ML20085J068
| ML20085J068 | |
| Person / Time | |
|---|---|
| Issue date: | 09/30/1983 |
| From: | Chang T Office of Nuclear Reactor Regulation |
| To: | |
| References | |
| REF-GTECI-A-46, REF-GTECI-SC, TASK-A-46, TASK-OR NUREG-1018, NUDOCS 8310110345 | |
| Download: ML20085J068 (87) | |
Text
NUREG-1018 Seismic Qualification of Equipment in Operating Plants Status Report Unresolved Safety issue A-46 U.S. Nuclear Regulatory Commission Office of Nuclear Reactor Regulation T. Y. Chang
,9* "%,
N
/
$$RkbOS$5930930 1016 R pyg
i NOTICE Availability of Reference Materials Cited in NRC Publications Most documents cited in NRC publications will be availabla from one of the following sources:
- 1. The NRC Public Document Room,1717 H Street, N.W.
Washington, DC 20555
- 2. The NRC/GPO Sales Program, U.S. Nuclear Regulatory Commission, Washington, DC 20555
- 3. The National Technical information Service, Springfield, VA 22161 Although 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 and copying for a fee from the NRC Public Docu-ment Room include NRC correspondence and internal NRC memoranda; NRC 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 licensee documents and correspondence.
The following documents in the NUREG series are available for purchase from the NRC/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 Regulations, and Nuclear Regulatory Commission issuances.
' Documents available from the National Technical Information Service include NUREG series reports and technical reports prepared by other federal agencies and reports prepared by the Atomic Energy Commission, forerunner agency to the Nuclear Regulatory Commission.
Documents available from public and special technical libraries include all open literature items, such as books, journal and periodical articles, and transactions. Federal Register notices, federal and state legislation, 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.
Single copies of NRC draft reports are available free upon written request to the Division of Tech-nical Information and Document Control, U.S. Nuclear Regulatory Commission, Washington, DC 20555.
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 purchased from the originating organization or, if they are American National Standards, from the American National Standards Institute,1430 Broadway, New York, NY 10018.
GPO Printed copy price: $4.50
,,n-~.
NUREG-1018 Seismic Qualification of Equipment in Operating Plants Status Report Unrcsolved Safety Issue A-46 M:nuscript Completed: September 1983 D ta Published: September 1983 T. Y. Chang Division of Safety Technology Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Wcshington, D.C. 20555
,e %,
f t
ABSTRACT The margin of safety provided in existing nuclear power plant equipment to resist seismically induced loads and perform their intended safety functions
)
may vary considerably, because of significant changes in design criteria and methods for the seismic qualification of equipment over the years.
Therefore, the seismic qualification of equipment in operating plants should be reassessed to determine whether requalification is necessary.
The objective 'f technical studies being performed under the Task Action Plan for Unresolved Safety Issue (USI) A-46 is to establish an ex11icit set of guidelines and acceptance criteria to judge the adequacy of the Seismic qualification of equipment at all operating plants, in lieu of requiring current qualification criteria which are applied to new plants.
This report summarizes the status of work accomplished on USI A-46 by the Nuclear Regulatory Commission staff and its contractors, Idaho National Engineering Laborctory (INEL), Southwest Research Institute (SWRI),
Brookhaven National Laboratory (BNL) and Lawrence Livermore National Laboratory (LLNL) that is applicable to USI A-46.
This assessment leads to the conclusion that the use of seismic experience data for equipment qualification provides the only 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 status of continuing efforts to establish requirements for an experience data base is provided in this report.
iii
TABLE OF CONTENTS Page ABSTRACT iii 1.
INTRODUCTION 1
1.1 Background
1 1.2 Staff Plan for Resolution of the Issue 2
2.
SUMMARY
OF TECHNICAL WORK ACCOMPLISHED 6
2.1 Identification of Seismic Risk Sensitive Systems and Equipment.
6 2.1.1 Background 6
2.1.2 Summary of Task 6
2.1.3 Staff Position on Task 6
2.2 Assessment of Adequacy of Existing. Seismic Qualification 7
2.2.1 Background 7
2.2.2 Status of Work as of June 1983 7
2.2.3 Staff Conclusions 13 2.3 Development and Assessment of In-Situ Testing to Assist in Qualification of Equipment 13 2.3.1 Background 13 2.3.2 Status of Work as of June 1983 14 2.3.2.1 Summary of Contractor Report "The Use of In-Situ Procedures for Seismic Equipment Qualification in Currently Operating Plants 14 2.3.2.2 Summary of Contractor Report " Preliminary Guidance and Acceptance Criteria for Application of Combined In-Situ and Analysis Procedures in Operating Plant Equipment Qualification" 21 2.3.2.3 Summary of Contractor Report " Summary of Work Performed to Date on Qualification Cost Estimate Task" 26 2.3.3 Staff Conclusions 27 v
~
2.4 SEISMIC QUALIFICATION OF EQUIPMENT USING SEISMIC EXPERIENCE DATA BASEL 27 2.4.1 Background l.
27 2.4.2 Summary of LLNL Report " Correlation of Seismic Experience Data in Non-Nuclear Facilities with
. Seismic Equipment' Qualification'in Nuclear Plants" 31 2.4.3 Summary of EQE Report.." Pilot Program Report-Program for the Development of-an Alternative Approach to Seismic Equipment-Qualification" 51-2.4.3.1 Methods Used in the Pilot Program 51 2.4.3.2 Conclusion and NRC Staff Comments 58 2.~ 5 DEVELOPMENT OF METHODS TO GENERATE GENERIC FLOOR RESPONSE SPECTRA 66 2.5.1 Background-66 2.5.2 Summary of Work Completed 67 2.5.3 Staff Conclusion 70 3.
REFERENCES 73 4.
APPENDIX A - Related Topics Covered by the INEL Contractor's Report on In-Situ Testing A-1
-vi
1 l
I 1.
INTRODUCTION
1.1 Background
General Design Criterion (GDC) 2 of 10 CFR Part 50, Appendix A for Nuclear Power Plants states that structures, systems and components important to safety shall be designed to withstand the effects of natural phenomena, such as earthquakes, without a loss _of capability to perform their safety functions.
Appendix B,Section III of 10 CFR Part 50 also states that design control measures shall provide for verifying or checking the adequacy of design by the performance of a suitable testing program.
Furthermore, it requires that suitable qualification _ testing under the most adverse design conditions shall be included.
These requirements point to the need for seismic qualification of safety-related electrical and mechanical equipment in' order to ensure structural integrity and functional capability during and after a seismic event.
Current criteria and methods of compliance are contained in Revision 2 to Standard Review Plan Section 3.10, " Seismic and j
Dynamic Qualification of Mechanical and Electrical Equipment," and Regulatory Guide 1.100, " Seismic Qualification of Electric Equipment for Nuclear Power l~
Plants," which, with some exceptions, basically endorses IEEE Standard 344-1975, "IEEE Recommended Practices for Seismic Qualification of Class 1E Equipment for Nuclear Power Generating Stations."
Based on the requirements and recommendations from these current 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 assured by its structural integrity.
Otherwise, some testing is required.. Seismic input motion to equipment is specified by required response spectra or by time histories.
When the test method is utilized, the equipment is mounted on a shake table and subjected to certain types of excitation corresponding-to a test response spectrum which envelopes the required response spectra.
The equipment should be tested in the operating condition.
For equipment too large to fit on a shake table, a combined analysis and test procedure is adopted.
i Since commercial nuclear power plants were first introduced, significant l
changes in_ seismic qualification criteria have occurred.
The analytical and l
experimental methods used to qualify equipment have also changed.
The margins of safety provided in existing nuclear power plant equipment to l
resist seismically induced loads and perform their intended safety functions may vary considerably, and may not meet current seismic qualification criteria. Therefore, there is a recognized need to reassess the seismic qualification of equipment in operating plants to ensure its performance during and after a seismic event.
It was also recognized that it may not be practical to cualify operating plant equipment using current seismic qualification criteria and methods due l
l i -_,
~
l to excessive _ plant down tire, difficulties in shipping irradiated equipment to a test laboratory.and in acquiring identical old vintage equipment for laboratory testing.
In December 1980, the Nuclear Regulatory Commission
< designated " Seismic Qualification of Equipment in Operating Plants" as Unresolved Safety Issue (USI) A-46.
The objective of USI A-46 is to develop-alternative seismic qualification methods and acceptance criteria that 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 Staff Plan for Resolution of the Issue A Task Action Plan (TAP) was developed for USI A-46 in the Spring of 1981.
Tasks selected 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 re qualify equipment using procedures required for new plants.
Only i
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 requireaents.
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 qualification of equipment; i
(4) Seismic q~ualification 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 approach to develop a qualification method 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 =(Ref. 1) which concluded that use of seismic
. experience data is feasible and can be as effective as current qualification methods. 'This study is discussed in more detail later in Section 2.4.2.
In j
addition, a utilities group, Seismic Qualification Utility Group (SQUG) conducted a pilot program to independently demonstrate the feasibility of using seismic experience data.
Their report (Ref. 2) published in September
-1982, was reviewed by the staff, and preliminary-requirements for an acceptable data base were transmitted to the SQUG.
A more detailed discussion'of this effort is presented in Section 2.4.3. L
In March 1983, the SQUG proposed to the NRC management the formation of Senior Seism'c Review Advisory Panel (SSRAP) to provide consulting services and expert opinion on the use of experience data. This idea was endorsed by the NRC management and the SSRAP was subsequently formed in June 1983. The staff is continuing to work closely with the SQUG and the SSRAP to develop an acceptable approach to using seismic experience data.
The SSRAP has completed the review of the SQUG pilot program report and has made the following preliminary recommendations and conclusions:
-(1) SSRAP endorsed the SQUG pilot program in general, and agrees with the program that the SQUG activity should be limited to the seven classes of equipment (see Table 2.4-6).
(2) The goal of the SSRAP review, if possible, will be to establish a set of screening criteria for the seven classes of equipment. The intent is to avoid piece by piece comparison of equipment in the data base with equipment-in the operating nuclear plants.
No further seismic qualification of equipment would 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, the SSRAP believes that a significant amount of data will be needed for each of the seven classes of equipment.
(3) Similarity and operability of e juipment are the two most important issues to be resolved in developing the screening criteria.
(4) Adequate equipment anchorage should be established prior to the screening of equipment.
The NRC staff is in general agreement with the SSRAP preliminary recommendations and conclusions outlined above. The requirements on equipment anchorage will be addressed separately by the staff in USI A-46.
A possible approach could be similar to that used by the Systematic Evaluation Program (SEP) phase II.
Equipment not in the seven classes or which can be eliminated from formal qualification by applying the screening criteria will be qualified using the more rigorous approach outlined in Figure 1.2.1.
Tasks 3 and 5, " Development and Assessment of In-situ Testing Methods to Assist in Qualification of Equipment" and " Development of Methods to Generate Generic Floor Response Spectra," play a strong supporting role.
The emphasis on both tasks was focused to support use of an experience data base..
^
/
Task 2 " Assessment of the Adequacy of Existing Seismic Qualifi:stion," 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 terms o,f expected multiple axis response.
A procedure was developed by Southwest Research Institute, but is of limited practical 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 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 Brookhsven National Laboratory'(BNL), was conducted on a hybrid model of a PWR plant and a hybrid model of a 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 into the risk importance of systems and components and demonstrated the effect of varying equipment fragility on overall risk, it is of limited usefulness in defining an equipment-list at cther plants.
The major conclusion of BNL was that they had demonstrated a methodology that could be applied on a plant specific basis to develop a risk based minimum equipment list.
For plants where 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 risk contribution.
This task is described in more detail in Section 2.1.
At the conclusion of the USI A-46 study, the proposed implementation requirements will be reviewed by the NRC Committee to Review Generic Requirements (CRGR) and will be issued for public comment prior to issuing final requirements.
The NRC intents to implement the final requirements in the form of a generic letter.
A rulemaking is not anticipated. __
De uc/op R e s ( In.T;/u fegas la#caft ~,
Tes//Ana1 sis or Nof N'2 '" d 7
Generic Spe c 'tr~a )
In Si/a ks' f%3aalfy Yo Tes/
Grnet Reyuirene t
(
RRS b pa-fa y
Sim/< No///?cafin,,
a 8ase Specfra E fr0VIWC In S eyar<irey
.g,;,,,,og, ;,,, =
o,c g#'"/'^ re E u,p, aa,,1
,p,,
c 9
fa,/e),,e,f Dag 3g be ve/op E_gsqmen((~) k'oYf.Dafa 8ase iIs f A l Ejuryment 7;;-
EJfn 44 'sf No, teeplacemmt
^
r
.C,witars.q No ey aua//A ed' _-
E uspmeist S
A'eguah/;enfbi, Complele ofAe,. MefAods, Da/a E.G., Compansag Basc ol' o/de, auatiwcan Me/Aod an/A Curres,r-R<g er it e then-f S ALTERNATWE SEISMIC CUALIFICATION PROCEDURE _.FOR USE WITli U_SI A-46 RESULT FIGURE 1.2.1 I
2.
SUMMARY
OF TECHNICAL WORK ACCOMPLISHED In the remaining sections of this report, each of the tasks are described.
A summary of work done and major conclusions are 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' viewpoint and recommendations.
The final staff positions on A-46 will be developed following completion of Task 4.
2.1 Identification of Seismic Risk Sensitive Systems and Equipment 2.1.1 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 then equipment could be' ordered by the contribution to risk.
