ML20199H265

Technical Evaluation Rept - Structural Design Issues Long- Term Svc Seismic Reevaluation,San Onofre Nuclear Power Generating Station 1

ML20199H265
Person / Time
Site:	San Onofre
Issue date:	05/31/1986
From:	Shieh L LAWRENCE LIVERMORE NATIONAL LABORATORY
To:	Office of Nuclear Reactor Regulation
Shared Package
ML13324A929	List: ML20199H233 ML20199H251 ML20199H265 ML20199H271
References
UCID-20769, NUDOCS 8607030244
Download: ML20199H265 (44)

Text

._ ._

, i .

Attachmer.t 3 -

UCID-20769 TECHNICAL EVALUATION REPCRT - STRUCTURAL DESIGN ISSUES LONG-TERM-SERVICE SEISMIC REEVALUATION SOUTHERN CALIFORNIA EDISON COMPANY SAN ONOFRE NUCLEAR POWER GENERATING STATION UNIT 1 Preparec by L.C. Shien, J.C. Chen, N.C. Tsai Lawrence Liver =cre National Laboratory Lawrence Liver = ore National Laboratory 7000 East Avenue Liver =cre, CA 93550 N.C. Tsai NOT Engineering, n:.

3650 Mt. Diablo 31vc.

?.0. Box 1937 Laf ayette, CA 94549 Prepared for Integrated Saf ety Assess =en; Progra:

Office of Nuclear Reac cr Regulation U. S. Nuclear Regulatory Cc==ission Was.*.ing::n, D. C. 20555 May , 1986 fN

$0' yVj40

$b g'lD a

,w- av ,- --4-y

e c.

= i TABLE OF CONTENTS

1. Introduction ..................................................... 1 1.7 P ur pos e an d Ba c k gr o un d . '. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Eva l uat i o n C r i t er i a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.3 References .................................................. 2

2. Evaluation of the R efueling Water Storage Tank . . . . . . . . . . . . . . . . . . . 3 2.1 Introduction ................................................ 3

2. 2 Discussion................................................... 3 2.2.1 Mo d eli n g of t he Tan k . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2.2.2 Modeling of the Soil Foundation ..................... 4 2.2.3 Analysis Method for, and Results of, the Soil-Structure Interacticn Anal. ................................... 4 2.3 T an k an d F o un da t i o n E val u a ti o ns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2.4 Confirmatory Analysis ,....................................... 6 2.5 Comparison of LLNL and Impell r es ponses r es ults . . . . . . . . . . . . . 6

~

2.6 Conclusions .................................................

2.~ References ..................................................

3 Eva'uation of Turbine Suilding Flocr-Response Spectra Generation ....................................................... 6 31 Introduction ................................................ 8 3.2 Discussion .................................................. 8 3.3 C o n c l u s i o ns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 3.4 References .................................................. 11

...,,,- a

_. . 2.. .,s

.. c.e . m.o.

.. e . , s.eu .e ,- .n. ,,

.. . < p , a . yna. :.a

.. . . . . .s , s.

Electrical Duct .................................................. 12 a.. i I n. ...n. s .w.

. C ~. <.w-. ,................................................. ,4-t.2 Discussion .................................................. *2

a. 3 Conclusions

~

................................................... 13

• .. .i e.,e."ences .................................................. 3 5 Evaluation of the ' lent Stack ..................................... 15

... t. .>.ns.,an.e..a,n................................................

. v o , a.

5.2 Discussion .................................................. 15 53 Conclusions ................................................. 'S

.. e_ v. ' ua . *. .a r. o .' C * ..a .a. ' *a=-= oor, '- c. -*.i..3 .D .i, a.'..e .cv=.=._s

.. .............. 1 .'

6.* Introduction ................................................ 16 i G. 9

_ Da. s .e . .s s

s . nve. . ..................................................

63 C o n c l u s i o ns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

,. oo..c...s.a.

o . a ...................................................... _-..

Appendix A ........................................................... 23 A.1.0 Introduction ................................................ 2h 111 e

,, - - - - , , . , , - ,,,m

< M i.,,~.~ - . =

. I e  %,

2 i I l

1 A.2.0 S tr uc t ur al- Fl ui d M o de ls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 4 "-' l A.3.0 calculation of Foundation Impedances . . . . . . . . . . . . . . . . . . . . . . . . 24 A.3 1 Soil Profile ................................................ 24 l A.3 2 F o und a t io n I m pe dan c e s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 )

A.4.0 Input Motion ................................................ 28 l A.5.0 Results..................................................... 32 ,

A.6.0 Comparison of LLNL and Impell response results . . . . . . . . . . . . . . 43 )

1 I

1 l

l O

iv

• s 3

f 1.0 Introduction 1.1 PURPOSE AND BACKGROUND This report provides technical evaluations of the Licensee's analysis of safety-related structures (including modeling techniques, methods of analysis, and results) with respect to the compliance with the Systematic Evaluation Program criteria, applicable industrial codes, and the applicable Nuclear Regulatory Co= mission (NRC) criteria. In the event that upgrading was necessary, the upgraded plan's methods and procedures were reviewed for acceptability.

The NRC determined that the modified Housner ground response spectra (a 10%

increase in the 0.07 to 0.2; second period range for the horizontal component of acceleration and a 10% increase in 0.05 to 0.15 second period range for the vertical component), anchored at 0.67g was the appropriate ground m0 tion for the seismic reevaluation cf the San Cnofre Nuclear Generating Statien Unit 1 (SONOS 1). This spectru: is eferrec Oc as One C.6'E 50dified-H0usner spectrum.

SCNGS 1 was shut down frc: early 19S2 to late 1960 Many structures , systems ,

and cc ;0nents were upgraded during this time by a program known as the Return-To-Service (RTS) program. The plant was then al10wed to resu=e operation, provided that the remainder of the seis=ic reevaluation program and the resulting plant modifications were completed prior to startup frcm tne current ref ueling cutage.

n a meeting with the NRC staff en February 12, 19S5 (Ref. 1.1}, ar.d thr0ugn a letter dated March 12,1985 (Ref.1.2) , the Licensee pr0p: sed their criteria and analysis methocciogy f 0r the Leng-Tern-Service *,LTS) upgrading. Further o meetings and sutrittals c'. arified and , in s ce cases, modified the Licensee's

rc;0sals.

Section 2 of this re;;rt presents an evaluation of the refueling water s Orage tan <, and Section 3 gives an evaluation of the turtine-tuilding ficer-res p:nse-spectrun generati:n. Section evaluates grade tes:s, electrical duct tanks, and the turbine tuilding scutn extensicr. Secti:n 5 evaluates ne vent Stack, Section 6 evaluates the steel-mencer design calcula 1:ns, Se0;ien

? presents our conclusions.

i . c. ...a 3. ., c 2. ,. . :. ~. . ..a

. . r.A.n. . _v .

SCNOS ' is one cf the NRC designated Systematic Eva'uation ?regra: 'SEPi _

plants , which were not designed to current codes and standards, and t0 current NRC licensing requirements. Therefore, the NRC has developed a set of criteria and guidelines for use in reviewing these plants. The following documents were used during the review of the SCNGS 1 structure design:

o. .,

f

a. NUREG/CR-0058, " Development of Criteria for Seismic Review of Selected Nuclear Power Plants," by N. M. Newmark and W. J. Hall, May,1978. )

b. "SEP Guidelines for Soil-Structure Interaction Review," by SEP Senior Seismic Review Team, December 8,1980.

c. Letter from W. Paulson, NRC, to R. Dietch, the Southern California Edison Company (SCE), " Systematic Evaluation Program Position Re:

Consideration of Inelastic Response Using NRC NUREG/CR-0098 Ductility Factor Approach," June 23, 1982.

d. Letter from W. Paulson, NRC, to R. Dietch, SCE, "SEP Topic III-6, Seismic Design Considerations, Staff Guidelines for Seismic Evaluation Criteria for the SEP Group II Plants-Revision I," September 20, 1982.

