ML20086P316

Rev 0 to, Response to NRC Questions on Fermi 2 Plant-Unique Analysis Rept

ML20086P316
Person / Time
Site:	Fermi
Issue date:	11/30/1983
From:	Baskin J, Kumar V, Steinert L NUTECH ENGINEERS, INC.
To:
Shared Package
ML20086P296	List: ML20086P293 ML20086P313 ML20086P316 ML20086P328 ML20086P336 ML20086P343
References
DET-04-028-1A, DET-04-028-1A-R00, DET-4-28-1A, DET-4-28-1A-R, NUDOCS 8402270098
Download: ML20086P316 (107)

Text

-. ..... . -.- .- ...... -.=..-.._....-...-....- .-_ - ._..-._- - -

t I

DET-04-028-1A Revision 0 November 1983 ,

i i.

APPENDIX A l i'

RESPONSES TO NRC-QUESTIONS ON THE FERMI 2 PLANT UNIOUE ANALYSIS REPORT 1

i i.

i.

l@

f 3

e i

f Maam,& nutar,b

REVISION CONTROL SEEET O

V TITLE: Appendix A REPORT NUMBER: DET-04-028-1A Responses to NRC Questions on Revision 0  !

the Fermi 2 Plant Unique Analysis Report L. D. Steinert/ Project Manager INITIALS 5h Birskin , P .E./ Engineering Manager INI$ALS v& L V . J ." KtYma r , P.E./ Engineering Manager vg INITIALS R. A. S4ttehtiz , P ./ Principal Engineer

$N INITIALS A82.

A. K.

/d & '

Lim / Systems Engineer ML INITIALS U ~ US y O. K. Earle/ Supervisor - Licensing INITIALS

~

b PRE- ACCURACY CRITERIA f- pg,y PRE- ACCURACY CRITERIA R"'r f A N PARED CHECK CHECK l PAGb (S) PAGE (S) l i

1 through 0

(($ /Jf k 24 through 0

de&

/

G[

l 7 41 l 8 VK 42 through

$0S YV. YK l Y k.

! through 63 17 64 l

18 through through 21 { h M 70 4

g 4 71 y 22-23 b d.$ VK VK gh y g 4 l CEP-001.4-00 y A-ii nutaq,h

ABSTRACT This appendix provides responses to NRC requests for additional i information on Volumes 1 through 5 of the Fermi 2 Plant Unique Analysis Report. NRC requests are identified in References 1 through 4. Detroit Edison has previously responded to these NRC requests in' References 5 through 8. This appendix is in a question-and-answer format and addresses topics in each of the ,

five volumes of the report.

l t

i i

i l

DET-04-028-1A A-iii Revision'O l

l nutagh r- e ,m ,.-v..-----,- +--v , ,,,,,--v --.m--,ye- m,, m e e w , , e,---#,_,,---,m---__w,,w-,. w w w-, - . wwm, m e we.m,,- ,

TABLE OF CONTENTS Page ABSTRACT A-iii REFERENCES A-v NRC REQUESTS FOR ADDITIONAL INFORMATION Ouestion Reference 1 1 A-1 2 1 A-8 3 1 A-15 4 1 A-16 5 1 A-18 6 1 A-22 7 1 A-23 8 1 A-24 9 1 A-30 10.1 2 A-42 10.2 2 A-44 10.3a 2 A-47 10.3b 2 A-59 10.4 2 A-60 10.5 2 A-61 10.6 2 A-63 11.1 2 A-64 11.2 2 A-65 11.3 2 A-69 11.4 2 A-70 11.5 2 A-71 11.6 2 A-74 12.1 2 A-76 12.2 2,3 A-77 12.3 2 A-80 13 4 A-82 14 4 A-90 DET-04-028-1A A-iv Revision 0 nutsch

' REFERENCES

1. Letter from B. J. Youngblood (NRC) to Harry Tauber (Detroit Edison), " Fermi 2 Mark I Containment - Plant Unique Analysis Report" dated June 29, 1982.

2. Letter from B. J. Youngblood (NRC) to Harry Tauber ( Detroit

- Edison), " Fermi 2 Mark I Containment - Plant Unique Analysis Report" dated July 19, 1982.

3. Letter from B. J. Youngblood (NRC) to Harry Tauber (Detroit "dison), " Fermi 2 Mark I Containment Plant Unique Analysis Report" dated July 21, 1982.

4. Letter from B. J. Youngblood (NRC) to Harry Tauber (Detroit Edison), " Fermi 2 Mark I Containment -

Plant Unique Analysis Report" dated June 30, 1982.

5. Letter EF2-58,955 from Harry Tauber (Detroit Edison) to B. J.

Youngblood (NRC), " Mark I Containment Request for Additional Information," dated August 2, 1982.

6. Letter EF2-59,222 from Harry Tauber ( Detroit Edison) to B. J.

Youngblood (NRC), " Mark I Containment Request for Additional Information," dated August 2, 1982.

[Q \

/

7. Letter EF2-59,268 from Harry Tauber (Detroit Edison) to L. L.

Kintner (NRC), " Mark I Containment Request for Additional Information," dated September 9, 1982.

8. Letter EF2-59,281 from Harry Tauber (Detroit Edison) to L. L.

Kintner (NRC), " Mark I Containment Request for Additional Information," dated September 9, 1982.

, fN N

DET-04-028-1A A-v Revision 0 nutggh

fN Ouestion 1 D

Published acceleration drag volumes were used to determine the drag loads on sharp cornered submerged structures instead of the equivalent cylinder procedure specified in the acceptance criteria. Provide a list of structures which were treated in this manner. For the ring beam, provide specific dimensions of the structure, as well as the local acceleration and velocity for the post-chug -loading condition. A copy of K. T. Patton's MS thesis from the University of Rhode Island (1965) would be useful in resolving this issue if it is available.

Response to Ouestion 1 x) 1. The alternate method for calculating acceleration drag volumes was used for the following structures:

Ring beam T-quencher support beam

- T-quencher support pedestal T-quencher support gusset plates The ring beam was divided into the segments shown in PUAR Table 2-2.2-9-for analysis of post-chug submerged structure loads.

Oi DET-04-028-1A A-1

\ /

Revision 0 nutggb

Post-chug submerged structure loads on the ring beam were calculated on the basis of the two nearest downcomers chugging at the maximum source strength, with the downcomer phasing selected to maximize the local acceleration. Segment 7 of the ring beam experiences the highest loads, as shown in PUAR Tible 2-2.2-9.

Forces are calculated for the 50 frequencies and corresponding source strengths listed in PUAR Table 1-4.1-15.

The following information presents sample calculations of post-chug submerged structure loads on Segment 7 of the ring beam.

For submerged structure loads, the contribution due to velocity drag is negligible compared to acceleration drag. Figure 1-1 shows the cross section of the ring beam at Segment 7, used for calculating the acceleration drag volume. For the flow direction normal to the web, the beam is idealized as a rectangular cross section, as shown by the dotted lines in Figure 1-1. From LDR l Table 4.3.4-1, the acceleration drag volume, V, for a rectangular cross section is V = A, x L (4ab + 1.33 x a 2),

where Ay = Wall interference factor and L = Len;th of the segment.

DET-04-020-1A A-2 Revision 0 nutp_Qh

y For Segment 7, Aw = 2.0 and L = 2.72 feet. This results in a U drag volume equal to 90.36 ft 3 for Segment 7.

For the bounding load case (two downcomers chugging out of phase), the acceleration, A g, on Segment 7 normal to the direction of the web for a unit source strength was calculated as 2 Therefore, the force, 0.009342 ft/sec . F, for the unit source strength is F= = 1.636 lbs.

9c Table 1-1 shows the results of sample calculations for the dynamic force in each f requency range from 0 to 50 Hz. Dynamic load factors are calculated corresponding to the 48.5 Hz natural

/

(j f requency . of the ring beam given in PUAR Figure 2-2.4-3. The dynamic force in each frequency range is absolutely summed and multiplied by a factor of 0.65 to account for randomness in phasing. The surface area of Segment 7 was calculated precisely as 1,160 in 2 from the finite element model shown in PUAR Figure 2-2.4-1. Therefore, the pressure on the web at Segment 7 of the ring beam, as shown in PUAR Table 2-2.2-9 (without tha FSI effect), is calculated as l 50 0.65 x I (F x DLF) ressure = egment 7 '

Area o and is equal to 24.1 psi, as shown in Table 1-1.

i

/m s

\

\

DET-04-028-1A A-3 Revision 0 nutggb

As discussed earlier, the acceleration drag vo'.ume for various segments of the ring beam for the flow direction normal to the web has been calculated by idealizing the I-section by a rectangular section. In the equation of acceleration drag volume, the area of the cross section was conservatively added as

! the area of the rectangular cross section rather than the actual I-section. Overall, the submerged structure loads in the PUAR have been calculated conservatively.

l i

l The right-hand coordinate system is:

r l

l TORUS CROSS SECTION l

- TO RPV q l

l hY x

e. _ _ _ - _ -

K. T. Patton's Master's thesis is not available from either the University of Rhode Island or the author.

l l

l l

DET-04-028-1A A-4 Revision 0 nutp_qh

Table 1-1 7-- DYNAMIC FORCE ON SEGMENT 7 DUE TO

( j POST-CHUG SUBMERGED STRUCTURE LOADS FORCE CORRESPONDING TO AMPLITUDE DYNAMIC DYNAMIC FORCE FREQUENCY AT EACH FREQUENCY LOAD FACTOR (F x DLF)

(Hz) (F) (lbs) (DLF) (lbs) 0-1 19.6 1.0 19.6 1-2 19.6 1.0 19.6 2-3 16.9 1.0 16.9 3-4 16.1 1.0 16.1 4-5 28.5 1.0 28.5 5-6 27.8 1.0 27.8 6-7 30.9 1.0 30.9 7-8 30.9 1.0 30.9 8-9 30.9 1.0 30.9

^

9-10 30.9 1.0 30.9

/)

\m,/

10-11 11-12 143.8 124.6 1.0 1.1 143.8 137.1 12-13 67.1 1.1 73.8 13-14 58.7 1.1 64.6 ,

14-15 11.2 1.1 12.3 15-16 10.1 1.1 11.1 9

16-17 5.1 1.1 5.6 17-18 6.8 1.1 7.5 18-19 4.8 1.2 5.8 19-20 27.5 1.2 33.0 20-21 28.7 1.2 34.4 21-22 50.2 1.2 60.2 22-23 151.2 1.3 196.6 23-24 151.2 1.3 196.6 24-25 220.0 1.3 286.0 25-26 513.4 1.4 718.8 26-27 618.1 1.4 865.3 i

DET-04-028-1A A-5 (U)

~

Revision 0 nutggh

A Table 1-1 DYNAMIC FORCE ON SEGMENT 7 DUE TO POST-CHUG SUBMERGED STRUCTURE LOADS (Concluded)

FORCE CORRESPONDING TO AMPLITUDE DYNAMIC DYNAMIC FORCE FREQUENCY AT EACH FREQUENCY LOAD FACTOR (F x DLF)

(Hz) (F) (lbs) (DLF) (lbs) 27-28 412.1 1.5 618.2 28-29 267.2 1.5 400.8 29-30 190.8 1.6 305.3 30-31 70.6 1.6 113.0 31-32 35.3 1.7 60.0 32-33 62.0 1.8 111.6 33-34 82.7 1.9 157.1 34-35 69.6 2.0 139.2 35-36 101.2 2.1 212.5 36-37 68.6 2.3 157.8 37-38 34.3 2.5 85.8 38-39 40.0 2.7 108.0 39-40 48.0 3.0 144.0 40-41 367.9 3.3 1214.1 41-42 367.9 3.7 1361.2 42-43 367.9 4.3 1582.0 43-44 367.9 5.0 1839.5 44-45 367.9 6.2 2281.0 45-46 367.9 8.0 2943.2 46-47 367.9 11.2 4120.5 47-48 367.9 17.7 6511.8 48-49 367.9 25.0 9197.5 49-50 367.9 17.1 6291.1 Total 43059.8

=

059 = 24.1 psi.