Equipment whose failure resulted in a small change in risk co11d then be culled from the qualification list.
2.1.2 Summary of Task Bookhaven National Laboratory (BNL) under contract to the NRC conducted a study (Ref. 3)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 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 davelop 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 equipaent list.
2.1,3 Staff Position on Task For plants with existing seismic PRA studies, the staff believes it may be possible in some 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 1
6~-
developed by BNL, the staff would consider it.to be-an acceptable method subject to the analysis assumptions and inherent uncertainties.
The staff was unsuccessful in developing a generic minimum equipment list.
2.2 Assessment of Adequacy of Existing Seismic Qualification 2.2.1
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 " Correlation of Methodologies for Seismic Qualification Tests of Nuclear Power Equipment" (Ref. 4).
Some examples demonstrating the application of this approach are included in that report.
2.2.2 Status of Work as of June 1983 The concept of vibrational equivalence is a key factor in development of the correlation of methodologies for seismic qualification of equipment.
Vibrational equivalence forms the basis for a damage comparison between two different motions.
In the qualification of nuclear power plant equipment, a great variety of physical failure mechanisms may occur.
Therefore, the concept 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 t'ne 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 2.2-1.
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 an 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 exists at that point in the equipment._ This arbitrary failure mechanism is dependent on the response amplitude at location y, and nay also be 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 F This function representsasurface,anypointohw(f,t)isgenerated.
hich corresponds to failure of the equiptent.
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 2.2-1 for excitation conditions 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 equipment at specified excitation levels is required for qualification.
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 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 2.2-2 as a fragility surface.
This surface can be represented as a function F f, t) = M fragilityIdr(face,canbe('f,t),whereMintermsofth$a(mp,litudeoftheexcitation, f t), measured at the the response spectrum, 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 conservatively 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 M (f, t)
D fr Mf (f,t)
- 1
=
where M(f,t) is the value of the actual excitation function and Mf (f,t) is the value of the fragility function at the same conditions of frequency and time. This is shown in Figure 2.2-3.
A damage fragility
. equivalence similar to that described in Figure 2.2-1 can then be stated as:
2.d M(f,t)
M (f,t)
~
M (f,t )
f 2 2 f
This is the' general basis for comparing various test motions.
The report then proceeded to define simple systems and complex systems.
A simple system is one whose fragility function is influenced by a single resonance, and therefore can be generated by a slowly swept sine or narrow band random excitation.
A complex system is one where
- Y(f,:)
location Failure F
Locatien H,y( f) x y
Meenanism Failure
~
Amclitude 1 fn Excitation Level 1
Response
Freq. 1 1
Level 1 F,y(f).t) 3 l
I 4
Time Duration 1
~
~'
h __
- h I
Both Peints
_g Scecimen Constitute Failure Transfer Func: ion I
g____q i
Failure I
1 1
~
~
Amplitude 2 Excitatica l
l
[#*9IIiU runctica Level 2
.J 4
Respense Level 2 F,y(f '*2) 2 Freq. 2 l
Time l
~'
~
Duration 2 j
Figure 2.2_1 Conceptual Approach to Vibration Correlation l
_g-
several failure modes can occur as the result of multiaxis and/or multimode response, and interaction between responses is included.
Due to 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
.QuadratiE 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 system.
In developing a fragility surface for existing qualification data, it was-recommended by the contractor that a correction factor, generated from resonance search data, be used to modify the level of qualification excitation in order to develop an approximate lower bound fragility function.-
The next step is to establish a correlation between the approximate fragi_lity function (namely existing qualification information) and the qualification corresponding to a different set of criteria, e.g.,
current criteria.
The possible pairs of comparison are listed in Figure 2.2-4.
In a specific application, some judgement 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.
In summary, the results of a previous qualifications are used first to establish some form of an approximate or acceptable fragility function.
Then, the new criteria are compared to'this acceptable fragility function to determine whether a greater 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 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 contractor suggested that it may be more practical to consider a corrplete new requalification.
It was also surmised by the contractor 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 significantly more potential damage than do typical random motion simulations that have been more generally used after 1975.
I Magnituce i
Actual
/s
- Surface s'
s ~
s' - tiiin HW.
s f; :u l
l I
i Acceptable l
Sur' ace l
/
l I,
4
\\
/
d s
M N
/
gd Figure 2.2-2 Comparison of Actual with Acceptable Fragility Surface M (f,t) p
. 9
..:'.?g::
~
c.
49 I
Nih'$f!'
.%( f),t))
n.
5 # !!s, 1
M(f),t))
...s..
i Mii' y,
\\
l'
's
!i
/
N t
\\
t'2' g (f,t )
2 2 M(f *t )
2 2 Figure 2.2-3 Basis for Da= age Fragility Ratio 1
7: agility Furetion Qualifica-J.on
?araneters Paranaters Single Axis Single Axis 1
Nanow Sand Narrow Band 2
=
Excitation Excitation Single Axis Single Axis 3
3 road Band Broad Band 4
Excitation Excitation Multi-Axes Multi-Axes 5
Na=cv Band Narrow Band 6
Excitation Exc1'tation Multi-Axes Multi-Axes 7.
Broad Band Broad Band 8
Excitation Excitation Includes sinusoidal excitation rigure 2.2-4 Possible Cembinations of 7: agility 7 unction and Qualification Parameters l - _ - -
2.2.3 Staff Conclusions.
The' technical: basis and'generalLmethodology to correlate seismic qualification' tests have been developed and demonstrated.
From what is
}
1shown in the contractor.'s report, this is a -promising methodology.
However,;in order for nuclear plant owners to be able to. apply this methodology,more specific guidelines and acceptance criteria need to be generated.
Simpler guidelines and criteria are to be developed by SWRI s
by August 1983.
i f 2. 3 Development and Assessment of In-Situ Testing Methods to Assist In Qualification of Equipment 4
2.3.1.. Background This ' task:was selected.for A-46,: due to the potential that in-situ testing can be a promising tool in. assisting the seismic qualification of equipment in operating plants.
The task is conducted by Idaho-National Engineering Laboratory (INEL), and was started in early 1982.
The intent of this task is to investigate present in-situ testing methods and to evaluate the feasibility of using these-methods to assist in're qualifying equipment, and to develop methods, guidelines and acceptance criteria for their use.
More spe'cifically, the work scope for this task consisted of the following topics:
(1). Basic review of existing approaches to in-situ testing and identification of preliminary in-situ test methods forz the j
qualification of equipment.in plants which are currently licensed and operating.
(2) Review of approaches to. laboratory testing and simulation.of.
seismic events in the laboratory for qualification of equipment.
Limitations on the use of current guidance was also studied.
_(3) Review of the analysis procedures fundamental to in-situ testing i
methods.
Review of use of subcomponent proof test and/or subcomponent' fragility-tests in the qualification process.
Review c
f7
.of the qualification requirements for anchors.
-(4). Investigate.techniq'ues for assessing / monitoring the effects of chemical or. metallurgic aging, mechanical fatigue, and wear during plant operation.
(5) Address adequacy, limitations and inherent shortcomings, and nonconservatisms of the various approaches above.
T 4
d,
. ~,
(6)- Development of guidelines and acceptance criteria for use of in-situ testing to support alternative methods of-seismic qualification of safety'related equipment.
(7) Define requirements'for a test data base in support of seismic qualification of existing equipment in currently licensed operating plants.
(8) Develop cost estimate.for alternate siesmic qualification methods.
(9) : Verification and.further development of combined in situ and analysis methods suitable for equipment qualification.
Examine limitations and pitfalls of applying in-situ testing methods in determining dynamic characteristics and evaluating component mountings of structures which support, contain, or position safety-related equipment in operating plants.
Develop guidelines for minimum testing requirements and reporting requirements in qualification documentation.
2.3.2 Status of Work as of June 1983 Results of work on topics 1, 2, 3, 4, and 5 of Section 2.3.1 above are covered in an interim contractor report titled "The Use of In-Situ Procedures for Seismic Equipment Qualification in Currently Operating Plants (Ref. 5)" issued in December 1982.
Another contractor progress report titled, " Preliminary Guidance and Acceptance Criteria for Application of Combined In-Situ and Analysis Procedures in Operating Plant Equipment Qualification" (Ref. 6) and issued in April 1983, provides the preliminary results on topics 6 and 9.
Finally, a separate draft contractor report titled, " Summary of Work Performed to Date on Qualification Cost Estimate Task" (Ref. 7) covers topic 8 and was issued also in April 1983.
All three reports have been sent to the staff for review and comments.
Following is a summary of these tasks.
2.3.2.1 Summary of Contractor Report "The Use of In-Situ Procedures for Seismic Equipment Qualification in Currently Operating Plants" 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 manners:
(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, only 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 conditions.
The required conditions are that:
(1) The motion applied to the equipment supporting structure should not excessively load the appurtenances, the components mounted tnereon or in the vicinity and the equipment supportir.g structure itself.
(2) Sufficient access must exist in crder 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, such that component can be employed in service for its nominal useful lifetime.
These conditions severely limit the usefulness of full load level in-situ tests.
Valve operators are one equipment type that have been dynamically qualified 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 changes significant to operability of safety related equipment (particularly in a seismic environment) cannot 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 was recorded..
The contractor concludes from his study that in-situ testing will be useful in the following area related to equipment qualification:
~ establishment ~of similarity between equipment with consideration of failure modes prediction of component specific Required Response Spectra (RRS) component mounting evaluation 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:
to establish component / equipment seismic capacity to support a continuing surveilliance 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 method to seismic qualification of equipment in operating plants probably is to establish dynamic similarity between equipment.
As mentioned in Section 1.2, after reviewing the results to date of all the tasks of A-46, it was concluded by the NRC staff that seismic qualification using seismic experience data probably is the most likely approach to develop a qualification 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 to help qualifying equipment in operating plant.
They are:
(a) 'To establish that RRS of equipment in operating plant to be requalified 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 imme~diately following and also by Section 2.5.
Condition (b), the question of similarity between equipment, has been touched upon by No. 4 of the staff comments to SQUG Pilot Program Report (see Section 2.4).
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 an equipment consists 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 report).
A characterization of viscous damping is also possible by using in-situ tests which 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 (Ref. 8) 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 necessary to establish the specific RRS.
For equipment mounted on a floor, the response can be predicted by the floor response spectra.
However, numerous safety-related. components are mounted on or attached to the equipment supporting structures (such as electrical cabinets, racks, etc.), the RRS for these components will thus 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, (1) a synthetic time history is developed from a specific floor response, (2) the modes, frequencies, and modal participation factors are calculated from the model, (3) a time history analysis is performed on each significant mode, (4) the modes are algebraicly combined to determine total time histories, and (5) the time histories are converted to RRS for the "t
components of interest.
It is felt by the contractor that this basic procedure is potentially unreliable because of system complexity and unreliability of boundary condition modeling.
Consequently, it can only be used if the equipment is already installed and in-situ procedures are used to verify the calculated modal parameters.
A major disadvantage of the approach is that it is relatively costly 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 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 procedures directly in constructing a numerical solution.
There is no need to develop a finite element model.
As with the finite element approach, the response of' individual modes is calculated and then reperimposed for the total response.
The_ contractor offered several comments on using this method.
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 factors.
Consequently accurate calculation of only the lower mode shapes will probably be necessary.
The situation must be checked for every individual case.
The second comment concerns closely. spaced modes.
The decomposition of the total frequency response into a modal frequency response function is one step in the development of the mode shapes.
Closely spaced mode shapes reduce the accuracy with 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 infrequently in which case the alternative of method "a" above can be used to determine RRS.
A final comment is that 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 method will require rubstantially more cost.
Consequently, analysis procedures wi.ich use experimentally determined modal parameters are recommended to be the prime candidate for predicting RRS in operating plants.
i (c) The third method involves direct response spectra transfer.
Procedures (a) and (b) discussed above employed variations of time history analysis where a synthetic time history is used to define the load.
Using these procedures an input response spectra can be transferred to an output location yielding an output response spectra.
Since the input is initially specified by a response spectra, the use of time history analysis in transfering response spectra is essentially artificial and the output response spectra is not uniquely defined by the input spectra. Methods for transferring the input response spectra in a unique, more meaningful, and less costly way are preferrable.
Direct methods for response spectra transfer have been looked at by various investigators.
A direct method uses the input or floor response spectra in combination with the modal parameters and modal participation factor to determine output response spectra.
The associated e
analytic procedures are algebraic.
Any direct method will eliminate the time history analysis portions of the transfer process.
In addition, by using mode shapes, frequencies and transfer functions determined frem in-situ procedures the need for a finite element model can be eliminated, yielding a very cost effective method.
However, one distinct mode of dispersion, i.e., the feature by which the transferred response spectra become non-unique, seems to exist.
This happens where the spectral frequency is near one of the structure frequencies, i.e., tuned conditions.
The acceptance of a method for direct response spectra transfer, in the contractor's opinion, awaits a firm resolution to predicting response at tuned conditions.
A final contractor's position on this matter is expected in August 1983 and may possibly be included in the final guideline for A-46.
(3) Compontat 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 straight forward process to assess existing designs.
The main distraction is the large number of mountings that exist.
Enveloping the maximum acceleration could be an approach to reducing this workload.
Examining mountings on a theoretical basis may not address some (perhaps the major) problems.
It is pointed out by the contractor that quality of installation or use of problem prone designs may be a stronger influence on mounting adequacy than strength considerations.