For cases not specifically covered by these criteria, the following Standard Review Plan (SRP) sections and Regulatory Guides were used:

a. Standard Review Plan, Sections 2.5, 3 7, 3.5, 3.9, and 3 10.

b. Regulatory Guides 1.26, 1.29, 1.60, 1.92, 1.100, and 1.122.

If the Licensee's proposed methodology and criteria deviated from these review criteria and guidelines, we reviewed and evaluated the justification presented by the Licensee, based on our experience and best engineering judg=ent. We understand that plant-specific deviations may be found acceptable on a case-Ly-case basis, as long as they reasonably meet the intent of the SEP review guidelines.

1.3 REFERENCES

1.1 Memerandum frcm E. McKenna to C. I. Grimes , dated Fe:ruary 12, 1965.

1.2 Letter frem Mark Medford, SCE, to J.' A. Zwolinski, NRC, dated March 12, 1965.

e- s. -

~

l 2.0 Evaluation of the Refueling Water Storage Tank 2.1 Introduction' The results of the Licensee's Long-Ter=-Service (LTS) reevaluation of the refueling water storage tank (RWST) are presented in Ref. 2.1. The results of the Licensee's calculatiens/ analyses were reviewed and audited during the following review meetings: December 10-12, 1985: January 7-9 and 29, and February 6 and 18-19,1986. To assist the assessment of the Licensee's methodology and seismic responses (including the soil-structure interaction (SSI) analysis of the tank) we perfor:ed an independent confir:atory analysis of the tank for the hori:Ontal excitation. The RWST reevaluation was required for the Safety Evaluation Report (SER) issued for the San Onofre Nuclear Generating Station Unit I (SCNGS-I) Long-Ter: Service (LTS) (R ef. 2.2, as Review Item 3 7). The findings and conclusions of our evaluation are summarized below.

2.2 Discussion Our review of the Licensee's LTS reevaluation of the RWST was perfer:ec in three =ajcr areas;

a. A modeling of the tank, including the s10shing effect and soil foundation

t. A soil-structure interaction (SS!) analysis, and O. A reevaluation of the tank anc foundation.

These areas are discussed separately in the following three subsections.

2. 2.1 Modeling of the Tank - For the heri Ontal analysis, the tank structure was representec Oy a nree-= ass stick :odel, and the contained water by two a00itional lumped 21sses. These twc = asses included a sicsning mass M, and a

~

risic = ass M r. Tney were determined by using H0usner's simplified the0ry

Ref. 2.3). We found the tan < an0 fiuit 200el fcr the hori

Ontal analysis :

ce acce; table.

Fcr the vertical analysis, the tank shell an: water were represented by One rigid mass. We found this vertica; 200el of the tank and water also to te acce ptable .

2.2.2 Modeling of the 5011 Foundation - The ::pe;; version Of the CLASS: 00 de .as used for the hori ental SSI analys *

• the RWST. We found the application cf the CLASS: ethodology to the heri SSI analysis of the RWST to be acceptable. This finding was docums ' in Ref . 2. 2. In the Licensee's  !

horizontal SSI analys'is, the foundation soil, including the backfill, was l assumed to be one uniform material. The shear modulus of 1390 kilopounds per 1

l l

• a

- i square foot (ksf) for the soil was consistent with that used in the SSI analysis of the reactor building reevaluation, which was acceptable (Ref. 2.4). The soil material damping is taken to be 8%, which is acceptable to the NRC per the SONGS-I LTS SER, Item 3.10.2.2 (Ref. 2.2) . We therefore found that the soil properties used in the horizontal SSI analysis of the RWST were acceptable.

For the Licensee's vertical SSI analysis, the soil impecance is represented by a constant vertical soil spring (determined in accord with Ref.

2.5). A 10% composite modal damping was assumed, although the actual computed value (per Ref. 2.1) exceeded 10%. The vertical SSI model is tnerefore a single degree-of-freedom system having a 10% damping and appropriate soil stiffness. In the vertical SSI analysis, additional conservatis: was introduced by taking the spectral peak acceleration from the ground response spectra as the vertical response of the tank. Therefore, we fcund the results of the vertical SSI analysis to be acceptable.

2.2.3 Analysis Me ned for, and Results of, the Soil-Structure Interacti0n Analysis The five-mass (including two water masses) stick =0 del, with CLASSi-cc:puted soil impedance, was used for the RWST horizontal SSI analysis. This analysis was perfer:ed by using Impell's version of the CLASSI code. The vertical seismic response was calculated by using a one-mass model with soil stiffness and a 10% system damping. This analysis was perfer ed by hand calculation.

The free-field codified Housner response spectrum for the ,0.67g seismic event was used as input for the horizontal SSI analysis; 0.45g was used for the vertical analysis.

The res;0nses of the tank were in-structure response spectra, the max *-" "ase

ents and shears, stresses, and everturning soments; all resulting fect hori:Ontal excitation. The in-structure peak accelerati0ns were 0.62g, C. 525, and 1.2? g at the base ,. rigid fluid mass location, ar.d r:0f of the tank, respectively. (The remaining results are discussec in secti ns 2.3.)

23 Tank and Foundation Evaluations - The tank and founcation were reevalua:e ty the _icensee, :ase: On One seismi 10 ads generate: fr:: :ne hori::ntal an: -

vertical analyses. The audit was performed for the following areas: tan <

snel', anchcrage, soil pressure, c0n: rete cat (cue 50 a postulated 1.5" seismically incace: settlemen; cf the backfill, and the seis:ic leads),

no::les and be110ws, sliding and Overturning stability, and tank re:f. They are discussed separately in the felicwing list.

a. Shell buckling was evaluate: by 00: paring the maximum 00:;ressive stress in tne tank shell with that allowable for buckling, ac00rding to the rules of ASME Code Case N-284 The maximum stress accounted for included the dead weight, seismic inertial loads, hydrestatic pressure, and hydrodynamic pressure. The resulting safety f actor agair.st buckling was greater than the l'.34 rece:: ended by the code for Level D events. We found the buckling evaluation acceptable.

.u- .

. _ _ _ _ _ . , _ _ _ .__._.-__._m -

_ . _ , . , , , _-my -- - . - , . y

. . ~ .

. 4 e

b. The maximum principal stress in'the tank shell was evaluated against the Level-D-event allowable of the ASME B&PV Code,Section III, Division I, Subarticle NC-3800. The maximum principal stress was determined by combining the raximum tensile hoop stress (due to the hydrostatic pressure, hydrodynamic pressure, and vertical seismic load); the maximum tensile longitudinal stress (due to overturning moment from the horizontal seismic analysis); and the maximum shear stress derived from the horizontal seismic analysis. The evaluation criteria and analysis procedures. appear adequate. The stress results are within the allowables.

c. The concrete mat foundstion was evaluated against the seismic loads.

In addition, a postulated 1.5-inch seismically induced settlement of the backfill soil was assumed beneath the northern and western portion of the tank base. The evaluation was done using vertical soil springs to represent the soil. The effects of soil settlement on the base =at were evaluated by using an equivalent bea odel for the baserat. For the settlement evaluation, the mat was modeled as teing supper ed by an elastic foundation. A unifer: load, equal to the dead weight Of the tank and fluid, was applied to the basemat. The stiffness of tne equivalent soil springs under one-half of the tank was reduced until j the maximum displacement in the :odel was equal to the predicted j displace:ent of 1.5-inches. A static analysis was then performed.

Another analysis was then performed, which considered the mat as being supported by a uniform elastic foundation, and lended oy the seismic loads and dead weight. The stresses in the concrete basemat fec: the two ar.alyses were then ec tined by the algebraic su: =ethod, and evaluated against the ACI-349 Ocde. The evaluation results a;; ear adequate. :n addition, the maxi =u: soil pressure, as generated fro:

the seccnd analysis, was well within the alicwable.

d. The ancher bolts were evaluated for tensile and shearing f ailures against tne rules of the ASME Code, Appendix F, for Level D events.

