ML20080H540

Responses to Request for Info from BNL Re Oyster Creek Nuclear Generating Station:Mark I Containment Long-Term Program Plant-Unique Analysis Repts

ML20080H540
Person / Time
Site:	Oyster Creek
Issue date:	08/31/1983
From:	MPR ASSOCIATES, INC.
To:
Shared Package
ML19277D762	List: ML20080H519 ML20080H527 ML20080H540
References
NUDOCS 8309210394
Download: ML20080H540 (49)

Text

r __

'* *i MPR ASSOCIATES. INC.

. RESPONSES TO A REQUEST FOR INFORMATION FROM THE BROOKHAVEN NATIONAL LABORATORY CONCERNING THE OYSTER CREEK NUCLEAR GENERATING STATION MARK I CONTAINMENT LONG-TERM PROGRAM PLANT-dNIQUE ANALYSIS REPORT 3 l

Prepared for:

l General Public utilities Nuclear Parsippany, New Jersey August 1983 l

l l

8309210394 830914 PDR ADOCK 05000219 i P PDR WASHINGTON. D.C. 20036 202 659 2320 1C50 CONNECTICUT AvtNut. N.W.

. s .

s M P R ASSOCIATES. INC.

I TABLE OF CONTENTS

1.0 INTRODUCTION

2.0 RESPONSES 3.0 . REFERENCES 4.0 APPENDIX A Brookhaven National Laboratory Request for Information (Forwarded by NRC letter L505-83-04-030 to Oyster Creek dated April 14,1983)

A l

O l

I l

1.0 INTRODUCTION

This document provides responses to a request for more information about '

the Oyster Creek Nuclear Generating Station Mark I Containment Long-Term Program Reports submitted to the U.S. Nuclear Regulatory Commission (NRC) in August 1982 (References 3.2 and 3.3). This request was

. prepared by the Brookhaven National Laboratory as a contractor to the NRC assigned to review the plant-unique analysis reports (PUAR) of the ,

Mark I Program (Reference 3.1). These responses were discussed with NRC and Brookhaven personnel in a meeting at MPR Associates on July 14, 1983. A copy of the Brookhaven request, as transmitted to GPU Nuclear, is contained in Appendix A of this document.

e

/

2.0 RESPONSES TO REQUEST FOR INFORMATION ,

This section contains the responses for each of the 12 items included in the Erockhaven request for information.

4 e

1 t

e I

sP i

~

N.

~

e J

, --,.r- ,,--y . - , ~,.-.---.---,4,, . . , -m . - - , ,., -, v.-,y,, . . _ _.m., e---, , , ,,,,, - , . . - - . ,,-, . _ _ _ , . - , , - - , .-,.-,....w--ev.,- - - . -

Recuest:

Item 1:. Section 2.1 of the Acceptance Criteria (Appendix A of Reference 3.1) states that " as part of the PUA, each licensee shall specify procedures (including the primary system parameters monitored) by which the operator will identify the SBA, to assure manual operation of the ADS within the specified time period. Longer time periods may be assumed for the SBA in any specific PUA, provided (1) the chugging load duration is correspondingly increased, (2) the procedures to assure manual operation, within the assumed time period are specified, and (3) the potential for thermal stratification and asymmetry effects are addressed in the PUA."

The PUAR does not specifically address the above requirement. Clarification is needed.

References:

Section 2.1 of Appendix A of Reference 3.1

Response

A generic agreement has been reached between the Mark I Owners Group and the U.S. Nuclear Regulatory Commission (NRC) to eliminate the need for discussion of these procedures in the plant-unique analysis reports (PUAR). The development of these procedures is now the responsibility of the Energency Procedures Committee of the BWR Owners Group and their work is being reviewed by the Procedures and Test Review Pranch of the NRC. This agreement is documented in an October 16, 1981 letter from the NRC to General Electric (Reference 3.4). .

l l

l l

l l 2-2 1

Recuest:

Item 2: At the vent line/drywell intersection of the Oyster Creek Plant, the vent line is reduced in diameter. Some thrust loads can be expected to occur at this reduction. No rention of a thrust load at this location is made in Section 4.4.1 of the PUAR (Reference 3.2) or the PULD (Reference 3.5). Section 6.2.2 of the PUAR does not make it completely clear whether the loadings applied to the finite element analysis of the intersection accounted for a possible thrust load. Clarification is needed on this point.

References:

Section 2.2 of Appendix A of Reference 3.1 Section 4.4.1 and 6.2.2 of Reference 3.2

_ Response:

At the Oyster Creek vent line/drywell intersection, an internal pressure load produces a net axial force on the vent line as a result of the reduction in diameter of the vent line just adjacent to the drywell.

This net axial force is not mentioned as a net thrust load in Section 4.4.1 of the PUAR or in the PULD because the methods for determining thrust loads in Section 4.2 of the GE Load Definition Report (Reference 3.6) and Section 2.2 of Appendix A of NUREG-0661 (Reference 3.1) do not include this location in the scope of " thrust loads". However, this net

^

axial force was accounted for in the Oyster Creek structural analysis.

This was accomplished using the finite element model of the vent line/drywell intersection shown in Figure 6.2.2-2 of the PUAR.

One of the unit load cases analyzed using this finite element model was internal pressure. In this analysis the pressure was applied to all internal surfaces of the vent line and drywell. This pressure acting on the section of vent line with the diameter reduction produced a net axial force on the vent line. We confirmed that the results of this analysis showed that the difference between the net force carried by the vent line on the drywell side of the diameter reduction and the net 2-3

~

force carried by the vent line on the vent. system side of the diameter reduction was equal to P x A, where P is the pressure and A is the projected area of the diarneter reducing transition piece (i.e., in annular area).

e e

9 i

s O

l

[

i i

L 2-4 l

l

Request:

Iten 3: The LDR and NUREG-0661 specify a minimum of four tests as a data base for obtaining net torus vertical loads. The PUAR states that Oyster Creek will operate with Dop between drywell and wetwell and this plant condition was the ene selected for calculating net torus vertical loads. It appears that only two tests at Oop with the same subnergence were conducted for Oyster Creek (Test 5 of NEDE-21944P and Test 10 of NEDE 24615P). The PUAR states that "Olp . loads were increased to account for the larger statistical variance associated with the smaller number of tests at the Gap conditions." Further justi-fication and clarification for this excepticn from the acceptance criteria is needed: How many tests are the loads based on? What is the statistical basis for the load increase and by what amount were they increased?

The above comments apply not only to the net torus vertical loads but to all loads for which the QSTF tests at OLp played a crucial role, such as impact and drag loads, etc.

