DCL-87-082, Applicant Exhibit A-7,consisting of Forwarding Three-Dimensional Studies of High Density Spent Fuel Racks (Acorn 10 & Acorn 12), in Response to NRC 870401 Request for Addl Info Re Spent Fuel Rack Integrity

Applicant Exhibit A-7,consisting of Forwarding Three-Dimensional Studies of High Density Spent Fuel Racks (Acorn 10 & Acorn 12), in Response to NRC 870401 Request for Addl Info Re Spent Fuel Rack Integrity

ML20237J299
Person / Time
Site:	Diablo Canyon
Issue date:	06/17/1987
From:	Shiffer J PACIFIC GAS & ELECTRIC CO.
To:	NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM)
References
DCL-87-082, DCL-87-82, OLA-A-007, OLA-A-7, NUDOCS 8709030548
Download: ML20237J299 (24)

Text

.

~~ ~

T ~ }p 6[/7 g 75l[3 F7 2.3 - OtA PGandE Exhibit No. 7 M- 7 W O PACIFIC GAS AND E LE C T RI C C O M PANY V '87 liUG 26 P3 :29 FGuG j 77 BEALE STREET . SAN FR ANCISCO, C ALIFORNI A 94106 . (415)781 4211 . TWX 910 372-6567 "Ea ,2,"f"" April 23, 1987 PGandE Letter No.: DCL-87-082 Nucitar Regulatory Commission ATTN: Document Control Desk Hashington D.C. 20555 Re: Docket No. 50-275, OL-DPR-80 Docket No. 50-323, OL-DPR-82 Diablo Canyon Units I and 2 Additional Information on Spent Fuel Rack Analysis Gentlemen:

NRC letter dated April 1, 1987, requested additional information on high -

density spent fuel rack analysis based on the three-dimensional single-rack model presented in the Reracking Report (PGandE Letter No. DCL-85-306, dated September 19, 1985).

Specifically, the NRC requested a reanalysis of two computer analyses (ACORN 10 and ACORN 12) performed for rack qualification as described in the Reracking Report. The results of PGandE's reanalysis are summarized 11 the enclosure. These results confirm that rack impact loads and stresses are significantly below allowable values, thereby ensuring the integrity of the racks and fuel for the postulated Hosgri event. 1 Kindly acknowledge receipt of this material on the enclosed copy of this  !

letter and return it in the enclosed addressed envelope. '

Sincerely, fC J. D. Shiffer i Enclosure cc: L. J. Chandler J. B. Martin M. M. Mendonca  ;

P. P. Narbut B. Norton '

C. Trammell i r CPUC i Diablo Distribution Reracking Service List 1388S/0049K/MLM/1998 8709030548 870617 5 DR ADDCK 0500

HCLEAR REGULATORY CONutist0N

$~ -

Officti Lah. No.

(kchet No. 50 ~

b C D Wy1 1 in the matter of _ .

lDENilflED __

$taff _ - - -

- - RECElVED _

Applicant _ _ -

l REJECTED __..

Intervenot _

Cont's Off's

--Daft b Contractor --

- .Wttnen _ _

Other _

M M A#D Dr TV N Reporter __

\

Q v

n N)~

l

6+ .e d H e-PGandE Letter No.: DCL-87-082 l

ENCLOSURE THREE-DIMENSIONAL STUDIES

~

, (ACORN 10 and ACORN 12) l O

• I 1

l J

l O  :

i 1

l 1388S/0049K i

,,j,.

DIABLO CANYON POWER PLANT' UNITS 1 AND 2 THREE-DIMENSIONAL STUDIES OF HIGH DENSITY SPENT FUEL RACKS

, (ACORN 10 AND ACORN 12)

O 1

Pacific Gas and Electric Company April 23, 1987 1

0  ;

i 13BBS/0049K l

_ _ _ _ _ _ _ _ _ _ _ O

---,. i y ,./

TABLE OF CONTENTS .,'

O

. , 'i

1. BACKGROUND / ./

, . /

) /!