Total Pressure 16 A-6 O1 DET-04-028-1A Revision 0 nutech

e 3.25" I

2.0" l

2a = 3.4. . 2.0= C o

o C=

.. h o i i 4,

- C

- C = 1.5*

v iP I g g o

2.292' 2h = 16 "

a. Actual Geometry

% a

' 3 J, e o a' e

' l C

' l e .

I 8 Flow Direction - -- C = 1.5"

• l

. . e e e l e s

i '

. l C l <r I I 9

- 2s = 3.4' 2b = 16

b. Model Geometry l Figure 1-1 RING GIRDER CEOSS SECTION AT SEGMENT 7 OF l ACCELERATION DRAG VOLUME CALCULATION O

('

DET-04-028-1A Revision 0 A-7 nutgsb 1 . - . ._. - _ _ _ - . . . - . - . _ , . , . _ - - - - _ _ . . _ - - _ . _ - _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _

m Ouestion 2 A statistical basis was used to account for random phasing of the loading harmonics for condensation oscillation and chugging loadings. The random phasing approach consists of multiplying the absolute sum of the responses (i.e., the AC accepted approach) by a scale factor determined from the FSTF data.

Provide more detailed documentation for the justification of the 0.65 value of the scale factor and comment on the remaining conservatism after application of this factor for both the condensation oscillation and chugging loadings. List all loads (such as CO, post-chug, etc.) and all structures (such as torus shell, ring beam, etc.) for which the scale factor is used. In addition, provide the basis for the statement that Alternate 4 leads to a 20% increase in the loads and verify the numbers given in Table 1-4.1-4 on page 1-4.48. In particular, check the consistency of these numbers with those given in the FSTF Letter Report MI-LR-81-01.

Response to Question 2 The loads for which the random phasing methods were used to combine harmonic response are:

a. DBA Condensation Oscillation Loads on the Torus Shell DET-04-028-1A A-8 Revision 0

. nutg.q, h

b. DBA Condensation Oscillation Loads on Submerged Structures

c. Post-Chug Loads on the Torus Shell

d. Post-Chug Loads on Submerged Structures The components of the torus and vent system affected by the above loads are identified in Mrmi 2 PUAR Tables 2-2.2-1 and 3-2.2-1.

The components of the SRS piping and T-quenchers affected include the submerged portion of the SRV piping, the T-quenchers, and their supports.

For combining harmonic responses, recommendations are made in NEDE-24840 for the use of a 50% non-exceedance probability (NEP) value based on random phasing cumulative distribution function (CDP) curves as a means to provide an appropriate level of conservatism for the combined response. The approach used in the Fermi 2 plant unique analysis (PUA) consists of multiplying the absolute sum of the harmonic responses by a scale factor of 0.65, determined from the data contained in NEDE-24840. The method estimates the response at an 84% NEP with a 90% confidence level, as described in Section 1-4.1.7.1 of the PUAR. This method of combining harmonic responses is more conservative than that recommended in NEDE-24840.

The scale factor of 0.65 was derived from the 14 response quantities given in Tables 6-1 through 6-3 of NEDE-24840. Ratios DET-04-028-1A A-9 9j j Revision 0 l nutp_qh

p of the absolute sum and 84% NEP response values from these tables

\') were calculated as shown in Table 2-1. The mean ( ) and standard deviation (o) values were calculated for the ratios of these 14 responses. Using a Gaussian (normal) distribution, the tolerance limit ( Ra ,y ) is then calculated as:

Ra , y = go where a = Confidence level, y = Non-exceedance probability (NEP), and e = Tolerance factor for normal distribution depends on a,y and the sample size.

Using the values in' Table 2-1, a tolerance limit (Ra,y) with an 84% NEP and a 90% confidence level is determined to be 1.53.

n

( \

.Q Thus, an 84% NEP response with a 90% confidence level can be calculated using a scale factor of 0.65 (reciprocal of 1.53) applied to the absolute sum response.

A - comparison of FSTF responses calculated using the Fermi 2 methodolgy with the maximum measured FSTP responses in various tests is provided in PUAR Table 1-4.1-4. A copy of this table is l

j attached. The comparison provides an assessment of the con-l servatism which results when applying the Fermi 2 methodology.

l The values for maximum measured FSTF response listed in PUAR Table 1-4.1-4 were obtained using the same methods as for l NEDE-24840. The FSTF letter report MI-LR-81-01 may have used DET-04-028-1A A-10 c n/ Revision 0 l

L nutash

preliminary test results, slightly different data reduction, or analytical techniques. The differences between the PUAR and MI-LR-81-01 are very slight, and do not affect the conclusions of the comparison of the analysis to the data. The analytical techniques used in the Fermi 2 PUA are conservative and result in predictions that bound the maximum measured FSTF response by a wide margin.

Measured pressure amplitudes from FSTF Test M12 were used in the Fermi 2 PUA as a fourth alternate in calculating the response due to the condensaton oscillation load. It has been observed that FSTF Test M12 condensation oscillation torus shell pressures at certain frequencies are higher than the pressures for the three alternates specified in the LDR. A comparison of the torus responses for Fermi 2 due to application of FSTF Test M12 amplitudes and amplitudes for the three LDR alternates shows that the additional conservatism in the response due to M12 is location dependent. At some locations, the response is as much as 27% greater than that due to the LDR alternates, as shown in Table 2-2. It is estimated that the fourth alternate (M12) has added about 10% to 30% of additional conservatism to the Fermi 2 condensation oscillation response.

DET-04-028-1A A-ll Revision 0 nutp_gh

Table 2-1 O TORUS RESPONSE RATIOS FROM TABLES 6-1 THROUGH 6-3 OF NEDE-24840 RESPONSE OUANTITY RESPONSE RATIO (ABS. SUM / 84% NEP) 1 1.53 2 1.55 3 1.69 4 ' 70 5 1.72 6 1.74 7 1.78 8 1.80 9 1.86 10 1.88 11 1.94 12 2.01 l

13 2.01 14 2.02 Mean (p) value 1.80 Standard Deviation (e) 0.16 DET-04-028-1A A-12

- ' Revision 0 nutggh

Table 1-4.1-4 FSTF RESPONSE TO CONDENSATION OSCILLATION MAXIMUM MEASURED CALCULATED FSTF RESPCNSE RESPONSE OUANTITY FSTFRESPONp{I AT 84% NEP M8 Mlla M12 BOTTOM DEAD CENTER 3.0 2.3 1.6 2.7 AXIAL STRESS (ksi)

BOTTOM DEAD CENTER 3.7 2.6 1.4 2.9 HOOP STRESS (ksi)

BOTTOM DEAD CENTER 0.17 0.11 0.08 0.14 DISPLACEMENT (in.)

INSIDE COLUMN 184 93 68 109 FORCE (kips)

OUTSIDE COLUMN 208 110 81 141 FORCE (kips)

NOTE:

(1) USING CO LOAD ALTERNATES 1, 2, AND 3.

DET-04-028-1A A-13 Revision 0 nutp_qh

' Table 2-2 COMPARISON OF FERMI 2 TORUS RESPONSE DUE TO CONDENSATION l

OSCILLATION LOAD ALTERNATE 4 (M12) AND THE THREE LDR ALTERNATES l I

l MAXIMUM RESPONSE '

RESPONSE QUANTITY CONSERVATISM DUE ALTERNATE 4 ALTERNATES 1,2, TO ALTERNATE 4 i (M12) AND 3 FROM LDR (%)

INSIDE COLUMN DOWNWARD FORCE (KIPS) 206.76 182.58 13.2 OUTSIDE COLUMN DOWNWARD FORCE (KIPS) 225.44 200.31 12.5 INSIDE SADDLE DOWNWARD 289.48 261.38 10.8 FORCE (KIPS)

OUTSIDE SADDLE DOWNWARD FORCE (KIPS) 350.23 309.98 13.0

[ MEMBRANE STRESS INTENSITY Q )' AT BOTTOM DEAD CENTER 5.81 13.1 NEAR MID BAY (KSI) 6.57 MEMBRANE STRESS INTENSITY AT ABOUT 60* BELOW EQUATOR AT MITER ON OUTSIDE 8.8 7.81 12.7 MEMBRANE STRESS INTENSITY AT 30' ABOVE EQUATOR NEAR MITER 6.44 5.07 27.0 l

l l

l I

I DET-04-028-1A A-14 Revision 0

< nutggh l

. Question 3 8.v The downcomer dynamic load methodology derived from the supple-mental FSTF tests was for tied downcomers. Justify the use of the methodology for the untied downcomers, as shown in the PUAR.

Response to Question 3 Section 4.4.3.1 of the LDR, Revision 2, states that the load definition for condensation oscillation downcomer loads is applicable to downcomer pairs which are tied by lateral bracing, or where the downcomer-to-vent header intersection is stiffened with gussets or other means. The Fermi 2 downcomer pairs are stiffened at each intersection by a crotch plate and by outer stiffener plates shown in PUAR Figure 3-2.1-12. A frequency analysis of Fermi 2 downcomers shows that the predominant, fundamental mode of vibration is the sway mode, i.e., both downcomers in a pair simultaneously deflecting in the same direction. This results in the Fermi 2 downcomers responding as if they were tied by lateral bracing at the ends of the l downcomers. This behavior is identical to that of the FSTF tied l

downcomer pairs.

1 C DET-04-028-1A A-15 Revision 0 nutgsb

Question 4 The acceptance criteria specified that for multiple downcomer chugging, the force per downcomer shall be based on an exceedance-probability of 10-4 per LOCA. A correlation between load magnitude and probability level derived from a statistical analysis of FSTF data was utilized in the PUA. Provide the details of the correlation and justification for the use of the correlation.

Revised Response to Question 4 The methodology used to compute the probabilities of exceedance for the Fermi 2 multiple downcomer chugging loads shown in PUAR v Table 3-2.2-15 is based upon the understanding that the chugging duration of 512 seconds and the number of downcomer chugs of 313 were obtained from FSTF test results.

Further study of the FSTF chugging data report (General Electric Report NEDE-24539-P, dated April 1979) indicated that a chugging duration of 512 seconds represents a realistic duration for an actual plant. By dividing the chugging duration of 512 seconds by a conservative chugging period of approximately 1.63 seconds observed in FSTF, a total number of 313 chugs was obtained.

Also, it was observed that not all of the 313 chugs were synchronized pool chugs.

(

DET-04-028-1A A-16 Revision 0 nutggh

From FSTF Test M1, which is representative of Fermi 2 plant conditions, it was observed that about 33 percent of all the chugs were synchronized pool chugs. The rest of the chugs were not well synchronized pool chugs and would not result in any multiple downcomer lateral load having the force in the same direction occurring at the same time. Therefore, based upon FSTF Test M1, out of 313 chugs, only about 104 chugs (33%) were synchronized pool chugs resulting in a number of downcomers having the lateral force in the same direction at the same time.

Scaling the above information for the conservative Fermi 2 chugging duration of 900 seconds, the number of synchronized pool chugs for Fermi 2 will be about 182. In acevrdance with NUREG-0661, the probability of exceedance for calculating the force per downcomer in multiple downcomer chugging is based on the premise that the force per downcomer would exceed the design load once per LOCA. Thus, for Fermi 2, the probability that the force per downcomer in a pool chug can be exceeded once per LOCA will be the reciprocal of 182, or 5.5x10-3 This probability level is applicable for any number of downcomers considered to be loaded with the same force in the same direction at the same time.

Based upon the above probability of exceedance, the chugging forces per downcomer presented in PUAR Table 3-2. 2-15 are bound-ing for different numbers of downcomers considered to have the lateral force in the same direction occurring at the same time.

