar to data already collected. This study was performed, however, to assure it is most unlikely in the future that numerous equipment failures will occur (during earthquakes) which have not been studied by the project. A summary of the report follows.

This study addressed the'same seven equipment types considered in the SQUG pilot program. The study was undertaken in three parts:

(1) A survey of U.S. experts (2) A literature survey of equipment performance in the 1964 Alaska i earthquake (3) A literature survey of equipment performance in foreign earthquakes.

NUREG-1030 2-50

{ _, .':*. .* f N

2 c-  % *'"*".

.~ ..

o i;l N . .

rri . ....

G) hsN,,, ,_ Scale in H11es

~

o ,) -

- l

- 9 i aa 5 s #

W

-. , ' EPICENTRAL AREA

~~'(E -- - ". . . '

t.

• MAIN SHOCK 2/9/71 e Record & peak

'j . , . .-

, hortrontal ground acceleration c., y N. 4[

$ Visited Isc111ty x

10tivt v1 Y " 88 , 'kPACOIHA dan O Other Factitty sylstAu ce>NVEltTLR . ,

l. g/l.24g of Interest

,y,,, ,

STATinte 0.50g ,

.n I ' 's3,,,,4 a % f,,g,,g, '

. ,0A - ( <

[si :ici st(test c stalloal i \ ,1

• f, - f> =r f, " . Y . C . *

, a

. -$ya ~

\

s val.I.EY STEAM *

1. s N

""~

.*- )_ 4 s r A Pl. ANT 0.30g

'^

) s.g\3 , i g

l11u casscorts ,

• h.e .. q ,

g\s. -Q

• i ORION BIND. ,

' i ' s" ev [ .1 l ..' * ~ h.g

, O.27s/0.I's .... / i !_ .. I - ly,'f,'ll',

m

-te J .

. . i ,e -

. auRi m e

,/; , -::"~

', {I lin uvis' '[ ' m'~"*

c ( ' ;' i Ti -

, l .,,. . ..- - -

' "ff:

• ]: ,j D HUMBANK rtMJElt s Cl.ENI)Al.E PtNER

-j ;

)

ls~ I rl. ANT 0.32g i} Pt. ANT 0.27g . y, , .j g,.

. A 5 ./ q,p= 0.28g/0.23s -'*"

( 'g;t'., *- ij . . .

(.'

, ~., ,i rAsAtaNA riutu i

',. / ( , ift;T . . t .

y,, rs u r 0.IBa I

ib  ! (*.*{: r g \ '

I'#

.l m -. 1 m i

, s if I ,

* ""i M.... .,, Y )fia u .iI P

1

/ HOLLYWOOD '

,\. - stususu .j )

! .J1 4 l,' :l....'l ,,J I .. . g b L-.. l 4 zji i. *

. .<>> h . 3 ": _ ' i-Figure 2.1-10 Location of the San Fernando Valley data base sites and the ground motion records which are the basis for the estimated average peak horizontal ground accelerations

3.0 -

2.8 [

? 2.6 --

l i s A i a e 2.4 -

I dl 0

e 2.2 2.0 l

'h $l'\ i a

i I

i ii i b H76W component II4W zea i

l)3'f7 i I i i i r, ^ ,' ', n76W Average i.24 1.245

/ t s

^*c' e 1,8 --

lnl g \ r 1st 't' r i V .,Y u s oc

1. \

E {,h .. I I V '

g d

I'4 , ,,

\

v / sN \s 's t i i 1.2 - S16E 's

,s ll _ ___ _ _ _ __ __ __ _ _ _

v 0 1.0 -

NI'

= n ia 0.8 -

WI G 0.6 --

h]IV 0,4 -)

0.2 -1 0,0 l'  :  :  :  :  :  :  :  :  :  :

0 4 8 12 16 20 24 '28 32 36 41 Frequency,Hz Figure 2.1-11 Response spectra for the ground motion record at Pacoima Dam, 5% damping, two horizontal components and their average

3.0 --

! 2,8 --

5 a 2,6 --

o n a e 2,4 --

c 2,2 --

e 1

2.0 --

e 1,8 --

P 1,6 --

a t I'4

i 1,2 --

T 0 1.0 --

\

= n 4 0,8 --

ZPA = 0.50g 0,4 -

N 0,2 -

y 0,0  :  :  :  :  :  :  :  :  :  :

0 4 8 12 16 20 24 28- 32 36 41 Frepency,Hz l Figure 2.1-12 Average horizontal response spectrum, Sylmar Converter Station, 5% damping, based on the ground motion record at Pacoima Dam, scaled by a factor of 0.50/1.25 i

I

2.1.4.3.1 Survey of U.S.' Experts In this survey, 36 U.S. experts with expertise and experience in seismic capability of the equipment under consideration was developed. A questionnaire was prepared and mailed to all the experts, inquiring whether they have seen or are aware of failures of any of the equipment of concern in past earthquakes. A sample questionnaire is included as Figure 2.1-13. The questionnaires were followed up with telephone calls. When an expert returned the questionnaire indicating having seen or having knowledge of equipment failures, the indicated failures were researched further by EQE; most of the reported failures were for equipment other than those seven types included in the scope of this survey (such as substation equipment, ceramic failures,etc.)

or were failures of equipment without adequate anchorage or no anchorage at all. Few failures were found which were relevant to the A-46 program.

In general, the experts expressed a very strong interest in the survey. They also typically expressed a strong lack of confidence in the completeness of their observations. 'Most did not look at anything more than the performance of anchorages and attachments, such as pipes, nozzles, and valve operators.

All of the experts were specifically asked if they were aware of severely damaged internals of equipment in the seven classes. Typically they were unaware of such damage and felt that they would have probably been informed at the time by the owners / operators of the equipment or would have certainly noticed gross failures. Several of the experts from California utilities had specifically sought such information. Their comments have also been included in the findings of the survey.

EQE concluded that, on the basis of the returned questionnaires and the conver-sations with the experts, no major failures of the seven types of equipment in past earthquakes have been uncovered.

2.1.4.3.2 Literature Survey of Equipment Performance in the 1964 Alaska Earthquake In this literature survey, EQE reviewed available investigative reports on the 1964 Alaska earthquake. Specifically, the effects of the earthquake on the power and industrial facilities were studied; any reports of equipment failures were noted. In addition, EQE made telephone inquiries to personnel from power plants affected by the 1964 earthquake.

The Alaska earthquake occurred at 5:36 p.m. AST on March 27, 1964. It had a Richter magnitude of about 8.4. The energy release was probably about twice that of the 1906 San Francisco earthquake. The quake produced shaking that lasted, at some locations, for 3 to 4 minutes with two strong segments caused by at least 6 separate earthquakes. The damaging portion of the earthquake is estimated to have lasted about a minute. No strong motion records were obtained.

What is known about the nature of the earthquake was obtained or deduced from witnesses, damage investigations, and seismographic information. The peak ground acceleration at Anchorage was estimated to be about 0.2 g. The quake was caused 1 by faulting along a plane extending from Kodiak Island to the vicinity of Valdez. The epicenter was in the Prince William Sound area and the main energy release was somewhat south of Montague Island.

NUREG-1030 2-54 l

QUESTIONAIRE

1. Have you visited any of the areas affected by the following earthquakes?

^

Earthquake Yes No_

1964 Alaska.(Prince William Sound) 1972 Managua, Nicaragua 1976 Friuli, Italy 1978 Miyagi-Ken-0ki, Japan 1980 Campania-Basilicata, Italy 1983 Akita Prefecture, Japan

2. Have you visited any other earthquake areas? Please list:

3. In the areas you visited, did you survey any power plants or facilities housing the equipment of interest (see Question 4 for a list)?

Yes No If "Yes", how many facilities surveyed?

4. Have~you seen or are you aware of failures due to earthquake of any of the.following equipment types common to power plants?

Any Failures?

Equipment Yes N_o_ Where?

Motor control centers Low voltage (480 V) switchgear Metal-clad (2.4-4 kV) switchgear Motor-operated valves Air-operated valves Horizontal pumps and motors Vertical pumps and motors

5. Please provide, to the extent possible, a description of the failures if you marked "Yes" for any of the equipment listed in Question 4 above.

References to published material would be very helpful. If uncertain, please tell us all you can; we will search for the information, j

Figure 2.1-13 l

NUREG-1030 2-55

- - - _ . . . . . - _s ..

In spite of the great magnitude and destructive power of the earthquake, the number of communities and facilities affected was relatively small because the event occurred in a region with a population of only 140,000. It did cause considerable destruction and 131 people died,115 in Alaska and 16 in British Columbia, Oregon, and California. A major portion of the damage was due to landslides and soil subsidence. Further damage was caused by tsunamis, particu-larly at Valdez, Seward, and Kodiak, in Alaska, and in Crescent City, Califor-nia.

Some facilities were directly damaged by vibration, which, although sig-nificant, was small compared with that from other causes. Many large and small buildings were severely damaged or collapsed as a result of the vibratory effects of the earthquake. Some of these buildings were presumably well l designed and constructed.

Eight significant power facilities (all relatively small) and many minor facil-ities in the area were affected by the earthquake. The eight facilities were:

(1) City of Anchorage gas turbine plant in Anchorage (2) Chugach Power Plant at Knik Arm (3) Fort Richardson Heat and Power Plant (on northeast outskirts of Anchorage)

(4) Elmendorf Field Heat and Power Plant (on northwest outskirts of Anchorage)

(5) Bureau of Reclamation Eklutna hydroelectric plant at the end of Knik Arm (6) Chugach Bernice Lake gas turbines near Kenai (7) Chugach Cooper Lake hydroelectric plant 25 miles from Seward (8) Port of Whittier Heating and Power Plant in Whittier.

The performance of these facilities was summarized by F.F. Mautz (EQE, November 1983a) reporting on his April 22-28, 1964, visit to Alaska:

In summary, electric generating and distributing facilities in Alaska withstood the earthquake quite well. Except for very local damage to certain plants none of it was severe enough to cause the plant to be out of action for any length of time. All plants continued to operate for some length of time during and after the earthquake until forced off the line, generally by some circumstance or problem outside of the plant proper. In one case damage was due to earth sliding rather than earthquake shaking proper. All plants were quickly restored to operating condition, and at the time of inspection all were in full operating capability, even though emergency repairs were still being carried out in some cases.

In Appendix B the purpose of each of the eight electric power facilities listed above is given.

The reported instances of failures in Alaskan power plants related to the seven basic types of equipment being reviewed in the SQUG project are the following:

(1) Control panel at Elmendorf Air Force Base power plant upset. It had been anchored with one 1/2-in. bolt in each corner (NAS,1973, p. 946).

(2) Air-operated valves at Elmendorf Air Force base power plant opened auto-matically on four tanks when an instrument air control line broke. This resulted in the loss of 60,000 gallons of treated boiler feedwater (NCEL, June 26, 1964, p. 4).

NUREG-1030 2-56 s

(3) Motor starter circuit breaker damaged at Fort Richardson (NAS, 1973,

p. 402).

(4) A fan motor starter and two small motors burned out at Fort Richardson Power Plant. " Motor burnouts were also reported in other buildings where motors were damaged by falling debris but most burnouts were probably caused by motor-starting under low-voltage conditions when power was restored after the earthquake." (NAS,1973, p. 402).

(5) A control cabinet at the City of Anchorage gas turbine plant tore loose at floor anchors and fell over (NAS,1973, p.1053, & DC,1967, p. 29).

There were other failures in the plants and substations which caused problems, e.g., broken water lines, an ash hopper which fell, damaged buildings, and toppled regulators and transformers. All of these items had to be repaired before operations were restored to normal.

-Some equipment failures were reported in buildings and facilities other than power plants:

(1) A control panel fell over at Elmendorf AFB hospital. It was not bolted down (NAS, 1973, p. 353).

(2) Relays were damaged at an L Street apartment building when a selector panel door on which they were mounted swung open (NAS, 1973, p. 353).

(3) A circuit breaker panel at Fort Richardson Barracks failed when the copper-bronze bolts failed on a copper bus (NAS, 1973, p. 399).

(4) Control panels toppled over at various places (unnamed). They were not anchored (NAS, 1973, p. 400).

(5) Valve (hand operated) opened slightly as a result of shaking at Whittier Union Oil Company West Camp area. This resulted in 3,000 barrels of leakage through a broken pressure relief line (NAS,1973, p.1101).

2.1.4.3.3 Literature Survey of Equipment Performance in Foreign Earthquakes The purpose of the study was to investigate, from a review of the literature, the performance of power plant equipment in some significant earthquakes in foreign countries. The study is concerned specifically with major failures of the seven types of equipment currently considered by the SQUG program.

Four significant earthquakes were studied: 1972 Managua, Nicaragua (magni-tude 6.2); 1976 Friuli, Italy (magnitude 6.5); 1978 Miyagi-Ken-Oki, Japan (magnitude 7.4); and the 1980 Campania-Basilicata, Italy (magnitude 6.8). For each of these, the earthquake characteristics, ground motion records, if any, and the general effects of the earthquake are reviewed. The effects of the

- earthquakes on power and industrial facilities are studied in more detail.

Sunnary of this study is presented in Appendix C.

The study did not uncover any mass failures of the equipment of interest.

Reports of failures were found; however, most of the reported damage for equip-I ment was attributable to anchorage failures. Those few reports of failures that would concern the SQUG project are highlighted in the report.

l

. NUREG-1030 2-57 1

1 2.1.4.3.4 Conclusions and Staff Comments on Alaskan and Foreign Earthquakes The purpose of this effort was to determine if earthquakes not studied in the SQUG pilot program have caused equipment failures pertinent to A-46.

The study failed to discover significant failures.

Through the survey of experts and the literature review studies, some reported failures were discovered. However, such failures were very few and did not indicate any trend. Whenever detailed information about the failures was avail-able, such information was recorded.

Descriptions given in most reports were found incomplete, and in some cases, contradictory to other reports. This is mainly because, especially in the in-vestigations of the earlier earthquakes, the investigators were rarely inter-ested specifically in equipment performance other than anchorage failures.

Also, very few reports are available specifically on equipment performance.

Details of reported failures can be obtained only through onsite investigations.

2.1.5 Sumary of SSRAP Report, "Use of Past Earthquake Experience Data to Show Seismic Ruggedness of Certain Classes of Equipment in Nuclear Power Plants" The SSRAP completed its study and made an oral presentation to the NRC staff and SQUG/EQE on December 15, 1983, to outline its conclusions and recommenda-tions on the use of non-nuclear seismic experience data. The SSRAP conclusions and recommendations were documented in a report titled "Use of Past Earthquake Experience Data to Show Seismic Ruggedness of Certain Classes of Equipment in Nuclear Power Plants," which was published in January 1985 (SSRAP, January 1985).

The SQUG pilot program studied only seven classes of equipment. SSRAP, after its study, concluded that there are adequate data on " unit substation trans-formers" and included them in their recommendations.

The SSRAP assessment was primarily based upon past earthquake performance data provided to SSRAP by the Seismic Qualification Utility Group (SQUG) through its consultant, EQE Incorporated. Detailed reviews were conducted by EQE on the performance of the eight classes of equipment at:

(1) Several conventional power plants (Valley Steam Plant, Burbank Power Plant, Glendale Power Plant, and Pasadena Power Plant) and the Sylmar Converter Station subjected to the 1971 San Fernando earthquake (magnitude 6.5).

(2) The El Centro Steam Plant subjected to the 1979 Imperial Valley earthquake (magnitude 6.6).

(3) Pumping stations and refineries near the 1983 Coalinga earthquake (magni-tude 6.5).

In addition, much more limited reviews were conducted at several electrical substations for the 1971 San Fernando earthquake, the Ormond Beach Plant and o;ie substation subjected to the 1973 Point Mugu earthquake (magnitude 5.9), and at the Ellwood Peaker Plant and the Goleta Substation subjected to the 1978 Santa Barbara earthquake (magnitude 5.1). Limited literature reviews searching for reported failures of equipment in these eight classes were conducted for NUREG-1030 2-58

the 1964 Alaska (magnitude 8.4),1952 Kern County (magnitude 7.4),1978 Miyagi-Ken-0ki, Japan (magnitude 7.4),1976 Friuli, Italy (magnitude 6.5), and 1972 Managua, Nicaragua (magnitude 6.2), earthquakes.

Some of this work was initiated at the request of SSRAP and all of this work was carefully reviewed by SSRAP.

All members of SSRAP performed walk-throughs of the Sylmar Converter Station, Valley Steam Plant, and Glendale Power Plant, and the SSRAP members spoke with operators present at the Sylmar Converter Station and the Glendale Power Plant during or shortly after the 1971 San Fernando earthquake. In addition, at least one SSRAP member is familiar with equipment in these eight classes at the El Centro Steam Plant and at some of the pumping stations and refineries used in the Coalinga data base. All members of SSRAP have conducted walk-throughs of at least three different types of nuclear power units for the pur-pose of reviewing these eight classes of equipment. Several members have con-ducted similar walk-throughs of many additional nuclear power plant units. The purpose of these walk-throughs was to judge similarity between the equipment in nuclear power plants and that in the conventional plants from which past earth-quake experience data were collected. SSRAP and vendors of some of these classes of equipment discussed the similarity between equipment installed in nuclear plants and equipment in conventional plants. A partial list of the material reviewed by SSRAP is given in the bibliography to the report (SSRAP, February 1984). Lastly, SSRAP relied on the extensive collective experience of its five members with these eight classes of equipment.

After a detailed and careful review of the full range of the available experi-ence data base, the SSRAP conclusions for these eight classes of equipment are:

(1) Equipment installed in nuclear power plants is generally similar to and at least as rugged as that installed in conventional power plants.

(2) This equipment, when properly anchored, and with some reservations (to be discussed later), has an inherent seismic ruggedness and a demonstrated capability to withstand significant seismic motion without structural damage.

(3) For this equipment, functionality after the strong shaking has ended has also been demonstrated, but the absence of relay chatter during strong shaking has not been demonstrated.

Therefore, with several important caveats and exclusions as discussed below, it is SSRAP's judgment that for excitations below the defined seismic motion bounds, it is unnecessary to perform explicit seismic qualification of existing equipment in these eight classes for operating nuclear power plants to demon-strate functionality after the strong shaking has ended. The existing data base reasonably demonstrates the seismic ruggedness of this equipment up to these seismic motion bounds. Secondly, it only applies to functionality after the strong shaking has ended. Third, there are exceptions as denoted in sub-sequent sections. Fourth, the conclusion is only applicable to those eight classes. However, SSRAP believes that similar conclusions might be applicable for some other classes of equipment, but such an extrapolation should only be made after a very detailed and careful review.

NUREG-1030 2-59

The datd base is inadequate to preclude the possibility of an inadvertent change of function (breaker trip, etc.) oue to causes such as relay chotter.

This does not mean that SSRAP expects these problems to occur. It simply medns that their preclusion has not been demonstrated by the availdble dato base. The data base does demonstrate the breakers can be properly reset ond the equipment functions properly ofter the earthquake.

SSRAP is particularly concerned with equipment anchorage and feels thot any dttempt to justify equipment for aCCeptdble seismic performance must ensure adequate engineered anchorage. There are many exemples of equipment sliding or overturning in earthquake exposure because of no anchorage or inadequate anchorage. Inddequate enchorage con include short, _ loose, or poorly installed bolts or expansion enchors, and improper welding or bending of-sheet metal frames et anchors. SSRAP believes thdt equipment enchorage must not only be strong enough to resist the anticipated forces but also stiff enough to prevent excessive movement of the equipment end potentidl resonant response with the structure. It is SSRAP's opinion that any review program should include con-sideration of both strength and stiffness of the anchorage and its component parts.

Excluded from essessment in the SSRAP study are air, power, fuel, and cooling systems which connect to the equipment.

2.1.5.1 Seismic Motion Bounds SSRAP uses three different seismic motion bounds (Type A, B, and C) in its report. These bounds are defined in terms of the 5% damped horizontal ground response spectro shown in Figure 2.1-14. The seismic motion bounds may be used for the equipment class as defined in the table that follows.

These spectra bounds are intended for compdrison with the 5% domped design horizontal ground response spectrum et a given nuclear power plant. Alter-ndtely, one mdy compare 1.5 times these spectrd with a given 5% damped horizontal floor spectra in the nuclear plant.

The comparison of these seismic bounds with design horizontal ground response spectro is judged by SSRAP to be acceptable for equipment mounted less than 40 feet above grade (the top of the ground surrounding the building) and for moderately stiff structures (fundamental frequency gredter than 2 Hz). For equipment mounted more then 40 feet dbove grade, comparisons of ,1.5 times these spectro with horizontal floor spectra is necessary. In all coses a comperison with floor spectra is acceptable.

The criteria dre met so long as the 5% damped design horizontal spectrum lies below the appropriate bounding spectrum et frequencies greater then or equal to the fundamental frequency range of the equipment.

2.1.5.2 Motor Control Centers On the basis of a review of the data bdse and onticipoted variations in condi-i tions. SSRAP is of the opinion that motor control centers are sufficiently rugged to survive a seismic event and remdin operational thereafter, provided the following conditions exist in the nuclear facility:

NUREG-1030 2-60 l

~ __- ._ _ _ . .

Seismic Motion Bound Equipment Class Seismic Motion Bound Derived From*

Motor control centers Type B Sylmar Converter Low-voltage (480-V) switchgear Station (San Fernando Metal-clad (2.4 to 4-kV) switchgear earthquake)

Unit substation transformers Motor-operated valves with Type C Valley Steam Plant and large eccentric operator Burbank Power Plant lengths to pipe diameter (both for San Fernando ratios (see Figure 2.1-14) earthquake)

Motor-operated valves Type A El Centro Steam Plant (exclusive of those with (Imperial Valley large eccentric operator earthquake) lengths to pipe diameter ratios) Pleasant Valley Pumping Air-operated valves Plant (Coalinga Horizontal pumps earthquakes)

Vertical pumps

• Based on smoothed averaged horizontal ground 5% damped response spectra from actual ground motion records divided by 1.5.

(1) Motor control centers of the 600-V class (actual voltage is 480 V) are considered. The style of cabinets must be similar to those specified in NEMA Standards. This requirement is imposed to preclude unusual designs not covered in the data base. SSRAP feels that cabinets which are styled after NEMA Standards will perform well if they are properly anchored.

Cabinet dimensions and material gauges need not match NEMA Standards.

(2) The cabinets have engineered anchorage. Both the strength and stiffness of the anchorage and its component parts must be considered. Stiffness can be evaluated by engineering judgment on the basis of the cabinet con-struction and the location and type of anchorage, giving special attention to the potential flexibility between the tiedown anchorage and the rigid walls of the cabinet. Adequate stiffness can also be shown by determining that the fundamental frequencies of the anchored cabinet under significant shaking in both horizontal directions is above approximately 8 Hz. It is the opinion of SSRAP that properly anchored cabinets will have a fundamen-tal frequency greater than about 8 Hz.

(3) The intent of this requirement is to ensure that under earthquake excita-tions the natural frequency of the installed cabinet will not be in resonance with both the frequency content of the earthquake and the fundamental frequency of the structure.

(4) Cutouts in the cabinet sheathing are less than 6 inches wide and 12 inches high.

NUREG-1030 2-61

.n

1.2 i i i i i I

$ 1.0 -

5% DAMPING -

z 9

l-

< 0. 8 -

TYPE B A w

00.6 -

U

< TYPE A & B J

TYPE C y 0.4 -

F O.33 0

W

@ 0.2 0.20 0.0 I I ' I ' '

O 4 8 12 16 20 24 28 FREQUENCY (Hz)

Figure 2.1-14 Seismic motion bounding spectra (5) All internal subassemblies are securely attached to the motor control cabinets which contain them.

(6) Adjacent sections of multi-bay cabinet assemblies are bolted together.

(7) Equipment and their enclosures mounted externally to motor control center cabinets and supported by them have a total weight of less than 100 pounds.

SSRAP does not consider the functionality, that is, inadvertent changing or failure to change state on command of relays during an earthquake. The func-tionality must be established by other means. The structural integrity of relays contained in the motor control centers and their ability to function properly after earthquakes, as defined in Section 2, has been demonstrated.

2.1.5.3 Low-Voltage Switchgear Low-voltage switchgear of the 600-V class (actual voltage is 480 V) is consid-ered. The style of cabinets must be similar to those specified in ANSI C37.

This requirement is imposed to preclude unusual designs not covered in the data NUREG-1030 2-62

+ _ _ ._

base. SSRAP feels that cabinets.which are styled after ANSI Standards will perform well'if they are properly anchored. Cabinet dimensions and material gauge need not match the ANSI Standard.

All the conclusions, limitations, and bounding spectra for motor control centers are applicable to low-voltage switchgear.

2.1.5.4 Metal' Clad Switchgear Metal-clad switchgear of 2.4 kV and 4.16 kV is considered. The style of cabi-nets must be similar to those specified in ANSI C37 Standards. This requirement is imposed to preclude unusual designs not covered in the data base. SSRAP feels that cabinets which are styled after ANSI Standards will perform well if they are properly anchored. Cabinet dimensions and material gauges need not match NEMA Standards.

All the conclusions, limitations, and bounding spectra for motor control centers are applicable to metal-clad switchgear, except that the cutouts in the cabinet sheathing shall be less than 12 inches by 12 inches.

2.1.5.5 Unit Substation Transformers Unit substation transformers convert the distribution voltage to low voltage.

Unit substation transformers which convert 2.4-kV or 4.16-kV distribution voltages to 480 V are considered.

On the basis of a review of the data base and anticipated variation in condi-tions, SSRAP is of the opinion that unit substation transformers are suffi-ciently rugged to survive a seismic event and remain operational thereafter, provided that in the nuclear facility both unit substation transfonner enclo-sures and the transformer itself have engineered anchorage.

The functionality of properly anchored unit substation transformers during and after earthquakes, as defined above, has been demonstrated.

2.1.5.6 Motor-0perated Valves On the basis of a review of the data base and anticipated variations in condi-tions, SSRAP is of the opinion that motor-operated valves are sufficiently rugged to survive a seismic event and remain operational thereafter, provided the following conditions exist in the nuclear facility:

(1) The valve housing and yolk construction is not of cast iron.

(2) The valve is mounted on at least a 2-inch pipe.

(3) The actuator is supported by the pipe and not independently braced to or supported by the structure unless the pipe is also braced, immediately adjacent to the valve, to a common structure.