The anchor belts were also evaluated agair.st pullout failure free the concrete. We found the ancher tolt evaluatier. to be acceptatie.

e. Lccal stresses in the shell at the nc::le locations were evaluated against the allowables frc: the ASME 35?V Code, Section !!I, Division I, Summer 19c3 Adder.da. The evaluation accounted for piping icads and the effect of tne pcstulated 1.5-inch settlement of the backfill soil. We fcund the nc::le-load evaluation for the shell to be j adequate.-

f. The sliding stability of the tank and concrete mat was evaluated, based on a 0.59 sliding friction coefficient, because if the seis:ic base shear exceeded the friction resistance (nor=al force times friction coefficient), the concrete mat would slide. To determine the maximum distance of a slide, the Licensee performed a nonlinear time- l l

i 1

• e history analysis of the tank using the ANSYS code. The tank was represented by a simplified one-mass stick model. The horizontal and vertical ground motions were simultaneously considered in the analysis, which showed that the maximum sliding distance is on the order of one-half inch. We found the sliding evaluation acceptable, as long as the bellows on the piping attached to the tank can withstand the effect of the tank's sliding, as discussed in Item h.

g. The overturning stability of the tank was evaluated by comparing the kinetic energy induced by the seismic motions to the potential energy required for the tank to tip over at an edge of the concrete mat (Ref.

2.5). The energy comparison showed a safety factor of about ten aginst overturning. We found the overturning evaluation to be s uf fici ent .

h. Bellows and No :les - There are four bellows connected to the piping systems near the lower part of the tank. Two of them (G-50A and G-503, respectively) are located between the tank and the safety injection pump on the two fourteen-inch saf ety injection lines. Cne (G-27) is located between the tank and the refueling water st: rage i'

pump on the eight-inch miscellaneous water line. The fourth one is on the four-inch drain line. These bellows and the corresponding no::les were evaluated against the seismic loads and soil settlement. The acceptance criteria for the bellows were based on the Standards of the Expansion Joint Manufacturers Association, Inc. The acceptance criteria for the no::les were based on ASME B&PV Code,Section III, Division I, Summer 1983 Addenda. Bellows on the t- and

• 6"-inch lines are being replaced with newly designed ones in order to acco==odate the predicted soil settlement and sliding displacement. We found the bellows evaluation and local stresses in the shell at the no::la loactions to be acceptable.

i. Tank r00f - The licensee did not evaluate the tank roof, based on their belief that a roof failure will not result in a less of function of the refueling water storage tank. This was found to De acceptable.

2.0 CONF 1RMATORY ANALYS13 -

To assist in assessing the Licensee's cetnodology and seis ic response of tne soil-structure interaction effects, we perfor:ed an independent confir:atcry analysis of ne tank, in the heri:onal direction. The vertical respenses Of the tank are =uch s= aller than th0se in the horizontal cirection, and were estimated conservatively by hand ca10ulation. Theref ore, we consideret the confirmatory analysis performed in the heri:Ontal direction adequate f:r assessing the methocology and moceling technique usec by the Licensee.

The tank was modeled using 3-di=ensional bea: ele =ents of the SA?4 cocputer code (Ref. 2.6). Masses were lumped along the axis of the tank at the ,

appropriate heights, with two = asses representing the contained fluid. The dynamic fluid model was developed using the analysis procedures of Housner 6-

1 r

8 p

l l

l (Ref. 2 3). The mass of the fluid is divided into two parts in this procedure: mass associated with the first sloshing mode of the fluid (convective mass), and mass associated with the ground motion --(rigid or i impulsive) mass. The rigid mass of the fluid was lumped with the tank shell at the calculated height, according to the procedure. At the same ti=e, the convective mass was connected to the shell with a spring, so that the vibrational frequency of this mass-spring system equaled the sloshing frequency predicted by the Housner procedure.

We calculated the foundation impedance functions of the the tank by using two methods: the CLASSI approach (frequency-dependent impedance), Ref. 2 3, and the constant-soil-spring method, Ref. 2.2. For both methods, a uniform elastic half-space medium was assumed for the soil under the tank.

The constant soil-spring methodology is acceptable to the NRC, and has been used as a reference for evaluating other SSI methodologies.

The dynami: 50dal pecperties of the fixed-base structure were calculatec by using the SA?u code. For the constant-spring approach, the cc:;0 site =odal j damping values were calculated by using the COMDAM? code. The LLNL versi0ns l

of the 00 puter codes, CLASSI and RESPNS, were used to generate structural l

responses for the CLASSI and constant-soil-spring approaches, respectively.

j CLASSI is a c0=puter code for sicultaneously analyzing the soil-structure l effects and 00 puting structural responses.

With the CLASSI approach, the peak accelerations were calculated as being

' 225, 0.53g, 0.62g at the roef, the rigid fluid = ass, and the base, respectivel). For the Constant-spring method, the peak accelerati0ns were calculated as being 1 39g, 0.91g, and 0.o5g, at the corresponcing locations, respectively. The maximum difference between these two metheds is at0ut 141. ( Appendix A descr10es the confirmatory analysis.) Theref 0re , the CLASS:

approacn is acceptatie.

2.5 COMPARISCN OF LLNL AND IMPELL RESPCNSE RESULTS In order t0 assess the accepta ility of the Licensee's methecology anc res.ults for generating the response Of tne tank, we cc pared responses generated c:. a 00: parable =eth0d: the CLASS! appr0ach. As a result, we saw cifferences only between LLSL's and 1:pell's results in the second place af ter the decimal point. We therefore concluded that the Licensee's methodo10gy and the results f 0r the structural res;cnse of the refueling water s:Orage tan < are acce ptable .

4.o --,,..L.-.....ouvos

.va Our evaluation found the LTS reevaluation of the RWST to te 1

acceptable.

2. 7 REFERENCES

-7 1

1

e a -

2.1 Impell Report, San Onofre Nuclear Generating Station, Unit 1, Evaluation of the Refueling Water Storage Tank for Long Term Service, Impell Corporation, Walnut Creek, California, Impell Report No. OL-0310-1392, Revision 2,1986 (two volumes). Transmitted by letter dated 3/31/86.

i 2.2 Safety Evaluation Report by the Office of Nuclear Reactor Regulation, Long Term Service Plan - SEP Seismic Reevaluation, Criteria and Methodology, San Onofre Nuclear Generating Station Unit No.1.,

Docket No. 50-206, September 19, 1985.

2.3 Nuclear Reactors and Earthquakes, TID-7024, August 1983: " Chapter 6 -

Dynamic Pressure on Fluid Containers." +

2.4 0. R. Maslenikov, et al., "SMACS - A System of Computer Programs for s Probabilistic Seismic Analysis of Structures and Subsystems",

j Lawrence Livermore National Laboratory, Livermore, California, UCID i 20t13, Vol. I, Maren 1985.

l 2.5 Tocical Report: Seismic Analyses of Structures and Ecuicser.t for i

Nuciear Pcwer Plants, Topical Report No. 30-TO?-4-A, Rev. 3, Sect.tel l Power Corp., Novester 1974 2.6 S. J. Sackett, User's Manual for SAP 4 (A Modified and Extended Version of the U. C. Berkeley SAPIV Code), Lawrence Liverscre National Laboratory, Livermore, California, UCID-18226 (1979).

3.0 Evaluation cf Turbine Building Flocr-Rescense-Scectra Generation 31 INTRODUCTION The ficer response spectra generated by.the Licensee for the turbine building (T/3) were audited during the review meetings held en July 1-2 anc December 10-12, 1985. This audit was required by the Safety Evaluation Report (SER), Section 3 10.3 (see Ref. 3 14 The Licensee also pr:vided a report er

,Se;tember 24, 1955 (Ref. 3 2) .

