ML20237K399
| ML20237K399 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 06/17/1987 |
| From: | Carfagno A, Okaily A, Tishman H CALSPAN CORP. |
| To: | NRC |
| References | |
| CON-NRC-03-81-130, CON-NRC-3-81-130 OLA-S-001A, OLA-S-1A, TER-C5506-625, NUDOCS 8709040331 | |
| Download: ML20237K399 (99) | |
Text
88 - 2 75 3 2.3 - 66A NRC Staff Exhibit 1-A 6//7/r7
,: :.m w
3_
'87 AUG 26 P4 :09 GF?
j DOCK i
i b;-
FRANKLIN RESEARCH CENTER l
DIVISION OF ARVIN/CALSPAN EVALUATION OF SPENT FUEL RACKS STRUCTURAL ANALYSIS l
PACIFIC GAS & ELECTRIC COMPANY l
DIABLO CANYON UNITS 1 AND 2 TER-C5506-625
)
TECHNICAL REPORT r
20TH & RACE STREETS PHILADELPHIA, PA 19103 TWX 710-670+1889 TEL (215) 448-1000 (v h 9709040331 870617
- Ca n
. gy, 1-p 0
MCliAR #CGltlATORT C00135!05 A
Dxtet No. 504%D@
grrgjg gg, gg,.
b4Y mPnW la the matter of
/
IDENTIFl!D 5tsH _
t REttlVED Apphtafd -
RUECTED 3818m* -
Cont *g Offt _..
b~( ~ ]
DME Contractor _
Witness __
Other -
hYY4 dd6 Reporter t
h8 f
I f
l
- 44 l
l I
l k
l l
L---
..)
n.. g j8 9
TECHNICAL EVALUATION REPORT 3
S NRC DOCKET NO. 50-275, 50-323 FRC PROJECT C5506 3]
NRC TAC NO. --
FRC ASSIGNMENT 26 1
NRC CONTRACT NO. NRC-03 91 130 FRCTASK 625
)
3 EVALUATION OF SPENT FUEL RACKS STRUCTURAL ANALYSIS s.
EACIFIC GAS & ELECTRIC COMPANY n
DIABLO CANYON UNITS 1 AND 2 TER-C5506-625 d
Prepared for q
Nuclear Regulatory Commission FRC Group Leader:
A. A. Okaily j
Washington, D.C. 20555 NRC Lead Engineer:
H. Ashar
- ]'
April 30, 1986 Revised May 28, 1987 3
f This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United Etates Government nor any agency thereof, or any of their employees, makes any warsiatty, expressed or impiled, or assumes any legal liability or responsibility for any third party's use, or the results of such use, of any information, appa-ratus, product er process disclosed in this report, or represents that its use by such third party would not infringe privately owned rights.
Prepared by:
Reviewed by:
Approved by:
b 'l W
U Principal Author ~
Department Di[ectq/
~
ZO Nb(A '07 Date: 5 $S l$G Date:
0~~ E
Date:
D U
FRANKLIN RESEARCH CENTER DIVISION OF ARVIN/CALSPAN 20th & R ACE STRfffl.PMILADf LPHIA. PA 19105 t
1 l=
TER-C5506-625 l
l I
CONTENTS Section Title Page 1
INTRODUCTION 1
1.1 Purpose of the Review.
1 e
)
1.2 Generic Background.
1 I
l 2
ACCEPTANCE CRITERIA.
3 I
I
[
2.1 Applicable Criteria 3
2.2 Principal Acceptance Criteria.
3 3
TECHNICAL REVII'd 6
l 3.1 Modeling of Spent Fuel Rack Modules for Seismic Analysis 6
b 3.2 Evaluation of the Nonlinear Dynamic Displacement Analysis.
29 j
3.3 Rack Module Impact and Stress Analysis.
32 3.4 Spent Fuel Pool Analysis for High Density Fuel Racks 41 3.5 Fuel Handling Accident Analysis 70 l
4 CONCLUSIONS.
'l i e
{
l 5
REFERENCES.
76 l
l APPENDIX A - LICENSEE RESPONSE, PER REFERENCE 13, FOR BACKGROUND
)
AND VERIFICATION OF DISPLACEMENT ANALYSIS METHOD DYNAHIS l
l I
l
\\
iii
)
e e
TER-C5506-625 r,,
Ti s 2-FOREWORD
~
...J This' Technical Evaluation Report.was prepared by Franklin Research Center 7
. under a contraf:t with the U.S. Nuclear Regulatory Commission (Office of y
b Nuclear Reactor Regulation, Division of Operating Reactors) for technical assistance in support of NRC operating reactor licensing actions. The g
.4
.;1 technical evaluation was conducted in accordance with criteria established by r
the NRC.
?
m lh ce b
1 p
V l
l r ';
Ps J
n
- ^a!
L L..
y 15 Pi
~
1
- h. 3
[
1 D
k l]
v i '.;
P l
l'.!
ie-J
s TER-C5506-625 1.
INTRODUCTION 1.1 PURPOSE OE THE REVIEW This technical evaluation report (TER) covers an independent review of' the Pacific Gas and Electric Company's licensing report (1), including its l
main supporting document (2) and supplementary studies (3, 4), on high-density spent fuel racks for Diablo Canyon Units 1 and 2 with respect to the evaluation of the spent fuel racks' structural analyses, the fuel racks' design, and the pool's structural analysis.
The objective of this review was to determine the structural adequacy of the Licensee's high-density spent fuel e
y racks and spent fuel pool.
I I
p 1.2 GENERIC BACKGROUND Many licensees have entered into a program of introducing modified fusi racks to their spent fuel pools that will accept higher density loadings of g
spent fuel in order to provide additional storage capacity.
However, before n
the higher density racks may be used, the licensees are required to submit rigorous analysis or experimental data verifying that the structural design of the fuel rack is adequate and that the spent fuel pool structure can accommodate the increased loads, p
The analysis is complicated by the fact that the fuel racks are fully n
f immersed in the spent fuel pool.
During a seismic event, the water in the l
h pool, as well as the rack structure, will be set in motion resulting in fluid-6
)
structure interaction.
The hydrodynamic coupling between the fuel assemblies I
y and the rack cells, as well as between adjacent racks, plays a significant
- t role in affecting the dynamic behavior of the racks.
In addition, the racks are freestanding.
Since the racks are not anchored to the pool floor or the
[
pool walls, the motion of the racks during a seismic event is governed by the P
static / dynamic friction between the rack's mounting feet and the pool floor, l
I and by the hydrodynamic coupling to cijacent racks and the pool walls.
]
P l
Accordingly, this report covers the review and evaluation of analyses b
submitted for Diablo Canyon Units 1 and 2 by the Licensee (1, 2, 3, 4),
wherein the structural analysis of the spent fuel racks under~ seismic loadings
]
is of primary soncern due to the nonlinearity of gap elements and W
l h,
a TER-C5506-625 static / dynamic friction, as well as fluid-structure interaction.
In addition to the evaluation of the dynamic structural analysis for seismic loadings, the design of the spent' fuel racks and the analysis of the spent fuel pool structure under the increased fuel load are reviewed.
?
4 l
l i
l l
l 5
l l.
--, q w.
TER-C5506-625-m, I
~ '
2.
ACCEPTANCE CRITERIA I
id 2.1 APPL'ICABLE CRITERIA l
The criteria and guidelines used to determine the adequacy of the high-Ld density spent fuel racks and pool structures are provided in the following q
documents:
o OT Position for Review and Acceptance of Spent Fuel Storage and.
Handling Applications, U.S. Nuclear Regulatory Commission, January 18, 1979 [5]
Standard Review Plan, NUREG-0800, U.S. Nuclear Regulatory Commission o
laf Section 3.7, Seismic Design Section 3.8.4, Other Category I Structures 3
Appendix D to Section 3.8.4, Technical Position on Spent Fuel
[j Pool Racks Section 9.1, Fuel Storage and Handling o ASME Boiler and Pressure Vessel Code, American Society of Mechanical
.1 Engineers,Section III, Division 1 l.,]
~
Building Code Requirements for Reinforced Concrete (ACI 318-63),.
o American Concrete Institute o Regulatory Guides, U.S. Nuclear Regulatory Commission I
1.29 - Seismic Design Classification 1.60 - Design Response Spectra for Seismic-Design of Nuclear Power Plants 1.61 - Damping Values for Seismic Design of Nuclear Power Plants 1.92 - Combining Modal Responses and Spatial Components in Seismic Response Analysis 1
o Other Industry Codes and Standards American Institutes of Steel Construction.
')
..l 2.2 PRINCIPAL ACCEPTANCE CRITERIA I
i
.The principal acceptance criteria for the evaluation of the spent fuel 3j racks' structural analysis for Diablo Canyon Units 1 and 2 are set forth by the NRC's OT Position for Review and Acceptance of Spent Fuel Storage and i
?
. )
J
TER-C5506-625 Hand,ing Applications (OT Position Paper) (5).
Section IV of the document describes the mechanical, material, and structural considerations for the fuel racks and their analysis.
The main' safety function of the spent fuel pool and the fuel racks, as stated in that document, is "to maintain the spent fuel assemblies in a safe configuration through all environmental and abnormal Icadings, such as earth-quake, and impact due to spent fuel cask drop, drop of a spent fuel assembly, or drop of any other heavy object during routine spent fuel handling."
Specific applicable codes and standards are defined as follows:
" Construction materials should conform to Section III, Subsection NF of the ASME* Code.
All materials should be selected to be compatible with the fuel pool environment to minimize corrosion and galvanic effects.
Design, fabrication, and installation of spent fuel racks of stainless steel materials may be performed based upon the AISC** specification or Subsection NF requirements of Section III of the ASME B&PV Code for Class 3 component supports.
Once a code is chosen its provisions must be followed in entirety.
When the AISC specification procedures are adopted, the yield stress values for stainless steel base metal may be obtained from the Section III of the A!XE B&PV Code, and the design stresses defined in the AISC specifications as percentages of the yield stress may be used.
Permissible stresses for stainless steel welds used in accordance with the AISC Code may be obtained frcm Table NF-3292.1-1 of ASME Section III Code.
Other materials, design procedures, and fabrication techniques will be reviewed on a case-by-case basis."
Criteria for seismic and impact loads are provided by Section IV-3 of the OT Position Paper, which requires the following:
Seismic excitation along three orthogonal directions should be o
imposed simultaneously.
The peak response from each direction should be combined by the o
square root of the sum of the squares.
If response spectra are available for vertical and horizontal directions only the same horizontal response spectra may be applied along the other horizontal direction.
- American Society of Mechanical Engineers Boiler and Pressure Vessel Codes, Latest Edition.
- American Institute of Steel Construction, Latest Edition. __-__
y.
3 A
TER-C5506-625
- o F
Increased damping of, fuel racks due to submergence in the spent fuel
~
o pool is not acceptatlo without applicable test data and/or detailed analytical results.
- s
\\
o Local impact of a fuel assembly within a spent fuel rack cell should be considered.
i d
Temperature gradients and mechanical load combinations are to be
considered in accordance with Section IV-4 of the OT Position Papnt, The structural acceptance criteria are provided by Section IV's of the OT.
Position paper.
For sliding, tilting, and rack impact during seismic events, Section IV-6 of the OT position Paper provides the following:
a "For impact loading the ductility ratios utilized to absorb kinetic energy in the tensile, flexural, compressive, and shearing modes should be quantified.
When considering the effects of seismic loads, factors of
(
safety against gross sliding and overturning of racks and racks modules' under all probable service conditions shall be in accordance with the s
P Section 3.8.5.II-5 of the Standard Review Plan.
Thispositiononfac) tors l
of safety against sliding and tilting need not be met provided any one of i )
s the following conditions is met:
y 4
(a) it can be shown by detailed nonlinear dynamic analysii that the a
amplitudes of sliding motion are minimal, and impact between i'
adjacent rack modules or between a rack module and the po'oi walls is l
prevented provided that the factors of safety against tilting are within the' values permitted by Section 3.8.5.II.5 of the Standard Review Plan l
(b) i: can be shown that any sliding and tilting motion will be cantained within suitable geometric constraints such as thermal
,(
clearances, and that any impact due to the clearances is incorporated."
i.g I
l 9
3 31 3
i U
l s
t
- I ia l
'=e A 2
1 TER-C5506-G25 3.
TECHNICAL REVIEW
\\
I 3.1 MODELING OF SPENT FUEL RACK MODULES FOR SEISMIC ANALYSIS 3.1.1 Overview of Rack Module Dynamic Anal'ysi's I)
The fully submerged high-density fuel rack modules, as reviewed herein, s
are freestanding on the spent fuel pool floor, i.e., they are neither anchored to the pool floor laner, attached to the pool walls, nor fastened to each other, but are constrained by friction between the module support pads and the 1
J' pool 2iner.
As a consequence, the modules may exhibit highly complex motions
)
under seismic excitation, thus requiring comprehensive, nonlinear, structural
}
dynamics analyses to simulate, or predict, their maximum displacements and f
associated stresses.
1 il Sources of nonlinearity in the analyses include the following:
!l Hydrodynamic coupling through the water between each fuel assembly l
o and the storage cell walls of a module, as well as that between i
adjacent rack modules and between rack modules and the pool walls.
o Friction (stetic and sliding) between the base of the rack module ar"
)
the pool floor liner, which acts together with hydrodynamic coupling i
forces between adjecent modules and the pool walls to both excite and restrain the module during seismic events.
1 4
1 Impact between the fuel assemblies and the storage cells of a rack o
module, wherein each fuel assembly moves through its clearance space to impact, repeatedly, the cell walls.
{
Impact of a rack module with adjacent modules and/or the spent fuel o
pool walls.
q i
Lift-off of one or more support pads of a rack module from the pool o
floor and the subsequent impact of that pad (or pads) with the pool
- l floor.
i To determine the maximum rack module displacements, impact forces, and i
stresses, the Licensee modeled selected rack modules, performed the nonlinear l
)
dynamic displacement analysis, and, using the results of the displacement analysis, computed the maximum stresses in the rack module and compared the l
stresses to allowable values.
The nonlinear displacement analysis was a full i
three-dimensional time-history analysis based on earthquake acceleration time l
1 histories derived from the accepted earthquake spectrums defined in the updated i
~6-
l 1
j
~
TER-C5506-625 Final Safety Analysis Report (6).
The use of three-dimensional analysis
}
resolved any question regarding adequate summation of separate two-dimensional analyses'through the square-root-of-the-sum-of-squares technique.
4 7;
I j
3.1.2 Rack Module Orientation in the Spent Fuel Dool J
l Diablo Canyon Units 1 and 2 plan te utilize 16 high-density fuel racks comprising 13 rack nodule storage configuration designs as defined in Table 1 n
for each spent fuel pool.
The 16 rack modules are to be arranged in the pools
^
of Units 1 snd 2 as shown in Figures 1 and 2 (which were adapted from Reference 1).
I i
Clearance between the rack modules and the pool walls is a minimum of 3.0 inches and generally grea,ter (up to 37 inches) for most modules.
Clearance I.i between exterior walls if ad]acent modules is 2 1/4 inches; at the 7/8-to a
1-1/16-inch-thick girdle bars, which are provided to protect each rack from rack-to-rack and rack-to-wall impacts, the minimum clearance is 1/8 inch.
>t Note that the rack modules shown as Region 1 in the spent fuel pools of l
Units 1 and 2 are provided with neutron absorbing material (Boraflex) to accept spent fuel that has not achieved sufficient burnup for Region 2.
The fuel rack modules of Region 2 do not contain the neutron absorbing material.
3.1.3 Rack Module Construction I
As provided by the Licensee (1), an elevation view of a typical rack 1
module with girdle bar is shown in Figure 3.
Fixed and adjustable mounting
,a d
supports are shown in Figures 4-a and 4-b.
l l
g Horizontal and vertical cross sections showing typicel construction of d
the rack modules are provided in Figures 5 and 6.
Note that Figures 5 and 6 include the neutron absorbing material required for rack modules in Region 1.
e Typical construction details of Region 2 rack modules were stated by the Licensee to be the same except for the deletion of the neutron absorbing m
l l
material.
1 Typical module storage cells or Loth Region 1 and Region 2 racks have an i
8.85-inch-square cross section.
The Licensee stated that this cross sectional dimension is sufficient to ensure that fuel assemblies with the maximum expected axial bow can be inserted and removed from the storage cells without I
damage to the fuel assemblies or the rack module (1). :.)
TER-C5506-625
'i f
Table'1. Rack Module Storage Configuration (Table 2.2 of Ref. 1) l am
. c.