The limitations on operator weight and eccentric length relative to pipe diameter are derived from the data base for motor-operated valves that was provided by SQUG. The data base contains relatively few heavy operators and NUREG-1030 2-63 a_ - m

small pipe diameters subjected to severe ground shaking. These limitations.

could be less restrictive if more motor-operated valves had been located and ~

. documented-in the areas of. higher shaking. It is felt that additional data, either from other earthquake experience or seismic qualification tests', can expand the scope of these recomendations. These limitations are shown in Figures 2.1-15 and 2.1-16.

Recent data from the Chilean earthquake (March 3, 1985), however, did add about 120 MOVs to the data base inventory, including a number of motor operators on smaller lines. In addition, shake table test data is being compiled for motor operators through the EPRI - sponsored program. Therefore, it is expected that the limitations shown in Figures 2.1-15 and 2.1-16 will be

. relaxed somewhat eventually.

For motor-operated valves not complying with the above limitations, the seismic

. ruggedness for ground motion not exceeding the Type A bounding spectrum may be demonstrated by static tests. In these -tests, a static force equal to -three times the approximate operator weight shall be applied non-concurrently in each of the three orthogonal principal axes of the yoke. The limitations other than those related to the operator weight and distance from the top of- the operator to the centerline of the pipe, given above, shall remain in effect.

2.1.5.7 Air-0perated Valves On.the basis of a review of the data base and anticipated variations in condi-tions, SSRAP is of the opinion that air-operated valves are sufficiently rugged to survive a seismic event and remain operational thereafter, provided the fol-lowing conditions exist in the nuclear facility:

(1) The valve ho'using is not of cast iron.

(2) The-valve is mounted on a pipe of 1-inch diameter or greater.

(3) Limitations on pipe dianeter versus distance from centerline of pipe to top of operation are shown in Figure 2.1-17.

(4) The actuator is supported by the pipe and not independently braced to the structure or supported by the structure unless the pipe is also braced imediately adjacent to the valve to a common structure.

The air line and its connection are not included in this assessm nt.

For air-operated valves not complying with the above limitations, the seismic ruggedness for ground motion not exceeding the Type A bounding spectrum may be demonstrated by static tests. In these tests, a static force equal to three times the approximate operator weight shall be applied non-concurrently in each of the three orthogonal principal axes of the yoke. The limitations other than those related to the distance of the top of the operator to the centerline of the pipe, given above, shall remain in effect. However, the recent Chilean carthquake of 1985 did add more data on air-operated valves to the data. base inventory. There is a possibility that the limitations on the air-ope' rated valves mentioned above will be relaxed somewhat in view of this. .

2.1.5.8 Horizontal and Vertical Pumps NUREG-1030 2-64 ,

n - - -

~

i i e

y OUTSIDE EXPERIENCE DATA j v WITHIN EXPERIENCE DATA E

liw j100 -

h ;rs50p 8

8 n.

h 60 - --- ~-

"R ~- ;

h $[4dOk w Ele i  !!l-llll 8 lili w qs

! iflf tc sit

>- ($

5 ll?

o  ;;;l:!! in.g .y 3: -  :::: .

12 24 PIPE DIAMETER (inches)

' APPROXIMATE MAXIMUM OPERATOR WEIGHT Figure 2.1-15 Motor-operated valves for which Type A spectrum is to be used SSRAP feels that horizontal pumps in their entirety, and vertical pumps above their flange are relatively stiff and very rugged devices because of their inherent design and operating requirements. However, the applicability of the data base is subject to the limitations set forth below.

For horizontal pumps, one must ensure that the drive (electric motor, turbine, etc.) and pump are rigidly connected through their base so as to prevent damag-ing relative motion. Of concern are intermediate flexible bases; these must be evaluated separately. Proper horizontal thrust load capacity must also be ensured in both axial directions. The data base covers pumps up to 2500 hp.

However, SSRAP feels that the conclusions are equally valid for horizontal pumps of greater horsepower.

For vertical pumps, the data base has many entries up to 700 hp and several up to 6000 hp; however, SSRAP feels that safety-related vertical pumps, above the flange, of any size are sufficiently rugged to meet the Type A bounding spectrum.

NUREG-1030 2-65 n-- m

i i a i  : i 9

2 O OUTSIDE EXPERIENCE DATA

• WITHIN EXPERIENCE DATA k

$ 100 -

8

$ ?kYh$

l 8 70' _

amagwagw:qw--  !

n.

O O

w 50 -

se

"" 40 -

j 650[*E g!A O 100**

m 30 - --

sw E

b 5

.:.:w. . :.:

M f:: S. . . .YN i:s 2 4 6 8 10 12 PIPE DIAMETER (inches) l

• APPROXIMATE MAXIMUM OPERATOR WElGHT Figure 2.1-16 Motor-operated valves for which Type C spectrum is to be used SSRAP feels that the variety of vertical pump configurations and shaft lengths, below the flange, and the relatively small number of data base points in several l categories, preclude the use of the data base to screen all vertical pumps.

l Vertical turbine pumps with cantilevered casings up to 20 feet in length and with bottom bearing support of the turbine to the casing appear well enough represented to meet the bounding criteria below the flange as well. SSRAP recommends either individual analysis or use of another method as a means of evaluating of other vertical pumps below the flange. The chief concerns would be damage to bearings from excessive loads, damage to the impeller from exces-sive displacement, and damage from inter-floor displacement on multi-floor sup-ported pumps.

l 2.1.5.9 Conclusion and NRC Staff Comments General conclusions arrived at by SSRAP after its study of the data base for the eight classes of equipment are summarized in Section 2.1.5.

NUREG-1030 2-66

_ _ _ _ . _ _. b

..-__..A

i i

.E OUTSIDE EXPERIENCE DATA h *? WITHIN EXPERIENCE DATA h '

g <

g .,

tr h

w o 60 -

f. ,

w:

O  :.c. ,

a " gs. s h 45 - N O 'n W k ii.

.  :?Ei.,

32, >j

- . " iM[j j O N  ::!e j  : :h.+

s 3

@ ' 2, '

2 s.... . s m  ::it i:

O . . . . .  :!?V N q ::  :-:: c 1 4 PIPE DIAMETER (inches) s Figure 2.1-17 Air-operated valyes for which Type'A spectrum is to t*e used SSRAP envisions that a seismic review of these items of equipment in an existing nuclear power plant will require a walk-through of the plant (1) to determine which equipment is within the limitations of these recommendations dad, (2) to a evaluate judgmentally other factors that may affect the seismic performance of" the equipment, such as the evaluation of adjacent equipment and conditions to verify that impacts during a seismic event which might dan: age the safety-related equipment are precluded. It is expected that this evaluation will flag for special review any unusual or non-typical conditions such as major modifica-tions to standard equipment or equipment that is unique.

The SSRAP recommendstions are based on experience data thich confirm that the equipment included within the limitations is'rogged enough, to maintain function-ality after the strong shaking has ended. However, it has been brought to the NUREG-1030 2-67

a attention of SSRAP that there apparently have been cases where maintenance per-  !

sonnel have noted increased wear in bearings of vertical pump shafts several l weeks after the earthquake exposure. Because wear of bearings is a normal con- l dition and because these pumps did operate for weeks after the earthquake before i maintenance was required, SSRAP feels that this potential situation is within routine maintenance and not a matter of concern. It is mentioned only as an additional consideration for post-earthquake maintenance checks.

Much of the data base equipment was over 20 years old at the time of the earth-quake exposure and some of this equipment is located in reasonably high thermal and corrosive environments, so the data base undoubtedly does address these

. aspects of equipnent aging. However, none of the data base equipment was ex-posed to radiation, so the aging effects from radiation exposure upon the equipment are beyond the scope of this program.

As part of the data development for this program, literature reviews of several significant earthquakes were conducted to determine if failures had occurred which might contradict the lack of failures within the data base. References in Appendices B and C reported several isolated failures of equipment in the s 1964 Alaskan earthquake and the 1972 Managua, Nicaragua earthquake. The origi-nal reports contain incomplete data, poorly documented, and most failures can conceivably be explained by conditions such as improper anchorage. Nevertheless, because of.the overwhelming evidence in the data base, SSRAP has not altered its conclusions on the basis of these reports, but it is suggested that at-tcmpts be made to determine if more detailed information does exist to properly evaluate these reports.

The conclusions of the SSRAP study have been based largely on the data base that was provided. As previously noted, some items or portions of equipment have been excluded from the scope of the recommendations because of lack of informatidn within the data base. For example, some motor-operated valves as well as the functionality of relays during the earthquake have been excluded.

SSRAP believes that the limitations imposed on some of the eight classes of equipment can be relaxed w!th the use of seismic equipment qualification tests b which undoubte'lly have been performed or could be performed on an industrywide basis, o

SSRAP believes that the approach to equipment evaluation for seismic performance St'ilized in this sttdy can be extended to other classes of equipment. It is recomended that future studies utilize both earthquake experience data as well es seismic equipment qualification test data as appropriate. Each class of

't

quipment must be carefully addressed on an individual basis to consider poten-tial vulnerabilities and appropriate limitations. Although no detailed studies have been performed, SSRAP suggests that the following are examples of classes of equipment that may be amenable to this approach: air compressors, conduit and cable tray raceways, diesel generators, electric motors, fans, heat ex-changers, and HVAC systems. SQUG is currently working on documenting the seismic adequacy of all equipment needed for hot shutdown, which utilizes information from both seismic experience and test data collection (EPRI project, see Section 2.4), and possibly seismic ruggedness developed from

/ other evidences. This study will be reviewed and approved by SSRAP and the

i. NR2 staft before A-46 implementation.

The staff is in general agreement with the SSRAP conclusions, recomended caveati, and exclusions (as outlined in Sections 2.1.5, and 2.1.5.1 to 2.1.5.8).

i/ NUREp-1030 2-68 1

-- m

y-,. ._. m __ , ,

. ~ . .m.

k,r W

!  : 2.2 ' Developmentsand Assessment of In-Situ Testing Methods To Assist 1in-Qualification of Equipment b ~2.2.1 Background v +

j

~

p LThis tssk wasr selected for A-464.because the potential exists that in situ.

. testing can be:a promising toolTin assisting the seismic. qualification.of equip-

~

(

ment in: operating plants. 'The' task i's conducted by Idaho National Engineering Laboratory (INEL), ind _was , started in early 1982. . The:. intent of this task is-to7i nvestigate presenthin-situ. testing methods and.to evaluate the feasibility A of using;these methods to assist-in requalifying equipment, and to develop methods, guidelines, and acceptance criteria for their'use. '

-More'specif'icallyi t'he work. scope for this task consisted of the following topics:

. e

! ' ?;(1) ~ Basic review o'f exisf.ing approaches to in-situ testing and identification

! - of preliminary in-situ test methods for the qualification of equipment in plants which are' currently licensed and operating.

t . U

-(2)L Review of approaches to laboratory testing and simulation of seismic

^

s events .in the , laboratory, for qualification of equiprent. Limitations -on' -

p the'use of current guidance was also studied. 1 i (3) Review of the analysis procedures fundamental to in-situ testing methods.

Review. of use of ~ subcomponent proof test and/or subcomponent fragility tests in the qualification process. . Review of the qualification require -

~

i: - ments for anchors.

(

h (4) --Investigation of techniques'for-assessing / monitoring the effects of

- chemical or metallurgic aging, mechanical fatigue, and wear during plant operation.

(5) Address adequa'cy, limitations.and inherent. shortcomings, and nonconserva-1 tisms of the various approaches above.

,y

-(6) Development of guidelines and acceptance criteria for use of in-situ test- /

, _ ing to-support alternative methods of seismic qualification of safety-

[p related equipment.

.O (7) Definition of requirements for a test data base in support of seismic

qualification of existing equipment in currently licensed operating plants.

L (8) ' Development of cost estimate for alternate seismic qualification methods.

(9) Verification and further development of combined in-situ and analysis-methods suitable for equipment qualification. Examination of limitations and pitfalls of applying in-situ testing methods in determining dynamic characteristics and evaluating component mountings of structures which i support, contain, or position safety-related equipment in operating plants. I Development of guidelines for minimum testing requirements and reporting requirements in qualification documentation.

b NUREG-1030 2-69 j; ,

og --ww-ew y- -

1.e e e  % e m.iyvi - w v ---y- , - yyry

2.2.2 Summary of INEL Report, "The Use of In-Situ Procedures for Seismic Qualification of Equipment in Currently Operating Plants" Results of work on topics 1, 2, 3, 4, 5, 7, and 8 of Section 2.2.1 above are covered in the contractor report titled "The Use of In-Situ Procedures for Seis-mic Qualification of Equipment in Currently Operating Plants" (NUREG/CR-3575)

.(NRC, June 1984). This report is divided into four parts, each of which addresses a specific area. Following is a summary of these four parts.

2.2.2.1 Summary of Part A and Part B, " Preliminary Study of the Use of In-Situ Procedures for Seismic Equipment Qualification in Currently Operating Plants" and " Improved In-Situ Procedures and Analysis Methods" The goal of this study was to examine the most important uses of in-situ testing employed to assist in requalification of safety-related equipment.

Theoretically, in-situ test procedures could be applied in the following three methods:

(1) Testing at full load level with equipment in place.

(2) Low load level testing with equipment in place.

(3) Periodic intermediate or low load level testing to support a continuing surveillance data base.

It is the conclusion of this study that among the three potential methods of in-situ test, cnly method 2 is normally practical and feasible. Method 1, which applies the dynamic load up to the safe shutdown earthquake (SSE) level, has to satisfy certain condities. The required conditions are that:

(1) The motion applied to the equipment-supporting structure should not ex-cessively load the appurtenances, te components mounted thereon or in the vicinity, and the equipment-supporting structure itself.

(2) Sufficient access must exist in order to load the equipment mounting.

(3) No damage occurs to the local area where load is applied.

(4) No significant mechanical aging degradation has occurred during testing, so that component can be employed in service for its nominally useful lifetime.

These conditions severely limit the usefulness of full load level in-situ tests. s Valve operators are one equipment type that have been dynamically qualified I in-situ by using a static load to perform an interference evaluation. However, the potential for performing full load level in-situ testing is so limited that it is not considered further.

Method 3 above could, in principle, be useful for identifying aging degradation.

However, the contractor concluded that for the types of equipment of interest in this program, no potential applications are apparent. This is because NUREG-1030 2-70

changes significant to operability of safety-related equipment (particularly in a seismic environment) can not generally be detected by in-situ procedures.

The low load level in-situ tests are normally performed by applying hammer impact on equipment or supporting structures. Portable electromagnetic or  !

hydraulic shakers can also be applied to equipment or equipment-supporting structures in place, in order to dynamically test them. The input force and output, normally acceleration, are recorded as loads are applied at various positions. The recorded quantities are converted from time histories to a frequency representation by use of the Fourier transform. Using the frequency representation, transfer functions are calculated between points of input and output. These calculations are typically performed with minicomputers which are part of the modal analyzer system. Software internal to these computers then identifies natural frequencies and mode shapes. The mode shapes encompass points on the structure where data were recorded.

The contractor concludes from his study that in-situ testing will be useful in the following areas related to equipment qualification:

(1) establishment of similarity between equipment with consideration of failure modes (2) preoiction of component-specific required response spectra (RRS)

(3) component mounting evaluation (4) comparison of fundamental building frequency with equipment-supporting structure frequency It was also concluded that in-situ testing will not be feasible and suitable for the following applications:

(1) to establish component / equipment seismic capacity (2) to support a continuing surveillance data base The applications of in-situ testing methods is further discussed below. Other related topics covered by this contractor's report are described in Appendix B.

(1) Establishment of Similarity Between Equipment With Consideration of Failure Modes The most obvious application of in-situ testing to seismic qualification of equipment in operating plants is to establish dynamic similarity between pieces of equipment.

As mentioned in Section 1.2, after reviewing the results of all the tasks of A-46, the NRC staff concluded that seismic qualification using seismic experience data probably is the most likely approach to develop a quali-fication method which is both economically attractive to the plant owners and would be acceptable from a public safety viewpoint. Two conditions will have to be established before the experience data base can be utilized NUREG-1030 2-71

to help assess ~ seismic adequacy of equipment in operating plants. They are:

(a) 'To establish that RRS of equipment in operating plant to be re-qualified is enveloped by the pertinent experience data base response spectra.

(b) To establish similarity between operating plant equipment to be requalified and equipment in the experience data base.

Condition (a) is addressed by No. 2 (immediately following) and also by Section 2.5. The staff's position on the definition of similarity was described as "for equipment to be similar for the purpose of qualifying an equipment item on the basis of experience data on another item, the safety function as well as the dynamic characteristics, should be similar.

This means that the experience data must include data on performance both during and after a seismic event. Similarity parameters must include mass distribution, material, size, stiffness, configuration, restraints, and anchorage details...."

Similarity of dynamic characteristics can most effectively be addressed by conducting an in-situ test. Dynamic-characteristics of equipment consist of mode shapes, natural frequencies, mass distribution, and damping. In-situ procedures identify the natural frequencies and mode shapes. In certain cases the mass distribution can also be estimated (alternate methods for determining the mass distribution are proposed by the contractor in his reports). It is also possible to characterize viscous damping by using in-situ tests that represent the damping that actually occurred during the test. Since damping may depend on response level, the contractor proposed that values obtained from low level in-situ tests may not necessarily be valid and Regulatory Guide 1.61 (NRC) is recommended for damping values.

The safety function aspect (operability and failure modes) of similarity is further discussed in paragraph 1 of Appendix B.

(2) Prediction of Component-Specific RRS In order to seismically qualify a piece of equipment, it is first neces-sary to establish the specific RRS. For equipment mounted on a floor, the response can be predicted by the floor response spectra. However, because safety-related components are mounted on or attached to the equipment-supporting structuros (such as electrical cabinets, racks, etc.), the RRS for these components will be different from the floor response spectra.

In situations like these, three methods are studied and proposed by the contractor to establish component-specific RRS. Each method will utilize in-situ testing to a different extent.

(a) The first approach is to develop a finite element computer model of the equipment-supporting structure and the mounted equipment. The analysis procedures involved here are those of the typical time history method. In this process, (i) a synthetic time history is developed from a specific floor response, (ii) the modes, frequen-cies, and modal participation factors are calculated from the model, i

NUREG-1030 2-72

_ . - _ . . ~ . . _ _ _ . __. . - _. __

E>

n (iii) a: time history analysis is performed on each significant mode,

. (iv) the modes are algebraicly combined to determine total time _

histories, and (v).'the time histories are converted to RRS-for the

components of-interest. The contractor feels that this basic proce-

[ ' dure is potentially unreliable because the system is complex and boundary condition modeling is unreliable. Consequantly, it can only be used -if the equipment is already installed and in-situ proce- ,

dures are used to verify the calculated modal parameters. A major l disadvantage of the approach is that it is relatively expensive because of the cost associated with. developing a finite element model. An advantage is that if minor equipment modifications are made at a later date, the model can be updated and a new set of RRS

+

can be calculated.

'(b) The second-method to-generate component-specific RRS is an analysis method by utilizing modal parameters directly. .The process involves using the frequencies and mode shapes determined from in-situ proce-

,. dures directly in constructing a numerical solution. In this approach, the modal participation factors can either be estimated by using the definition for the modal participation factor and approximating-it with discrete mode shape and modal mass, or using an approach which

. is based on reconstructing the force vector using the significant modes of the structure. The second method is judged by the contrac-tor to provide the best possible estimate of.the modal participation

, factor and is recommended by the contractor. When using this method to generate the RRS, there.is no need to develop a finite element

~

j

. model. As.with the finite element approach, the response of indivi-j dual modes is calculated and then superimposed for the total response.

The contractor offered several comments about using this method.

i. First, as the natural frequency increases it becomes more difficult for in-situ procedures to resolve the associated mode shapes. For seismic analysis it is felt that higher modes, or modes with several antinodes will result in low or negligible modal participation fac-

~

tors. Consequently, it will probably only be necessary to accurately calculate the lower mode shapes. The situation must be checked for every individual case. The second comment concerns closely spaced j' modes. The decomposition of the total frequency response into a modal frequency response function is one step in the development of 4 the mode shapes. Closely spaced mode shapes reduce the accuracy with l' which the modal frequency response functions are' calculated from the experimental transfer functions. The existence of closely spaced significant modes could render the direct use of modal parameters .

infeasible. It is anticipated that this situation will occur infre-quently in which case the alternative of method "a" above can be used to determine RRS. The advantage of the direct use of modal parameters is that the' modal parameters are relatively inexpensive to generate experimentally. Generation of modal parameters by the finite element L method will require substantially more expense.

t (c) The third method involvec response spectra transfer based on the application of random vibration theory. When applied to seismic

, environments, this normally implies that the mean square response is

!. used as the basis for predicting peak response values. The applica-tion of random vibration theory to a particular process is simplified l

I

' NUREG-1030 2-73 f

,-m,~ -..m, _ _ , , . . . . ,,-.,_-,,_,.v----,.,-.,,_-,---,.--,_m,,-m-

- - _ . - . . . - - ---,.m- _ _ . . , - , ,,_,_,,.,._m---e ,..- m-,_,.,,

if the process is Gaussian, zero mean, and stationary, because power spectra density (PSD) function completely defines the process under-these restrictions. The contractor suggested that earthquakes are Gaussian in character because of broad frequency content and the random phasing of the frequency components, and they are obviously zero mean. Furthermore, the contractor suggested that earthquakes may be considered as a finite duration segment in a stationary process and corrections can be applied to structural response for the non-stationary effect of duration. Under these conditions, the statisti -

cal properties of the output can, in theory, be inferred from the input using the properties (natural frequencies, mode shapes, modal participation factors, and modal dampings) of the intervening structures.

One difficulty in seismic analysis arises from the structural motion starting from zero initial conditions. Correction factors must be used to correct for the differences between steady state response and response from realistic initial conditions.

On the basis of the above discussion, the contractor proposed a procedure for response spectra transfer using random vibration theory. The recom-mended procedure is to develop a response-spectrum-consistent PSD using an appropriate correction for duration, calculating the output PSD includ-ing the effects of all cross-modal terms and multiple directions of ex-citation through the use of transfer functions, integrating this PSD to determine the mean square response, and finally determining the response spectrum value from the root mean square response and an appropriate peak value factor. Details of the procedure are described in the INEL report of October 1983, " Improved In-Situ Procedures and Analysis Methods for Seismic Equipment Qualification in Currently Operating Nuclear Power Plants."

(3) Component Mounting Evaluations Mounting inadequacy has been a major cause of retrofit and retest in qualification programs. The current qualification process essentially

. qualifies mountings during shake table testing. For operating plants several options are available. Analysis procedures using data from in-situ testing can predict the maximum acceleration of equipment. Thus, the loads that mountings must transmit can be predicted. It should be a straightforward process to assess existing designs. The main distraction is the large number of mountings that exist. Enveloping the maximum ac-celeration could be an approach to reducing this workload.

Examining mountings on a theoretical basis may not address some (perhaps the major) problems. The contractor points out that quality of installa-tion or use of problem prone designs may be a stronger influence on mount-ing adequacy than strength considerations. To address these concerns, the contractor suggests a physical mounting review by practitioners experi-enced in both seismic qualification testing and current mounting design practice would be an effective mounting evaluation measure. This process would be enhanced if the reviewers were supplied with an equipment table identifying an enveloping acceleration, equipment weight, and a simple description of the mounting. The plant walkthrough would then screen NUREG-1030 2-74

mountings for those requiring in-depth review or retrofit. 'The effective--

ness of this process is that it screens out items which are clearly ade--

quate and concentrates more costly review on questionable items.

_(4) Comparison of Fundamental Building Frequencies With Equipment-Supporting Structure Frequencies The level of equipment-supporting structure response during a seismic event can be related to the corresponding floor response: spectra. The design floor response will generally contain a region with significantly amplified magnitude. The center of this amplified region will. generally lie between 2 and 10 Hz and coincides' with the fundamental frequency of the building. The motion of the equipment-supporting structure is reckoned as a combination of its free vibration modes whose maximum values are determined from the floor-response spectra. Generally the first mode has

.the largest modal participation factor and is the most important. -Knowing

-the first mode frequency and its modal participation factor, the maximum response is estimated readily from the floor-response spectra.

Tuning of the equipment-supporting structure and the building containing it occurs when a natural modal frequency of this equipment-supporting structure coincides with the fundamental building modal frequency. As an example, cabinet frequencies between 5-15 Hz are typical so that tuning is possible. In case tuning occurs, the floor-response spectra may result in a response level 2-5 times the predicted non-tuned response. A com-plicating factor is that the lowest natural frequency of an equipment-supporting structure depends on how it is attached to the floor as well as its physical properties. For instance a welded mounting will result in a higher frequency than a mounting with a minimum number of bolts.

Thus, for operating plants, uncertainties relating to equipment-supporting structures include both physical properties and the mounting boundary condition.

Hence, the design environment of equipment will depend heavily on the relationship between the equipment-supporting structure and a building's fundamental frequencies. It is clear that most of the safety-related systems were not intentionally designed to function in highly amplified dynamic environments (i.e., tuned conditions). The contractor suggests that systems which may be subject to these loads should be identified by in-situ procedures. Here an abbreviated process can be followed in which all the equipment-supporting structure's natural frequencies below 15 Hz are experimentally determined. Mode shape need not be determined. A modal analysis crew should be able to check a number of cabinets in a single day, so cost is not an overwhelming burden. Where amplified equipment-supporting structure response is identified, two options are recommended. Regardless of the criteria applied to other equipment in operating plants, the contractor recommends that this equipment should be qualified vigorously. The first option is to determine the design-basis environment (or component-specific RRS) and qualify equipment to that environment. The second option is to modify the equipment-supporting structure, depending upon which is appropriate. That a lower response is assured should be verified by in-situ procedures.

NUREG-1030 2-75

2.2.2.2 Summary of _ Part C, " Guidance and Acceptance Criteria for Application of Combined In-Situ and Analysis Procedures" This part covers Topics 6 and 9 defined in Section 2.2.1 of this report. Four-teen technical areas are identified.

Following is a summary of the guidance end acceptance criteria in the fourteen technical areas. Details can be found in the INEL report (NRC, June 1984).

(1) Dynamic Parameters From Tests. Guidance is required on the number and position of nodal paints for describing the mode shape. Node points are to be located at all significant masses (>5% of total system mass), and there should be no less than four node points between modal antinodes for the significant mode with the largest natural frequency.

Assurance must be provided that all modes in the frequency range of interest have been determined. Additional guidance concerning natural frequencies is included in Iter.a 8 and 14, that follow.