DET-04-028-1A A-17 Revision 0 h

nutp_g.h

i O Ouestion 5 l

1 (a) On page 1-4.113, it is stated that the peak positive '

bubble pressure and maximum bubble pressure differential from the Monticello T-quencher test data are 9.9 psid and 18.1 psid, respectively. Our information (Table 3-3, Page 3-10, NEDE-21878-P) indicates that these l

values are 9.3 psid and 17.4 psid. Provide information to permit clarification of this discrepancy.

(b) We require additional information to determine whether modification of the bubble pressure bounding factor from the LDR value of 2.5 to the proposed value of 1.75 is justified. Specifically, the peak positive and negative k, bubble pressures predicted by the SRV bubble pressure methodology when the 1.75 multiplier is employed should be reported. The initial conditions for this calcula-tion are to correspond to the CP, NWL, SVA case as l

listed in Table 3-2 of NEDE-21878-P.

l l Response to Question 5 l

l (a) The peak positive bubble pressure and maximum bubble l

l ' pressure differential from the Monticello test c' ata are 9.3 psid and 17.4 psid, respectively. The statement in the PUAR regarding the values of 9.9 psid and 18.1 psid I \ DET-04-028-1A A-18 Revision 0 nutggb l

f-

refers to the calculated values of peak positive and maximum bubble pressure differential using the bubble pressure bounding factor of 1.75.

(b) The predicted values using the 1.75 factor for the CP, NWL, SVA case listed in Table 3-2 of NEDE-21878-P are:

Peak positive bubble pressure = 9.9 psid Peak negative bubble pressure = 8.2 psid The techniques used to model the Fermi 2 T-quencher water jets are the same as those used for the Mark I T-quencher (General Electric Report NEDE-25090-1-P), except the Fermi 2 T-quencher geometric characteristics are used (PUAR Figure 1-4.2-6). The model described in NEDE-25090-1-P was approved in NUREG-0661 and is based on steady-state submerged jet theory, published literature on jets, and test data.

The Fermi 2 T-quencher has the same hole size and hole spacing as the Mark I T-quencher. However, the Fermi 2 T-quencher arm diameter is 20" and the hole distribution is slightly dif ferent from the Mark I T-quencher. Therefore, the model described in NEDE-25090-1-P was utilized with slight modifications to account for the Fermi 2 geometric differences.

DET-04-028-1A A-19 Revision 0 nutgch

O The dif ference in arm diameter, and hence water volume, is taken into account in the SRV clearing model described in the PUAR.

This model provides the mass flow rate information needed to calculate the water velocity in the T-quencher arm from which the hole velocity in each jet section is calculated. (A jet section is defined as the portion of the arm where the number of holes per column is equal.)

Since the Fermi 2 T-quencher hole size and hole spacing are equal to the Mark I T-quencher, the jet phenomena will be similar.

4 That is, orifice jets will be formed first. These orifice jets will then merge into rectangular column jets which, in turn, will merge into quencher arm jets. The widths and heights of the jets for Fermi 2 are based on the Fermi 2 T-quencher geometry. This b is the same procedure used to determine jet widths and heights for the Mark I T-quencher. The jet velocities are derived from the jet width and height and the principle of conservation of momentum up to.the time (to) when all the water has been cleared from the T-quencher. After all water has been cleared from the T-quencher, the quencher arm jet velocity is assumed to decrease linearly to zero in a time equal to the clearing time (to), as described in NEDE-25090-1-P.

i The T-quencher water jet model described in NEDE-25090-1-P is t

based mainly on steady-state jet theory and published literature on jets. The test data were used only to confirm the prediction D DET-04-028-1A Revision 0 A-20

, nutagh l

of the point where orifice jets merge and to estimate the time (to) required for the quencher arm jet to decay to a negligible velocity. The Fermi 2 T-quencher water jet model uses the same principles and assumptions while properly incorporating the Fermi 2 T-quencher geometry.

O !

l 1

l I

I l

I l

I I

l l

DET-04-028-1A A-21 )

i Revision 0 l

nutg,gh  !

i

i Ouestion_6_

sO)-

v l Th'e post-chug submerged structure loads, as specified in the acceptance . criteria, were to be computed on the basis of the two nearest downcomers chugging at the maximum source strength, with phasing between _the downcomers that maximizes the local acceleration. On PUAR page 2-2.39, it is stated that the loads were developed using the average source strength. Please clarify the situation by documenting the calculation in detail for the ring beam, giving the source strengths used and their locations.

Response to Ouestion 6

.g)

(

The statement on PUAR page 2-2.39, Revision 0, which stated that

' ~

the average V loads were, developed using the post-chug source strength was incorrect. PUAR page 2-2.39, Revision 1, correctly states that the maximum source strengths calculated using the LDR

~

The source strengths

~

method,s were used in the Fermi 2 analyses.

as a ' function of frequency used for the Fermi 2 post-chug sub-merged structure load' analyses are listed in PUAR Table 1-4.1-15.

These post-chug submerged structure loads were computed on the l

basis of the two nearest downcoiders chugging at the maximum source strengths, with phasing between the downcomers that i

maximizes the local acceleration as required in Appendix A of NUREG-0661. The response to Question 2 provides additional details on the calculation of post-chug loads on the ring beam.

, O DET-04-028-1A A-22 Revision 0 nutggb

Question 7 Provide a more detailed discussion of the method used to account for FSI effects on condensation oscillation and chugging submerged structure loads. Include an explanation of how the local pool fluid accelerations are determined.

Response to Ouestion 7 A detailed discussion of the method used to account for FSI effects on condensation oscillation and chugging submerged structure loads is provided in the Continuum Dynamics, Inc. Tech Note No. 82-15, Revision 0, " Mark I Methodology for FSI Induced Submerged Structure Fluid Acceleration Drag Loads."

(

j DET-04-028-1A A-23 Revision 0 nutggh

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

_ - _ - . = = - _ - . - .

Question 8 Provide a complete description of the bases for the local-to-bulk pool temperature differences which are presented in Section 1-5.1 of the PUAR. The AC (Section 2.13.8.2) stipulate that this parameter should be supported either by existing Monticello pool temperature data or in-plant tests. If the first of these options is employed, the applicant mast demonstrate the applicability of the Monticello data base by providing a detailed comparison of the respective quencher and suppression pool geometries. Also, since credit for RHR effectiveness in reducing the local-to-bulk temperature difference is being taken by the applicant, comparison of the suction and discharge geometries of

, g the 'espective RHR systems should also be provided.

If the

' Monticello data base is used in conjunction with any analytical

+

.modeling to estimate plant unique values of local to bulk temperature differences, a complete description of the analyses should be supplied together with a demonstration of the credibility -of the model in terms of its ability to accurately 1

predict experimental suppression pool temperature responses to j

extended SRV discharges.

1 Response to Questions 8 GE had analyzed seven, postulated long-term SRV events in Fermi 2 to demonstrate the plant's conformance with the local pool

\ -DET-04-028-1A A-24 Revision 0

__. . , , _ . _ _ . . _ _ . , . . , - _ . . _ . , _ . , _ _ . - . _ _ . _ . , . _ _ . _ . , _ . . ~ . _ _ . . , , . . . -

l temperature limit as defined by the NRC (c.f., Reference 15 of the Fermi 2 PUAR). The asbomptions and results of this analysis are presented in Section 1-5.1 of the PUAR.

The analyJis is based on properly modeling each of the seven events using two GE proprietary computer codes to evaluate local and bulk pool temperatures as a function of time.

The first code is a coupled RPV and suppression pool thermo-dynamics model which calculates the transient response of the suppression pool during long-term events that add heat to the pool. This model performs fluid mass and energy balances in the reactor primary system and the supprossion pool and calculates the reactor vessel water level, pressure, and the long-term response of "che suppression pool bulk temperature. The various modes of operation of all important auxiliary systems, such as the SRVs, main steam isolatior. valves (MSIVs), emergency core cooling system (ECCS), residual heat removal (PHR) system, and feedwater are modeled. To simulate a specified reactor cooldown rate or depressurization rate, a rate of change of temperature or pressure may be imposed on the reactor vessel. In addition, the model also simulates system set points (automatic and manual),

and specified operator actions. The calculated maximum suppression pool bulk temperatures of each event are tabulated in Table 1-5.1-1 of the PUAR.

DET-04-028-1A A-25 Revision 0 nutp_qh

i l

l l The second computer code is a local pool temperature model which

1. celc.dlates the water temperature in the vicinity of the quencher during SRV discharge events that add heat to the pool. This code was developed under the Mark I Program expressly to model Mark I plants equipped with T-quencher discharge devices. The computer code is applicable to the Fermi 2 T-quencher.

The model is calibrated to the Monticello test results.

Pool temperature distributions predicted by the model have been compared with the Monticello T-quencher test results and the Monticello T-quencher thermal mixing test results. Results of the comparison indicate that the model predicts the local quencher temperature during SRV actuation.

V R sults from the first model, such as the mass and energy added to and removed from the pool during each transient (i.e., RHR and SRV flows), are input into the second model along with pool geometry, submerged structures geometry, and initial pool conditions.

The overall local temperature analysis consists of two major, coupled components: a momentum balance to solve for the bulk pool velocity and a two-dimensional energy model which determines the temperature distribution in the pool by superimposing the circulation of pool water induced by the SRV discharge on the bl DET-04-028-1A Revision 0 A-26 nutggh

bulk motion of the pool. The energy model is of sufficient generality to accomodate multiple SRV actuations for random patterns of T-quencher discharge at selected points in time.

The energy model is applied locally at uniformly distributed nodes throughout the pool. One axial node is assigned to each half bay for each of eight horizontal layers. Thus, a total of 16 nodes per bay are used to describe the temperature distribu-tion in the pool, Application of the models to these nodes results in a coupled set of algebraic equations which are solved by successive substitution at each time step. A simple iterative scheme is employed to ensure conservation of energy.

The local temperatures of interest in this analysis are calcuiated by averaging the temperature of the nodes directly ,

above and below the T-quencher in the downstream portion of the bay. The local temperatures tabulated in Table 1-5.1-1 of the PUAR correspond to the bay with the highest temperature throughout each event calculated in this manner.

The remaining portion of this response pertains to the oral request to address the basis of the Fermi 2 local pool temperature limit curve given in Figure 1-5.1-1 of the PUAR.

DET-04-028-1A A-27 Revision 0 nutgc_h

% The curve is based on NUREG-0783 (c.f., Reference 15 of the Fermi 2 PUAR), which states:

1. For all plant transients involving SRV operations during which the steam flux through the quencher perforations exceeds 94 lbm/ft 2-sec, the suppres-sion pool local temperature shall not exceed 200*F.

1

2. For all plant transients involving SRV operations during which the steam flux through the quencher perforations is less than 42 lbm/ft 2-sec, the suppression pool temperature shall be at least 20*F subcooled.

.,x 3. For all plant transients involving SRV operations during which the steam flux through the quencher perforations exceeds 42 lbm/ft 2-sec, but is less than 94 lbm/ft 2-sec, the suppression pool local temperature is obtained by linearly interpolating the local temperatures established in Items 1 and 2.

l

[

Fermi 2 T-quenchers are submerged in 10 feet of water, corresponding to 18.9 psia. The saturation temperature at 18.9 psia is 224.8*F., Thus, for Limit 2, a 20*F subcooling translates into a suppression pool local temperature limit of 204.8'F.

DET-04-028-1A A-28 Revision 0 nutggb

Since the steam mass flux through the quencher perforations is O

directly dependent on reactor vessel pressure, mass fluxes of 42 lbm/ft 2-sec and 94 lbm/ft2 -sec correspond to reactor vessel pressures of 273 psia and 615 psia, respectively.