Approximate Module Quantity Cells per Array weight 1
(1b/ module)
Region Type Per Unit Module Size 4
1 A
2 100 10 x 10 21,500' 1
B 1
90 9 x 10 19,500 2
C 1
100 10 x 10 25,500 2
D 3
90 9 x 10 23,000 2
Es*
1 C' 6 11 x 6 17,C30 2
F 1
72 9x8 18,500 Ea
.i 2
G 1
80 10 x 8 20,500
{
15,000
'l I 2
H
'1 24 + 10 2
J 1
96 9 x 10'+ 6 24,500 2
K 1
54 6x9 14,000 2
L 1
81 9x9 21,000 2
M 1
110 11 x 10 28,000 2
N 1
81 9x9 21,000 s
i 3
1 l
s.l c;
.i
.j j
1 k.
e TER-C5506-625
,l' 6*
N 3
I k p.3
_ Id-o" 89 b
.y
,3%
1+
y n:
Ref.
Ret.
2/41 i
q) go 9,,,y ; 88 10 8*
}
s o
y I3
,,pDi
%e g*_
- E 9X9 g
me w.
~y
(
q, h
M
Di i -
M
B^
,e-N p
E X 2+02 l
M{
F!us IDX9
'K Region 1
-{
[f,8f,j_
i
[;,.
9 x s-iOx 9-
=,
9 a.
"~
,c"
'Di
'4 L,
i..
8 X 10 10 X 10 9 X 10 Region 1 g
58-4 m
go x go I
R e f.
M
'x 4
1
" ~
.c 3
.y-
.g..
o-
. g, 6X9 IOX9+6 9 X 10 Region 1 I '*
Ilt II" lO X IO
- S 6 iG-
- e e
h~
==:=-5=bR f
m
'E'
'L'
'M" g:
H d
Cask Pit '
6 X il 9X9 10 X li
'm
/
/
~
g l53 3
26'/16" e
r 4
.l 4
R e f.
rg d
ld-B" 64 [8 9716 108" - *9 I
I/
g 9 sk 2'/k 2 /[
I n
.e Ref.
3 5'- O '
=
C E
1
.I 1.1 9
Figure 1.
Pool Rack Module Layout, Unit 1 (Figure 7..la of Ref. 1)
_9 f'
TER-C5506-625 3 S'- 0" N
916 2 '/.
2k n
4 R eT.
f-i*
u 10-8"
! 64 8
/
_9716 10 8
, g.. j.
\\ Ref.
I 4 I
" gY
- *,' .'.6 "
g
.y.
,.,i N Cask Pit 6 x il 9X9 IO X il e
N P 'S E
a N
2b "F*
Q" "D"
Region 1 5
$,~
8X9 IOX9+6 9 X 10 10 X IO o
.g S
o
'_q g
a e
~,
i N
34:*,
v'
- 'c
'Di Region 1 Ref.
8 X IO IO X IO S X IO IO X IO 9
'Ai o.'s
,n I w
.Ds
-n-a n (Plus 10 GX244X2 9x6 Region 1 10 X 9 4.
3ox 9
~
e-
, Misc Cey ac].
'B' e
j_ y.
u Wi
'N'
'4 m2
,0 9X9 6
_ 88 6 108"
'd t*
l 97 f,6 e
\\
RIf.
2 '/[
I 15" 5
. 3"-
19ga
- 3 Ref.
Re f.
~"
~
_ 6-O" l
Figure 2.
Pool Rack Module Layout, Unit 2 (Figure 2.1.b of Ref. 1) -
TER-C5506-625
~
.a
)-
sy N
t t-0 l
1 4.
v-vy-w-
w r ww w-w r,cervyry
.(
. ;a.
I4 i..
i ei g 8i l ii l.
\\ rdle l
l a
Gi Bar e
s i
e e
i I
a e
9 e
s
.o a
e,-
1,
.1 I
I e
i l
i i
i 1.
N i
1 k
l T
l 7
N Figure 3.
Typical Rack Module. Elevation 3
(Figure 2.2 of Ref. 1) d
. t t'
lO>
TER-C55064 25
,1
..i i
~]
, :1 4
g:
Baseplate I
5'[
ll l
=
=
13.
i ll If l
I 7
V J
{V a
(8) 1* Dia.
Flow Holes n
v
- E.
12" Sch 160
((
Pipe k
h I
k l
1*
o 8
t 14' O.D.
c t1
- )
i i
1 l
Figure 4-a.
Fixed Mounting Support (Figure 3.6a of Ref. 1)
- q V.
.x
TER-C5506-625
>S5 fj
.y o
Y:
{,,
6 '/ %
Il Baseplate 8
j 4
l
.l 5', G le I
.4 I '/8 e
'? 5
((
l l
l Y
l t
g,8 s'
2 s
o I8) 1' Dia, Flow Holes
/ :
a "4 E
~
~o=<
- u u
bz g\\ 5'-4Unc.
o Class 1 A W
w m
~$
W,W 7
\\
O i *mP l
[l ljj l
Stop 6
gi I
,a
-4/2%-+
.]
- 6 '/[%
1"% Hole
]
9*9 3
8 l1 b
R
?
8 Figure 4-b.
Adjustable Mounting Support 11 (Figure 3.6b of Ref. 1) il4 V, i
7 ER-C5506-625 i
i i
4 t.0 5 0'
.050' 10.93*
Inner Box
=
(Internal Square Tube)
Pitch
,i i
1
__.m, r1...
m,.
1
):n j
jlsuij g
{i
)
i d
i s
i g
I s
s
_,q._.
4 g
5e d
)
l-a e
s l
p=i[
,m l
~;
i;
,s. _ _ - - ~...
_v=f m s,
.. --m>
l r-.m i
=
t l
4
,s
(
a l
l 8
J=
limi t
4 i
'h ll'uli lluulk
[
e l
l l
l a
l yt vs l
,J i
ta>.
n.
a
- 1,,,r.,s m ll Connecting Channel.
Cover Sheet (Gap Channel) i Neutron Absorber Sheet (Boraflex) i i
I Figure 5.
3 x 3 Array of Typical Storage Cell Construction - Region.1 (Poisoned Cells)
(Figure 3.1 of Ref. 1) -
TER-C5506-625 t
2h1V Y9' 3
Weld 7 4 Lp.
i
-- i
[j T
- v
t i
!. O_
. O' b A!
es
[3 h
2 t,
].
e s__ l.' V, ][
~'
f
(./
'E
( /
^
- s t
?.
(' r.e.
Baseplate
-- 8.8 50"?
sq.
E Z
lil (Typ.)
Fuel Assembly f'
pl j2 Ref.
n f
_ _D Noth 8
l l
~:-
3 i
_ _ _,i l
l F- - ~]
i l L_
l
_ _1 l
L_ __, _ s,I a
cn I
e Poison l
l l
l y
l l
22 l
cn l
l l
c p
m l
..y i
i
\\
l S
L _
_ _ _4 r---
l 25 l
7 l
i l
1 1/2" Dia.
8 L~
i l
l Flow Hole I
_I.i y
l l
l l.,
s 8 l _.'
Baseplate
~
-.S y$
y.:..-).
l i1 l
i s
w 1
(
li l
I D.
t g
6' Dia Hole Typ.
~]
(5' At Support Leg Locations)
.i
. ~2 *,
Ift lu
.a Figure 6.
Typical Cell Elevation - Region 1 r.-
(Poisoned Cells)
(Figure 3.5a of Ref. 1)
,).
C
TER-C5506-625 t
i Materials for both Region 1 and Region 2 rack modules were stated to be:
- Stainless steel sheet and plate ASTM A-240-304L Forging naterial ASTM A-182 j
t 1 Weld filler material ASME SFA-5.9, Type 308L and 308 LSI I
3.1.4 Nonlinear Dynamic Displacement Model j
i The Licensee's nonlinear model of a spent fuel rack module for dynamic displacement analysis is shown in Figures 7, 8, 9, and 10.
Figure 7 represents the rack module as a single-cell, or stick, model on a rigid base and four sppports.
Figure 8 shows the addition of impact springs acting through gap' elements to simulate compression-only contact (impact) with j
adjacent rack modules and/or the pool walls.
The springs registered the
! I impact force of such contacts in the analysis.
Sixteen contact springs were
- l used, two at each top and bottom corner, to resolve the horizontal forces at the corner.
Figure 9 represents the modeling of a fuel assembly in a rack storage cell. Springs acting through gap elements were used to represent movement of the fuel mass in clearance space before impacting the cell walls.
Although the Licensee showed the presence of fluid dampers, fluid damping was not used. The dumper elements, however, may be used to represent hydrodynamics coupling between the fuel assembly mass and the cell walls as evaluated in a later section of this report.
Figure 10 (Figure 6.3.1 of Reference 1) was used by the Licensee to clarify the model concept by showing a two-dimensional view of the three-dimensional model. Note the relationship of the fuel assembly mass (mass 2*)
to the model, and that the fuel assembly masses (1* and 2*) may be offset from the rack module centroid, as shown in Figure 7, to investigate the effects of spent fuel being placed on one side of a rack module.
Figure 10 also provides greater detail in the modeling of the rack module base and support.
At the base plate, springs K,are the rack-to-rack or rack-to-wall impact springs.
Springs K6 model the vertical stiffness of the lower rack ar.c supports.
They are connected to gap elements to accommodate liftoff.
s l
TER-C5506-625 9
i*
J 2:
J r -
M j
Coupling Elements y.
(
2 f._,
/
d
-q, j]
n A-
'1
/, 2, j
/
i
/
\\
,a Q7 l
.I i
A3l y
H d
l
. ()l a
I A y.._,
L 1
AX 1-f
'q 2 y
=
_ /
Q4 5
s n
i Support h
N On S1 x
S2 +
o
)
g nw
'X h.,
,[
XB, YB - Location of fuel rod q,
group mass centroid - relative to l
centerline of fuel rack a
a
.1j i
2 l
(
't b
i Figure'7.
Fuel Rack Module Dynamic Model (Figure 6.2.1 of Ref. 1) a I j
4
..J TER-C5506-625 i
.1 i
Typ. Top Impact i
Element m
m I
I W
(
r!
1 1
Rack Structure
/
i
.o i
Typ. Bottom impact Element i
f t
- 3 i )
4 i
Figure 8.
Gap Elements'to Simulate Inter-rack Impacts (Figure 6.2.2 of Ref. 1)
'b
c.
b TER-C5506-625 r
3.
t
- q F'
i 1
j.
0 I;
h y
Impact Springs i l i
a '
3 ' en re Mass 2*
$a LJ
\\ Fluid f.'
/
4 Dampers (not used in
-"~
i this analysis) i l'
higid Frame r.
1' W
/
7-x l
E9 l3 1
1 Ll l
.f 1
Figure 9.
Fuel Assembly Dynamic.Model 1
(Figure 6.2.3 of Ref. 1)
J e
. j i
1
. -.... - - ~
,__~w.
- n -.
.,,r. w.._.sa.,=
n =.;-,
1 TER-C5505-625 l
i..
g, 2"
i:
n l
K, K
f w
l 2 d
_l yM n,
2 l
l 1
Gap Elements j
To Simulate Inter-Rack Irnpacts (4 f or 2-D Motion -
Rack Centroid il 16 for 3-D Motion)
(Aisumsd At H/2)
E H
1 l
[ Rigid Rack l
l Baseplate l
l K
K
% }<
Q3 1
K K
$g y
h K
wq
[
c
- K K
7 h
4er R sw f
i Figure 10.
Two-Dimensional Representation of Rac.k Dynamic Model (Figure 6.3.1 of Ref.'1) 20-4
J I
TER-C5506-625
{
.l
\\
Springs K are special elements that, like springs, sense loads up to some j
f I
predefined limit, after which motion is permitted to take place with resist-ance at that limiting value, and thus simulate sliding fri.: tion at a constant
'i coefficient of' friction.
Springs K are the torsional counterpart of linear l
r friction springs K and are used by the Licensee to account for any g
torsional frictional moment that may be developed at a support due to
{
E rotational displacement of the rack module.
It was noted that the Licensee's model includes two mass points comprising 8 degrees-of-freedom (DOF).
Mass 1 is located in the rack module and has 6 degrees-of-freedom, i.e., three-dimensional space with three linear transla-l tions and three rotations.
Mass 2* is located at the top of the fuel assembly and moves with 2 transnational degrees-of-freedom.
The lower mass point of H
the fuel assembly is lumped with the rack module mass.
a The Licensee's 2-mass, 8-degree-of-freedom model was formulated in an I
effort to reduce the degrees-of-freedom to a minimum to facilitate simul-
]
taneous analysis in three-dimensional space, while addressing impacts between e,
the fuel assemblies and the storage cell walls, and between adjacent rack i
e t, modules. With the rack module modeled as one mass point with 6 degrees-of-1 freedom, the Licensee assured that the rack module was a rigid body, and justified this assumption by citing the high natural frequencies of the rack 5
module in contrast with the much lower seismic excitation frequencies.
The mass of the fuel assemblies (the major portion of mass of a loaded l
]
rack module) was divided into one mass at the top of the rack model and one j
e' lumped into the base.
Kinetic energy principles were used first to estimate i
I the portion of the total fuel assemblies mass that was assigned to the top i
,,u d
fuel mass. Refinements to the proportioning procedure were made by comparisons with a parallel solution using a 32 degree-of-freedom model that
- .q; included a number of mass points assigned to the fuel assembly.
The i
a comparison to the 32 degree-of< freedom model served both as a refinement of a
l the procedure and a chech on the magnitude of mass assigned to the top fuel 6
L assembly mass point.
l3 In order to validate the dynamic results from the limited model, tha
)
Licensee coc' pared the computed dynamic response of the 8 degree-of-freedom model to that for a 32 degree-of-freedom model under the same acceleration o
(' d i
i TER-C5506-625 l
time history.
The following table was supplied by the Licensee in response to a request for more justification of.the 8 degree-of-freedom model.,[?):
4 I
32 DOF 8 DOF Time Step 1 x'10-5.sec*
5 x 10-6 se:*
5 x 10-6,,e l
Max. x-displacement-2.595 in 2.710 in 2.793 in Max. y-displacement 2.147 in 2.213 in 2.279 in Max. support load 1.464 x 105 lb 1.484 x 105 lb 1'.22 x 105 lb
~
on the foot l,)
Max. R6 in support "
0.272 0.279 0.280 leg i
Max. R6 in the 0.115 0.106 0.123 e
rack body y
- Two solutions, using different integration time steps Both the 8-and 32 degree-of-freedom models were tased on a Diablo Canyon rack module with a coefficient of friction of 0.2 between the rack modu:e supports and the pool floor liner using zero structural damping.
Note thr.c R S
denotes the ratio of the resulting stress to the allowable valv.e as discussed later in this report.
The close agreement between x and y dirflacement, support loads, and stress ratios indicates that the 8 degree of-freedom model based upon the assumption of a rigid-body rack module provices satisfcetory
~
analysis of the dynamic displacements under seismic excitation.
3.1.5 Frictional Force Between Rack Supports pads and the Pool Liner i
The Licensee used.a maximum valu'. of 0.8 and a minimum value of 0.2 for l
i the range of static friction coefficient between the rack support' pads and the pool liner (1).
Rabinowicz, in a report to the General Electric Company [8),
focused attention on the mean and the lowest coefficient of friction to be used in these circumstances.
While Rabinowicz supported the range of co-efficient used.by the Licensee, he also indicated that the cynamic, or sliding, coeff2cient of friction is inversely proportional to velocity.
The Licensee used an initial static coefficient of friction equal ~to a dynamic coefficient of friction.
However, the range (0.8 and 0.2).of friction coefficient used by,
\\
i
.TER-C5506-625 l
i the Licensee is considered to be sufficient to bound the influence of friction i
,1 coefficient.
Thus, the Licensee's use of friction coefficient between the
)
1 support pads and the pool floor liner is acceptable.
?.l.6 Hydrodynamic Coupling
' Hydrodynamic coupling is the effect upon adjacent rack modules, or upon a D
fuel assembly within a fuel cell of a rack module, resulting from fluid (water) motion induced by the movement of the racks, or fuel within racks.
As rack modules accelerate toward each other, or as a fuel assembly accelerates toward the cell walls, the fluid between the bodies is accelerated to flow out of that space.
The fluid mass causes fluid pressures to develop upon both bodies bounding the fluid.
As large bodies approach each other across a small P
fluid space, these fluid forces become large.
l The effective fluid masses, or coupling coefficients, were considered by the Licensee for motion between adjacent rack modules, between rack modules and the pool walls, and between fuel assemblies and the storage cell walls.