2. DESCRIPTION OF NRC-REQUESTED 3-D STUDIET/ ,, .r

3. DESIGN ALLOWABLES j i,

).1 Gross Section Stresses ,

j n

3.2 Local Stresses l

3.3 Fuel Assemblies

\

4. ANALYSIS OF RESULTS o' j y,

4.1 NRC-requested 3-D Studies 4.2 Comparison.with 2-D Parametric Studies ,

4.3 Summary O 5. CONCLUSION

6. REFERENCES ,

< /

l

\

\

l t e O  :

13BBS/0049K i ,

i

1

1. BACKGROUND The analytical methodology used for obtaining rack interaction impact loads for the high density racks at Diablo Canyon was described in the Reracking Report, submitted by PGandE on September 19, 1985 (Reference 1). The methodology, in general, utilizes conservative inputs 7n a three-dimensional (3-D) single-rack model that lead to conservative

/

i'mpact loads.: PGandE determined that these conservatively determined loads, combined with use of ample design margins, provide an adequate i design ~ basis to acconwdate the effects of multi-rack impacts. This i

determination was furthar confirmed by the results of recent tWo-dimensional (2-D) panmetric studies using both multi-rack and single-rack models (Reference 2) which showed that the conservative models, in general, predicted conservative results. Furthermore, the i calcuhted loads in all cases were found to be substantially less than the loads used for rack qualification.

, I l Following a technical meeting on March 26, 1987, to review the 2-D _

parametric studies, the XRC Staff requested (Reference 3) that PGandE l perform further.3 D single-rack studies which would incorporate the realistic asmmptions used for some of the 2-D parametric studies.

Specifically, the NRC Staff requested that computer runs ACORN 10 and i'

tCOM 12 (Table 6.8.2 of the Reracking Report) be terun with these assumptions. In addition, the NRC Staff requested that the results of these studies'be sumarized in a table similar to Table 6.8.2 of the i

Reracking Report, including the stress ratios (Rj through R6}*

1388S/0049K o____

2. DESCRIPTION OF THE NRC-REOUESTED 3-D STUDIES Computer runs ACORN 10 and ACORN 12 analyzed the behavior of'a fully loaded 10x11 rack located at a' corner of the spent fuel pool for friction coefficients of 0.8 and 0.2, respectively. Based on a review of the

,results of numerous cases reported in the Reracking Report and the

$eismic Analysis Report (Reference 4), it was determined that the case analyzed by ACORN 10 resulted in the highest stresses in the rack, and hence ACORN 10 was selected for this study by the NRC. Computer run ACORN 12 was selected to facilitate comparisons with ACORN 10 in order to study the behavior of the same rack assuming a lower friction coefficient.

The model used in these NRC-requested 3-D studies was the same as the

model used in the design basis analysis reported in the Reracking Report (Figures 6.2.1, 6.2.2, and 6.2.3) except for the following assumptions

• Spring constants: The design basis analysis used spring constants significantly higher (over 10 times) than the calculated values. -

l The parametric studies used spring constants consistent with those used in the recent 2-D IIarametric studies (e.g.,1.5 times the calculated values for the fuel assembly to cell wall impact spring).

• Hydrodynamic coupling: In the design basis analysis external hydrodynamic coupling effects were assumed to be a function of only the horizontal transnational motion of the rack centroid. The forces developed due to these hydrodynamic effects were located at 1388S/0049K __ ____ ____ _ _ __ -

i s ,

' \ * %.

' 'i ,

• s Q

V the rack mid-htight and diretted through tre rack centroid. This s

assumption conservatively neglected the hydrodynamic coupling s.

( effects induced by the roci'.ing and torsional motion of the racks.

Consiste'it with the 2-D parametric studies, the hydrodynamic j coupling terms due to rocking and torsion were included in these NRC-requestsd 3-D studies to more realist'ically represent rack motion'. .