References:

Section 2.3 of Appendix A of Reference 3.1 Section 4.3.1 of Reference 3.2

Response

, There were single tests run at OAP conditions for the maximum and minimum water levels in the Oyster Creek quarter-scale (OSTF) tests.

Therefore, at the controlling high water level condition, the shell load l

definition was based on a single test.

General Electric has recognized that some plants may elect to operate at conditions for which only one QSTF test was run. For these plants, such as Oyster Creek, GE has provided additional generic factors to apply to the plant-unique shell load definitions from single tests to account for the increased statistical uncertainty of a single test. It is our understanding that these factors are based on the judgement that 2-5 C

decreasing the number of tests in the data base by a factor of 4 (from 4 tests to one test) increases the uncertainty of the load by a factor of the square-root of 4, or 2.

As a result, GE has specified that these additional margins be included with the NRC rargins of Section 2.3 cf Appendix A of Reference 3.1 as shown on the attached Figure 3-1 for pool swell shell loads based on a single QSTF test. GE has also concluded that no increased margins are required for any other loads derived from a single QSTF test. It is our understanding that this is based on their judgement that the existing margins imposed on these loads are sufficient to cover this.

e e

l l

l 5

2-6 i

l INCREASED MARGIN FOR P0OL SWELL LOAD BASED ON SINGLE OSTF TEST o OaP POOL SWELL SHELL LOADS BASED ON A SINGLE QSTF TEST FOR OYSTER CREEK o GENERAL ELECTRIC HAS PROVIDED GENERIC FACTORS FOR SINGLE TEST STATISTICS BASED ON DOUBLING THE UNCER-TAINTY DUE TO HAVING 1 TEST INSTEAD OF 4 IN DATA BASE

~

o GENERAL ELECTRIC SPECIFIES THE FOLLOWING:

UPLOAD -

NRC MARGIN FOR 4 TESTS IS 6.5% FOR UNCERTAINTY 15% FOR 3D EFFECTS TOTAL MRGIN FOR 1 TEST IS 2X6.5% = 13% FOR UNCERTAINTY 15% FOR 3D EFFECTS

^

' DOWNLOAD -

FOR 1 TEST, DOUBLE THE NRC MARGIN USED FOR 4 TESTS (9.6% FOR OYSTER CREEK)

SHELL NRC MARGIN TOTAL MARGIN  % INCREASE IN MARGIN LOAD (4 TESTS) (1 TEST) 1 TEST VS. 4 TESTS l UPLOAD 1.215 1.28 5.3%

DOWNLOAD 1.096 1.192 8.8%

l 1

o GENERAL ELECTRIC HAS CONCLUDED THAT NO INCREASED l MRGINS ARE REQUIRED FOR ANY OTHER LOADS DERIVED FROM l QSTF SINGLE TEST FIGURE 3-1

Recuest:

Item 4: The AC specification for multiple downcongr ltading is baseo on an exceedance probability of 10- per LOCA. The PUAR states that only two cases of multiple loadin'g were considered and that 1100 lbs per downconer bounded the results in both cases. Clarification is needed why only 2 cases were necessary and what exceedance probability 1100 lbs corresponds to.

References:

Section 2.12.2 of Appendix A of Reference 3.1 Section 4.4.3 of Reference 3.2

_ Response:

~

The multiple downcomer loads'were determined in accordance with Section 4.5.3 of the GE Load Definition Report, Reference 3.6, and Section 2.12.2 of Appendix A of NUREG-0661, Reference 3.1. An exceedance prcbability of 10-4 per LOCA was used in all multiple downcomer load calculations. The calculations were based on curves relating the load per downcomer to the number of downcomers considered and to the proba-bility of exceedance. These curves were derived from the FSlF data and are shown in Figure 3.9-3 of Reference 3.1. As in the case of single downconer loading, the results calculated from these curves were multiplied by a factor to account for differences in dynamic amplifi-cation at Oyster Creek. As mentioned in the PDAR, this factor was 2.0.

The Oyster Creek Plant has 120 downcomers. The calculation method that was used determined the load per downcomer (FDC) for N downcomers chugging in the same direction simultaneously, where 1 < H < 120. The calculation method showed that the load per downcomer decreased as the number of dcwncomers increased, but that the total load (i.e., NFDC) increased with the number of downcomers. Hence, the maximum net lateral load is obtained for 120 downcomers, and this was one of the cases 2-7

considered in the structural analysis. A load of 1100 lbs per downcomer was determined for 120 downcomers chugging in the same direction simul-taneously, based on a probability of exceedance of 10-4 per LOCA.

I It was recognized that local structural reactions might be affected more l

severely by fewer downcomers chugging in the same direction simultane-ously but with a higher load. For example, in the limit, the down-comer / vent header intersection is most severely affected by considering the maximum load on one downcomer. These local structural responses are sensitive not only to the number of downcomers considered but also to the locations of these downcomers. For example, eight downcomers chugging in the sa'me direction simultaneously but dispersed around the torus produce far less severe local structural responses than eight downcomers chugging in the same direction simultaneously in one bay of the torus. It was. found that local effects such as bending and torsion in the vent header and vent line, forces in the vent header support columns, etc., were most severely affected by the total load in the two-bay span between adjacent vent lines.

Accordingly, an analysis was performed to determine the maximum load for 12 downcomers located in a span between vent lines chugging in the same direction simultaneously. This analysis gave the most severe load for j the local structural evalua.tions. Consideration was given in this analysis to the limited likelihood that a group of downcomers chugging striultaneously will have 12 members of the group located in one span. A total probability of exceedance of 10-4 per LOCA was used, where this probability was the product of the probability that a force F would be obtained in N downcomers times the probability that 12 of the group of N are located in a single span. This latter probability is zero for N < 12 and un'ity for N > 108. A value of F was determined 'or each possible value of N and the maximum F value was considered cs the final load. As expected, the final result showed a force lower than would be 2-8

calpulated for a randca set of 12 downcemers chugging in the same direction simultaneously. The result was 1100 lbs per downcomer. This value is actually slightly higher than the number calculated for'120 downcocers; however, to the precision of this analysis the two numbers were equal.

In summary, two cases were considered because these two cases produced the maximum structural responses in vent system members. The result of' 1100 lbs which bounded the results of both cases is based on a proba-bility of exceedance of 10-4 per LOCA.