7 Fritz (9) provided data on the coupling coefficients for various body shapes and arrangements, as well as provided limited experimental verification of ths L.
technique.
Fritz's data were used by the Licensee for analysis of the fluid coupling in the equations of motion for fuel racks and fuel assemblies.
It should be noted that fluid coupling, as discussed here and used by the 1
Licensee, is coupling based upon fluid inertial forces and does not constitute n.
damping.
7
}
Fritz's work, as applied to the rack modules, is limited in that his experimental work, which compared wel.1 with his theory, was accomplished for l
1
).
l infinitesimal vibratory displacements and for relatively large fluid spaces i
between the vibrating body and the fluid boundary wall. While this does not j
exactly correspond to the conditions that prevail for spent fuel rack modules, the technique is based upon well established principles in fluid mechanics and l
serves to provide a lower bounding estimate of the fluid coupling for rock ri y
module analysis.
j TheLicensee'sconsiderationofhydrodynamiccouhilingassumesthateach Y
sdjacent rack module is vibrating in a manner equal and opposite to the ra:k j
module under analysis. This assumption permits the consideration of a rigid l ;
1
TER-C5506-625 1
1
' l boundary halfway between the two adjacent racks, and permits the analysis to j
proceed usig one-hr.lf of the real clearances between modules.
While l
techniques such as this are useful in making problems tractable, they isolate
_ j the rack module under analysis so that the differences in rack response with adjacent dissimilar racks, or racks vith different fuel loadings, are not 5j simulated by the analysis.
The Licensee has derived [3] hydrodynamic coupling terms for racks in a multi-rack environment, including the effects of finite I
l lateral spacing, rack rocking and twisting, and vertical flow.
Supplementary analyses [3, 4) demonstrated that the incorporation of these phenomena yielded smaller impact forces than analyses based on the hydrodynamic coupling t
I assumptions of the licensing report [1], for which infinite lateral spacing of adjacent racks and only the horizontal translation of the rack centroid were k
1 considered.
- 1 I
The Licensee's use of state-of-the art analysis to estimate the q
hydrodynamic,c mass and coupling is acceptable for analysis of the impact forces:
the t; thod includes the virtual mass of tLe water in the analysis of i
dynamic response, and conserva'ively underestimates the co pling forces to yield higher impact forces with adjacent structures.
)
l 3.1.7 Fuel Assembly Impact Within a Storage Cell Clearance is provided between a fuel assembly and the storage cell walls to permit bowed fuel assemblics to be inserted and removed safely.
- Thus, 1
under seismic excitation, the fuel assembly will vibrate within the clearance j
space and, when excited to higher amplitudes, will move through the clearance space before impacting the storage cell walls.
The model for fuel assembly dynamics is shown in Figure 9 (Figure 6..'.3 of Reference 1).
The stiffness of the impact springs was derived by the Licensee from the stiffness of only a small portion of the storage cell wall, instead of _ including the lesser stiffness of the more limber fuel auembly.
Thus, the impact springs used by the Licensee in the analysie were exceedingly
?
stiff. Two-and three-dimensional supplementary analyses [3, 4) were-performed comparing the impact spring constants of the design basis analysis
[1] with the more realistic values, confirming that the more realistic values for the impact springs yielded smaller impact forces and stresses.
1 j
l t
TER-C5506-625 j
Pluid damping was not used, but hydrodynamic coupling as discussed in the preceding section was employed.
Thus, the fluid coupling forces acted upon both the rack module and the fuel assemblies prior to impact with cell walls.
The Licens'ee demonstrated (10) significant additional conservatism in the Q
computation of the hydrodynamic coupling coefficients by using the nominal
{
rather than the actual gap size.
Structural or impact damping was not used in the analysis of fuel assembly impact within a storage cell.
The use of stiff impact springs and no impact damping resulted in very l
1 high computed impact forces that were moderated only by the hydrodynamic coup-ling.
Actually, the limber fuel assemblies usually have considerable hystere-I sis in cross-sectional deformation from which the Licensee could have derived d
considerable damping and a much lower stiffness for the impact springs.
The use of high stiffness impact springs and no damping were cited by the Licensee as added conservatism in the analysis.
While conservatism is warranted, undue conservatism in a dynamic simulation analysis should generally be avoided in order to simulate the dynamic interactions of the structural components as realistically as possible.
Evaluation of the Licensee's use of stiffer impact springs indicated, in this case, that the major effect was to predict higher impact forces which acted for a shorter period of time, but which did not alter, significantly, the energy transferre.d during the impact.
Another aspect of fuel assembly impact, within a storage cell, was th2.c the consideration of a single call, or stick, model dictated that only one fuel assembly, or one eg'ivalent fuel assembly, could be considered in the analysis.
This was tantamount to assuming that all fuel assemblies in a given rack moved in harmony.
However, when one considers that the analysis of tne racks is primarily concerned with large displacements and that large rack
?,
displacements are usually associated with liftoff and tipping of a rack in a definite mode (say, rocking about an axis parallel to its longer dimension),
it is more eanily accepted that the fuel assemblies can move in near harmony.
9 However, impacts with other rack modules, especially if impact occurs at a
^
corner, can distort the more uniform rocking rooticn.
While rack module impacts are' predicted for a Hosgri earthquake at the Diablo Canyon plant, the assumption of uniform motion of fuel assemblies adds a small degree of con-servatism to the maximum impact forces predicted.,
TER-C5506-625 3.1.8 Damping With respect to damping in the nonlinear dynamic displacement analysis, the Licensee supplied the following [1):
"In reality, damping of the rack motion arises from material hysteresis (material damping), relative intercomponent motion in structures (structural damping), and fluid drag effects (fluid damping).
In the analysis, a maximum of 4% structural damping is imposed on elements of the rack structure during HE seismic simulations.
This is in accordance with the FSAR and NRC guidelines.* Materials and fluid damping are I
conservatively neg3ected.
The dynamic model has the provision to incorporate fluid damping effects; however, no fluid damping has been used for this analysis."
3.1.9 Rack Module Impact Modeling The initial design analysis and early verification runs of the non-linear dynamic displacement analysis by the Licensee indicated that inter-rack impacts were possible.
The Licensee designed the racks to accommodate such impacts and incorporated impact springs and gap elements at each corner I
in the dynamic model.
Elements 10 through 24 of Table 2 list the inter-rack impact elements.
The following statement by the Licensee is taken from Reference 1:
"During these verification runs, it became apparent that due to the high level of slab acceleration associated with the Hosgri event, inter-rack impact could be anticipated to occur, especially for low values of the friction coefficient between the support and the pool liner.
To account for this potential, yet still retain the simplicity of simulating only a single rack, gap elements were located at the corners of the rack at the top and at the baseplate.
[ Figure 8) shows the location of these gap elements.
Loads in these elements, computed during the dynamic analysis, are used to assess rack integrity if inter-rack impact occurs. The S DOP mocel is used to avoid possible numerical problems due to the large I
number of nonlinear elements that would be required to model inter-rack impact with the 32 DOF model."
Supplementary two-and three-dimensional analyses [3, 4) demonstrated the overall conservatism of the single rack model.
Although impact between rack I
and wall was not predicted in the analyses of the licensing report [1], large l
impacts were predicted between rack and rack at the girdle bar and baseplate.
These are further discussed in Sections 3.1.12 and 3.3.2.1.
- US NRC Regulatory Guide 1.61, " Damping Values for Seismic Design of Nuclear Power Plants," 1973.
l ;
l E _ _____________________ ________ ____ _ _
l
I TER-C5506-625 Table 2.
Gap Elements anc Friction Elements
]
(Table 6.3.1 of Ref. 1) 9
.,cj Gi
[
I.
Nonlinear Sprinos (Cap Elements) (24 total) i Number Node Location Description 1
Support S1 2 compression only element 2
Suoport 52 2 :ompression only element 3
Sui rert S3 I compression only element 4
Support S4 1 compression only element
'n 5
2,2*
X rack / fuel assemoly impact element 6
2,2*
X rack / fuel assembly impact element 7
2,2*
Y rack / fuel assemoly impact element 8
2,2*
Y rack / fuel assembly impact element 9
Top cross-section Inter-rack impact elements il of rack (corners)
~-
10 Inter-rack impact elements 11 Inter-rack impact elements 12 Inter-rack impact elements 13 Inter-rack impact elements 1
=a 14 Inter-rack impact elements i
g,f Inter-rack impact elements l
15 16 Inter-rack impact elements 17 Bottom cross-section Inter-rack impact elements 18 of rack (corners)
Inter-rack impact elements
{
19 Inter-rack impact elements I
20 Inter-rack impact elements I
21 Inter-rack impact elements 22 Inter-rack impact elements 23 Inter-rack impact elements s,-
s 24 Inter-rack impact elements i
II.
Friction Elements (16 total)
, Number Node Location Description l
1 Support 51 X direction support friction J
2 Support S1 Y direction friction 3
Support 52 X direction friction 4
Support 52 Y direction friction l
P 5
Support 53 X direction friction j
6 Support S3 Y direction friction i
d 7
Support S4 X direction friction 9
Support S4 Y direction friction 9
S1 X Slab moment i
i 10 S1 Y Slab moment 11 S2 X Slab moment i
12 S2 Y Slab moment l
13 S3 X Slab moment 1
- q 14 S3 Y Slab moment 15 S4 X Slab moment 16 S4 Y Slab moment a
i a
J i
I L
TER-C5506-625 3.1.10 Seismic Loading The acceleration time-histories used to input seismic excitation to the nonlinear dynamic displacement analyses are shown as Figures 6.1.1 through While the updated FSAR for Diablo Canyon Units 1 and 2 6.1.6 of Reference 1.
indicates that the design of structures must satisfy three eartnquake conditions (DE = design eartheaake, DDE = double design earthquake, and HE =
Hosgri earthquake), studies p..rformed on a typical rack module indicated that the displacements and strr ses were more severe for the Hosgri event [1].
Therefore, the Licensee used DE and HE for the design for the rack modules.
The acceleration time-histories used in the Licensee's analysis included horizontal x, horizontal y, and vertical components for the DE and HE events.
The three simultaneous earthquake acceleration time-history -components were input to the three-dimensional, nonlinear displacement analysis of a rack module to produce the results evaluated later in this report.
3.1.11 Selection of Rack Modules for Analvsic The Licensee selected two rack modules for dynamic displacement analysis:
the largest module (10 x 11 cell configuration) and the module with the largest aspect ratio of its horizontal dimensions (6 x 11 cell configuration)
The 6 x 11 rack module was also the rack with the shortest horizontal
[1].
dimension.
Experience in the review of spent fuel rack dynamic analysis (11, 12] has indicated that the largest displacement response in a tipping, or rocking, mode generally is realized for rack modules that have the highest ratio of height to width.
Since all of the spent fuel rack modules for Diablo Canyon Units 1 and 2 are of the same height, then the racks having the shortest cross-sectional dimensional should be analyzed. The reason for this is that the natural frequency of a rack in a rocking, or tipping, mode, with liftoff of the support pads from the pool floor liner, is lowest for racks that have the
~
highest ratio of height to width. Two rack modules of the Diablo Canyon set have the same high ratio of height to width, the 6 x 11 and 6 x 9 module designs.
The Licensee chose the 6 x 11 rack for analysis. The 6 x 9 design, by contrast, would be expected to have similar dynamic response from the same 3
horizontal earthquake acceleration component directed across its 6 cell._
1 TER-C5506-625 l
direction. However, the 6 x 9 design could suffer somewhat greater response in the 9 cell direction.
Thus, the simultaneous response of the 6 x 9 design
]
from the two horizontal and one vertical earthquake components could produce I
somewhat greater total displacements, with possibly more nonlinear interaction
(
i between the two linear and one rotational motions in the horizontal plane.
In summary, any differences between the responses of the 6 x 11 and 6 x 9 f
rack module designs would not be expected to be large enough to cause any of i
the design criteria to be exceeded, given the conservative results that were f
I l
presented by the Licensee.
j i
)
3.1.12 Confirmatory Two-and Three-Dimensional Parametric Studies The reracking report (1) described the analytical methodology used for the time history analysis of the high density spent fuel racks that used generally conservative input for three-dimensional (3-D) single-rack models.
l l
The Licensee claimed that these analyses yielded conservatively estimated r
results, which, combined with the use of ample design margins, provided an i
adequate design basis to accommodate or envelope the effects of multi-rack i
behavior.
This conservatism was confirmed by the results of two-dimensional parametric studies (3) performed by the Licensee using both multi-rack and single-rack models.
Following the evaluation of the two-dimensional parametric studies, the NRC staff requested the Licensee to perform L
three-dimensional single-rack studies that would apply consistent realistic 1
1 l
1 i
impact spring and hydrodynamic coupling assumptions used for some of the i
s two-dimensional parametric studies to two specific design basis computer l
runs.
The NRC-requested three-dimen-sional study (4) further demonstrated that the design basis loads used for qualification of the racks (1) envelop l
t t
the results based on more realistic assumptions.
l 3.2 EVALUATION OF THE NONLINEAR DYNAMIC DISPLACEMENT ANALYSIS l
I j
3.2.1 Evaluation of the DYNAHIS Analvsis Method j
y f
The time-history solutien of the B degree-of-freedom mathematical model f
of the rack modules was described by the Licensee as follows (1):
I "Having assembled the structural model, the dynamic equations of motion
(
corresponding to each degree-of-freedom can be written by using Newton's L f
.TER-C5506-625 second law of motion; or by using Lagrange's equation.
The system of
~
equations can be represented in matrix notation as:
i
[M) {q) = (Q) + {G}
wher9 thd vector (Q) is a function of nodal displacements and velocities, j
s and 1G) depends on the coupling inertia'an'd the ground acceleration.
j Premultiplying the above equations by (M)-1 renders the resulting equation uncoupled in mass.
1 We have:- {q}-= [M)-1 (Q) + (M]-1 (G}.
This equation set!is mass uncoupled, displacement coupled, and is ideally suited for numerical solution using the central difference scheme.
The-computer program 'DYNAHIS' is utilized for this purpose.
Stresses in various portions of the structure are computed from known element forces at each instant of time."
i'!
In a later communication (13), the Licensee added that the algorithm of the code DYNAHIS is documented in the book, "The Component Element'~ Method in Dynamics" by Levy and Wilkinson, McGraw-Hill (1976), and provided the
.a ri following description:
j "O
"There are no numerical simulations.or modeling assumptions unique to DYNAHIS. Moreover, in contrast to a general-purpose finite element code, DYNAHIS does not internally generate the equations of motion'.
These equations are written out explicitly.and can be verified independently.-
Such verifications have been carried out internally at Oat, by several.
A/E firms, and, at the NRC's request, by Franklin Research Center in 1984."
i A detailed review of DYNAHIS was made during the evaluation of two I
previous spent fuel rack module designs [11, 12).
F i
In that review, DYNAHIS was found to be categorized as an equation y
solver, ?s opposed to well known finite-element engineering analysis computer programs. As the Licensee stated above, DYNAHIS does'not generate the equations of motion internally.
Rather, in the course of the analysis, the
,i analyst formulates the set of' dynamic equations from the mathematical model.
that was constructed (as discussed'in Section 3.1), and inputs those' q
equations,-along witn the initial conditions,-boundary conditions, and
-l excitation time-histories,'to DYNAHIS for solution.
l l
TER-C5506-625 e
DYNAHIS was determined, in the previous reviews [11, 12), to use a b
time-steg integration method based upon the central-difference method.
In-'
j 1
association with those reviews, a number of computer solutions were performed to verify that the central-difference solution method.was both stable and
'~
4 convergent.
It was determined that the integration method was.both stable and convergent, provided that the integration time step was maintained within 2
certain bounds unique to the specific problem.
Outside of those bounds, the displacement was seen to rise.very rapidly in magnitude, as is characteristic of central-difference integration.
Identification of integration time-step values within the acceptable region for a stable, convergent solution was accomplished satisfactorily by comparing the displacement solutions for two or more time-step values that differed by at least a factor of 2.
If the computed displacements we're the same, or nearly the same, then the solution j>
H-was found to be acceptable.
l In response to a request, the Licensee provided the following comparison of displacement computed using two time-step values [7):
7
" Convergency and stability of the runs are assured by running the same l'
problem at different time steps and observing the close agreement between the results thereof.