(2) Analytically Determined Dynamic Parameters. Guidance relating to analyti-cally determined equipment-supporting structure models is that these models are to be verified by comparing computed and experimentally deter-mined natural frequencies. The analytic and experimental frequencies must correlate to a reasonable tolerance - say 10%, for frequencies in the range of interest.

(3) Analysis Methods for Generating Device Location Required Response Spectra (RRS)

The time history analysis method is currently accepted (NRC, RG 1.92) and

, the same guidance should be applied to operating plants. Response-spectra transfer using random vibration methods is acceptable; the complete mean square response must be employed, the peak value factors must be justified, th? modal participation factors employed must meet the criteria in Item 7,

and all significant modes must be included in the structural model. Addi-tional details are available in the INEL report of October 1983, " Improved In-Situ Procedures and Analysis Methods for Seismic Equipment Qualification in Currently Operating Nuclear Power Plants."

(4) Modal Participation Factor (MPF)

Proposed guidance is to determine the mass matrix ([M]) from physical characteristics of the system and calculate MPF according to the following equation

{$}jT [M] {1} = MPF 9 An alternative method is to use the equation (where {f} is the vector of MPFs where [$]* represent the incomplete modes)

{f}* = ([$],T g,3,) 1 g,3,T gy)

NUREG-1030 2-76 l

and verify that the body force load is well simulated, i.e.,

{R}/{I} $ 0.05.

where {R} is an error vector, Other methods for approximating the MPF must be justified and will be evaluated on a case-by-case basis.

(5) Determination of Fundamental Frequency of Equipment- hpporting Structure The frequencies of equipment-supporting structures are acceptable if the transfer function in the frequency range of interest is determined from data maintaining a coherence of 0.8 or greater at the natural frequencies.

Another acceptable approach is to document that the magnitude and phase angle of the driving point frequency response functions (FRF) follow rules consistent with the absence of a natural frequency.

Other methods of establishing the low frequency range containing no natural frequencies will be evaluated on a case-by-case basis until experience warrants the development of general guidelines.

(6) Frequency Margin As stated in Item 8 the exact value for the fundamental frequency can play a large role when modal parameters are combined with analysis procedures near a floor response spectrum peak, small errors in the in-situ frequency estimates can result in significant errors in the calculated RRS. There are potential sources of uncertainty in the frequency estimate, and the introduction of margin may be required to ensure conservative results.

The approach incorporates an uncertainty of 110% in dynamic parameters determined using in-situ procedures. In this guidance it is assumed a time history or PSD cc.isistent with an unbroadened floor response spectrum is employed.

In Figure 2.2-1 several frequency regions are defined on a line graph. If w

s is the best estimate of a building's fundamental frequency and w c is the best estimate of a support structure's frequency, then Region 1 is 0.85 m3 5 w $ 1.15 ms , Region 2 is 0.9 mc 5 w $ 1.1 wc , and AD is the distance, measured in frequency (Hz) between the two regions as shown in Figure 2.2-1. If AD > 0.1 w cthen the two regions are considered to be well spaced, otherwise they are considered to be coupled. One st.t of guidance applies if the regions are well spaced and a separate set applies to coupled regions. As noted earlier all guidance presented herein is based on unbroadened floor response.

For well spaced frequencies, either time history or mean square response (i.e., during PSD function) analysis procedures may be used. The input to the support structure is consistent with the unbroadened response spectra with peak at ws The structure for which an in-structure response NUREG-1030 2-77 t

e,= Building Frequency wc= Support Structure Frequency Region 1 Region 2 0.86 w, 1.15 w, 0.9 e c l' "c V V AD  :

l w,

A  ;

l wc J w mz Figure 2.2-1 Line graph definition of Region 1, Region 2, and frequency separation AD spectrum is sought is modeled with its best estimate modal properties.

These estimates must be consistent with guidelines presented elsewhere in this document or in existing regulatory guidelines. The required in-structure responses are predicted using time history or root mean square procedures. Figure 2.2-2 shows the expected features of the in-structure response spectrum. The response spectrum peaks are horizontally extended across Region 1 and Region 2 to apply margin, and the remainder of the spectrum is formed in conformance with NRC Regulatory Guide 1.122.

For the situation in which Region 1 and Region 2 couple, the procedure is somewhat different. Time history methods are not practical because three separate spectra-consistent floor time histories are required to use the procedures to be described. Coupling or tuning of building and support structure is not expected to occur frequently. SQUG experience data in-vestigations show support structures natural frequencies above 6 Hz to be the typical situation. This is signficant because incorporating margins for building modal parameters and support structure modal parameters is relatively more complicated for the condition where Region 1 and Region 2 couple.

The methodology for estimating secondary response spectra with the incor-poration of margin on support structure frequency is now described. Two procedures are required. One for the case in which coupling occurs without overlapping. Three floor response spectra are defined. These response spectra have peaks at w s

, 0.85 ms , and 1.15 ws , respectively. A spectrum-consistent PSD is calculated (NRC, June 1984) for each response spectrum.

Several versions of the support structure's modal model are generated. The NUREG-1030 2-78

a w,= Building Frequency w = Supp rt Structure Frequency c

cm l> -

R E 1 [

- / \

\ e

/

/ \N N

I I "s "c co - H z

]

.i Figure 2.2-2 Best estimate in structure response spectra and broadened response spectra mode shapes are not modified. One modal model has a set of natural fre-quencies in which the first mode frequency is 0.90 mc. A second model employs a first mode natural frequency of 1.1 wc. If Region 1 and Region 2 do not overlap, no other support structure models need be considered. The floor input PSD for 0.85 w, is combined with the support device structural model using 0.90 wc as its fundamental frequency and a response spectrum is generated using the root mean square approach. A second in-structure response spectrum using a PSD for 1.15 m s and fundamental support structure frequency of 1.1 mcis constructed. A third in-structure response spectrum using a PSD for 1.15 ms and fundamental support structure frequency of 0.9 ws is constructed. Finally, a combined response spectrum enveloping these two response spectra is formed and this response spectrum incorporates margin on both building properties and support structure properties. These three spectra are employed to generate the enveloping RS.

If Region 1 and Region 2 overlap, then a calculation in addition to the two described above is required. It is assumed the only practical situation is shown in Figure 2.2-3. An input PSD is generated for floor response spectra whose peak is at 1.15 ms . This input is applied to a structural model with fundamental frequency also at 1.15 m3. A second in-structure RS is calculated as follows. An input PSD is generated for floor response spectra whose peak is at 0/9 mc. This input is applied to a structural model with fundamental frequency also at 0.9 wc . As before, the RS are superimposed and an envelope is formed NUREG-1030 2-79

w,= Building Frequency

• c= Support Structure Frequency 0.9 w, 1.1 w e V >A 0.86 m, 1.15 m, V A m, w w-Hz e

Figure 2.2-3 Coupled building and support structure natural frequencies (7) Equipment-Supporting Structure Linearity Support structure attached to the floor using bolt attachments must justify that installation preloads are not reduced by more than 70% during the SSE environment.

(8) Enveloping Criteria As with current criteria, the experience response spectra (ERS) for rigid equipment must envelope the RRS at the ZPA. Envelopment at lower frequen-cies is not essential. For the structural integrity of equipment-supporting structure, envelopment is required only at frequencies greater than the fundamental frequency of support structures (with 15% margin on frequency).

See Figure 2.2-4.

If justification can be provided that equipment is not specifically sensi-tive to low frequency inputs (i.e., so that the input does not have to be rich in low frequency content to perform a qualification test), envelop-ment can be restricted to the remaining frequency range.

(9) Component Mounting Structural Integrity Loads on component mounting can be calculated using dynamic parameters developed from in-situ procedures. An acceptable maximum acceleration is calculated using the peak-broadened FRS, the modal parameters, and the NUREG-1030 2-80

< i kV RB undamental Requency of NPP Building W,, = NPP Equipment Supporting Structure Fundamental Frequency W,g = Experience Data Equipment Supporting

.= Structure Fundamental Frequency

= l 6 I N l l l

l l 1 -

1 I I l l l I I l Was Wio Win Frequency - Hz Figure 2.2-4 Comparison of envelopment analysis methods discussed in NRC Regulatory Guide 1.92. The mass is taken as the sum of the components and mounting fixture masses.

(10) Calibrational Certification of Equipment, Instrumentation, and Computer Software Guidance with respect to calibration of equipment and instruments is that the calibration procedures used must be recorded and included with the test documentation. These procedures should be referenced to an applicable testing standard if possible. The methods of calibration (system or com-ponent), the instrument calibrations and the calibrated range, and manu-facturer's specifications for calibration should be included in test documentation. Manufacturers' specifications for instruments (including weight and rated operating range) and equipment should be included with test documentation. A driving point frequency response function measured during the initial stages of testing should be repeated at the completion of testing. These two measurements of the same frequency response func-tion at the driving point must compare within acceptable limits to verify -

stability of measurements. The modal extraction software employed should have been certified by the solution of a standard problem. Software certi-fication is discussed further in Item 14. A sketch of the system tested showing overall dimensions, location of seismic Category I equipment, instrumented positions, and detailing of anchorage must be included with documentation.

NUREG-1030 2-81

(11) Pretest Evaluations The major item to be resolved during pretest evaluations is identifying the appropriate method, locations, and directions for exciting the struc-ture. To ensure that all natural frequencies have t,een determined, ex-citation must be applied at a minimum of three positions for each principal horizontal direction. At these positions, frequency response functions at the driving point should provide the complete set of natural frequencies.

The excitor location to be used in generating the complete set of FRFs should maintain an acceptable value of coherence over the frequency range of interest (0.8 or greater). A coherence check at the natural frequen-cies between the input point and a remote accelerometer position is also required. In this case it is expected that the coherence will be lower in frequency ranges where the FRF indicates an antinode (a small modal coefficient for a given mode). Over the remainder of the frequency range of interest, the coherence must meet the same standard as the standard imposed at the driving point.

The reciprocity (output at 1 for an input at 2 versus output at 2 for an input at 1) between excitation location and a remote point should be verified. The comparison between FRFs should be sufficiently close to indicate that the same load paths are operating for both cases. Finally, the most representative frequency response function at the driving point should be evaluated at several levels of loading. The purpose is to demonstrate, in combination with the reciprocity check, that the natural frequencies and mode shapes will remain relatively invariant with excita-tion level.

(12) Data Collection The qualification documentation should record the following information:

(a) total number of data points in sample, (b) number of samples used to develop FRFs, (c) anti-aliasing filter employed, (d) windowing (if used) to prevent leakage in data, and (e) the sampling frequency.

(13) Calculation of Frequency Response Functions (FRFs) From Recorded Data It is considered that no special guidance or acceptance criterion is neces-sary. A requirement to develop FRFs for a standard set of data could be imposed if the NRC staff felt that this level of certification was neces-sary. If the NRC staff felt certification of software was necessary, then a one-time requirement for development of accurate FRFs from a standard set of data could be imposed.

-(14) Modal Extraction The contractor should identify the developer of the software and the basis for choosing the modal extraction process used.

The major item in auditability of the modal extraction process is valida-tion of the software used in modal extraction. The theory of steady state linear vibrations, Fourier transforms, linear algebra, etc., provides the common basis for modal extraction. However, numerous details are involved NUREG-1030 2-82

in developing computer software for application to modal extraction. Hence a direct check on software accuracy is desirable. In-situ test contractors should certify their software to one or more standard problems. This certification should be maintained by the utility for each such contractor retained for performance of in-situ investigations. Furthermore, it is recommended that the standard problem use data recorded during testing of an equipment-supporting structure typical to those found in nuclear power plants.

2.2.2.3 Summary of Part D, " Seismic Qualification Cost Estimating Task" The objective of this task was to estimate costs associated with the steps of implementation of alternative seismic qualification methods as depicted in Figure 2.2-5. A table of estimated costs is given in this report (NRC, June 1984) and is shown here as Table 2.2-1. It should be cautioned, however, that initial comments on this cost table by an industry group indicate that equip-ment replacement costs are low by a factor of 3 to 5 and in some cases as high as 9.

Assumptions used to develop the cost estimates are described below.

Equipment List-The equipment list was obtained by modifying the list offered in the report

" Survey of Methods for Seismic Qualification on Nuclear Plant Equipment and Components." The modifications resulted from a comparison of the list with two complete lists of safety-related equipment for two new plants--one PWR, one BWR.

Analysis The " analysis" cost estimates were based on experience in estimating analysis jobs and on reviews of such analyses performed during staff audits of new plants for licensing reviews. Equipment which has no estimate for analysis is not suitable for qualification by analysis.

Test and Analysis The numbers under " test and analysis" represent the cost to determine equipment /

support dynamic characteristics via in-situ testing. These numbers were based on an attachment to the contractor's report (NRC, June 1984). Cost of labor, travel of personnel, and transportation of test equipment are included in the estimates.

Replacement

" Replacement" is the cost incurred to replace equipment with qualified equipment.

This includes purchase of the equipment with qualification documentation and installation. It does not include freight charges. Estimates are primarily 4 based on " Process Plant Construction Estimating Standards," by Richardson Engineering Services, Inc. (RES). Two editions of the standard were used, one dated 1975 and the other 1981. Estimates taken from the 1975 edition were increased by 30% to account for inflation. Two components on the list (MSIV &

CROM) were not covered by the standard. Estimates for these two were obtained from equipment vendors.

NUREG-1030 2-83

_- , ,--_ -----,.-.-.__-_____,__.---rw-4 r m-v--- - -- - - - - + - - - > - --

Equipment Screened Out by Date Bose y ISetisfies 8 Classee Recommended by sSRAPI  ;

A

?

w Equipment Outside Umits of Data Bose (Cavents and o Bounding Spectre) or Not

. Belonging to the 3 Classes .

I r Ucensee Develops Recommended by Equipment Plant-specific Compare Ust SSRAP in Date Bose Screened Out Seismic Adequacy Equipment ust

• With Emperience  :  ; of Equipment From Fun-tional Date Bese" Means t Assured Requirement *

'From Section 1.3.2 (beyond tt.e scope of this work).

• • An estimate was made for the cost of comparing dynamic and functional characteristics of equipment in plant and that in the data base.

ta. Extend experience data which are comparable to SSRAP guidance and caveats,

b. Find test data which are applicable to equipment.

c. Develop other evidence of seismic ruggedness.

d. Test protntype.

ro e. Perform analysis and/or in-situ test to show seismic ruggedness or similarity with data base or test data d (see NOTE 1, below).

A f. Simple modification to provide similarity with data base 2 (see NOTE 2, below).

g. Replacement by qualified equipment (an estimate of replacement cost was made).

h. Qualify to current requirement.

NOTE 1: An estimate was made of the cost of determining equipment / support dynamic characteristics via in-situ testing.

Supports are typically either included in the qualification of equipment (e.g., diesel generator skid) or qualified at separate equipment (e.g., panels, racks, cabinets).

NOTE 2: A cost estimate of simple support modifications to obtain similarity with the data base was made. These numbers represent the cost of providing simple support modifications to obtain similarity with the data base equipment.

I They were calculated using the following formula:

Cost = (1.5 Lj x W) + 0.1 Cy + 200 where Lg = the number of. manhours required for installation of a new piece of equipment (the " average" L9 is twice the " low" L and one-half the "high" L )

u 9 W = hourly wage of installation labor ($20/hr was used)

C, = base cost of a new piece of equipment.

The first term of the equation (1.5 L xg W) represents the labor cost to make the modification. The second term (0.1 Cy ) is the material cost. The third term (200) represents four hours of an engineer's time at $50/hr.

Figure 2.2-5 USI A-46 screening procedure

2:

g Table 2.2-1 Cost estimates

• m k

H Analysis Test and Analysis Replacement Comparison Support Modification 8

o Equipment Type High Low Average High Low Average High Low Average High Low Average High Low Average Air Circ Fan / Motor 10,000 6,000 8,000 44,500 9,900 15,300 75,000 3,500 13,500 600 100 200 7,000 1,300 2,600 Air Cond Unit 200,000 75,000 100,000 118,000 26,200 40,600 260,000 28,000 115,000 1,600 400 800 15,000 2.400 7,000 Cabinet 13,000 7,000 9,000 44,500 9,900 15.300 4,500 1,000 2,500 600 100 200 850 350 500 Circuit Board -- -- -- -- -- --

600 90 400 600 100 200 350 230 275 C

CRDM -- -- --

44,500 9.900 15,300 32,580K 2,450K 27,000K 600 100 200 33,700 5,800 13,800 Diesel Generator 200,000 75,000 100,000 118,000 26,200 40,600 750,000 250,000 500,000 2,000 400 1,200 88,600 24,800 49,400

, Inverter -- -- -- -- -- --

1,300 200 900 600 100 200 370 240 300 a

MSIV 18,000 12,000 15,000 53,600 11,900 18,400 350,000 140,000 200,000 600 100 200 37,400 13,100 21,600 Panels 13,000 7,000 9,000 44,500 9.900 15,300 30,000 1,000 7,000 600 100 200 1,870 360 710 Small Horiz Pump / 23,000 14,000 17,000 44,500 9,900 15,300 95,000 6,000 54,000 1,200 200 400 8,100 1,460 - 4,400 4

Motor Medium Horiz Pump / 23,000 14,000 17,000 44,500 9.900 15,300 160,000 17,000 78,000 1,200 200 400 16,800- 3,400 8,400 Motor Large Horiz Pump / 23,000 14,000 17,000 44,500 9.900 15,300 245,000 31,000 125,000 1,200 200 400 25,200 5,200 12,800 Motor Sma11 Vert Pump / 26,000 17,500 22,000 44,500 9,900 15,300 42,000 7,000 24,000 900 100 300 12,100 3,040 6,300 Motor Medium Vert Pump / 26,000 17,500 22,000 44,500 9.900 15,300 87,000 30,000 59,000 900 100 300 18,900 5,200 10,200 ro Motor e Large Vert Pump / 26,000 17,500 22,000 44,500 9.900 15,300 160,000 50,000 100,000 900 100 300 31,800 8,500 16,800 w

$ Motor Racks (Instr.) 13,000 7,000 9,000 44,500 9,900 15.300 3,300 750 .1,900 600 100 200 800 350 510 Racks (Bat.) 13,000 7,000 9,000 44,500 9,900 15,300 5,000 1.100 2,800 600 100 200 870 360 540 Strip Chart Rec. -- -- -- -- -- --

7,500 800 3,400 600 100 200 970 350 570 Relays -- -- -- -- -- --

800 130 560 600 100 200 350 230 280 Metal-Clad -- -- --

53,600 11,900 18,400 73,000 12,000 42,500 600 100 200 9,000 2,140 4,800 Switchgear d

Voltage Switchgear -- -- -- -- -- --

7,100 300, 3,200 600 100 200 680 230 430 Motor Control -- -- -- -- -- --

10,700 350 3,650 600 100 200 1,270 270 410 Center

Transducer -- -- -- -- -- --

1,300 500 1,000 600 100 200 370 250 300 i Transformer -- -- --

27,400 6,100 9,400 8,500 1,500 5,500 600 100 200 1,530 500 920 Check Valve 6,000 2,000 4,000 27,400 6,100 9,400 9,000 150 4,800 600 100 200 1,150 350 700 Small Instr. Valve 6,400 3.200 4,800 26,800 6,000 9,200 300 90 125 600 100 200 330 230 260 Small Reiter Valve 13,000 8,500 11,000 44,500 9,900 15,300 15,000 1,300 8,000 600 100 200 1.150 340 700 Large Relief Valve 13,000 8,500 11,000 53,600 11,900 18,400 45,000 5,200 25,500 600 100 200 3,400 760 1,920 Small Safety Valve 11,000 6,500 9,000 44,500 9,900 15,300 6,000 2,800 4,500 600 100 200 1,030 460 670 Large Safety Valve 11,000 6,500 9,000 53,600 11,900 18,400 35,000 6,000 14,000 600 100 200 2,500 660 1,200

a. Equipment with no estimate for a particular method is not suitable for qualification by that method,

b. Cabinet only. Contents of cabinet not included

c. K = x 1,000

d. 15 amp-240 V ac 3 pole circuit breaker,

e. 600 V 3 phase ac 9 2 hp motor starter.

I 4

5

Qualification documentation was assumed to cost 150% of the cost of the unquali-fied components for all but three of the components--small instrument valves, transducers, and relays. These components are produced in large quantities and required in large quantities in typical plants. Their qualification documentation is assumed to be less costly--50% of the cost of the unqualified component.

Comparison The " comparison" estimate is the cost of comparing dynamic and functional characteristics between equipment in plant and that in the data base. The estimate is based on the assumption that necessary data are readily available.

Therefore, no costs resulting from analysis or in-situ testing have been included.

Table 2.2-1 is a summary of cost estimates taken from this contractor's report.

2.2.3 Staff Concl.usions As mentioned in Section 1.3.3, if there are items of equipment that can not be screened out by the data base, either because they are outside the limits of the data base (caveats and bounding spectra) or they do not belong to the eight classes of equipment recommended by SSRAP in the data base, then one of the alternatives is to perform analysis and/or in-situ tests to show seismic rugged-ness or similarity with data base or test data. Section 2.2.2.1 addressed this alternative. Figure 2.2-5 shows schematically the steps suggested by the staff if equipment is not covered by the existing data base.

2.3 Development of Methods To Generate Generic Floor Response Spectra 2.3.1 Background In the current practice of seismic qualification of safety-related equipment (either by analysis or by testing), when the dynamic characteristics of a piece of equipment are known, the required input seismic loading to the equip-ment, or more exactly, the information necessary to evaluate the response of the equipment to a seismic loading, usually is contained in the form of a set of required response spectra (RRS). If this equipment or component is attached to a floor, these RRS are the same as the " floor response spectra." In the case that this equipment or component is attached to an equipment-supporting struc-ture (such as a rack, a cabinet, etc.), floor response spectra usually are still the starting point of analysis whereby the RRS at the equipment or component attachment locations can be obtained. Floor response spectra, therefore, are essential elements for the qualification of equipment in nuclear power plants.

To determine specific floor motion or equipment-supporting structure motion which is applicable to the development of equipment or component RRS, an ex-pensive and time-consuming time history finite element analysis generally is required. For many operating nuclear power plants, the information on floor response spectra may not have been developed according to the current require-ments. In other cases, the information is simply no longer available. The objective of this task was to develop a set of " generic floor response spectra" which can be utilized for qualifying equipment.

NUREG-1030 2-86

The task of developing generic floor response spectra was undertaken by Brookhaven National !.aboratory (BNL). The task now is complete. NRC issued a report in September 1983. Followirg is a summary of this contractor report (NUREG/CR-3266).

2.3.2 Summary of BNL Report, " Seismic and Dynamic Qualification of Safety-Related Electrical and Mechanical Equipment in Operating Nuclear Power Plants,"

The development of generic floor response spectra starts with the concept that there is a degree of boundedness to the structural responses. This report I (NUREG/CR 266) (NRC, September 1983) follows this concept and shows that the response can be bounded within a useful range.

The general approach was to study the effects on the dynamic characteristics of each of the elements in the chain of events that goes between the applied loads and the responses. This includes the seismic loads, the soils, and the structures. Two actual structural models, one BWR and one PWR, were used in the study. For the BWR model (Model 3), a Mark I containment structure is modeled as a single stick, as shown in Figure 2.3-1. For the PWR model (Model 4), the system is modeled as three separate structures on a common foundation.

Three stick models are used to represent the shield structure, the steel con-tainment, and the internal structure. Figure 2.3-2 shows this PWR model.

Free-field earthquake response spectra from the El Centro earthquake were used to generate horizontal earthquake time histories. Vertical spectra were not developed in this program. The peak acceleration of this input time history was scaled to a 1 g level as a normalization procedure to study the response.

In reporting the proposed generic response spectra, the peak values were normal-ized to a more realistic time history peak of 0.1 g. The excitation was applied through the soil and into the various structures to produce responses in equip-ment at each level. An entire range of soil conditions was used with each structure, from soft soil (with a shear wave velocity of 800 ft/sec) to solid rock (shear wave velocity of infinity) in seven steps. For both the BWR and PWR models, stiffness properties were varied, with the same mass, to extend the fundamental base structure natural frequency from 2 Hz to 36 Hz. This resulted in fundamental mode coupled natural frequencies as low as 0.86 Hz and as high as 30 Hz. From all of these models of soils and structures, floor response spectra were generated at each floor level.

The proposed spectra were reported for the top level of a generic structure, based on an earthquake time history with a peak acceleration of 0.1 g. Reduc-tion factors are applied to the peak accelerations to account for the site-specific time history maximum acceleration. A second factor was obtained which recognizes a reduced level of acceleration for equipment located at lower elevations.

Figure 2.3-3 is the maximum generic floor response spectra which were deduced from this study. The curves apply to the top of the structure, which is the point of maximum acceleration. They were normalized from an eartnquake time history with a peak acceleration of 0.1 g. These spectra are for five different classes of soils (shear wave velocity from 800 ft/sec to infinity). As shown in the figure, curves A through E are associated with interaction frequencies NUREG-1030 2-87

EL.147*-2_.5_"_ _ _ % , g t,a A

E L.129'-0" u

>8 a

e EL.108*-6" Ru

>7 a

so N

E L. 82*-9_" _ _ _ u~ ,6

'  : n o

N E L. 65'-9_" ._ _ _".o~ i5 n

M N

EL. 42'-6" u

---

~ >4 a

o*

N '

u E L. 14*-6_" _ _. _ _% ,3

e a E L. O'-0"_ _ _ _._e% ,2 JL o*

M EL. 26*-0" u 1 g i. A n. h c__

Figure 2.3-1 Model 3 4

NUREG-1030 2-88

l SHIELD STRUCTURE STEEL CONTAINMENT

---

<>1 h 11 0 - - - - -

J6 e

.- o g .

Ru

't

---

o 2 12 o ---

a e

ir

---

0 3 13 < >  !

n 1 o

R v

- - - - - <>4 14 < >

,a o

R v

g

---

o5 15 o o

Ru

---

t >6 16 4 >

o '

Ru

---

t >7 17 < >

o.

