ML20045B398

Rev 1 to Millstone Unit 2 Sf Rack Seismic Analysis Subtask Design Rept.

ML20045B398
Person / Time
Site:	Millstone
Issue date:	06/12/1985
From:	ELECTRIC POWER RESEARCH INSTITUTE, NUS CORP.
To:
Shared Package
ML20045B367	List: ML20045B366 ML20045B396 ML20045B398 ML20045B402 ML20045B407
References
NUDOCS 9306170300
Download: ML20045B398 (21)

Text

.

~ ~ - - . .- . - . .- ;

P f

NUSC0/EPRI ,

h i, :

MILLSTONE UNIT HO. 2 SPENT FUEL RACK SEISMIC ANALYSIS SUBTASK DESIGN REPORT i

i I

JUNE 12, 1985 i

REVISION 01 i

l 9306170300 930610 7 ,

PDR ADOCK 05000336 r$

P PDR l - -_- .- - . _. .

CONTENTS SECTION PAGE 1 OBJECTIVES 1-1 2 APPROACH TO PROBLEM 2-1 3 ANALYSIS 3-1 ,

ANALYSIS METHODS USED 3-1 SEISMIC INPUT 3-2  ;

L LATERAL SEISMIC ANALYSIS 3-3 VERTICAL SEISMIC ANALYSIS 3-4 i VERTICAL IMPACT ANALYSIS 3-5 i

i j

4 RESULTS 4-1 i

5 REFERENCES 5-I .,

l l

f l

I j

(

SECTION 1 OBJECTIVES The objective of the seismic analysis subtask is to perform horizontal and i

vertical seismic analyses of the Millstone Unit No. 2 high density spent fuel storage racks. The purpose of these analyses is to provide seismic i

loads to verify the adequacy of the fuel rack design for consolidated fuel and standard fuel assembly storage, and interface loads to the spent fuel pool. Separate analyses are performed to provide loads associated with the Operating Basis Earthquake (OBE) and the Safe Shutdown Earthquake .

(SSE). .

?

i r

I l

r I

1-1 1

I

SECTION 2 i-APPROACH TO PROBLEM c 0

i I

Seismic analyses were performed to quantify module base shear loads, module sliding response, and module tipping response. Nonlinear time-history analyses were performed for the two horizontal directions and the vertical I b

impact due to tipping. A response spectrum analysis was used'in the vertical direction.

Because it was anticipated that the seismic loads due to the storage of consolidated fuel would otherwise have been large for the Millstone 2 earthquake, the conclusions and recommendations of the interface modeling subtask, Reference (1), for reducing seismic loads were incorporated in the basic fuel rack module design. For example, the recomendations for detuning the rack module frequency from the frequency of the earthquake response spectrum peak included the use of a flexible module support scheme. The use of the flexible support design resulted in coupled rack module natural frequencies that were significantly lower than the fre-quency of the earthquake response spectrum peak, thereby reducing the rack seismic loads.

In developing analytical models of the consolidated fuel canister, the results of the Consolidated Fuel Canister (CFC) Model-Test Correlation Subtask, Reference (2), and the computer code CFCGEN, from the Interface Modeling Subtask, Reference (1), were utilized. The model-test correlation was used to obtain the natural frequency of the canister in water and CFCGEN was used to develop the CESHOCK lumped-mass model of the canister.

2-1

I i

To determine the seismic adequacy of the fuel consolidation rack design for f consolidated fuel and standard fuel assembly storage, the following conditions were analyzed:

1. Earthquake excitation in two horizontal directions and the vertical direction.

2. SSE and OBE.

3. Racks not permitted to slide to determine maximum base shear.

4. Racks allowed to slide to determine maximum displacement.

5. Maximum tipping.

6. Vertical impact due to tipping.

The modules analyzed were selected so that the loads, displacements, and resultant stresses would be maximized for the conditions listed above. The modules analyzed were:

Region I - 8 x 10 module i

Region II - 7 x 8 and 7 x 9 modules Fully loaded, partially loaded, and empty modules were considered in the analyses.

s 2-2 ,

1

1

~

l 4

SECTION 3-q ANALYSIS h' 1

3 n

il

.d ANALYSIS METHODS USED h Horizontal seismic analyses of fuel rack modules with consolidated fuel. ,

or standard fuel assemblies consisted of nonlinear time history analyses using the CESHOCK computer code. CESHOCK is a C-E proprietary' version of.