The output data of an 8 DOF run for a fully loaded 10 x 11 rack with coefficient of friction (COF) = 0.8 for two time steps are shown below.
l Maximum Maximum Max. R6 Maximum i
Time Step x-displacement y-displacement in Rack Floor Load u
5 x 10-6 sec 0.3402 in 0.4780 in 1.49 6.735 x 105 13 3
2.5 x 10-6 sec 0.3401 in 0.4783 in 1.52 6.727 x 105 lb q
The above table confirms the excellent convergence characteristics of the 8 DOF model."
g 1w lg In support of DYNAHIS, the Licensee submitted the background of the
'~
analysis method as well as four problems in which the solution by DYNAHIS was 1
verified by other recognized analysis methods [13).
This supporting material
.o 1
is included with this review as Appendix A.
1 Solutions by DYNAHIS for the spent' fuel rack modules for Diablo Canyon l
[
Units 1 and 2 ere considered'to be acceptable.
?)
I d
- ) l
[
_.______ _____________ _ _ ___ _ _ __ _ d
~~
TER-C5506-625 3.2.2 Ra.ek Displacements During the design and analysis of the rack modules, it was determined that the earthquake excitation was sufficient to cause impact of the rack modules.
Thus, instead of submitting displacement data, which is generally to show that the rack modules would not impact, the Licensee submitted a'n' impact I
analysis for the racks that were analyzed and compared the impact forces and stresses to allowable values. The resulting impact forces are reviewed in Section 3.3.2 of this report.
3.3 RACK MODULE IMPACT AND STRESS ANALYSIS 3.3.1 Structural Design Criteria With respect to load and stress criteria, the Licensee developed the allowable loads and stresses in accordance with Appendix D, Section 3.8.4 of j
the USNRC Standard Review Plan (14).
However, the NRC staff have indicated that operating reactors should be based upon the staff's OT position (5) rather than the Appendix D, Section 3.8.4 of the Standard Review plan.
In response, the Licensee submitted the following comparison of the loading combinations and limiting values of the OT position and Appendix D [7):
"The analysis presented in the Reracking Report meets OT Position guidelines.
As stated in the Reracking Report (1), the stress levels in the rack feet greatly exceed those in the rack modules proper.
Therefore, in essence, support feet stresses govern the rack design.
Since there is no interaction between To, Ta loads, and inertia loads (D, L, E, E') in the support feet, the Standard Review Plan 3.8.4 (NUREG-0800) and OT Position criteria both yield the following effective limits:
Loading D+L+E Normal limits as stat'ed in Appendix XVII-200 D + L + E' Normal condition stress limits increased by Appendix F-1370.
Thus, although there are differences in the stress criteria between the l
OT position paper and Standard Review Plan 3.8.4, Appendix D, thel' amount to identical limits for Diablo Canyon." -
TER-C5506-625 In reporting the computed stresses and their comparison to allowable values, the Licensee reported the ratio of the computed stress to the allowable I
I l
value as R through R '
""8'
- "9 6
y 6
1 h
follow [1):
b F
"R2=
Ratio of direct tensile or compressive stress on a net section to its allowable value (note support feet only support compression)
R2=
Ratio of gross shear on a net section to its allowable value R3=
Ratio of maximum bending stress due to bending about the x-axis to its allowable value for the section R4=
Ratio of maximum bending stress due to bending about the y-axis to its allowable value R5=
Combined flexure and compressive factor ij R6=
Combined flexure and tension (or compression) factor."
Evaluation of the Licensee's position indicated that it is acceptable, as are the allowable stresses derived from this position as shown in Reference 1.
i' ee 3.3.2 Impact Loads and Stresses Impact Between Fuel Assembly and Cell Wall 3.3.2.1 The Licensee provided the following description of acceptable loads [1):
"The local stress in a cell wall is (stimated from ped impact loads obtained from the dynamic simulations.
Plastic analysis is.used to obtain the limiting impact load that can be tolerated.
Including a a
d safety margin of 2.0, we find that the total limit load for the number of cells (NC) is:
NC Limit Load (Ib) 66 530100 i
100 883501 5
From the results of the dynamic analyses, we find the actual impact loads y
do not exceed 251,000 lb for NC = 110 and do not exceed 136,000 lb for NC 4
= 66."
As discussed in Section 3.1.7, the Licensee overestimated the spring constants of the impact springs associated with a storage cell by considering i
d -
5
'TER-C5506-625 only the compliance of the cell wall without including the compliance of the Ifuel assembly. Thus, where impact does occur, the computed force's'are nigher
)
than wouid be expected.
Actually, the energy of the impacting fuel assembly is the same; a' lower spring constant would spread the impact over a longer e,
time period at a lower net force.
~
&J The Licensee's impact forces and their comparison to allowables are shown in Tables 3 and 4.
It should be noted that the runs summarized in Tables 3 and 4 were originally performed and reported in the reracking report (Tables 6.8.1 and 6.8.2.of Reference 1) using nominal va'ues for Pic1d stresses and 7
section properties based on an earlier design for the upper support foot.
The tabulated results of load factors, R, were modifitd later through the g
application of calculated correction factors as deasiled in the seismic l
analysis report, Rev. 3 [2].
A review performed b'/ FRC of the technical basis of those correction factors concluded that one of ahim, the inertia factor, was applied instead of the section modulus factor to modify bending stresses 1
a for some load factors reported in Tables 6.8.1 and 6.8.2 [1].
Based on FRC calculations of the most critical load factor (R of Run No. acor: 10), it 6
was concluded that this error would not result in a significant increase of tnat critical load factor (R should be 1.548 instead of 1.436). Neverthe-6 less, Tables 3 and 4 show that all impact forces are below the allowable values and are acceptable.
The two-dimensional parametric studies (3) concluded that the use of conse-"ative spring constants and fluid coupling inputs in the reracking report analysis [1] yielded conservative rack and fuel assembly impact loads.
t f.
These studies provide a high 1cvel of confidence that the design basis model
~
[1] has predicted conservative rack qualification loads, which ensures compliance with all design criteria. Table 5 summarizes the maximum impact load values determined by the licensing design basis as well as those corresponding to the parametric studies (7 cases).
l 3.3.2.2 Impacts.Between Adjacent Rack Modules The Licensee provided for rack-to-rack impact as follows [1):
j..
"All of the dynamic analyses assume, conservatively, that adjacent racks move completely out of phase.
Thus, the highest potential for inter-rack impact is achieved.
Based on the dynamic *oads obtained in the gap elements simulating adjacent racks, we can study rack integrity in the,
TER-C5506-625 i,$.
J sicinity of the impact point.
The use of high-yield stress framing material around the top of the rack allows us to permit impact loads of up to 175,000 lb.
The maximum reported value in the tables in this i
~~'
di report is 71,400 lb.
Thus, impacts between racks can be accommodated without violating rack integrity.
We also study the case where the g
corner of one rack impacts an adjacent rack away from a corner. We show, p
under such a condition, that the stress levels remain below.the yield H.
value."
1 l
[
LU The Licensee compared impact forces to allowable values as shown in 1
4 Tables 3, 4, and 5.
In d
In response to discussion with the NRC staff concerning the effect of partially filled fuel racks, the Licensee supplied the following [13):
)
"The loading cases evaluated for the Diablo Canyon limiting rack modules I
as reported in the reracking report (PG and E letter DCL-85-306 dated 3
September 19, 1985) are as follows.
f, o
Racks fully loaded e
d o
Racks empty t
i o
Racks half full (for the 6 x 11 rack module only) y
.i o
Racks partially full (11 cells).
p Tables 3 and 4 (Tables 6.S.1
- nd 6.8.2, respe::tively, in the reracking report) also show that significant design margins exist for the above i
load cases.
L f;
The selection of load cases was based on past experience of Oat with cellular rack geometries (Fermi Unit 2, Quad Cities Units 1 and 2, Rancho Seco, Grand Gulf Unit 1, Oyster Creek, pilgrim, and V. C. Summer).
d Specifically, the selection process included the following:
o Review of licensing submittals from Quad Cities (Docket Nos. 50-254 and 50-265) 1 o
Table 1 [of Ref. 13) (a copy of Table 6.5 of a licensing submittal G
for Quad Cities) shows the results cf a number of cases evaluated for 2
Quad Cities.
In all cases, a fully loaded rack with H = 0.8 results in maximum stress response.
This conclusion is valid for the l
3 Diablo Canycn rack design. Accordingly, this load was evaluated for i
3 the Diablo Canyon racks.
C o
The kinematic response (maximum displacement) could be greater for l
asymmetrically loaded racks (other than the cases evaluated for l
Diablo Canyon). However, such loading cases are not critical for j
11 11 P I I
l
s 4gbUc t$u
(
i C
c 1
9 4
4 8
3 9
4 3
6 i
9 9
6 8
3 1
3 8
7 7
t 6
2 7
1 2
3 2
5 2
7 e
s 1
R h
a e
eTt l
s l( e p
a E
.'foh B
M t
e l
i t
ks 6
7 4
5 9
6 6
9 7
6 retw a
ct 2
3 6
5 1
8 7
6 8
3 ol t
i or 1
2 0
0 0
0 1
0 4
fbot t
1 Ro R
api i
p ewpm n
oui t g l
ou bl sl
- i fS al d
warl l
gr 5
4 2
5 3
9 4
8 7
4 o
ee-l 0. o e o'
u Ro 8
4 9
6 7
9 6
8 6
6 wv 1
3 f
1 s
R 1
4 0
0 1
0 0
2 1
3 l
w rd a2ll
(
h x
on
- 0. t h H eE c
t a i
6 c
4 8
3 0
5 7
9 5
7 4
2 et h
a 1
0 2
3 3
4 5
3 9
0 1 e 7
w e
F 2
p R
1 2
0 0
0 0
0 1
0 2
6 f
e s d
y d
stl1 a
a T
are o
o bovd l
L k
pee 1
6 3
3 4
0 7
7 9
5 k pl r e
c) 9 7
5 8
8 3
5 2
6 4
cu h
a2 R
i 0
1 0
0 0
1 0
1 0
1 a sd u t
R R
r aq f
roe r
e rol r e
of v
sR f
l e o
t sah l f srct 2
uo roi s
s*
8 5
ottt f
e1 0
0 0
0 t ci e o
no 10 1
0 1
1 cace
)
R b
et x1 x
1 x
x afcm y
6 f
e 5x 9
x 0
0 f
t i
wkk 65 1
7 5
1 l
oe tcc 57 deh e
l s
eaa 6
7 9
2 darc f
yb d
BRR 01 0
3 3
5 ol mh s
aooi a
ra a
L w
aT o
me f
m(
L e,u h s o
l 5
- 5 5
rmt r u
t ey 0
0 0
ai o
r S
c neul 10 1
0 1
xet o
a e gF b u1 x
1 a
0 navc t
p ea/m 8 4 1
x 0
1 mmaa c
m wrl e 60 9
6 0
x u
hf a
I t ol s 33 9
6 3
0 3
l r
f et es oatd BSCA 15 0
1 2
6 cl ea a
e eo l
i t
hmf b b
h a
/ c ci t
ast t i
T kk a s
s s
s s
e t e w
- +
ccp e
e e
e e
oov aam Y
Y Ys Ys Y
nn po e
O RMi l
l iapn u
l l
u 1
l e r e r f
sssh. a l
l c a c a l
ee cn v
g l
l e
e a
uueui eu eu e1 n) el nl e
nh llhmg t
kf kf k1 k
kI k
k i
aat t
c
+
a a
a sc a sc a
Y vv ea a
s u,
u, u,
ea u, ee u
l f vm p
h q8 q2 q2 t t q8 tr q8 l e ffoa m
t a
a uv oo ht i
a i
i i
l y i l y fi t
n m
r=
r=
r=
ut t = ut t=
t ttrsa e
e g
g g
mp g mp g
fi eeanc l
R sF nF sF i m
oF i m sF l s sspoi b.
o0 oO oO se oO se oO ao if an N0 H
N
(
HC
(
NC hp rrrti wo eeean oi pw pc g l t popoi l
c 1
2 3
4 S
U bul s Aa n.
0 0
0 0
O io e
e e
e e
a RN e
o e
e e
d, $
s
.l1l il l;i I
illl l!ll l
l
r.
e
\\
'a, TER-C5506-625
)
a m
m e e
N m
o e
e J
4 w
w C&
es N
es em N
m o
em N
gwgg e
g e
e e
o e
m 3
=a O
9 U e
en a
em DVw La e
==
O *=*
g Ja(.
9"C e
e e
vi.
m o
e m
a ga m
o e
e m
C 4
e W&
D -
ue
==
w O
o O
O O
Ekd*
6 E
e e
e e
- W#
W d
uw w
e 2
>r.
g C owd.
C
' l y,a we w
=
- )-
O G "3 om e
w N
e n
m e
,*9 w
w N
e e
m m
e e
ea w
w e
e,e e e
o N
N O
C "D
mo g
e e
o e
ba N
D e3&w e
gw 30D e 3
':A:
C*nn M
em o.ie wC N
M N
e W
W m
e 4
~j O#
e N
N
- 8b*
U N
o N.
- o. o.
m.
O.
O.
. t> C m
=
y a,
a g
e a
e
- 425 Nw T
,f no
@ U m
e e
o m
e N
e em m
- 4e O
O en e e
N N
O e
m 99Ww a
m,e ee5 en m
O aus Q
.a em O
e e
e o
e tw e
p.e g
e e
e D ee D
Xn 9~p e
l aw9 j
h.,
u n 00>
w W
tc 4 G"S W D
I w
3 j
y 3
0 I W w eT O l,..
f*
m%
O eC N
we e
twwC=
U
- e Om a w
y%
4 0
o*
o o
o ew -
o CC
-C
- a w *" E a
U N
.C GG N on N
N N
O@ @=
7'"
wN O
O N
& wQ" u
M
% c*
3aa
-e M
M M
00*
w G UU W&
9 RGGG w
l t e4 e e e
w Um 4 e
W S.)
O.gj T
aaa e=
e
. N ew*
a 4
9*O
>p 3
e e
a w
m A 'c *
- O eO
(*-
"H e
e e
,k LI j2 W
cc c
c c
wo w
u Ce -
- a e
4
@ D b J)
MM M
N W
C w
2 W
C 4%s WC W
P EEtO O
w em N
e M
$4=
c I
U E
a o "= m
.N.. e e
3Esz w
{
ee w
w
=
- t ea@m e e e
e
>e3 o-3 if.
EMUd Ne N
N N
OC
.s * @
e T
E=C e>E Q
&ER w W
W em2 4C s
s o 0 w o
aa.
...o e.au 3
oa e
e
> ens a ec to u.E C-.m e
E, 1
=
a a
..a.
1 w t. A l
E.-
,4 gS.
gN.
gw.
Sw.
e.
N.
- u. =e a
e.
e=
e.
o.
,..f
- s. :
u w
w
-w w
w w
.d.
a w.
w..
w..
D OC "9
OC "T
w&
C
.a=
1 E
O "t
o h. " 9 a e awhae e.bme Qec=
s q
eb=e E
OC2C CO3C C0030 00030 a4 u
c EOh" 3Ob=
EUUb=
2 U U h. =
uewI t 4.
O>
vo
=
wc]sC A.U E
w e
=
O N
w C ba 3
=
t aGO Cr l
4Mw>w
c e,
N C
C L@ eem
" "a j
h]
i C
C w
he
= m GK E d
C
- O C
C C
.i DC e
a v
U e
l
[;;
EE 4
4 4
4 e
I Yk I
1 s.
l.
l p,
4 2 7 5 7 3
1 7 3 1 8 2 4 E
e no e
1 9 2 9 1 9 e
5 4 r
0 3 1 0 o
4 t
f kr A
eo s
p e
p u
u rS 3 6 0
1 l
o 9 3 7 2 a
1 v
0 3 1
3 f r o I mu
.i f m
.dn 5 4 3 7
)
)
i ra 5 6 2 5 k
k x
2 c c a
o 0 1 0 0 a
a m
t R
c r
r e
e r r h
r o
o t
d 7 3 3 8 l
i 3 5 6 9 F
r s
mR 0 1 0 1 e e e
g t
t m
n n i i
I t
(
(
8 S 8 S
O0 0
0 I1 1
)
e t
xx x
ra n
79 8
o s
78 6
C 0
e 1
61 1
c
)
no
(
b et x
0 r
1 e
o 1x f
(
wkk 4
6 4
tcc 9
s r
e eoa e
)
e d
BRR 4
i 4
e e
l ut n
l a
b lo t
i b
m k
ao o l
l m
a I
ey '
e 0
c vf o
T t
neul 0
s 1
o f
e r
.1 n
c e gF b 1
s a
a/m x
a 7
n 1
a p
rl e 6
1 y
ef l
3 3
mo f
p l
t m
tol s e
p o
(
.e t e s i
SCA 2
u 2
m e
n f
e tu e i el u
t gk s
ea l
e
/ c nc e
fv a
l kka s
i a s
t v
b ccp e
nr e
a 4m i
aas Y
i Y
t u
m s
u RRI ay u
fm u s
m oi m
o tt s
k np k
l g
x i p
c ome c
ee n
ma x
i uM a m
,k a
cek o
a d
S M
u r
a r
esu eu s
m s
r8 t eQ r
Q o =
= i x
k o
tt o
t l
r i
eai i
ei A
E a
a r=
cl r r
cr r
m.
m e
ug e
g o
k k
t e
tP 1 mn t
e o
c c
eo R
nD 1t o n
o l
a o ho iC
( eH i
H F
R R
Tf 3
4 0
0 0
0 n.
oo a
a RN a
a e$:
~
l TER-C5506-625 Table 5.