R u

---

<>8 18 o INTERNAL STRUCTURE o

='

21 t i- - --

R $ 6'-6" u 22 o ---

---

t>9 19 i> 0 6'-0"

. 23 4 i---

a g

R 20'-0" v

---

1>10 20 0 U h 24 t >- - -

o u -

$ 6*-6" a

e 25 4

w EL. 25'-6" u

///////////

Figure 2.3-2 Model 4 NUREG-1030 2-89

l 2.2 CPS 8 CPS 7.2 -

7.2 I

7.05 7

6.8 INTERACTION CURVE FREQUENCY 6.3 A 2 6.0 - B 3 C 5 D 10 E 50 4.8 -

6 2

O g 3.6 U

U A B C u E 4

2.4 -

1.2 -

0.8 0.4 I i i i a i t 1 2 3 4 6 10 100 FREQUENCY (Hz)

Figure 2.3-3 Generic floor response spectra NUREG-1030 2-90

(a natural frequency calculation obtained by taking the square root of the ratio of soil stiffness to an equivalent mass of the soil and structure) of 2 Hz through greater than 50 Hz, or from soft soil through solid rock, respectively.

Figure 2.3-4 shows the reduced peak acceleration values that apply to the accelerations in the response spectra at different floor levels. This figure corresponds to soil condition of solid rock (Case E) which has a maximum peak acceleration of 7.2 g at the top level for a 0.1 g earthquake. The peak was calculated to be 6.0 g for a 0.1-g earthquake. This was increased by 20% to ,

7.2 g because only one earthquake time history was used for the horizontal l spectra. As the shear wave velocity of the soil decreases (softer soil), the i maximum floor response acceleration decreases. The peak acceleration at the top level of a structure on soft soil was taken to be 5.0 g. This is 30% less than the peak floor response acceleration of 7.2 g at the same elevation for a solid rock soil.

In summary, this report established a procedure for generating the horizontal generic floor response spectra to any operating plant. The procedure allows a utility to use as much or as little information as is available. The conserva-tisms of the spectra generated increase if little seismic data are available.

Generic spectra in the vertical direction were not developed in this program.

Because of the conservatism accumulated by this approach every step along the way, the NRC staff believes that con.ervative vertical generic floor spectra can be reasonably estimated by taking two-thirds of the values of generic floor spectra in the horizontal direction.

2.3.3 Staff Conclusions Required response spectra (RRS) are needed whether analysis, test, or experience data are used for the qualification. If equipment is attached to the floor, the floor response spectra will be the RRS. If equipment is attached to a supporting structure, the RRS at the point where the equipment is attached can be generated by a variety of ways (see Section 2.2) from the floor response spectra.

By using the methodology described in this section, the floor response spectra can conceivably be generated with reasonable conservatism without having to go through the rigorous time history and finite element analyses normally required.

However, the staff believes that this approach will have its limitations, and these limitations should be spelled out clearly.

NUREG-1030 2-91

l 0 2.0g 4.05 6.0g 8.0g i i i i 1 5.0 6.3 6.8.7.2g TOP co 10 3

2 CPS 3.3 4.3 6.0 MIDDLE - - - - -

TOP i

I f

l MIDDLE X BOTTOM BOTTOM (b) (a)

Figure 2.3-4 Generic peak responses at top, middle, and bottom levels NUREG-1030 2-92

2.4 Seismic Qualification of Equipment Using Existing Test Data 2.4.1' Background.

In addition to seismic experience date, another' type of experience data is the large amount of information generated by the nuclear industry during seismic qualification testing of equipment in the past several decades. In 1984, the Electric Power Research Institute (EPRI) initiated a project to collect and evaluate seismic test date. SQUG, SSRAP and the NRC staff worked very closely

with EPRI since the beginning of this projec't to make sure that the results could be considered in resolution of USI A-46. The specific goals of the project are to establish 1) the classes of equipment for which sufficient qualification test data exist, 2) the generic seismic ruggedness level for each equipment class, and 3) the functionality of equipment required to operate during seismic motion (" operability" or " functionality") and after-seismic motion (" survivability"). The anticipated program accomplishments are

1) both operability and survivability ruggedness levels for each identified equipment class, 2) inclusion rules and cautions for each equipment class, and

3) field checklists for screening of equipment for class applicability.

2.4.2 -Sununary of EPRI Report, " Seismic Equipment Qualification Using Existing Test Data" The scope of the first phase of the program (EPRI, October 1985) was to develop the methodology for establishing generic ruggedness levels and to demonstrate the methodology for the following equipment classes:

• Motor Valve Operators *

• Motor Control Centers *

• Switchgear*

• Batteries and Battery Racks

• Inverters Battery Chargers

• Relays (The asterisked items were addressed in the SQUG pilot program. Data on electrical penetration assemblies were also collected and evaluated during the first phase of the EPRI program.) The second phase involves extension of the methodology to additional equipment classes.

The methodology involves data collection from utilities, test labs, and other sources. The initial work showed that there is extensive test data available with relatively high input motion levels, that proprietary issues can be handled without difficulty, and that it is feasible to collect the data.

The collection procedure consists of certain iaformation being extracted from test reports and evaluated. This information includes equipment description, test anchorage information, representative Test Response Spectra (TRS), and the results of functional tests (if performed), including failures (if any).

Code numbers are assigned to protect proprietary interests. Once the data have been edited and checked, they are transmitted to a computerized data base.

The data base is being augmented by means of a cooperative effort, in which RES (Office of Nuclear Regulatory Research of NRC) is collecting the same type of data for use in a separate program concerning equipment fragility.

NUREG-1030 2-93

In the EPRI program, data are evaluated in the following manner. The data base is accessed to aggregrate data corresponding to specific parameters of interest. The spectral data are standardized to 5% spectral damping, and the TRS are weighted according to whether they are biaxial or single-axis excitation and random or narrow-banded input motions. The diversity of the equipment represented by the test data is established and subclasses are defined, as appropriate, which have low diversity (i.e., similar physical systems and potential failure modes). The final step of the evaluation is to construct a Generic Equipment Ruggedness Spectrum (GERS) for a specific subclass of equipment and a specific performance level. The GERS is defined as the input motion at the base or support point for which equipment of a given class has been demonstrated, on the basis of test experience, to have sufficient ruggedness to perform as required.

The procedure for constructing a GERS is intended to produce a qualification spectrum that has a confidence level comparable to qualification spectra generated by the test or analysis methods in current industry standards.

Therefore, it can be used to qualify a particular equipment item that (1) satisfies specific inclusion rules (to be discussed further below), and (2) has a Required Response Spectrum (RRS) that is enveloped by the GERS.

Note that the RRS is specified at the base of the equipment item, i.e., it is a floor (or mounting) response spectrum, and therefore differs in nature from the SQUG/SSRAP bounding spectra, which refer to ground motion. The manner in which equipment input motion should be specified for application of the GERS for resolution of USI A-46 is beyond the scope of this EPRI program; this question will be addressed by SQUG/SSRAP and the NRC staff prior to the implementation of A-46.

In the first phase of the work, the collected data were adequate to develop preliminary GERS for five of the eight equipment classes (batteries on racks, battery chargers, inverters, motor valve operators, and electrical penetration assemblies). These preliminary GERS may change depending upon evaluation of additional data in the second phase of the EPRI program. Also, more work is needed to establish the applicability cf the preliminary GERS to older vintage equipment. The GERS developed to date have peak spectral amplitudes which are in the range from 3 to 24 g (when normalized to five percent spectral damping).

The preliminary GERS for motor valve operators is shown in figure 2.4-1.

Associated with each final GERS will be inclusion rules which define the characteristics of the equipment included in the subclass and covered by the GERS. In general, the inclusion rules will specify the characteristics (weight, size, etc.) of the equipment comprising the data base and, perhaps, limitations on the manner in which the equipment is installed.

The GERS provides a measure of equipment capability based on available test data. It does not address (1) the issue of in-plant anchorage capacity or (2) any plant-specific situation which could affect equi during an earthquake (e.g., impact of nearby structures)pmentInspection performance procedures to be developed in the second phase of the work will include cautions concerning equipment condition and proximity of adjacent equipment and structures. Guidelines for anchorage adequacy are being developed in other EPRI projects.

NUREG-1030 2-94

4

\

flo , flodels i,, s T

~

4 40 y, -',

- .c,

' i  ! -

l -

- 1 ne -

30 -

g/ e 6 s.

f 3

- I -

2 I

g - GERS -

)

i b

~ / i /' _

/i1 -

y /

[ 20 -

7 I -

I

, s I -

E I

& / sint D**'I -

- a C' '

'G I .

- 1

' - ~~

10 l

1 l j >

l Valve

- I ~

I 0perator  ;

1 -

0 1 10 ZPA Frtquency (Hz) ( -100 Hz) N Equipment Base for T Yoke g ri -

GERL L A

• . .gkya lve ~ "

Figure 2.4-1 Genuric Equipment Ruggedness Spectrum (GERS) for operability of motor valve operators l

)

NUREG-1030 2-95

__ ________ ___ ~.

s * ,3' m.

a -

, ,y < - ,

'~

/ ',.

For three of the equipment classes (motor control centers, switchgear, and relays), additionail f ata are needed to define ruggedness levels. The current phase of the work ilcludes the evaluation of additional data for the first phase equipFpnt classes, collection and evaluation of data '-* approximately tventy additional classes, construction of additional GERS sod corresponding inclusion rules and cautions, and developnent of a deconvolution methodology for' comparing in-equipment GERS to floor motion (this methodology would be used, for example, for qualifying relays with known GERS for use in an

^

electrical enclosure). Final results from this EPRI generic qualification program are scheduled to be available for review by SQUG/SSRAP and the NRC staff (yDecember1986.

2.4.3 ,Cohclusion end NRC Staff Connents Based op the kikiel review of a number of equipment GERS packages, the NRC staff it in agreement with the SSRAP that these GERS will supplement the seismic ' experience data in the resolution of USI A-46. This project is especially helpful in the areas of addressing the concerns on the cperability (functionality) of essential relays, and in possibly raising the bounding sp s spectra for certain equipment. Both these two and other on-going activities will be reviewed by SSRAP and the NRC staff before their use in the w . implementation of USI A-46.

b t

s k.

/,

"((

m s

b i

y ,.

-[

d4 h'

e ,e i

t i

NUREG-1030 2-96

.3 REFERENCES-ANSI (American National Standards Institute) 816.41-1981, " Functional Qualification _ Requirements for Power-0perated Active Valve Assemblies for Nuclear Power Plants." Draft 3, Rev. II.

N41.9-1976, (or,'IEEE Std. 334-1974), "IEEE Standard for Type Tests of Continuous

-Duty Class IE Motors for Nuclear Power Generating' Stations."

N278.1-1975, "Self-Operated and Power-0perated Safety-Related Valves Functional

. Specification Standard."

DC (U.S. Department of Commerce)

(1967), Wood, F. J.'(Ed.), "The Prince William Sound, Alaska, Earthquake of 1964 and Aftershocks," Vol. II, Part A. Washington, DC.

EERI (Earthquake Engineering Research' Institute)

(May'1973),Meehan,J.F.,L.S.Cluff,H.J.Degenkolb,G.A. Carver,D.F.

Moran, R. - B. Matthiesen, K. V. Steinbrugge, C. F. Knudson, "Managua, Nicaragua Earthquake of December 23,1972," Reconnaissance Report, Berkeley, CA.

(December 1978), Yanev, P.I.-(Ed.), "Miyagi-Ken-Oki, Japan Earthquake, June 12,

'1978," Reconnaissance Report, Berkeley, CA.

(1981a), Anirheh. 3 E. , G. A. Hegemier, and G. Krishnamoorthy " Performance of Native Constructwa. Masonry Structures and Special Structures in Managua,

~ Nicaragua Earthquake of December 23,,1972," EERI Conference Proceedings: Managua, Nicaragua Earthquake of December 23, 1972, November 29 and 30, 1973, Vol. I,

-San Francisco, CA.

. (198'b)', Cajina, A., "The Managua Earthquake and Its Effects on the Water E'opl" System," EERI Conference Proceedings: Managua, Nicaragua Earthquake of Jecember 23, 1972, November 29 and 30, 1973, Vol. II, San Francisco, CA.

(1981c),FerverG.W.,."Managua: Effects on Systems," EERI Conference Proceedings: Managua, Nicaragua Earthquake of December 23, 1972, November 29 and 30, 1973, Vol. II, San Francisco, CA.

(1981d),Hanson,R.D., ands.C.Goel,"BehavioroftheENALUFOfficeBuilding in the Managua Earthquake of December 23, 1972," EERI Conference Proceedings:

Managua, Nicaragua Earthquake of December 23, 1972, November 29 and 30, 1973, Vol. II, San Francisco, CA.

(1981e), Klopfenstein, A., and B. V. Palk, " Effects of the Managua Earthquake on the Electrical Power System," EERI Conference Proceedings: Managua, Nicaragua Earthquake of December 23, 1972, November 29 and 30, 1973, Vol. II,

-San Francisco, CA.

NUREG-1030 3-1

(1981f), Knudsen, C.F., and H. A. Francisco, "Accelerograph and Seismoscope Records from Managua, Nicaragua Earthquakes," EERI Conference Proceedings:

Managua, hicaragua Earthquake of December 23, 1972, November 29 and 30, 1973, Vol. I, San Francisco, CA.

(19819), Yanev, P.I., " Industrial Damage," EERI Conference Proceedings: Managua, Nicaragua Earthquake of December 23, 1972, November 29 and 30, 1973, Vol. II, San Francisco, CA.

(July 1981), Lagorio, H. J., and G. G. Mader, " Earthquake in Campaia-Basilicata.

Italy, November 23, 1980, Architectural and Planning Aspects," Berkeley, CA.

EPRI (Electric Power Research Institute)

(October 1985), Smith, C.B., Merz, K.L., " Seismic Equipment Qualification Using Existing Test Data," EPRI Interim Report NP-4297, Palo Alto, California.

EQE (EQE Incorporated)

(September 1982), Yanev, P. I., S. W. Swan, " Program for the Development of an Alternative Approach to Seismic Equipment Qualification," Vols. I and 2, San Francisco, CA.

(November 1983a), " Investigation of Equipment Performance in Foreign Earthquakes and the 1964 Alaska Earthquake," San Francisco, CA.

(November 1983b), " Seismic Experience Data Base--Average Horizontal Data Base Site Response Spectra," San Francisco, CA.

(November 1983c), " Seismic Experience Data Base--Data Base Tables for Seven Types of Equipment," San Francisco, CA.

(August 1984), "The Performance of Industrial Facilities and Their Equip' ment in the Coalinga, California, Earthquake of May 2, 1983," with Addendum Summary of Ground Motion Intensities from the Coalinga, California, Earthquake of May 2, 1983: Based on Observed Ground, Structural, and Equipment Response," San Francisco, CA.

IEEE (Institute of Electrical and Electronics Engineers)

Std. 344-1974 (or ANSI N41.9-1976), "IEEE Standard for Type Tests of Continuous Duty Class 1E Motors for Nuclear Power Generating Stations."

Std. 344-1975, "IEEE Recommended Practices for Seismic Qualification of Class 1E Equipment for Nuclear Power Generating Stations."

Std. 382-1980, "IEEE Standard for Qualification of Safety-Related Valve Actuators."

Std. 501-1978, "IEEE Standard Seismic Testing of Relays."

Std. 649-1980, "IEEE Standard for Qualifying Class 1E Motor Control Centers for Nuclear Power Generating Stations."

(November / December 1980), Carfagno, S. P. , G. H. Herberlein, Jr. , "A Study of the Effect of Aging on the Operating of Switching Devices," IEEE Transactions on Power Apparatus and Systems, Vol . PAS-99, No. 6.

NUREG-1030 3-2

JAERI (Japan Atomic Energy Research Institute)

(August 1979), Uga, T., K. Shiraki, T. Homma, H. Inazuka, N. Nakagima, " Operating Function Tests of the PWR Type RHR Pump for Engineered Safety System Under Simulated Strong Ground Excitation," JAERI-M8354.

LLNL (Lawrence Livermore National Laboratory)

) (July 1981), Johnson, J. J., G. L. Goudreau, S. E. Bumpus, 0. R. Maslenikov,

" Seismic Safety Margins Research Program, Phase I Final Report-SMACS."

NAS (National Academy of Sciences)

(1973), Comittee on the Alaska Earthquake, Division of Earth Sciences, National Research Council, The Great Alaska Earthquake of 1964 Engineering Volume, Washington, DC.

NCEL (U.S. Naval Civil Engineering Laboratory)

(June 26,1964), Stephenson, J. M., " Earthquake Damage to Anchorage Area Utilities - March 1964," Technical Note N-607, Port Hueneme, CA.

NRC (U.S. Nuclear Regulatory Commission)

(October 1975), " Reactor Safety Study--An Assessment of Accident Risks in U.S.

Commercial Nuclear Power Plants," WASH-1400 (NUREG-75/014), Washington, DC.

(July 1981), " Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants," LWR Edition, NUREG-0800, Washington, DC.

(June 1983), Brookhaven National Laboratory, " Identification of Seismically Risk Sensitive Systems and Components in Nuclear Power Plants Feasibility Study," NUREG/CR-3357, Washington, DC.

(August 1983), Lawrence Livermore National Laboratory, " Correlation of Seismic Experience Data in Non-Nuclear Facilities With Seismic Equipment Qualification in Nuclear Plants (A-46)," NUREG/CR-3017, Washington, DC.

(September 1983), Brookhaven National Laboratory, " Seismic and Dynamic Qualifi-cation of Safety Related Electrical and Mechanical Equipment in Operating Nuclear Power Plants," NUREG/CR-3266, Washington, DC.

(June 1984), Idaho National Engineering Laboratory, "The Use of In-Situ Proce-dures for Seismic Qualification of Equipment in Currently Operating Plants,"

NUREG/CR-3875, Washington, DC.

(August 1984), Southwest Research Institute, "A Research Program for Seismic Qualification of Nuclear Plant Electrical and Mechanical Equipment,"

NUREG/CR-3892, Washington, DC.

(RG 1.29), " Seismic Design Classification," Rev. 3, Washington, DC, Sept. 1978.

(RG 1.40), " Qualification Tests of Continuous-Duty Motors Installed Inside the Containment of Water-Cooled Nuclear Power Plants," Washington, DC, March 1973.

(RG-1.61), " Damping Values for Seismic Design of Nuclear Power Plants,"

Washington, DC, October 1973.

(RG 1.73), " Qualification Tests of Electric Valve Operators Installed Inside the Containment of Nuclear Power Plants," Washington, DC, January 1974.

(RG 1.92), " Combining Modal Responses and Spatial Components in Seismic Response Analysis," Revision 1 Washington, DC, February 1976.

(RG 1.100), " Seismic Qualification of Electric Equipment for Nuclear Power l Plants," Revision 1, Washington, DC, August 1977.

(RG 1.122), " Development of Floor Design Reponse Spectra for Seismic Design of Floor-Supported Equipment or Components," Revision 1, Washington, DC, February 1978.

(RG 1.148), " Functional Specification for Active Valve Assemblies in Systems Important to Safety in Nuclear Power Plants," Washington, DC, March 1981.

RES (Richardson Engineering Services, Inc.)

(1975), " Process Plant Constrction Estimating Standards."

(1981), " Process Plant Construction Estimating Standards."

SSRAP (Senior Seismic Review Advisory Panel)

(January 1985), Kennedy, R. F., W. A. Von Riesmann, P. Ibanez, A. J. Schiff, L. A. Wyllie, Jr., "Use of Past Earthquake Experience Data to Show Seismic Ruggedness of Certain Classes of Equipment in Nuclear Power Plants."

i NUREG-1030 3-4

APPENDIX A

SUMMARY

OF TECHNICAL WORK COMPLETED THAT IS NOT-IMPLEMENTED IN USI A-46 RESOLUTION In this. appendix a summary of work done and major conclusions is presented.

Detailed discussions of certain tasks are then included as separate appendices.

The following sections summarize contractors' results and conclusions of the various tasks. Unless otherwise stated, they represent the contractors' viewpoints and recommendations.

A.1 Identification of Seismic Risk Sensitive Systems and Equipment A.1.1 Background The objective of this task was to investigate possible methods of developing a

. generic minimum equipment list. If a methodology could be developed to evaluate the risk importance of safety systems and equipment, equipment could be ordered by the contribution to risk. Equipment whose failure resulted in a small change in risk could then be culled from the qualification list.

A.1.2 Summary of Task Brookhaven National Laboratory (BNL) under contract to the NRC conducted a study (NRC, June 1983) to evaluate the seismic risk sensitivity of system and components in a PWR and a BWR. Both plant models used were hybrids in that they are not representative of any existing plant. The PWR model consisted of modified Surry Plant fault trees and event trees from the WASH-1400 study and used fragility data developed for the Zion plant. The BWR model consisted of modified WASH-1400 (NRC, October 1975) Peach Bottom risk models and Oyster Creek fragility data.

The intent of this study was initially to develop a generic risk-ordered list of plant equipment which could be applied to specific plants with some additional guidelines to develop plant-specific minimum equipment lists. However, BNL concluded, and the staff agrees, that results of the study should not be used generically. BNL's conclusion states that the study presents a methodology that can be applied on a plant-specific basis to develop a risk-ordered equip-ment list.

A.1.3 Staff Position on Task For plants with existing seismic probabilistic risk assessment (PRA) studies, the staff believes it may be possible in sume cases to eliminate components from the seismic qualification program on the basis of low risk sensitivity.

If a utility should decide to conduct a PRA study using the methodology developed by BNL, the staff would consider it to be an acceptable method subject to the analysis assumptions and inherent uncertainties.

NUREG-1030 A-1

_ . . . . . i

A.2 Assessment of Adequacy of Existing Seismic Qualification A.2.1 Background This task involves a study by Southwest Research Institute (SWRI) to evaluate past and present methods to qualify mechanical and electrical equipment to withstand seismic events. Conclusions have been documented in a contractor report titled "A Research Program for Seismic Qualification of Nuclear Plant Electrical and Mechanical Equipment" (NUREG/CR-3892) (NRC, August 1984). Some examples demonstrating the application of this approach are included in that report.

A.2.2 Summary of Work Accomplished The concept of vibration equivalence is a key factor in development of the correlation of methodologies for seismic qualification of equipment. Vibra-tional equivalence forms the basis for a damage comparison between two dif-ferent motions. In the qualification of nuclear power plant equipment, a great variety of physical failure mechanisms may occur. Therefore, the con-cept of vibration equivalence was generalized to include an arbitrary type of failure or malfunction, that can always be established by input vibrational conditions denoted as the fragility levels. It is understood that the failure or malfunction may or may not impart permanent damage to the equipment.

The conceptual approach for applying vibrational equivalence to correlation of equipment qualification by test is shown in Figure A.2-1. The upper and lower halves of the diagram (conditions 1 and 2, respectively) each represent the independent establishment of a fragility, or threshold of failure level, in equipment which is subject to a dynamic excitation at location x. The effect of the response at location y is to actuate a failure mechanism which depends on the equipment. This arbitrary failure mechanism is dependent on the response amplitude failure mechanism and is dependent on time. Thus, the failure is indirectly dependent on the excitation amplitude, frequency, and

time. If the excitation is manipulated so that failure barely occurs, then the threshold of failure, or fragility function Fxy (f,t) is generated. This function represents a surface, any point on which corresponds to failure of the equipment. If more than one physical failure mechanism at more than one response point is present, then each possesses a failure surface, and the minimum value composite failure surface becomes of concern. The central assumption of the vibration equivalence concept is then postulated

the establishment of failure conditions (see Figure A.2-1 for excitation condi-tions 1 and 2) is possible by various types of vibration excitations, and the corresponding amplitude, frequencies, and time durations constitute equivalent excitations.

Generally, the information on failure, or malfunction, is not required as part of an equipment qualification process. On the other hand, functionality of an item of equipment at specified excitation levels is required for qualifi-cation. Functionality and fragility are very much related--fragility is the upper limit of functionality. Conversely, existing qualification data, which include excitation levels and functionality data, may be useful as a lower bound for fragility. Thus, since fragility data are necessary for a general application of the vibrational equivalence concept, use of such existing NUREG-1030 A-2

Location L cati n Failure x Hmy(f) y Mechanism F (f,t) my

~ Failure ~

Amplitude 1 Excitation Fragility Response _ _ Function vel

] Level 1 F,,1f3.t3) l ~

Time ~

Duration 1 l

--- l 1

Specimen  !

Transfer Both Points Function Constitute Failure ~ ~ +l g

i

~

Failure ~

l Amplitude 2 l Exc n (e e Response _ _ F ncti n Fr q . 2 Level 2 F,y(f22.t I Time -

Duration 2 Figure A.2-1 Conceptual approach to vibration correlation qualification data, where possible, is highly desirable to avoid the necessity of generating or collecting more precise fragility information for the great variety of equipment typically contained in a nuclear power plant.

The most general description of a fragility concept is shown in Figure A.2-2 as a fragility surface. This surface can be represented as a function Fxy (f, t) = M7 (f, t), where Mf (f, t), measured at the fragility surface, can be in terms of the amplitude of the excitation, the response spectrm power spectrum, or a variety of other parameters which may be used, or have been used, in typical equipment qualification procedures. The true surface may be quite complex, but a simpler lower bound surface can be defined conser-vatively from existing qualification information which is acceptable for practical engineering purposes.

A convenient method of measuring the onset of failure is proposed by the contractor as the damage fragility ratio Dfr _ M(f,t) -< 1 M (f,t) f where M(f,t) is the value of the actual excitation function and M7 (f, t) is the value of the fragility function at the same conditions of frequency and NUREG-1030 A-3

time. This is shown in Figure A.2-3. A damage fragility equivalence similar to that described in Figure A.2-1 can then be stated as:

1 M(f,t) __ M(f,,t,) i M (f,t) ' M (f ,t ) '

f f 2 2 This is the general basis for comparing various test motions. j The report then proceeded to define simple systems and complex systems. A simple' system is one whose_. fragility function is influenced by a single reso-nance, and therefore can be generated by a slowly swept sine or narrow band random excitation. A complex system is one where several failure modes can occur as the result of multiaxis and/or multimode response, and interaction between responses is included. Because of the difficulties involved when considering complex systems, it is advantageous to develop approximations as required to reduce the system to a simple one.

A number of procedures have been developed in structural analysis to look at the combined effects of multiaxis and multimode response. These procedures, such as absolute sum method, square root of the sum of the squares (SRSS) method, double sum method, closely spaced modes method, grouping method, ten percent method, Lin's method, and complete quadratic combination (CQC) method, are all generally based on modal or response spectrum analysis. Any one of

these methods will give an estimation of the combined maximum peak response of a complex systems. In developing a fragility surface for existing qualifi-cation data, it was recommended by the contractor that a correction factor, generated from resonance search data, be used to modify the level of qualifi-cation excitation in order to develop an approximate lower bound fragility function.