• Gap valces for hydrodynamic coupling: The design basis runs were l

(, performed by assumi'1g conservative (higher) values for the effective gap used in calculating the hydrodynamic coupling effect. The -

effective gap, as derived based on continuity and energy principles,

[ !c given by the fc11 cuing expression:

1 1 1  !

-- .- + .  ;

o 9, 9L Sg t  :

l 7- where g, is the effetivr; gap, and gg and gg are the fluid o

channel widths on ths.left and right side of the rack, perpendicular

s I

ll -

to the direction of rack motion. A value of g, = 2.0 inches was

~

assumed in the design bksis analysis for the x and y directions. In the NRC-requested 3-D studies, the effective gaps yere calculated to i

b. consistint with real): tic situations. The corresponding n'ffective caps were g,,.f 1,0 inch sad g,y = 0.864 inch. Use of N .s 13 BC/0(k9K s\ '

s.  : s

the realistic gap values is appropriate and consistent with the 2-D parametric studies. The hydrodynamic coupling coefficients for transnational motions were developed using an equivalent lateral gap (h0) of 7.5 inches.

,a Support foot construction: The design basis ;nalyses were originally. performed using a smaller support foot geometry than the final foot configuration selected for use in the racks. The computer results were subsequently modified to account for the substantial increase in foot cross-sectional area and inertial properties. During the NRC structural audit of March 24-25, 1986, I l

the NRC Staff noted that the final configuration resulted in lower.

stress ratios, thereby improving design margins (Section 4.4g of Reference 5). These modified stress ratios are included in Tables 4.1 and 4.2. The results of the NRC-requested 3-D studies reported herein reflect the final configuration of the rack support feet.

1 e'

l

'l O

138BS/0049K o_

1

3. DESIGN ALLOHABLES l

3.1 GROSS SECTION STRESSES The acceptance criteria for the Hosgri load combinations are the Level D i Service Limits as specified in Section III of the ASME Boiler and EressureVesselCode. Section F-1370 (Section IIIe Appendix F) states that the limits for Level D Service are the minimum of 1.2 (Sy/Ft ) 0#

1 (0.7 Su/Ft) times the corresponding Level A Service Limits.

Substituting the appropriate value for F t the corresponding multiplying factor is 2.0 for the rack material and the upper portion of the rack feet. For the lower part of the support feet, the multiplying.

factor is 1.67. Table 3-1 summarizes the acceptance criteria for O >*re e.

I l

Using the criteria described above, PGandE examined nine critical sections of the rack, including:

~

• The rack body immediately above the baseplate where maximaa shear and moment will occur. *

• The upper portion of each of the four rack support feet where compression, shear, and moment will be the highest.

1388!i/0049K TABLE 3-1 DESIGN ALLOHABLES Suncort Feet Rack Unoer Comoonent Lower Comoonent

' Material ASTM A240-304L ASTM 479-S21800 SA 564-630 S"

y 23.150 psi 44,900 psi 94,350 psi Su 68.100 psi 101,040 psi 145,000 psi l ASME Code, Table I-2.2 Table I-2.2 Table I-2.2 Section III. Table I-3.2 Table I-3.1 Table I-3.1 Reference O

i t

1 O

i 13885/0049K  ;

~

, e The lower portion'of each of the four rack support feet where shear and. compressive stresses will be the highest due to the reduced cross-sectional area.