Maximum Impact Loads (Table 7.1 of'Ref. 3)
~
i I. 4 0
Rack Zapact Loads
.3 h
emet wedel(a)
Assuretion(b) ggg1(c) gggg(d)
,9 Licensing Basis SR C
249 105 n
1 SR C
129 90
^'.,
2 SR R
67 78 Q
3 MR R
72 74 4
MR R
78 75 J:
5 MR R
71 107 6
MR R
77 63 7
MR R
75 80 h::
843(e) 175(r)
D Allowables 200(v) v f
W 4
?f (a) SR. $1ngle Rack Analysis MR = Multiple Rack Analysis
-i (b) C = Conservative Assumptions Used in Analysis (high spring constants and
%)
i:)
targe hydrodynamic gaps)
R. Realistic Assumptions Used in Analysis l
(c) Maximum fuel tapact load (rack-to-fuel), all loads in kips l
(d) Maximum of rack-to-rack or rack-to-wa11 loads, all loads in kips f
E)
(e) Allowable for 10 X 11 rack; Allowable for 10 X 10 rack is 803; (r). rack
]
and (w) = wall l
J l
U i
r i
i hl F i l
4 1
M9h 2 5 em 7
06/03/87 40:42 8'. 02 -
'JUN.23 '67 11:29 FRANKLIN RES. CENTER. lH8LA. PA.
~
TER-C5506-625
]
i Diablo Canyon since the racks are intentionally engineered to accommodate rock-to-rack impact, with a high mergP of safety against plastig deformation.
As a result of discussion with the NRC staff on February 20, 1986, PG
]
and 2 svaluated en additional loading case involving an asymmetrically loaded 10 x 11 rac': moduls.
One quadrant of the moduie le assumed to
.j be loaded with fuel assemblies. The results are as follows:
j l
m!n=
4 DISPLACINDff LOAD FACTORS I
(Too corner)
(Maxingn valves occur in sunoort loos) j
)
R-R2 83 24 R5 86 RW CCF SQ Dg D
1 y
\\
N04d
.4 DE
.0748"
.0900"
.079
.070
.173
.164
.303
.345 j
N01
.8 HD
.1320"
.2627"
.105
.125
.293
.240
.455
.519 N03d
.2 DE
.0292"
.0353"
.042
.014
.033
.021
.065
.073 f
N02
.2 HE 1.0920"
.0664"
.194
.064
.141
.099
.310
.333 A conparleen of the above results with the displacements and load factors f
documented in Table 4 (Table 6.8.3 of the raracking report (11) confirme PGandE's judgment that a fully lo,nded rack with W = 0.8 provides maximum response "
3.3.3 Rack Weld stresses The Licensee indicated that the maximum weld stress in the rack occurred in the supports as follows (1):
"The critical weld locations under seismic loading are at the connection of the rock to the baseplate and in the support leg welds.
For the rack welds, the allowable weld stress is.the ASME Code va2ue of 18,520 psi.
The welda at the rack base are fillet welds.. Accounting for skip-welding in this location, the. maximum principal stress is 7151 psi..
Wald stresses due to heating of an isolated het esil are also computed.
The sesumption used is that a single cell is heated, over its entire length, to a temperature above the value associated with all surrounding cells. No thermal gradient in the vertical direction is assumed so that the resulte are conservative. Using the tettparatures associated'with this unit,.we show that the spot velds along the entire se11' length de not exceed the allowable value for a thermal loading condition.
The maximum computed shear stress in the spot welds, at the most critical location, is less than 6,400 pai under the HE condition."
3ased on the largest loads cortputed by the Licensee [2] for the junction between the upper support leg and the t<ase plate, weld stresses were computed
~40-c
06/03/87 -
10:43 NO.005 003
-I JUN.03 '67 18:33 FRANKLIN RES. CENTER, PH8LAs PA.
P.03 j
TER-C5506-625 e
~
by F3td, in accordance with generally accepted methods, f.or the combined groove and fillet welds.
These stresses, shear and normal, were found to be less than ASME allow-ables, which vary depending on stress classification. Minimum factors of safety of 1.1 in the inside groove weld and 1.3 in the outside weld were calcu-1 lated by FAC, taking no credit for pinsticity as was done by the Licenses in a l
supplementary submittal in which a minimum factor of saflety of 1.92 was computed.
Rvaluation indicated that the wald stresses are ecooptable.
3.3.4 Rack Module Timeine stability The Licensee provided the following as evidence of rack module stability in the tipping mode (13]1 "With regard to the safety margin for overturning for the limiting rack module, the OT position paper, by reference to the Standard Review plan, atipulates that a factor of safety of 1.1 over the extreme condition ground motion (Hosgri earthquake for Diablo Canyon racks) he provided against rack overturning.
Towards this end, the rack module with the worst aspect ratio (6 x 11 redulus was run with 1.1 times the Hosgri seismic excitation.
A coefficionb of friction of H = 0.8 was used to produce the maximum overturning condition.
The x-axis is parallel to the short side and the y-axis is parn11e1 to the long side.
The x, y origin is at the rack centerlina.
Two cases of eccentric 20ading are considered.
Both cases have 50% of the colla filled with fuel assemblies.
Case 1 loads all fuel in the positive x half of the racks case 2 loads all fuel in the negative y half of the enck.
The following table sunmarises the results of the two analyses.
Rotation in x-s olane. daarnes Eptation in v-g_ plane, deernes Value for Factor value for Tsotor Load Calculated Incipient of Calculated Incipient of Casa Maximum Tioolne 5$1ej:V Maximum Tionine Safety 1
0.58 19.36 33.4 1.15 35.94 21.2.
1 0.77 21.14 27.4 1.26 33.38 26.5 These results show that a large rsergin of safety against kinematic l
lastability exists in all rack raodules."
3.4 SPENT FUEL p00L ANALYSIS FOR MGt DDJS2TY FIAE. RACK 8 3.4.1 $sent Fuel pool Overview the 1.icenses provided the following description of the spent fuel pools at the Diablo Canyon Units 1 and 2 an well as a statement of the modifications made to the pool to accommodate the high density spent fuel rack modules (7):
i TER-C5506-625 i
"The spent fuel pools are located on each side of the east end of the j
auxiliary building.
Figure 11 shows the plant layout and foundation j
elevations.
Generally, one half of the auxiliary building is symmetrical j
to the~other, with each half of the structure containing equipment for j
one unit.
The Unit 1 spent fuel pool is shown in Figure 12.
The walls of the spent fuel pool are 6 feet thick except for local areas around the fuel transfer tube, as sho.wn on (Figure 12). The foundation slab under s
the spent fuel pool has a minimum thickness of 5 ft and is founded on approximately 5 ft of lean concrete placed on rock.
The spent fuel pool sides and bottom are lined with stainless steel, 1/4 in. thick on bottom and 1/8 in, nominal thickness on the sides. -Figures 5 and 6 (of Ref. 7) show the typical layout and details of the liner.
F The spent fuel pool structural analysis, addressed in the FSAR, is i
affected by the high density fuel racks due to the following:
r o
Weight of the proposed high-density racks o
Thermal gradient across the walls o
Interaction between the rackr snd the pool structure.
There are no physical changes to the spent fuel pool concrete structure addressed in the FSAR analysis.
However, minor modifications to the.
liner plate will be made to accommodate the high-density racks. The modifications include relocation of spent fuel handling tool brackets to the east end of the north and south walls of Units 1 and 2, respectively.
[ Figure 11] shows typical details of the modified brackets."
eo 3.4.2 Spent Fuel Pool Analyses The following analysis, a part of the updated FSAR [6], was cited by the Licensee as pertinent to the spent fuel pool, although no change was required t
i to these analyses (7):
"As described in the FSAR, Section 3.7.2.1.7.1, the seismic inertia loads
,)
were obtained using a time-history analysis of a spring and lumped mass model of the auxiliary building. Two horizontal models and a vertical model, shown in Figure 8, were used in the analysis.
[j A detailed analytical static model of the auxiliary building (Figure 9) was used to distribute seismic inertia forces and moments to various walls, diaphragms, and columns, as described in the FSAR, Section
.l 3.8.2.4."
r.
t Other analyses were updated to consider the change in fuel mass and floor
['
'l loads associated with the high density spent fuel racks.
Included in analyses that were updated was the dynamic response of the spent fuel pool under i
seismic excitation.
While the change in global mass was determined to be less 1
J
((. ]
I-( ' '
TER-C5506-625 l. ;'
fl.)
?)
RELOCATED
~'
m BRACKET A
n i
T g
OLD B ACKET k f
fj d
I a
UNIT"2 POOL SYMETRIL4L jj Ascu THis UNEJ l
PUN D)
SPENT FUEL Pool Q
101 UNIT *I
, I 9".
L I' ~
rPOOL LINER O
'}
, {*
f EL. /39'- 9" h*E
,p00t.UNER
, c.
d g {
- j
]B
\\
C M n
EL.IB9'-3' 0
CN!
\\ 5~~
'2,
~3 V f:))
/g ExtST.
EMBEOGO d SPUT FUEL ASSEMBLY
/
L TOOL BRACKET '
( 3{x /Lo**1 3*)
SEC TION - B ~0 f]_
SECTION A-A I
- -)
b LA Figure 11.
Related Spent Fuel Pool Brackets (Figure 7 of Ref. 7)
'i W
"$)"
'M
7 6
TER-C5506-625
)
e*'
N
~
~'
J L Oy.
3, y
,l 24 o*
32'O' l
.l
- 4. )
-y 4;
99
~
~ ~ ' '
.f\\
x r.
4
/
s
,/
N E
- l
'g O
h C
'f
'i.'..
4 fa',
/s's" 4
N
${l g
i X
.+
m N
e.*
Y 1
.J t
io F /
G H 5 pea sg f.
_h
-f t
.I !
4 a
Y h
e 4,,
1
+
w g
6
)
f (s,'..'
t lll:
9
'O Gbe G L'E.W
/
e I
.i 5
)
/
8
'I F.
N 4
~
E' L
l 6
Os ef g4 a
t>
!j j i-j i
e.
z'o* l4'0 260' 7' o' c' o' 2*.o
- J!f*. *k 7
--w-l Figure 12.
Plan of Spent Fuel Pool Walls (Figure 3 of Ref. 7) b e,
c
TER-C5506-625 n
than 1%, the Licensee included analysis to document the fact'that changes in-response were insignificant.
Other analyses that were modified were described by the Licensee as follows [7):
"The pool walls were analyzed for local out-of-plane effects due to hydrostatic, hydrodynamic, and seismic loads using U.S. Bureau of Reclamation tables for plates (Ref. 1).
Hydrodynamic effects were determined based on the methods in TID 7024, " Nuclear Reactors and.
Earthquakes" [Ref. 15), and " Dynamics of Fixed-Base Liquid-Storage Tanks" (Ref. 16).
Thermal effects in the pool elements due to the increased thermal gradient generated from the-use of the high-density racks were analyzed considering. cracked concrete section.
The liner has been evaluated for loads and load combinations described in the [following section).
The mathematical model used in the thermal evaluation consists of a set of springs representing the axial stiffness of the liner and flexural stiffness of the embedded anchors."
Other considerations included the following:
o pool floor vertical dynamic response o
pool wall impact o
pool floor bearing loads o
pool leak chase system o
cask pit.
J For the analyses unchanged from the FSAR, the Licensee provided the following summary (7):
" Global loads - Contro11'ing regions in the walls of the auxiliary building due to in-plane effects are summarized in FSAR Updated Tables 3.8-14, 15, and 16 for the DE,'DDE, and Hosgri events respectively.
The pool structure wall stress ratios from these tables are tabulated in Table 1 (reproduced in this report as Table 6)."
3.4.3 Load Combinations and Acceptance Criteria The Licensee indicated that the following load combinations and acceptance criteria were used for the analysis of the spent fuel pool [7):
" Concrete Structure Load Combinations The loads and load combinations used for the evaluation of the spent fuel pool structure are in accordance with the FSAR Update, Section 3.8.2.3, and are listed below.
A.
Normal Conditions Dead load, live load, loads from the DE, thermal loads, and pipe reactions are considered in all possible combinations.
Inasmuch- -
l
.TER-C5506-625 Table.6.
Auxiliary Building Spent Puel Pool Concrete Walls Stress Ratios (Table 1 of Ref. 7)
Shear. esi Noment r 103. k f t Stress Wall location (a)
W Q3sgj Cavacitv(b)
Q3snd Caeacity(b)
M (c)
A.
DESIGN EARTHQUAKE HALL AD 1Q0 li 220 lil 143 2J 85 85 220 164 664 2.6 HALL BC 100 45 155 20 113 3.4 NALL CD 100 80 250 54 146 2.7 HALL AB 100 80 240 54 145 2.7 i
B.
DOUBLE DESIGN EARTHQUAKE HALL AD 120 140 1!D 221 22 2J 85 170 370 316 1,086 2.2 HALL BC 100 85 260 40 244 3.0 MALL CD 100 160 420 105 280 2.6 HALL AB 100 160 410 106 268 2.5 C.
HOSGRI EARTHQUAKE WALL AD 100 11Q 210
.411 110 IJ 85 230 320 493 979 1.4 MALL BC 100 130 190 60 197 1.5 '
HALL CD 100 330 480 220 339 1.4 HALL AB 100 270 460 199 324 1.6
\\
(a) Forwallidentification,seeFigure(12). Counterparts in Unit 2 are similar.
(b) Axial demand effect is included in the capacities.
(c) Stress ratio = capacity / demand for shear or moment, whichever is smaller.
TER-C5506-625 as working stress design is used for normal operating loads, the factored load approach is not used.
For each structural member, the combination of these loads that produces the maximum stress is used for design.
Stated in equation form:
d = D + L + DE + To+Ro (1) where:
C = Required capacity of member based on working stress of ACI 318-63 as supplemented by Section 3.8.2.5.3 of the FSAR Update, which is described in response to Item 15d (of Ref.
7)
D = Dead load of structure and equipment loads including hydrostatic loads L = Live load DE = Loads resulting from the design earthquake To = Thermal loads during normal operating conditions (inside face of wall 176*F, exposed face 91'F, and bottom face of lean concrete 101'F)
Ro = Pipe reactions during normal operating condition's.
B.
Abnormal Conditions Dead load, live load, earthquake loads, and loads associated with accidental pipe rupture are considered in the following combinations; for each structural member, the combination that produces the maximum stress is used for design:
l UeD+L+TA+RA + 1.5 PA (2)
U=D+L+TA+RA + 1.25 PA+
1.0 (Yj + Ym+Y) + 1.25 DE (3) r U=D+L+TA+RA + 1.0 PA+
m + Y ) + DDE (4) 1.0 (Yj + Y r
U=D+L+TA+RA + 1.0 PA+
1.0 (Y3+Yn+Y) + HE (5) r where:
AA = Thermal loads on structure generated by a postulated pipe o (for T, inside face of walls 214*F, break, including T A
exposed face 93*F, and bottom face of. lean concrete 113*F) i RA = Pipe reactions on structure from unbroken pipe generated I
by postulated pipe break conditions, including R o. _ _ _ _ _ _ - _ _ _ _ _ _ _ _ - _ _ - -
TCR-C5506-625 l
r PA = Pressure load within or across a compartment and/6r
~
building generated by.a postulated pipe break, and including an appropriate dynamic factor.(DLF) to account for the dynamic nature of the load t
l Yj = Jet load on structure generated by a post. lated pipe break, including an appropriate DLF Ym = Missile impact load on a structure generated by,'or during, a postulated pipe break,-such as a whipping pipe, l
including an appropriate DLF Yr = Reaction on structure from broken pipe generated by a postulated pipe break, including an appropriate DLP i
U = Ultimate strength required to resist design loa'ds' based on j
the methods described in ACI 318-63.