I The next step is to establish a correlation between the approximate fragility function (namely, existing qualification information) and the qualification corresponding to a different set of criteria. In a specific application, some judgment must be used, the detail of which may vary with each case. Several examples which demonstrate the application of these methodologies are included

! in the contractor's report. (See Figure A.2-4 for possible combinations of fragility function and qualification parameters.)

L In summary, the results of a previous qualification are used first to establish l some form of an approximate or acceptable fragility function. Then, the new

j. criteria are compared to this acceptable fragility function to determine whether a more severe or less severe test is implied. If result shows a less severe test is implied by applying the_new criteria, then it can be concluded

! that this equipment is still qualified to the new set of criteria. In some l cases, a more accurate fragility function may need to be established in order to provide a final determination of the comparison. In these cases, the con-l tractor suggested that it may be more practical to consider a completely new j requalification.

L The contractor also surmised that much of the previously qualified equipment will be able to be requalified to new criteria by the analytical method developed. His belief is based on the fact that many qualification tests prior to 1975 included sine wave and sine beat excitations of some form. The comparison of relative damage severity indicated that such motions produce NUREG-1030 A-4

p Magnitude n

Actual n Surface

/

-, hiliii

= me =s=,r s

s

...;iMil:. s, 4

y - n:,1 72!!.J.,+

i;;g;

' 1I y- --.=u...

\ 2:;-

l l  %: _9 -a l l .

j i I l

l Acceptable 1 Surface l

1 I /  %,

/s '

'd '

Figure A.2-2 Comparison of actual with acceptable fragility surface M ,(f.t) n ii!fi:: h!!ii::

ji!{iji -

.!!!!5 s:g:s

=

hf!k s: . ~ M ,(f3.t3) i!i!!!: I stiGi) 3

'::.f!!!$  ; .;.ii I

/ 's s #**1 1

jid ' ,-

f'*2 2 / MF2II 'Y M(f ,t I 22 Figure A.2-3 Basis for damage fragility ratio NUREG-1030 A-5

Fragility Function Qualification Parameters Parameters Single Axis Single Axis 1 Narrow Band Narrow Band 2 4- Excitation

• Excitation Single Axis Single Axis 3 Broad Band Broad Band 4 Excitation Excitation Multi-Axes Multi-Axes 5 Narrow Band Narrow Band 6 Excitation Excitation Multi-Axes Multi-Axes 7 Broad Band Narrow Band 8 Excitation Excitation

' includes sinusoidal excitation Figure A.2-4 Possible combinations of fragility function and qualification parameters significantly more potential damage than do typical random motion simulations that have been more generally used after 1975.

A.2.3 Staff Conclusion The technical basis and general methodology to correlate seismic qualification

tests have been developed and demonstrated, but are of limited practical value l in their present form because of the need to either know the fragility level or estimate the fragility of the equipment and know the required response spectra. It may be useful in special cases.

l A.3 Related Topics Covered by the INEL Contractor's Report on In-Situ lesting Even though the contractor report (NUREG/CR-3875) (NRC, June 1984) is concerned mainly with how to utilize in-situ testing to assist in performing seismic qualification of equipment, the contractor studied other related topics. Among them are the following.

NUREG-1030 A-6

A.3.1 Operability and Failure Modes:

In order to develop methods to utilize experience data to qualify equipment, the contractor suggested that a systematic treatment of operability is necessary. The failure modes which result in inoperability, from the contractor's viewpoint, are an essentia'. ingredient to these methods. The contractor first defined inoperability and its causes and then identified all possible failure modes that may cause inoperability during an earthquake.

Inoperability is defined as any action or interaction of component parts or interfaces which prevents a com,9onent from performing an active operation or maintaining a state continuously. Inoperability can result from: l l

inability to monitor the control condition inability to change states when so directed inability to maintain the current state when no change of state is directed The contractor suggested that inoperability during an earthquake occurs through the following modes:

structural integrity - stress limits are exceeded, permanent deformation occurs, flaw initiation or extension occurs.

operability loss due to temporary or permanent reconfiguration -

vibratory elastic motion results in a change of state or prevents a change of state from occurring.

structural interference - excessive relative motion rasults in a tolerance mismatch.

nonstructural changes in state peizoelectric effects, effects of dynamics on contact resistance, and others; anywhere a fundamental nonstructural response is affected by vibration or stress.

The contractor then proposed that similarity between two equipment designs can be defined as similarity in potential failure modes. The basic premise involves two pieces of non-identical equipment having a common critical failure mode.

The first piece has been qualification proof tested and its controlling design features are either identical to or inherently more fragile than the equipment in question. In that case, qualifying the first, amounts to qualifying the other to the same environment. The contractor suggests the procedures below to establish seismic capacity based on similarity.

Specify operability requirements, take into account whether equipment is required to operate and/or maintain a continuous state during earthquakes.

If there are no requirements during the earthquake, certain failure modes will be eliminated and qualification is simplied.

Identify the design features /subcomponents which affect operability. The procedure will be impractical if there are too many.

Identify similar pieces of equipment, i.e., equipment with nominally the same or less seismic capacity in the potential failure model(s). Some form NUREG-1030 A-7

of design eva'luation/ comparison will be required in making this assessment.

Equipment used for comparison must be of known seismic capacity. The staff believes that in-situ testing will be a valuable tool to establish dynamic.

similarity between equipment through the comparison of the dynamic character-istics (mode shapes, natural frequencies, damping, size, shape, weight, etc.).  ;

1 A.3.2 Environmental Aging Consideration:

The environmental history of a piece of equipment can produce changes in properties and dimensions which affect its seismic capacity. Addressing the total environmental qualification of equipment in operating plants is imprac-tical. The contractor adopted an approach based on the interaction of aging and seismic capacity. Such an approach suggests that since some aging mechanisms will not affect seismic capacity, these cases need not be considered in seismic qualification.

~The contractor considered the use of sn-situ testing in evaluating the effects of aging on seismic qualification, nowever, no well developed technologies were identified. Consequently, aging has been examined in a broader context where:

The consequences of aging degradation are examined. This allows the relation-ship between dynamic qualification and aging degradation to be organized in a fashion which more clearly demonstrates the interaction.

Alternate criteria based on failure mode and similarity analysis. This provides both an organized aging assessment procedure and a method for using test data from "similar" equipment.

Equipment without specific operability requirements during seismic events

.has been identified as less vulnerable to aging.

The effect of aging on seismic capacity is illustrated in Figure A.3-1. A l systematic basis for evaluating aging degradation is provided by the failure mode analysis and the procedures embodied in Figure A.3-1. This method as proposed by the contractor is as follows. First, a determination of any aging l effects produced by the design-basis environments should be conducted. This

involves listing all vulnerable materials and examining environmental data for l each. Presently, such data are only available for some materials. Those l components demonstrating no environmental aging require no further examination.

For components containing materials affected by the design environments, the aging mechanisms are defined and categorized by the contractor as follows.

Category I aging: This includes all aging mechanisms which modify the

! dynamic response. The changes in dynamic response can affect all four l failure modes defined earlier. Each failure mode must be examined in light l of the anticipated degradation. If it cannot be established that no signi-

! ficant change in seismic capacity occurs, then the critical failure modes should be established. A similar system with a known aged seismic capacity may provide data on which to base the aged seismic capacity. Adversely affected items should be qualified to current criteria.

NUREG-1030 A-8 l

l Environmental Aging No Yes Seismic Capapity Unaffected 1

Dynamic Dynamic Response Response Affected Unaffected 1f Operability Affected Non-structural Degradation espon Affecting Seismic Capacity

/ \

Load l Seismic Load Magnitudes, Path Environment Frequency a Contributor Content 1 f 1 f Operation During Structural Normal Integrity Environment Affected i f I f Not a Concern of Structural Dynamic Qualification: In Interference Province of Routine in-service Surveillance i f Reconfiguration Figure A.3-1 Effect of aging on seismic capacity NUREG-1030 A-9

. Category II aging: This is any aging mechanist. which could affect the operability of safety equipment when combined with the predicted seismic loads. It is assumed that the dynamic response has not been affected. This is a type of aging mechanism which impacts only the nonstructural effects.

It need only be examined if a known aging effect exists in a component.

Again, seismic capacity can be inferred from tests on similar equipment.

However, the requirements on similarity are somewhat more stringent in this case. Any loss of seismic capacity will be due to degradation combined with local structural dynamics. Thus, similarity requires that both be simulated.

Category III aging: The mechanisms of this category are those identified which have no effect on seismic qualification (IEEE, November / December 1980).

For a typical component many mechanisms would fall in this category.

The application of the above approach would probably be most economical if conducted in stages. The contractor proposed that initially all equipment would have a cursory examination for (a) no aging, (b) some aging, though with no effect on seismic capacity, (c) aging with a potential effect on seismic capacity, or (d) too complex to determine easily. For situations where further consideration is warranted, the steps are similar to those as described in the first paragraph of this appendix. The failure modes are used to establish similarity, and data from similar equipment are transferred to the equipment in question. The important factor is that much equipment will exhibit no signifi-cant seismic aging interaction of concern and, thus, screening can narrow the field effectively without overlooking substantial aging degradation.

l NUREG-1030 A-10

APPENDIX B )

-PERFORMANCE OF POWER FACILITIES DURING THE 1964 ALASKA EARTHQUAKE City of Ancho'r age Gas Turbine Plant The plant contained two gas turbines rated at 15,000 kW and six older diesel generators. Three reports give different versions of what happened at the plant (NAS, 1973, p. 1053; NCEL, June 26, 1964; and F.F. Mautz, cited in EQE, November 1983a). . Apparently, at least one unit operated through the earthquake even though a control cabinet or transformer toppled over. One unit was switched to diesel oil and started to supply power, but it was shut down when diesel fuel supply was lost because the fuel storage tank failed. Unit 2 became unbalanced several weeks later because of aftershocks. It was realigned and put back in operation.

Chugach Power Plant at Knik Arm The plant had three coal-fired boilers. Only one unit was operating at the time of the earthquake. It continued to operate for about five minutes after the earthquake and then was shut down by some disturbance outside the plant.

There was no other power available so it could not be restarted (Mautz, in EQE, November 1983a). The facility suffered structural damage in the coal bunker bay and an ash hopper fell (NAS, 1973, p. 255). One boiler fired up in 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. One week later 3 boilers were in operation, and 5 days later the plant was back in normal production while structural repairs progressed. The turbine bay was undamaged. There was no electrical damage. One compressor was put out of service either by vibration or foundation settlement. Filter mixing tanks which were not bolted down fell over and ripped out piping. Pipe hangers also failed.

Fort Richardson Coal-Fired Steam Plant This plant had a generating capacity of 18,000 kW and a heating capacity of 1,080,000 lb/hr of steam; it had five turbine generators and eight coal-fired boilers (NAS, 1973, p. 911). It produced steam without interruption and could have produced power but the receiving stations were not functioning. The plant had limited structural damage and extensive nonstructural damage, mainly to the superheater tubes and tile bricks.

Elmendorf AFB Coal-Fired Steam Plant This plant had a capacity of 22,500 kW electric and a heating capacity of 950,000 lb/hr steam (NAS, 1973, p. 934). It had three 7,500-kW generators and six boilers. The plant operated through the earthquake, was shut down about an hour later (Mautz, in EQE, November 1983a), and was back in operation in 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />. The shutdown was caused by a loss of circulating water from the failure of buried transit pipe. A break in one air control line also contri-buted to the shutdown. Structural and nonstructural damage was similar to that at Fort Richardson.

NUREG-1030 B-1

Eklutna Hydroelectric Plant This.is a 30,000-kW hydroelectric plant. The plant apparently operated through the earthquake.(NAS, 1973, p. 464 and NCEL, June 26, 1964, p. 3). The intake structure and conduit were damaged by soil consolidation and there was air circuit breaker and transformer damage to porcelain insulator columns. j Chugach Bernice Lake Gas Turbines Continued.to operate (NAS, 1973, p. 1073).

Chugach Cooper Lake Hydroelectric Plant Undamaged, but transmission line down (NAS, 1973, p. 1062).

Port of Whittier Heating and Power Plant This plant had three steam turbine generators--two 2,000 kW and one 2,500 kW.

Damage to the plant was minor. One condensate line and two 10-in, water supply lines were broken (NAS, 1973, p. 1079).

Cordova Diesel Engine Plant No report of damage-(NAS, 1973, p. 1067).

Kodiak No report of damage to power generators (NAS, 1973, p. 1070).

Homer No report of damage to power generators (NAS, 1973, p. 1071).

Seldovia No report of damage to power generators (NAS, 1973, p. 1071).

Kenai This town had an old substation with skid-mounted transformers and regulators.

Some transformers and regulators tipped over causing short circuit.ing (NAS, i 1973, p. 1072).

l l

l l

l NUREG-1030 B-2 a

+w

APPENDIX C PERFORMANCE OF POWER AND INDUSTRIAL FACILITIES DURING SOME FOREIGN EARTHQUAKES C.1 Managua, Nicaragua, Earthquake of December 23, 1972 C.1-1 Earthquake Data Magnitude: Ms = 6.2, Mb = 5.6 Time: December 23, 1972, at 6 hr 29 min GMT Location: Beneath the center of Managua Depth: 8 km C.1-2 Ground Motion Records Four strong motion accelerograms and nine seismoscope records were obtained from a series of earthquakes that occurred in December 1972 and January 1973.

One accelerograph and 13 seismoscopes recorded the main shock on December 23 (EERI, 1981f). The only accelerograph from this shock was recorded at the Esso Refinery; peak ground accelerations were 0.38 g in E-W direction,- 0.34 g in N-S direction, and 0.33 g vertically.

An estimate of the ground motion to which Managua's major industrial facilities were exposed is provided in a report by P. I. Yanev (EERI,1981g): "It is the author's estimate, based on the accelerogram taken at Esso refinery and on judgment, that the industrial facilities experienced an earthquake of moderate duration with the peak ground acceleration exceeding .25 g. Some facilities experienced accelerations exceeding .60 g."

C.1-3 General Effects of the Earthquake Approximately 10,000 people died. Many structures collapsed completely and economic loss was heavy. The downtown business area, the industrial areas, and the surrounding residential areas were most seriously affected. The downtown area was almost totally destroyed, but most modern high-rise structures sustained the shock without collapse and often without significant structural damage. The architectural and other nonstructural components of these newer buildings were often damaged severely. Mechanical systems in buildings were generally inoperative after the earthquake (EERI, 1981g).

Power and industrial facilities suffered considerably lower losses. Damage to equipment and equipment systems was responsible for the greatest part of the industrial loss. Much of the damage and consequent delays in operation could have been prevented with improved equipment anchorages and other minor details.

Few industrial facilities were left undisturbed by the earthquake (EERI, 1981g).

C.1-4 Electric Power Facilities Two hydroelectric plants, each with two 25-mW units, were located 80 and 100 km from Managua. 100 km northeast of Managua there was a 15-mW gas turbine NUREG-1030 C-1

,- - .. . . - --- - , .- . -. -- .- . ~ . _ -

. generator. None of these plants were damaged; however, they all disconnected

-electrically from Managua.

The Enaluf power plant in the City of Managua is a thermal electric power plant with one 40-mW and two 15-WW steam turbine generators. The plant is located on -the shore of Lake Managua, immediately _ adjacent to (or possibly even on top of) the Tiscapa fault, which caused the event. ' Displacements of

, 10 in. along this'section of the fault were reported within 200 m of the plant site.

l It is reported that the facility was designed for a static-equivalent lateral 1 load coefficient of 10%. Most of the equipment was anchored to the floor and experienced no damage. Some of the worst damage occurred to unanchored equipment l which was free to displace or fall (EERI, 1981g).

The main shock-caused generators to trip off-line by protective relays either through legitimate protective measures or through malfunctions due to vibration of mechanical contacts (EERI, 1981e). One of the units was back in . service in two weeks and the second in three weeks. The third unit was not operative for 3

several months because of greater damage and misalignment of the turbine shaft (EERI,.1981g).

Arturo Roja, General Manager of Enaluf (EERI, 1981c), prepared a list of the equipment damaged in the earthquake. This list is presented in Table C.1-1.

Some of the reported damage relevant to the SQUG project is discussed below.

4 4

All three deaerators moved on their bases. The Unit 3 deaerator also ,

sustained a broken air pipe connected to the deareator and damage to ~'

refactory lining (EERI,1981e).

Draft fans, motors, and vents associated with the boiler and exhaust system did not suffer significant damage. Several of these shifted out of alignment (EERI, 1981e).

All three steam turbine generators sustained sufficient damage to incapaci-1 tate them. Bearings in Unit 3 were badly worn when the emergency oil pump motors lost their DC power when the battery racks failed and the batteries broke. Misalignment and broken turbine blades were common to all three generators. There was also some relative movement between the turbine i generator supports and the floor which resulted in further damage and i misalignments (EERI, 1981e).

The condensers associated with the 15-mW generators shifted 6 inches. This broke the valve between the pump and the pipe to the condenser (EERI, 1981e, Figure 19).

I The obvious pipe damage discovered on Unit 3 steam system included broken

piping in the boiler. A pipe connected to the saturated vapor valve of the deaerator was broken. The high pressure pipe of the primary element of the three recirculating valves for the water seating pumps was bent. In

addition, three recirculating valves suffered cracks on their interior sections. The condenser had air pipe damage. '

4 r NUREG-1030 C-2 l

Table C.1-1 Damage to Enaluf Steam Plant Siemens Unit No. 1 (15 mW)

1. Generator case: Shafts displaced.

2. Forced draft fan out of alignment.

3. Induced draft fan out of alignment.

4. Condensate pump: Burned-out bearing.*

5. 440-V ac Panel No. 2: Fallen.*

6. Condensate pump intake valve broken.*

7. Boiler No. 1: Tubing broken and refractory walls fallen.

8. Deaerator No. 1: Fallen from its base.

9. Chimney of Boiler No. 1: Anchor bolts broken and stack leaning.

Siemens Unit No. 2 (15 mW)

1. Generator case: Shafts displaced.

2. Draft fan forced out of alignment.

3. Induced draft fan out of alignment.

4. Boiler No. 2: Refractory walls fallen.

5. Deaerator No. 2: Fallen from its base.

Intake valve of condensate pump broken.*

E 6.

Franco Tossi Unit No. 3 (40 mW)

1. 440-V ac control center: Fallen.*

2. Main transformer bushings broken.

3. Starting transformer bushings broken.

4. Exciter transformer bushings broken.

5. Unit transformer bushings broken.

6. Ljungstrom pre-heater seals damaged.

7. Four turbine bearings burned out. (Batteries broken, cutting off supply to DC powered emergency lube oil pump.)

8. 69-kV switch bushings broken.

• - Denotes equipment failures of particular interest to SQUG.

A motor control center (EERI,1981e, Figure 23) fell over with many of the drawers coming loose from the main cabinets. After the earthquake, the cabinets were uprighted and the system was checked out and placed back into service. Before the earthquake, the cabinets had been secured in place with small bolts in concrete anchors which were not capable of resisting the overturning forces.

C.1-5 Industrial Facilities Throughout the area the performance of industrial buildings ranged from complete collapse, such as the Pepsi-Cola Building, to structures with no damage. such as the Esso and Siemens industrial buildings. The degree of damage to the buildings related directly to the quality of desiga and construction, the distance from the fault, and the ground accelerations (EERI, 1981a).

NUREG-1030 C-3

Esso Refinery - The Esso oil refinery is located on the east side of Lake Asososca. Two seismoscopes and an AR240 strong motion seismograph located at this refinery provided the only records of the Managua, Nicaragua, earthquake.

The record from the AR240 seismcgraph indicated a 30% to 40% ground accelera-tion both horizontally and vertically (EERI, 1981a).

The plant was built in two stages during the mid-1950s and early 1960s and was-designed to meet UBC (Uniform Building Code) requirements. All detailing reflected the latest U.S. design procedures. At that time no specific provi-sions were added for existing seismic hazards. All equipment was tied to its foundations, piping systems were braced, etc. Some difficulties arose after a 1968 earthquake; consequently, the plant was apparently redesigned to withstand 20 g.

Damage at the refinery was minimal. At the time of the shock, half of the facility was shut down for maintenance. Damage to administration and equipment facilities was not significant and operations were resumed within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.

Many grout pads at the supports of vertical steel vessels were spalled. Some piping in the low ground-level pipeway trenches jumped from saddle supports.

Piping on the second floor of the concrete pipeway structure and floor drains for a heat exchanger shifted (EERI, 1981g).

Fabritex Textile Mill Complex - This facility is composed of several large industrial buildin;s of various sizes and construction types. None of the buildings suffered serious damage. Acoustical tiles fell, creating problems in putting the equipment back on line. Whole and broken tiles showered on equipment, falling inside intricate machinery. Inadequately braced machines were thrown out of alignment. The machines themselves were unharmed, but many bobbins and spools fell to the floor and were damaged. Some equipment displace-ments (sliding) caused pipe breaks throughout the system (EERI, 1981g).

Tanic Cigarette Factory - This factory is located 3 miles east of Managua and was less than 5 years old at the time of the earthquake. Construction is a heavy reinforced-concrete frame with fragile curtain walls of hollow clay tile blocks. Some shear cracking occurred in the walls, but the cracks usually did not penetrate the concrete frames and damage was minimal. Most equipment in the factory was light, low profile, unanchored equipment. Movements of 3 to 4 inches at the bases of the equipment were common. No severe mechanical damage was incurred (EERI, 1981g).

Pepsi-Cola Bottling Plant - This plant suffered extensive structural and equip-ment damage and was inoperative for more than 1 month. The long shutdown was caused primarily by the failure of the reinforced concrete building which housed bottling and production equipment. Anchored equioment not damaged by the falling debris generally survived without significant damage (EERI, 1981g).

C.1-6 Water Supply System Potable water is pumped up several hundred feet from Lake Asosoca, a caldera located about 3 miles southwest of downtown Managua, by five 500-hp vertical turbine pumps submerged several meters into the water (EERI, 1981g). The main pumping plant is located at the lake level and pumps water up the steep incline of the caldera in two steel pipes. Some earth slides occurred on the steep slopes and partially blocked an access road but did not damage the pipes or NUREG-1030 C-4

l pumps. Anchor bolts holding down the surge tank were elongated as a result of the earthquake. The roof of the materials warehouse at the pumping plant collapsed (EERI, May 1973).

The 2500-kVA transformers of from 13,200 to 24,000 V, which feed the principal pumping station and the transformers of the booster stations, suffered damage in the secondary insulators. The GE control boards of the main pumping station suffered misalignments impeding the starting of the equipment and were repaired provisionally by means of flexible jumpers (EERI, 1981b).

C.1-7 Enaluf Office Building In the penthouse, equipment that was sitting on the floor but not connected to the structure was displaced. The air conditioning unit slipped from its isolation pads causing the base to translate and rotate relative to its floor support and causing the metal cabinet to move relative to the base. An electric motor fell from its support, but the switch racks to which it was connected were not displaced. Some pipes failed. Roof acceleration is estimated at 1.16 g. Overall building performance was excellent. Nonstructural damage was minimal and structural damage was isolated to floor diaphragm cracks through the weakest part of the floor system. Damage to equipment in the penthouse could have been reduced by appropriate connection of the equipment to the 1 structure (EERI, 1981d). t C.2 Friuli, Italy, Earthquake of May 6, 1976 -

C.2-1 Earthquake Data Magnitude: Richter Scale 6.5 for the May 6 shock and 6.0 for the two aftershocks of September 15.

Time: Main shock on May 5,1976 at 9:00 p.m. , local time; aftershocks on September 15,1976 at 3:15 a.m. and 9:21 a.m. , local time.

Location: Northeastern Italy C.2-2 Ground Motion Records The ground motions from the main shock and the aftershocks were recorded by a number of accelerograph stations. A maximum peak ground acceleration of 0.37 g was recorded at Forgaria from the May 6 event.

C.2-3 General Effects of the Earthquake The event was centered in an area of high density of towns and villages. 1,000 deathr and 5,000 injuries were reported. Most construction was old (approxi-mately 100 years); however, there were some new industrial and residential complexes in the area. In total, 42,000 structures were destroyed. The pre-ponderance of damage was in residential areas and to older homes.

C.2-4 Electric Power Facilities There are a number of steam generating stations and hydroelectric power plants in the region owned by ENEL. All generating stations in the region tripped.

NUREG-1030 C-5

The same happened to the interconnecting and distribution transformers. Fol-lowing is a description of the effects of the earthquake on these facilities.

Somplago Plant - This is a 180-mVA hydroelectric power plant. The plant buildings suffered some damage in the form of cracks in the roof of the switch-board room, the workshop, the dining hall, and the storehouse; there was also damage from landslide action.

Electrical switchyard equipment was severely damaged: 18 out of the 21 single-

. pole oil circuit breakers came down; the same happened to 209 insulator elements out of a total-of 580 because of porcelain cracking. There were also breaks in the contact sections of the disconnecting devices, in the pneumatic operating mechanisms, and associated pressure lines. The busses were overstressed at the joints.

.Compagnola Hydroelectric Plant - The main damage was incurred by the brickwork 1 and by the hydraulic structures with splits and displacements along the head '

race, in the wicket gates of the overtaking duct, and in the control building.

A set of batteries fell off its stand. The upsetting and displacement of transformers was also noted.

Pireda Plant - There were breaks along the wall of the bypass canal and in the control building. There was minor equipment damage.

Campolessi Plant - There was some building damage. Damaged batteries were also reported.

C.2-5 Power Distribution Systems In the S. Daniele and Buia primary cabins the high voltage transformers weighing 70 to 100 tons were displaced and derailed; in other substations the destruction of insulating elements and the overturning of the battery racks was almost complete. The distribution cabins suffered substantial damage.

C.3 Miyagi-Ken-Oki, Japan, Earthquake of June 12, 1978

. C.3-1 Earthquake Data Magnitude: Richter Scale 7.4 Time: June 12, 1979, at 17h 14m Japanese Standard Time (8h 14m GMT)

Location: 38 degrees 09 minutes N latitude 142 degrees 13 minutes E longitude Focal Depth: 30 km C.3-2 Ground Motion Records Many strong motion instruments recorded the event. The maximum recorded peak ground acceleration was about 0.4 g at Sendai Kokuketsu Building (NRC, June 1983). Intensities 4-5 on the Japan Meteorological Agency scale or 7-8 in MMI occurred in worst hit areas (T. Okubo and O. Masamitsu, cited in EQE, November 1983a).