0 1

i I

2-9 l

l

Recuest:

Item 5: In the analyses of structures for C0 loads the PUAR indicates that the individual load harmonic were cocbined using a random phasing methodology instead of absolute summation. Is this the methodology of NEDE-24840, " Mark I Containment Program, Evaluation of Harmonic Phasing for Mark I Torus Shell Condensation Oscillation loads"? More detail is needed on how the randem phasing was used to obtain the torus shell respcnse. What are the remaining margins after appli-caticn of this nethod? How much reduction from the absolute sun of the responses does the phasing metho-dolocy provide? The above comments apply to all loads for which the randon phasing methodology was used.

References:

Section 2.11.1 of Appendix A of Reference 3.1 Section 4.3.2 of Reference 3.2 Respense:

The random phasing rethod used for the Oyster Creek DBA Condensation Oscillation torus shell load is tiie random phase angle summation method discussed in Gereral Electric Report NEDE-24840, Reference 3.7. The shell pressure . load definition, including the Alternate 2 pressure amplitudes, from Figure 4.4.1-1 of the Load Definition Report, Reference 3.6, was selected. The method of analyzing the shell response to this DBA Condensation Oscillation load is described below.

First, the natural frequencies of the torus shell were found using a dynamic finite element nodel with added water mass. Then, several steady state frequency response analyses were performed with the dynamic model. Each analysis used the DBA(CO) shell pressure spatial distri-butien nonmalized to 1.0 psi at bottom dead center and applied at unit frequencies between 1 to 30 Hz, including the shell natural frequencies.

The resulting responses at each frequency are a measure of the amplified response or transmissibility of the structure for the unit loads.

2-10

The response of the torus shell to the DBA(CO) load amplitude in each 1.0 Hz frequency band is found by multiplying the load amplitude hy the largest torus response in that band. If one or more torus natural i frequencies occur in the band, the load amplitude is divided equally and applied at each frequency.

The total response of the shell to the DBA(CO) load is found hy assign-ing random phase angles to the scaled responses at each frequency and performing a sunmation in accordance with the method of Reference 3.7.

i The largest Mark I Program load combination for the shell which involves DBA Condensation Oscillation is the DBA(CO) + EQ(0) combination. For this case, the allowable general primary membrane stress intensity in the shell is 19.3 ksi and the calculated stress intensity is 18.1 ksi, of which about 8 ksi is due to the DBA(CO) shell load. In addition, it should be noted that this allowable is based on material properties which are the minimum properties permitted by the ASME Code, Reference 3.8. The material of the torus shell at Oyster Creek is significantly stronger than the Code minimum (e.g., the yield strength is at least 32%

higher than Code minimum).

1 As shown in Reference 3.7, use of the random phasing method results in DBA(CO) shell stresses about 45% to 55% lower than stresses obtained by the worst case absolute phasing method for Oyster Creek. The DBA(CO) l load is the only shell load for which the random phasing methodology is used in the Oyster Creek PUA.

i .

2-11 .

D

/

- , - - - , - , . - - , - . - - - - ,, . ,n.---- __ _, . - , - - . . - - - - - . - - - - - - - - . . . , - - - . . _ - . _ _ , - - - - - - - , _ .

~

P.ecuest:

Item 6: Describe how the longitudinal and azimuthal multipliers (LDR Table 4.3.2-1) were used in conjunction with the submerged pressure histories to perform the torus shell evaluations. Provide an example of a time history at a particular location (e.g.,0 = 180* at z/1 = 0.0) to illustrate their use.

References:

Section 2.4 of Appendix A of Reference 3.1 Section 4.3.1 of Reference 3.2

Response

Longitudinal and azimuthal multipliers are given in Table 4.3.2-1 of the

~

Load Definition Report, Reference 3.6. These multipliers are applied to the plant-unique average submerged pressure time history for the pool swell load to obtain the local pressure time history at any desired location within the wetted surface of the torus shell.

For 0yster Creek, the average submerged pressure time history for the ifmiting plant condition is shown in Figure 6-1. For example, to find the local pressure at the bottom of the shell in the middle of the vent bay at the beginning of the transient, the following steps are used.

The multipliers are found in Table 4.3.2-1 of Reference 3.6 for this shell location (z/1 = 0.0, 0 = 180*, as shown in Figure 4.3.2-3 of Reference 3.6). For this time, the " Start of Accident" values of My = 1.0 and M3 = 1.0 are selected. The value of the average submerged pressure at this time is 0.0 psi from Figure 6-1, so that the local pressure is calculated as 1.0 x 1.0 x 0.0 = 0.0 psi for this case. This process is repeated for all desired shell locations and all load tite l points using spatial and time interpolations where necessary for inter-mediate points between values listed in Table 4.3.2-1 of Reference 3.6.

l 2-12

~.

An additional step is taken for Oyster Creek to streamline the dynamic finite element analysis process. After the local submerged pressure is obtained for a time point, the value.of the airspace pressure at 'that time as shown in Figure 6-2 is subtracted from it to obtain the inputs for the dynamic model. This is possible since the airspace pressure transient is slew enough that the torus responds to it statically.

After the dynamic analysis is performed, the airspace pressure effect is added back as a scaled static response at each time point. An example '

of the pressure time history input to the dynamic analysis is shown in Figure 6-3. This input is for a location 30* off the bottom at about quarter bay in the non-vent bay.

. After the 1ccal pressure time history inputs were prepared for the dynamic finite element model, the inputs were checked by calculating the resulting torus vertical load time history over the vent bay and non-vent bay areas and plotting it in Figure 6-4. On the same plot, the vertical load time history calculated directly from the net vertical load curve in the PULD, Reference 3.5, was also plotted for compari-son. The Figure shows that the load inputs were correct since the applied vertical load has a similar shape to the PULD curve and is larger during the download and upload phases by the amount of the NRC pool swell margins of Reference 3.1, as expected.

l 2-13

/

- - s -. ._. ._ , - _ _ . , , , _ _ . _ _ _ _ _ _ _ _ _ _

' i

=

. 0 0

0 f

5 1

0 T 0 -

N A

L P D

=

f P /

G

)

D r

NS I IT S TO P AL EC /

RP C 0 E NE [

NY 0) l ER ER 0 0 c 6 GU t GO T ESRS / 1 s

n i

E R

RS ME e U

G AI EH DR L UP S

[ n t

i I

F CE 7

, UM 6E D /

_ NI 0G K

T 4E

.R /

EO M EL B RU CP U S u 0 0

0 T

N R E 5 E E G I

S T A N S R A

R Y E T O V E A A R l

U S

E R

- P C

' 0 0 0 4 2

• "w E

l4 ili l! . ! j II t ii ji ji

. l

. 0 0

0 5

. 1 T

N A 0 y L .

P 0 G

NS P

D

=

/ .

I T TO AL _

RP E C )D 0 E

NY N S E P I

/ 0 0)C 2

ER G 0 E 6 GO 1 RS E T R C

E

/ S N

(

E R

U .

