ML20043A692
| ML20043A692 | |
| Person / Time | |
|---|---|
| Site: | Catawba  |
| Issue date: | 05/17/1990 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20043A689 | List: |
| References | |
| NUDOCS 9005220407 | |
| Download: ML20043A692 (4) | |
Text
,
9; 4#
h
- UNITED $TATES' v.1 NUCLEAR REGULATORY COMMISSION E
WAsHINoTON, D; C. 20665
\\....+/
i SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION DIRECT GENERATION METHOD OF GENERATING FLOOR RESPONSE SPECTRA FOR SNUBBER REDUCTION PROGRAM DUKE POWER COMPANY, ET AL.'
j CATAWBA NUCLEAR STATION, UNITS 1 AND 2 DOCKET N05. 50-413 AND 50-414-t I.
INTRODUCTION
. By Reference 1, the licensee informed the.NRC of-its: intent to use: Direct..
Generation Method (DGM) of generating-floor response spectra for reducing-the snubber supports for safety?related piping systems.
It.was the-licensee's contention that'because of.the significant margin (i.ei, 30%=to.-
. 40%).between the base input response; spectrum and the spectrum' generated
' using the synthetic time histories (Reference.15), it is adequate to ' remove ;
the additional conservatism-in the reassessment of various snubber supports in most of the safety related piping systems.
1 Revision 1 (July 1981) of the Standard Review Plan:(SRP) Section 3.7.2
~
includes methods other than'tne. time-history method for generating floor response spectra..but requires the staff to assure the conservatism 11n the proposed approaches.
The draft Revision 2 (July 1988) of the SRP'Section explicitly allowed the use of DGM for generating floor response-spectra provided the adequate theoretical basis and equivalency with'the time-history method are demonstrated.
Revision 2 of'the SRP Section^was.
published in its' final form in August 1989,' and it includes the DGM'as
(
an acceptable method.
II.
EVALUATION
't The staff had two basic concerns with the licensee's' proposed method'for generating floor response spectra:
(1) use of_an: appropriate input ground response spectrum, and (2) the adequacy of the ' software,. " Equipment-Dynamic Analysis Package" (EDASP) for generating floor response spectra.
Those?
concerns and their resolutions are discussed in the following paragraphs.
.l 1
l.
n G
pp q
.p p
+
e.
b]
w.
- - Input Ground Response-Spectra The licensing basis' Design Response Spectra (DRS) for Catawba' Nuclear Station (CNS)= utilizes a set of-Newmark spectra anchored at 0.1Sg peak ground acceleration for the design of structures. _For the design of equipment, piping and other components, the licensee used_four-synthetic tirne-histories.
The average spectrum (at!S% damping) of.these time-histories enveloped the DRS with a margin of 30% to_40%.
-In Supplement I to NUREG-0954 (Reference 16) Section.2.5.2.3,:the staff pointed out that a set of site-specific spectra (SSS) based on Wolf Creek and Perry site analyses would provide better representation,of the. site characteristics for ground motion, at.CNS.
A" comparison'of-the SSS_and the DRS at 5% damping indicates that between the frequencies of.
s 3.and.1DHz, the SSS exceeds the DRS by 15% to 16%.
Initially, the licensee' proposed to use DRS for generattag the floor response spectra.
However, after extensive discussions (References 1 to 14), the licensee proposed a conservative-composite of the DRS and-SSS.
The licensee also developed a Power' Spectral Density Function (PSDF)?
corresponding to the Composite Response. Spectrum and demonstrated.that the' PSDF exceeds the minimum required PSDF of_ Appendix A of.SRP 3.7.1~(Rev.-2).
Thus the use of the Composite Response Spectra to develop the floor
. response spectra utilizing an appropriate DGM-is acceptable-to the staff.
Equipment Dynamic Analysis Packaae (EDASP)
EDASP is an interactive computer aid in the dynamic: analysis'of structures and equipment.
Given the model properties of' a structure, it is capable of developing the floor response spectra directly-from.a set of base-(or ground) response spectra without resorting.to the development of the enveloping time-histories.
The licensee and its consultant' Stevenson_& Associates, provided-the theoretical bases for the algorithm' utilized in the package and provided the following comparisons to justify the use of EDASP:
(1) Floor-Response Spectra generated using a time history '(TH) (enveloping Regulatory Guide (RG) 1.60 spectrum),and those generated directly from the response spectrum using EDASP.
TH spectra-vs._EDASP-spectra.
(2) Floor Response Spectra generated using an actual-tiine history (ATH)
(earthquake record) (whose high frequency peak.is centered at -20 Hz) and the corresponding response spectrum at 5% damping =using E0 ASP.
ATH spectra vs. EDASP spectra.
4
____i______-_a
. lp A
i i
y Response spectra were computed at 0.5% and 5% damping for six floor elevations in.the Catawba Reactor Building.