NUREG-1030 C-6

t C.3-3 General Effects of the Earthquake There were 28 deaths and 11,028 injuries (almost all occurred in Miyagi Prefecture) as a result of the seismic event. Sendai, a modern city of 615,000 people, suffered surprisingly small damage. Most of the damage seemed to correlate with poor local geologic and soil conditions (EERI, December 1978).

C.3-4 Electric Power Facilities Sendai, a large industrial city, had more than 6500 business and manufacturing firms at the time of the earthquake; the facilities investigated represent only a small sample of the structures that were damaged by the earthquake.

The degree of damage observed ranged from negligible (at the Fukushima Nuclear Power Plant) to severe (at the Sendai Gas Facility).

Electric power system damage to utilities was concentrated in Miyagi Prefecture.

Before the earthquake, the Tohoku Electric Power Co. was delivering 4900 mW to the northern portion of Honshu Island. There was approximately a 1,500-mW decrease in demand after the earthquake, including the interruption of some 1,130 mW of supply. System frequency momentarily fluctuated from 50.00 Hz to 50.58 Hz, then returned to normal in 5 minutes. Power service of an estimated 681,600 customers was affected by seismic damage to power system facilities and by operation of relays triggered by the earthquake. These relays were reported to have normally operated and protected the equipment from electrical faults in the system before any equipment was structurally damaged.

Fukushima Nuclear Power Plant Complex - The site is on the Pacific coast, approximately 140 km from the epicenter. Faulting may have extended 60 km west of the epicenter, in which case the plant site may be located about 80 km from the nearest source of energy.

The complex has six nuclear units for a total of 4,700 mW and is the largest nuclear power complex in the world. Units 1 and 6 were instrumented with between 20 and 30 strong motion accelerometers and much valuable information was obtained from the earthquake. The recorded peak ground accleration, which could be considered to be a " free field" acceleration, was 0.125 g. The corresponding accelerations in the north / south direction and up/down directions were 0.100 g and 0.050 g. The strong motion exceeded 30 seconds in duration.

The records were obtained from instruments located on the base slabs of the two units and at downhole instruments, about 30 to 40 m below two of the containments.

The reported maximum response accelerations in the buildings were about 0.50 g.

At the time of the visit by a U.S. reconnaissance team (June 23, 1978 -11 days after the earthquake), Units 1, 2, 3, and 6 were operating; Unit 5 was still under construction but was essentially completed, and it is believed that Unit 4 was scheduled to go into commercial production soon (EERI, December 1978). The plants are founded on a competent soft mudstone formation with a thickness in excess of 300 m. Unit 1 was designed for a peak ground acceleration of 0.18 g and a response spectrum based on Taft record from the southern California (Kern County) earthquake of 1952.

The reconnaissance team inspected the exterior of Unit 1 and the exterior and interior of Unit 6, including the containment structure, the reactor vessel NUREG-1030 C-7

pedestal, some of the equipment on the refueling floor, some of the equipment in the reactor building, the underside of the control rod drive in the contain-ment, miscellaneous critical and non-critical piping, various critical and non-critical cable trays, the reactor building, the turbine building, the overhead crane, and various auxiliary buildings, the turbines and tanks. There was no damage or evidence of working of connections in any of the inspected areas.

The only reported damage to the complex was to some non-critical electrical insulators (EERI, December 1978, Figure 57) some distance to the west of Units 1 and 2.

New Sendai Power Plant, Tohoku Electric Power Co. - This plant is located on the Pacific coast and has two Mitsubishi oil-fired boilers. Unit I was completed in 1971 and has a generating capacity of 350 mW; the 600-mW Unit 2 was completed in 1973 and was the largest of the company's units. The plant's seismic alarm located at the level of the turbine operating floor was trig-gered at approximately 0.15 g. Because the plant is closer.to the epicenter of the earthquake and the assumed area of faulting, it may be assumed that the ground motion was somewhat stronger at the plant than at the city of Sendai, where the recorded peak ground accelerations varied between 0.2 and 0.4 g. The plant is located in an area of recent alluvium and on filled land; the depth of unconsolidated sand is approximately 15 m.

Both units were damaged and the plant was shut down for 6 days. Three types of damage occurred at the plant: (1) damage from local, minor settlement, (2) damage to the structural and architectural elements of buildings which was minor, and (3) damage to the equipment, which constituted the bulk of the loss.

In Units 1 and 2 tubing inside the boilers was damaged. A small furnace platen cooler tube inside the slag screen was sheared in the Unit 1 boiler. A similar failure occurred in the boiler of Unit 2 to one of the reheater spacer tubes.

The suspended boilers and their structural supports also pounded against one another and sustained some damage. There was no other reported significant damage. The turbine pedestal and operating floor of 6he turbine buildings in Japan are usually separated by a 3- to 4-in. gap, and, in this case, there was no pounding between the two structures.

C.3-5 Electrical Substations A total of 18 substations sustained equipment damage to varying degrees, including two 275-kV, seven 154-kV, and nine 66-kV. or lower voltage substations.

The primary cause of extensive power outages in the Sendai area was severe damage to electrical equipment at two of the key bulk power substations, including Sendai Substation. Most of the damage to equipment at these sub-stations was associated with failures of porcelain components.

Sendai Substations, Izumi - This is a multilevel facility built on a site with extensive cut and fill work. Yard equipment in all parts of the facility was extensively damaged. Most damage occurred to various ceramic insulators, lightning arrestors, cicuit breakers, and transformers.

NUREG-1030 C-8

C.3-6 Industrial Facilities l Haranomachi Plant of Sendai City Gas Bureau, Sendai - This facility suffered major damage. The total collapse of a large propane gas holder was primarily responsible for the stoppage of gas service for the city. The collapsed tank caught fire shortly after failure, and all the stored gas was consumed. The fire was extinguished about-25 minutes later. The collapsing tank struck nearby pipeways and other piping systems and equipment, causing much additional damage to the facility. There was evidence of other kinds of damage throughout the facility; however, none of the other tanks at the facility are believed to have suffered major damage.

Sendai Refinery, Tohoku Oil Co., Ltd. - This facility suffered extensive damage from tank ruptures and massive oil spills on the site.

C.3-7 Water Supply System Sendai City bureau of water supply provides potable water to some 200,000 customers from 3 treatment facilities, having a maximum daily capacity of 320,000 cubic meters. Facilities for collection, storage, transmission, and treatment work survived the earthquake without any substantial damage. Power required at treatment facilities was obtained from emergency power units and power outages did not affect service to customers.

C.3-8 Sewer System The sewer system of Sendai serves approximately 60% of the city's population.

The system has 11 main pumping stations where sewage is boosted to a single treatment plant. Although various types of damage were inflicted upon the sewerage system, the single most important seismic effect was the disablement of several pumping stations caused by power outages.

C.4 Campania-Basilicata, Italy, Earthquake of November 23, 1980 C.4-1 Earthquake Data Magnitude: Richter Scale 6.8 Time: November 23, 1980, at 19h 34m local time (18h 34m GMT)

Location: 40 degrees 46 minutes latitude 15 degrees 18 minutes longitude 100 km east of Naples Depth: 10 km C.4 Ground Motion Records The earthquake triggered a number of strong motion accelerographs. There were five shocks in less than 2.5 minutes. The strongest rece-ded motion was 0.35 g at Sturno. The range of recorded ground motions varied from 0.1 g to 0.35 g. The first shock of the five was the largest. The total duration of the five shocks (acceleration greater than 0.05 g) was 147 seconds. A peak ground acceleration of between 0.6 g and 0.7 g was estimated at the epicenter of the event (EERI, July 1981).

NUREG-1030 C-9

C.4-3 General Effects of the Earthquake The earthquake killed approximately 3,000 people and injured about 9,000. The damaged area covered more than 10,000 square kilometers. Damage to housing was severe because of the multiple strong shocks and the lack of seismic resistance for the structures. Much of the damge to lifeline facilities was caused by building failures and the movement of building debris down slopes in the mountain villages.

C.4-4 Electric Power Facilities Most power outages (caused when insulators and conductors broke in the epicentral region) were caused by buildings falling on distribution lines. Two hydroelectric power plants, Tanagro Hydrostation with 27-km epicentral distance and Agri Gener-ating Plant at 100-km epicentral distance, suffered no damage or interruptions.

At Calore Generation Station (43-km ericentral distance) lightni..g arrestors were damaged and conductors were broken.

The Garigliano Nuclear Power Plant located at Sessa Aurunca (125-km epicentral distance) felt the earthquake. The plant is a 150 net MWe General Electric BWR completed in 1962 and is similar to the Dresden 1 (U.S.) plant. Although the plant was in a shutdown condition, the control rod scram system, set at 0.05 g, was actuated by a 0.051 g signal from the vectorial sum seismic device.

This plant was not damaged.

The earthquake was also felt at the Lat;na Nuclear Power Plant, a 150 net MWe, graphite-moderated, gas-cooled reactor unit completed in 1962, located 217 km away from the epicenter. This plant.was also in a shutdown condition for maintenance. However, the safety system was actuated by spurious signals below the set value of 0.03 g, causing the insertion of the (safety) control rods. No evidence of damage or malfunction was found at the plant.

k NUREG-1030 C-10 m _ _ _ _

' 1 l l

N O

S T N N O E I M T M U O L C O S

C E D I R L

X B D I U E D P S N O E F P P O O P R A N P O

I 6 T 4 A -

R A E

D I I S S U N

O C

'.2 f Il 1 11l ll!l l

PUBLIC C0pmENTS ON.NUREG-1030 AND ASSOCIATED' REG. ANALYSISj(A-46)

ORIGINATOR' .NRC STAFF RESP 0MSE COPMENTS A) Applicability of Backfit Rule and Justification for A-46 Review:

A-46 Regulatory Analysis, Page 21 -11* 1. The NRC' staff. position is thatLthe new backfit rule.

1. as required by 10 CFR 50.109 dces: apply to USI-The Group believes that the scope and cost of the proposed USI A-46 resolution have A-46.- The regulatory analysis of A-46 (NUREG-1211)*

not been justified by an appropriate was changed to incorporate.a backfit analysisfwhich-backfitting analysis. Nor has the presents justification for performing!the:

value-impact assessment performed by seismic adequacy' review. Backfit analysis for the Staff been adequate. This-is due to the the correction'of'any deficiency will be; inherent seismic capability of equipment which . performed on'a case by case basis if. required; has been demonstrated by the development of the. follcwing cocpletion of. the review. Based on experience data bases and the resulting' the regulatory analysis performedito support' substantial reduction in the safety the-seismic.cdequacy review, the staff _has:

significance of this issue. concluded that substantial increase in the .

overall. protection of the public health and The Group notes that the NRC Staff has not safety can be achieved and that the direct or conducted a backfitting analysis as required. indirect costs of implementation are justified:

a by 10 CFR 50.109. Section 50.109(a) (3) in view of'this increased protection.'

j, provides: "The Commission shall require the backfitting of a facility only when it detendines, based on the analysis descril ed in paragraph (c) of this section, that there is a substantial increase in the overall protection of the public health and safety...

and that the direct and indirect costs of implementation for that facility are justified in view of this increased protection."

Section 50.109(c), which identifies the various factors which the NRC must adoress, requires. ,

-l consideration of the " potential change in the 1 risk to the public" due to accident releases.

The Commission backfitting policy clearly contemplates more than a simple qualitative evaluation. However, the draft USA A-46 J regulatory analysis states that "[t]he safety benefit of verifying the seismic adequacy of equipment in operating plants was not _

• The numbers here refer to the originators of comments at the end of this document.

n ---

'PUBLIC COMENTS ON NUREG-1030 AND ASSOCIATED-#M. ANALYSIS (A-46)' Pcge 2 COMMENTS ORIGINATOR- NRC STAFF RESPONSE guantified in terms of risk reduction."

LA-46 Regulatory Analysi::, page 21] and further: "Although the incremental risk has not been quantified in this study, the potential for safety improvement exists. The Staff concludes-that the inspection dr.d l verification program outlined would result in I significant safety improvement."

[ Ibid, page 36] The Group does not believe j that these conclusion statements are what is )

contemplated by the backfitting rule. j Aside from the obligations imposed by the backfitting rule, the Group recommends that the Staff perform a quantitative value-impact assessment to justify ti e proposed licensee expenditures. The assessment should consider the risk reduction results of the various P

" seismic PRAs conducted to date including NUREG/CR-3568, "A Handbook for Value-Impact Assessment," December 1983, and NRR Office Letter No.16, Rev. 2, " Regulatory Analysis Guidelines," October 3, 1984. The Group believes that an appropriate Staff analysis meeting the intent of the value-impact guidelines must provide, at a minimum, some quantitative basis which demonstrates that the NRC estimated costs of $400,000 to

$800,000 per plant are reasonable and warranted and result in some justifiable reduction in core-melt probability or uncertainty.

In this regard, a comparison with the NRC Staff analysis in Enclosure 2 to NUREG-1109,

" Regulatory Analysis for the Resolution of Unresolved Safety Issue A-44, Station Blackout," may be helpful . While the Group

PUBLIC COMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 3 COMMENIS ORIGINATOR NRC STAFF RESPONSE Cannot Coment on the accuracy of the-NUREG-1109 analysis, the use of a quantitative value-impact assessment is preferred to the assessment provided for USI A-46 which is devoid of any risk reduction considerations.

The Group does note, however, that the NUREG-1109 analysis suggests that for station a mean reduction of core-melt of blackoug,per 3 x 10- reactor-year justifies, at

$1,000 per averted man-rem, an expenditure of roughly $1,200,000 per plant. That analysis would suggest that for USI A-46 core-melt reductions in the order of 10 5 per reactor-year and greater would be necessary to justify the NRC estimated implementation costs. Because of the inherent rugaedness of the equipment and the overall quality of anchorage design and installation, it is

?

likely that such significant reductions in core-melt cannot be achieved, and therefore a less ambitious generic resolution must be i considered.

2. The new backfit rule, 10 CFR 50.109, which 2. See response in A) 1.

became effective on October 21, 1985, should be applied to the resolution of USI A-46. The regulatory analysis dated August 1985 (attached to NUREG-1030) does not properly quantify the benefits and costs of proposed licensee requirements nor does it justify all required licensee actions, as required by the backfit rule. Specific comments in this regard are as follows:

FUBLIC C0f41ENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 4 COMliENTS ORIGINATOR hkC STAFF RESPONSE

a. The benefits of the required licensee actions are not quantified at all.

NUREG-1030 states thot " margins of safety in existing nuclear power plant equipment to resist seismically induced loads...may vary considerably, and may I not meet current seismic qualification i criteria. Therefore.... seismic Cdpability of equipment in operating plants must be reassessed." This conclusion is not supported by the I

l results of the SQUG program ror is it l

justifieo by a cost / benefit analysis.

b. The costs may be low by a factor of two l due to a low estimate for labor costs (e.g., consultant rates are typically

$100/ hour plus expenses or about

?

$200,000/ man-year not $100,000/ mon-year as assumed in the regulatory analysis).

c. The proposed requirement to conduct a plant-specific review of relay operability has not been justified. Earthquake experience demonstrates that plants may trip during a seismic event, but they can be quickly restored. No benefit from this significant effort has been quantified and the requirement shculd be deleted.

d. The proposed requirerent to conduct a plant-specific ver1'scation of anchorage of tanks and heat exchangers has not been justified. Neither the potential radiation exposure and costs for this inspection nor the assumed benefits

- s PUBLIC C0!EENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-45)_ . Pig 5 5 C0feENTS ORIGINATOR NRC STAFF RESPONSE have been quantified as required by the backfit rule. No problem.in this area has b'en e identified to exist in operating plants to justify this_ effort. These requirements should be deleted.

e. The proposed requirement to document the seismic adequacy of adoitional types of equipment beyond the original eight classes in the SQUG Pilot Program is not justified. Earthquake experience to date demonstrates that industrial-grade power plant equipment is inherently rugged.

In sumary, it is coubtful that anything beyond anchorage verification (walkdowns) of actual equipment can be justified by a quanticative cost / benefit 6nalysis.

• 3. Draft NUREG-1030 should be revised to 3 3. See response in A) 1.

conform to the new requirements of 10 CFR 50.109. Specifically, the document should provide additional details indicating how the new requirements will provide "a substantial increase in the overall protection of the public health 6nd safety or the comon defense" and how the " direct and indirect costs of implementation...are justified in view of this increased protection." For this purpose, the Staff should prepare the evaluation based on the costs associated with implementing these proposed requirements on older plants, thus, dssuring a Conservative Cost / benefit analysis.

I x PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A l5) Page 6 COMMENTS ORIGINATOR NRC STAFF RESPONSE

4. The SQUG effort has provided significant 7 4. See response in A) 1.

evidence that the potential of realizing a safety benefit from further work is l limited to anchorage and relay functionality l verifications. However, implementation of that work should only be required if a quantitative value - impact assessment meeting the intent of 10 CFR 50.109 Backfitting justifies the proposed licensee expenditures. In this regard, Duke fully endorses the comments provided by the Nuclear Utility Group on Equipment Qualification.

5. The first paragraph of Item VII on page 39 9 5. See response in A) 1.

of the Regulatory Analysis indicates an attempt was made by Brookhaven National Laboratory (BNL) a to develop a quantitative basis for estimating the 4 risk reduction due to qualifying equipment but the results were inconclusive. However, item two on page 40 of the Regulatory Analysis states BNL concluded that "...overall seismic induced structural failures and random failures (due to nonseismic causes) contribute more significantly to core melt probability and risk then seismic induced equipment failure." The NRC staff concluded in the last paragraph on page 41 "...that it is not feasible to provide a quantitative estimate of net safety benefit in terms of risk to the public."

Therefore, we recommend that the NRC apply the new Backfit Rule which states in part

"...The Commission shall require the backfitting of a facility Only when it determines...that there is a substantial increase in the overall protection of the public health and safety...

,_ __ ~ , . . . . - . ,

PUBLIC COMMENTS ON NUREG-1030'AND ASSOCIATED REG. ANALYSIS (A-46)' ..Page 7 COPMENTS ORIGINATOR NRC STAFF RESPONSE.

to be derived from the backfit and that the <

direct and indirect costs of implementation for that facility is justified in view of this increased protection."'

Conclusion:

The Draft Generic Letter would impose requirements on utilities to verify seismic adequacy of equipment at a high. cost with questionable safety benefit. We'believe'that this requirement would place an unnecessary burden on the utilities and increase worker exposure to radiation. Therefore, we recommend that the requirements described in the Draft Generic letter be thoroughly tested with the new Backfit Rule before p being imposed.

y

6. The new or revised backfit rule, 10 CFR 50.109 12 6. See response.in A) 1.

which became effective October 21, 1985, has not been applied to the-resolution of USI A-46.

NUREG-1030 and its attendant Regulatory Analysis document does not properly quantify the benefits nor adequately estimate the costs of the proposed licensing requirements.

7. We agree that seismic qualification of equipment 4 7, 8 and 9.

using earthquake experience data is the most Coments 7,' 8 and 9 question the need to do a A-46 reasonable and cost-effective alternative for review since earthquake experience shows'that verifying seismic adequacy. We also agree, as . equipment is inherently rugged and_ not susceptible stated in the Regulatory Analysis.'that equipment tol seismic damage. .

installed in nuclear ' plants is inherently' rugged

- _ - _ = _

-PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46). Page 8-COMMENTS OR161NATOR NRC STAFF RESPONSE and not susceptible to seismic damage. Because The staff ocknowledges that equipnent is inherently of these factors and because the incremental rugged. However, the seismic experience dota risk of significant seismic damage hos not been indicateo that there are still three areas of concern:

quantified, we believe the potential for safety improvement needs to be carefully evaluated. a. odequdcy of equipment onchorage.

b. functional capability of essential relays (essential We believe thot using seismic experience dat6 relays are reldys which must remain functional

~

to resolve this issue will lead to the without chotter during a SSE).

conclusion that no action is necessary at this c. Outlier equipment time, other then possible verifying odequate ,

anchorage. If continued collection of. The scope of A-46 review is reduced to the above three areas experience data indicates a certain equipment of concern for equipment needed to bring 'the plant to hot type or types are subject to seismic damage, shntdown and paintain it there for at least 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. The dpproprldte action can then be token. Until stoff believes thdt the SGUG is in agreement with these then, however, the experience data reported three areas of concern.

in NUREG-1030 and the recognition as stated on page 41 of the Regulatory Analysis "...that is a is not feasible to provide d quantitative e estimate of net safety benefit in terms of risk to the public," do not appear to justify significant expenditures to resolve this issue under the requirements of the recently revised backfitting process.

8. The AIF's Connittee on Power Plant Design, 9 Construction and Operation developed a position paper titled " Seismic Qualification of Safety-Related Equipment in Operating Nucleor Power Plants" and submitteo it to Mr. William Dircks in a letter of June 17, 1983. In that paper, the Committee concluded from all evidence that there is reasonable assurance that the seismic ddequacy of equipment in nucleor pldnts is acceptable with odequate margin for a design basis earthquake. We recommended that the NRC not proceed with ony new requirements for seismic evaluation of equipment in operating-plants since the potential for a significant e

q PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 9 COMMENTS ORIGINATOR NRC STAFF RESPONSE safety problem does not exist. The evidence from more recent earthquakes continues to bear out our conclusions that power plant equipment is inherently rugged and will survive acceleration levels far in excess of building responses associated with the safe shutdown earthquake regardless of whether the equipment was qualified for seismic service.

NUREG-1030 and the accompanying Regulatory Analysis provide a means of seismic qualification of equipinent using seismic experience c:sta.

Although we still consider the proposed requirements to be excessive, we support the use of earthquake experience data as a means to close Unresolved Safety Issue A-46.

c 9. The first sentence of item (2) on page 7 of the 9 4 Regulatory Analysis states, " Seismic experience data collected by SQUG and reviewed by SSRAP, supplemented by reviews and literature surveys of strong motion earthquakes indicate that mechanical and electrical equipment of types commonly used in nuclear power plants are unlikely to fail at earthquake levels typical of SSEs at U.S. plants east of California."

Therefore, it is questionable whether this subject would be treated as a generic issue.

B) Implementation Schedule:

1. The proposed requirement that the program "be 3 1,4,5,6,8.

completed no later than 28 months from the The staff agrees that each utility whether a member of date of issuance of the USI A-46 final resolution" the Generic Group or not, should negotiate with the (attachment of NUREG-1030, page 38) should be staff on the implementation schedule. The proper modified to read, "be ccmpleted in accordance integration of the proposed work scope into the

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46)

Page 10 COMMENTS -ORIGINATOR NRC STAFF RESPONSE-with a schedule to be negotiated with' individual indivicual plant's living schedule for plont licensees. The schedule will consider other modification will be considered.- The text is regulatory activities ongoing et the facility." modified to reflect this.

I 10 CFR 50.109(c) requires the staff to consider other regulatory activities when imposing a new requirement.

2. The last sentence of the second paragraph on 9 2, 3, 9.

page 2 of the Draft Generic Letter states that The 28 months mentioned in the original Regulatory dssessment of seismic dGequacy and ful.Jtional Analysis is now changed to 36 months. This allows Cdpability can be delayed, if DeCessary, ur+,il time for the generic group (SQUG) to fine tune I Wdlk implementation schedule will be negotiated with walk down procedure and conduct workshop.for the-d test data bdse has been developed by EPRI/RES. member utilities before the start of implementation.  !

The actucl implementation schedule will be negotiated This statement is in possible conflict with the with individual utilities. As commented by EFRI in B) 3,.

3. all data and Generic Equipment Ruggedness Spectra stotement in the fourth paragraph on page 13 of the Regulatory Analysis which states that all (GERS) from the EPRI. program are scheduled to be a work will be completed within 28 months of the available by December 1986, therefore this informo-i'

.' issuance of USI A-46 final resolution. The tion will be avoilable before the 36 Ironths mentioned latter statement assuiaes additional test experience above, data will be available from the EPRI/RES program.

There is concern that the utilities would have

difficulty meeting their commitments to the NRC l if the EPRI/RES Program data is not available on schedule.

3. The fourth paragraph on page 13 of the Regulatory 10 Analysis states that additional test experience data will be avdilable from the EPRI/RES program 28 months from the date of issuance of the USI A-46 final resolution. Actually, all data and Generic Ruggedness Spectra from the EPRI program are scheduled to be available by December 1986. This schedule may aiffer from the schedule for completion of the RES test data collection program. It does not appear appropriate to imply, as this paragraph does, that the implementation schedule is closely tied to completion of research programs. The apparent tie-in is stored even more explicitly in the- fourth poregraph of the Drof t Generic Letter.

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 11 COMMENTS ORIGINATOR NRC STAFF RESPONSE

4. Regulatory Analysis, Page 15 11 The Staff's schedular requirements for completion of the seismic verification within 28 months is inconsistent with the NRC policy on integrated schedules for plant modifications and the low safety significance of the proposed plant reviews.

The Group recommends that general schedular goals, rather than specific deadlines, be established for the resolution of this issue.

The final A-46 plant specific schedule should then be based on proper integration of the proposed work scope into the individual plant's living schedule for plant modifications and discussions with individual licensees or SQUG.

a 5. Regulatory Analysis, Page 13: The fourth 12

. paragraph on this page indicates that resolution of outliers can be deferred for a period not to exceed 28 months from the date of issuance of the USI A-46 final resolution. Considering the plant-specific nature of the implementation program, a specific time period of 28 months is considered inappropriate. It is recomacaded that completion dates be established on a plant-specific basis based on plant integrated schedules and other considerations. In any case, considering the fact that the seismic qualification has been determined not to be a significant safety issue, the imposition of a 28 month completion date is not justified. This comment also applies to Paragraph 3 on page 15, page 5 of Appendix A.

6. We endorse the attached comments offered by the 8 Seismic Qualificatian Utility Group, dated November 12, 1985, particularly those regarding the Generic Letter r eference to the schedule for

Pag? 12 PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46)

ORIGINATOR NRC STAFF RESPONSE COMMENTS implementation of requirements (consistency must be maintained with pages 17 and 20 of the Regulatory Analysis) and the recommendation that completion dates be established on a plant-specific basis in the Regulatory Analysis (fourth paragraph of page 13),

since 28 months may not be appropriate for all utilities.

Implementation schedules for resolution of 5 7. Staff agrees with this comment. Text'is modified !