M E G AI EH U B R U

/ r, N

I I

F T

L S CE S S UM NI T

S E O R

/

.P -

K 4 ,

ED R EL RU CP I

A // 0 0

S 0 T

R U / 5 N E

R E

T O / I S

N A

S T R Y T Q J, _

E R

U S

S E

R P

O 0 0 - 0 1

1 2

' {h 5*

\ $ i : I,\ l

1', 1\.;)l );!;)! J

4 OYSTER CREEK NUCLEAR GENERATING PLANT ,

POOL SWELL TIME HISTORY - OYNRE1 AMPLITUDE VS TIME INPUT -

4.06 FT SUBMERGENCE SP=0.0 NON-VENT BAY .

) .

2 20.0 i  : ' -

E

\

j g -

im i

f VT i \ - - -

l O _.A 1 \ -_ e

& ~ - . - - -'

i /

l 0 .50 1.00 1.50 TIMc (scc) -

ricunc 6-3 Pacssunc inansicwr z/1 = 0.724 i Theta = 150 degrees

t

• OYSTER CREEK NUCLEAR GENERATING STATION -

So (1/40) SECTOR TORUS SUPPRESSION CHAMBER MODEL

\

COMPARISON OF NET VERTICAL FORCE TRANSIENT TO NET TORUS VERTICAL LOAD SUMMATION OF FORCES OVER VENT AND NON-VENT BAYS 4.06.FT DOWNCOMER SUBMERGENCE - SP = 0.O PSI l

_ _ _ p _ _ _ _ _ _ _ _ __ __

,./_

N wn h _

~ "^

%Q ~: -

2 3=

0 *

, ,,,, ,,m r n.% _

w 3

/ _ _ _. _ _ _

E -

i j u i < ai i <

4/

J[

?

-1.00X10 -

0 . .2 .4 .6 .0 1.0 1.2 1.4 1.6 TINE (SEC) ,

FIGURE .6-4 NET TORUS VERTICAL LOAD MULTIPLIED BY HORIZ. PROJECTED AREA AT WATER LEVEL

---

T O T A L N E T COMPUTED FORCE IN VERTICAL (Y) DIRECTION

Request:

Item 7:- The PUAR states that FSI effects are accounted for in the submerged structure loadings. Additional detail is needed on how this was done. Is the criteria for including FSI effects the same as that stated in the AC? How were the FSI loadings obtained? Is the boundary acceleration added to the local fluid acceleration as suggested in the AC or has another method been used?

References:

Section 2.14.5 of Appendix A of Reference 3.1 Section 4.5.2 of Reference 3.2

Response

~

For the Oyster Creek plant-unique analysis, the criteria of NUREG-0661, Reference 3.1, were used to determine the need to include fluid structure interacticn (FSI) effects. Specifically, FSI was included for any structural segment for which the local fluid acceleration is less than twice the torus boundary acceleration.

FSI loads on subnerged torus structures were obtained using a two-step procedure. First a simple, very conservative and non-mechanistic bounding FSI load was defined for each submerged structure. Then, for

, those structures for which preliminary structural analysis indicated that a more realistic definition of FSI was required, a more complex and mechanistic analysis'was performed to better define FSI loads. Each of l these methods is described in detail below:

i l A. Bounding FSI Load Definition l

The bounding effect of fluid-structure interaction (FSI) of the vibrating torus shell and torus pool water on submerged structure drag loads was conservatively accounted for as follows:

2-14 l

. 1. The acceleration of the torus shell was calculated over the range of frequencies for CO, pre-chug, and post-chug.

2. The calculated torus shell acceleration was applied directly to each structure as a fluid acceleration neglecting attenua-

, tion of the acceleration field by distance from the shell or by orientation of the structure with respect to the shell.

B. Mechanistic FSI Load Definition For subnerged structures for which preliminary structural analysis indicated that a more realistic definition of FSI loads was required, local fluid acceleration vectors induced by shell boundary motion were calculated by solving Laplace's equation for the pressures in the pool. The pressures in the pool were deter-mined for all given frequency data. Numerical differentiation was used to calculate the vector components of fluid acceleration.

Torus shell accelerations were calculated using the finite element model of the torus shell described in the PUAR, Reference 3.2.

These were provided to GE which performed the analysis for pool acceleraticn vectors, and provided the results to GPU. The resulting fluid acceleration vectors were used to define FSI loads, for CO, pre-chug, and post-chug.

l t

l l

2-15 .

e b

Recuest:

Item 8: Provide the details of a post-chug submerged structure load calculation for a given segment of the ring girder.

Include numerical values of source strength and DLF as a function of frequency. In addition, provide the accel-eration volume, drag coefficient, interference effect maltiplier and pertinent geometric parameters and configuration used in the calculation.

References:

Section 2.14.6 of Appendix A of Reference 3.1 Section 4.5.2 of Refere.nce 3.2

Response

Post-chug drag loads for Oyster Creek submerged structures were defined based on thetark I Long-Term Program generic load definition per the l LDR, Reference 3.6. Tlie generic load definition was applied to Oyster Creek as follows:

1. Calculate input for GE computer code CONDFOR. (Described in the PUAR, Reference 3.2.)

2. Run CONDFOR. CONDFOR output needed for post-chug load definition:

Method of inages information on all downcomers and each structure segrent in the pool.

3. Use CONCFOR output to calculate unit vector drag loads for post-chug assuming unit source strengths at the two downcomers nearest the structure and assuming the sources are:

(a) in-phase (b) 180* out-of-phase

4. Cotapare total resultant load on entire structure for Cases 3(a) and 3(b) above. Choose case with larger resultant load as design case.

2-16

5. Determine if velocity drag is significant. If so, correct design case loads to account for velocity drag. .

6. Determine if shell or structure interference effects are signifi-cant. If so, correct design case loads to account for interference effects.

7. Transform coordinates if desired by structural analyst.

8. Result of Steps 1 through 8 is unit vector drag loads for post-

~

chug.

9. Combine unit vector drag loads with post-chug source strength data to obtain post-chug drag load.

10. Combine post-chug drag load with FSI load to obtain total post-chug

. drag load.

Numerical values of source strengths as a function of frequency used for post-chug drag analyses are tabulated in Table 8-1. Numerical values of acceleration volume, drag coefficient, interference effect multiplier, and pertinent geometric parameters and configuration used in the above calculations for the ring girder Segment 5, are summarized in Table 8-2 and Figures 8-1 and 8-2. Dynamic load factors for the ring girder are tabulated in Table 8-3.