3 2 D:SCUSSION

~l The structural analysis model, including scil foundation f'exibility for the T/B, is basec on a secel previously developed by Bechtel. A t hr e e-dimensional finite-element structural model was used to represent the T/3 structure. The 3echtel SS: model uses frequency-indepencent soil springs, i:. accordance with the Licensee's proposed methodology for an SSI analysis of the T/B during the Systematic Evaluation Program (SEP) phase of the seismic reevaluation (Ref. 3.3). This approach was considered acceptable by NRC staff during the RTS evaluation (Ref. 3.5). For the given model,.

the Licensee previously proposed to generate the floor response spectra -

for the Long-Term-Service (LTS) phase using the direct generation technique implemented in Impell's FLORA computer code. The FLORA wide-8-

l i

i 1

i y--.wwy- --

• a I

band solution approach has been accepted by the NRC (Ref. 31). However, the Licensee changed the methodology to use a modified time-history-analysis technique in lieu of the direct generation technique.

For the modified time-history-analysis technique, a time-history modal analysis of the T/B model was performed with the I=pell code EDSGAP, to generate the floor response spectra at required locations. The floor spectra from the time-history analyses were then =ultiplied by certain

" correction factors" to obtain modified floor spectra, to account for the difference between the input synthetic time-history spectra and the modified Housner spectra. The correction factor varies with the structural frequency. At each frequency, the correction factor represents the ratio of the floor spectrum generated from the input (a synthetic time history) to the corresponding floor spectrum directly generated from the

=odified Housner spectra. The direct generation 1: based on the FLORA methodology, but the actual computation of the correction factors was done with another computer code, FACTOR, a special version of FLORA developed by Impell. The modified floor spectrum so generated excludes the interaction between Ope piping systems and the T/3 structure, and is equivaler.; to using a ':ero value for the =cdal interaction mass, 214, when the flocr spectrum is directly generated using the FLORA code. The Licensee indicated during the review =eeting that they cay consider using a non-zero mik wnerever the piping system and structure interacticn effect is deemed significant enougn to warrant using it.

The modified time-history-analysis technique just described appears acceptable. H0 wever, we were concerned that the. correct 10n f actor may be sensitive to whether the FLCRA narrow-band or wide-band solution is actually used. We alSO felt it necessary to review the magnitudes Of the correction fact as at representative locations in :ne T/3, and to review the ordinates of the floor spectra at these same locations those that would have been directly generated frcm the wide-band soluti.cn of FLCRA, basec on either the =cdified Housner or the synthetic-tire-history spectra being input. In addition, noting that the current floor spectra for the LTS typically are lower than the ones for the SEP, we felt that it was' necessary to qualitatively c0= pare the current uncorrected ficer spectra to the corresponding spectra previously generated by Be:htel (during the SEP phase of tne seis=ic reevaluation), particularly wnere a large discrepancy ccourred between the two. C nsecuently, as a result of tne July 1-2,1935 audit review, we requested the Licensee to provide the folicwing aceitional information:

a. C0rrebtion fact 0rs generated from both the narrew- and wide-band sclutiens for the 2% damping spectra at these 100ations ir the T/5:

I D ir'ection Elevation Location Nede Numbers )

i

)

N-S 42' Area 2 Deck (A-53) 580, 586, 611 '

Vertical 35.5' Area 6 Deck (A-60) 29, 71 , 86 l

l 1

x

+ 's.

E-W 35'5'

. Area 6 Deck ( A-61) 29, 71, 86

b. The 25-damping, floor-response spectrum was obtained by enveloping the three floor-response spectra at the corresponding three locations specified in Item a above, which were directly generated using the FLORA wide-band solution and based on either the modified Housner or synthetic time-history ground spectrum. For example, the floor response spectrum at the area-two deck (A-53) at an elevation of 42' is the envelope of three spectra (at nodes 580, 586, and

' 611 ) . Contributions frem the three earthquake components were to be combined with the SRSS technique _for each of the three specified locations .

c. A comparison of the 2%-damping Bechtel-design floor spectra to the uncorrected floor spectra, as generated from the Impell time-histcry-analysis of the SSI :odel, at the three locations specified in Ite: a.

The Licensee provided Item c in Ref. 3 2, and Items a and b during the audit of December 10-12, 1985. The information provided appears sufficient to resolve our concerns on the acceptability of the T/3 LTS floor spectrum.

3 3 CONCLUSIONS Based on the outcome of the two audit review meetings and the additional information subsequently provided by the Licensee, we conclude that:

a. The structural =cdeling ethodology appears acceptable because it is consistent with that used previously in the SE? phase of the reevaluation.

b. The ccrrection f acter approach for generating the modified (corrected) ficer response spectrum fec: the time-history-generated spectru= appears acceptable fer tne case of =gg - C.

i (c) The LTS floor spectra are qualitatively censistent with these l previously generated by Bechtel during the SE? phase and, hence,  ;

j appear adequate.

3.4 REFERENCES

3.1 Safety Evaluation Report by the Office of Nuclear Reacter ,

Regulation, Long Term Service Plan - SIP Seismic Reevaluatien, Criteria and Methodology, San Onofre Nuclear Generating Station, l Unit No.1, Docket No. 50-206, Septe:ter 15, 1985. i 3.2 " SONGS-1 Responses to NRC Request for Information, Turbine Building Response Spectra," Encicsure three to letter frc: M. D. I Medford, SCE, to J. A. Zwolinski, NRC, September 24, 1985.

3 3 " Balance of Plant Structures Seismic Reevaluation Program, Turbine Building and Turbine-Generator Pedestal, San Onofre Nuclear Cenerating Station, Unit 1," Enclosure two to letter from K. P.

Baskin, SCE, to D. M. Crutchfield, NRC, April 30, 1982.

3.4 " Balance of Plant Structures Seismic Reevaluation Criteria, San Onofre Nuclear Generating Station, Unit 1," Enclosure to letter from K. P. Baskin, SCE, to D. M. Crutchfield, NRC, February 23, 1921.

3.5 Safety Evaluation Report, by the Office of Nuclear Reactor Regulation, Return to Service, Criteria and Methodology San Onofre Nuclear Generating Station, Unit No.1, Docket 50-20e, Plan November 21, 1984 Ii w -y,-

0 e.

4.0 Evaluations of the Grade Beam Design, and the Buried Electrical Duct Bank, and the As-Built Reevaluation of the Turbine Building South Extension.

4.1 Introduction The Licensee's results with regard to the design of grade beams for the Auxiliary Feedwater Pu=p foundations, the reevaluation of the buried electrical duct banks, and the as-built reevaluation'of the Turbine Building south extension are presented in Refs. 4.1 and 4.2. The results were reviewed by LLNL and its subconsultant, and an audit, was conducted at the Norwalk Office of the Bechtel Power Corporation on December 13, 1985. Our findings from reviewing Refs. 4.1 and 4.2, and frc: the audit =eeting, are summarized in Sections 4.2 and 4.3 4.2 DISCUSSION Grade Beam for Auxiliary Feedwater Pu=o Foundation - Two new grade beams were designec anc constructed in June, 1964 to ad:ress the ef fect of seismically induced settlement of the backfill soil supporting the Auxiliary Feedwater Pump foundations. Each of the two new grade bears are in turn supp;rted on new concrete piers at the ends. The design is based on the assumption that the backfill soil does not provide any support to the pu=p foundations, and that the new grade tea:s and concrete piers provide the necessary support.

The two ends were assumed to be a pin pin connection: a conservative ass u=pti on. The analysis of the grade bea:s took into account all applicable dead loads, seismic loads, jet impingement loads, and pipe reaction loads during n r:al 0;eration cr in the shutdown condition. Hand calculatior.s were used for the analysis. The design is based on ACI 318-1977. Both the analysis and the design appear to te adequate.