7. to incorporate this. l l

Unresolved Safety Issue A-46 should permit l completion of inspections and any subsequent i

modifications during planned outages when I required. Unplanned shutdowns and greatly extended outages to complete modifications would not appear to be justified by the level of risk that exists from this issue, Regulatory Analysis, page 38: the estimated 12 a' 8.

.. schedule requirements given on this page are

" considered optimistic and are not justified.

It is recommended that schedules for implementation of A-46 be based on discussions with the Generic Group and ultimately with individual utilities. Imposition of specific completion dates is inconsistent with NRC policy on integrated schedules as indicated in Generic Letter 83-20 dated May 9,1983, and is not warranted by the safety importance of this unresolved safety issue.

9. Regulatory Analysis, Page 13 11 Equipment walkdowns should not be required until the EPRI/RES test experience data programs and the SQUG efforts to address the remaining classes of equipment have been completed, reviewed and accepted. This will minimize needless replication of licensee efforts and is justified based on the existing equipment's seismic capability.

n PUBLIC COMMENTS ON NURCG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 13-COIMENTS ORIGINATOR NRC STAFF RESPONSE-

10. The Generic Group has been given 90 days f rom 5 10.This has been revised to reflect that each utility,:

receipt of the A-46 Generic Letter to submit to whether men 4ber of' the Generic Group or not, will' the hRC a generic schedule for implementation, be given 60 days from receipt of the A-46 Generic but individual utilities wculd have only 45 days Letter to subnilt to the NRC 4 schedule for to provide the some response, what is the implementation. Utilities who may not have justificotton of this difference?

occess to SQUG implenentation procedures or dato base niay have difficulty in establishing implementation schedules within 60 days. For

' these utilities the NRC will negotiate time 1

extensions on a case by cose basis.

11. The Generic Group on which the generic letter 5 11.The anchurdge review guideline dna generic depends, in part, hos not yet finalized many implementation procedure should be available at the of its required action items and these shoulo time the ger.eric letter is issued. Test dato and be ir. place before imposing the generic letter. GERS will be availoble arouno December 1986, and a deloy of review of equipment outside the seismic experience ond test dato base is allowed in the Regulatory Analysis.

U 12. In the draft Generic Letter, submittals for a 12 12.See B) 10 obove.

schedule for implementation of requirements should be compatible with that specified on pages 17 and 20 of the Regulatory Analysis.

C) Relay Review Guidelines:

1. Page 33 of Regulatory Analysis next to the last 2 1. This is our best estimate.

item: It is assumed that 10-30 relays would neea to be identified as requiring to function... Is this a good assumption?

2. The third paragraph on page 14 and the first 9 2, 3. The staff ogrees with the comments. The text of paragraph of page 15 of the Regulatory Analysis the Regulatory Anolysis has been revised. See indicate that reloys cre of particulor concern response in C) 6.

due to chattering during a seismic event.

The occeptobility of relay chatter should be determined by an evaluation of the effects of relay chatter on cownstrean equipment. For some applications relay chatter is inconsequential.

,a- -

Page 14 PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46)

ORIGlHATOR NRC STAFF RESPONSE COMMENTS I

The last paragraph on page 14 of the Regulatory 9 3.

Analysis states that relays must function durino l the first 30 seconds of an earthquake. l The time duration for relay functioning should be determined by the safety application of the relay. Some relays need only function following d seismic event while others are needed to dCluate protection functions during the seismic event.

12 4, 5.

4. Page 14 of Regulatory Analysis: the review of Sesed on the seismic experience data gothereo to date, equipment functional capability as presently defined, appears to include equipment or it appears thot the only concern thot remains on components in addition to relays. Based on the equipment functional capability is that regarding experience data available to date, we believe reldys (including contdctors and switches). Unless that the review of equipment functional the test data being collected by EPRI/RES reveals capability should be limited to relays. There anything otherwise, the text is revised to indicate a

,L is no need for such a review of equipment that certain types of relays are the only equipment

• included in the SQUG data base such as motors, whose functional capability needs to be verified, pumps, valves, etc.

Page 9 of Regulatory Anolysis: Figure 1 12 5.

identifies review of relays and other equipment needed for functional capability as an essential cart of the A-46 implementation. It is suggested that relays are the only types of equipment whose only functional capability needs to be verified l

ds part of the implementation of A-46. Functional Cdpobility of other equipment is covered by the SQuG dato base. Examples of such equipment include motors, power operated valves, pumps, motor generator sets, etc. It should be made clear thdt these equipments do not require aoditional functional capability review.

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Pcge 15 COMMENTS ORIGINATOR NRC STAFF RESPONSE

6. Fcges 14 and 15 of Regulatory Analysis: The 12 6, 7.

requirement for review of electrical relays should The staff is in agreement with the comments.

be revised. First, not all relays identified in The section on " Review of Equipment Functional Paragraphs (1) and (2) need be identified as Capability" in the Regulatory Analysis (page 14 to indicated. Instead, the SQUG plan is to (1) page 15) and related entries in the text have been identify essential safe shutdown functions, revised to reflect the present position of SQUG (2) identify associated systems and electrical and the NRC Relay Review Team.

circuits required to provide these functions, (3) evaluate these circuits to determine those relays which are in fact essential (i.e., those relays whose chatter could result in unacceptable consequences), and (4) evaluate the seismic adequacy of only those relays which are determined to be essential. (Steps 3 and 4 may be done in parallel or in a reverse order.) As a consequence of this approach, it is only necessary that the final

" essential" relays be identified. Second, other a alternatives to relay qualification, comparison

,L with test data or replacement are available. These include redesign of circuitry, strengthening of relay supports / cabinets to reduce seismic demand, and relocation of relays to reduce demand. Such other alternatives should be permitted.

7. Page 14 of Regulatory Analysis, last paragraph: 2 Will this significantly change due to the work that SQUG is currently doing? If so, will the A-46 criteria be revised?

8. A-46 Regulatory Analysis, Page 23. 11 8, 9, 10.

As a result of industry efforts, a substantial The staff agrees that credit can be taken for timely basis exists for demonstrating equipment operator action to restore systems and equipment to functionally after an SSE. Furthermore, operable status, provided the following requirements short-tenn equipment inoperability due to are met:

possible relay chatter will likely be generically resolved by crediting operator a. Detailed relay resetting procedures should be action prior to the need of most equipment developed.

post-SSE. This coupled with the inherent b. There is sufficient time for resetting the relays.

Page 16/

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED. REG. ANALYSIS-(A-46)

ORIGINATOR NRC STAFF RESPONSE ,

COMMENTS 4

equipment ruggedness strongly suggests that The text has been revised accordingly,

~

the A-46 resolution process need only address the adequacy of equipment anchorage.

The Group recommends that the proposed plant specific actions necessary to resolve A-46 be 'l limited to such anchorage reviews. The other l issues (e.g., functionality, relay chatter),

if necessary, can be adequately resolved on either a generic or sample basis. l A-46 Regulatory Analysis, Page 14 11 9.

The Group believes that the relay chatter concern will be generically resolved by demonstrating that these effects will not affect required system performance. To the extent relay chatter may temporarily preclude the use of shutdown equipment, timely operator action a should be credited with restoring systems and

' equipment to operable status, cn

10. the proposed requirement to conduct a plant- 1 specific review of relay operability has not been justified. Earthquake experience demonstrates that plants may trip during a seismic event, but they can be quickly restored. No benefit from this significant effort has been quantified and the requirement should be deleted.

D) Scope of Review: 1. As a result of the A-46 study the staff has already The Group commends the NRC and industry efforts 11

1. narrowed down the scope of review to address only to address the adequacy of seismic qualification

.the following:

of equipment by the generic use of seismic experience and text experience data. To date adequacy of equipment anchorage these efforts have provided a substantial a. .

technical basis to support the prevalent industry b. functional capability of essential _ relays (those

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 17 COMMENTS ORIGINATOR NRC STAFF RESPONSE perspective that mechanical and electrical relays which must remain functional without equipment, when properly anchored, can chatter during a SSE) inherently be exoected to function at typical c. Outlier equipment safe shutdown earthquake levels.

The existence of this data base has both significantly increased the level of knowledge on the ability of equipment to function during earthquakes and substantially reduced the safety significance of this topic as an unresolved safety issue. The scope and costs of the efforts necessary to implement the final resolution must consequently reflect the minor significance of any remaining issues associated with USI A-46.

2. The assumptions in NUP.EG-1030 which dictate the 1 2, 3. The staff agrees with the comments. The text systems and equipment within the scope of USI A-46 has been revised to mention that the assumption is a state that a seismic event does not cause a LOCA now that SSE does not cause LOCA, HELB or SLBA. This

,L and that LOCA does not occur simultaneously with is based on:

a seismic event. This assumption should be clarified to include steam-line-break accident a. the seismic margin of piping observed from seismic (SLBA) and other high-energy-line breaks (HELB's), experience data.

This addition is justified by earthquake experience b. IE Bulletin 79-02, 79-07, 79-14 already required which shows that power plant piping systems are review of safety related piping, rugged yet ductile and flexible and do not rupture under typical seismic loads.

3. Page 4 of Regulatory Analysis: Paragraph 2.(1) 12 indicates that the seismic event does not cause a LOCA and a LOCA does not occur simultaneously with or as the result of a seismic event. This paragraph should be further clarified to indicate that main steam and other high energy line breaks are also not assumed to occur simultaneously with the seismic event.

I v i

Page 18 PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46)

ORIGINATOR NRC STAFF RESPONSE COMMENTS

4. A-46 Regulatory Analysis, Page 6 4. The staff agrees with the comment. The reasoning The Group agrees with the NRC's assessment that 11 is the same as those mentioned in response D) 2, 3. l accident mitigation system equipment need not be considered as part of the A-46 resolution process due to the extensive piping system design margins which exist and the inherent ruggedness of electrical and mechanical equipment. The NRC assessment should be clarified to clearly state that such piping margins exist in all seismically supported systems including the Reactor Coolant System.

12 5. The staff agrees with the comment. The text has

5. Page 11 of Regulatory Analysis: A statement been revised accordingly.

should be added after the first sentence on this page which indicates that "This list is based on SQUG polls of member utilities and is expected to include all of the types of a safe shutdown equipment in nuclear power

,. plants; plant-specific lists to be generated

• as part of the implementation of A-46 are expected to be shorter."

12 6. The staff is in agreement with the comment.

6. Page 15 of Regulatory Analysis: With regard to The text has been modified accordingly, the review of equipment unique to the nuclear plants, it noted that some power operated relief valves are unique to nuclear plants while some are not. Examples of those power operated relief valves which are represented in conventional power plants include solenoid actuated valves and Dresser / Crosby type electromatic relief valves.

7. The staff agrees with the comment. However, the staff

7. The scope of equipment in USI A-46 covers active 1 maintains that seismic adequacy can be verified for both mechanical and electrical components. Since electrical penetration assemblies and neutron detectors electrical penetration assemblies and neutron by using seismic experience data or test data. The NRC detectors are passive (i.e., no moving parts), has no objection to their inclusion.

they should be deleted from Table 1, " Typical Equipment List for USI A-46" (Page 10 of the Regulatory Analysis). This equipment should not be required to be seismically qualified.

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 19 COP 94ENTS ORIGINATOR NRC STAFF RESPONSE

8. The proposed requirement to conduct a plant- 1 8. The staff disagrees with the comment. Tanks and specific verification of anchorage of tanks heat exchangers in the scope of A-46 are important and heat exchangers has not been justified. to hot shutdown. Seismic experience data have Neither the potential radiation exposure shown that certain type of tank anchorage is prone and costs for this inspection nor the assumed to failure under seismic loading, therefore the benefits have been quantified as required by review of tank and heat exchanger anchorage is the backfit rule. No problem in this area has necessary.

been identified to exist in operating plants to justify this effort. These requirements should be deleted.

4 9. The proposed requirement to document the seismic adequacy of additional types of equipment beyond the original eight classes in the SQUG Pilot Program is not justified.

Earthquake experience to date demonstrates that industrial grade power plant equipment is inherently rugged.

U 10. Additional documentation of the seismic 1 9, 10.

adequacy of additional classes of equipment Because of inherent ruggedness demonstrated by beyond the original eight classes is required the original eight classes of equipment, no on Page 26 of Appendix A in the Regulatory collection of additional seismic experience Analysis (i.e., Draft Generic Letter). We do data is required by the staff on equipment not believe this is necessary or can be other than the original eight classes. However, justified based on quantitative cost / benefit from the experience of eight classes of equipment, analysis. Limited utility resources could the caveats and exclusions, as well as appror fate be spent to achieve much greater safety bounding spectra have to be identified for al.

benefit by concentrating on the verification equipment classes. In addition, a technical of adequate equipment anchorages. basis for seismic adequacy must be developed for each type.

E) Scope of Walk-Through Inspection:

U PUBLIC COMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 20 COMENTS ORIGINATOR NRC STAFF RESPONSE

1. A-46 Regulatory Analysis, Page 12 11 1, 2, 3. I 10 CFR 50.109(c)(8) presently requires the The SSRAP is considering the option of sampling. I Staff to consider any facility differences This issue will be decided by the staff after I as part of its backfitting review. The the SSRAP has reached a recommendation. There is a i anchorage review presently proposed by the likelihood that the final guideline on this issue

~

l Staff requires a review for all selected will be included in the forthcoming anchorage review equipment in all plants under the scope of guidelines currently being developed by EPRI and USI A-46. This generic resolution does not reviewed by the SSRAP, NRC and SQUG.

address the possible variations in anchorage design and installation in different facilities which could occur based on the age of the facility or the level of design and installation control which were applied.

The proposed anchorage review can adequately address such differences by permitting for each facility a statistical sampling of the adequacy a of equipment anchorage. If deficiencies are A, identified as part of this sampling process a o larger review scope would then be undertaken.

The use of such a sampling method would both appropriately limit the anchorage review for those licensees with adequate installations and insure a full review of those facilities with demonstrated anchorage anomalies.

2. Page 12 of Regulatory Analysis, Anchorage Review: 2 This indicates that all equipment will have anchorage verified. Will we not be able to consider any sampling? Also, will we be able to rely upon documentation from previous walkdowns for other purposes wherein we were able to verify the anchorage of a particular item of equipment?

3. DRAFT NUREG-1030 proposes "a plant walk-through 3 and visual inspection of all items in the equipment scope." The "all items in the equipment scope" should be modified to read

"(1) a representative sample of equipment belonging to the eight types in the seismic

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 21 COMMENTS ORIGINATOR NRC STAFF RESPONSE experience' data base, (2) a representative sample of equipment of type not incluaed in the eight types in the seismic experience data base but which are present in the data base plants, (i.e.,

no seismic experience available.)"

4. The "walkdowns" to verify anchorages should not 4. The staff disagrees with the comment.

be needed if the anchorage information is available Engineering documents do not necessarily from engineering documents. Walkdowns increase reflect the as-built condition. Defects radiation exposure and should be avoided when other in installation can only be detected by a alternatives are available. They also have to be walk-through inspection. However, the delayed until outages. Prior to initiating either of staff agrees that if radiation is a concern the above, more guidance as to what are acceptable for the inspection of certain equipment the anchorages should be provided, inspection can be performed during outages.

See response in B) 7. Also, the anchorage review guidelines should be available before the commencement of A-46 implementation.

[5.

~

The second paragraph on page 1 and the first paragraph on page 7 of the Regulatory Analysis 9 5. It should be noted that:

a. IE Information Notice 80-21 is for Class 1E make reference to IE Notice 80-21 and IE electrical equipment only, and it is only Bulletins 79-02 and 79-14 respectively and for information.

state that these documents address the b. IE Bulletins 79-02 and 79-14 are for piping verification of anchorage issue. The second review only. They supposedly do not cover sentence of the second paragraph of page 2 of equipment anchorage review.

the Draft Generic Letter requires a walkdown to verify anchorage. However, the staff will give credit to previous equipment walk-through inspection to the extent Anchorage of equipment during a seismic event that they address the concerns in the anchorage is a recognized issue that the industry has review guidelines under development by EPRI.

been addressing for sometime through the above mentioned NRC documents. Therefore the NRC should consider substituting activities already conducted with respect to these documents in place of additional walkdown requirements.

6. Page 12 of the Regulatory Analysis, 12 6, 7, 8.

(1) Anchorage Review: The staff endorses The staff is in agreement with the comments. The that the Generic Group will develop a text has been modified to reflect this. Also see detailed walk-tnrough procedure. This response in E) 1, 2, 3.

vr -

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 22 COMMENTS ORIGINATOR NRC STAFF RESPONSE procedure is still under development and may allow a sampling type of anchorage verification depending on site specific considerations. However, the word all is underlined and appears to preclude

! ongoing efforts to simplify the walk-downs.

I Suggest rewording to just state that " equipment anchorage must be verified in accordance with walkdown procedures to be developed by SQUG and approved by SSRAP." l

7. Page 17 of Regulatory Analysis, first 12 paragraph: Comment 20 applies to this paragraph as well in regard to use of "all."

8. Page 20 of Regulatory Analysis, Last 12 Paragraph: Again the word all. Not only does the staff appear unwilTiiig to simplify a

plant walkdowns (see comment 20), it appears 4 bent on conducting complete plant reviews.

m This appears unnecessary. Suggest deleting the word all.

F) Requirement to Submit Justification for Continued Operation:

1. The requirement to submit a justification for 1 1,2,3,4,5.

continued operation (JCO) (Page 18 of the a. The staff believes that the concern on the Regulatory Analysis) for each identified requirement to submit JC0 eminates from the deficiency, most of which are expected to nisconception of the definition of the word be minor in nature, is a waste of valuable " deficiency." " Deficiency" is to be engineering resources. These resources would differentiated from " outlier," and this is be better spent correcting the deficiency in clarified further in the revised section of a more timely manner. Reporting requirements " Identification and Review of Deficiencies already exist in 10 CFR 50.73 and operability and Outliers" (page 12 of the Regulatory requirements are written into the plant Technical Analysis).

Specifications to handle significant deficiencies, b. The JC0 will be required for proven deficiencies, which compromise plant safety. The requirement if these deficiencies are not corrected within for JCO's should be deleted. 30 days.

1

l.

lPage 231 PUBLIC C0094ENTS ON NUREG-1030 AND ASSOCIATED' REG. ANALYSIS-(A-46)- ,

C009 TENTS ORIGINATOR NRC STAFF RESPONSE-

2. Page 18, paragraph G of Regulatory Analysis: 2 This paragraph discusses the writing of. JCO's for identified deficiencies. I do not: feel this is necessary. Rather we should promise to resolve-the deficiency by a certain date.. A JCO, in my opinion, is not necessary because, as the NRC has concluded, the equipment is inherently rugged.and not susceptible to seismic damage. Promises to correct the deficiencies in a rea ,onable amount of time should be a reasonable alternative to JCO's.

3. The proposed regulation requires submittal to the 3 NRC of justification (s) for continued operation (JCO) for identified deficiencies, modifications and replacements of equipment / anchorages as a result of the reviews, and the proposed. schedule for a required modifications and replacements not completed A, at the time of the report submittal. In our view, w there should not be a requirement that these JC0's be submitted to the NRC, but rather be kept at the licensee's file where they would be available for NRC review. Reportability of deficiencies should be handled similar to the guidance provided in Generic Letter 85-15, which states "... licensees should report the finding if the condition found meets the reporting criteria of 10 CFR 50.72 (Prompt Notification) or 10 CFR 50.73-(Licensee Reporting System)..."

4. A-46 Regulatory Analysis Appendix A, Pages 11-8 & 10.

The Group does not believe that justifications , _

for continued operation (JCO) need be submitted to the NRC for all identified deficiencies.and outliers. This requirement' duplicates, for major

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page?24' COMMENTS ORIGINATOR NRC STAFF RESPONSE deficiencies, the reporting requirements of 10 CFR 50.72 and 50.73 and requires unnecessary submittals for all minor discrepancies. When one recognizes that major equipment failures were not found during the development of the SQUG experience data base, such minor discrepancies have little impact on operability conclusions.

The Group recommends that the reporting requirements for deficiencies be limited to that -

presently required by 10 CFR 50.72 and 50.73.

5. Page 18 of Regulatory Analysis: - Paragraph (g) 12

! This paragraph seems to state that JC0's are required for all identified deficiencies. This l should be clarified in that it is antic.g.ted that i most deficiencies will not present a serious safety concern and therefore should not require a JC0's. This comment also applies to paragraph 43 (4) on Page 8, Appendix A.

a G) Cost Estimate:

1. A-46 Regulatory Analysis, Page 26 11 1. If radiation is a concern for the walk-through

As part of its cost analysis, the Staff failed inspection of certain equipment, then the inspection to address anticipated facility downtime of these equipment can be delayed until the planned resulting from the proposed seismic revie outage. See responses in B) 7 and E) 4.

or any resulting equipment modifications.ys The 2. The staff is in disagreement with the comment. The.

Group agrees that extended outages should not costs for the review of the Systematic Evaluation

typically be required for the A-46 implementa- Program (SEP) plants should be more expensive as tion due to the minor safety improvements that compared to later plants because

i

! 2/The cost of facility downtime is an appropriate l consideration for purposes of backfitting analysis under 10 CFR 50.109(c)(5).

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page:25 COMMENTS ORIGINATOR NRC STAFF RESPONSE will result and the appropriate use of integrated 4. SEP plants are typically older. Equipment and modification schedules. Absent any specific anchorage in.later plants have better seismic inclusion of facility downtire in the backfitting capability.

dnalysis, the Staff should clearly state that b. A-46 approach is for analysis before review.

plant specific implementation schedules need not Guidelines and procedures so developed will require extended outages in order to meet plant minimize review time and cost, specific deadlines.

2. A-46 Regulatory Analysis, Page 33. The NRC's 11 i

estimated cost to licensees appears to under-estimate the scope of effort required. In

! particular, the walkdown anchorage reviews and miscellaneous equipment modifications are collectively estimated at.$242,000 to

$500,000 per plant. As part of the Systematic Evaluation Program, the review of equipment

. structural adequacy required expenditures of c, approximately $750,000 to $1,000,000 per plant.

3 A, This value, which is substantially higher than o' the values tabulated in the NRC cost estimate, suggests that average plant costs will likely be in excess of $1,200,000.

3. Page 33 of Reguletory Analysis - The cost for defining systems, subsystems,'and components 2 3, 4, 5.

estimated that $17,000-$35,000 seems low. The review cost of A-46 is based on the staff's best estimate. the fact that extensive analysis and

4. Page 26-35 of Regulatory Analysis -Present cost guidelines are developeo "up front" before the review estimates which are part of the NRC cost-benefit should reduce sizable labor cost.

dnalysis. While we agree that the selected approach offers considerable advantages over meeting current licensing requirements, we concur with the comment in the first paragraph of page 30 that the NRC cost estimates are low by a substantial amount.

)

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) :Page 26?

COMMENTS ORIGINATOR NRC STAFF RESPONSE

5. The costs may be low by a factor of two due to a 1 low estimate for labor costs (e.g., consultant rates are typically $100/ hour plus expenses or about $200,000/ man-year not $100,000/ man-year as assumed in the regulatory analysis). .

H) Guidelines for Replacement Equipment:

1. The NRC needs to clarify their position on 2 1,2,3,4 whether or not the USI A-46 position can be The staff is in agreement with the comments.

used to seismically qualify new equipment. Regulatory Analysis has been revised to state The wording on this topic appears differently that A-46 criteria are acceptable to seismically in two locations in the Regulatory Analysis. qualify replacement equipment regardless of The wording on Page 4 indicates that reason of replacement.

qualification by current licensing criteria is an option and that new equipment can be seismically qualified by using A-46 criteria.

c, However, on Page 6 of Appendix A, the wording ,

A, is such that the seismic qualifications would I o' be applicable only to items existing or replaced as a result of the implementation

, of A-46. It is desirable and would be far reaching if we could seismically qualify replacement equipment by using A-46 criteria regardless of the reason of the replacement.

l Current licensing criteria is often far more severe.

2. Guidelines for the future replacement, 5 modification or relocation of equipment

initially qualified by the Experience Data Base should be included.

3. Page 4 of Regulatory Analysis: The third 12

! paragraph on this page indicates that for replacement of equipment and/or parts in plants subject tc A-46 requirements, future replacements must be verified for seismic 4

PUBLIC C0Pt4ENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46)' Page 27 COMMENTS ORIGINATOR NRC STAFF RESPONSE

- adequacy either by using A-46 criteria and methods or as an option, qualification by current licensing criteria. In discussions with NRC and CRGR staff representatives in i Williamsburg, Virginia, in July 1985, it was clarified that this requirement should apply l to replacement of equipment and/or parts made necessary as a result of resolution of USI A-46, as well as to replacements performed for other reasons - both regulatc,ry and non-regulatory . It is recommended that this paragraph be clarified to indicate that these alternative criteria apply to replacements made for any reason. This comment also applies to the first paragraph of page 6, Appendix A.

4. Page 15 of Regulatory Analysis: The discussion 12 e, of requirements for replacement parts should be A3 clarified to indicate that replacement parts

" installed as a result of the A-46 review or for other reasons may be verified using A-46 criteria or as an option, qualification by current licensing criteria.

I) Safe Shutdown Requirement:

1. A-46 Regulatory Analysis, Page 5. 11 1. The staff agrees that the utilities can take The proposed resolution to A-46 should permit credit for timely operator actions to the use of timely operator actions to demonstrate demonstrate the achievement and maintenance of the achievement and maintenance of hot shutdown hot shutdown so long as if procedures are for 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. This is consistent with the available and there is sufficient time.

plant systems analysis approach used to meet See response in C) 8, 9, 10.

the NRC fire protection requirements (10 CFR 50.48 and Appendix R) which also demonstrated that plants can achieve and maintain safe shutdown for 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.

2. A-46 Regulatory Analysis, Page 7. 11 2. The staff disagrees with the comment. The Staff The requirement to demonstrate the achievement maintains that, in the event that maintaining safe and maintenance of safe shutdown, assuming a shutdown is dependent on a single (not redundant) non-seismically related single random component component whose failure, either due to seismic

PUBLIC COPEENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 28 COMMENTS 0RIGINATOR NRC STAFF RESPONSE failure, should be eliminated. Given the low loads or random failure, would preclude decay heat probability of potentially disabling earthquakes removal by the identified means, the licensee and the inherent ruggedness of equipment, a should show that at least one practical alternative focused review limited to a single train to for achieving and maintaining safe shutdown exists safe shutdown would provide the necessary which is not dependent on that component, assurance that facilities can be safely shut down.