2-17

(

TABLE 8-l PAGE1 OF 1

! OYSTER CREEK checkea b /$/ 5/.1/52- '

l AMPLITUDES AND FREQUENCIES l FOR POST-CHUG l

F R E O L E *JC Y FREQUENCY R ANGE I / AMPLITUDE 2/ - RANGE I /

AMPLITUDE 2 ' -

(HZ) - 3 CFT /SEC2) (HZ)- (FT 3/SEC 2 )f 0-2 11.98 20 - 21 17.53 2-3 10.36 21 - 22 30.67 3-4 9.87 22 - 24 92.39 4-5 17.40 24 - 25 134.50 5-6 17.00 25 - 26 313.84 6 - 10 18.88 26 - 27 377.83 10 - 11 87.90 27 - 28 251.89 11 - 12 76.18 28 - 29 163.32 12 - 13 41.01 29 - 30 116.66 13 - 14 35.89 30 - 31 43.14 14 - 15 6.82 31 - 32 21.57 15 - 16 6.20 32 - 33 37.91 16 - 17 3.14 33 - 34 50.54 17 - 18 4.18 34 - 35 42.54 18 - 19 2.94 35 - 36 61.87 '

19 - 20 16.82 36 - 37 41.95 37 - 38 20.97 38 - 39 24.47 39 - 40 29.37 40 - 50 224.90 NOTES:

If Se:act the freq ency within each range which will result in the greatest respense.

2 2f Total of all a=plitudes is 2464.43 ft 3/sec ,

. _ , .s,, _ - _ , , _ , . _ , , _ . . , . _ . . , - _ ,. ,

REACTOR VESSEL Y-AXIS I k .

2. 6 5 PT -

T.

I ,,"

+ X-AXIS

,,~. 7,~-~.

~%

9 1 10.63 10.63*e 10

,e ~,

2 9 s 's 3 8 e

g, . .

4 4 0 14.4' s' t 7- 4 0 14.4*'

57.6' TOTAL i 5. 6 :1 57.6

• TOTAL LOCATION IN X-Y PLANE OF 2 0 21.77' SOURCE STRINGT?.S FOR 43.54' TOTAL POST - CHUG DRA0 LCAD CALCULATION FOR RING GIRDER i

l FIGURE 8- 1 l

l OYSTER CREEK l

RING GIRDER SECTION 5 -

AND SOURCE STRENGTHS FOR

. POST -CHUG DRAG LOAD ANALYSIS

O O $

e 0

o - LOCATIONS IN X-Z PLANE OF SOURCE STRENGTHS FOR POST-FRRH G RD R EG E T U

W I

\

x.g 73 j Q '

Z- IS ll .

'N ! / 9/

ll- / ,

e M

_ _ . . S _T ,

FIGURE 8-2 OYSTER CREEK RING GIRDER SECTION 5 AND SOURCE STRENGTHS FOR POST-CHUG DRAG LOAD ANALYSIS

TABLE 8-2 Oyster Creek Plant-Uniaue Analysis Sur. mary of Post-Chug Underwater Draq Load Definition Analysis Inouts for Ring Girder Segment 5 (See Figure 8-2)

Acceleration Drag Volume: 46.8 ft3 Drag Coefficient Not applicable UT.< <2. 7 4 Ir Interference Effect Multiplier: 2.0 Pertinent Geometric Pararaeters -

X Y Z

~~

and Configuration (See Figure 8-2 for coordinate system)

Segment Center Coordinates (f t) -2.7 -14.1 158.7 Segment Direction Cosines 1.0 0 0 (non-dimensional)

Downcomer Bubble Coordinates (f t) -3.1 -6.2 156.1

-3.1 -6.2 162.2 9

Y t

o

~

TABLE 8-3 OYSTER CREEK RING GIRDER o DYNAMIC LOAD FACTOR 5 (SEE NOTE 1)

FREQUENCY FREQUENCY RANGE DYNAMIC LOAD RANGE DYNAMIC LOAD (Hz) FACTOR (Hz) FACTOR

~

0-2 1.01 20-21 3.96 2-3 1.03* 21-22 2.64 3-4 1.05 22-24 1.95 4-5 1.08 24-25 1.25 5-6 1.13 25-26 1.05 6-10 1.45 26-27 .90 10-11 1.61 27-28 .78 11-12 1.81 28-29 .69 12-13 2.11 29-30 .61 13-14 2.57 30-31 .55 14-15 3.34 31-32 .50 15-16 4.90 32-33 .46 16-17 9.51 33-34 .42 17-18 25.00 34-35 .38 l

18-19 23.95 35-36 .35 -

19-20 7.48 36-37 .33 l 37-38 .31 38-39 .29 39-40 .27 40-50 .25 NOTE 1:

This table is slightly different from the table presented during the meeting between the NRC, GPU, and MPR on July 14, 1983. Inad-vertently, the DLF's presented during the meeting were not those used for the ring girder analysis.

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

. - - - -- -,._._w7-9 , _ . . _ __,-.-_.-m, , _ ,- , , , , - . - - - , , y

1 Reouest:

Item 9: The vent system loads occurring during condensation oscillation and chugging as discussed in Sections ,4.4.4 and 4.5.4, respectively, of the LDR, are only centioned very briefly in the PUAR. Were these loads applied as specified in the LDR? What is meant by the statement in Section 4.4.3 of the PUAR that net load effects (for vent system chugging loads) are covered by the dowr. comer tip

. loads?

References:

Section 2.11.3 and 2.12.3 of Appendix A of Reference 3.1-Section 4.4.2 and 4.4.3 of Reference 3.2

Response

~

The vent system loads occurring during condensation oscillation and chugging as discussed in Sections 4.4.4 and 4.5.4, respectively, of the LDR, Reference 3.6, consist of oscillating internal pressure loads applied to the vent lines, vent header and downcomers. These loads were applied to these structures as specified in the LDR. The extreme responses to these loads were then used in the compilation of total responses for C0 and chugging load combinations.

For the pressure loads in the downcomers mentioned in Sections 4.4.4 and 4.5.4, respectively, of the LDR, the LDR states that these pressures should be used to calculate only the hoop stress of the downcomers.- The effects of thrust or other net loads transmitted through the downcomers to other components are covered by separate load definitions ('the downcomer dynamic load for C0 and the downcomer tip load for chugging).

The meaning of the statement in Section 4.4.3 of the PUAR that net load effects (for vent system chugging loads) are covered by the down60mer l tip loads is that the downcomer tip loads were considered to be the source of the net loads transmitted from the downcomers to the vent header, rather than the oscillating internal pressure in the downcomer.