Consistent with the original design methodology, the stresses in the above locations were computed at each time step of the analysis for each critical section,. and their maximum values are reported in the form of stress ratios. These ratios are defined (Reference 1) as follows:

R; - Ratio of direct. tensile or compressive stress on a net section to its allowable value (note: support feet only resist compressive loads) -

R 2- Ratio of gross shear on a net section to its allowable value R

3- Ratio of maximum bending stress about the x-axis to its allowable  !

value for the section s

R 4- Ratio of maximum bending stress about the y-axis to its allowable i value R

5- Ratio of combined flexure and compression R

6- Ratio of combined flexure and tensics (or compression) j i

i 13BBS/0049K l

~

l 3.2 LOCAL STRESSES O

Impact loads were examined for local effects on rack components. Areas examined included contact areas between the fuel assembly and cell wall,

'and contact areas (girdle bar and baseplate) which are subjected to rack-to-rack and rack-to-wall impact loads. Standard Review Plan 3.8.4, ippendix D Table 1 states that "... deformation limits should preclude damage to the fuel assemblies." Local effects on Diablo Canyon racks were conservatively evaluated to ensure that no permanent deformation will occur.

3.2.1 Fuel Assemb1v to Cell Hall Imnact .

Using the principle of virtual work, it was determined that the cell wall limit load for fuel assembly impact is approximately 16 kips per cell or 1,766 kips for a 10x11 rack. Allowing for a factor of safety of 2, the corresponding values are approximately 8 kips and 883 kips, respectively.

~

3.2.2 Rack Imnact The local effect on girdle bars due to rack-to-rack impact was evaluated for a corner impact and for a line' load impact. It was determined that for a corner impact, the impact load on the girdle bar would need to exceed 175 kips before any local deformation would occur. The corresponding load necessary to cause deformation for a line load impact is higher.

l l

13885/0049K l

.a Impact loads on rack baseplates are resisted by the in-plane load bearing capacity of the baseplates. The impact capacity of the baseplate is substantially higher than 175 kips.

3.3 FUEL ASSEMBLIES 2achfuelassemblycontains264fuelrodsina17x17 array. The rods are supported by eight fuel assembly grids spaced along the length of the fuel assembly. During rack movement, some of the fuel assembly grids will impact the storage cell walls. The fuel r.ssembly manufacturer, Westinghouse, has determined that fuel assembly integrity will be maintained for an impact force in excess of 3,400 pounds per fuel .

assembly grid.

O

,- i l

O 1388S/0049K _ _ _ _ _ _ - - _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .

l

4. RESULTS OF ANALYSIS

~

l 4.1 NRC-REQUESTED 3-D STUDIES l

' Tables 4-1 and 4-2 show a comparison of the results for the design basis analysis (ACORN 10 and ACORN 12) and the NRC-requested 3-D studies performed using realistic assumptions. In order to establish a common basis for comparison, the stress ratios reported in the.Reracking Report (Table 6.8.2) for the design basis analyses were revised to account for the larger support feet. The results of the comparison indicate the following:

l

a. All stresses obtained from the NRC-requested 3-D studies are enveloped by the maximum stresses as reported for the design basis analyses (ACORN 10). The stresses from the NRC-requested 3-D studies show that the rack and supporting feet have factors of safety in excess of 2.5 over the allowables.

=

i

b. The fuel-to-rack impact load is reduced significantly. The revised load corresponds to approximately 600 pounds per fuel assembly, thus  !

providing significant factors of safety and ensuring cell wall and l fuel assembly integrity (see Sections 3.2.1 and 3.3). l i

c. As indicated in Table 4-1, the maximum stress ratio (1.436 froth l ACORN 10) occurs in the upper component of a support foot. The O

1388S/0049K l _:___-_--____

R 4. N c 1( do al o

lt moo

)

1 3 5 8 7 uf e 2 6 3 5 m t 5 3 2 2 6 it o R xr N

(

1 ao mp p

t r

ru as o l o l r c

u 4 4

3 6

0 9

8 5

3 i eew S R 1 5 0 2 so

/ l s

e e a

re ah t

B' s

- e; 1 7 6 2 us s 3 3 3 2 6 ll o A 2 7 1 3 ae l vv t e a fl

_ R o d

s

_ t a)

_ s eo .