To address these concerns, the contractor suggests a physical mounting review by practitioners experienced in seismic qualification testing as well as 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 walk-down would then screen mountings for those requiring in-depth review or retrofit.
The effectiveness of this process is that it screens out items which are clearly adequate and concentrates more costly review on questionable items.
(4) Compariscn of fundamental building frequencies with equipment supporting structure frequencies.
The level of equipment supporting structure response during a seis;nic 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 hertz 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 hertz 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 complicating 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, equipment design environment will depend heavily on the relationship between the equipment supporting structure and building 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).
1 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 where all equipment supporting structure natural frequencies below 15 hertz are experimentally determined.
Mode shape determination is not required.
A codal analysis crew should be able to check a number of cabinets ir 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.
2.3.2.2 Summary of Contractor Report " Preliminary Guidance and Acceptance Criteria for Application of Combined In-Situ and Analysis Procedures in Operating Plant Equipment Qualification" This preliminary contractor report covers the progress to date on Topics 6 and 9 defined in Section 2.3.1 of this report.
Twelve technical areas are identified by the contractor that require guidance and acceptance criteria. The guidance in in-situ testing procedures and pitfalls will be addressed in a final version of this document, presently scheduled to be published in August 1983.
Following is a summary of the preliminary guidance and acceptance criteria in the twelve technical areas. The contractor has been requested to continue his investigation and provide more definitive guidance by August 1983.
(1) Dynamic Parameters from Tests.
Guidance is required on the number and position of nodal points for mode shape description.
Node points are to be located at all significant masses, and there should be no less than four node points between local maximums and minimums of all significant modes.
(2) Analytically Determined Dynamic Parameters.
Guidance relating to analytically determined equipment supporting structure models is that these models are to be verified by comparing computed and experimentally determined 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.
T'ie time history analysis methe ' is currently accepted and the same guidance can be applied to operating plant application.
There are no currently accepted direct transfer methods.
Stochastically/ statistically based methods look very promising.
These methods will be reviewed to determine if their use can be justified.
Providing that the methods are justifiable, corresponding acceptance criteria will be developed.
(4) Functional Similarity and Functional Requirements.
Experience data must identify the functional requirehtents tested during the real seismic events.
Claims that these functional requirements were tested must be supported by documentation of the methods used to establish these functional requirements.
Acceptance criteria in these areas will be developed in the follow-on effort and available in August 1983.
(5) Experience Data Floor Response Spectra (FRS).
Currently accepted guidance is to estimate floor spectra by analysis followed by peak broadening (Regulatory Guide 1.122, Ref. 9).
This procedure is applicable to experience data minus the peak broadening requirement.
Guidance on the use of the ground response spectra (or some fraction of it) to provide a conservative estimate of the FRS has not been developed.
Specific guidance and acceptance criteria will be provided in the follow-on effort.
The criteria will be directed at equipment near the ground level.
See Figure 2.3-1.
(6) Damping. Guidance relating to damping is that experience data for equipment in supporting structures be estimated based on a range of damping ratio values (2%, 4%, 6%, and 10%).
Guidance for correlating operating nuclear power plant equipment supporting structures damping with experience data damping has not been developed yet.
Guidance will be forthcoming in the follow-on effort.
(7) 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:
$9 [M][1] = MPFj Additional guidance will be developed allowing the calculation of MPFs from dynamic parameters determined from in-situ testing.
l _ _
Wj = Fundamental building frequency Floor response spectra I
la I
Ground response spectra en j
i I
.u is j
Region in which the floor response spectra may not I
envelope the ground response l
spectra 1
- )(
I N
1 Frequency - hz Figure 2.3-1 Comparison of floor and ground response spectra.
4 l
r (8) Fundamental Frequency Determination.
The lowest equipment supporting structure frequency is acceptable if the transfer function in the low frequency range is determined from data maintaining a coherence of 0.8 or greater.
(9) Margin.
Additional margin against uncertainties in equipment supporting structure frequencies is not required if peak broadening (Regulatory Guide 1.122) is accounted for in predicting RRS (i.e.,
if time histories used for analysis are consistent with peak broadened FRS).
The RRS can also be predicted using a floor time history without peak broadsning.
In this case, two analyses are performed.
In the first anal sis, a RRS is calculated using the modal parameters from 3
in-situ terting.
In the second analysis any natural frequencies in the vicinity of floor peaks are modified by 15% in the direction of the peaks.
The remaining modal parameters are unchanged and another RRS
.'s calculated.
A final composite RRS is developed by enveloping t :e two response spectra.
(10) Equipment Supporting Structure Linearity.
Equipment supporting structure attached to the floor using bolt attachments must be secured such that installation preloads are not reduced by greater than 90% during the SSE environment.
(11) Enveloping Criteria.
As with current criteria, the Test Response Spectra (TRS) for rigid equipment must envelop the RRS at the Zero Period Acceleration (ZPA).
Envelopment at lower frequencies is not essential.
For equipment supporting structures, envelopment is required only at frequencies greater than the equipment supporting structure fundamental frequency (with 15% margin).
See Figure 2.3-2.
If justification can be provided that equipment is not specifically sensitive 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), then envelopment can be restricted to the remaining frequency range.
(12) 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 methods from Regulatory Guide 1.92.
The mass is taken as the sum of the component and mounting fixture masses.
Assurance must be provided that at least 80% of the component will move as a rigid body during the dynamic loads.
fundamental frequency W
pg = of NPP building WIN = NPP equipment supporting structure fundamental frequency experience data equipment supporting l
W ID =tructure fundamental s
frequency 8
I l
l i
l I
I i
I i
j l
I i
l i
NRB ID IN
[
l
(
Frequency - bz Figure 2.3-2 Comparison of Envelopment.
l -
2.3.2.3 Summary of Contractor Report " Summary of Work Performed to Date on Qualification Cost Estimate 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 2.3-3.
A table of estimated costs is given in the contractor report.
These costs will be used in development a regulatory analysis to support proposed requirements developed by the staff.
Assumptions used to develop the cost estimates are described below.
Eguipment List The equipment list was obtained by modifying the list offered in the report " Survey of Methods for Seismic Qualification on Nuclear Plant Equipment Components (Ref. 10)."
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 estiinates were based on experience in estimating analysis jobs and on reviews of such analyses performed during staff audits of new plants during 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. "In-Situ Structural Characterization Test Cost Estimates." Cost of labor, travel of personnel, 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 based on " Process Plant Construction Estimating Standards," by Richardson Engineering Services, Inc.
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 & CRDM) were not covered by the standard.
Estimates for these two were obtained by contact with equipment vendors.
Qualification documentation was assumed to cost 150% of the cost of the unqualified components for all but three of the components -- small instrument valves, transducers, and relays.
These components are s
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 is readily available.
Therefore, no costs resulting from analysis or in-situ testing have been included.
The tables in the contractor's report are not reproduced here since they are of preliminary nature.
The final version of the estimation will be included in the regulatory analysis supporting staff implementation requirements.
2.3.3 Staff Conclusions As mentioned in Section 1.2 of this report, as work progressed it became increasingly apparent that Task 4, " Seismic Qualification of Equipment Using Seismic Experience Data" was the most likely approach to develop a qualification method which is both economically attractive to the plant owners and acceptable from a public safety viewpoint.
The application of experience data to qualify equipment in operating plants involves the confirmation of two items.
First, the experience response spectra should envelope the nuclear equipment RRS for frequencies within the range of interest; second, similarity between the equipment in the nuclear plant and equipment in the experience data base has to be established.
The study conducted by the contractor indicated that in-situ testing can assist to provide information for both items.
The staff agrees with this conclusion arrived by the contractor.
- However, the staff believes that more specific guidelines and acceptance criteria for conducting the in-situ testing (such as in the areas of test setup, data acquisition and data processing) are needed in order to achieve meaningful and valid information from the in-situ testing.
INEL is presently working on this topic and results are expected in September 1983.
2.4 Seismic Qualification of Equipment Using Seismic Experience Data Base 2.4.1
Background
It is well known that numerous non-nuclear power plants and industrial facilities containing equipment similar to those in nuclear power plants underwent major earthquakes in various parts of the world.
It is also recognized that during the course of qualifying safety related equipment for licensing nucleaf plants in the last decade or so, numerous equipment Dcve/op RRs C zn
.rir'u
"" I'
~
Ter//Anal7 sis or Nof Repa re d Generic Specfra ) y In Silu
=
x 3f N Te f%9 un ///y' Ya Clieck i,,,,,,e. gap;,,,,:y Enve/op, mad or RRS B Da-fet y
Sim,/e MeliRes/>'m 8ase 3 ecfr; s
7 E I'!OVIlC In /~r ey e>>cy 00*/aee Eyaa).,e,,1 g,,,, op
,y,3,,i/aef/y /Va*f/s w
wy,,
Efayme,,f ( ) k'sYb.bant 8a1<
In/*res/
Dafu Bose K ) g3,,y,,4
( N*l' 2 )
Lrst n
No ireplacernent a
[
- ""##W No
_fh Br Q<<a/<'A*ed
~U E uipines# W 3
Se2uaNOctifon coiny/ ele DOlQ Offet- /Wef/ rods, Base rs<,te t)
E.G., Compasv~ son __.
of orden auadacorrav, N*k % C os-f e.sf,,,,s.,,, y, _ m h
- g w itj, cu,,,N
/rjeyA,
,,e 2
MS ( N N 1
Cast e s s,4.,,
y, seynj,,cy., p yo., 4,q 2 Coat e rna a no,,
is e. + mo rte b e ce,use if is n ey/,yUe 3.
Cost e rAwaA ha, is s,a-r m ade de cause dis ca..M c-oce a Jery,',p _
y
- 4. Ces f g sh%ahhn is m n e/c-Asr m ofysis w>,ty FIGURE 2.3-3 ALTERNATIVE SEISMIC OllALIFICATION PROCGXIIE FOR USE WITil USI A-46 REStA.T
TABLE 2.4-1 SEISMIC CUALIFICATION tfTILITY GROUP MEMBEP.S BALTIMORE GAS & ELECTRIC CCMPANY 30STCN EDISCN CCMPANY CCmCNWEALTa EDISCN CCMPANY CCNSOLIDArs EDISCN CCMPANY CcNSLP.ERS PCWER COMPANY Deir:0IT EDIScN COMPANY DUKE PCbER CCPPANY FLcRIDA PCWER CORFCRATICN W NUCLEAR CCRFORATICN NE3RASKA PusLIC POWER DISTRICT NORTHERN STATES PCWER CCMPANY PENNSYLVANIA POWER & LIGHT COMPANY Roc-: ESTER GAS & ELECTRIC CCMPANY SOUTH.=_R CALIFORNIA EDISON CCPPANY TcLEDo EDISCN CCMPANY YANKEE ATmIc ELECTRIC CCPPANY __
i items were tested for seismic capability on shake tables in laboratories.
Therefore, there is a wealth of information regarding seismic experience that potentially can be' utilized as an alternative to formal qualification of equipment in operating plants.
To use this information the data must be collected and organized and guidelines and criteria developed.
Two independent efforts to develop a seismic experience data base were initiated.
The SQUG (Table 2.4-1) conducted a pilot program, " Program for Development of an Alternative Approach to Seismic Equipment Qualification," The pilot program has been completed by their contractor, EQE Incorporated.
Results of this pilot program were recorded in a two-volume' report issued in September 1982 (Ref. 2).
The second effort was one initiated by the NRC staff, with Lawrence Livermore National Laboratory (LLNL) as the contractor.
A draft report
" Correlation of Seismic Experience Data in Non-nuclear Facilities with Seismic Equipment Qualification Nuclear Plants" (Ref. 1) was published in November 1982.
The results of both studies confirmed the feasibility of utilizing non-nuclear seismic experience data to qualify equipment in operating nuclear power plants.
Staff comments on the pilot program report were principally an assessment of what further data collection efforts were needed and suggested guide-lines for acceptability of an experience data base.
The staff's assessment is that use of experience data provides the only viable alternative to current qualification criteria.
Several of the other A-46 tasks will directly support the use of an experience data base.
As mentioned in Section 1.2, SSRAP was formed by the SQUG in June 1983 to provide consulting services and expert opinion on the use of experience data.
The staff is continuing to work closely with the SQUG and the SSRAP to develop an acceptable approach to using seismic experience data.
In the coming sections the two studies mentioned above will be described first, followed by staff comments on the SQUG pilot program.
u 2.4.2 Summary of LLNL Report " Correlation of Seismic Experience Data in Non-nuclear Facilities with Seismic Equipment Qualification in Nuclear Plants" The study was completed by LLNL and a report issued in November 1982.
This study was intended to answer the question:
Is it feasible to use experience data on the performance of equipment in non-nuclear facilities during earthquakes in addressing issues concerning the seismic qualification of equipment in operating nuclear power plants located in the eastern United States?
The study shows that the answer to this posed question is affirmative.
LLNL's general approach to the feasibility determination is based on the conservative assumption that if experience data can be shown to be equivalent to current seismic equipment qualification requirements, then it is feasible to use experience data.
The basic approach was to develop an overall summary statement evaluating seismic experience data and current requirements, as embodied in twelve different NRC Standard Review Plan sections, Regulatory Guides and national standards.
A comparison of the two summary statements provides the basis for the feasibility determination.
In LLNL's approach, thirty categories (issues) of possible seismic equipment qualification requirements are identified.