the SHOCK computer code, Reference (3). CESHOCK's features. include the- .;

capability of modeling nonlinear structural behavior, including. impacting [

of adjacent structures, the capability of modeling friction between I structures with a slip-stick friction element, and the_ ability to model water effects, including hydrodynamic coupling between structures. l The CESHOCK model for the horizontal seismic analysis of the' spent fuel rack module loaded with consolidated fuel is shown in Figure 1. The spent f fuel pool, the spent fuel rack structure, and the consolidated fuel- 'l canister are represented as discrete structural elements'and are coupled

.by nonlinear impact spring representations, friction elements, and- l hydrodynamic mass elements. A nonlinear torsion spring between the rack base and the pool floor models the_ tipping characteristics of the rack module. j Because the vertical seismic input.is not severe enough to cause the' fuel {

to lift off of'the storage modules or the modules to lift off_ of the pool floor, the vertical models behave linearly and response spectrum. analyses  !

are sufficient to quantify vertical seismic loads.

4 3-1 -l l

.l q

-7 ""

"sTN--r-"vf w i i-. - . ,, y,_ ,.,,, ,_ , ,_

The spent fuel storage modules are free-standing and therefore can slide and/or rock on the pool floor. As part of the seismic analysis it is  ;.

necessary to calculate the maximum module sliding displacement and the f' maximum module rotation due to rocking to ensure that adjacent modules do i not contact each other. Separate CESHOCK analyses were perfomed to detemine these quantities. Because a module partially loaded with consolidated fuel is less stable for rocking responses than either a ' fully loaded module or an empty module, analyses for these loading conditions I

were also performed.

Separate vertical impact analyses were performed using nonlinear CESHOCK models to determine the maximum impact forces when a module rocked enough to lift off the pool floor and subsequently impact the floor. The CESHOCK model for the vertical impact analysis of the spent fuel rack module is shown in Figure 2. Results of the horizontal seismic analyses, which included tipping, determined the peak module vertical lift-off heights used as input in the vertical impact analyses.

. SEISMIC INPUT The seismic input provided by Northeast Utilities, Reference (4), consisted of OBE acceleration time histories for the Millstone 2 spent fuel pool. ,

Separate horizontal acceleration time histories were provided for the pool floor and the pool walls in both the north-south and east-west directions. .

SSE time histories were obtained by scaling the OBE data by a factor of 1.89.

Figures 3 and 4 are SSE response spectra plots for the north-south and the i east-west directions, respectively. Figure 5 is an SSE response spectrum plot for the vertical direction.

3-2

. -. -._ . a. n . ,

1. IY[ W'M ," *--+'g*e*,r,e era.

LATERAL SEISMIC ANALYSIS The first step in the lateral seismic analysis was the determination of the dynamic characteristics of empty rack modules with the appropriate support conditions. These characteristics were obtained from frequency analyses of linear three-dimensional models of the rack modules using the computer code SAPIV, Reference (5). SAPIV is a general purpose finite element computer code for static and dynamic analyses. The results of these frequency analyses were then incorporated into nonlinear representations of the fuel rack structures based in part on the methods and models developed in the Interface Modeling Subtask. .

These nonlinear CESHOCK models included a mathematical model of the consolidated fuel canister (CFC), based upon the results of both the Consolidated Fuel Canister (CFC) Model-Test Correlations Subtask and the Interface Modeling Subtask. Because of the close proximity of the fuel storage canister to the rack structure and of the rack structures to the spent fuel pool walls, the hydrodynamic effects are accentuated and were therefore included in the nonlinear models. In addition, for Region I modules a third separate structural element, the spent fuel poison box assembly, and its associated hydrodynamic effects were included in the nonlinear models of the Region I modules.

The CESHOCK models of modules loaded with consolidated fuel included a consolidated fuel canister model developed by the CFCGEN computer code.

The CFCGEN computer code was developed and is described in the Interface Modeling Subtask Design Report, Reference (1). The code develops a lumped-mass stick model of a consolidated fuel canister based on the results and conclusions of the Consolidated Fuel Canister (CFC) Model-Test Correlations Subtask Design Report, Reference (2).