_ e 6 7 0 9 sl n

)

r 2 5 4 8 9 i

_ t R 1 3 0 1 rl g 8 S ear 0

wca oim lt

= it p ern hca

( 2 5 7 5 t c t 2 0 1 1 ei 1 R 1 3 1 2 ;rf 1 ) oi m l emng 1

t b ei 0 2 ehs 1 rt wt ooe 6 2 o h E

P Y

pot pF o u N 9

2 0

2 l et l vi aaw

_ O1

- T S ( h 4 7 E

K C 1 t

c 0. t 6 2e L A / 2 a ( e1

_ s R kl p f A cl e a 8 e f

_ T - ) aat l 4 sto

_ sRWo o ar S o N borpo T l ( N L k k pt

_ U ( 1 cu:

_ S / 2 osa

_ E skk r f

_ R dcce 6 5 e aaat 7 8 ehd F oRRo hta

_ O L N t e

_ ( fl

_ Y t ro

_ ~

R c s o E

% al e ftHr g 9

P pel mui e sag i n

_ U IFb 0 9 rpn r

_ S / a 5 5 e i p l e 2 t rt s

_ l s cei os apm t CA ' f pl c uI a p

d aee a i t

/ t ohh tTt kkc cca ( n aoa s

e s

e e r

t e

a RRs Y Y a .e r I

we ) m o f

olh

) rbc e ei u E - hwh cow a l

e I

( al v r el s el n d gys ar m kl g aui k;nhn a o u uf s s cdi l 0.t m q ei aardi l

c l h ;d s ropnp . s2a a e(f a tk rc;l y

r l sau o v ltd m

a aada el csc aea e s

a eran ea n

rait tnc vee h t

fl C i rl of sai f re s c it m otr t O

gnli ll sa re n sres ;l ann tow ooua Eueoy epo Hcfb Hfrcd s pl u

rsh e

peu r c

ph m ut N

o 0 e d 1 erv a n N hoa o Tfh L u R 8 R O 2 C 0 .

A 3 1 2

R 1 2 5 c

2. (N do al o

lt

)

1 mo o e uf 7 3 4 7 m t 5 4 7 4 0 it o R 1 2 2 5 nr N ao

( mp p

t r ru es o l d l r n 5 3 0 8 ee u 4 6 8 8 6 t w S R 0 0 0 1 so

/ l e e i re la I

ah t

s

- e; s 3 1

8 8

9 2 9 us 0 6 li o R 0 0 1 1 ae l

vv t

a e fl R o s d s ta) e 3 5 4 3 eo .

r 2 2 sl n

_ ) 4 6 9 i t R 0 0 0 0 rl g 2 S ear 0 wca

_ lt oim

_ = it

_ p ern

( 1 3 3 7 hca

_ y 5 3 7 7 t c 1 R 0 1 1 2 ;rf ei 1

)oi m l emng 1

_ 0 t 2 b ei

_ 1 rt ehs ooe 6 wt E pot 2 1

6 o h P pFo 1 2 l et Y l vi

_ u N aaw

_ O2

_ - T S (

4 h K 0. t 76 t

_ E C 1 c L A / 2 a 2e.

( e1

_ B A

R kl p f

_ cl e a 7 e f

_ T - ) aat t 3 sRWo sto ar S o N o T l ( N bor L k pe

_ U ( 1 k pt S / 2 cuc E skk osa R d cce 7 5 r f aaat 2 4 e

_ F oRRo ehd O M hta L_ ( t e fl Y

R m

dal e s ro o E g

p oel *< ftH P sui r n U sag i S IFd /

l e 3

1 2

1 6

rpn e i r

p l s trt s es cei CA '