That is, seismic equipment qualification standards might be (but presently are not) formulated in terms of requirements and criteria that addresses each of the thirty issues.
Each of the thirty issues was ranked and a minimum set identified.
Table 2.4-2 lists the thirty issues and a brief description of each issue.
The twelve " current requirements" documents which are considered most important in terms of seismic equipment qualification for new plants are listed in Table 2.4-3.
LLNL's evaluation was performed by first reviewing the twelve current requirements in each of the thirty categories in Table 2.4-2, followed by an overall evaluation of these requirements.
The evaluation was performed by ranking the current requirements in the thirty categories using the following numerical weights:
- 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 should not be interpreted as " ideally" or " perfectly" or that it " addresses the issue as perfectly as can be conceived.".
i Table 2.4-2 Categories of Possible Seismic EQ Requirements Category of p'ossible seismic E0 requirement Brief description of category Physt. cal attributes 1.
Sampling For equipment items qualified by testing, only.a limited number of ti e items installed in a plant is tested.
2.
Simila rity The EQ for one item of equipment is sometimes-extended to similar but different 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 piping, and so forth are often attached to the major item of equipment.
5.
Dummy components Equipment is sometimes qualified by testing with a durny 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 particular category of equipment) are sometimes defined.
- 7. Enveloping load assumption It is often assumed that if an item of equipment is qualified for load L1, then it is also qualified for load L, where L1 is greater than L -
2 2
- 8. Required design load Here the question is whether the required design load and parameters adequately reflect EQ issues and Concerns..
Table 2.4-2 (Cont. ).
Categories of Possible Seismic EQ Requirements
- 9. Margin Here the question is whether there is sufficient margin in the capacity of the equipment.
- 10. Tolerances Here the question is whether tolerances are specified for the required l
qualification load.
- 11. Single vs. multi-axis testing Here the question is what number of l
independent test excitation axes are requi red.
- 12. Wave form A number of issues are related to the f
wavefom of the test motion imparted to equipment.
1
- 13. Fatigue-The fatigue requirements are considered I
here. An example is 5 OBE plus 1 SSE.
i Strength / capacity
- 14. Fragility This category _ addresses whether or not the EQ requirements address ~ the strength. of equipment, and if so, how.
i l
- 15. Failures This category addresses failures that occur during qualification testing.
- 16. Functional requirements This category addresses the functional Werfomance of the equipment before, during, and after qualification testing.
- 17. Critical parameters This category addresses the parameters that are most important to the surviyability or functionality of equipment.
- 18. Degradation under test Here the question is whether the qualification testing has been so j
severe that the capacity of the equipment to perform as required in the l
future can be questioned.
- 19. Response This category addresses the observed response of the equipment during qualification testing.
l l
t Table ' 2.4 2. (Cont. ).
Categories of Possible seismic EQ Requirements
- 20.
Unexpected 'resul ts (Jnexpected results include failures at unexpectedly low levels, un ual response patterns, and behavior that is inconsistent with predictions.
Seismic and other. loads
- 21.. Load combinationj This category. relates to ' appropriate combinations of loads such as' seismic, thenna1, and pressure.
- 22.
Load sequencing Load sequencing is a variant of load combination.
Miscell aneous 2 3.-
Errors This category includes design, qualification, construction, mounting, and maintenance errors.
- 24. : Maintenance lhis category includes consideration of how n6nnal' (rather than erroneous) maintenance might affect.the
' qualification. status of equipment.
c
- 25. - Mounting adequacy '
This category addresses the adequacy of
~
' ~ ~
.the equipment mounting.
26.
Post Earthquake lhis c,ategory addresses the issue of assessing EQ subsequent to an earthquake.
= 2 7.
Value/ impact This category addresses the benefit of seismic EQ in risk reduction (value) versus the cost of such requirements (impact).
E
- 28. EQ by. analysis This category addresses the issue of performing EQ by analysis rather' than testing.
29.
EQ by. testing and analysis This category addresses the-issue of performing EQ by a combination of.
testing and analysis.
- 30. -In-situ testing This category addresses the issue of the possible role of in-situ testing in EQ.
l l
TABLE 2.4-3 Documents Most Important for Seismic Equipment Qualification U.S. Nuclear Regulatory Commission, Standard Review Plan, Section o
3.10, " Seismic and Dynamic Qualification of Mechanical and Electrical Equipment,"'NUREG-0800, Rev. 2, July 1981.
U.S. Nuclear Regulatory Commission, Regulatory Guides:-
o 1.40
" Qualification Tests of Continuous-Duty Motors Installed o
Inside the Containment of Water-Cooled Nuclear Power Plants,"
March 16,1973.
j o
1.73
" Qualification Tests of Electric Valve Operators Installed Inside the Containment of Nuclear Power Plants,"
January 1974.
o 1.100
" Seismic Qualification of Electric Equipment for
. Nuclear Power Plants," Rev.1, August 1977.
1.148
" Functional Specification for Active Valve Asse'mblies o
in Systems Important to Safety in Nuclear Power Plants," March 1981.
b
-IEEE Standard for Type Tests of Continuous Duty Class 1E Motors for o
Nuclear Power Generating Stations, ANSI N41.9-1976, IEEE Std.
334-19/4.
'IEEE Recommended Practices for Seismic Qualification of Class 1E o
- Equipment for Nuclear Power Generating Stations, ANSI /IEEE Std. 344-1975, IEEE Standard for Qualification of Safety-Related Valve Actuators, o
'o IEEE Standard Seismic Testing of Relays, IEEE Std. 501-1978.
IEEE Standard for Qualifying Class 1E Motor Control Centers for o
Nuclear Power Generating Stations, IEEE Std. 649-1980.
Sel f-Operated and. Power-Operate.d Safety-Related Valves Functional o
Specification Standard, ANSI N2/8.1-19/5.. _,, _
.~
m y
Lh 1,
. TABLE 2.4-4
. Summary of reasibility Evaivation.
--r.i_rs.t.Se.ve. n..S..o.u.r.ces Category SRP 3.10 l
R.G. 1.40 1
R.G. 1.13 l
R.G. 1.100 l
R.G. 1.148 llEEE 5td. 334-1974l IEEE 5td.'344-1975 l.
Sampling Sampling is IA
- prototype unit" A aprototype unit" IA
- prototype unit" Implicit acceptance of l acceptable.
Ito be tested under ' to be tested under lto be tested under sampling, at least for lSamplesizeis -lmost adverse most adverse lmost adverse cases where fragility.
lnotdefined.
l design conditions. Idesign conditions. Idesign conditions.
I ltesting is performed.
I I
I I
2.
Similarity l l
l l
l
. Extension of EQ by test l
l l
l l
Ito similar equipment is l
ll l
l l
l iallowed using a combl.
l l
l l
l l nation of test and l
l ll l
lanalysis.
I
~
I 3.
Mounting lThe flature l
l l Ihe equipment shall be simulation
' design shculd l
l l
1 mounted in a manner simulate the l
l lthat simulates the
- actual service lintended service s
l mounting.
i
, mounting.
l 4.
Peripheral l Major peripheral l l
The ef fects of attachments l attachments are l
peripheral attachaents laddressed.
l l
l
'l Imust be considered.
5.
Dummy 1Dummy specimens l l
l IUse of dummy specimens compor.ents are allowed to.
l l
l Is allowed.
siellate the l
mass effects and l
l l
l l
dynamic coupling l
l
- to the supports.
6.
Generic l
loads l
i i
l ll l
l 7.
Enveloping lNot clear l
ll lThe assumption l
l
- The assumption is made, load luhether the l
ll lIs made.
l l
l assumption lassumption I
i l
l l
l Is made.
l i
l l
t i
8.
Required l
l l
l design load i
1 l
l l
l l
- A blank Indicates that no requirement was found.
TABLE 2.4-4 (cont.). summary of reasibility Evaivation.
First Seven Sources
,1 I
I f
~ ~ ~
I l
l Category l
SitP 3.10
}
R.G. 1.40 l
R.G. 1,73 l
R.G. 1.100 ll R.G. 1.l48 llEEE 5td. 334-1974, IEEE Std. 344 1975 l
l l
l 1
1 I
i g.
Margin lMarglas are l
l l Margins are l
l l10% margins are spect.
l required but l
l Irequired but l
l'
- fled for the response lnot specified.
l l
Inot specified.
l l
lspectrum at the
~
i I
l 1,
mounting point of the l
l
-l l
equipment.
i I
l l
l
- 10. Tolerances l I
l I
I l
l-1 i
- 11. Single vs. l Two simultaneous l
. lMulttamis testing is multlauls lames of input I
i l
lsuggested. Single atis testing lare generally
~ l l
l testing is allowed if trequired.
l l
l conservative, or if the l General proce-
)
i responses in the ames Fdures are l
l l
1
-l are independent, e
l'specified.
I I
l I
us I
N
- 12. Idave form 1 The character-I l
I Requirements for simu-listics of the l
l l
l l
lating earthquake are required input
-l l
- given. Specific should be l
l l
l requirements for proof lspecified by l
l l
testing are specified.
tresponse l
l i
i
(
l spectrum or time'I, l
l i
1,
- l
'l l
thistory methods.
l lihe requirement is
- 13. Fatigue istructural I
l iPerformance must I
ifive OBEs plus arf '.5E.
l Integrity and l
.l lbe assured during ll l
loperability l
ll land af ter an SSE l
Imust be assumed l l preceded by l
l Junder an SSE I
lseveral OBEs.
l l
preceded by l
l li l
iseveral 08Es.
l
- l li
- 14. Fragility I
i l
l l
Fragility testing i
recomended, but not i
required, for equipment l
l l
l to be used in a sua6er l
l l
1 l
of appilcations, l
l I
l l
l l
l
TABLE 2.4-4 (Cont.). 5-aary of feasibliity tvaluation.
First Se_ven Sour.ces Category l
SRP 3.10 I
R.G. l.40 l
R.G. 1.13 l
R.G. 1.100
. R.G.
1.14 8 llEEE Std. 334-1914 - IEEE 5td. 344-1915 l
- 15. Failures, I
l l
l l
l l
l l
l 1
l l
l g
- 16. Functional 'I 0perationality l General, Indirect, Reference is made 5eismic input is For devices (relays, require.
Ishould be l
l l references to jto ANSI N218.l.
l assumed to occur l motors, sensors), it is ments lvertfled during l l
l functionality are 11975.
Iwlth motor stand-l assumed that the land /or after l
lgiven.
l lstill, starting,
- selsmic input.can be l testing.
l 1
l running,or imposed while simul-
)
1-l l coasting down.
- lating normal operative I
l l
l l and sensing perform-l l
l
.ance, l
l
- 11. Critical l
l i some Parameters are.
parameters l
l sugge*.ted as possibly critical and are
,recomended for identi-c.a l
['
- fIcation, 00 g
i i
i
- 18. Degradation l
l l
JUpon completl6n under test l
l l
lof the test, the ll l
l l
l l
[ motor shall be l
l l
l l
l
.l dismantled and l
l' I
l l
l l Inspected.
i
- 19. Response I
l l
l l Monitoring is required, I
l l
l
' lbut specific require-1 l
cents are not given.
l l
l l
- 20. Unexpected l
l l
l l-Analysis Alght be used results l
l-l I:
l llto emplain unexpected l
l l'
.I l behavior during a test.
l l
1 I
- 21. Load lIt is not clear l l
l l
l Normal operating loads combination lwhat combina-l l.
l lwhich adversely affect ltions are l
l l
l function must be acceptable.
I I
l l
l combined with seismic l
l l loads.
l l
l l
1
s TABLE 2.4-4 (cont.). 5 ary or reasibility Evalsstion.
First Seven Sources 1
,SRP 3.10 R.G.~ 1.40 R.G. 1.73 R.G. I.100 R.G. 1.148 IEEE Std. 334-1974 IEEE Std. 344 1975
- 22. Load l Load sequencing l l
lload sequencing l
l sequencing lIs to follow l
lis Indirectly
.l l
lIEEE Std. 323-I laddressed.
I l
l1974 I
l l-I i
1 1
1 l
l 1
1 1
1.
- 23. Errors l
l l
l l
l l
1 1;
I l
l l
l
- 24. Maintenance 1 l
l l-l
- 25. Mounting l Requirements l
l l
l Jhe mounting method afeguacy lon mounting l
l l
l lshall be the same as ladequacy are I
1 l
l l
Ithat recommended for (given with 1
I l
l l
l active service.
Jrespect to l
l l
l 1testing and/or I
l l
l analysis l
l g
- assessments.
1
- 26. Post earthquake l
l l
l l
1
- 27. Value/
I l
l l
l I
l l
l l
lapact l
l l.
l l
ll Igenerally r,ecommended analysis lIs preferred.
l l implicitly l
1 without test exc/pt luhere structural l
l laccepted by l Integrity alone can lIEEE Std. 344-l 11975.
- l ensure equip =ert l
l function.
I I
l l
l 5
gt y 5
in pl ee 7
t cu n
9 s cg-a y~
1 eaa cb t
v g
- 4 s
gQn 4
diy nEi 3
e l
i t.
nsn tf ss i o soei d
ibs e
t s t
m yt t t y
- S ol u.
rdl cabd uaea
.E n
e t pnn E
ya
.n i
ia E
b ei
- s. a b I
dlf md' Qnbe ne on' ll E aad ibc a 4
llill 7
9 1
4 3
3 d
t S
E E
. E I
llIIil ll 8
4 1
n.