The maximum local pool temperature of 202*F calculated for Case 3A (small break accident (SBA) accident mode assuming one RHR loop available) occurs at a time when the quencher mass fluxes are far below 42 lbm/ft 2-sec, which defines the region where the NRC limit is 204.8*F. Therefore, the maximum local pool temperature for this case lies below the NRC limit. As shown in Table 1-5.1-1 of the Fermi 2 PUAR, the maximum local pool temperatures of all other cases also remained below the NRC limit throughout the transient. Considering the degraded assumptions employed for each case and the conservatism of the NRC limit, the results ate considered acceptable, and unstable steam condensation would not be expected.

DET-04-028-1A A-29 Revision 0 nutggh

'Ouestion 9 l

The description of the suppression pool temperature monitoring system (SPTMS) which is provided in the PUAR is inadequate.

Additional information is needed to provide a clear demonstration that the Fermi 2 SPTMS design is in accordance with the requirements of AC Section 2.13.8.3.

Response to Question 9 The Fermi 2 suppression pool temperature monitoring system (SPTMS) design is discussed in the Fermi 2 PUAR, Section 1-5.2. .

In the design of the SPTMS, particular attention was given to the placement. of the temperature sensors to ensure that the system

]

would provide a conservative measure of bulk pool temperature and early operator notification of energy discharges into the pool.

i The considerations provided in the design included functional redundancy (dual element thermocouples), potential pool i circulation patterns, . the location of the RHR discharges, and identification of energy discharges (SRV, HPCI turbine exhaust, and RCIC turbine exhaust). Additional information to demonstrate that the SPTMS design is in accordance with the requirements of NUREG-0661, Appendix A, Article 2.13.8.3 has been provided in Table 1 and Figures 1 and 2 of this response.

P DET-04-028-1A A-30 Revision 0 t

nutagh

TABLE 1 Comparison of the FERMI 2 Suppression Pool Temperature Monitoring System with NRC Acceptance Criteria NRC Acceptance Criteria Fermi 2 Suppression Pool NUREG-0661 Appendix A Temperature Monitoring Article 2.13.8.3 System Design a) Each licensee shall demonstrate a) The Fermi 2 Suppression Pool that there is a sufficient number Temperature Monitoring System and distribution of pool tempera- (SPTMS) utilizes twelve (12) ture sensors to porvide a reason- dual element thermocouples able measure of the bulk tempera- installed in the suppression ture. Alternatively, redundant pool water space. In addition, pool temperature monitors may be there are four (4) dual element located at each quencher, either thermocouples installed in the on the quencher support or on the suppression pool air space.

torus shell, to provide a measure The distribution of the water of local pool temperature for each space thermocouples relative to quencher device. In such cases, Safety Relief Valve Discharge the Technical Specification limits (SRVD) quencher position is for local pool temperature shall shown in Figure #1. Eight (8) be derived frcm the calculated of water space thermocouples bulk pool temperature and the have been installed on the bulk to local pool temperature . torus shell at elevation 556'-1".

difference transient. The other four (4) thermo-couples are installed at eleva-tion 551'-4". Figure #2 provides a section elevation of the torus and shows the relative orienta-tion of the quencher and the water space thermocouples.

DET-04-028-1A A-31 R; vision 0 nutech

TABLE 1 (Cont'd)

Figure #1 shows that the place-ment of the thermocouples essentially provides an even distribution of temperature sensors. A thermocouple is located in the vicinity of a pair of quencher arms. The placement of the eight thermo-couples in the upper region of the pool will ensure measure-ment of the hotter water that is essentially pumped to the pool surface due to the re-circulation flow pattern in the pool established during-quencher discharge. These

,-m\

, temperature elements have also

( j) been appropriately placed in positions downstream of the RHR discharges. Therefore, the Fermi 2 SPTMS design will provide a conservative measure of the suppression pool bulk temper-ature.

7-t 1

('~' JET-04-028-1A A-32 Revision 0 nutp_gh

l TABLE 1 (Cont'd) w b) Sensors shall be installed b) The elevation of the thermo-sufficiently below the minimum couple placements relative to water level, as specified in the the suppression pool water level plant Technical Specifications, is shown in Figure 2. The to assure that the sensor pro- normal operating suppression perly monitors pool temperature. water level is at elevation 557'-0". The low-low supprs ,-

sion pool water level (minir water level) is at elevation 556'-10". The minimum sub-mergence of the SPTMS thermo-couples would be nine (9) inches.

Therefore, the temperature sen-sors are installed sufficiently below the minimum water level to provide proper temperature monitoring.

DET-04-028-1A A-33 Revision 0 nutp_qh

i l

}

TABLE 1 (Cont'd)

-- c) Pool temperature shall be indi- c) The Fermi 2 SPTMS design cated and recorded in the control provides the suppression pool room. Where the suppression pool temperature indication and temperature limits are based on recording in the main control bulk pool temperature, operating room. The indication and procedures or analyzing equip- recording devices are located ment shall be used to minimize on the panel providing the SRV t:e actions required by the actuation controls and poeition operator to determina the bulk indication. The panel display pool temperature. Operating is arranged to allow quick procedures and alarm set points operator identification of shall consider the relative suppression pool energy inputs, accuracy of the measurement reading of pool temperature system. values, and determination of the bulk pool temperature.

The suppression pool bulk temp-orature will be calculated based

( ,

on the eight (8) temperature sensor readings available in the upper region of the pool water space. The recirculation flow pattern established during quencher discharge moves the warmer water toward the pool surface. Therefore, utilizing the temperature values measured by the eight thermocouples will ensure a conservatively cal-culated bulk pool temperature.

An operating procedure will provide details and necessary steps such that minimum oper-ator action will be required to determine the pool bulk

[ _.aET-04-028-1A temperature.

A-34 R; vision 0 nutgch

TABLE 1 (Cont'd) d) Instrument set points for alarm d) The Fermi 2 technical speci-e shall be established, such that fications provides that the the plant will operate within plant will be operated under the suppression pool temperature the appropriate suppression limits discussed above. pool temperature limits. A copy of the applicable draft standard technical specifica-tion under development between the NRC staff and Detroit Edison is attached. In addition, an alarm in the main control room is annun-ciated when any one of the thermocouples at elevation 556'-1" measure a temperature of 105'F. The 105*F alarm set point will alert the operator very early during plant transient conditions of energy discharges into the pool and consequently will ensure that the suppression pool will be maintained within allowable suppression pool temperature limits.

DET-04-028-1A A- 3 5 R:: vision 0 nutggb

TABLE 1 (Cont ' d)

-- a ) All sensors shall be designed to e) The suppression pool temperature seismic Category 1, quality Group sensors (thermocouples) are B, and energized from onsite emer- seismically qualified. The gency power supplies. sensors are a passive element and do not require any power supply. The sensors are mounted on seismically qualified channels and supports, and the signal cables are routed in seismically qualified and supported trays and conduits to the main control room recorders. There are three multi-pen (12 pens) Strip-chart Recorders in the main control room and they are powered from onsite emergency

p. bus power supplies. These

( ) recorders are seismically qualified.

/ S i JT-04-028-1A A-36

~~R; vision 0 nutp_gh

O

- SUPPRESSION x N i CHAMBER 1

l E - E l l

E ,

E i

270 t 90*

6'-3' hN

@ d- '

RHR DIV.I i , RHR DIV.II DISCHARGE .

DISCHARGE SUPPRESSION / SP.V DISCHARGE CHAMBER T-QUENCH CENTERLINE 118 0 '

PLAN VIEW  !

hDUAL ELEMENT THERMOCOUPLES IN THE '#ATER SPACE AT EL. 556'-l' hDUAL ELEMENT THERMOCOUPLES IN THE WATER SP hDUAL ELEMENT THERMOCOUPLES IN THE AIR SPACE Figure 1 DISTRIBUTION OF SUPPRESSION POOL THERMOCOUPLES DET-04-028-1A Revision 0 A-37 nutp_qh

O

f. SUPPRESSION CHAMBER l

1

- 6' h NE t N 4

y SUPPRESSJ_ON i _

j _

E CHAMBER  !---

L / *,, m ~4 '

EL 557'-O' NORMAL EL 556'-1~ _ _

EL 556'-I- WgERL T EL 551'-4~ TE I~

4.EL_54B'4 B e-QUENCHER

1. 'DISHARGE DEV!CE E [ (f i

VERTICAL QUENCHER SUPPORT BEAM hDUAL ELEMENT THERMOCOUPLES IN THE WATER SPACE AT EL 55 hUAL ELEMENT THERMOCOUPLES IN THE WATER S hDUAL ELEMENT THERMOCOUPLES IN THE AIR SPACE Figure 2 PARTIAL SUPPRESSION CHAMBER SECTION RELATIVE ELEVATION OF SUPPRESSION PCOL TEMPERATURE SENSORS DET-04-028-1A Revision 0 A-38 nutggh

' T.::3 mir. urm rr.nu;..a .m. . .

• su .-

MhfE./. TION .: R0".THE AP2U0fdT . a t;. .: a CCNTAlf2fNf $YSTU*5 3 /4. 6. 7 PfpRE5 FUR 12AT1M SYSTD*5 $UPP't551M CMAnetts ITRTTio c afTP.DWc T6YTtT

.t._. S.....m........-..........;;=_....

a. Tha pool watar:

3 I

1. Volume between (87.600) ft and (39.500) ft . equfM1ent ta a level between (22' 0*) and (24'C*). and &

2. Maximum average temperature of 95'F during CPERATIDMAL CDWDITION l 1 or 2. except that the maxism everage tas9erature may be permitted ta increase to:

a) '105'F during testing which adds heat ta the suppression l chamber.

b) 110*F with THERMAL PCr=TR less than er equal to 2% of RATED l T1:E*. ML PC".TR.

c) 120*F with the main steaa line isolation valves closed l fe11cwin2 a scr:.c.

h. A total 1estage between the suppression chamber and drywell of less than the equivalent leakage through a 1 inch dia ster erifice at a differential pressure of 1 psig.

APPtict?tLTTY: OPERATIDMAL CQnDITIDMS 1. 2 and 3.

1* With the suppression chat.ber water level outside the above limits, restore the water level to within the limita m'tthin 1 have or be in at least ICT SHUTCa.N within the next 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> and in COLD SHUTDOWN wit.hin

, the. follcwin2 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.-

h. In CPERATIONAL'CCICITION'1 or 2 with the suppression et.a-mer average water toeperature geeste'r than $5'F. testore the average temperature to less than or et:ual to $5'F within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> or be in at least ICT SHIJTD0'.W within the next 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> and in CDLA SK11TC.H within the fe11swin2 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, sacept, as permitted above:

1. With the suppression chaftbar average water tesserature greater than 105*F during testing wnich asas heat to the suppression l chamber, stop all tasting which adds heat to the suppression chamber and restore the average temperature ta less than 15'F l within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> or be in at least NOT SHUTM.M within the next 12 heurs and in COLD SHUTDOW within the following 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.

2. With the suppression charher averagt water temperature greater than:

a) 90*F for more than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> and THERML POWER greater than 2% of RATED THEP %L PChit, or b) 210*F.

place the reactor mode switch in the Shutdown position and operate at Isast one residual heat removal locp (si the suppres-sten pool tooling mode.

3. With the suppressten chaser average water temperature greater vepressurize the reactor pressure vessel ta les2 l than 120*F.ig than 2r,0 ps within 12 taurs.

l "See 5pecificatten 3.5.3 for ECC3 requiraments.