apm t f pi c ul e d p aee s

/ t ohh i lTt kkr cc aa s

e s

e e

(

t n

a RR- I Y Y a .e r

e r we ) m o f

olh

) rbc e E ai u M - hwh l

_ ( e cow al a v

_ r el s

_ e1 n d ar m k1 g  ; gyg n e u aui k;nhn i 0. ct i um

_ uf ss q cdi l ei aardi s2a

_ h ;d s ropnp e(f a

_ y tk rc;l r l sau o u e aada el csc ltd

_ s eranoa neit aea e

_ a rut nc vee fl h

t

_ C irl ofsai f re gnli s c it m otr t ll sa re n sres ;l ann

_O tow epo e ooua Eueoy

_ Hcfb Hfrcd s pl u

rsh e

peu r c

ph m ut

_ o 2 e d N 1 erv n N hoa e u Tfh L R 2

_ R O 2 C 0 A 3 1 2

, ll'

NRC-requested 3-D studies show that the corresponding stress ratio is significantly reduced (0.743). This decrease is due to the introduction of realistic spring constants. The revised stress ratio provides a factor of safety greater than 2.5 over the allowable (2.0).

For a friction coefficient of 0.2, the maximum stress ratio (0.551) is higher than the corresponding stress ratio (0.298) for the design ,

I basis analysis. However, it is less than the stress ratio (0.743) reported for the friction coefficient of 0.8 and provides a factor of safety of 3.6 over the allowable.

d. The rack-to-rack impact force on the girdle bar (85 kips) is higher than the force (76 kips) reported for the design basis analysis (ACORN 10). However, the resulting stress ratios are significantly lower than those reported for the design basis analysis. In addition, the 85 kip load is enveloped by other design basis

_ analyses documented in the Seismic Report, and the revised girdle -

bar impact force results in a factor of safety greater than 2.0 l against local deformat16n.

1

e. Although the NRC-requested 3-D studies indicated rack-to-wall impacts, the impact force (48 kips) is less than the corresponding force (85 kips) resulting from rack-to-rack impact. Accordingly, rack qualification is not affected. Additionally, the load is substantially lower than the maximum impact load the wall is capable of resisting as reported in PGandE's submittal dated April 7,1987.

1388S/0049K

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

4.2 COMPARISON HITH 2-D PARAMETRIC STUDIFE There are some similarities between the 2-D single-rack parametric studies reported earlier (Reference 3) and the NRC-requested 3-D studies

" discussed herein. These similarities are:

e The fuel-to-rack impact load in the 2-D parametric studies decreased l significantly when the 2-D model was revised to incorporate realistic assumptions. Consistent with this trend, the NRC-requested 3-D studies show a similar reduction in impact loads.

• For the governing cases (p = 0.8), both the 2-D parametric and -

NRC-requested 3-D studies show a reduction of floor impact loads  ;

when realistic assumptions are used. In the 2-D parametric studies, the floor impact load (not previously reported) for the support foot of a 10x10 rack decreased from 339 kips (Case 1 of Reference 2) to 166 kips (Case 2), and in the NRC-requested 3-D studies the load (for a 10x11 rack) decreased from 296 kips to 202 kips (Table 4-1):

l e Forthefrictioncoefflibent0.2,the2-Dparametricstudyshowedan increase of the support foot load (not previously reported) from i 103 kips .(Case 1) to 155 kips (Case 2); however, the increased load i was enveloped by the load obtained from the conservative 2-D model with a coefficient of friction of 0.8. A siellar trend was observed in the NRC-requested 3-D analysis in that the maximum support foot i

O 1388S/0049K

/ load increased from 126 kips (ACORN 12) to 261 kips. This load is si!illenvelopedbytheloadpredictedbytheconservativedesign basis analysis (ACORN 10). '

i

= The maximum rack-to-rack and rack-to-wall impact forces obtained for

2-D and 3-D single-rack models using realistic assumptions are

~

comparable after accounting for differences in rack size (10 x 10 vs. 10 x 11). These loads are further enveloped by the maximum desige basis analysis loads used for qualification of the racks.

t 4.3

SUMMARY

A key decision in predicting the seismic response of the high density O < ck i voiv > i eti c r *iv ai

• i a#* < * *> aich will produce conservative design loads in lieu of developing a very complex model with more realistic input data sets. The modeling of the groups of racks to simulate nonlinear behavior is an extremely complex and time-consuming task. As a practical solution, a 3-D single rack -

model was selected for the design basis analysis. Such a rodel, however, required that the motion of the adjacent racks be predefined.