3-3 l

l

The nonlinear CESHOCK models used to calculate sliding included friction i elements between the consolidated fuel canisters and the rack modules, and between the rack modules and the pool floor. Nonlinear torsion springs between the base of the rack modules and the pool floor, representing the rigid body tipping or rocking characteristics of the modules, were also included. Nonlinear impact springs were included in the models to represent the impact between the fuel and the poison box, and the poison box and the rack in Region I models. The results of Reference (2) were used to determine the appropriate stiffness values for fuel impacting elements.

r VERTICAL SEISMIC ANALYSIS The vertical seismic analysis was performed using a response spectrum analysis. As discussed previously, the horizontal analyses of the spent fuel racks required the use of time history analyses due to the nonlinearities introduced into those models by the impacting between the consolidated fuel canisters and the rack, the sliding of the fuel in the racks, and the nonlinear rocking characteristics of the rack modules. 3 These nonlinearities are not present in the vertical analysis and the use of a response spectrum analysis for the linear vertical analysis is justified. ,

t OBE and SSE analyses were performed for both fully loaded Region I and Region II modules. The vertical response spectrum used for the SSE analyses is shown in Figure 5.

3-4 e

4 P

VERTICAL IMPACT ANALYSIS Because the east-west SSE horizontal analyses indicated that the modules l would rock and lift off of the pool floor, separate vertical impact analyses were performed to determine the maximum impact forces. i Nonlinear time history analyses using the CESHOCK code were used. The ,

CESHOCK model used in these analyses is shown in Figure 2. This model 1 includes the effects of module and fuel vertical stiffness as well as the ;,

stiffness of the impacted pool floor. The vertical lift-off heights, as L

determined in the horizontal analyses, were used as input in the vertical impact analyses. f i

i 3-5 ,

_ _ _ ~ _ . _ .

m i

1 l

SECTION 4 RESULTS i

r A series of nonlinear OBE and SSE time history analyses were performed in both horizontal directions. The analyses included 8x10 modules in Region' l I, and 7x8 and 7x9 modules in Region II. In addition, sliding of an empty I module and tipping of full, partially loaded and empty modules were examined in separate analyses. A linear response spectrum analysis provided the vertical seismic loads. The maximum base shear and vertical loads calculated in these analyses are summarized in Table 1. f Fully loaded modules were found to tip and lift off of the pool floor in +

the east-west direction SSE analyses. Since fuel rack modules were found to lift off in the east-west direction analyses, a separate vertical impact analysis was performed. The maximum vertical loads, per support pad and per module, including dead weight and impact force, are provided in Table 2. Fuel rack modules did not lift off in the north-south SSE analysis or in the OBE analysis. .

The maximum relative displacement of adjacent modules, based on displace- l ments caused by rocking (tipping) and elastic displacements, was 1.776 inches. This maximum relative displacement is less than the nominal spacing between modules.

Sliding analyses revealed that fully loaded modules did not slide and that the maximum sliding displacement was 0.562 inches, for an empty Region II 7x8 module.

4-1

1i h,

1 Il

'l TABLE 1 I

Maximum Base Shear and Vertical Loads (1bsf / cell) l Region I ii Dead Weight (incl. bouyancy) for consolidated storage = 2846 ji Vertical OBE for consolidated storage = 486 ,

- Vertical SSE = 875

" " = 880 Horizontal SSE " l t

Region II t

Dead Weight (incl. bouyancy) for consolidated storage = 2621 i Vertical OBE for consolidated storage = 446  ;

" = 802 Vertical SSE "

" " = 603 Horizontal OBE "

" =  !

Horizontal N-S SSE for " 977

" " " = 971 Horizontal E-W SSE 4-2 ,

i 1

~ - _ _..v, TABLE 2 li b

Maximum Vertical Loads on the Pool Floor -[I Due to Rack Tipping in the E-W Direction (includes Dead Weight and Vertical Impact)

- !f Maximum load per pad (to assess localized pool floor loading) = 116,250 lb f Maximum load per module (to assess overall pool floor loading) = 305.100 lb f NOTE: A summary of the maximum vertical loads on the pool floor, itemized by seismic component, appears in the table titled Seismic and Impact Floor Loads in Appendix III of C-E Report No.18767-RCE-401, Rev. 01, " Fuel Rack Structural Analysis Summary for the Northeast Utilities Millstone Nuclear Power Station Unit No. 2 Spent Fuel Storage Racks," D. Tirsun, 5/15/85. The loads in Table 2 above do not include vertical loads due to the vertical seismic component or the' horizontal seismic component in the north-south direction.