1 io G.
ta R
r s
t e
I
'l ncr E
uo y S t
a n
i ev 0
lc e 0
S 1
sa t
1 e
s r
r G.
i r
F e
R yr llIIll a
l ll mu s
3
)
1 t
1 no c
G.
(
R 4
1 l
ll l
4 2
0 4
E l
L B
A G.
T R
l l!
l lIl g.
nd i e 0
t r 1
sie u 3
t qs ei.
P ur d
R t
t e iti w o o s.ntl n
ul isba llIIli1ll lill1 yr g
is ug o
n s
tn g
yi y
ii e
ht l
st t
sda
- s a
Qenn ne C
Et aa It 9
0 2
3 I
bo 8
TABLE 2.4-4 (cont.). scmary of reasibility evaluation.
Sources 8-12 and Other Data l
Score on l l
l l current l
l $ core on
, IEEE 5td.
1[EF $td.
IEEE 5td.
ANSI N218.1-ANSI B16.41-
, require-
'[
,esperience 382-1980 501-1918 649-1980 1975 1981 l ment.*
Esperience data data
- 1.
Sampling lA procedure I A minimum of lAt least one I
liesting of at 1
3 Several units are 1 6 lsaggested for lthree specimens idevice must be l
lmost one sample. l (commnly excited at l
l selecting the lls required.
Itested, but not l lls acceptable.
I jonce by an earthquake. I,
, test units is l
lone motor lTherefore,emperience lgiven in App. A. l l control center.
Idata are potentially l
l l
l l
l rich in sagling.
ll l
I 2.
51milarity 15tmllarity is lExtension of l General guide-i Guidelines are l 3 I; Equipment among non-6 laddressed in lquallfled relays llines are given,
Iqualification lusually quite slallar, i
given to catend l l nuclear facilities is l l terms of generic,to relays not lto eatend the l
l groups of valve l tested is l qualification l
lof value A casual comparison l actuators from l allowed.
lof motor control
' assemblies to also indicates that Ishich test units l l centers to other similar units.
the equipment is also lare drawn.
l lunits.
l
.quite stellar to l
I I
I that in nuclear l
facilities'.
i
~
e 3.
Mounting iThe valve actu-lihe relay must 11he motor l
4 Experience data 6
M simulation lator is required lbe mounted as it.l control center i reflect the true lto be mounted to lnormally would Imust be mounted l l
l l mounting conditions.
l lthe shaker table lIn servlCe.
las it would be l l
1 l Therefore, mounting is las it would be
'I lin a plant.
I I
Inot an issue for such l
I data, attached to the i
valve.
I 4.
Peripheral IElectrical, 11 l Anticipated l
l Electrical, 4
lThe credibilty of l 6 attachments l hydraulic,or ladditional l
lhydr'aulic,or.
I leffects from l
l pneumatic l
l weight and l
[ pneumatic
] peripheral attach l
l connections must I testernal connec-l l connections l
lments is not an issue l lbe attached.
l ltions shall be l
.shall be l
Ifor experience data.
l l
Istmulated.
I required.
1 I
5.
Dumy
]
I l
4 lDumy. specimens do not I; 6 components i
l i
represent an issue for l l
1 l
'l experience data.
l l
I 7
6.
Generic l Generic loads ifragility l Generic load EQ lNot Not loads lfor valve actu- ' testing is itechniques are
[ required.
required.
lators are required for l allowed for
[
lestabIlshed for relays; there-Igroups of equip-I i
imost plants, ifere generic ument.
I 1
l lloads are I
l l
l lessentially l
l l
l Irequired.
.I l
l l
l l
1 l
l l
1
- 5ee Table 81.
.~
4 TABLE 2.4-4(coat.). Summary af feasibility avaisation.
Sources 8-12 and Other Data l
.I I
i; 15 core on l l
l l
l current-1-
Score on i IEEE Std.
l IEEE Std.
I IEEE Std.
ANSI N278.1-ANSI 816.41-Irequire.
' emperience l 382-1940 1 501-19181 l 649-1900 l'
1975 i
1981
- ment.*
Imperience data Idata*
l I
7.
Enveloping IEnveloping is l
.I I
l 2~
Espertence data could. I 2 lagd 1probably estab-l ll l
l provide an Indication I assumption Illshed through of equipment perform-l generic loads, ance at loads that 1
i 1
l l
, envelop required loads l l
l l
l Ifor EQ.
l.
8.
Required l The required 1
6 l Although loads esperi-3 design l design load l
l I
enced are realistic, l
load lmay be deficient.1 1
Ithe adequate reflec-l Itton of such loads to
.l-
'l l
l l
areas of concern in J
l l
- Eq of nuclear plant l
l l equipment day be l
l 1
lacking.
I 9.
Margin (Margins are (fragility
. Margins are l
3 15cee evaluations l 6 included in the testing includes specified la l
l Indicate that some non-j nuclear facilities genefic load..
lthe concept of Table 1.
have empertenced Jaargins.
I A
l l
l
- lselsmic loadings '
l l
1 in excess of design l
- 1oadings in tuclear l
8 l
l 1
- facilities.
l l
l l
l l
l 1
I
\\
I 1
I i
- 10. Tolerances l
l Tolerances are I
i l
lNot l
lNot required.
required.
l lspecified for i
i l
linstrumentation.
l l
l l
l 1
I I
I i
- 11. Single vs.
l8tantal testing ITriasial testing l l
6 I Expertence data in -
6 l
multlaxis Irequired.
Ils desired, but I
general consist of lblaxlal testing I three-dimensional lis acceptable.
l I
excitation.
- 12. Wave form l Requirements are lTwo multi-I g
Inputs in experience l 6 l consistent with l frequency, l
. 1 I data can be either I
llEEE 5td. 344-standard I
narrou banded if the
!1974.
l response spectra equipment is mounted Jare specified l
on a structure or lfor qualifica-l piping system, or ll l
'l l
l lbroad banded if ltion of relays.
l l
l mounted on the l
l l
l l foundation.
l 1
l l
a TABLE 2,4-4 (Eent.). 5 - ary af reasibility evairatloa.
Sources 8-12 and Other Data l
l l
.I
}
l5 core on l l
4 I
l l
Il current licore on i
l IEEE Std.
I IEEE Std.
l IEEE std.
ANSI N278.1-1 AN51 816.41-l require-I lcaperience 1 382-1980 l 501-1978 l 649-1980 1
1975 l
1981
, ment.*
i Esperience data Idata*
I l
I I
I l
- 13. Fatigue 108E and SSE
.,Five OBE plus an 3
Low-cycle fatigue may I 3
l testing are 1 SSE testing are I
l he revealed by esper t-
- reqalred. Each l required. Mini-
,ence data. '
l l test must be laum duration is l l
L l
l-115-s minimum.
l15-s per test l
l l
l l
l-l l
l l
l 1
I i
,Present Indications i 3 l
- 14. Fragility I
- Fragility l
l l
l 3
. lfrom a limited review l l
ltesting is l
1 l
l l
l required for l
i l
l lof esperience data l
l
,l relays.
1 l
l lsuggest that few or
)'
l l
l l
Ino fallures of equip. l l
l.
I iment will be observed.
I l
I i
- 15. Failures l Determination of I.
l 1
3 I Failure information 3
lwhat consitutes l
,may be limited.
Ifailure for 1
l l
Irelays is given. 1 l
l l
l A
- 16. Functional lvalve actuators Relays must be
, Motor control l valve assemb11es
, Functional I g
.Esperience data on 1
6-W requirements joust t,e func-l tested in the l center opera-Imust be operable l requirements are l I,the functionality of i e
ltional before, ltransttlon from ltional capa-lduring and af ter (given for valve I lequipment may be l
Iduring,and (nonoper4 ting (mustbe Ithe' test.
lassemblies.
[relatively scarce.
l lafter testing.
lLo operating
. demonstrated.
l l
l l condition.
l l
l l
l ll l
- 17. Eritical l
l l
1 lSince few, or no.
1 0
parameters l l
l lfallures have been l
l l
\\
lobserved, it is l
l l
l l
,unilkely that esperi-l l
l l
l l
lence data will reveal l
l l
l ll l
l l critical parameters.
l l
l l
l l
l lfhe most important l
l I
lfallures observed l
1 lhave been f ailures of l mountings or attach-l i
1 1
I iments.
I I
1 l
1 i
l l
- 18. Degradation !
L I
lingpection of 0
l Degradation is 3
under test I tvalve assemblies l generally not an l
1 lshall be lissue for esperience l
l l
performed before i data.
-1 I
I l and after l
l
, testing.
I U
1
- 5ee Table 81.
1
Table 2.4-4 (cont.). Suunry er faasibiitty evaivatten.
Sources 8 12 and Other Data i
i l
15 core on l
l l
l l current
-15 core on IEEE 5td.
IEEE Std.
l IEEE Std. 1 ANSI N218.1-
. 1 ANSI 816.41-l require-1 l eaperience 382-1900 l 501-1918 1 649-1960 l'
1975 l
1981
'iment.*
I Esperience data data *.
- 19. Respor.se l
l l
l l
lNot I
(Not l
l l
l l
l required.
' required.
- 20. IJnexpected l
l i
ll INot lNot results l
l l
l Required. L required.
i
- 21. Load I
15elsmic testing I
l l
6 Normal operating loads l 4 combination l Iof relays can be l l
l ll loads are expected to l ll l performed under I
I be present already l
lprevilling I
1 lwhen an earthquake l ambient condi-
, occurs.
ltions of the test l l laboratory.
I l
l 1
l I
i 1
- 22. Load lA standard load l l Sequencing of I
l A sequence of 6
. Equipment in operating I' 4 i
sequenclag l sequence is l
lpreaging and l testing is l
l plants can be espected l 8
required.
.lselsmic testing specified for I
to have exper.enced I
lls specified.
valve assemblies, normal environments, I
ae i
l l
. transients, and l
l l
l l
l lIn-situ vibration.
l e
- 23. Errors 1
0 l Equipment in plants' 2
l l presumably has been 1
l linstalle'.;lth a more 1; I
l l
l lor less typical set i
l 1
ll l
l l
lof errors.
l 1
l I
i
- 24. Malatenance ' Maintenance to belMaintenance can l Modifications lif maintenance 0
Emperience data should 2
l performed (uring the performed lduring testing l lor adjustments l
lhe valuable in assess-l lthe test must be laf ter a given lshall be evalu-l lare required l
lIng if, and how, main-1 ltenance affects seismic l lspecified.
l fragility test. lated to deter-l Iduring testing, l performance.
lmine their l
l acceptance of l
leffect on the l
lthe test must be lEQ.
l evaluated.
l l
- 25. Plounting IThe valve actu-l Recommended l Mounting must be IThe valve 9
l Fallure of mountings
.l 6
lappears to be the adequacy lstor must be l mounting hardware lby welding or l assembly must be mounted to the Imust be used, lbolting for supported as single most important l
, shake table as l
lselselc testing.
. required to
.fallure; therefore.
l I
l lt would be l
l I per'elt testing lemperience data can be l mounted to a l
l lIn accordance lempected to provide l
l valve.
j l
1, lwith the luseful information l
- regarding mounting l
l l
l 1
l standard.
I adequacy.
l l
l l.
l 1
l l
l l
H I
Table 2,4-4 (cont.). Sumary or feasibility evaivation.
Sources 8 12 and Oth'er Data T
i C" "-
~ ~ l-I score on i-i l
l I
-1 I
icurrent I l$ core on lesperience l IEEE Std.
l IEEE Std.
I IEEE Std.
ll ANSI N218.l-ANSI 816.41-require. l' Emperience data data
- 649 1980 1
1915 1981
' ment.*
1 382-1980 501-1918 1
I
- 26. Post l
l 2
Equipment esposed to l
I earthquake l:
I an earthquake is subse.l.
I quently subjected to normal operation, itranslents, etc.
I therefore, experlence l
l
'l l data should be useful l l
I for assessing post.
)
l' I
learthquake behavior.
l ll l
l lbut only partially.
I I
I
- 27. Value/ impact I
i
! Not not I
l i
Required.
Required.
- 28. EQ by
,EQ by analysis lEQ by analysis lEQ by analysis IEQ by analysis.,,
1 Expertence data are at 1
analysis its allowed to lls assumed to be lis allowed.
I ls allowed.
i i least as amenable to lentend quellff-possible.
I i
l analysis as fQ is 1
cation of a through ordinary means.
gener,1c group to I,a specific 8
lapplication.
l l
- 29. EQ by lEQ by combinatiblGeneral require-EQ by combination l
((Q by combination 1
- fhe use of combined I
i testing and lof test and iments are given.
of test.and I
lof test and I
test and analysis in analysis lanalysis is l
lanalysis is l
l analysis is esperience data would lallowedtoextendl l allowed.
l l allowed, i
- have to be defined in imore detall to make EQ of a generic l l
l l
i l group to specific l l
Ia good evaluation of i
its value.
applications.
I
- 30. In.sttu l
I i
l l
1 It should be possible 3
testing 1
I l
l 1to develop acceptable J
I lin.sttu techniques i
i l for nonnuclear l
Ifacilities and nuclear 4
facilities alike, i
i TOTAL 5 l
l 1
l l 91 97 j
l I
l' I
l I
_l l
i s
A 4
Moderately Adequate - 2:
This is 'the next highest ranking.
- Marginally Adequate - 1:
This is a. poor ranking.