3uried Electrical Duct Sanks - The buried electrical duct tanks were re-evaluated te address the effect of a seis=ically induced settle snt of the backfil; soil. In the reevaluation, the duct tanks were assumed to have no support frc: the backfill where they traversed the backfill areas. With the backfill soil discounted, the duct banks were analyzed as beams having simply supported end conditions at the native soil /bacxfill interface, er of having fixed end ccnditions where the duct tan <s are ente:ded in con: rete. :n determining the =oment capacities of One cuc; tanks, the a:becde conduits were taken as being reinforcements in aedition to the actual reinf:rce:ents in the teams , if any. The =odeling and =ethod 10gy for analysis appear :: ce acceptatie. The evaluation criterion is AC: 3*S-77. During the audit, we examined the design of the South and North Ducts, and the East-cf-Pump-Well Duct (calculations LSC-CC 2.0, p.p. 7/60 and 5/60, 4/2/ 93; LSC-CC 2.0, p.

12/60, 5/6/S3 ; and L30-CC 2.0, p. 4;/60, t/1/ 53, respe::ively) . The reevaluation results appear to be adequate.

As-Built Reevaluation of Turbine Building South Extension - The as-built

=odifications to the Turbine Building (T/S) South Extension are identified in Ref. 1. The evaluation is based on Ref. 4.3 The analysis is based on a finite-ele =ent structural =odel consisting of the South Extension, the turbine 4

- f.u.nc.:.m.