4-3

-e -

SECTION 5 REFERENCES

1. NUSC0/EPRI Fuel Consolidation Program Subtask Technical Report, "NUSCO/EPRI Interface Modeling Design Report"

2. NUSC0/EPRI Fuel Consolidation Program Subtask Technical Report, "NUSC0/EPRI Consolidated Fuel Canister (CFC) Model-Test Correlations Design Report"

3. " SHOCK - A Computer Code for Solving Lumped-Mass Dynamic Systems",

V. K. Gabrielson, Sandia Corp., Livermore Laboratory, Report No. SCL-DR-65-34, January 1966

4. Time Histories on Computer Cards, Bechtel Job 7604, File 7604-C-85, transmitted by Bechtel Power Corporation letter MB-5370, " Seismic Input Accelerations for Millstone Spent Fuel Pool", H. S. Kassel to R. D. Hart, January 5, 1976.

5. K. J. Bathe, R. L. Wilson and F. E. Peterson, "SAPIV - A Structural l Analysis Program for Static and Dynamic Response of Linear Systems",

Earthquake Engineering Research Center Report No. EERC 73-11, University of California, Berkeley, June 1973 5-1

h Hoatiourfrt. '6E5Noud maoEL - REmu n monote wirw cret.  :

P pc 963 ,'

I ~ jd#b,s .#)

g  ;

nuwss -

&EnM3 $ l I) *

~

% V WeD A l

h*P<rM@7 "'f,',*"  !

p -y,y4 s ume s

/

L

-ws-

• * ~~A'WSS

== x.<= N 'n t/tt4 Q to ._

f M-UMID>C gtc.cA) etEMENI wgic,y is pluuto M" 71//(MC ARMW y M CwNCCit0AJ  ;

-l; Figure 1 l I

l l

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

1 b

'I e

m I

J

.i i

fh Mer.h u) L .c.T 9 CESWocK Moded .RestaI moduk A W-~ il 4

f 40W50t.14Mra NEL l u @ n @

e o at s s , ,= ,

l 1

RALIC IrmM cT N j '

ucW , , , , ,

is .

9 l

l i

W I$ -

-I l

1 I

P!LLSTONE SSE RESPONSE SPECTRUM. MSS POINT 1.M-S 4 7J OAM#Mb i'

10 , , i . . i  !.

i i i-0 1

! t I o / \

- / \  !\

' ~ k t

@ < s'--_ #%um g

5 w

w w

u u

CE 1

i

, . .- i , , ,,.

O I I I 1 i i i ii  !

l  !

I l I I 01

. e

_ C.

. C.

FRECUENCY (CFS1 Figure 3

'A. -

T P!LLStCNE SSE REEFONSE SPECT*UM.? ASS PC[tJT 1.E-W 4 7e MA/*

10 > ,

4 i s

i ,- .

i:

\ 1, .

i, C

( l JI  ;

\

5

- ,~

) Q .

5 c

/ -

m--.

= f

.~

v

]t U

C 1

i . .

t i

i I

01

- . a

• - a

- c_ _

FDECUENCY tCPS1 Figure 4

4 Meeenee m m m ,

. . . **w % ^

A e-- ,

to.

d 1.

O H

H C f '

sYs

{

J -

f x w \/ ~ h v

k

.1 i

.01

-

• O e C M $ -

FREQUENCY I CPS )

Figure 5

I I

4 i

i t

Spent Fuel Rack Structural Report  !

Rev. 01 .

b

?

5 4

l I

f l

i t

I f

i i

e i

i I

1 1

4 I

')

m 7 , _-.

t M MP2-85-208 l June 18,1985 l

)

Mr. J. F. Opeka Senior Vice President Northeast Utilities Service Company P. O. Box 270 ,

Hartford, CT 06101 ,

SUBJECT:

NUSCD/EPRI Fuel Consolidation Demonstration Program.

Fuel Rack Structural Analysis Report Submittal s>-

i ENCLOSURE: Fuel Rack Structural Analysis Sumary. -

Report No. 18767-RCE-401, Revision 01

Dear Mr. Opeka:

The enclosed report has been revised to incorocrate seismic and impact I loads transmitted to the floor by the spent fuel storage racks. Changes to the report are indicated by a vertical line on the page right margin i

next to the change area. f Please feel free to contact me if you have any questions.

Very truly yours, COMBUSTI ENGINEERING, INC. j

.n ~

S. R. . Jacques  !

', Project Manager RCJ/RLM:amb CRD-85r201 .

cc: G. N. Betancourt (W-24)

T. J. Mawson (W-24)

R. T. Harris ,

G. L. Johnson ,

J. J. Kelley  !

E. J. Mroczka W. H.' Sta s -

Power Systems ,

Comuusbon Ergrws,. Inc.

J