Accordingly, in our analyses of the vent header, vent line, support columns, etc., for chugging load combinations, we included the response 2-18 l

, . - , . - . _ _ . _ . _ .-_ _. mm._,., , . _ _

due to the loads applied to the downcomer tips. This response was included over and above the response to the vent header and vent line oscillating internal pressures, as described in the above paragraph.

This approach is consistent with Section 4.5.4 of the LDR.

9 m

4 6

l -

I

,~

t l

l ~

l l

! 2-19 b

. -. .- . . - . - . - . = -

9 Request:

Item 10: The PUAR states that the contribution of harmonics above 30 Hz was stall for the Oyster Creek chugging toru's shell loads. What is the size of tha contribution between 30 and 50 Hz as a percent of the total load? Also, the AC requires the use of a coupled fluis!-structure analytical model for applying the " post-chug" load specification of

. Section 4.5.1 of the LDR. Was such a coupled model used for Dyster Creek?

References:

Section 2.12.1 of Appendix A of Reference 3.1 Section 4.3.3 of Reference 3.2

Response

Most structures analyzed considered chugging loads to 50 Hz, except the shell model, which considered chugging loads to 30 Hz. The stress calculations for the torus shell used absolute summing (LDR-phasing) of the response to chugging loads. However, the generic study of chugging showed that random phasing (Kennedy-phasing) bounded the FSTF data. By using LDR-phasing on Oyster Creek in lieu of random-phasing, a conser-

. vatisa of about 40% was added to the total chugging response. Evalua-tion of Oyster Creek for chugging loads to 50 Hz indicates that the j

centribution frem 30 to 50 Hz would add only 1-4% to the total chugging i response. Therefore, the shell stress analysis is conservative since l the 30 to 50 Hz contribution is much smaller than the conservatism introduced by using LDR-phasing.

The torus shell model used in the Oyster Creek analyses was a coupled I fluid-structure inudel which included the torus water mass. Section

! 5.1.1.1 of the PUAR, Reference 3.2, provides a description of the model.

1 l

l t

2-20 1

O t

o Reauest:

Item 11: The PUAR states that the 1/40 sector torus shell model discussed in Section 5.1.1 was used to analyze overall shell and support response for static and hydrodynamic Mark I symmetric loads. What about loads not symmetric around the torus such as SRV loads? Was the 1/40 sector model used for a load whose source extends over several bays and whose effects are clearly non-symmetric over the length of the torus? If a different model was used, details need to be provided. Also, what is meant by

" synthesizing the vent . system model mass and stiffness terms into the explicit torus shell model? Convergence studies demonstrate that this coupling adequately represents the participation of the vent system in the torus response for Mark I static and dynamic loads?"

_ Elaboration on the terms " synthesizing" and " convergence studies" is needed.

Reference:

Section 5.1 of Reference 3.2

Response

These two structural nodeling questions were previously addressed by MPR in response to the request for information from the Franklin Research Center. It was agreed at the neeting between the NRC, Brookhaven National Laboratory, Franklin Research Center, GPUN, and MPR on July 14, 1983, that no additional response to these questions is required.

e 2-21

-Request:

Item 12i The Oyster Creek SRY system is unusual for two reasons:

several SRV lines converge to a common header, and plant-unique Y-quenchers are used instead of GE T-quenchers.

The only presently available information for evaluating this unique system is contained in the PUAR itself and Reference 8.3.4 of the PUAR, the Oyster Creek Test Report of May 11, 1978. Reference 8.3.4 does not address questions of load methodology. While the discussion in

, Section 4.6.2 of the PUAR indicates that the Oyster Creek methodology was developed in accordance with Section 2.13.9 of the AC, it is felt that more details are needed to adequately assess the methodology. Details such as instrumentation location and initial conditions for the seven tests used for the definition of the SRV loads are needed. Specific numerical examples of model calibration from the base case and extrapolation to design case amplitudes and frequencies should be provided.

Finally, Section 2.13.8 of the AC states that as part of

the PUAR each licensee is required to either demonstrate that previously submitted pool temperature analyses are sufficient or provide plant-specific pool temperature i'

~

response analyses to assure that SRY discharge transients will not exceed certain pool temperature limits. No such

' discussion has been found in the PUAR and must therefore be provided.

i

References:

Section 2.13 of Appendix A of Reference 3.1 l Section 4.6 of Reference 3.2

Response

The location of the instrumentation for the in-plant SRV test at Oyster Creek is shown in Figures IV-2, IV-3, and IV-7 of the "0yster Creek Nuclear Generating Station Test Report on the Modified Electromatic Relief Valve Discharge Device", dated May 11, 1978. The initial conditions for the tests are defined on page IV-1 and Table IV-1 of the same reference.

2-22

The procedure for model calibration and results for the design case SRV transients involves the following four steps.

1. Calibration of stress analysis for base case test conditions. The base case test condition selected was the 2 valve, cold pipe transient at 1035 psia reactor pressure, zero delta p betw.'en drywell and wetwell, and 12.2 feet of torus water. The average -

stress at each torus shell location and the average column loads for the tests were obtained from the test instruments. These results were compared to the stresses and loads obtained from a

- stress analysis for a hydrostatic load distribution with a peak pressure corresponding to 12.8 feet of water. The ratio of test to analysis was calculated for various areas of the structure using the 95% confidence value from the test data. Bounding values of the ratios we're defined for each area of the structure. Two exatples are tabulated in the attached Table 12-1.

2. Extrapolation for bubble. amplitude. The bubble amplitudes were calculated for the test conditions and for the design case conditions using the Mark I Program generic codes, RVFOR and QBVBS. The ratios of design case to test case bubble pressures were calculated for both positive and negative bubble peak pressures. The larger of the positive and negative ratio was used for all subsequent calculations for each design case condition.

The results for the five design case conditions are listed on the attached Table 12-2.

3. Extrapolation for dynamic response to bubble frequency. The bubble frequencies were obtained from the same RVFOR and QBUBS calcula-tions noted in 2, above. The frequency margins specified in the acceptance criteria of NUREG 0661, Reference 3.1, were added to the i calculated frequencies. The bubble pressure time history was 2-23

applied to a coupled fluid-structure analytical model of the torus at the frequency of the test bubble and at the frequency of each design case bubble. The response of the torus shell stresses and column loads were calculated for each frequency and compared to the static response to obtain a dynamic load factor at each frequency.