- ,._ _._- ~

~~~

generator and pedestal, and the gantry crane. Structural damping is assumed to be 75 and soil material damping,135. Soil springs were used to represent the soil-structure interaction effect, and a maximum composite modal damping of 20% was used whenever the computed value exceeded 205. Apparently, the soil =aterial damping, used for the evaluation of the South Extension, exceeds the 8% value specified in the SONGS 1 LTS Saf ety Evaluation Report (Ref.

4.4). However, soil radiation damping is usually much higher than soil material da= ping. In addition, the licensee already conservatively limited the composite modal damping, which is a combination of structural damping and soil material and soil radiction damping values, to 20%. Based on the above mentioned reasoning and our sound engineering judgement, we concluded that the 13% of the soil material da= ping for the evaluation of the South Extension is acceptable. The BSAP computer code was used for the analysis. The analysis model and methodology appear to be adequate. Reference 4.2 provides the results of Licensee's additional evaluation of the structure for the effect of the shif ting crane weight to the intact leg, and for the effect of i= pact when the uplif ted leg is lowered onto the support rail. The addition evaluation indicates a safety =argin o'r 1.63, which appears sufficient.

0 3 00NCLUS10NS Eased on our review of the infermation presented in Refs. 4.1 and 4.2, l and the out:0:e of the audit, our conclusions are as fellows:

a. The design of the two new grade beams, and their supporting concrete piers, are sufficient to address the concern about the seistically induced settling of tne backfill soil.

. The reevaluation of the four turied electrical duct banks appears adequate to adcress the concern at0ut the seistically inducet tackfill soil stttle ent.

O. The reevaluation 7f tne T/3 South Extension including the gnatry crane appears sufficient.

.: REFERENCES 4.1 Letter frc: M. O. Medford of SCE to J. A. Zwclinski of One L'SNRC ,

dated September 24, 1935.

L. 2 Letter frc: M. 3. Medford of SCE to G. E. Lear cf the USSRC, dated March 27,1986. ,

i L.3 Safety Evaluatien Report by the Office of Nuclear Reseter Regulation, Return- -Service Flan, Criteria and Metncd:1:gy, San Onofre Nuclear Generating Station, Unit No.1, Decket 50-206, 1 Nove=ber 21, 1984 u.4 Safety Evaluation Report by the Office of Nuclear Reactor Regulation, Long-Ter:-Service, Criteria and Methodology, San Oncfre l

1

• e ,

Nuclear Generating Station, Unit No.1, Docket 50-206, September 18, 1985.

9 l

• . s.

5.0 Evaluation of the Vent Stack

5.1 INTRODUCTION

During the December 10-12, 1985 review meeting held at Impell, we reviewed the evaluation results for the vent stack. The review was conducted by auditing the Calculation No. DC-1663 of SCE, dated 7/20/84 Our findings are su=marized in Sections 5.2 and 5.3 5.2 DISCUSSION The stack is constructed of A36 steel plate. It is tapered from 4.5 feet in diameter at the top to 8 feet in diameter at the base, which is anchored with ASTM A193 anchor bolts (Grade B7,yF = 105 ksi) to a 20 foot-diameter octagonal concreta foundation. The vent stack was evaluated for 0.67g modified Housner earthquake loads. The analysis model consisted of an 11-= ass stick model run on the SAPIV code. A fixed-base analysis was made, which was s uf f icient , j udging by the low first-mode frequency.

F0r the seismic evaluation, the allowables are 1.6 times the AISO coce allowables. The seismically induced stresses at the base of the stack are within the allowables by a large margin. We reviewed the stress in the 06 one-inch diameter anchor bolts, which are pre-tensioned to 10 Ksi during i nstallation. The anchor bolts have a large margin of safety agair.st seismic loads. We also audited the duct-opening stress condition, the stability of the vent stack against sliding and overturning, the maximum soil bearing pressure, and the anchor bolt pull-out capacity. Sufficient safety margins exist in all of these aucitec areas.

33 CONCLUS:CNS The vent-stack analysis for seismic loading indicated that there is a large saf ety margin agair.st the al10wables. Our audit f 0un'd the vent s ,ac< to be syfficient for the seismic condition.

l

o, n. -

6.0 EVALUATION OF STEEL BEAMS SUPPORTING PIPING SYSTEMS

6.1 INTRODUCTION

According to the preliminary results of the LTS seismic reevaluation, the Licensee identified twenty eight of the steel beams supporting the safety-related piping systems as having exceeded the AISC-code elastic limits. They included two beams in the reactor building and twenty six in the turbine building, as listed in Table 6.1. Of these twenty eight steel beams, seven in the turbine building were shown to be within the elastic limits, as based on the final reevaluation results. The remaining twenty one beams have been (or are being) upgraded to meet the LTS elastic limit criteria, as identified in Table 1 We conducted an on-site inspection of the steel teams on April 14, 1986 and followed up with an audit at Impell's offices in Walnut Creek, California on April 16 - 18, 1986. Our findings are sur arized in Sections 6.2 and 6.3 6.2 D:SCUSSION Site inspection - The site visit was to inspect the as-tuilt conditiens , anc the status of the modification design with regard to meeting the intencec purpose of the =odification. We also inspected most of the steel teams that had been upgraded prior to the LTS, as listed in Table 6.2.

Most of the modifications were accomplished using any one or a combination of the following types of modification design:

a. Reinforcing the teams for loading in the major axis by adding cover plates at the top and /cr bottom flanges, ad:ing web stif fener plates ,

or adding a structural T-section to the bott0: fl an ge .

t. Reinforcing the beams for leading in this minor axes by adding lateral tracings, and for loading causing torsional stresses by adding torsion-resisting assemblies.

c. Reinforcing the eccent-resisting capacity at the enc connections of the tea:s with additional welding.

In addition to upgrading the steel teams, five new columns , fcur of which are laterally traced, are teing installed to provide additional structural support to the steel tears of the north extension me::anine. The 50cification designs appear to be reasonable for serving their intended f unction.

Audit - The audit evaluated the moce'.ing technique, meth0c of analysis, anc whether the calculated stress met the LTS elastic li=it criterien, fcr these beams with (or without) modification. This criterion is that the stresses, in both the beam mester and end connections, induced by the LTS 1 cads must stay within 1.6 times the allowable stresses that are specified in the AISC code, Part 1.

~3 A

a. Beams Requiring No Modification - For the seven steel beams that were qualified as elastic, and hence required no modification, we audited the Impell calculations for the beam member qualification, and the Bechtel calculations for the end connection qualification.

For the bea: member qualifications, hand calculations were used for the analysis. All beams were considered as one-span, except for Beam No.

EMP-B22 which was analyzed as a two-span beam. The beam ends were assumed to be simply supported, except for Beam Nos. NE-B4.4 and NE-B4.8, for which one end was assumed to be fixed (because of the moment connection design). The stresses from the building seismic loads were combined with those from the pipe support seismic loads using the square-root-of-the-su=-of-the-squares (SRSS) rule. The total seismic stresses were co=bined with the static load stresses using the absolute sum ( ABS) rule. The stEess induced by the pipe support seismic loads frc: the same (or different) piping systems were combined using the ABS rule. The safety factor, defined here as the margin of the LTS elastic limit allowe over the LTS load-induced maximum stress in the me::er, is listed in Table 6.1 for the seven beams in questi0n. The results indicated very narrow margins.

For the end-connection qualificaticns, according to the i=pell calculations, the end reactions from the LTS loads are smaller than the corresponding reactions previously used by Bechtel in their design. We audited the Bechtel design calculations in Calculations Nos. IPTC-CC-03 2 to CC-03.h, which showed that the existing end connections were within the elastic li=it under the LTS loads.

b. Beams Requiring Modification - On a samplir.g basis, out of the twenty-one , steel tears requiring =0dification, we audited the codified desigr.s of sixteen of the tes:s cor.tained in 3e0htel's Calculation M003--00.

Hand calculation was used in the analysis. The analytical method was the same as that applied to the seven bears requiring no

= edification. Primard '. "..e modification designs appear acceptable for the intended funct. n. The modifica; ion designs primarily used one (or a 00 tination) of the three =ethods previously described.

6.3 CON LUS10NS i Based on the results of the on-site inspecti0n and the subsequer.: sudit review, we found that the LTS reevaluation cf the twenty eight steel beams ,

(listed in Tatle 6.1) appear to be reasonable. Mcwever, the extensive l

odifications of the Turbine Building. North Extension, including tr.e  !