For load combinations involving Yj Y, and Y, local stresses due to those m
r concentrated' loads may exceed the allowable provided there is not loss of function.
See Reference 17, Enclosure 3, Document (B), for more detailed information concerning the-i i acceptance criteria for these load combinations.
DDE = Loads resulting from the double design earthquake HE = Loads resulting from a Hosgri earthquake.
7 i
For the spent fuel pool structural evaluation, P, Yj, Y e A
m and Y are not applicable.
r Liner Load Combinations To assure leaktightness of the liner, the following loads and load combinations are used in the evaluation of the liner:
D+L+T (6) o D+L+Ta + HE
'(7)
Concrete Structure Acceptance Criteria I
'r The acceptance criteria as described in the FSAR Update, Section 3.8.2.5, are summarized below, f
l For DE and DDE load combinations, the nominal design strength for concrete and specified yield strength for reinforcing and structural steel are considered.
For load combinations including HE, however, the actual material properties are used.
1.
Normal Loads For normal loads, the auxiliary building is designed for the allowable working stresses of ACI 318-63, as supplemented by Item 3 below.
TER-C5506-625 2.
Abnormal Loads For abnormal loads, the-auxiliary building is designed for overall elastic behavior.
For concrete elements, the strength design of ACI 318-63 applies, as supplemented by Item 3 below.
Y, and Y as For load combinations involving Yj, hose concentrated loads may m
r applicable, local stresses due to t exceed the allowables provided there is no loss of function.
See Reference 17, Enclosure 3, Document (B), for more detailed information concerning the acceptance criteria for these load combinations.
3.
In-Plane Loads on Concrete Elements The design of slab diaphragms and shear walls for in-plane forces is not explicitly covered by ACI 318-63.
Section 104 of ACI 318-63 allows criteria based on test data to be used for the design of elements not covered by its provisions.
Consequently, the document titled " Recommended Evaluation Criteria for Diablo Canyon Nuclear Power Plant Auxiliary Building Walls and Diaphragms" (Ref. 18) is developed to provide criteria for evaluation of auxiliary building shear walls and floor diaphragms for in-plane seismic forces, including the simultaneous effects of out-of-plane forces.
Accordingly, the structural elements are evaluated as follows:
(i) The columns are evaluated by the provisions of ACI 318-63 for all loading conditions.
(2) The slabs and walls are evaluated for out-of-plane loads according to ACI 318-63, and for in-plane loads according to Reference 17.
Liner Acceptance Criteria The allowable concrete bearing pressure under rack dead loads is in accordance with ACI 318-63.
The design for local impact loads is not covered by ACI 318-63.
ACI 349-80 is used to determine the allowable bearing pressure on contrete due to rack impact.
The allowable foundation pressure is in accordance with FSAR Update Section 3.8.4.1.4.
The allowable liner strains and anchor displacements for normal operating and accident thermal conditions are in accordance with ASME Section III, Division II, 1983."
l 3.4.4 Pool Floor Vertical Dynamic Response The Licensee provided the following summary of the vertical impacts on the pool floor structure (7):
TER-C5506-625 "The interaction between the pool structure and the rack modules is primarily between the support feet and the pool slab.
The table below shows the maximum cumulative suppott reaction (sum of 4 feet) for the two representative modules studied (Hosgri earthquake):
Module -
Maximal value of Impact
'Tppe Dead Load (b) cumulative reaction (a)
Factor = a/b 10 x 11 158,676 lb 677,000 4.27 6 x 11 97,652 lb 405,500 4.153 An impact factor f of 4.5 can be safely taken to bound the data for all modules.
Since there are 13 types of modules in a pool containing 16 modules, these maximal reactions will occur at different instants of time, and statistical summation of these reactions is appropriate.
Dynamic amplification of loads due to soil-structure interaction is determined to be insignificant because the structure is supported on bedrock."
s 3.4.5 Pool Wall Impacts Analysis by the Licensee determined that (7):
"Those empty racks which are located on the outer periphery of the pool to within 4 in of the pool wall may impact the wall under the Hosgri earthquake. The associated impact loads are small:
the 10 x 11 module (empty) develops a maximum load of 3P 550 lb with a total duration of
.001 sec. The wall impact loads for s11 proximate modules may be conservatively estimated to be 80,000 lb under HE conditions, applied as a line load over the length of the girdle bar facing the wall."
In response to discussions about the frequency of wall impacts by the spent fuel rack modules, the Licensee submitted the following (13):
"The rocking frequencies of the racks as determined by Oat
- are less than 10 Hz.
The fundamental frequencies of the pool walls are greuter then 30 Hz.
The two frequencies are far apart, hence the potential interaction effect between the rack modules and the pool wall is insignificant."
Supplementary analyses (3, 4) performed by the Licensee indicated that impacts may occur between rack and pool wall either at the girdle bar or base plate level of the rack module. These impacts were enveloped by the design basis rack impacts: however, they may lead to additional loads on the pool wall. The Licensee previously computed (7) that the most critical wall
- Joseph Oat Corporation, Camden, NJ, spent fuel rack module vendor. ___ _ _ _.
TER-C5506-625 subjected to the Hosgri Earthquake was Wall AD with a capacity to demand ratio of 1.4 (Table 9).
Accounting for the largest girdle bar impact predicted in
,the supplementary analyses, this ratio would be reduced to approximately 1.1.
The fluid effects on the wall are much more signficant than the' effects of maximum possible rack impact.
With the loads and stresses lumped into the values shown'in Tables 7, 8, and 9, evaluation in this review indicated that the response of the wall to impacts is satisfactory.
3.4.6 Pool Floor Bearing Loads Maximum bearing loads on the spent fuel floor were investigated by the Licensee as follows (7):
"The maximum vertical impact load under the support leg of rack modules is 189 kips for the Hosgri load combination. This load is transmitted to foundation slab and bedrock via bearing. Two cases are considered:
1 (1) 6-1/2 in, diameter adjustable rack foot over leak chase l
(2) 6-1/2 in, diameter adjustable rack foot away from leak chase The bearing pressure'is computed by discounting the area of leak chase drains below the rack leg. The maximum bearing pressure computed is 7.5 ksi, which is below the concrete (8.4 ksi) and steel (24'ksi) allowables.
However, for ease of installation, provisions are being made to provide bearing plates with minimum dimensions of 12.5 in x 12.5 in x 1 :n under l
each foot of the racks.
This would further reduce the bearing pressure l
to 4.5 ksi and increase the factor of safety against the allowable value.
The effect of the impact load on the foundation bearing pressure is determined to be insignificant because the maximum computed contact pressure underneath the fuel pool is 7.2 ksf, which is significantly lower than the allowable foundation bearing pressure of 80 ksf used in design of adjacent structures (FSAR Update Section 3.8.4.1.4).
y i
The maximum horizontal sliding force on a support leg of rack modules is estimated to be 151 kips.
This sliding force is resisted by frictional i
resistance between the steel liner and concrete floor slab, and'by bearing of the stiffener angles embedded in concrete.
Two loading cases J
are considered:
(1) Sliding force resulting from four legs of interior rack modules clustered together (2) Sliding force resulting from one leg of a module. l
l a
TER-C5506-625 Table 7.-
Auxiliary Building Spent Fuel Pool Concrete Structure Stress Ratios-(DE)(a)
(Table 2 of Ref. 7)
Demand Values Cacacity Values Moment Shear Moment-Shear.
Stress Hill (b) k-ft/ft kltt_
k-ft/ft 1/.f,t.,
jltti;L, AB 195 31 220 56 1.1 s
225 42 245
'56 1.1 CE 145 20-220 56 1.5 ED 180 27 250 65 1.4 I
GF-215 27 260 46 1.2 1
i Basemat 210 small 275
!arge 1.3 (a)
Stress ratio = capacity / demand for shear or moment, whichever is smaller.
4 (b)
For wall identification, see Figure 12.
i 4
l e l
TER-C5506-625 Table 8.
Auxiliary Building Spent Fuel Pool Concrete Structure Stress Ratios (DDE)(a)
(Table 3'of Ref. 7)
Demand Values Caeacity Values Moment.
Shear Moment.
Shear Stress j
Eg11(b) k-ft/ft gl.f.1 L.lifft Lt.t.1 11110 AB 265 38 420 85 1.6
{
8C 295 42 460 85 1.6 AD 305 52 460 85 1.5 CE 195 25 420 85 2.2 ED 235 30, 470 99 2.0 I
GF 285 32 490 71 1.7 Basemat 265 small 520 large 2.0 (a) Stress ratio = capacity / demand for shear or moment, whichever is smaller.
(b) For wall identification, see Figure 12.
)
i '
- l TER-C5506-625 Table 9.
Auxiliary Building Spent Fuel Pool Concrete Structure Stress Ratios (HE)(a) i (Table 4 of Ref. 7)'
?
i r
l Demand Values canacitv Values Malj,(b)
AB 300 45 500 92 1.7 BC 355 55 550 92 1.5 AD 365 68 550 92 1.4 l
CE 200 31 500 92 2.5 ED 245 34 570 107 2.3 5
GF 295 37 590 76 2.0
)
Basemat 345 small 630 large 1.8 k
e6 l
l (a) Stress ratio = capacity / demand for shear or' moment, whichever is-smaller.
(b) For wall identification, see Figure 12, i
k s
)
I l
?
i l
i l
l TER-C5t' 625 I
The liner plate' weld seams are checked for the maxin.um axial force developed as a result of diference between the rack sliding forces and the resisting forces.
The maximum force computed in the liner weld seam is 1.5 k/in, compared with allowable force of 6 k/in.
3.4.7 Summary of Concrete Loads and Stresses The Licensee provided a summary of computed loads and their comparison to allowable values in Tables 7, 8, and 9 [7].
It was noted that Tables 7, 8, and 9 include loads that changed as a result of introducing higher density l
fuel storage as opposed to Table 6 that summarized the unchanged FSAR position.
Evaluation in this review indicated that the' load values reported are satisfactory.
l l
3.4.8 Pool Dynamic Response Under the Increased Spent Fuel Mass i
Although the change in mass due to the increased density of fuel storage i
only increased the mass of the spent fuel pool dynamic system by less than 1%,
1 the Licensee choose to demonstrate by the following that the change in j
response was exceedingly small (13):
l "As described in FSAR Section 3.7.2.1.7.1 and in PGandE's previous j
response to Item 15, the seismic inertia loads for the auxiliary building i
were obtained using a time-history analysis of a spring and lumped mass model.
(Figure 13) shows the mathematical models used in the previous
- analysis, i
1 The added weight of the racks and fuel assemblies represents approxi-i mately 1% of the global mass of the auxiliary building.
In order to evaluate the effect of this added weight, the original mathematical model was revised to include the additional mass consistent with its geometric location in the building.
[ Figure 14) illustrates the revised seismic model.
The effects of the Hosgri earthquake was evaluated using the Newmark time-histories which govern the building response.
The results of the evaluation are as follows:
The dynamic characteristics of the auxiliary building, in l
o general, remain unchanged as shown in (Tables 10 and 11).
(Table 12) compares the significant shear forces and overturning o
moments of the revised analysis *lth the previous analysis.
The revised analysis results show an increase of approximately 1%
over the original forces which is-consistent with the increase in l
the building mass.
As indicated in PGandE's previous response to Item 15b, these minor increases are accommodated by the design margins.
- TER-C5506-625'
,i
- I 3
l LEGEND:
2 - NOOE NUMBER l
@ - ELEMENT NUMBER l - ROTATIONAL DEGREE OF FREEDOM f
l
/ -TRANSNATIONAL DEGREE OF FREEDOM 0 - MASS POINT
@E W
_ 2.4 _ss s-
' s t.s' l
n
\\
6 E L. iss-6 4 >6 o
50j ) H 9
E L.1s3 9
I m l I-
[
I E L 140' 2N1 2.
1 RIGID LINK TY.-
g IG[D y
I I
I
'i1*3,,8% TOR $10N AL 88f SOIL SPRING 3-J 27 5'sjg 4
i w
yp ANg -
I 4
I A gg, L ATION AL p
-r
-+
T 17,7 *.
T SOIL SPRING
{ggg g
EL.95.Q* NEE 6
i 6
C NOTE NOTE s2.3'.
N S MODEL E W MODEL VERTICAL MODEL
]
1 l
t.
I Figure 13.
Auxiliary Building Mathematical Models for Transnational, Torsional, and Vertical Analysis (Figure 11 of Ref. 13) '
l l
~
TER-C5506-(25
<4 LEGEND:
2 - NODE NUMBER
@ - ELEMENT NUMBER
- f. - ACTATIONAL DEGREE OF FMEEDOM
/ -TRANSNATIONAL DEGREE OF FREEDOM Q - MASS POINT.
8 W--E I d' H B-W1.5' o
0 6
6 0
4 >6 il L.1st p f,O') )
g 9
,g iL 18Y I
I m l I I
t I
3 E L 140' In Iai
~
f y RfCfD RIGID LINK TV.
f g
o LM8KTYP.
.w,ygggion Ag P
b>
Il 91I' Soll SPRIN G d
~11J')
j
~*
g t, 3 bf $ 7,I I_
4;j 8-44 Sea Smwo[5 MO I'.
av'A L L
l 7 T r
iu-1 E L 35 Q Mit i
5
- asOTE 8,"
l K?
)
8 2 7 __.
13 8.28',13 9.30'_ l l
N S MODEL EW MODEL VERTICAL MODEL i
1 l
i i
l 2
Figure 14.
Auxiliary Building Revised Mathematical Models To Include High Density Racks j
(Figure 12 of Ref. 13) i s
0 0 0 0 0 0 0 0 0 0 n
3 3 4 2 3 7 4 5 2 3 o
)
1 2 7 6 3 4 1
4 3 i
b 0 7 2 8 5 2 5 1
1 1
t r(
1
- 6 7 1
3 1
2 ao R
pt e
t h
i t
ca iF) 0 9 4 0 9 8 3 4 1
5 t
t a
c r
(
a T
0 8 1
5 9 3 7 8 4 3 e
P 4 5 1
1 1
1 rr a
l s t
el o
di n
n oy o
Ml s
a e
i )
h n y
0 8 9 3 8 2 4 o
c 6 6 1
tl t A_
d ae u
n) 5 0 3 4 0 8 9 1
0 4 od ez rd S e uH 1
3 4 6 8 2 9 1
6 0 r
1 1
2 3 3 4 6 7 o
b o
- l q(
t i M hl e
c t i r
V a
re F
f h
oR f t N
s ou i
sS h
o t
e -
1 7 0 1
5 0 5 4 5 4 i h N
d) 4 2 7 6 3 3 2 2 1
1 er ct O
oc 6 3 0 0 0 0 0 0 0 0 o
nr I
ie S
rs 0 0 0 0 0 0 0 0 0 0 f
e o )3 R
e(
e uN O
P r
q T
e 1
h e,
t 5
ri f
f t
F r i
t ge P
u
,sR p
L so A
n f
i df N
l o( o Q
l U
a ir s A
0 0 0 0 0 0 0 0 0 0 no er a L
P o2 S
n 2 3 1
6 4 5 3 0 8 7 t
N o
)
1 2 9 3 5 5 8 6 2 0 s
b 0 7 2 8 5 2 5 8 1
2 r
t t
i f c e A
t r(
1
- 6 7 1
3 2
o l
l ao R
t oa I_
F b pt i t o
n a
n ca onT D_
iF) s 0 9 2 7 7 8 3 6 8 9 so(
[
t a
i L_
4 5 1
1 1
1 e
r
(
rt U
P.
T 0 8 1
4 9 3 5 9 2 3 ii P
a aa O
re C
pp hT mi l s
(
oc ei r
d s Ci oy o
t t
Ml r
t y
c r
a o
hn c
a f
t tA n
u e) 6 6 1
1 9 9 5 3 8 1
c 0P oL uz 5 0 3 4 0 8 0 2 0 1
o f
n a
1 d
St qH
- n e(
1 3 4 6 8 2 0 1
6 1
i en hi r
1 1
2 3 4 4 6 7 t
n a
o l a t g F
p i b
ri i
t or c
a a
NO i
p T
t i
r c
a i p
tr
'd )
oc 1
7 0 1
6 0 5 4 5 4 l
a i e 4 2 7 6 3 3 2 2 1
1 a
p n
6 3 0 0 0 0 0 0 0 0 rs o
l e(
i a)
P 0 0 0 0 0 0 0 0 0 0 t
n a
os l
it s
tl n
au a
t s r
oe T
R r i
I I e.