FEEMI - tmIT 2 3/4 G-12 MAY Z 41982 (

l DET-04-028-1A Revision 0 A-39 nutd :

CONTAINMfWT SYSTEMS LIMITING CONOTTION FOR OPERATION (Continued 1 3: (Continued)

3. ifith the suppression chamber everage water temperature greater

.than 120*F, depressurize the resctor pressure vessel to less than 200 psig within 12 hoves.

s. ifith one suppression chamber water level instrumentation channel inoperable and/or with one suppression pool water temperature, 3 instrumentation channel inoperable, restare the (noperable cha.v nel(s) to CPERABLE status within 7 days er verify suppression.

chamber water level and/or temperature to be within the 1 faits at least once per 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />.

d. iff th both suppression. chamber water level instrumentation channels inoperable ar.4/or with more than one suppression p'ool water temperature instrumentation channel inoperable, restore at least one inoperable water level channel and seven temperature instru entation channels to OPERAILE status within 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> or 1,e in at least NOT SHUTDOWN within the next 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> and in CDLD SHUTDOWN within the fs11 ewing 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.

e. iff th the dryve11-to-suppression chaeber bypass leakage in exce?s of the 11 sit, restare the bypass leakage to wittin the limit prior to increasing reactor coolant temperature above 200*F.

l ) $URVE7LLA'rtf REQUIREMENTS J 4.5.2.1 The suppression chamber shall be demonstrated OPERABLE:

l a. By verifying the suppression chamber water volume to = within the limits at least once per 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.

b. At least once per 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> in OPERATIONAL CONDITION 1 or 2 by verifying the suppression chamber average water temperature to be i

less than er equal to 95'F, except:

L At least once per 5 minutes during testing which adds heat to the suppression chamber, by verifying the suppression chamber average water temperature less taan or equal to 105'F.

L At least once per have when suppression chamber average water temperature is greater than or equal to 95'F, by verifying:

a)~' Suppression cha:aber hverage water temperature to be less than or equal to 110*F and b) THERMAL POWER to be less than er equal to 2 of RATED l l THERMAL POWER after suppression chamber everage water l

temperature has escoeded 95'F for more than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.

3. At least once per 30 minutes following a scras with suppression chamber average water temperature greater than or equal to 95'F, by verifying suppression chamber average water temperature '

less than er equal ta 120*F.

fem! - unzT 2 3/4 6-13 WN ! T 1981 DET-04-028-1A Revision 0 A-40 l

nutggh 1

i

. .r.GCE ??EN PE*'Xu .il2PI U ' i

.  ; a, 2.- ;.

,isrC:l.'.l.Tl00. TOM THEd??!.Gi ..f.J iU-J ';

CO n w !aT SYSTEMS SURVEILLANCE REQUIREMENTS (Continued)

c. By an external visual examination of the suppression chaeter after l safety / relief valve operation with the suppression chammer average water temperature greater than or equal to 160'F and reactor coolant system pressure greater than 200 psig.

d. By verifying two suppression chamber water level instrumentation channels and eight suppression pool water temperature instrumentation channels CPER/JLE by performance of .a:

1. CHANNEL CMECK st least once per 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />,

2. CHANNEL FL?:CTIC!:AL TEST at Icest onen per 31 cays, and

3. CHtJ0t!L CALILAATIDM st le:st once per 18 e* nths, with the water level and taperature alarm setpoint for:

1. High water level 1 ( ).

2. Low water level 1 ( ),and

3. . High water t:sperature 1 (I:3)*F. (

e. At'16ast once per 15 months by concusting a cryvell-to-suppression chamber bypass laat test at en initial sifferential pressure of 1 psi and verifying it.:t the cifferential pressure does not Cecrease by more than 0.25 inchas of water p:r cir.ute for a period of 10 minutes.

If any dryweil-to-su;pression chtr.bc.r bypass lese test fails to aset the speciff sd lir.it, the test sc%gule f or subsequent tests shall be reviewed and approved by the Cr.nissiren. If two consecutive tests fail to seet the sp::ified licit, & test shall be perforr.ed at least every 9 ranths until two cc,nsecutive tests cert the specified limit, at which stee the is conth test schedule say be resusad.

l l

l l

l FERMI - UMIT 2 3/4 6-14 **f24IM l l

DET-04-028-1A Revision 0 A-41 nutggh

C)

N, /

' Ouestion 10.1 With regard to suppression cha.aber analysis, provide justifica-tion for not analyzing a 180' beam segment including the torus, columns, and seismic restraints as required by the criteria for considering the effect of seismic and other lateral loads. Also, discuss the implications of this approach with regard to stresses in the suppression chamber in the region surrounding the support columns.

Response to Question 10.1 The approach to evaluate suppression chamber lateral loads used e' x in the' Fermi PUAR results in total lateral loads which envelop m- those which would be obtained by a 180' beam model. As discussed in PUAR Section 2-2.4.2, maximum accelerations and dynamic load factors are used to develop bounding values of lateral loads for seismic loads and for asymmetric torus shell loads due to SRV discharge and pre-chug, irrespective of the suppression chamber frequency. Specifically, the maximum OBE spectral acceleration of 0.23g is used for seismic loads, the maximum dynamic load factor of 2.60 is used for SRV discharge loads, and the maximum dynamic load factor of 13.9 is used for pre-chug loads, each of which occurs at a different frequency. The resulting lateral loads for the seismic, SRV discharge, and pre-chug loadings are added absolutely to obtain a bounding value of the total f^'\

s- DET-04-028-1A A-42 l Revision 0 L

nutggb

suppression chamber lateral load, which is conservatively assumed to be transferred by two of the four seismic restraints.

The total lateral load which would be produced by a 180' beam model would be less since the response to each loading would primarily be determined at the dominant lateral frequency of the suppression chamber.

Lateral loads result in a shear effect and an overturning moment effect on the suppression chamber. The horizontal shear effect is the more significant component and is resisted by the seismic restraints shown in Figure 2-2.1-10. The overturning moment effect results in vertical loads which are resisted at each mitered joint by the suppression chamber columns and mitered joint saddles shown in Figure 2-2.1-4. The vertical leads on any one column / saddle assembly are small compared with those caused by the major torus shell loadings which primarily tct in the vertical direction, the results of which are shown in Tables 2-2.5-4. The corresponding stresses in the suppression chamber shell adjacent to the column / saddle assembly due to the overturning moment would also be small.

i DET-04-028-1A A-43 Revision 0 nutec_h

,m

./

Question 10.2 O)

With regard to the assumption that only 20% of the total mass of water in the suppression chamber contributes to lateral seismic loads, provide justification to indicate the applicability of the tests cited in the PUA report to support this assumption.

Response to Ouestion 10.2 For a suppression chamber partially filled with water subjected to a horizontal seismic excitation, a portion of the total water mass acts as a rigidly attached mass, while the remaining water mass acts in sloshing modes. The effective weight of water which acts as a rigidly attached mass was determined from 1/30 scale

\

N- generic tests performed as part of the Mark I program effort.

The scismic slosh test facility is identical and the test proce-dures used to determine the effective water weight are similar to the testo described in General Electric Report NEDC-23702-P,

" Mark I Containment Program Seismic Slosh Evaluation," dated March 1978.

The 1/30 scale model test facility is based on a prototypical Mark I suppression chamber whose geometric characteristics are very close to those of Fermi 2. Tests were performed with three different support stiffnesses (rigid, medium, and flexible),

which covered the range of stiffnesses and frequencies for all p

\

L DET-04-028-1A A-44 Revision 0 nutg.gh

Mark I plants, including Fermi 2. The analytical model developed for use in the referenced study predicts that 20% of the total mass of water acts as a rigidly attached mass with the suppres-sion chamber. This prediction was verified with the 1/30 scale model by comparing the measured frequencies of the test facility with those obtained from the analytical model. This was done by first adjusting the emptied test facility support stiffness to that necessary to obtain the frequency of the empty suppression chamber predicted by the analytical model. The test facility was subsequently filled with water to a height below the equator, and a series of tests were performed to determine the frequency. The resulting frequencies compared favorably with those obtained using the analytical model with the same assumed water height.

Therefore, the test results confirm the analytical results which showed that 20% of the total water mass acts as a rigidly attached mass. These results are considered applicable for use in evaluating the Fermi 2 suppression chamber response to seismic loadings.

The evaluation of the Fermi 2 suppression chamber for horizontal seismic loads is discussed in Section 2-2.4.2 of the PUAR. The seismic lateral load is conservatively calculated assuming that 20% of the total water mass acts at a maximum spectral accelera-tion of 0.23g. The remaining 80% of water is assumed to act at the maximum accelerations in the range of sloshing frequencies.

The methodology accounts for 100% of the water and results in a DET-04-028-1A A-45 Revision 0 nutp_qh

1 i

l

)

i i

i bounding value for the suppression chamber lateral load due to seismic loads.

t t

! i

!. . f i .

h- l i- i

' I i

1 i -

e: j t' -

1 i

l  !

i.  ;

i=

t I

I i

l

i 1

I.

1-4 i

it-b e

1 a 4

f. f

1.  :

i -

t j- I DET-04-028-1A A [

Revision 0  !:

e i  ?

t- r j-- . - . - . . - . - . _ _ . _ _ _ _ - ______ . . __ . _ _ . . -

) Question 10.3a d

Provide detailed calculations to indicate how the modal correc-tion factors given in Sections 1-4.2.3 and 2-2.4.1 of the PUA report are obtained.

Response to Question 10.3a Modal correction factors used in calculating the response due to SRV tcrus shell loads are obtained by dividing the response of an initial value or free vibration problem by that of a transient forced vibration response problem. The physical representations for the analogs used in developing these correction factors are shown in Figure 10.3-1. Two representations of source pressure

\s in a rigid tank are shown. The transient response problem consists of a rigid torus with a pressure source, P,B which is prescribed as a decaying cosine function, The initial value problera consists of a similar torus, which contains a spring and disk mechanism for providing an impulse to the surrounding pool water.

The analogs, described in terms of masses and springs, are shown in Figure 10.3-2 for the transient response or forced vibration problem and the initial value or free vibration problem. The torus system is described in terms of a generalized stiffness, ks, and a generalized mass, m's (i

t V DET-04-028-1A A-47 Revision 0 nutgdl

The forced vibration analog is subjected to an applied loading, described as

" e sw '

PB(D B B where P = bubt'.le force at time equal to zero, B

A = attenuatior. constant, w = frequency of the SRV bubble, and B

t = time.

The free vibration analog incorporates an additional mass and spring, representing the bubble system, which is utilized to establish a frequency of the bubble oscillation. The apparent or effective mass of the bubble is defined as m. g The numerical value of an apparent bubble mass is estimated by averaging the hydrodynamic mass for the case of an oscillating sphere in a still fluid and a fixed sphere in an oscillating fluid. The bubble stiffness is computed by multiplying the bubble mass by the bubble frequency squared.

Four spherical bubbles are assumed for the T-quencher discharge device. Single bubble stiffnesses are additive since the bubbles are assumed to act in phase (i.e., parallel springs).

The initial conditions for the free vibration analog are prescribed by compressing the spring, k B, a distance, al , such DET-04-028-1A A-48 Revision 0 nutp_gh

that when time is equal to zero the force in the spring is equal to P.

B Therefore, when time is equal to zero, both the forced vibration ano free vibration analogs have the same applied load magnitude.

Damping for the torus system can be described in terms of load attenuation and structural damping. Based upon test observa-tions, it is assumed that the structural response will be decayed in the same manner as the prescribed pressure. Accordingly, a decaying exponential function of the form e -A D is used to represent load attenuation and structural damping in the solution of the forced and free vibration system, respectively.

The equations of motion for the free vibration analog described in Figure 10.3-2 are obtained from free body diagrams for the structure mass and bubble mass as m - +

• ss (ks+ kB} s BB and m

BB -kXBB+kBs" '

where Xs = structure mass displacement, 5, = structure mass acceleration, X

B

= bubble mass displacement, and 5B = bubble mass acceleration.

e U DET-04-028-1A A-49 Revision 0 nutagh

The solution to these equations expressed in terms of the structure response, X s, is given as the following function, X

s

• fIl

' ' I' s s s s where e = bubble frequency and B

w, = structure frequency.

The other variables have been previously defined.