3. Page 7, Section 3, third sentence of Regulatory 12 Analysis: This statement appears to refer to 3. The intention is to include only active components passive not active components. The concern in this statement. The text has been revised accordingly.

should not affect resolution of USI A-46.

It is the basis for plant licensing considering single failure criteria.

4. Page 8, Paragraph 3 of Regulatory Analysis, Requirements for Plant Shutdown, lists four 12 4. The staff is in agreement with the comment. The functions required to be perfonned in text has been revised to read " Provide AC and/or conjunction with an earthquake. These DC emergency power as needed on a plant-specific a basis."

4 include bringing the plant to hot shutdown, m maintaining support systems necessary for hot shutdown, maintaining control room functions, including instrumentation and controls, necessary to monitor hot shutdown and providing AC and DC emergency power. the latter

' function should not be included in the basic functions listed, since the need for AC and/or DC emergency power in order to meet the first three functions is a plant-specific consideration.

As an example, AC emergency power may not be i required to meet the basic safe shutdown functions in some plants for a considerable time after the earthquake. It is recommended that item (4) therefore be deleted from this paragraph.

J) Equipment Seismic Demand and Seismic Capacity:

1. Regulatory Analysis, Page 11, bottom paragraph, 2 2/3 way down: " Appropriate generic bounding 1. The appropriate generic bounding spectra for spectra are under development." When are equipment other than the eight classes should' these to be available, and will this affect the be available the early part of 1987 for use in the implementation program.

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 29 COMMENTS ORIGINATOR NRC STAFF RESPONSE schedule that is later spelled out in this document? Do I conclude from this paragraph that if these bounding spectra are not available, we would use type A for equipment not included in the original eight classes? This should be clarified.

2. Appendix A, Page 6, of Regulatory Analysis, second 2 2. Yes, floor spectra can be used throughout for to the last paragraph: Can we conclude from this the verification of anchorage. The proper that alternatively we can use floor spectra application of this floor spectra to the estimation.

throughout for the verification of anchorage? of anchorage capacity will be discribed in the ,

anchorage review guidelines currently being developed by EPRI.

3. Utilities should not have to adopt the " Generic 5 3. The staff is in agreement with this comment.

Floor Response Spectra" if they can show that their own spectra are less.

) 0 4. Page 11 of Regulatory Analysis: The discussion 12 4, 5.

k regarding use of floor response spectra on this The staff will consider the SQUG proposal when

• page should recognize that SQUG is considering available. The SSRAP recommendation on this efforts to extend application of SSRAP bounding matter will be factored in the consideration.

spectra to equipment installed at elevations It is premature to include this in the text.

greater than 40 feet above grade by the use of 1 appropriate amplification factors. This approach should be permitted for equipment mounted over 40 feet above grade with SSRAP/NRC approval. <

l 5. Page 16 of Regulatory Analysis: comment 5 above 12 l applies to the third paragraph on this page l regarding treatment of equipment mounted

' greater than 40 feet above grade.

4

6. The degree of rigidity of the component supports 5 6. The staff agrees that there is no need to analyze should not have to be analysed because the effort the rigidity of each component support within an to determine any possible amplification of response equipment. However, obviously weak component spectra would not provide meaningful information supports will be identified during the walk-since the components they support are being through inspection.

qualified from an Experience Data Base which does not include this type of support information. Also, it would imply the " qualification" of non-rigid supports with the possibility of having to maintain that " qualification" during future plant modifications.

PUBLIC COMENTS ON NUREG-1030 AND ASSOCIATED REG ANALYSIS (A-46) Page 30 COMMENTS ORIGINATOR NRC STAFF RESPONSE-

7. NUREG-1030 Page 11: The statement that "These 13 7. If only seismic experience data are used for A-46 generic bounding spectra will not exceed the review, then type A bounding spectra are the type A bounding spectra...," should be justified. upper limit. This limit may be raised if justifiable information, such as test data, can be provided.

K) Make-up of Walkthough Inspection Team:  ;

1. In the specified composition of the plant 1 1, 2, ? , 4.

walkthrough inspection team on Page 12 of the a. The staff is in agreement with the comment, that it is Draft Generic Letter, there appears to be no not required to have an operations supervisor or SR0 need for an operations supervisor /SR0 on the on the walk-through inspection team. However, they team. Although a SR0 could be useful in should be available for consultation before and during generating the list of required equipment, this the walk-through process, list would have already been established before b. There is no attempt on the part of the staff to restrict' the walkthrough starts. In addition, the the walk-through inspection team members tc degreed engineering members of the team should not be engineers. The emphasis here is on " relevant experience."

restricted to engineers degreed in mechanical c. The staff agrees that all members of the inspection team c and electrical engineering, as long as they are not required to participate in all parts of the

, have relevant knowledge and experience, walk-through. However, appropriate technical expertise o should be included for each review area, and a person with' proper structural background should always be present to inspect the anchorage for all equipment.

2. The NRC should not require a plant operations 3 supervisor or a licensed Senior Reactor Operator The text has been revised to include the above points.

to be part of the inspection team. The knowledge of such an individual would be useful during the development of the list of equipment. However, once the list has been developed and the equipment identified, the marginal value of the knowledge of such an individual may be smaller than the marginal value of an individual experienced in the areas of seismic anchorage, plant mechanical equipment, and plant electrical equipment.

In addition, the requirements should not limit team members to degreed engineers. In our judgement, individuals with degrees'in other physical sciences and/or with considerable years of experience in these areas may have the knowledge to perfonn the walk-through inspection.

PUBLIC COP 9lENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46)' Page 31 COMMENTS ORIGINATOR NRC STAFF RESPONSE l 3. A-46 Regulatory Analysis, Page 12 The Group does not agree that the walk-through 11 team must be composed of the members suggested 4

by the NRC. Walkdowns of electrical equipment need not require the participation of a mechanical engineer. Similarly, electrical engineers need i not participate in the mechanical equipment walkdowns. Finally, while an operations

. superintendent, or Senior Reactor Operator, may i participate in the equipment selection process, j their contribution to the walk-through. anchorage i review remains questionable. The Group recommends that the team composition be based on the requirements of the final walk-through procedures instead of the Staff proposed team composition.

l a 4. Appendix A, Page 3, of Regulatory Analysis, Last 12 2, Paragraph: It should be clarified that all members

~ of the inspection team are not required to participate in all parts of the walk-through.

L) Expansion of Seismic Experience Data Base:

i a 1. NUREG-1030, Page 1-5, paragraph 1.3.3.3: SQUG 2 1. Yes, the 21 classes will be sufficient to cover l

has increased the database from eight classes everything required by A-46 review, except possibly t to 21. Can we conclude that the 21 classes will for outliers.

l now be sufficient to cover everything required?

l 2. The documents refer to the caveats described by 2 2. The Staff agrees with the coment. A statement to SSRAP. These caveats are bound to change and that effect will be added in the text.

most probably relaxed because of the investigation j

of the Chilean earthquake.

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 32 COMMENTS ORIGINATOR NRC STAFF RESPONSE

3. Pages 1-5 of NUREG 1030: Paragraph 1.3.3, 12 3. The staff is in agreement with the comment. The pertaining to equipment outside the scope of the text will be changed to reflect this.

seismic experience data base, implies that the scope of the equipment covered by the experience data base is limited to the initial eight equipment classes covered explicitly in the SQUG pilot program. This paragraph should be clarified consistent with our understanding of the final staff position endorsed by CRGR. Specifically, it should indicate that gathering of additional experience data on equipment not in the original eight classes of equipment may be done by documentation of engineering judgment that the eight classes studied in the SQUG pilot program are similar to and representative of a broader cross section of typical mechanical and electrical equipment.

Documentation of specific extension of experience e data to other classes of equipment by this method a would be reviewed and concurred in by SSRAP and the N NRC staff.

M) Role of SQUG in Generic Implement 3 tion:

1. I am concerned about the possibility of SQUG 2 1. The staff agrees with this comment.

becoming involvea as an enforcement agency for the NRC. SQUG should not be in the position to enforce the requirements of the A-46 position, however, it should be in a position to provide implementation criteria and assistance. SQUG should not be in a position to require a utility to correct a deficiency; such enforcement should be done by the NRC.

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 33-COMMENTS ORIGINATOR NRC STAFF RESPONSE

2. Page 17: Paragraph 6 (3) indicates that the 12 2. The staff is in agreement with the comment. Each Generic Groups will assume responsibility for utility, whether it is a member of the Generic the implementation and will make provision Group or not, is now required to negotiate for systematic and consistent plant-specific individual implementation schedule with the NRC.

reviews. Since it is not possible for the See response to comments B) 1, 4, 5, 6, 8.

Generic Group to assume responsibility for the implementation on specific plants, this should be revised as follows: "The Generic Group would i develop generic implementation procedures for the systematic and consistent plant-specific reviews."

Similarly, Subparagraph (a) of this paragraph should indicate that the Generic Group would submit to the NRC a generic schedule for development of implementation procedures and for training seminars for participating utilities, in lieu of requiring the Generic Group to submit a generic schedule for a implementation of A-46 requirements.

L

" 3. Page 19: Subparagraph (h) specifies 12 3. The positions of SQUG and SSRAP suggested certification requirements for the Generic Group. here are acceptable to the staff; as long Since each participating utility is responsible as SQUG and SSRAP report on their reviews and for his own plant, the requirement for Generic audits to the NRC.

Group certification of the completion of walk-through certification or the completion of walk-through inspections by individual utilities is not considered appropriate. The Generic Group would provide a report of audits performed and results of this audits. Similarly, rather than requiring endorsement of the Generic Group audit report by SSRAP, it is suggested that SSRAP should be required to report on results of any reviews and audits performed by them.

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 34 COMMENTS ORIGINATOR NRC STAFF RESPONSE N) Accessibility of SQUG Results to Non-SQUG Members:

10 Page 8, paragraph 4 of Regulatory Analysis: 2 1, 2. l The document reads "The results of the SQUG The text in the Regulatory Analysis has been revised I (generic group) and EPRI/RES Study will be to reflect this. However, work sponsored by NRC/RES l (Office of Nuclear Regulatory Research) is publicly l accessible to all utilities." The results of the SQUG effort should only be available available.

to those who paid for it. That is, SQUG members.

2. Page 8 of Regulatory Analysis: Paragraph 4 12 indicates that the results of the SQUG (Generic Group) study will be accessible to all utilities. It is not clear at this time that results of the SQUG program will be accessible to non-member utilities. Further, c, the generic implementation procedures will g, likely not be available to non-member utilities.

4= This paragraph should be revised accordingly.

0) Plant Specific SERs

1. Regulatory Analysis: Page 19, paragraph J: Will 2 1, 2.

the NRC write plant-specific SER's to closeout The staff intends to prepare plant-specific SER's the A-46 issue? to close out the A-46 issues.

2. Regulatory Analysis: Page 19, the generic 12 resolution plan should indicate whether plant-specific SERs will be prepared by the NRC staff.

PUBLIC COMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 35 COP 9 TENTS ORIGINATOR NRC STAFF RESPONSE

, P) Applicability of A-46 to New Plants /New i Equipment:

le A-46 Regulatory Analysis, Page 21

The Group notes that the Staff did not 11 1. In the proposed change to IEEE Standard 344, a 4

specifically address 10 CFR 50.109(c)(9) section is added on the use of seismic experience which requires a clear statement that the data for seismic qualification of electrical j proposed backfit is either an interim or a equipment. NRC may accept this position through i final resolution. We assume that the proposed the endorsement of the upcoming revised standard.

resolution of USI A-46 is a final resolution It should be noted however, that A-46 does not and will, absent additional information, form address changes to current requirements, and it is the Staff guidance on the methods which can intended to be applied to operating plants only, be used to support the seismic qualification of plant equipment including new and replacement equipment. Since the methodology is considered as an acceptable substitute for current licensing a criteria, the Group recommends that the Staff 4, clearly state that the USI A-46 methodology is 4

m considered as an acceptable method of complying with Regulatory Guide 1.100 and Standard Review Plan Section 3.10 for there plants. Appropriate revisions to these two documents should be initiated to incorporate the A-46 methodology.

Q) Applicability of A-46 to Specific plants.

1. We believe that draft NUREG-1030 fails to justify 5 1. The staff disagrees with the comment. The seismic i

the generic application of the expensive walk- design bases is not evaluated by A-46.

1 downs and analyses that are outlined in the

! attachment to NUREG-1030. If the problems envisioned by Unresolved Safety Issue A-46 exists at all, they would exist only for a few plants I

in high seismic risk locations. Also, plants located in a relatively aseismic area such as Florida should be excluded from consideration in the resolution of any seismic risk issue.

f

PUBLIC C0titENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 36 COMMENTS ORIGINATOR NRC STAFF RESPONSE

2. With regard to the plants that are identified 7 2. If evidence is presented to and agreed upon by the as being within the scope of USI A-46 Duke Power Staff, then McGuire will be removed from the A-46 Company notes that the McGuire Nuclear Station plant list.

should not be included. Consistent with Paragraph I of Section IV of the Regulatory Analysis (pages 3 & 4), the McGuire Nuclear Station was evaluated to current requirements (IEEE 344-1975 and Regulatory Guide 1.100) and found to be in compliance as documented in NUREG-0422, (McGuire SER dated March 1978, i and SER Supplement 6, dated February,1983).

t R) Specific Coments on NUREG-1030:

1. Page 2-2, Table 2.1-1: Only 20 SQUG members 2 1. This table has been updated according to coment a are listed. This table should be updated. R) 8.

2. Page 2-13, Table 2.1-4B, Item No. 11: This 2 2. This conclusion is based on the result of a LLNL table indicates that the score on experience study sponsored by the NRC (NUREG/CR-3017.

data is equal to the score on current " Correlation of Seismic Experience Data in requirement. I feel the score on experience Non-nuclear Facilities with Seismic Equipment data should be higher than the score on Qualification in Nuclear Plants," August 1983).

current requirement.

3. Page 2-27, Goal 3, Conclusion, first line: The 2 3. This error is corrected.

word " equipment" is misspelled.

4. Page 2-61: The un-numbered table at the top of 2 4. The unit substation transformers should use the the page does not include unit substation type B seismic motion bound. A typographic error transfomers; the eighth class. Where would in this table caused the confusion, these fit in the seismic motion bound criteria?

5. Recorrend modification to the abstract to 12 5. The recommendation is accepted and the text include the first paragraph of section 1.3 on has been modified accordingly.

page 1-3.

6. Page 1-3, top of page: "In February,1984, SSRAP 12 6. "In February, 1984,..." has been changed to released its draft report..." Please include the "In January , 1985. . . "

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 37 ComENTS ORIGINATOR NRC STAFF RESPONSE word draft if February,1984 is the date otherwise use February,1985 for release of the SSRAP report. Please check the document '

for similar errors elsewhere.

7. Page 1-4, Section 1.3.1: It is recommended 12 7. The staff believes it is more appropriate to list that the complete set of conclusions be caveats under each class of equipment. However, presented here, including caveats and a sentence has been added in Section 1.3.1 to exclusions. Later sections could then refer specific caveats to Sections 2.1.5.2 to i refer to this section. 2.1.5.8. l

8. Page 2-2. Table 2.1-1: The list of SQUG 12 8. The old Table 2.1-1 has been replaced by the new members is out of date and contains errors, one suggested.

Please use the list provided below, o SEISHIC QUALIFICATION UTILITY GROUP LIST American Electric Power Co. Nebraska Public Power District Arkansas Power & Light Co. New York Power Authority Baltimore Gas & Electric Co. Niagara Mohawk Power Corp.

Boston Edison Co. Northeast Utilities Service Co.

Carolina Power & Light Co. Northern States Power Co.

Central Electricity Generating Board Omaha Public Power District Comonwealth Edison Co. Philadelphia Electric Co.

Consolidated Edison Co. Public Service Electric & Gas Co.

Consumers Power Co. Rochester ' Gas & Electric Co.

Detroit Edison Co. Sacramento Municipal Utility District Duke Power Co. Southern California Edison Co.

ENEL ctn/NIRA Tennessee Valley Authority Florida Power Corp. Toledo Edison Co.

Georgia Power Co. Vermont Yankee Nuclear Power Corp.

GPU Nuclear Corp. Virginia Power Co.

INTERCOM /Electrobel Wisconsin Electric Power Co.

Iowa Elec. Light & Power Co. Wisconsin Public Service Corp.

Maine Yankee Atomic Power Co. Yankee Atomic Electric Co.

PUBLIC COMMENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 38 C0f9 TENTS ORIGINATOR NRC STAFF RESPONSE

9. Page 2-29, top of page: Please revise as 12 9. The text has now been corrected of these underlined. typographical errors.

Chairman - Robert P. Kennedy (Structural Mechanics Associates Vice-Chairman - Walter A. von Riesemann (Sandia NationaT Laboratories

10. Page 2-54, Section 2.1.4.3.1: The discussion of 12 10.

the questionnaire and the attendant responses a. The questionaire was limited to the original tend to be misleading. Reporting of failures 7 types of equipment only. However, as was not limited to the original 8 classes of mentioned in the NUREG, the responses revealed the Pilot Program. Nor is it clear how the more than performance of the original 7 word " study" is being applied. types of equipment.

2$ b. The word " study" has now been changed to ca " survey."

11. Page 2-56: There are no conclusions derived 12 11. Nothing more can be said now, because the records from the citing of the failures enumerated on the 1964 Alaskan earthquake were found in most at the bottom of page 2-56. It appears that part, incomplete. However, it was indicated in some explanation is in order. Section 2.1.4.3.4 of NUREG that "such failures were very few and did not indicate any trend."

12. Page 2-63 and 2-64, Section 2.1.5.6, Section 12 12. The staff agrees with the comment. The text has 2.1.5.7: Data is being expanded to include been modified to reflect this, results from Chilean earthquake. These sections should reflect that fact.

13. There is frequent mention of using previous 10 13. A summary description of the EPRI test data program qualification test data in the implementation has been added to NUREG-1030.

of USI A-46 (p.1-2,1-5, 2-1, etc.) and reference is made to test data developed by EPRI in both the Regulatory Analysis (p. 8, 9) and Draft Generic Letter (p. 2).

PUBLIC CON 9ENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) P:ge 39 CGOIENTS ORIGINATOR MRC STAFF RESPONSE However, there is no summary description of the EPRI work similar to the susanary description in NUREG-1030 of other technical work which supports USI A-46 technical resolution. It would appear that a sununary description of the EPRI test data program would benefit the user of NUREG-1030.

14. Abstract 13 14. This has been done in the text. See responses to R) 5.

I suggest modification to the abstract to include details from the first paragraph of 1.3 and that walkdowns will be required.

15. P. 1-4 Section 1.3.1: I strongly recomunend 13 15. Sec responses to comunents R) 7 and R) 6.

a that the full set of conclusions be presented 4, here and later sections refer to this section.

• Also the January 1985 revised version of the SSRAP report should be used.

16. Page 1-4. Section 1.3.2: .... bring the plant 13 16. The text has been modified to safe hot shutdown.... according to the comument.

17. Page 2-29: Sandia National Laboratories 13 17. See response to comument R) 9.

18. Page 2-54: It is stated that "few failures 13 18. See response to comument R) 11.

were found relevant to" the study of the eight classes of equipment. Since the study is being extended, the failures are now of interest.

Explanations are in order.

19. Page 2-56: For the reported failures related to 13 19. See response to comunent R) 11. <

the seven basic types of equipment, what conclusions are made?

20. The Seismic Experisce Data Base should be expanded 5 20. The experience data base to include all equipment requiring qualification. (including both seismic experience Equipment unique to nuclear plants should be included data and test experience data) is by analysis and comparison with non-nuclear equipment now expanded to 21 classes, which having similarities. All equipment would be should cover all equipment in the

PUBLIC C0f7 TENTS ON NUREG-1030 AND ASSOCIATED REG. ANALYSIS (A-46) Page 40 COW.Et<TS ORIGINATOR hRC STAFF RESPONSE qualifiec from the Seismic Experience Data Base to scope of A 46 review, except for avoid individual utilities having to do extensive outliers: See response to analysis and duplication of effort. comecent L) 1.

S) Specific Cox.ents on Regulatary Analysis:

1. Page 42, 43, and 44: These figures ought to be enlarged 2 1. These figures v:ere reproduced from for readability. the BNL report (NUREG/CR-3357,

" Identification of seismically Risk Sensitive Systems and Components in Nuclear Power Plants Feasibility I

Study", June 1983). A better copy is included in the final print to assure better readability.

i

2. Enclosure irrediately following PaSe 44: It would be 2 2. The staff agrees that this is a helpful if the plants covered by the SQUG program were very good suggestion. The text has i c identified on this table. been modified to incorporate this.

l 1 o 3. Appendix A, Page 2, last paragraph: This paragraph should 2 3. The "45 days" stated in the text is I be rephrased to reflect the different requirements for the being changed to "60 days". This

! certers of the generic grcup (SQUG utilities). requirement now applies to all utilites.

4. Page 27: The middle of the page discusses that it is 2 4. This is the staff's best estimate.

estimated that the qualitication procedure will involve approxicately 110 pieces of equipeent. This number seems icw.

1 5. Page 13: The guidelines provided in Paragraph 6 for the 12 5. The staff agrees with the coment.

review and identification " outliers" was developed by The text has been modified i SSRAP, not SEP and the SQR Team. accordingly.

6. Appendix A: In general, the coments outlined above also 12 6. a. The staff disagrees with the apply to the draft Generic Letter and its enclosure which coment. The SSRAP general i

are included as Appendix A to the NRC regulatory analysis. conclusions listed on page 1-4 of In addition, the draft Generic Letter should be clarified NUREG-1030 are based on the SQUG l to indicate the folicwing: sponsered EQE report (reference 1),

and the staff position to require

l PUBLIC COMENTS ON hVREG-1030 AhD ASSOCIATED REG. ANALYSIS '(A-46) Page 41 CGMENTS ORIGINATOR NRC STAFF RESPONSE

a. The technical basis for the NRC's conclusion that anchorage review, relay functional seismic adequacy of equipment must be verified is capability check and review of documented in Reference 3, not References 1 and 2, outliers are all in turn based on the SQUG/SSRAP reports. the SSRAP recommendations.

b. The statement that the requirement for verification b. The text has been revised to of seismic adequacy is based principally on work incorporate this point.

performed by SCUG is not correct. Instead, the statement should be revised to indicate that the technical resolution of USI A-46 is based principally on works performed by SOUG.

The SQUG program results do not support the NRC requirement for verification of the seismic adequacy of equipment in operating nuclear power plants.

7. hTREG 1030 and Appendix A to the proposed generic letter 12 7. The staff agrees with the comment.

a include SSRAP conclusions, caveats or exclusions. It The text has been modified a

~

should be made clear that these are typical and subject accordingly.

to change as more data are gathered and evaluated.

8. Page 6-7 13 8. The staff disagrees with the comment. We think the text is The argument for excluding piping should be strengthened in clear.

the area of support motion-induced failures.

9. Page 11 13 9. The staff agrees with the comment.

The definition of " grade level" has

" Grade level" should be defined. been added in the text.

Originater ef Cossnents

1. Wisconsin Electric Power Company

2. Baltimore Gas and Electric Company

3. Carolina Power & Light Coc:pany

4. Tennessee Valley Authority

5. Florida Power & Light Company

6. Cocr:cnwealth Edison Company

7. Duke Power Capany

8. Public Service Electric and Gas Company

9. Atomic Industrial Form , Inc.

10. Electric Power Research Institute

11. Nuclear Utility Group on Equipment Qualification

12. Seismic Qualification Utility Group

13. Sandia National Laboratcries

?

h

g r,o mi c. =oct... ...ot. 1 > . oa r u .. . . ... ,

E'O'# essuoGRAPHIC DATA SHEET NUREG-1030 un i=st.oerio~i on t . .gviau 3 Tsftt .4. SutflTLE 3 44.v t SL.N.L S21smic Qualific ion of Equipment in Operating Nuclear eso e afety sue A-46 ,,,' #'""**",,,,

. .v1 oa<i. February! 1987 f . o.u . .oa r .uuio T. Y. Chang p~ r- 'eaa l

Feb q/ry 1987 Division of Safety Revie and Oversight Office of Nuclear Reactor egulation oa caa' av*a U. S. Nuclear Regulatory C miission Washington, D.C. 20555

.. .,o~ o..~o o.o o,4.rio. ... .~o . m. o .o .u ,, u c , ,i. 1,..o...,0.,

Technical Report Same as 7. above. . , ,oo co . . o ,,.,, ,

il Su,,ttweg .m,40ftS The margin of safety provided in ex ting nuclear power plant equipment to resist seismically induced loads and perfor their inte ded safety functions may vary considerably ,

because of significant changes in des n criter a and methods for the seismic qualification of equipment over the years. Therefor , the s ismic qualification of equipment in operatin ;

plants must be reassessed to determine hethe requalification is necessary.

The objective of USI A-46 is to establis at explicit set of guidelines and acceptance criteria to judge the seismic adequacy of- .uipment at all operating plants, in lieu of requiring qualification to the current cri cria that are applied to new plants.

This report surmiarizes the work accompli he on USI A-46 by the Nuclear Regulatory Commissi in staff and its contractors. In addition the collection and review of seismic experience data and existing seismic test data by he S G and EPRI respectively, and the review and recommendations of the SSRAP are prest ted. e principal technical finding of USI A-46 is that seismic experience data, sup emented existing seismic test data, applied in accordance with the quidelines deve ped, provi es the most reasonable alternative to current qualification criteria to ' frify the se smic adequacy of equipment in operating nuclear plants. Explicit seismic ualification iould be required on y if seismic experience data or existing test ata on similar omponents cannot be shown to apply.

i. oow .. , . ..........o o o..< ..,0..

.. .,,.,. g .,

Unlinited Unresolved tafety Issue -46 Availability

't SICust f, C4.ll ,4C.t#0%

a v. ,

. .o.= e ,. , a o . i ~o o n a** Unclassifieri

,r,,.,,

V,ric

, w ..l a . s c s.i.f

.oi i ed

.e .. ( 4

UNffED STATES NUCLEAR RESULATORY C00H400000N N T M ""

WASHINGTON, D.C. 2005 waSs c.

Peatert ses. 04F OPPICIAL sutwetSS .

PENALTY POR PfWVATE USE,4300 t