me::anine, may substantially change the LTS seismic loads in the steel be ams . In view of the very narrow margins in the seven beams not currently l requiring modification, and the possible increase in the structural frequencies, the Licensee was required to evaluate the impact of the North-Extension structural modifications on the piping and on the building LTS l

1 i

.e,

~

ME**;_

.. ___.~

e, seismic loads in the steel beams. Therefore, the acceptability of the steel-beam evaluation for the Turbine Building North-Extension, is contingent upon the outcome of the north extension reevaluation of the final modified configuration. For the steel beams in the Reactor Building, the Turbine-Building East-Heater Platform, and the Turbine-Building West Heater Platform, the modification is much less extensive and the current evaluation / modification appears to be acceptable.

e

_ _ _ _ . ~_. _ . _ _ - . _ . . . . _

M =-

o TABLE 6.1 Twenty-eight former inelastic beams.

Location Beam No. Reevaluation Acceptance Status RTS-B11 modified yes Reactor Building RTS-B18 modified yes NE-B4 3 modified yes**

NE-B4.4 elas tic (FS=1.05)

• yes**

T/B North Extension NE-B4.7 modified yes**

NE-34.8 elastic (FS=1.04) yes

NE-35.2 modified yes

T/B North Extension NEM-B2.9 modified yes**

NEM-34 modified yes**

NEM-32.10 modified yes**

NEM-32.11 modified yes**

Me::anine NEM-B2.2 modified yes**

NEM-32.4 modified yes**

NEM-B2.5 modified yes**

NEM-B2.8 modified yes**

NEM-35 :odified yes**

KEM-B6 modified yes**

EHP-35 elastic (FS=1.01 yes EHP-36.2 elas ti c (FS=1. 7) yes EHP-37 modified yes T/B East Heater EH?-32 4 elas ti c (FS= 1. 2 2 yes

?latfor: EHP-32 :odified ses EHP-34 _: 0dified yes EHP-322 elas tic (FS=1.01 ) yes EM?-33 modified yes T/B West Heater WHP-St.1 e' as tic (FS-1.00)

. yes

?latf0r WHP-323.1 modified yes WHP-36 modified yes

• FS= saf ety f a:ter of =aximu LTS stress in bean ce::er agair.st 1.6 A:S-Part I alicwatle.

• • Acceptance is contingent upon the outcome of the reevaluation of the codified configuration.

O o.

TABLE 6.2 Additional Steel Beams Inspected During Site Visit Location Beam Number RTS-B4 RTS-38 ~

W18x96 (NEAR RTS-B19)

Reacter Building RTS-B19 RTS-B57 R_TS-B57 RTS-B72 NE-B5.1 T/B North Extension NE-B5.4 NEM-B1.1 NEM-B1.2 NEM-B2.3 NEM-32.6 T/B North Extension NEM-B2.7 Me::anine NEM-B2.12 NEM-B3.1 NEM-B3.1 NEM-B100 NEM-B105 T/B West Heater WHP-35 Platfer WP-37 T/B T3-3103 E-5

?

9 A -

7.0 CONCLUSION

S A detailed review was performed'io provide technical evaluations of the structural upgrading design, analysis, and load generation at the San Onofre Nuclear Generation Power Station, Unit 1. The structures reviewed in this report include the refueling water storage tank, turbine building, buried grade beams, electrical duct banks, turbine building south extension, vent stack, and steel members. Reviews of the Licensee's criteria, methodologies, and results, along with additional information provided by the Licensee led to the following conclusions,

a. Refueling Water Storage Tank:

o For the tank and soil models, the CLASSI method of analysis for soil-i structure for interaction effects appears to be acceptable, o The evaluation of the tank shell, concrete mat, anchor bolts, bellows, and no::les appear to be adequate,

b. Turbine Building Floor Restonse Soectrum Generation:

o The SS: modeling methodology appears acceptable, o The correction-factor approach for generating the modified flocr-response-spectra from the time history generated spectra appears acceptable,

c. Grade Beams inside tf.e Turbine Building:

o The evaluation of tr.e buried teams appears to be acceptatie,

d. Electrical Duct 3anks:

o The evaluation of the electrical duct banks appears tc te acceptable,

e. As-Built Turbine Building Scuth Extension:

o The evaluation of One Turtine Building South Extensicr. appears tc te acceptable,

f. Vent Stack:

o The evaluation of the vent stack appears to be acceptable,

g. Secondary Steel 3eams Suppcrting Piping Systems:

o The evaluation of the secondary steel beams appears to be acceptable.

However, due to the extensive modification of the Turbine Building North Extension (including the mezzanine) and the very narrow margins in the seven teams not currently requiring modification, the Licensee 21-i e- w yw. --- * ,,---

.- -,-a. g---- - -my v - w

O h*

was required to evaluate the impact of the modification on both the

. piping and the building seismic loads in the beams.

1. e e

.. . - ~

O +

i f APPENDIX A l

4 CONFIRMATORY SOIL-STRUCTURE INTERACTION ANALYSIS d

i 1 0F THE REFUELING WATER STORAGE TANK i

i

-l FOR i

I LONG-TERM SERVICE l

4

. SAN ONOFRE NUCLEAR GENERATING STATION, UNIT 1 1

i j

.g i

i I

i 4

l 4

4 i

l .

1 t

k 3

I 4

m. p-+ - - -- s im e ,,-%-eey w--.-~4 -y-w---se--wea 9 - eme t eens a. gp %m=.w-+- - - - - -aywm,%-,q,e ~y-w----+ y v y w g e s,- y,- e v -- w,e w w- e w c r -79e r~y-N+W " e' t

6 S A.

1.0 INTRODUCTION

This appendix provides the results of an independent soil / structure / fluid interaction analysis of the refueling water storage tank (RWST), using Licensee supplied input motions for the San Onofre Nuclear Generating Station (SCNGS), Unit 1, Long-Term-Service (LTS) seismic reevaluation. It describes the methodology, the development of the structured and fluid models, the input motions, and the results for the independent analysis. Finally, a comparison between the independent results and the Licensee's results for the responses is presented.

Two methods for modeling the soil / structure interaction were used. The first one is the CLASSI =ethodology, which was also the one adopted by the Licensee. The second is the constant i=pedance methodology. Both have been used as our technical evaluation basis in a previous Technical Evaluation Report for the SONGS-1 LTS program (Ref. 2.2).

A.2.0 STRUCTURAL-FLUID MODELS The SONGS-1 refueling water storage tank (RWST) is a surf ace-founded cylindrical steel shell with a conical roof (Fig.1). Of the base =at's 35.5 foot dia:eter, 40% sits on native San Mateo sand, and 60% on the shallow (up to 8-foot depth) backfill soil (Fig. 2). The tank is anchcred with 321-5/8 inch diameter anchor bolts e: bedded in the concrete base at to a steel base ring, stiffening ring, and stiffening plates which are welded to the tank shell. There are four piping lines with bellcws connection to the 10wer part of the tank. The tank was assu=ed to be full of water.

The tank was modeled using 3D tes: ele:ents of the SA?u co:puter code

(Ref. 2.6, Section 2.7) . Mass was lumped along the axis of the tank at the appropriate height.

The dyna:ic fluid :odel was developed using Housner's analytical procedures (Ref. 2 3 of Section 2.7). The mass of the fluid is first divided into two parts: a mass associated with the first sloshing mode of the fluid -- a convective mass; and a mass ass 0ciated with the ground =0 tion -- rigid (or impulsive ) = ass. FolicWing Housner's proceedure, the rigid cass of tne fluid was lumped with the tank shell at the calculated height. The ccnvective cass is connected to the shell with a spring, so that the frequency of vitration of this = ass / spring system tecc es equal to the frequency precicted by the procedure. This fixed-base tank / fluid =athematical =odel is sh0wn in Fig. 3 A.3.0 CALCULATICN CF FOUNDATICN IM?IDANCES A.3 1 Scil Profile The soil at the SCNGS-1 site is the unifor:ly dense San Mateo sand extending to about 1000 feet below site grade. However, 005 of the soil (up to depth of 8 feet) under the RWST is backfilled with San Mateo sand (at a relative compaction of 925), with the re=aining 40% of the foundation being on I

i

{

o- r l

l t

i I

i i j

t i

I shell 1

445" e

% e 0.25" I

i t -

I i W

e- C.329" 88" a ,,

l 1 -

-, ii l . T U baseplate * "

, gg.,

a 408" >

i Figure 1 The refueling water storage tanx.

, . ~

y

, f . .

I ~

maess% f M

i:

\L

\*

h* W W A4 \ p t

1 l

-t I 11, ,,m aes- I _

I -

p ' Backfill soil B. *t4J 4.

W W 4EEEE e w s%

-Fesung h 3 m- m w- em t fee C

3 * "

aw% e i

g L.

z

.w : - Backfill soil een w Settlement = 11"

&cCAL Sort CONDITIC4 t,weta RE7WEUhG DATER ITOMCI TMK Figure 2. Local soil conditions under the refueling water storage tank.

1 i

i

o . 900F MASS 2m .

• 2

$ " 16' (32.6 lb-SEC )

in i -

l > 15 (791') q) 34 SIDSHLNG FRSS "r C Q 12 13 K=2754 lb

, iii

()ll

$ 10

, (4223)

)9 RIGID FLUID FRSS D, h Os A 9* 7 e, .

h 06 S 3, '>5 m '

34 ,

' 7

~

g .h 93 -

$l2 7 7 Y  ? 1

//////// /<,/ = Y (E-W) l l

X Figure 3 Mathematical model of the refueling wate.a sto." age tank.

l l

b e native soil. The difference of properties, e.g., shear modulus, at the strain level of 0.67g earthquake, .is small, as indicated in Ref. 2.1. Therefore, a uniform elastic half-space medium was assumed for the analysis.

The soil properties used for the analysis were:

Shear modulus (kip /ft) 1203 Shear wave velocity (ft/sec) 578 Material damping (5) 8.0 Poisson's ratio 0.35 Weight density (kip /f t) 0.116 The average dynamic soil properties are compatible with shear strains at the site induced by the 0.67g design motion at the ground surf ace. These soil properties were used by the Licensee for their SSI Analysis.

A.3 2 Foundation Impedances The diameter of the tank circular foundation is 35.5 feet Cand 2 feet and a-1/4 inches in thickness.] We calculated impedance f unctions for the ref ueling water stcrage tank by using two =eth0ds: the CLASSI approach (Ref. 2.u) and the cor.stant impedance method (Ref. 2.2) .