A family of curves which represents dynamic load factors as a function of frequency (w B! "s) is generated by assuming either "U IS BI 's) as a constant. Based upon estimates of the (kB/ms) significant modal characteristics of the torus and the oscil-18 lating bubble, a range of values for (kB /*s) and (mB /*s) established. The range of (k B/ms) values is estimated to be about 160 to 1,600. The range of (mB /*s) values is estimated to be 0. 03 to 0. 3.

Figure 10.3-3 contains a comparison of DLF's for the cases with (kB/*s) equal to 160 and 1600, assuming a structural frequency of 20 Hz. The maximum DLF for the case with (k B /ms) equal to 160 is 2.8, whereas for (kB/*s) equal to 1600, the maximum DLF is 2.0.

It was determined that the DLF's for the case with (kB/*s) equal to a' constant are about 20% to 60% larger than the DLF's for the

. case with (mB /ms) equal to a constant.

DET-04-028-1A A-50 Revision 0 nutp_qh

The equation of motion for the forced vibration analog 'escribed in Figure 10.3-2 is obtained from the f ree body diagram for the structure mass as m,X, - k,X, = PB e cos w B

• The solution to this equation is given in the following function. .

Xs" s

• f 2I "B'"s'AI*

Figure 10.3-4 contains DLF's plotted as a function of (wB !"s

• A family of DLF curves is included for structure frequencies of 11, 14, and 20 Hz. The DLF's for the resonant condition range from 5.4 to 9.9 for structural frequencies between 11 and 20 h Hz. The forced vibration DLF's are approximately three to five times.the free vibration DLF's.

Correction factors are obtained by dividing the free vibration response to the system by the response of the forced vibration system. The sensitivity of the correction factor to the

variables (kb/"s'"B/m s,A) is evaluated in order to determine a valid set of correction factor curves to be used in design.

Based upon the above evaluation, it was determined that the lower range of (kg /ms ) should be used for determining conservative design basis modal correction factors. An attenuation factor of D

DET-04-028-1A A-51 l- Revision 0 nutagh

6 is selected since use of this factor results in correction factors which bound the response at resonance conditions for all structural frequencies.

A typical correction factor curve is shown in Figure 10.3-5 for a structural frequency of 20 Hz. For the plant unique analysis, a set of enveloping correction factors is generated for dif ferent modal frequencies of interest, as shown in PUAR Figure 2-2.4-5.

Table 10.3-1 contains a comparison of analytical results obtained using modal correction factors and measured results for Monti-cello. The results shown are obtained by dividing analytical results by test results for key response parameters. The compar-isons show that modal correction factors provide a conservative basis for calibrating the analytical model used to evaluate the response of the Fermi 2 suppression chamber for SRV torus shell loads. The modal correction factors are developed at test conditions and applied at design conditions in accordance with NUREG-0661.

DET-04-028-1A A-52 Revision 0 nutgch

Table 10.3-1 O' CORRECTED TRANSIENT RESPONSE ANALYSIS NORMALIZED BY TEST RESPONSE TEST CONDITION LOCATION COMPONENT COLD POP HOT POP a) STRESS INTENSITY 78* FROM INSIDE SHELL MEMBRANE 1.2 1.3 EQUATOR MIDBAY

.78

• FROM OUTSIDE SHELL MEMBRANE 1.4 2.3 EQUATOR MIDBAY b) COLUMN REACTION Og, INSIDE SUPPORT UPLOAD 3.7 3.0 COLUMN INSIDE SUPPORT DOWNLOAD 2.8 2.0 COLUMN OUTSIDE SUPPORT UPLOAD 4.4 3.5 COLUMN OUTSIDE SUPPORT DOWNLOAD 2.7 2.8 COLUMN DET-04-028-1A A-53 Revision 0 nutagh

e it) a P ge"A8co. m g 5

.Y@%%

/.

4 o r T >

....) j

' /

~

P.

.) TRANSaENT OEFMTON AT T!L*E TERO: Pg a  ! -- --

4 5

/

+ .

v.

b) NTIAL VALUE DEFMTION Figure 10.3-1 SRV LOAD DEFINITION DET-04-028-1A Revision 0 A-54 nutagh

.- . _ . . = - - . .. - - _ . . _.. _. _ . . . .

\

4 SU88LE P SIO *E

'I X g ks o

FLuto FLulO ms ms TORUS TORUS

=. =.

(d (t)

FORCED V18 RATION ANALOG FREE V!BRATION ANALOG i

i l

l l

Figure 10.3-2 l

STRUCTURAL RESPONSE ANALOGS DET-04-028-1A Revision 0 A-55 nutggb 1

l

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

O it,i,i, iii.o h 4 ;,i i, ii, i ,

! ,, l i i

• n- fTlTl' T lil Tnfb I!I 11

• i 44 O

g t

1... ..f

! t ii* f t *

. .t 4 * *

..I .. i t

, , i

=l j LEGEND: ll l l l 1'

= f, ... ......,.......  ;,; ;;,; ';;,;;  ;;;

,,, i lllllllil!Illlllli Ill Illi tilllill!

0 .5 1.0 1.5 2.0 er g , e's

)

i I

1 l

Figure 10.3-3 FREE VIBRATION ANALOG OYNAMIC LOAD N T FACTORS FOR (k B !*s DET-04-028-1A ,

Revision 0 A-56  :

)

i nutp_gh l

,/~N k

10.0 i ii.ii i! ej l . . .+

i iiiii

.,ii e i i i is i  %,. .ii 5

G

= t .O l  ! ii n ill!

W! lll11 a

$ LaceN0;

!!. i i ,,

3 T- i .. ii A y [ '"8 l lll ll ll lil!l ll; g ;a g...........".'*."t..., .

m ---- .. 20. n, Y _tii T1 ,i ..

Of )i l ll l)l l l

' f fi if 'i 0 .5 1.0 1.5 2.0 m, o w.

Figure 10.3-4 FORCED VIBRATION ANALOG DYNAMIC LOAD FACTORS

, DET-04-028-1A l

V Revision 0 A-57 l

i nutggb

- - - ---m-.,r. , _ , . _ . _ _ _ _ , . _ . , . _ , , _ . _ _ _ _ _ _ _

O 1.0 I I I II I- 1III I l l l l Il I l l l l NLJ ,s l I J. vr I i I 5

lIII IkI l/ lIllT IIII -l \1 1; 14 l 'IIII '

= 11 II N / l I I IIII E IIII'l l l\ l /- 1IIIIII I IIII I I T30l l l l l l l 8

III I IIIII I I l lI I I I II I I i i iI I iii 1 1 I i 0 .5 1.0 1.5 2.0 eB / #s Figure 10.3-5 TYPICAL MODAL CORRECTION FACTOR DET-04-028-1A Revision 0 A-58 nutsch

Ouestion 10.3b Y

Provide justification for the applicability of these factors to

" multi-degree of freedom" systems since the factors were developed using simple "one degree of freedom" systems.

Response to Question 10.3b The transient response of the Fermi 2 suppression chamber due to SRV torus shell loads is obtained using the modal superposition method. Using this approach, the equations of motion for a multi-degree of freedom system, such as the-suppression chamber, are decoupled into a set of equivalent single degree of freedom Each structural frequency or mode is represented by a

{ \

systems.

-single degree of freedom system. The responses of each single degree of freedom system are summed to obtain the total response of the suppression chamber.

During the summation process, modal correction factors obtained from PUAR Figure 2-2.4-5 are applied to the response of each single degree of freedom system for each suppression chamber frequency. As discussed in the response to Question 10.3a, the modal correction factors are developed using single degree of f reedom systems and are compatible for use in the modal super-position method since this method is completely linear.

O t}'

DET-04-028-1A A-59 Revision 0 nutagh

Question 10.4 .

Provide justification for not considering the effect of bending moments in' column analysis using interaction formulae.

Response to Question 10.4 Consideration of column bending moments using the interaction formula is necessary when compressive stresses or column buckling is a concern. Since the suppression chamber support columns for Fermi 2 are heavily reinforced and braced continuously along their length, as shown in Figure 2-2.1-7 of the PUAR, the effects of buckling are negligible. Furthermore, bending moments in the support columns are small since the columns are permitted to s, slide horizontally at their base. The support column loads shown in PUAR Table 2-2.5-4 are substantially below the allowable compressive loads for the support column wide-flange sections with cover plates.

a

/

y DET-04-028-1A A-60 Revision 0 nutggh

[ Ouestion 10.5 With regard to the suppression chamber columns, provide justifi-cation and/or additional information to indicate why a non-linear time history analysis was not performed as required by the cri-teria when net tensile forces are produced in the columns. Table 2-2.5-2 of the PUA report indicates that net tensile forces are produced in the columns.

Response to Question 10.5 The criteria requirements for performing a non-linear time history analysis are applicable for plants in which the suppression chamber and its supports are not anchored to the b) basemat. Such a condition would result in gross nonlinear behavior if uplift loads exceeded the weight of the suppression chamber and con'ained t water.

The Fermi 2 suppression chamber is fully anchored to the basemat at each mitered joint column and saddle base plate location, as shown in PUAR Figures 2-2.1-7 and 2-2.1-8. Although tensile forces are produced in the column and saddle supports, the tensile forces are within the allowable anchorage capacity of the support system, as shown in PUAR Table 2-2.5-4. The requirements for a nonlinear analysis, therefore, need not be evaluated for Fermi 2 since the suppression chamber is fully anchored to the

(

O)

!tv DET-04-028-1A A-61 Revision 0 nutggb

~

basemat and the effects of non-linearities on the overall suppression chamber response have been minimized.

O

- .e..e2e.1.

Revision 0

..., e nutgg.h

m

/

}

Question 10.6 v

Provide justification for using two different temperatures, 173*F for suppression chamber and vertical support systems and 100*F fo'r the base plate of the support system, for calculating the respective allowable stresses.

Responsa to Ouestion 10.6 The allowable stresses for the suppression chamber and its vertical sunports are -conservatively determined at 173*F since this is the maximum temperature specified for any LOCA event, as shown in PUAR Figures 2-2.2-4 through 2-2.2-6. The allowable stresses for the vertical support system base plates are

\v/ determined at-100*F, which bounds the maximum temperatures of the base plate expected during the specified events. There may be long-term conditions which result in higher base plate temperatures; however, base plate temperatures higher than 100*F are not expected to occur during times of peak transfer of hydrodynamic loads to the suppression chamber vertical support -

system. Furthermore, the allowable stresses at 100*F and 173*F are not significantly different.

O)

(

v' DET-04-028-1A A-63 Revision 0 nutgq.h

l l

l

, O\

.\

Ouestion 11.1 l

Provide justification for using the SRSS method to combine the SSE and LOCA responses for SRV piping analysis instead of the absolute sum or cumulative distribution function approaches, as required by the criteria.

Response to Question 11.1 The method of combining responses due to LOCA and SSE loads for SRV piping described in Section 5-2.2.3 of the PUAR is based on NUREG-0484, Revision 1, " Methodology for Combining Dynamic

' Responses," published in May 1980. The original issue of NUREG-4 0484 justified combination' of responses due to LOCA and SSE within the reactor coolant pressure boundary using the SRSS technique. The current Revision 1 has extended the application of this combination technique to include ASME Code,Section III, Class 1,2, and 3 systems, components, and supports. As described in Revision 1 of NUREG-0484, use of the SRSS technique provides a non-exceedance probability of 84% or higher. Since the Fermi 2 SRV piping is analyzed as Class 2, the use of the SRSS method is deemed acceptable based on Revision 1 of NUREG-0484.

OV DET-04-028-1A A-64 Revision 0 nutgrb

Question 11.2 Provide justification for using Markl's equation for fatigue analysis of SRV piping instead of the SN curve given in ASME Code,Section III, Division 1 Appendices.

Response to Question 11.2 The methodology for evaluating Fermi 2 SRV piping fatigue was originally presented to the NRC during a meeting in June 1981, and was followed by a letter submitted to the NRC shortly thereafter (Reference 11.2-1). Section 3.9.3 of Supplement 1 of the Fermi 2 Safety Evaluation Report (NUREG-0798) references the required fatigue evaluation and the proposed methodology.