C C t
d o I
I 0
t
)
)
o a
b oN I
2 3 4 5 6 7 8 9 1
N
(
(
M h'
g 0m,o em w 0 0 0 0 0 0 0 0 0 0 2 3 2 3 2 2 1
3 4 1
n
_)b 1
0 0 3 1
9 7 9 9 o
0 3 6 1
1 6 3 4 2 i
r(
1 6 7 1
2 2 1
1 tao R
s pt e
i c h
t c#
i[)
t a
6 0 4 4 1 1 2 1 3 0 tc r
(
a T
8 0 0 3 1
3 8 5 1
9 e
f P
- 6 3 2 1
1 1
fa l s t
ei o
n ds n
oy o
Ml s
i a
e t) tn yc 5 6 0 4 0 6 8 5 9 2 o
al s A_
d e
n) 4 5 6 1
8 0 5 4 1
7 re Nd ez bd
- e uH 1
1 3 6 5 3 5 2 8 9 r
ts q(
1 1
2 3 3 4 5 6 o
i o t
si e
c VM av r
a Ee F
f
~
ft R
os s
i e
h sW
~
t e -
it 8
1 4 2 9 0 8 4 7 4 e
cs N
d) 8 4 7 6 3 3 2 2 1
1 r
na)
O ot 6 6 0 0 0 0 0 0 0 0 o
I l e eE 3
S rs 0 0 0 0 0 0 0 0 0 0 f
)
e l
u R
a(
q
,1 O
P
.r a
e n
ei T_
h o
rr S
t t
f s
Fg U
r e
L s
t o
,oR P
u t
sl L
p f
n r
I d(
A i
e o
o N
h Q
i s l
t I
a o
rteob M
0 0 0 0 0 0 0 0 0 0 n
o e
Pt 2
[$
n 2 2 1
4 0 4 1
1 7 6 t
h c
N o
)
1 7 7 2 7 5 4 0 9 s
t e
A i
b 0 2 5 1
0 6 3 5 2 r
f a l
R t r(
1 6 7 1
2 2 1 1
o d
oF ao R
t n
b J_
a pt nn a o
ooT
=_
t c ta n
la si(
Q iF) s n
[
t a
6 0 2 2 8 1 9 8 3 0 i
o it l
r
(
ra P
a T
8 0 0 3 0 3 7 4 1
9 i
U P
6 3 2 1
1 1
e t
ap O
r a
e l
pi C
h s
mc s
T n
oi
(
a li Ct es r
r r
dy o
t a
ol t
.P Ma y
c r
e n
c a
o n
f t
o 1
t A_
n s
e) 5 6 1
5 2 7 4 5 8 2 c
(
1 d el uz 4 5 6 1
8 0 7 6 8 5 n
a n
o f
s Wa qH ea n
e(
1 1 3 6 5 3 5 2 8 0 i
e l
t1 r
1 1
2 3 3 4 5 7 t
n d
b s9 F
a o
o a
p i
m a
Ed i
t T
O c
a e
i p
r t
t u
r t
t a i c
p t
u r
r d) ot 8 1 4 2 9 0 8 3 7 4 l
a t
i e 8 4 7 6 3 3 2 2 1
1 a
p s
n r
6 6 0 0 0 0 0 0 0 0 rs o
l e
e(
i a) p P
0 0 0 0 0 0 0 0 0 0 t
n. u a
os s
l it s
tl l
n au e
a t s e
r oe t
T Rr S
n I
)
e.
C c do I
(
0 t
)
)
)
oN I
2 3 4 5 6 7 8 9 1
o a
b c
M N
(
(
f h'
TER-C5506-625 t
Table 12.
Comparison of Euilding Response 1
(Table 3 of Ref. 13)
I North-South Heritentat fneut Orieinal Amalvsit Revised Analvsis Overturning Overturning Element Shear Homent Shear Homent 3
6 3
6 (Kips x 10 )
(Kip-ft x 10 )
(Kips x 10 )
(Kip-ft'x 10 )
1 5
3.177 0.1525 3.178 0.1525 1
13.710 0.2949 13.750 0.2957 2
71.930 2.0476 72.220 2.0547 3
115.200 3.7320 115.700 3.7481 4
90.160 5.1101 91.090 5.1391 I
l East-West Heritental f neut Qr,1einal Analysis Revised Analysis j
Overturning Overturning Element Shear Homent Shear Homent 3
6 3
6 (Kips x 10 )
(Kip-ft : 10 )
(Kips x 10 )
(Kip-f t x 10 )
5 2.854 0.1370 2.855 0.1370 1
13.550 0.2913 13.580 0.2919 2
76.150 2.1104 76.250 2.1130 3
122.500 3.9026 122.700 3.9059 4
80.210 5.1456 80.950 5.1617 1 l L
I TER-C5506-625
[ Figures 15 and 16) show the comparison of typical response o
spectra at the location of the spent fuel pools (elevation 100 ft).
De original RRS was conservatively developed by adding the transnational and torsional spectra, neglecting the phase relationship of the two spectra at the location of the pools.
These spectra used the Blume and,Newmark enveloped time-histories.
The revised response spectra were developed using tww appforches:
(1) using the same methodology as described above - this resulted in an identical PRS: and (2) using the time-histories at mass point 7 [ Figure 15) of the revised model - this result is enveloped by the original RRS.
Based on the above, it is concluded that the effect of the added weight of racks and fuel assemblies is insignificant on the building response."
3.4.9 Analvsis of the Spent Puel Pool Liner l
Thermal analysis of the liner was performed by the Licensee in accordance with the analysis model shown in Figure 17 [7).
Liner strains and their com-parison to allowable values as performed by the Licensee are shown in Table 13.
{
i j
The effects of sliding forces on the liner were ir.vestigated by the Licensee as follows (7):
"The maximum horizontal sliding force on a support leg of rack modules is estimated to be 151 kips.
This sliding force is resisted by frictional r'esistance between the steel liner and concrete floor slab, and by bearing of the stiffener angles embedded in concrete.
Two loading cases are considered:
l l
1.
Sliding force resulting from four legs of interior rack rnodules l
l clustered together
]
2.
Sliding force resulting from one leg of a module.
The liner plate weld seams are checked for the maximum axial force developed as a result of diference between the rack sliding forces and the resist 2ng forces. The maximum force computed in the liner veld seam l
is 1.5 k/in, compared with allowable force of 6 k/in."
I When the possibility of an air gap forming between the liner and the concrete pool floor was questioned, the Licensee investigated its effect on i
the liner with rack sliding forces. The Licensee's response follows (13):
"The ascumption of an airspace underneath the liner was evaluated and found to have no impact on PGandE's previous response to Item 15f (of Ref. 7).
The floor liner plate is 1/4 inch thick and is welded to l i
TER-C5506-625 n
REQUIRED AND REVISED D
fRE3PONSE 3PECTRA y
l
.BY 3PECTRAL SUMMFloh 2.5 i
t a
9
[(
w 2.0 M
f
)fy
-\\
l d
[
e=.. -
fRESPONSE.$*ECTRA
=
o l '..... l y
l FROM REvtSto TME s
's.
Hi3 TORY ANALYJIS s.
n I
A T NODE 7
-J
/.5 g
y
~
1 2
l a.
i e
i 1.0
~~
/
~
s,
/
0.$
I
,i 00
.s 01 o.t
'0
' ^0 PEntoo (stc.)
i i
Figure 15.
N-S Horizontal Spectra l
(Figure 13 of Ref. 13) 1,
TER-C5506-625 8
30 l
}
[ l lilli I
REQUIRED AND R(V/j[O RLSPONSE JPicTRA BY 3PECTRAL $UMMATION 42.5
-A 2
k 9
i N
l \\
f 2.0
\\
7 y,s r,~i
~
N
\\!,,
-RESPONSE ' SPECTRA f lI v
p.)
q FROM REVISED TIME l
yl*0
~\\
I HIS 70R Y ANAL YS/3 I
AT NODE 7
?
g y
J f
\\
(
I. O J
^
l
/
\\
\\
\\
0,5 sN 0.0
=
ol 0.1 I.0 10 0 p,p,o
(,Ec }
I Figure 16.
E-W Horizontal Spectra (Figure 14 of Ref. 13) !
l
.I
1 1
'IER-C 5506-625
(
)
' 8"I84D AL O Jt/Afl#te
-n 45
' (
(
~
o
(
r-(
~ >
i 2.
c n-.
.Awet A
. A cc. _,
I_
"s* 8 a,6 n
\\
e
\\
/
[
f
~'
Ec8
-Mb I
l w
ANCHOR I-980 TAT 10sv
\\
$*AMCwCR DISPLACMT t
a -4=cuca snrracss c
z.,.snrracss or uwen cAuct 4 u,.snrracss or uwen et=ct a l
m t-l l
Figure 17.
Spent Fuel Pool Liner Model for Thermal Analysis (Figure 10 of Ref. 7) 1 I 1
L
TER-C5506-625
~
Table 13.
Auxiliary Building Spent Fuel Pool Line-Strains and Displacements j
(Table 5. of Ref. 7).
m LINER PLATE STRAINS l
Liner Governing Lecation Lead Ce*b Demaed Allowable (1)
Eafety Faeter(1) l Hall 0+Ta+HE 0.00112 0.005 4.5 Floor 0+Ta+HE 0.00156
.0.005 3.2 LINER ANCHOR DISPLACEMENTS Liner Governing Locatien Lead Cems Demaad A11owablg(3)-
Safety Facter(2)
Hall 0+Ta+HE 0.01220 0.06200 S.2 Floor D+Ta+HE 0.01375 0.06200 4.5 (1) from ASME BLPV Code Section !!I, 01 vision 2,1983 Edition, Table CC-3720-1.
(2) Safety facter is computed relative to the code' allowable.
(3) From ASME B&PV Code Section Ill. Division 2.1983 Edition. Table CC 3730-1, using 5, = 0.125 in, based on test resvits given in BC TOP 1 Appendix B (Ref.19).
1.
TER-C5506-625 embedded angles oriented in east-west and north-south directions as shown in
]
(Figure 18). The fuel pool was constructed using standard construction I
practices for mass concrete construction which included:
o control of pour rates to minimize the effect of plastic shrinkage o
controlled curing to minimize the effect of drying shrinkage o
ample reinforcement to minimize shrinkage o
a surface finish having a uniform texture that does not contain voids and is free of projecting aggregrate, high spots, and depressions.
Based on the above construction features, shrinkage of the concrete slab is expected to be minimal. However, it is conservatively assumed that the average shrinkage is 700 x 10-6 in/in for the entire depth (5 feet thick) of the slab.
This results in a postulated air gap of 0.042 inches.
The simplified model shown in (Figure 19] was used to determine the deformation of the liner under a hydrostatic head of 38 feet of water.
The load that would cause contact of the liner with the concrete was determined to be 0.6 psi which is small compared with the hydrostatic load of 16.5 psi.
The results, therefore, indicated that the liner would be in contact with the concrete under a hydrostatic head alone.
To check the liner contact under thermal effects, the same mathematical model as described in PGandE's previous response to item 15b was used.
The analysis indicated that in addition to hydrosti.
load, a normal load of less than 0.5 kip per rack leg is needed to achieve full liner contact with the coacrete under the rack leg.
This additional normal load is insignificant compared with the leg reaction load of 189 kips.
Therefore, under thermal conditions, the liner will remain in contact with the concrete under the rack leg.
l The liner strains were also determined and found to be within the l
allowable value.
Under hydrostatic loads, the liner strain resulting from the liner contacting the conc.ete due to a postulated air space is 0.00006.
The combined strain resulting from the governing load combination, D + Ta + HE, is 0.00156 compared with an allowable value of 0.005 (previously reported in PGandE's response to Item 15 (of Ref.
7)).
This provides a factor of safety of 3.2.
1 Based on the above, it is concluded that the liner will remain in contact with the concrete to develop adequate frictional resistance."
Sraluation of the Licensee's liner analysis indicated that the analysis is acceptable. _ - _ _ _ _ - _ _ _ _ - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
4 TER-C5506-625 N
2C*. c*
BLch U y* c _,
1 il
("
zuras l
di
- __. eq,.._I ip
~4 m
I o
y
- c. :r --,
t,,,,6 o
- =
a c
l _ _ _ _ ee_ _ _li I
I l_ _ _u ez.E _ ;_..p_ ggt.. r I
I
- t. _ _ _ e _ _ _, _
.re i _ _ _ _ e_ _ _i i
i l_ _ _ _se _,,_ _ gy_ _, _.,____w__.
u _ _ _.n _ V_ _ _x _, _ 8 '____eo_X
",n g'
i' _
.x _3 m.
i____e___
m
_ _1 - c _.
i w
(
i_ _ _ _se._J_n,___.@_^__'t. _ _e_ _!
rn
_ _ _os _
g._._
l _ _ _ eo_ _ _ g
._. _ _ _m _ _
_ _ no_. _ _\\
m~t 2 _ _.pc. y%
_J__>l_
<A l },
_ec
_ _ _g,7 _, _ _ _ _e.o _ _.b t _ _& _ _ _,L _ _ _ co -_ _y_
)
g i
j r/r N l_ _ g_, _. l____a,___ l
~
_ 80-
!_ _ _ _De. _ j e
.__f o
PL A N ( kl. 99'-o')
~
\\
L r2'.e* mrL SPENT FUEL POOL. FLOOR - UNIT f L
f F 4%'d Wuo
$g*i si (4"k iss) l **'m 9 s'-c'
~
'*8 * **
'* ' Q'T
' 6. *.
- Q 8. '.,'. d,,M e,*. *.*f ',* ** *f 7 kJ l
n ss a I
n McN/ Ton /NG
} Q xau%'tss)
)
V
'il rnsucy 1 wiog Li
~'
m a
(E M 304 sh
.,.. d,._
+ a cu'sruos
'a
( * ~* ",.- (
casou srt.
1.....
'.. -)
e t'.o o.c. crrn)
SECTIOkt A-A Figure 18.
Pool Floor Liner Details (Figure 15 of Ref. 13) -
TER-C5506-625
$" LINER HYDROSTATIC Low
- 1I I
I I
I sI I
ll**.*
Ig e
,s.
unau (7YP.)
P Figure 19.
Postulated Floor Liner Deflection (Figure 16 of Ref. 13) -
(-
TER-C5506-625 3.4.10 Spent Fuel Pool Leak Chase System In response to discussions concerning the leak chase' system of the spent I
fuel pool, the Licensee provided the following response (13):
"The spent fuel storage pool for each unit is a reinforced concrete structure with a seam-welded stainless steel plate liner. Monitoring trenches approximately 1 inch square are located behind the liner on each side of all seam welds connecting adjacent liner plates in both the walls and floor of the pool.
These trenches lead to one of six collection pipes which are valved and terminated at a common sump.
Leakage past the liner would be detected by opening these leak detection shutoff valves j
and observing any water accumulation or flow.
These valves are normally closed.
Prior to placing irradiated fuel in the spent fuel pool, Surveillance Test Procedure (STP) I-1C, " Routine Weekly Checks" will be revised.
STP 4
I-1C will require the operators to open each leak detection shutoff valve j
on a veekly basis and determine if any water has accumulated.
Plant Technical Specification 3/4.9.11 requires that at least 23 feet of water be maintained over the top of irradiated fuel. assemblies seated in the stor' age racks.
One assurance of this is provided by a low water level alarm which annunciated in the conctrol room when the pool level goes below 137'-8".
If and when a low level alarm occurs, operators
)
would refer to Volume 16, "Annuciator Response" of the plant manual.
i Volume 16 would direct the operators to open each leak detection shutoff valve and determine if any water had leaked through the fuel pool liner.
If leakage past the liner is detected, additional testing would be j
performed and the leak would be repaired as appropriate.
There are no safsty consequences to this since leakage past the liner woM d be j
terminated by closing the leak detection shutoff valver., and adequate makeup capability exists which exceeds any credible leakage through the liner (discussed in PGandE letter DCL-86-020 dated January 28, 1986)."
No detrimental effect of the higher density spent fuel loading was noted for the leak chase system.
3.4.11 Cask Pit and Cask Restraint The spent fuel pools of Units 1 and 2 each include a cask pit area in a corner of the pool; the cask pit is recessed below the floor of the spent fuel pool. Because spgnt N el rack modules E, F, and J are adjacent to the cask pit, the Licensee is providing a cask restraint, in accordance with the FSAR (6), that will prevent the cask from impacting with the adjacent spent fuel racks during cask operations (20).