The issue is addressed, but_not very satisfactorily.
Inadequate - 0:
This is the worst rankin'j.
The issue is either not addressed at.
all or, if it is addressed, it is addressed poorly.
Rar. king not required.
This ranking usually occurs when an issue does not have to be addressed, but is included for completeness.
Next, the use of experience data was also evaluated for each of the thirty categorios.
The same ranking as above was used.
These rankings were then weighted according to importance, and the two sums (current requiements and experience data) compared to arrive at a feasibility judgement.
The results of the evaluation is summarized in Table 2.4-4.
Table 2.4-4 shows that when the current requirements in existing NRC.and national standards were evaluated against the common set of thirty issues, they were estimated to score 91 out of 156 overall, or about 60%.
Experience data was estimated to score 97 out of 156 overall, or also about 60%.
Since the current requirements and experience data score about the same (60%), this 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 guidelines for the use of experience data.
For.all the categories considered to be the most important (those given an importance ranking of 3), guidelines were developed.
The categories considered are:
- Sampling
- Similarity
- Required design load Margin Single vs. multi-axis testing Wave form Fragility Failures e
Functional requirements Mounting adequacy l
J
)
The guidelines, as taken directly from the LLNL report, are
- combined under the.five headings as follows:
Sampling-(1) -Experience data should be gathered oniall 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.
We anticipate 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 j.
functional failure are included in the data base;_then we.do not recommend that the NRC accept experience data as fully as we have 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 4
items of each type and size of interest are found, then a Justification must be provided to extend the experience data.
Similarity (3) The issue of the similarity of equipment in non-nuclear facilities to equipment in nuclear facilities must be.
addressed.
However, exact similarity need not be established.
Rather, what is required is reasonable assurance that the equipment in non-nuclear, facilities
- is of the same type and basic' design
~
was manufactured by'the'same manufacturers in the same period as the. 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 facilities must be established in order to obtain i
a " rough" idea of the type of earthquake motion it i
experienced.
Rough" means that dynamic modeling or analysis j
is.not. required.
Two categories are suggested:
a.
Dimensionality. Was the earthquake motion affecting the equipment predominantly one, two,--or three-dimensional i
in nature?
t
~
a w
v s-.
e-
-, - - - ~,
r, 7-
,,,,,ym.e, yr,
,.,p-w-
,.q,-
-g
,,nm,p,-,.w,,,..e,--p
,c,..n+-
4
,.e--
-.wg-,ng,,,.--,,
i l
b.
Wave form.
Was the earthquake motion affecting to the equipment:
- 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 horizontal 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 significant 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 earthquakes 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 more detailed aspects of the motion affecting the equipment in non-nuclear facilities (for example, in-structure response spectra) and its relation to similar motions in nuclear facilities.
The authors believe that the above requiement for acceleration and duration 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, then three steps are. recommended:
6 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 Ref. 2, 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 Ref. 11, 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.
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 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 plan: i.
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 procedures in the check list should be drawn from three sources:
a.
Information gathered from the collection of experience data;
-b.
Information gathered from laboratories experienced in seismic equipment 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 anomolous behavior An example of such testing can be found in Ref. 12.
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 equipment 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 selection 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:
i I
125 y 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 alternative to 100 tests on only two categories of equipment, as outlined above, a minimum of five tes s on 20 or so categories would be acceptable.
2.4.3 Summary of EQE Report " Pilot Program Report-Program for the Development of an Alternative Approach to Seismic Equipment Qualification Numerous non-nuclear power plants and industrial facilities containing equipment similar to those in nuclear power plants have experienced major earthquakes in various parts of the world.
A sample of this experience is shown in Table 2.4-5.
The SQUG with help from EQE, initiated a pilot program to evaluate the potential for using experience data as the basis for qualification.
Stated goals of the pilot program i
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 equipment 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 earthquakes data to evaluate the necessity for seismic qualification of specific items of equipment by conventional methods.
2.4.3.1 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 earthquakes (also referred to as data base plants by the SQUG).
The steps involve'd in collecting data from the data base plants and the nuclear power plants and in comparing the data are shown in Figure 2.4-1.
Before walkdowns of the data base plants were conducted, available records of the seismic event at each site were collected.
I 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.
r i
I i 1 j
TABLE 2.4-5 SELECTED MAJOR EARTHQUAKES THAT HAVE AFFECTED POWER AND INDUSTRIAL FACILITIES (Yanev, 1981) l Recorded Estimated Peak Number Approxi-Ground of Power mate Accelera-Ground Plant Richter tion Motion Units Earthquake and Location Year Magnitude (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, Washington 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
+ Indicates equal to or greater than the number shown.
- Actual number not determined.
NUCLEAR RJWER PLANTS DATA BASE PUWrs REVIEW TYPE OF EQUIPMENT REVIEW RECmDS CN FACILITIES idilCH HAVE EXPERIENCED EARTN UAKES v
y SELECT REPRESENTATIVE PLANTS SELECT REPRESENTATIVE AND EQUIPMENT AND PERFORM PLANTS & PERFORM WALKDOWNS WALKDOWNS v
SELECT PLMTS & EQUIPMENT SELECT PLANTS AND EQUIPtENT FOR DETAILED SNfLING FoR DETAILED Suetite i
Y Y
COLLECT ECUIPMENT DATA COLLECT EQUIPMENT DATA AND AND Floor RESPONSE SPECTRA FLOOR RESPONSE SPECTRA X.
f C mPARE EQUIPMENT DATA AND RESPONSE SPECTRA DcicHM-INE IF EQUIPtENT REQUIRES DETAILED QUALIFICATION Figure 2.4-1 ETF00 USED IN PILOT STUDY
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 commonly encountered safety-related equipment.
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 walkdowns were used for collection of detailed data, including conducting in-situ dynamic testing.
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 2.4-6).
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 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 the 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. structures are relatively stiff, many are either light concrete structures with shear walls or braced steel-frame structures.
Therefore, the 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 conservatively 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.
1 -
TABLE 2.4-6 EQUIPENT SELECTION Ecu!FTENT SELECTED:
MOTOR CONTROL CENTERS 480 VOLT SWITCHGEAR 2.4 To 4KV SWITCHGEAR MOTOR-OPERATED VALVES AIR-OPERATED VALVES HORIZONTAL PLNPS VERTICAL PUMPS SEVEN NUCLEAR POWER PUwTS VISITED - THREE SELECTED FOR EQUIPMENT DATA COLLECTIONi DESIGN BASIS PLANT SSE DRESDEN 3 0.2G CALVERT CLIFFS 1 0,15G PILGRIM 0.15G l
l l
i - -
Table 2.'4-7 Comparison of equipment data.
ITEM:
480 Volt Motor Control Center Cabinets 480 Volt Motor Control Center 39-3 l _
IVA-6VA, P3A & P4A (Eight Units)
PLANT:
Sylmar Converter Station Dresden Nuclear Plant, Unit 3 MANUFACTURER:
General Electric 7700 Line Series, 1970 General Electric 7700 line Series, 1971 Sylmar Converter Station basement, facing Reactor Building Elevation 570, facing LOCATION:
northeast and southwest east (grade is at Elevation 517.5) i FUNCTION / SYSTEM:
Control of pumps and valves for rectifier Control of various Class I mechanical
{
cooling systems systems CABINET:
Each cabinet is four cubicle's wide; the Cabinet is six cubicles wide. The cabinet i
specific arrangement of starter units varies contains starter units'in cubicles of i
from cabinet to cabinet; they are otherwise various sizes.
i very similar.
u, f
COMPONENTS:
A typical starter unit consists.of a General A typical starter unit consists of a Electric CR-106 magnetic contactor, a cir-General Electric CR-106 or CR-105 magnetic cult breaker switch, a control transformer, contactor, a circuit breaker switch, i
on-off pushbuttons and a terminal block.
a control transformer, on-off pushbuttons, j
and a terminal block.
i ANCHORAGE:
The bottom channe) of the cabinet is tack The bottom channel is tack welded to welded to a baseplate embedded in the con-an embedded baseplate, two welds at the crete floor.
At least one cabinet was base of each stack of cubicles, front inadequately anchored at the time of the and back.
earthquake and slid a few inches.
APPLICABLE The records taken at Pacolma Dam are shown The calculated floor spectra for the
RESPONSE
scaled to 40% of the measured amplitudes as Reactor Building, Elevation 589 are shown.
SPECTRA:
a conservative estimate of the ground mo-Spectra at Elevation 570 were not generated.
tion at Sylmar.
EQUIPMENT The MCCs were in operation at the time of the earthquake.
No damage to either cabinet STATUS DURING or components was reported.
One cabinet slid a few inches due to lack of floor anchorage.
AND FOLLOWING Tile EARTHQUAKE:
Source: References 1, 10, 20, 25, 26, 27, 30 and 32 (Appendix A).
--. ----~2.
TABLE 2.4-8 Slft%RY OF DATA BASE PLNfiS
& EARTHOUAKES ESTIt%TED EARTHOUAKES FACILITY PGA SAN FERNANDO
- 1. SYU%R CONVERTER STATION 0.50 - 0.75 1971
- 2. VALLEY STEAM PLANT 0.40
- 3. BURBANK PWER PLANT 0.35 "
- 4. GLENDALE PcwER PLANT 0.30
- 5. PASADENA FC ER PLANT 0.20
- 6. RINALDI RECEIVING 0.50
- 7. VINCENT SUBSTATION 0.20 "
- 8. SAUGus SUBSTATION.
0.39 POINT MAGU
- 9. ORMOND BEACH PLANT 0.20 1973
- 10. SANTA CLARA SUBSTATION 0.10 1
l SANTA BARBARA
- 11. GOUETA SUBSTATIok 0.28 1978 I
- 12. ELLWOOD PEAKER PLANT 0.30 - 0.40 IlfERIAL VALEY 13 8. CENTRO STEAM PvwT 0.51 1979
- 14. f%irwux GEomERf%L PuwT 0.20 - 0.30 BASED ON:
Be00RDED PEAK GROUND Arm FRATION - AT Puwt StTE
~
l.0CATED NEAR Sm0NG (DTION RECORDS l
l l
The performance of data base equipment during past earthquake was evaluated and conclusions regarding the seismic resistance capability of similar nuclear equipment were reached.
A typical comparison is shown
'in Table 2.4-7.
i For the purpose of the pilot program non-nuclear power plants and other facilities in southern California where significant earthquakes have occurred was chosen for the study.
Table 2.4-8 shows the four earthquakes in southern California that were reviewed in detail in this program.
The facilities that contained the largest number of equipment items of interest and were reviewed in detail are the Sylmar converter station, Valley steam plant, Burbank power plant, Glendale 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 are Dresden Unit 3, Calvert Cliffs Unit 1 and Pilgrim.
These plants were selected so that the equipment reviewed for the project would be a representative sample of a variety of nuclear plant characteristics, including reactor type and vintcge.
Only equipment required for. safe shutdown was considered.
2.4.3.2 Conclusion and NRC Staff Comments The ' goals of this pilot program were evaluated by the SQUG against the results obtained from the study.
Table 2.4-9 listed the goals, findings and conclusion as seen by the SQUG.
Finally, the SQUG arrived at the following overall summary:
The SQUG has shown that the structural integrity of anchored power plant equipment and component is not compromised in strong earthquakes of up.to 0.50G peak ground acceleration.
The SQUG.has shown that typically power plant equipment operability is not compromised in strong earthquakes with peak ground acceleration of about 0.20G to 0.30G.
While the staff is in general agreement'with the SQUG on the first overall summary, the staff however has reservation on the second overall summary.
The NRC staff has completed the review of the pilot program report, and
~has concluded that it is feasible to accept experience data as a basis for seismic qualification.
The staff realizes that the SQUG pilot program was intended to be a feasibility study and reviewed it in that context.
Therefore, the comments are generally an assessment of what further work should be done to provide an acceptable experience data base. The comments, which were sent to the SQUG in December 1982, were
~
organized in eight subject areas as follows:
- i I
TABLE 2.4-9 MAJOR CONCLUSIONS OF SOUG GOAL 1 GOAL:
DEVELCP A HISTCRICAL DATA BASE CN THE PERFCRPMCE OF ECUIPfd:RT IN CCNVENTICNAL PCWER PLWTS DURING AND AritR STRONG EARThCUAKES.
FINDINGS:
SEVERAL PCWS PLANTS AND OTHER INDUSTRIAL FACILITIES HAVE EXPERIEMCED STRCNG EARTHCUAKES -D(CE:. DING me ent: rIELD SAFE-SHUTDCWN EARTHGUAKES REQUIRED FOR THE DESIGN OF MCST U.S.A. NUCLEAR PLANTS.
THE FLWTS RESPCNDED WELL TO THE EARTHCUAKES 2D USUALLY l
CCNTINUED TO C.FERATE OR WERE BACK CN LINE SHCRTLY Ar:::t THE EARTHQUAKES.
IAANY CF THE FAC*LITIES hERE IN CFERATICN AT THE TIME CF THE :ARTTUAKES; THUS THEIR ECUIPWT WAS SU2JECTED TO NOPPAL CPE:ATING LCADS IN ADDITICN TO THE SEISMIC LCADS FROM THE EARTHCUAKES.
WITH A FEW M:NCR EXCE:TIONS, THE ECUI?fdENT CONTAINED.IN l
THE P&ER FACILITIES WAS. UNDNdAGED AND WAS FUNCTICNAL l
Arick THE EARTHGUAKES.