'%..]

Since the SRV piping is a Class 2 system, the approach outlined in the presentation and letter was to evaluate fatigue using ASME Code, Class 2 piping rules as a guideline. The proposed methods included extension of the Class 2 equations and curves used for thermal fatigue evaluation to include all cyclic loads. A comparison of the extended Class 2 method to a Class 1 fatigue analysis was also provided, which showed that the two methods yield similar results.

i Using the propsed methods, a fatigue usage factor is determined for each of the cyclic loadings. For Mark I LOCA-related loads,

\_/ DET-04-028-1A A-65 Revision 0 nutggb

estimates of total stress cycles during plant life would be determined and associated fatigue usage would be calculated.

Since only very conservative estimates of the number of SRV discharge-related stress cycles were available, the approach proposed that SRV actuations would be monitored to assure that the allowable fatigue usage was not exceeded.

Following Detroit Edison Company's commitment to the NRC to perform a SRV piping fatigue evaluation, the matter was discussed between the NRC and the Mark I Owners Group. These discussions resulted in a commitment by the Mark I utilities to perform fatigue evaluations for SRV piping in the torus and for torus attached piping systems as part of the plant unique analyses.

Discussions among Mark I owners and their AE's followed and a task force was formed to develop a generic approach for fatigue evaluation. The approach agreed upon was a method which extended the Class 2 piping fatigue rules similar to the methods initially proposed for Fermi.

Refinements to the proposed Fermi 2 methods which have been incorpot ced into the generic approach consist of the following:

o Fatigue usage is evaluated based on considering critical loading combinations instead of on an individual load basis.

O DET-04-028-1A A-66 Revision 0 nu

o Total cumulative fatigue usage for all cyclic loadings is calculated in lieu of monitoring SRV actuations.

o The allowable number of stress cycles is determined by using Markl's equation (Reference 12.2-2) in lieu of the Class 2 thermal fatigue equation basis. (Markl's-equation forms the basis for Class 2 piping fatigue and was u-i in developing the Class 2 piping stress intensiti. on factors).

o Actual stress cycles for a given response time history are converted into equivalent full strass eg cycles using the methodology defined in Section NC-3611.2(e)(3) of the Code.

The SRV piping fatigue evaluation performed for Fermi 2 and documented in Volume 5 of the PUAR includes the extended Class 2 approach originally proposed for Fermi 2, and incorporates the additional refinements included in the generic Mark I approach.

The refinements result in a more practical, comprehensive method of evaluation for fatigue.

Reference 11.2-1 - Detroit Edison Company Letter EF2-53,824 to NRC, dated June 22, 1981.

, s DET-04-028-1A A-67 Revision 0 nutagh 1

- - - . ,m. . . . . -.---%,----w,.,---,-.--wv----%--,ww---+-e ,y,,we-,, , -. - - , - , , - - - - ,y y*,-,,-- v ---w-----tvv-------r---.-

Reference 11.2-2 - Markl, A.R.C., " Fatigue Test of Piping Components," Transactions ASME, Volume 74.

l l

l I

O l

DET-04-028-1A A-68 Revision 0 nutach .

Question 11.3 Provide justification for not using Equation 11 of ASME Code,Section III, Subsection NB for calculating the fatigue stresses, and explain the method used.

Response to Ouestion 11.3 Justification for not using Equation 11 of the ASME Code,Section III, Subsection NB (Class 1 pipino) is provided in the response to Ouestion 11.2. Equation 11 of Subsection NC (Class 2 piping) of the ASME Code provides combination methods for thermal and other sustained loads used in evaluating for fatigue. The 4

n V

methods applied in the Fermi PUAR extended the traditional usage of Equation 11 to the combination of stresses due to dynamic cyclic loadings, using the same method of absolute summation of stresses.

l l

l c

n V DET-04-028-1A Revision 0 A-69 l nutagh

Question 11.4 O.

Provide justification and reference for the maximum-stress cycle factors given in Table 5-2.4-4 of the PUA report.

Response to Question 11.4 See the response to Question 11.2 for a description of the methodology for evaluating Fermi 2 SRV piping fatigue. The basis for developing R factors used to determine maximum equivalent full stress cycles is derived from the Class 2 piping thermal fatigue techniques defined in Section NC-3611.2(e)(3) of the Code. The R factors for individual dynamic cyclic loadings also take into account consideration of loading characteristics such as frequency, time history, and random phasing of load Components.

f l

l l

6 i

l l

d DET-04-028-1A Revision 0 A-70 nutggh I

l 1

Question 11.5 Provide the magnitudes of the dynamic load factors used in Tables 5-3.2-1 and 5-3.2-3 of the PUA report and the justification.

Response to Question 11.5 The dynamic load factors (DLF's) included in the loads specified in Tables 5-3.2-1 and 5-3.2-3 of the PUAR are summarized in Table 11.5-1.

The loading functions for water jet impingement loads, T-quencher thrust loads, and T-quencher endcap thrust loads are defined as rectangular pulse loadings. The maximum DLF specified by standard structural dynamics handbooks for this type of load function is 2.0.

The DLF's for SRV air bubble drag loads were determined using Monticello in-plant test data, as permitted by NUREG-0661. The criteria state that actual measured pressure waveforms determined in tests may be used to develop a maximum structural amplifica-tion for resonant conditions. Using the measured Monticello pressure waveforms, a maximum DLF cf 3.0 at resonant conditions

(

was developed and is used for structures whose natural frequency is within the 4.0 to 14.0 Hz frequency range of the SRV air bubble drag loads. For structures whose natural frequency is f

i l

i - DET-04-028-1A A-71 Revision 0 nutsch

. ,_. ____.___ _,._ _ .,_ . . _ . _ _ _ _ . . _ _ _ _ _ _ _ _ _ _ _ _ _ _ - , - . _ _ _ _ _ _ , _ _ , _ _ _ _ _ _ _ , _ _ _ _ _ _ _ _ _ , _ . . _ , _ _ . ~

well above the maximum air bubble drag load frequency, a DLF of 2.0 is conservatively used. The natural frequencies of the T-quencher and it supports and the submerged SRV piping are sufficiently above the maximum air bubble drag load frequency, as shown in PUAR Figures 5-3.4-3 through 5-3.4-6.

O l

DET-04-028-1A A-72 Revision 0 nutggh

g Table 11.5-1 PUAR DYNAMIC LOAD FACTORS FOR SRV PIPING, T-OUENCHERS, AND T-QUENCHER SUPPORTS l

i PUAR TABLE LOAD TYPE NUMBER DYNAMIC LOAD FACTOR (DLF)

SRV WATER JET IMPINGE- 5-3.2-1 2.0 MENT SRV AIR BUBBLE DRAG 5-3.2-1 and 2.0 5-3.2-3 3 T-OUENCHER ENDCAP 5-3.2-2 2.0 THRUST LOAD l-t

/

DET-04-028-1A A-73 Revision 0 nutagh

Question 11.6 Provide the results of the analysis of bolted or welded connect!.ons associated with the SRV piping.

Response to Question 11.6 The tables in the PUAR which contain analysis results for wetwell SRV piping major support connections and welds are shown in Table 11.6-1. The referenced tables show that SRV piping wetwell support connection stresses are within allowable limits.

l l

l r

b l DET-04-028-1A A-74 -

Revision 0 .

l nutggb

(

Table 11.6-1 PUAR TABLE REFERENCES FOR WETWELL SRV PIPING SUPPORT CONNECTIONS AND WELDS l

SUPPORT PUAR TABLE NUMBER CONNECTION / WELD STRESS VENT LINE-SRV ~

~

PIPING PENETRATION AND WELDS SRV PIPING VENT ~

LINE AND VENT HEADER SUPPORTS RAMSHEAD AND 5-3.5-4 O

T-OUENCHER ARM SUPPORTS AND WELDS DET-04e028-1A A-75 Revision 0 nutech

l O Question 12.1 Provide justification for the method of lumping additional fluid masses along the ring beam, quencher beam (Page 2-2.103 of PUA report), submerged length of SRV piping, T-quencher, and supports (Page S-3.49 of PUA report) as indicated in the PUA report.

Response to Question 12.1 The hydrodynamic masses used for evaluating submerged structures are calculated using the relationships contained in PUAR Table 1-4.1-1, which are taken from LDR Table 4.3.4-1 (Reference 12.1-1). For the SRV piping, ramshead, T-quencher arms, 6" diameter lateral support members, and 20" diameter lateral

(/ support beam, the hydrodynamic mass for the T-quencher arm ring plate supports are calculated using the equation for a circular f

disk. For the ring beam and vertical quencher support beam, the hydrodynamic mass is calculated using the equation for a plate in the lateral direction and an I-beam in the vertical direction.

Reference 12.1-1 -

General Electric Report NEDO-21888, Revision 2, " Mark I Containment Program Load Definition Report," dated December 1981.

f

'l b DET-04-028-1A A-76 Revision 0 nutggb .

m

) Question 12.2

< J Provide justification for not considering the loads indicated in Table 1 which are required in the analysis according to NUREG-0661.

Response to Ouestion 12.2 All loads specified by NUREG-0661 were addressed in the PUAR.

The loads identified in Ta le 1 unich are not included in Table 1-4.3-1 of the PUAR can be categorized as being negligible, not applicable to the Fe rr.i plant, or considered in the analysis.

The loads circled in Table 1 are discussed in the following paragraphs.

V 4.3.5 Froth Impingement: The froth impingement loads on the torus shell are negligible as indicated in LDR Section 4.3.5.1. The torus support system will also have negli-gible effects due to froth impingement. For SRV piping, I the portion below the vent header is protected from pool swell impact loads by the vent header deflector. The i

portion below the vent line experiences negligible loads due to froth impingement.

t

! 4.3.8 LOCA Bubble Drag: The vent header support columns are the only structures above the bottom of the downcomers

'N DET-04-028-1A A-77 Revision 0 nutgsh

and below the normal water level. The LOCA bubble drag loads on these columns are contained in PUAR Table 3 -2. 2-9.

4.5.3 Chugging Vent System Loads: The chugging loads on the main vent and vent header were considered in the analysis and are contained in PUAR Table 3-2.2-19.

5.2.5 T-quencher Air Bubble Drag: The SRV air bubble drag loads on these structures were considered in the PUAR.

The SRV air bubble drag loads on the downcomers are given in PUAR Table 3-2.2-22. The SRV air bubble drag loads on the T-quencher and the SRV piping are given in PUAR Table 5-3.2-3, and the SRV air bubble drag loads on

'the T-quencher supports and vent header support columns are given in PUAR Tables 5-3.2-1 and 3-2.2-23, respectively.

5.2.6 Thrust Loads on T-quencher Arms: The thrust loads on T-quencher arms are given in PUAR Table 5-3.2-2.

5.2.7 SRVDL Environmental Temperatures: The SRV discharge line envirnnmental temperature loads are discussed on page 5-3.17 of the PUAR.

DET-04-028-1A A-78 Revision 0 nutgch

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

i l

i i-5.3 Ramshead Loads: Ramshead loads are not applicable for Fermi 2 since the SRV lines are equipped with T-quencher i

discharge devices rather than ramsheads. ,

4 i

i f

l 1

!G '

t i

i j

I t

I i

i e i

4 4

f 1

DET-04-028-1A A-79

. Revision 0

. nutagh

,O Question 12.3 i

\

Provide information on the analysis of the attachment welds of ,

regions connecting the internal structures to the torus shell, indicating whether the criteria requirements have been satisfied.

Response to Question 12.3

. The internal structure attachment welds to the torus shell have been evaluated in accordance with the criteria requirements.

Table 12.3-1 shows the most highly stressed catwalk and monorail support pad plate attachment welds to the torus shell. The load combinations for which the welds are evaluated are presented in PUAR Table 4-2.2-2. The welds are evaluated using the ASME Code s criteria contained in Subsection NE for Class MC components.