( CLASSI is a computer code for si=ultaneously analyzing the soil-structure effects and cc=puting structural responses. This code uses a three-step substructure approach -- a determination of the foundation input motion, and then the foundation impedance, follcwed by an anlysis of the coupled soil-structure system. Since the control otion is specified directly at the foundation, foundation input otion does not need to be determined. The foundatior. in;edances are calculated by using a continuu: cetnod. In geneaal, for a linear elastic or viscoelastic material and a unif:r: cr h:rizontal;y stratified soil deposit, each element of the impedance matrix is 00 plex-

. valued ar.d frequency dependent. The real part of the matrix represents the stiffness of the soil and the imaginary part represents the damping. The final step in the substructure approach is the actual SS analysis. The foundation input notion and the foundation i=;edances are cc bined with a cynami: =cde; Of the structures f r the soil-structure system. The respanse analysis is Onen performed in the frequency d:: sin. Fourier transfcr te n.iques are applied to cttain the tics history Of One respor.se.

Figures 5 througn 7 sh0w the impedance calculations by tne CLASS: approach, wnile the impedances calculated by using the constant spring method are 1.0-E5 Kip /f t for translational stiffness, 2.76E7 kip-f t/ radian for roc <ing stiffness , 2.07E3 xip-sec/ft for translational damping, and 2.12E5 Kip-f t-sec/ rad for rocking damping, respectively.

A.4.0 INPUT MOTION

+w a - --. a m - m 4_ x__ __- - __ -- _ _ _ , _ _ _ _ - _

0

• I 1

rett rette z-oia(t-ef j EL .

1C -

fikb]$i.s,.0;h..AAe>,,0.,v,,s.,,,,t.)

sC}4va ip p i.e r'un mp vcu vg s ;

-1C -

t l- i l 4

-2C '- '

1 C 2 4 6 8 1C 12 14 16 18 ftWC ($t 2 0$}

I i

i 1

i 1

4 Figu: e ha. The Input motion time history supplied by Impell.

1 l

1 9

g m--w-- , - -w y--,--,e- ~ "+-

r -,a

%9 py,. --.mwvb--.y.,w*-wr w+-

, ,m ,,wy w -

.g-.y7  ;- yw.p

4 a

45 -

4 40 -

m M

% M -

ar.:

\1 v

v <

PS -

>=

4 '

W b d M-W \ .

U U

< 1 W 2e -

I 3

J g 1$ -

I 1C b I

J i

s-C' *

-1 C 1 g 10 10

• to gg FREQUENCY (H:)

Figure Ub. The SONG 3-1 LTS design response spectru: in the east-west direction, with 5% damping.

b

~ ;&BO*d.** % '~

m.

2 stirrucss cocrricio<r x( 1.1) owius coctrictorr c( i. )

1 1

}

O

. o i 1

1 i -3 ,L -

t l

,\

(

I i i  ;

i

-2 ~  !

-2 E+08 C S @ E+08 0 o o m

o m

)

FWQDC (>Z)

NOTE: K=L3/E C=(LB-SEC/FT)/ frequency Figure 5. The horizontal translational impedances K(1,1) and C(1,1).

, . . - - - ,.n..

, , , . . , .,. - , , , - . . -- ,n.-

O m The SONGS-1 artificial time history in the east-west direction, which was generated by the Impell Corporation of Walnut Craek, California, was used for this soil / structure / fluid interaction analysis. The response spectrum corresponding to the time history envelopes the norizontal 0.67g modified Housner response spectrum. This time history is shown in Fig. 4a. The corresponding design response spectrum with 5% damping is shown in Fig. 4b.

The control motion was applied in the free-field at the ground surface level.

A.5.0 RESULTS The dynamic modal properties of the fixed-base structure / fluid combination were calculated by using SAP 4 code. The results are shown in Table 1. The second and third modes represent the structure's frequencies and the first mode represents the sloshing fluid's frequency. Note that all the results presented here are for the east-west direction only, since the geccetry of the tank is axisymmetric and the analysis was performed parallel to the base.

Recall that we used two different methods to calculate the four.dation impedances: frequency-dependent impedance (by the CLASS 1 approach) and frequency-independent impedance 'by the ocnstant soil-stiffness method}. In the following paragraph, we present the results fro: the frequency-dependent i impedance first, then the results from the frequency-independent i=pedance calculation. As a point of interest, the results of :odal analysis are then presented.

Figure 8 shows that, based on the CLASSI approach the in-structure response spectru: at the foundetion level is generally lower than that in the free-field. Figure 9 snows the peak accelerations along the structure elevation.

It can be seen frc Figure 9 that the accelerations decrease slightly at the icwer part of the structure, then increase all the way up to the tank rocf.

Figure 10 shows the peak accelerations along the structure elevation, based on the frequency-independent i pedance calculation. Again, the ac:elerat10r.s decrease at the lower part of the structure, then increase all the way up ::

the tank roof.

16 crder to explain the structural response tehavior of the soil / structure / fluid system, we 00 puted the 200al preperties f 0r ne case of I constant impedances. Tatle 2 st:ws the system frequencies and their participation factors. The frequency of tne first mode re=ains 0.2)~ hert:.

This reflects that the first :Cde's being the 10 cal =ote representing the slosning fluid. Table 3 consists of the ec=pesite =odal camping, ranging frc:

0.51 to 76.4%. Table 4 presents the ecce shapes ter the corresponding first three modes. The first code represents the 10 cal =0de Of the sicsning mass, while the second and third 50 des are the structural =0 des. Based On the i participating factors and frequencies, it is obvious the second mode dO=inates the tank response. This verifies the nature of the profile of the maximum accelerator of the tank along the height as described above, both for our confirmatory and the Licensee's analyses, because the acceleration profiles are similar to the shape of the second mode shape.

j  ;

i

y O e, l

stirrwtss cetrricion x( 4, 4) owiwo cotrricson c( 4, 4) 4 4>

4 -

2 -

1 l'

i

, 2; i' 1  !

i i

0 0

_1 .

l  !

-2 r -

t h

-2 r' + t i

)  : -4 e

. E+10 I

-3 C+10 0 C C C C n m C C C n m ratt3.!D<r (H2) FWC3.! Der (H2)

0TE

K=LE-FT/RC C= (L3-FT-SEC/RC)/frecuency Figure 6. The rocking impedances K(4.4) and C( 4,4),

~

stirrscss cocrricson x( t, s) me No coctricton c( i, s) 3 . . 3 2 2 i i 1- -

1r -

i 1

0 0 f

( -

t

-1 t y

/ j -1

, )

i .

N / l '

i

/  !

s

-2 -  ;

i l i 1

, j e 1 t

j l

-3 -3 ' '

r.q33 o e o o E+08 C O O O l

- - a m - m m I rRCQt,JtNCY (4) ramm (g)

.97E: K=G/RG C=(B-SEC/RO)/ frequency Figure 7. Translation / Rocking ccupling i:::pedar.0es K(1,5) and C(1,5).

l l

l

ga -

f;?f) ( -

e /

[ f ,

(

N 9

b 35 -

i f

l f --FFIEIELD w so -

I ift -

t b

w V

8" y\

~

.--FOUNDATION '

< \ .

w N gn -

8 E

" is -

to -

s -

1

' ' 3 I

0 to 10 10 w l

FREQUENCY 't l

Figure 8. The foundation-level and free-field in-structee .aesponse spectra.

s.uc.cc ~_^

~

e PEAX ACalATIm ROOF E S (g)

" gg

& 1.2:

1 > 15 15 13 m

- mel, s A

SICSIL'C

}%SS II ll 9' 10 e,

O RIGID o ,3 ,

O 8 FwTD IRSS d

~,

0 -

1)6

i

4. d

-p*"

C 4, 0.56 I

o4

.. b se il .e

=

. 0:

1 1 I T 0 s; _ Y , s-.

usii,,,...i X

Figtr'e 9. The Peak acceleration response based on frequency-deper. dent approach (CLASSI) i ;edance.

1 1

1 1

PEAK n 16 E*F HMIS 6

1.30

] ) 15 i s 9 14 13

~4 3 2)

' 12 j S:AEDC HMis .

O tt pJ 10 L

4

-,9 RIGID 0.);

08 FWD MS5 ,

h N

m 11 7 a6 T

j pp5 c,e;

1 64 Z

, i. -

< >3 h i at ,

02 1 1 I T 1

/////ooe ooss ^ t*: )

_Y l

I Figure 10. Peak acceleration response, based on tne frequency-indeper. dent impedance approacn (constant spring).

s i T 4

_,I _ _ -

Ccmparing the responses these two different impedance calculations, indicates that the general response trend is consistent, even though the responses from the frequency-independent approach are higher than those derived by the frequency-dependent method.

A.6.0 COMPARISON Of LLNL AND IMPELL RESPONSE RESULTS Impell's response results (Ref. 2.2) are shown in Fig.11. By comparing LLNL's CLASSI and Impell's response results, we see the second decimal place. Accordingly, we conclude that the results of the soil-structure-fluid analysis performed by Impell are acceptable.

I 0

0 1

l

i

~"

r (ft') ' Men Acceleration (9) 40 -

ROOF 14 O MA55 (1.21) 0$HING 30 -

mss 13 0--@ 15 (1.02) (0.571) 0 -

RIGID 12 ge -FLUID MIS I0'02) d

% . SHELL t=0.25' to -

11 O (0.59) h 5 HILL t=0.323'

< 10 0 -

m 7- (0.62)

EG 5%X EMl3G Figu: e 11 Impell's peak acceleration response, based on the frequency-dependent i=pedance approach.

Modes Frequency (Hz) 1 0.297 2 6.9 3 50.5 Table 1. Significant modes of the fixed base-structural-fluid model.

Modes Frequencies (Hz) Participation Factor 1 0.29" 3.1 2 6.9 7.6 3 50.5 2.t Tatie 2. Significant modes of the soil-stru ;ure-fluid interaction system

( constant impedar.ce ) .

MODES DAMPING 1 0.005 .

2 0.*19 3 0.764 Table 3 Ccaposite modal da= ping of the interaction System (constant impedance).

l i

l

I

. .s c Moda Shape Node Mode 1 Mode 2 Mode 3 1 1.0431e-04 4.2984e-02 2.7673e-01) BASEMENT 2 3.0350e-05 1.0248e-02 1.6230e-03 3 6.2372e-05 2.0969e-02 2.6895e-03 4 9.5918e-05 3. 2106e-02 3.1783e-03 5 1.3097e-04 4.3666e-02 3.0702e-03 6 2.1282e-04 7.0572e-02 2.0852e-03 7 2.9849e-04 9.8331e-02 - 1. 4014 e-03 8 3.8771e-04 1.2692e-01 -7.4801e-03 9 4.15 93 e-04 1 3589e-01 -9.8835e-03} RIGID MASS 10 4.7620e-Ou 1.4628e-01 -5.5379e-02 11 5.3521e-04 1.5638e-01 -1.0031e-01 12 5.9425e-ou 1.6642e-01 -1.4520e-01 13 3.2 n5 8e-01 -9. 9611 e-0 4 4.0045e-05) SLOSHING MASS 14 6 3652e-ca 1.7679e-01 -1.9088e-01 15 6.'7'4e-02 1.867he-01 - 2. 3 377 e-01 16 7.1575e-04 1.96228-01 -2.7257e-01} ROCF MASS Table 4 Mcde shapes of the interaction syster (constant impedance }.