The first comparison ~ indicated that..in general, the. floor spectra computed. using EDASP were' centered around the TH spectra in the frequency range of 0.2 Hz to 5.0 Hz, and exceeded the TH spectra at'high frequencies-The'second comparison-indicated.that
~
with 0.5% damping, the'EDASP spectra exceeded the;ATH-spectra at.all-
+
frequencies. 'At 5% damping,;the EDASP' spectra either equalled or'were' slightly below the ATH spectra. values-in low' frequency range (4 10Hz).
At high frequencies the EDASP-spectra generally exceeded the ATH spectra.:
1 The licensee also' submitted comparisons of the floor response, spectra l
for Perry Containment.model subjected to the January 1986 Leroy_ Earthquake.
The comparisons consisted of the following:
~
~
(1) -Floor-response spectra from the base-mat response spectra in al1~
three directions utilizing the proposed EDASP procedure.
(2) Floor response spectra from the; recorded time-history'at the base mat, s
(3) Actual recorded floor response spectra.
The EDASP generated floor spectra (in three directions) compared quite; closely with those from the base mat time-history; with the EDASP procedure resulting in slightly higher peaks.
However, comparisons'of both.(1)_and (2) with the actual-' recorded floor response spectra'were'not_ conclusive.
For example,lin the North-South direction,;the recorded floor _ spectrum was consistently higher than those from-(1) and'(2).
In the East-West direction, the recorded peaks were lower than those in (1) and'(2),'however, the recorded spectrum crossed the computed spectra four times between15 Hz to' L
30 Hz.
In tne vertical' direction, the recorded. spectrum'showed better consistency with'the computed spectra..The licensee's1 consultant attributed such deviations between the computed and the recorded spectra to..the modelling inadequacy and assumpti_ons in parameters.and maintained that the' deviations should be amply compensated.by the use;of conservative : seismic design criteria..The comparison ofjthe EDASP generated floor' spectra:with the time-history generated floor spectraLwas quite consistent.~..The staff accepts 1the licensee's basic conclusion on the comparison.
III.
Alternate Dampina Values
^
Il D ference 1, the licensee submitted an' update of FSAR Section 3.7.1.3,
=
%.1tical Damping Values," in which damping values for piping: analysis based.on ASME Code Case N-411 were provided as alternative values to those listed previously.
These damping values'were used in calculating;the floor response spectra based on the DGM.
In Reference >3, the licensee stated:that these values would be used subject to the restrictions of i
RG 1.84, Revision 24.
The staff finds that the use of damping values in accordance with.ASME Code. Case'N-411, as supplemented by RG 1.84, Revision 24, is acceptable.
i t
!- }
l
^*
- c. _
C J
IV.
Loadino Combination for Supports ~
-j In Reference 1, the licensee submitted an update;of FSAR Table 3.9.3-11
" Loading Conditions, Load Combinations and Allowable Stresses for Supports,_ Restraints'and Anchors, Duke Classes A,- B, C and F."
For those
?
supports falling within the ASME Section III NF jurisdictional' boundary,-
the-load combination corresponding to the faulted condition ~did not l
. include piping thermal. loads.
In Reference 5,-the li.censee committed L
to include these loads in,the~ snubber reduction program for both NF.
l and non-NF' supports when using variable. damped spectra based on the L
DGM.
These loads weresincluded in the results of the piping analysis problems which were-submitted in References 5.and 9.
~The staff finds that the licensee's commitment to include piping thermal.
l loads for both NF and non-NF' supports:is_ acceptable.
I L
CONCLUSION
. Based on the review'of the licensee's supplied information, pertinent FSAR l
sections, adequacy of the ground-motion input discussed in NUREG-0954 and
)
n the responses to-the staff requests for additional information, the staff.
l i
concludes that the proposed DGM for generating floor response' spectra using the licensee developed composite ground response spectratis acceptable for the planned Catawba; specific ~ snubber _ reduction program.
The details of the computed spectratto be used for the snubber reduction:
program should be available for any future audit by the staff.
REFERENCES l
1.
Duke Power Company's (DPC's) submittal d'ated' February' 24, 1988 2.
NRC's request for additional information (RAI) dated May 24,'1988 3.
DPC's submittal dated July 5, 1988 4.
Meeting summary dated July _ 28, 1988 5.
DPC's submittal dated September 16, 1988 6.
DPC's submittal dated.0ctober 27. 1988
-7.
DPC's submittal dated November 1, 1988-8.
NRC's RAI-dated November.-4, 1988 9.
DPC's submittal dated November 21, 1988
~,
10.
Meeting summary dated December 13, 1988 j
11.
DPC's' submittal dated May 10, 1989-
~
12.
NRC's RAI dated August 29, 1989 13.
DPC's submittal dated October 17, 1989 14.
DPC's-submittal' dated December 8,'1989 15.
Catawba's Final Safety Analysis Report.-(FSAR) Sections"3.7.1 and 3.7.2 16.
Supplement 1 to NUREG-0954 dated April 1983
.