This table shows that the internal structure attachment weld stresses are within allowable limits.

'DET-04-028-1A A-80 Revision 0 nutggb

Table 12.3-1 CATWALK AND MONORAIL SUPPORT PAD PLATE WELD STRESSES RATIO OF ITEM CALCULAT STRESS ALLOWABLE STRESS CALCULATED TO (ksi) ALLOWABLE I

CATHALK 3*46 15.01 0.23 PAD PLATE WELD MONORAIL 0.49 15.01 0.03 PAD PLATE WELD

\

O t

l DET-04-028-1A A-81 Revision 0 nutagh

i

[}

-% I Question 13 i PUAR Section 2.2.2.1 (Page 2.2-30), AC Sections 2.3 and 2.4.

Additional information is required concerning the torus shell pressures presented in Table 2-2.2-3 on page 2-2.48 of the PUAR. Provide the details of a specific torus shell pressure calculation at the two times specified in the table for a typical longitudinal location as a function of circumferential location (e.g., 2/1 = 0, 0 = 180, 150, 120, and 90 degrees). The follow-ing information should be included as part of the response, with and without the margins imposed by NUREG-0661.

(a) Net torus vertical load history

/"'g (b) Average submerged pressure history m- (c) Torus airspace pressure history Illustrate how these pressure histories are used in conjunction with the Load Definition Report (LDR) multipliers to arrive at the values presented in the table.

Response to Ouestion 13 As discussed in the response to Question 14, the transient structural analysis of the Fermi 2 suppression chamber for pool swell loads is performed using net pressure loadings which are obtained by subtracting the airspace pressure transient from the 1

es

)

\_ / DET-04-028-1A A-82 Revision 0 nutagh

l submerged pressure transient. The net pressures used in the structural analysis include the margins imposed by NUREG-0661.

PUAR Figure 2-2.2-8 shows an example of a resulting net pressure transient used in the analysis.

Table 13-1 provides additional information to clarify PUAR Table 2-2.2-3. Table 13-1 also shows the torus shell pressure values used in the Fermi 2 structural analysis, which were calculated in ,

1 accordance with NUREG-0661 requirements. The bases for the values in Table 13-1 and PUAR Table 2-2.2-3 are described in the following paragraphs.

The sample pressure values shown in PUAR Table 2-2.2-3 were obtained by taking the net pressure loads used in the structural analysis of the suppression chamber, including the NUREG-0661 margins, and adding the torus airspace pressures obtained directly from the PULD curves without NUREG-0661 margins. For ease of review, the pressures shown in Table 2-2.2-3 were reported at the same longitudinal locations as the locations at which the Load Definition Report (LDR) longitudinal multipliers are specified. The LDR specifies values at five 2/1 locations for calculating longitudinal multipliers. Intermediate values used in the Fermi 2 analysis are obtained by interpolating and enveloping the LDR values. Each longitudinal multiplier is conservatively applied over a range of 2/1 values.

DET-04-028-1A A-83 Revision 0 nutp_qh

/ ) Table 13-1 shows torus shell pressure components which include all the NUREG-0661 margins. Table '3-1 also shows locations at which the LDR longitudinal multipliers were obtained and the range over which each multiplier was applied. A description of how the values in Table 13-1 were obtained is provided in the following paragraphs.

The average submerged pressure time-history the torus airspace pressure time-history, and the net torus 5 nrtical load time-history without the margins imposed by NUREG-0661 are given in Fermi 2 PULD Figures 4.3.2-2, 4.3.2-4, and 4.3.1-2, respectively.

The pressure vslues at the time of peak download (t = 0.3 sec) and at the time of peak upload (t = 0.54 sec) are given in Table g-~ 13-2. The net torus load pressures shown in Table 13-2 are

\s_s/ calculated by subtracting the airspace pressure values from the average submerged pressures.

The average submerged pressures and the torus airspace pressures with the NUREG-0661 margins applied are provided in Table 13-3.

For the download, a margin of 10% of the net torus vertical pressure was conservatively applied. For the upload, a margin of 21.5% was applied as required by NUREG-0661. The download margin is applied to the average submerged pressure curve during that 4

portion of the time-history when a net download is acting on the torus, while the upload factor is applied to the torus airspace pressure curve during that portion of the time-history when a net upload is acting on the torus.

f_)

\

DET-04-028-1A A-84 Revision 0 nutggh

Torus shell pressures for a typical longitudinal location ( Z /1 =

0.552) are shown in Table 13-4. This table shows that local pressures at each circumferential location are calculated using the relationship Ploc = (P,yg)'g xMz*MO. The pressures (P avg I'M are obtained by subtracting the airspace pressure from the average submerged pressure, as shown in Table 13-3. The pressure values shown in Table 13-4 are the same as in Table 13-1.

Table 13-1 provides additional information to clarify PUAR Table 2-2.2-3 and shows the net pressures used in the structural analysis, which includes the margins imposed by NUREG-0661.

O 1

l DET-04-028-1A A-85 Revision 0 nutg,gh

^

)

O Table 13-1 l j TORUS SHELL PRESSURES DUE TO POOL SWELL AT KEY TIMES AND SELECTED LOCATIONS l o

n. o

- I _,.

8

_ _ _ 270* - 9o

_ _ _ _-J

n.

o.o c.s 1.0 fo o Fey Diagram LONGITUDINAL IDCATION (2/1)

CIRCUMFERENTIAL TORUS SHELL PRESSURE (psi)

IONGITUDINAL LOCATION (0)(deo)

FACIOR APPLICABLE PEAK DOWNIDAD PEAK UPIDAD IDCATION RANGE (t=0.30 see) (t=0.54 see)

CN 0.000 .361 '180 10.7 3.6 MAXIMCM AT 0.000 .361 150,210 9.7 3.4

(% 0.0 OR 0.361 0.000 .361 120,240 6.0 2.4 0.000 .361 0-90,270-0 0.3 7.5 FACIOR 0.361 .500 180 11.6 4.0 INTERPOIATED 0.361 .500 150,210 10.5 3.8 AT 0.50 0.361 .500 120,240 6.5 2.9 0.361 .500 0-90,270-0 0.3 7.5 0.500 .640 180 11.9 4.0 0.552 0.500 .640 150,210 10.8 3.9 l

0.500 .640 120,240 6.5 2.9

! 0.500 .640 0-90,270-0 0.3 7.5 FACIOR 0.640 .810 180 12.4 4.1

INTERPOIATED 0.640 .810 150,210 11.2 3.9

) AT 0.724 0.640 .810 120,240 6.9 3.0 0.640 .810 0-90,270-0 0.3 7.5 0.810 - 1.0 180 13.0 4.1 0.895 0.810 - 1.0 150,210 11.7 4.0 0.810 - 1.0 120,240 7.2 3.1 0.810 - 1.0 0-90,270-0 0.3 7.5 i

\

DET-04-028-lA A-86 Revision 0 nutggb

Table 13-2 PRESSURES AT TIME OF PEAK DOWNLOAD AND PEAK UPLOAD WITHOUT NUREG-0661 MARGINS AVERAGE SUBMERGED AIRSPACE NET TORUS PRESSURE PRESSURE LOAD PRESSURE TIME_

P avg (Psi) P a (Psi) P net (Psi)

PEAK DOWNLOAD 8.8 0.3 8.5 (t = 0.3 sec)

PEAK UPLOAD 3.6 6.8 -3.2 (t = 0.54 sec)

O MET.e.-e2e-1A Revision 0 A-e>

nut d

Table 13-3 PRESSURES AT TIME OF PEAK DOWNLOAD AND PEAK UPLOAD WITH NUREG-0661 MARGINS AVERAGE PRESSURE AVERAGE SUBMERGED AIRSPACE FOR CALCULATING PRESSURE PRESSURE LOCAL PRESSURES (psi)

TIME (P,yg) g (psi) (P aI M (Psi) (P,yg)'g=(P,yg)M-(P3)M 4

(1)

PEAK DOWNLOAD 8.8 + 0.1 x 8.5 0.3 9.3

(t = 0.3 sec) = 9.6 (2) 3.6 6.8 + 3.2 x 0.215 -3.9 PEAK UPLOAD = 7.5 (t = 0.54 sec)

(1) AT THE TIME OF PEAK DOWNLOAD (t = 0.3 sec)

(Pavg) M=Pavg + 0.1 x Pnet (2) AT THE TIME OF PEAK UPLOAD (t = 0.54 sec)

(P,)g =-P, + 0.215 x Pnet i

l DET-04-028-1A A-88 Revision 0 nutggb

,rw r,, . . , - - - - , , .ve-- .:..,,,e.,,..,.,.,,------a-._,---..,vr-,.- - , - - - . . . , , , - - . - , . . - - . - - + - - -

l l

Table 13-4 TORUS SHELL PRESSURE CALCULATIONS DUE TO POOL SWELL FOR A TYPICAL LONGITUDINAL LOCATION (Z/1 = 0.552)

CIRCUtlFERENTIAL LOCAL PRESSURES LOCAL PRESSURES TIME FACTOR FACTOR (psi) PLUS l (sec) g (2)

AIRSPACE LOCATION III g (2) 1 p

(0 ) (deo) Z 0 loc ,gpavo),Mxgzxg e PRESSURE (psi) 180 0.30 1.040 1.205 11.6 11.9 180 0.54 0.996 0.908 -3.5 4.0 150,210 0.30 1.040 1.083 10.5 10.8 150,210 0.54 0.996 0.940 -3.6 3.9 120,240 0.30 1.040 0.638 6.2 6.5 120,240 0.54 0.996 1.186 -4.6 2.9 0

0-90, 270-0 0.30 - -

0.3 0.3 0-90, 270-0 0.54 - -

7.3 7.5 (1) FOR CIRCUMFERENTIAL LOCATIONS, SEE PUAR TABLE 2-2.2-3.

(2) THESE FACTORS ARE TAKEN FROM THE LDR, TABLE 4.3.2-1.

DET-04-028-1A A-89 Revision 0 nutggb

r

[ Ouestion 14 V

s PUAR Section 2-2.2.1 (Page 2-2.30), AC Sections 2.3 and 2.4.

Describe in detail how the peak download and peak upload values presented in Figure 2-2.2-8 on page 2-2.65 of the PUAR were determined. Provide any additional inf orination required to duplicate these results which has not already been requested above. In addition, describe how this transient is used in the dynamic analysis of the torus shell loads.

Response to Question 14 The values presented in PUAR Figure 2-2.2-8 are for the mitered joint location (Z/1 = 0.5) and at bottom dead center (0 = 18@ ).

V For a given location, the maximum and minimum values are obtained at the times of peak download and upload (t = 0.3 and 0.54 sec) by subtracting airspace pressures from'the local submerged shell pressures. For example, at Z/1 = 0.552 and 0 = 18@ (See Table 13-4 in the response to Question 13) .

Peak Download = 11.9 - 0.3 = 11.6 psi and Peak Upload = 4.0 - 7.5 = -3.5 psi.

The factor M z at 2/1 = 0.552 is conservatively applied over the range of 2/1 values from 0.5 to 0.64. Therefore, the local pressure values at 2/1 = 0.5 shown in PUAR Figure 2-2.2-8 are the same as the values at 2/1 = 0.552.

V DET-04-028-1A A-90 Revision 0 nute_Ch

Pressure time-histories, such as the one shown in PUAR Figure 2-2.2-8, are calculated at 50 submerged torus shell locations in a 1/16 segment of the torus shell. These time-histories are used in performing a transient dynamic analysis of the torus using the methods discussed in Section 2-2.4.1 of the PUAR. The airspace pr' sure with the NUREG-0661 margin is applied statically tc the entire torus shell and added to the dynamic response to obtain the total response of the suppression chamber due to pool swell.

l 9

l DET-04-028-1A A-91 O

Revision 0 nutggh L