IHE ECUIPMENT WAS NOT K?ON TO BE I
MCOIFIED BECAUSE CF THE EARTdGUAKES.
' SUFFICIENT DATA EXIST TO ESTIMATE THE SPEt.ixA EXFiERIENCED BY THE PLWTS MD TrEIR EQJIN.
THERE IS A LARG, AVAILAa_E MTA 3ASE, CHLY A PCRTION CF
'rHICH WAS SNPLED IN TH'IS STT.!DY, CF PCWER Ft3HT EQJIFMENT THAT HAS BEEN SLE ICTED TO STRGIG EARTHQUAM S.
l CCNCLUSION -
l THERE IS A LARGE BODY CF AVAILABLE DATA ON THE PERFOfPANCE OF POWER PLANT EQJIIHENT IN STRGG EARTdGUAKES, INCLUDING BOTH l
tECFANICAL AND ELECTRICAL EQUIPMENT. IdANY CONVENTi MAL P G ER PLANTS AND INDUSTRIAL FACILITIES FAVE EXPERIENCED E3RTHQUAMS THAT SUBJECTED TIEIR EujIFWNT TO SEISMIC ENVIRGfENTS EQDL TO OR EXra nING SEI94IC LOADS ASSOCIAttu WITH SAi: H HJiDtMM EARTHQUAMS REQJIRED FOR 'lHE DESIGN OF MOST nun :E FCMER PLANTS.
l l
-l
TABLE 2.4-9 (C0t:TINUED)
GOAL 2 GOAL:
SHOW THAT MUCH OF TM ECUIPMENT INVESTIGAntu, WHICH FAS EXPERIENCED STRCNG EARTHQUAKES, IS SIMILAR TO EQUIPMS R FOUPO^
IN NUCLEAR PCWER PLAtGS.
FINDINGS:
A FEW MAJCR ECUIPMENT PMUFACTURERS SUPPLY I4)C}i CF TriE EQUIRENT FCR BOTH CGWENTICNAL AND t UCLEAR PCER PLMTS.
IdRE IS LITTLE CESERVABLE DIF:ERENCE BETiGEN T4 MEASURED DYtWi1C RESPCNSE FRECUENCIES OF ECUIP"ENT IN NUCLEAR PO GR PLANTS AtO THOSE IN CCNVENTICNAL PtA4TS.
THERE ARE NO GENERIC DIFFERENCES OTriER WAti AGE BEEN ECUIPMENT FOUND IN CONVENTICNAL MD NUCLEAR PCWER PLMTS.
1 CCNCUJSIONS:
CERTAIN TYPsS OF MEGANICAL AND ELECTRICAL EQJIPENT FOUND IN NUCLEAR PCWER PLANTS ARE VERY SIMILAR IN CCNFIGURATION, PUNCTION, MMUFACTURER, AND MCCEL TO Td TYPES FCUND IN CCNVENTICNAL PLANTS.
MUCH OF THE EQJIPMENT IN NUCLEAR PCHER
? LINTS MO CCNVENTICNAL PCWER PLANTS IS Td SME..
l TABLE 2.4-9 (CONTINUED)
GDAL 3 GDAL:
DETEFMINE WETHER ACTUAL EARTHCUAKE DATA ARE SUFFICIENT TO CONCLUDE THAT SEISMIC CUALIFICATION OF CERTAIN CLASSES OF EQJIPMENT BY CCtNENTICNAL MEWCDS !S NOT NECESSARY.
FINDIfGS:
EXCLUDING SCME UNN4 CHORED EQUIPMENT AND CNE AIR-OPERAle VALVE, NO FAILURES WERE REFORid IN ANY OF THE SEVEN TYPES OF ECUIPMENT ADDRESSED IN THIS STUDY.
WITH W.E POSSIBLE EXC:.-i10N OF ELECTRICAL R:J AYS, THERE IS NO EVIDENCE CF PALFUNCTICN OF Td REVIEnED EQUIPMENT l
DURING THE EARTHCUAKES.
IHE ESTIMAi c GRCUND-RESPCNSE SPECTRA FRCM SEVERAL CALIFCRNIA EARTriCUAKES AND TriE CONENTIONAL PCWER PLN4TS AFECTED BY ThEM ENVELCFE TriE FLOOR-RESPCNSE SPECTRA FCR TriE SAFE-SHUTDCW EARTFCUAKES REQJIRED FOR NUCLEAR POWER PLANTS IN DE Pf-NGES OF MOST EQJIfM.NT RESPONSE FREGLENCIES.
CCtNENTICt4AL PLNiTS TPAT kERE SLEJEt.:c TO EARTriCUAKES WIDi PEAK GROUND ACCELERATICtt OF ABCUT 0r3C(5 OR LCWER GENERALLY CCt4TINUED TO OPERATE THROUGY.UT TriE EARTriGUAKES.
CWCLUSION:
SEISMIC CUALIFICATION OF NUCLEAR ECUIFFENT SY CCfNENTICNAL PETriODS DCES NOT APPEAR _TO BE.NECESSARY FOR TriE CLASSES OF ECUIPMENT EVALUATED FOR MOST LEVELS OF SAFE-ShUTIXMI F.ARTHCUAKES.
l I
TABLE 2.4-9 (CONTINUED)
COAL h COAL:
DEVELCP A METHODOLOGY FCR WE USE OF ACTUAL EETHCUAKE DA CtitRMINE bHETHER SEISMIC CUALIFICATION CF SPECIFIC IT#S OF ECUIPMSIT BY CCtNFETICNAL tETMOS IS ffECESSARY.
FINDINGS:
THE SEIS4IC FERFCtTANCE OF THE REVIEWED ECUIPMEN~rAFFEA TO BE INDEFENCEfT FRCM N1Y OF THE FOLLvMING FACTCRS:
AGE OF ECUIPMENT YEARS OF SERVICE EkNUFACTURER AtO MCCEL YICUNTING CCNFIGURATION DYAN4IC PROPERTIES IHE PETHCIXX.Cc/ USED IN THE PILOT PROGR44 TO EVALUATE CUSSES OF EdJIRTNT WCULD BE ECUALLY APPLICABLE TO SPECIFIC ITEPS OF EQUIWSir.
CCNCLUSICN:
THE PILOT F,ROGRN4,HAS DEk.NSTRAico THE METHODCt.0GY.
THERE IS AN ASUNDANCE OF DATA THAT CAN BE USED TO IDENTIFY SPECIFIC ITEMS OF ECUIPMENT THAT DO NOT RECUIRE ACDITICNAL SEISi!C CUALIFICATICN.
~ 62 -
(1) Extent and Organization of the Experimental Data Base Five of the six data base plants included in the pilot study were subjected to the 1971 San Fernando earthquake.
The sixth plant, the El Centro Steam Plant, was subjected to the 1979 Imperial Valley event which was of relatively short duration and originated from a source very close to the site.
It is widely recognized that certain earthquakes are more damaging to certain types of structures and that duration time and travel path are important parameters.
The staff believes that the data base should be extended to include additional events.
The additional events should include to the extent possible, a wide range of durations, amplitudes, and frequency contents.
The data base should be organized such that all equipment in a "similar" group is catalogued along with a record of the iudividual event time histories and a corresponding response spectra which can be applied to members of this group.
A minimum of three separate and distinct earthquake histories should be used to construct a data base spectrum.
(2) Adequacy of Seismic Input For the two earthquakes considered in the pilot program, characterization of the shaking motion of the non-nuclear sites is adequate. The staff believes it is reasonable to assume that the seismic environment experienced by equipment located on the first floor or in the basement can be characterized by the estimated free-field motion at the building foundation, e.g., equipment located at higher levels in the plant can be assumed to have been subjected to the estimated free-field motion at the foundation except in cases where sharp, short duration peaks in the time history may have been filtered and " damped out" by SSI effects as may be the case for the El Centro steam plant.
If an acceptable dynamic analysis is performed on a data base plant to calculate response spectra at equipment attachment points, then these amplified floor spectra may be used as the experience d 'a input.
When comparing experience data, spectra to the required response spectra at a nuclear plant, the nuclear plant seismic analysis and development of floor spectra must meet current regulatory requirements properly accounting for the hazard and the spectrum shape or be approved by the staff for use in qualifying equipment on some other defined basis.
Recent information indicates that eastern earthquakes may exhibit significant frequency content above 15 hz.
The staff is continuing to evaluate this phenomenon.
In the event the staff determines that seismic design criteria should be modified, to account for the high frequencies, a change in criteria would be proposed.
(3) Adequacy of Equipment Anchorages Equipment qualification by comparison with experience data assumes the adequacy of equipment anchorages.
Deficiencies in the anchorages of electrical equipment were identified in the SEP phase II review and resulted in the issuance of IE Information Notice 80-21.
The staff believes that, regardless of the qualification procedure used, equipment anchorage is a key element and adequate anchorage must be proven.
This means adequate strength and no separation of anchor points during loading.
(4) Definition of Similarity An acceptable definition of similarity needs to be developed.
The staff believes that for equipment to be similar for the purpose of qualiijing an equipment item on the basis of experience data on another item, the safety function as well as the dynamic characteristic, 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.
For active mechanical equipment, such as pumps and valves, where long-term operability is needed after an earthquake, the possibility of incipient damage to similar equipment in data base plants should be investigated.
This could be done by reviewing plant maintenance records supplemented by detailed inspection of the in-situ data base components as required.
(5) In-Situ Testing Comparisons of frequencies between non-nuclear and nuclear equipment in the pilot program were based on low-level in-situ testing.
The staff believes that verification of the validity of low-level testing is required to show that there are not significant differences in response between low level tests ai,d excitation levels associated with significant earthquakes.
The staff is concerned that differences in anchorages may significantly affect equipment responses. Verification of anchorage integrity and similarity must be addressed. Anchorages are addressed further in item 3 above.
(6) Equipment Operability l
The primary design criteria of nuclear plant safety equipment during seismic events are (1) to operate on demand and (2) to have no spurious status change.
The experience data implies that electrical switchgear and motor control center equipment loacted in non-nuclear plants opened during actual seismic events and required subsequent operator action to l
1 l
be re-closed.
EQE stated during a briefing to the SQUG and the NRC staff that data base plant equipment actuated spuriously during the seismic event.
The data base should be more explicit regarding equipment operation during a seismic event, i.e., did it operate as intended or did it actuate spuriously?
To use the experience data as a basis for qualification of electrical equipment in nuclear plants, the required functions must be compared and/or detailed information on the state of the data base equipment must be known for comparison with the required funct. ion of the nuclear plant equipment.
Even though many of the electrical equipment components in fossile fuel plants are similar to those in nuclear power plants, safety systems in nuclear plants are considerably more complex and less accessible due to radiation hazard.
Therefore, the reset of circuit breakers, and relays, and the sequencing of the safety systems may not be possible in a timely fashion.
Operability following an earthquake should also be considered.
For instance, equipment in the residual heat removal system must operate on a long-term basis following an earthquake event:
the data base should include records which verify that equipment that appeared to be functional immediately following the earthquake was not partially damaged or approaching a " damage" threshold.
(7) Degradation of Seismic Performance Due to Aging Electrical components that include age-sensitivie elements, such as motor insulation, are likely to degrade and influence seismic performance.
The conclusion drawn in the EQE report that there is no observed degradation of seismic performance related to aging is based on limited data.
More definitive data on aging, including available test results, should be reviewed to verify the conclusion.
(8) Margin and Fragility The question of seismic margin has been of concern for many years.
Qualitatively it has been answered by acknowledging that some margin exists in the design procedure, and by the use of an enveloping response spectrum in addition to the " inherent strength" of structures and equipment. 'The fragility of equipment cannot be determined unless the failure level is known. Very little actual fragility data exists.
Most fragility curves are based on estimated failures levels either from consensus of expert opinion or by extrapolation of test data.
Even through some failure tests have been conducted, the amount of test data available is too sparse to be statistically significant.
i -
For the purpose of seismic qualification of equipment in operating plants, the staf f believes that three steps are necessary to demonstrate acceptable margin.
First, at the frequency ranges of interest the required response spectra at the attachment point of an equipment item should be completely enveloped by a spectrum developed from actual experience data or acceptable test data or some combination of the two, second the current criteria of 10% exceedance for this envelopment should still be applicable, third, relative movement of attachment points and anchors should be considered as a failure mechanism where appropriate.
A factor of safety of 2 should be used in evaluating relative movement of attachment and anchors.
The frequency ranges of interest must be established for a particular item of equipment by some combination of testing and analysis with due consideration given to the uncertainty in the methods used to estimate these frequency ranges.
The staff suggests that where possible, SQUG in the course of their data search, record acceleration levels on each peice of equipment and where failures have occurred, record the failure mode for use in estimating fragilities.
In addition, the SQUG participants are encouraged to pool all available test data to augment the experience data being collected under the current program.
2.5 Development of Methods to Generate Generic Floor Response Spectra 2.5.1 Background In the current practice of seismic qualification of safety-related equipuipment (either by analysis or by testing), when the dynamic characteristics of a piece of equipment is known, the required input seismic loading to the equipment, or more exactly, the information i
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, this RRS is the same as the " floor response spectra."
In the case that this equipment or component is attached to an equipment supporting structure (such as a rack, a cabinet, etc.), floor response spectra usually is still the starting point of analysis whereby the RRS at the equipment or component attachment locations can be obtained.
Floor response spectra, therefore, is an essential element for the qualification of equipment in nuclear power plants.