The dynamic load factor for each design case was compared to the dynamic load factor for the test conditic.) to obtain the correction factor for bubble frequency. The results for the five design case-2 conditions are listed on the attached Table 12-3. For the three-valve actuation cases correction factors were calculated for the hot condition only; these same factors were also used for the cold .

condition cases. Since the cold condition results in a lower bubble frequency than the hot condition, the dynamic amplification for this torus is lower for the cold condition. Thus, the use of

! the hot condition factors for the cold condition case is conser-vative.

1

4. ~ Addition for the effect of both quenchers discharging simulta-neously. To account for the possibility of all five SRVs I discharging simultaneously a calculation was made which adds l directly the peak responses from each of the two quenchers. This combined response is maximum in the bay with the 3-valve quencher.

The response in this bay to the discharge from the 2-valve quencher ~

was calculated by multiplying the response in the bay with the 2-valve quencher by an attenuation factor which was measured during the in-plant SRV test.

The procedure for combining the effects of all the above steps is

illustrated in the attached Table 12-4.

l 2-24 W

, , , , - - - , . , , , - . ,-,--,n-, . . , , , ----,,---,,.n-..;.,-, ,- - -....--- -- -,, -,-a. , , , - - , . . - , . - - - -

The request for infomation regarding the torus pool temperature analysis and temperature monitoring will be responded to by GPUN by shparate correspondence.

w 1

A l

\

l t

l l

2-25 i

I .

t

. .-- . , - . . w,, . ~, ,- ,- ..y-,,~ - -y .. - ,.w ....._,,y,, . .-. v--- y,-.- - - . ,..--y -_ . -- ,- . -,-,_.- .-- ._,--,-y,

TABLE 12-1 -

CAllBRATION OF BASE CASE SRV LOAD ANALYSIS SHELL OUTER FIBER COLUMN

~

STRESS INTENSITY LOAD

~

TEST, AVERAGE 1450 PSI 31.2 KIPS CALCULATED FOR 2513 PSI 131.4 KIPS HYDROSTATIC LOAD (TEST AVERAGE)

(CALCULATED HYDROSTATIC) 0.58 0.24 1

95% CONFIDENCE VALUE 0.85 0.35 USED IN ANALYSIS 0.88 0.37

/

- . _ _ _ __. , . . . , . - _ - - _____ m , , _ , . - _ .,,- .

~

TABLE 12-2 BUBBLE PRESSURE EXTRAPOLATION DESIGN CASE BASE CASE SRV EVENT + PEAK - PEAK DBA, 1 VALVE, COLD 0.57 0.69 NOC, 3 VALVE, COLD 1.46 1.29 NOC, 3 VALVE, HOT 1,64 1.29 SBA/lBA, 3 VALVE, COLD 2.19 2.23 SBA/IBA, 3 VALVE, HOT 2.21 2.23

TABLE 12-3 DYNAMIC RESPONSE TO BUBBLE FREQUENCY DLF FOR DESIGN CASE DLF FOR BASE CASE SRVEVEkT SHELL COLUMN DBA, 1 VALVE, COLD 0.68 0.65 NOC, 3 VALVE, COLD 0.68 0.77 i

NOC, 3 VALVE, HOT 0.68 0.77 SBA/IBA, 3 VALVE, COLD 1.0 0.92 SBA/IBA, 3 VALVE, HOT 1.0 0.92

TABLE 12-4 EXAMPLE OF SRV LOAD EXTRAPOLATION

~

SHELL STRESS ='(HYDROSTATIC STRESS)x(TEST CALIBRATION FACTOR)x (BUBBLE APPLITUDE FACTOR)x(FREQUENCY FACTOR)

SBA/IBA, 3 VALVE, HOT:

SHELL STRESS = (HYDROSTATIC STRESS)x0.88X2.23x1.0

= (HYDROSTATIC STRESS)x1.96 5 VALVE SIMULTANE0US = 3 VALVE +(2 VALVE)x(ATTENUATION FACTOR)

~

2

= 3 VALVE 1+ VALVE-(A.F.

3 VALVE. _

SBA/lBA, 5 VALVE, HOT 5 VALVE = 1.96 (1 + 0.64 x 0.4) = 2.47 i

3.0 REFEREN'OES I

3.1 U.S. Nuclear Regulatory Commission. Safety Evaluation Report, Mark I Containment Long-Term Program Resolution of Generic Technical Activity A-7. NUREG-0661, July 1980.

3.2 MPR Associates, Inc. Oyster Creek Nuclear Generating Station Mark I Contairr.ent Long-Term Program Plant-Unioue Analysis -

Report Suppression Chamber and Vent System. MPR-733, August 1982.

~

3.3 MPR Associates, Inc. Oyster Creek Nuclear Generating Station Mark I Containment Long-Term Program Plant-Unique Analysis Report Torus Attached Piping. MPR-734, August 1982.

3.4 U.S. Nuclear Regulatory Commission. Letter from Mr. T. A.

Ippolito to Mr. J. F. Quirk of General Electric, October 16, 1981.

3. 5 General Electric Company. Mark I Containment Program Plant-Unique Load Definition Oyster Creek Nuclear Generating

) Station. NED0-24572, Revision 3, November 1982.

3.6 General Electric Company. Mark I Containment Progra'n Load Definition Report. NED0-21888, Revision 2, November 1981.

3.7 General Electric Company. Mark I Containment Program Evaluation of Harmonic Phasing for Mark I Torus Shell Condensation Oscillation Loads, NEDE-24840, October 1980.

i 3.8 American Society of Mechanical Engineers. Boiler and Pressure Yessel Code Section III Nuclear Power Plant Components Division 1. 1977 Edition With Addenda Through Summer 1977.

/

. _ . . , , _ . _ . _ . . . . . . - . _ .._..,._,...__...___._.,__.._,_,..,....._,.__.,_...__,..-,,.._,...___,,,,,,m.,,,_

4.0 APPENDI'X A BROOKHAVEN NATIONAL LABORATORY REQUEST FOR IllFORMATION N

OYSTER CREEK Plant Unique Analysis Report Request for Infomation .

m 1

I L

l 4

/

),'Eage'To

,1.T.EM 1: A(section2.1 5 ction 2.1 of the Acceptance Criteria states that " as part of the FUA, eacn licensee shall specify procedures (including the prim-ary system para eters r.onitored) by which the operator will identify the SBA, to assure manual operation of the ADS within the specified tice period. Longer time periods may be assumed for the SBA in any specific PUA, provided (1) the chugging load duration is correspond-ingly increased, (2) the procedures to assure manual operation within the assu led time period are specified, and (3) the potential for ther- -

mal stratification and asyr. metry effe. cts are addressed in the PUA."

The PUAR does not specifically address the above requirement. Clar-ification is needed.