Accordingly, conservative (out-of-phase) motions of the adjacent rack were assumed, and this assumption was utilized in conjunction with several conservatism in other aspects of the model. In particular, increasing the stiffness of the impact springs and underestimating the value of the fluid coupling effects contributed to obtaining conservative results.

1388S/0049K l

In addition, a' number of cases were evaluated to study the effects of the differing rack sizes, variations of friction coefficients, variations in fuel loading, and location of the racks in the pool. The maximum values of loads obtained from these sets of analyses were used for design and l ' qualification of racks.

l  :

Yhe recent 2-D parametric studies confirm the validity of the above methodology in the context of planar motions. PGandE initially conducted planar motion studies of racks where a single rack was simulated by two degrees-of-freedom (DOF) containing one rattling mass. It was found that a single-rack, 2 DOF model produced conservative results when compared with those obtained by a multi-rack model with a more realistic set of parameters (Reference 6).

O The same conclusion was reached when the dynamic analysis was extended to a 4 DOF model (per rack) incorporating variations in friction coefficients, fabrication and installation tolerances, and layout and loading of the racks (Reference 2). -

The NRC-requested 3-D study reported herein shows that the design basis l loads used for qualification of the racks envelop the loads from the revised analyses.

Furthermore, it has also been demonstrated that PGandE's design of the rack hardware is conservative in that large design ,

margins exist for both gross section and local stresses. l l

O 1388S/0049K Finally, it should be noted that several aspects of conservatism in the design basis model (such as ignoring nonlinear coupling) were retained in all of the studies performed. Considering these aspects would show that design margins are even higher; s

8k.

I e

O O

l 1388S/0049K _ j7 ,

I

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

l

5. CONCLUSIONS The NRC-requested 3-D studies further demonstrate that the design basis analysis predicted conservative rack qualification loads. In trsponding

'to NRC Staff requests for additional information on rack behavior. PGandE

.has performed numerous parametric studies, including 2-D multi-rack

' analyses, using a spectrum of modeling assumptions. The results of these.

studies confirm in all cases that rack impact loads and stresses due to the Hosgri earthquake are significantly below allowable values.

Therefore, PGandE concludes, with a high degree of confidence, that the design basis evaluation was conservative and the high density spent fuel racks will maintain their integrity for the postulated Hosgri event. -

O d' I I

i i

O l

1388S/0049K -- -- _ _ __ _ _ o

^

f 6. REFERENCES

1. "Reracking of Spent Fuel Pools Diablo Canyon Units 1 and 2,"

Enclosure to PGandE Letter No. DCL-85-306, dated September 19, 1985.

2. " Additional Information on Rack-to-Rack Interactions," Enclosure 1 to PGandE Letter No. DCL-87-070, dated April 7, 1987.

3. NRC Staff letter to PGandE, dated April 1,1987 Re: Spent Fuel Rack Analysis - Request for Additional Information. ,

4. Seismic Analysis Report, " Seismic Analysis of High Density Fuel -

Racks for Pacific Gas and Electric for Diablo Canyon Nuclear Power O s**i a 3. sat 8r3.iens.4.Seier.Ta ,779.

5. Technical Evaluation Report, " Evaluation of Spent Fuel Racks Structural Analysis; Pacific Gas and Electric Company, Diablo Canyon Unitj 1 and 2." April 30,1986 R. C. Herrick, Franklin Research -

Center TER-C5506-625.

/

6. " Rack Interaction Studies " Enclosure to PGandE Letter No. DCL-87-022, dated February 6, 1987. j I

O  ;

1388S/0049K