To determine specific floor motion or equipment supporting structure motion which are applicable to the development of equipment or component RRS, an expensive and time consuming time history finite element analysis generally is required.
For many operating nuclear power plants, the information on floor response spectra may nrt have been s
developed according to the current requirements.
In other cases, the information is simply no longer available.
The objective of this task l
l was to develop a set of " generic floor response spectra" which can be utilized by the utilities for the purpose of qualifying equipment.
The task of developing generic floor response spectra was undertaken by Brookhaven National Laboratory (BNL).
The task now is essentially complete.
A draft report (Ref. 14) was issued in April 1983.
Following is a summary of this contractor report.
2.5.2 Summary of Wor'K Completed The development of a generic floor response spectra starts with the concept that there is a degree of boundedness to the structural responses.
The report 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 ona 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 2.5-1.
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 containment and the internal structure.
cigure 2.5-2 shows this PWR model.
A free-field earthquake response spectra from the El Centro earthquake was 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 normalized to a more realistic time history peak of 0.1 The excitation was applied through the soil and into the various g.structures to produce responses in equipment 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 cps to 36 cps.
This resulted in fundamental mode coupled natural frequencies as low as 0.86 cps and as high as 30 cps.
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.
Reduction 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. - - - _ _ _ _ _ _
EL.147_'-- 2J' __
,g n
N EL.129'-O' s !
52
. '8 e
s!O EL.108'- 6' q
'm I
'm N
EL.82'_Q'
'6
.g s
EL. 65'-9"_
D
,5
'd EL.42'- 6' 4
u
'O
'b N
EL.14'- 6' --
'3 e
si EL. O'-O' E
2
'O
'S N
EL.-26'- O' 1
Y Df LJ h
A al 8n
)//#
- /
I' V"'
- /# #/
Figure 2.5-1 Model 3.,
SHIELD STRUCTURE STEEL CONTAINMENT
_= - _ _ o g i
gjo____
'e
'o '
i i
'rn
'o N
N 2
12o
/
=
---o
'e
. o.
o~'
3 13" o
g' s
l'o N
4 14 o
_o s i g
1
'8 5
15"
,o/
-o
'b N
o6 16 "
l'o N
o7 17
/
i'o g
INTERNAL STRUCTURE o8 18 o
.f 21 o----
6'4" N
22o 6'-O' o9 19
/
'o 23"
'd 2CI-O' N
o10 20
= - - - -
24 o----
6'-6' t
- e
'A h25
_Q.:-23.6".L.
MODEL 4 fff/fr##/KU/
Figure 2.5-2 Model 4. _-
Figure 2.5-3 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 earthquake 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 (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 cps through greater than 50 cps, or from soft soil through solid rock, respectively.
Figure 2.5-4 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 7.2g at the top level for a 0.lg earthquake.
The peak was calculated to be 6.0g for 0.lg earthquake.
This was increased by 20 percent to 7.2g because only one earthquake time history was used for the horizontal specta.
As the shear wave velocity of the soil decreases (softer soil), the 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 percent 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 the generation of 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 conservatisms of the spectra generated increases if little seismic data, is available.
Generic spectra in the vertical direction were not developed in this program.
Due to the conservatism accumulated by this approach every step along the way, the NRC staff believes that a conservative vertical generic floor spectra can be reasonably estimated by taking two thirds of the values of generic floor spectra in the horizontal direction.
2.5.3 Staff Conclusions A Required Response Spectra (RRS) is needed whether analysis, test or experience data is 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 equipment attachment point can be generated by a variety of ways (see Section 2.3) from the floor response spectra. _ _ _ _ _ _ _ _ _ _ _
72 f 7.0 Sj
'/
INTERACTION
/
CURVE FREQUENCY A
2 6.0..
B 3
C 5
D 10 E
50 43.
m3
$_3.6-Q A
B E
O V a4 g..
.8
.4 i
8 3 k6 10 10 0 FREQUENCY. ( hie)
Fig. 2.5-3 Generic Floor Response Spectra.
3 0
2;0g 40g 6.0g a0g 50 63 63_12g TOP l
l l
10 INTERACTION 5
FREQUENCf 3
- - - -60 MIDDLE x
BOTTOM
_15g BOTTOM (b)
(a)
Fig. 2.5-4 Generic Peak Responses at Top, Middle and Bottom levels. _ _ _ _ _ _ _ _ -
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.
3.
REFERENCES 1.
" Correlation of Seismic Experience Data in Non-Nuclear Facilities with Seismic Equipment Qualification in Nuclear Plants (A-46)," by P. D. Smith and R. G. Dong.
LLNL draft report, November 1982.
2.
" Program for the Development of an Alternative Approach to Seismic Equipment Qualification," by P. I. Yanev, S. W. Swan, EQE report (2 Volumes) prepared for SQUG, September 1982.
3.
" Identification of Seismically Risk Sensitive Systems and Components in Nuclear Power Plants," by A. Azarm, J. Boccio, P. Farahzard, BNL draft report, November 1982 (revised April 1983).
4.
" Correlation of Methodologies for Seismic Qualification Tests of Nuclear Plant Equipment," by D. D. Kana and D. J. Pomerening, SWRI report no.
SWRI-6582-002, Contract NRC-04-81-185, Task 2 Summary Report, June 1, 1983.
5.
"The Use of In-Situ Procedures for Seismic Qualification in Currently Operating Plants," by S. Sadik and B. W. Dixon. INEL interim report, December 1982.
6.
" Preliminary Guidance and Acceptance Criteria for Application of Combined In-Situ and Analysis Procedures in Operating Plant Equipment Qualification," INEL preliminary report, April 1983.
7.
" Summary of Work Performed to Date on Qualification Cost Estimating Task," INEL preliminary report, April 1983.
8.
U. S. Nuclear Regulatory Commission Regulatory Guide 1.61, " Damping Values for Seismic Design of Nuclear Power Plants." October 1973.
9.
U. S. Nuclear Regulatory Commission Regulatory Guide 1.122, " Development of Floor Design Response Spectra for Seismic Design of Floor-Supported Equipment or Components," Revision 1, 1978.
10.
" Survey of Methods for Seismic Qualification of Nuclear Plant Equipment and Components," by D. D. Kana, J. C. Simonis, D. J. Pomerening.
SWRI report no. SWRI-6582-001-01, Contract NRC-04-81-185, Task 1 Summary Report Part I, October 1982.
l....
i 11.
' Seismic Safety Margins Research Program, Phase I Final Report-SMACS,"
by J. J. Johnson, G. L. Goudreau, S. E. Bumpus, O. R. Maslenikov, LLNL report, July 1981.
12.
" Operating Function Tests of the PWR Type RHR Pump for Engineered Safety System Under Simulated Strong Ground Excitation," by T. Uga, K. Shiraki, T. Homma, H. Inazuka, N. Nakagima, Japan Atomic Energy Research Institute, JAERI-M8354, August 1979.
13.
" Seismic and Dynamic Qualification of Safety Related Electrical and Mechanical Equipment in Operating Nuclear Power Plants," by J. Curreri, C. Costantino, M. Reich, BNL draf t report, April 1983.
14.
"A Study of the Effect of Aging on the Operating of Switching Devices,"
by S. P. Carfagno, G. H. Herberlein, Jr., IEEE Transactions on Power Apparatus and Systems, Volume PAS-99, No. 6, November / December 1980.
l APPENDIX A Related Topics Covered by the INEL Contractor's Report on In-Situ Testing Even though the contractor report is concerned mainly with how to utilize in-situ testing to assist in performing seismic qualification of equipment, other related topics were studied by the contractor. Among them are the following.
(1) Operability and Failure Modes:
In order.to develop methods to utilize experience data to qualify equipment, the contractor suggested that a systematic treatement of operability is necessary. The failure modes which result in inoperability, from the contractor's viewpoint, are an essential ingredient to these meti.ods. 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 prevent a component from performing an active operation or maintaining a state continuously.
Inoperability can result from:
inability to monitor the control condition inability to change states when so directed inability to maintain the current state when no state change is directed It is suggested by the contractor that inoperability during 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 state change or prevents a state change from occurring.
structural interfere 0ce - excessive relative motion results 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 A-1
l failure mode. The first piece has been qualification proof tested and its controlling design features are either identical or inherently more fragile than the equipment in question.
In that case qualifying the first amocnts to qualifying the other to the same environment. The procedure below is suggested by the contractor 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 than certain failure modes will be eliminateu and l
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 of design evaluation / comparison will be required in making this assessment.
Equipment used for comparison must be of known seismic capacity.
It is the staff's belief that in-situ testing will be a valuable tool to establish dynamic similarity between equipment through the comparison of the dynamic characteristics (mode shapes, natural frequencies, damping, size, shape, weight, etc.).
(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 impractical. An approach based on the interaction of aging with seismic capacity is adopted by the contractor.
Such an approach suggests that since some aging mechanisms will not affect seismic capacity these cases need not be considered in seismic qualification.
The use of in-situ testing in evaluating the effects of aging on seismic qualification has been considered by the contractor, however, no well developed technologias were identified. Consequently, aging has been examined in a broader context where:
The consequences of aging degradation are examined. This allows the relationship between dynamic qualification and aging degradation to be organized in a fashion which more clearly demonstrates the interaction.
A-2
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 have been identified as less vulnerable to aging.
The effect of aging on seismic capacity is illustrated in Figure A-1.
A systematic basis for evaluating aging degradation is provided by the failure mode analysis and the procedures emoodied in Figure A-1.
This method as proposed by the contractor is as follows.
First, a determination of any aging effccts produced by the design basis environments should be conducted.
This involves listing all vulnerable materials and examining environmental data for each.
Presently, such data is only available for some materials.
Those 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 failure modes defined earlier.
Each failure mode must be examined in light of the anticipated degradation.
If it cannot be established that no significant 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 shculd be qualified to current criteria.
Catetory II aging: This is any aging mechanism 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 (Ref. 14 ).
For a typical component many mechanisms would fall in this category.
(
A-3 l
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 paragraph (1) of this appendix.
The failure modes are used to establish similarity, and data from similar equipment is transferred to the equipment in question. The important factor is that much equipment will exhibit no significant seismic aging interaction of concern and thus screening can narrow the field effectively without overlooking substantial aging degradation.
A-4
Environmental Aging Yes Seismic Capacity Unaffected Dynamic Dynamic
Response
Response Affected Unaffected u
Operability Affected Non-structural Degradation by Dynam,ic Effecting Seismic Capacity
Response
N
/
Load Se.ismic Load Magnitudes, Environment Path Frequency Content a Contributor n
Operation During Structural Normal Integrity Environment Affected l
Structural Interference Not a Concern of Dynamic Qualification:In Province of Routine in-service Surveillance Reconfiguration Figure A-1 Effect of aging on seismic capacity A-5 l
N,:CPom M U.S. NUCLE A;) REGUL ATORY COMMISSION m e,i S!BLIOGRAPHIC DATA SHEET NUREG-1018
'~~*#
NeN$c $a'1Ne##"In'"'o7'I'qYipriM' in Operating Plants ati t
'"'"8^"'**' "" '
r lv d afety Issue A-46
- 7. AUTHOR (S)
- 5. DATE REPORT COMPLE TED I^"
T. Y. Chang September 1983
- 9. PERFORMING ORGANIZATION NAME AND MAILING ADORESS (laciude 20 Codel DATE REPORT ISSUED Division of Safety Technology l+1983 vO~ m Eaa Office of Nuclear Reactor Regulation September U.S. Nuclear Regulatory Comission 6 ft** e **aal Washington, DC 20555 8 (Leave Nanni
- 12. SPONSORING ORGANIZ ATION NAME AND M AILING ADDRESS (lactude l>p Codel Division of Safety Technology Office of Nuclear Reactor Regulation i, nn so.
U.S. Nuclear Regulatory Comission Washington, DC 20555
- 13. TYPE OF REPORT PE RIOD COV E RE D flectus,ve datest Technical Report
- 15. SUPPLEMEN TARY NOTES 14 (Leave o/e*l
- 16. ABSTR ACT (200 *ords or less)
This report sumarizes the status of work accomplished on USI A-46 by the U. S. Nuclear Regulatory Comission staff and its contractors, Idaho National Engineering Laboratory (INEL), Southwest Research Institute (SWRI), Brookhaven National Laboratory (BNL) and Lawrence Livermore National Laboratory (LLNL) that is applicable to USI A-46.
This assessment leads to the conclusion that the use of seismic experience data for equipment qualification provides the only 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 status of continuing efforts to establish requirements for an experience data base is provided in this report.
- 17. KEY WORDS AND DOCUMENT ANALYSIS 17e DESCRIPTORS Unresolved Safety Issue A-46 Seismic Qualification of Equipment in Operating Plants 17tt IDENTIFIERS OPEN ENDE D TERMS
+
Unlimited Availability 22 eRicE 2o 3I1DR TY MSS,Ta.s,,w 1
ass ed s
NEC FORM 335 111816
UNITED STATES
,,,,,c,,,,,,,,
NUCLEAR REGULATORY COMMISSION posta3s a etts pa o WASHINGTON, D.C. 20555
,jf,"jc PiAWif%e M OFFICIAL BUSINESS PENALTY FOR PRIVATE USE,5300 h{0 078877 1 1ANIAll1519X ADM-DIV OF TIDC hk0 Y& PUB MGT BR-POR NUREG l
WASHINGTON DC 20555 l
1