ITEM-2: PUAR section 4.4.1 and 6.2.2, AC section 2.2 At the vent line-drywell intersection of the Oyster Creek Plant, the vent line is reduced in diameter. Sene thrust loads can be expected to occur at this reduction. No mention of a thrust load at this location is made in section 4.4.1 of the PUAR or the PULD. Section 6.2.2 of the PUAR does not make it completely clear whether the loadings applied to the finite element analysis of the intersection accounted for a possi-ble thrust load.- Clarification is needed on this point.

ITEM 3: PUAR section 4.3.1, AC section 2.3 The LDR and NUREG-0551 specify a minimum of four tests as a data base for obtaining net torus vertical loads. The PUAR states that Oyster Creek will operate with Ozp between drywell and wetwell and this plant condition was the one selected for calculating net torus vertical leads. It appears that only two tests at Otp with the same submergence were conducted for Oyster Creek (Test 5 of NEDE-21944P and Test 10 of NEDE 24615P). The PUAR states that "Oop loads were increased to ac-count for the larger statistical variance associated with the smaller nunber of tests at the Dap conditions." Further justification and clarificatiori for this exception from the acceptance criteria is needed: How many tests are the loads based on? What is the statisti-cal basis for the load increase and by what amount were they in-creased?

The above comments apply not only to the net torus vertical loads but to all loads for which the QSTF tests at Oop played a crucial role, such as impact- and drag loads, etc.

O

Pa'ge ihree ITEM 4: PUAR section 4.4.3, AC section 2.12.2 The AC. specification for multiple downcomer loading is based on an ex-ceedance probability of 10-4 per LOCA. The PUAR states that odly two cases of multiple loading were considered and that 1100 lbs per down-comer bounded the results in both cases. Clarification is needed why only 2 cases were necessary and what exceedance probability 1100 lbs corresponds to.

, _ ITEM 5: PUAR section 4.3.2, AC section 2.11.1 In the analyses of structures for C0 loads the PUAR indicates that the individual load harmonics were combined using a' random phasing methodo-logy instead of absolute sunnation. Is this the methodology of NEDE-24840, " Mark I Containment Program, Evaluation of Harnonic Phasing for Mark I Torus Shell Condensation Oscillation Loads"? More detail is needed o'n how the random phasing was used to obtain the torus shell re-sponse. What are the remaining margins after application of this meth-od? How nuch reduction from the absolute sum of the responses does the phasing nethodology provide? The above comments apply to all loads for ,

which the randem phasing methodology was used. -

ITEM 6: PUAR section 4.3.1, AC section 2.4 Describe how the longitudinal and azimuthal multipliers (LDR Table 4.3.2-1) were used in conjunction with the submerged pressure histories to perform the torus shell evaluations. Provide an example of a time history at a particular location (e.g. e = 180* at Z/t = 0.0) to illus-trate their use.

ITEM 7: FUAR section 4.5.2, AC section 2.14.5 The PUAR states that FSI effects are accounted for in the submerged structure loadings. Additional detail is needed on how this was done.

Is the criteria for including FSI effects the same as that stated in the AC? How were the FSI loadings obtained? Is the boundary acceler-ation added to the local fluid acceleration as suggested in the AC or, has another metnod been used?

O O

e

.. ,?

• Page Four ' .

lTEM 8: FUAR section 4.5.2, AC section 2.14.6 Provide the details of a post chug submerged structure load calcula-tien for a given segment of the . ring girder. Include numerical values of source strength and DLF as a function of frequency. In addition, provide the acceleration volume, drag coeff tetent, interference effect cultiplier and pertinent geometric parameters and configuration used in the calculation.

ITEM 9: PUAR section 4.4.2 and 4.4.3, AC section 2.11.3 and 2.12.3 The vent system loads occurring during condensation oscillation and -

chugging as discussed in sections 4.4.4 and 4.5.4, respectively of the LDR are only centioned very briefly in the PUAR. Were these loads ap-plied as specified in the LDR? What is meant by the statement in sec-( .

tien 4.4.3 of the PUAR that net load effects (for vent system chugging loads) are covered by the downcemer tip loads?

• ITEM 10: PUAR section 4.3.3, AC section 2.12.1 The PUAR states that the contribution of harmonics above 30 Hz was small for the Uyster Creek chugging torus shell loads. What is the

. size of the centribution between 30 and 50 Hz as a percent of the total load? Also, the AC requires the use of a coupled fluid-struc-ture analytical model for applying the " post chug" load specification of section 4.5.1 of the LDR. Was such a coupled model used for Dyster Creek?

' TEM 11: PUAR section 5.1 The PUAR states that the 1/40 sector torus shell model discussed in section 5.1.1 was used to analyze overall shell and support response for static and hydrodynatic Hark I symmetric loads. What about loads not sycmetric around the torus such as SRV loads? Was the 1/40 sector model used for these also? If so, how could a model of 1/2 of a bay be used for a load whose source extends over several bays and whose effects are clearly non-symmetric over the-length of the torus? If a l different nodel was used, details need to be provided. Also, what is meant by " synthesizing the vent system model mass and stiffness terms i

e O

e y

s, . .

• Page Five into the explicit torus shell model. Convergence studies demonstrate that this coupling adequately represents the participation of the vent system in the torus response for Mark I static and dynamit loads?"

Elaboration on the terms " synthesizing" and " convergence studies" is needed.

ITEM 12: PUAR section 4.6, AC section 2.13 The Oyster Creek SRV system is unusual for two reasons: Several SRV lines converge to a ccamon header, and plant unique Y-quenchers 'are used instead of the GE T-Quenchers. The only presently available -

'i nfomation for evaluating this unique system is contained in the PUAR

_ itself and Reference 8.3.4 of the PUAR, the Oyster Creek Test Report of May 11,1978. Reference 8.3.4 does not address questions of load

. methodology. While the discussion in section 4.6.2 of the PUAR indi-cates that the Oyster Creek methodology was developed in accordance with section 2.13.9 of the AC, it is felt that more details are needed to adequately assess the methodology. Details such as instrumentation location and initial conditions for .. :e seven tests used for the def t-nition of the SRV loads are needed. Specific numerical examples of model calibration from the base case and extrapolation to design case acplitudes and frequencies should be provided.

Finally, section 2.13.8 of the AC states that as part of the PUAR each licensee is required to either demonstrate that previously submitted pool temperature ana' lyses are sufficient or provide plant-specific pool temperature response analyses to assure that SRV discharge trans-No such dis-ients will not exceed certain pool temperature limits.

cussion has been found in the PUAR and must therefore be provided.

( .

4 0

1 l -

l

/

- . . , - - - - , , ,- , - - - . - . . , , .--e