~
f TER-C5506-625 During seismic events, a cask in the cask pit will be restrained from lateral movement toward the fuel racks by the walls of the recess'til cask pit.
Via Reference 21, the Licensee provided a sketch of the cask restraint I
(Figure 20) plus an installation plan (Figure 21) showing the rel'ationship of the cask pit, the cask restraint, and the adjacent fuel rack modules.
Reference 21 confirms that the Licensee is mounting the cask restraint so that the centerline of the horizontal member of the restraint is at the elevation
\\
of the impact girdle bars on the rack modules.
Thus, should the fuel rack i
modules impact the cask restraint during a sosimic event,- the impacts will l
take place on the girdle bars of the rack modules that are provided for that
{
j purpose, as discussed earlier in this report.
3.5 FUEL HANDLING ACCIDENT ANALYSIS i
l In the Licensee's approach to fuel handling accident analysis, the l
following assumptions were used (1):
"1. The virtual mass of the body is conservatively assumed to be equal to j
its displaced fluid mass.
Evidence in the literature (Ref. 16),
indicates that the virtual mass can be many times higher.
4 2.
The minimum frontal area is used for evaluating the drag coefficient.
]
3.
The drag coefficients utilized in the analysis are the lower bound j
values reported in the literature (Ref. 17).
In particular, at the i
beginning of the fall when the velocity of the body is small, the corresponding Reynolds number is low, resulting in a large drag coefficient.
4.
The falling bodies are assumed to be rigid for the purposes of impact
. 1 stress calculation on the rack.
The solution of the immersed body motion problem is found analytically.
The impact velocity thus computed is used to determine the maximum stress generated due to stress wave propagation."
Using these assumptions, the Licensee provided the following with respect to handling accident considerations [1):
f
" Dropped Fuel Accident I j
A fuel assembly (weight = 1616 pounds with control rod assembly) is dropped from 36 inches above the module and impacts the base.
The final velocity of the dropped fuel assembly (just prior to impact) is calculated and, thus, the total energy at impact is known.
To study l m___-
d O mFGv5"
.-o y 1,
yL yE g
{
y-g}\\.~
1
-Q, '
l.Pr*
ng -
5 s
t )P t
f n
Ty
\\
t o
r N
TY u.-
F s (T
(
r
- pE a
gE W o.
i L
P r
M D1 K
E1 P
0 s.
Tr n
s
. ~
A 3 $,
Cp-n s.
- 4 n
ES W0 3
i p-a r-a U
T T
[r O
N F
N B
S f-tQ 5
I T
)
u M
),
1 a
],
e' x
K t
S 5
AC 9
33) r e
t L o.
D 3W 9
E, s
i t
f9 t
0 T M
O 3(
g ',
R g.
.t 3
e Tt 9s Wf m
T
$T1 y
N mt tt e
9 s
r r
E aa M
u f
P sme 3
Mf f
- a...
A t
Et
/
Wt a
2 N
Nt 3y Fe A
n
=
f tn l-tua
- R tt o.
N'2
.w *.%
M t
u s
6 m'
JY 8
I TER-C5506-625 ENCLOSURI ML
~
}
A}ow..
w omauscus sm 9)
ARE NC>W/NAL,..
e 4.*. o 2
aa rune,,
e
~
n EL.99'CY.
y 4
Y
\\
b Q
9
\\*h u
I A
,$+
- W o
Q y
u MacrsuPPorY
,, " )
I,)
% )*
L
/% ::
~'~ u
/
- %"a'~
/
c.
b T
f
~fy hyY o
y C",S)
Of CASM, ply-6&,,
- j 9
(
m h
- g 4
g 6
x k
E hb W
s
' ge
- Q s
N N
o 3
(.
to '- o *
,rCA5x K.E5YAr.4NY'-
L eo L s " e r r.
f, qq,'.g *
(m i of MwV
$lifg,"2CP~
i y'
ABL.C CR5K.
a
?
Ai?ES'YRA/NT)
".9 /p
- #%g
$p
(
)
UAIJY f SPEATr'FUF*/- POOL.
(s-W CORNEA OF PCO&
\\
SEf!ACL'A./G &COMT'SEP SGURE 2./a,.*
[yo 5
)
i Figure 21.
Cask Restraint Installation Plan (Sketch 1 of Ref. 21) - _ - _ _ _.
i TgR-C5506-625 i
baseplate-integrity, we assume that this energy is all directed toward U
punching of the baseplate in shear and t'hus transformed into work done by_ the supporting shear stresses.
It is determined that shearing deformation of the baseplate is-less than the thickness of the baseplate so that we conclude that local piercing of the baseplate will not occur.
Direct impact with the pool liner does not occur. The suberiticality of the adjacent fuel assemblies is not violated.-
Dropped Fuel Accident II' One fuel assembly drops from 36 inches above the rack and hits the top of-the rack.
Permanent deformation o'f the rack is found to be limited to
'I the top region such that the rack cross-sectional geometry lat the level of the top of the active fuel (and below) is not_ altered..The region of local permanent deformation does not extend below 6 inches from the racktop.
An energy balance approach is used here to obtain the results.
?
Jammed Fuel-Handlino Ecuipment and Horizontal Force A'4400-pound uplift force and an 1100-pound horizontal force are applied at the top of the rack at the ' weakest' storage location; the force is assumed to be applied on one wall of the storage cell boundary as an upward shear force.
The plastic deformation is found to be limited to the region well above the top of the active fuel."
For Drop Fuel Accident I, a possible mode of rack module deformation was not addressed by the Licensee.
In this mode, the kinetic energy of the falling fuel assembly is dissipated in the strain energy associated with the shearing of welds between the base plate and the storage cell walls and, as'a minimum, would shear the welds of all four walls of the immediate cell.
In the extreme, the worst that could happen would be the deflection of the base plate to the floor.
Since the fuel assembly does not ixtend through the base plate, and since there are no protrusions on the bottom of the base plate, there would be no damage to the pool liner should the bas 7 plate touch the liner.
Review of the fuel handling accident analyses indicates that they are acceptable.
a i
_-_m___.___.____________-______J
TER-C5506-625 s
4.
CONCLUSIONS
~
Based upon the review and evaluation, the following conclusions were
~
reached:
~
The Licensee used three-dimensional, nonlinear dynamic displacement o
analyses with three simultaneous, independent, orthogonal, earthquake j
acceleration time histories to provide greater resolution of the rack module displacements and impacts than is posrible with two-dimen-sional analyses combined by the square root of the sum of the squares method.
Seismic loadings included the simultaneous application of one o
vertical and two horizontal acceleration' time histories derived from the design, double design, and Hosgri earthquake excitations specified by the updated Final Safety Analysis Report (FSAR).
I o
Impact analysis of fuel assemblies within storage cells as well as 3
rack-to-rack and rack-to-wall impacts were integrated into the Licensee's dynamic displacement analysis when design studies indicated that impacts would take place under the specified earthquake excitations.
Fuel assembly impact forces within a rack storage cell were over-o estimated by use of impact springs based upon the storage cell wall compliance, neglecting the greater compliance of the fuel assembly.
However, the resulting stresses were within allowable values.
Two rack modules were selected by the Licensee as representative of o
bounding the range of 13 design configurations used in each spent fuel pool of Units 1 and 2.
These involved the 10 x 11 cell rack module as the largest and heaviest loaded module, and the 6 x 11 cell design as the module with the largest ratio of horizontal dimensions and, therefore, most likely to undergo the most displacement in the tipping mode.
While dynamic displacement analysis was not performed for the 6 x 9 cell design, its displacement response would be expected to be somewhat greater than the 6 x 11 cell design, but not enough greater to cause any of the design criteria to be exceeded, I
given the conservative results presented by the Licensee.
An 8 degree-of-freedom mathematical model was used by the Licensee I
o for dynamic displacement analysis after it was validated by compari-son of results with that of a 32 degree-of-freedom model. The 8 degree-of-freedom model was used to accommodate the added nonlinear complexity associated with the increased number of gap elements required for rack-to-rack impact.
The two-dimensional supplementary parametric studies of the o
multi-rack analysis indicated that the fuel-to-rack and rack-to-rack or rack-to-wall impact loads were enveloped by the fuel-to-rack and rack-to-rack values determined from the design basis analysis presented in the reracking report. - _ _ - _ _ _ _ _
I i
~
TER-C5506-625 o
The two-and three-dimensional parametric studies performed by the Licensee, using realistic dynamic analysis input, have demonstrated that the use of conservative elascic springs and fluid coupling input
'in the rera'cking report yielded conservative estimates of the internal rack stresses as well as rach and fuel assembly impact forces.
o Rack module stresses are within the allowable values, Rack module stability in the tipping mode under a Hosgri earthquake o
indicated a wide margin of safety in tipping.
o A cask iastraint structure was provided to protect the fuel rack modules from impact by the spent fuel cask during cask operations.
Lateral movement of a cask in the cask pit toward the fuel rack modules during a seismic event is restrained by the walls of the recessed cask pit.
j In summary, it was concluded that the Licensee's structural analyses of the spent fuel rack modules and the spent fuel pool are adequate and are acceptable:
they indicate the rack modules and pool structure to be satisfactory for high density fuel storage.
)
I i
1 l
j l
TER-C5506-625 5.
REFERENCES j
1.
Pac'ific Gas and Electric Company, Licensing Report on High-Density Spent Fuel Racks, "Reracking of Spent Fuel Fools, Diablo Canyon Units 1 and 2,"
September 1985
,j l
2.
Seismic Analysis Report, " Seismic Analysis of High Density Fuel Racks for j
Pacific Gas and Electric for Diablo Canyon Nuclear Power Station," Rev.
3, September 3, 1986, A. Soler, TM No. 779 3.
" Additional 7.nformation on Rack-to-Rack Interactions," Enclosure 1 to PG&E Letter No. DCL-87-070, April 7, 1987 4.
"Three-Dimensional Studies, Acorn 10 and Acorn 12," Enclosure to PG&E Letter No. DCL-87-082, April 23, 1987 5.
U.S. Nuclear Regulatory Commission OT Position for Review and Acceptance of Spent Fuel Storage and Handling Applications January 18, 1979 6.
Pacific Gas and Electric Company, Updated Final Safety Analysis Report (FSAR) for Diablo Canyon Units 1 and 2 7.
Pacific Gas and Electric Company, Letter No. DCL-86-019 to U.S. Nuclear Regulatory Commission, January 28, 1986 8.
E. Rabinowicz
" Friction Coefficient Value for a High-Density Fuel Storage System" Report to General Electric Nuclear Energy Programs Division November 23, 1977 9.
R. J. Fritz "The Effect of Liquids on the Dynamic Motions of Immersed Solids,"
Journal of Engineering for Industry, pp. 167-173, February 1972 10.
" Study of Non-Linear Coupling Effects," Holtec International Report No.
HI-87102, January 1987 i
11.
Franklin Research Center, Evaluation of Spent Fuel Racks Structural Analysis, Oyster Creek Nuclear Generating Station, Technical Evaluation Report TER-C5506-525, July 5, 1984 12.
Franklin Research Center, Evaluation of Spent 7uel Racks Structural Analysis, Virgil C. Summer Nuclear Station, Technical Evaluation Report TER-C5506-526, August 15, 1984 13.
Pacific Gas and Electric Company, Letter No. DCL-86-067 to U.S. Nuclear Regulatory Commission, March 11, 1986 TER-C5506-625 14.
U.S. Nuclear Regulatory Commission, Standard Review Plan, Section 3.8.4, Appendix D, NUREG-0800, Revised July 1981
- s 15.
U.S. Department of Commerce (NTIS), " Nuclear Reactors and Earthquakes,"'
TID 2024 16.
A. S. Veletsos and J. Y. Yang, " Dynamics of Fixed-Base Liquid-Storage Tanks," Proceedings of U.S.-Japan Seminar on Earthquake Engineering Research with Emphasis on Lifeline _ Systems, November 1976 17.
Letter (Docket Nos. 50-275 and 50-323) from A. Giambusso of the U.S.
Atomic Energy Commission to F. T. Searls of the. Pacific Gas and Electric Company, dated August 23,-1973, including enclosures 18.
" Recommended Evaluation Criteria for Diablo Canyon Nuclear Power Plant Auxiliary Building Walls and Diaphragms," Jack R. Benjamin & Associates, Inc., February 11, 1983 19.
" Containment Building Liner Plate Design Report, BC-TOP-1," Bechtel Power Corporation, San Francisco, California 20.
Pacific Gas and Electric Company, Letter DCL-86-109 to U.S. Nuclear Regulatory Commission, April 24, 1986
-l 21.
Pacific Gas and Electric Company, Letter DCL-86-108 to U.S. Nuclear Regulatory Commission, April 24, 1986
'I J
4 APPENDIX A t
LICENSEE RESPONSE, PER REFERENCE 13, FOR BACKGROUND
]
AND VERIFICATION OF DISPLACEMENT ANALYSIS METHOD DYNAHIS FRANKLIN RESEARCH CENTER DfVISiON OF ARVIN/CALSPAN 20th & RACE STREETS. PHILADELPHIA.PA 19103
)
1 I
l
PGandE Le%%er No.:
DCL-86-067 ENCLOSURE PGandE RESPONSES TO NRC REQUEST FOR ADDITIONAL INFORMATION Letter Item 26 Hith respect to the computer code. DYNAHIS, discuss specific efforts ' performed to verify that the code provides realistic seismic responses of the spent fuel racks subject to a set of complex loading conditions and geometric constraints.
For example, indicate if any experimental data were used as the basis for demonstrating the validity of the code.
Provide a discussion of the assumptions and limitations which are unique to the code.
i Meetino Item 26a Regarding Item 26 on the validation of the computer code DYNAHIS, the spuific benchmark problems should be identified and appropriately referenced.
The results, including response comparisons, of problems run with the DYNAHIS code and the ANSYS code should be provided. Discuss the extent of use of the DYNAHIS code in teracking applications by other utilities.
Reseense 26 aniljA The algorithm of the code DYNAHIS is documented in the book "The Component Element-Methed in Dynamics" by Levy and Hilkinson, McGraw Hill (1976).
The application of the code to a wide variety of linear and nonlinear problems
- illustrated in that text.
The Joseph Oat Corporation (Oat) version of this *"
code contains the added attribute of diagonalizing the mass matrix.
Howevco,
the code has been tested thoroughly at Oat's computer center.
The benchmarking effort consisted of running several nonlinear problems for which classical solutions exist. A dynamic problem involving a nondiagonal mass matrix was also constructed and analytically solved.
DYNAHIS was run to verify its accuracy versus the analytical solution.
Its accuracy has been further confirmed by comparing results of sample problems run on DYNAHIS and on general purpose codes such as ANSYS.
In fact, DYNAHIS has received a level of scrutiny which rivals that given to any general-purpose fin'ite element code used for nonlinear analysis of dynamic systems, including reviews of teracking applications for Fermi II, Quad Cities I and II, Rancho Seco, Grand Gulf Unit 1, Oyster Creek, Pilgrim, and V.C. Summer.
Since no perti~nent data on the dynamics of rack structures exist, benchmarking against experiments is not possible.
Instead, Dat has taken the proven recourse of using a mathematical model that is highly conservative in all respects, e.g., amount of damping and the mass lumping scheme.
A detailed description of the calculation method of DYNAHIS is presented in f
Section 6 of the reracking report.
There are no numerical simulations or Moreover, in contrast to a modeling assurpions unique to DYNAHIS.
general-purpose finite element code DYNAHIS does not internally generate the These equations are written out explicitly and can be equations of motion.
i Such verifications have been carried out internally verified independently.
at Oat, by several A/E firms, and, at the NRC's request, by Franklin Research Center in 1984. 07195/0042K t
1
-}
q As discussed in meet 16gs with the Staff, one DYNAHIS verification was conducted as part of Oat's work for Detroit Edison Company on the Enrico Fermi Unit II reracking application. As part of this DYNAHIS validation, several problems that would test the analytical features of DYNAHIS related to hydrodynamic response of rack modules were run on DYNAHIS by Oat and on ANSYS by Detroit Edison's architect / engineer.
These problems consisted of various combinations of masses, springs, dampers,-and gap elements. Comparisons of the DYNAHIS and ANSYS response predictions confirmed that the results from the two computer codes were in excellent agreement. An. additional, more complex problem, a 32-DOF model of a rack module, was also analyzed on ANSYS by the i
architect / engineer.
These results were also compared with Oat's DYNAHIS response predictions and the comparisons were in good agreement.
These evaluations were not formally documented or submitted on the Fermi II docket.
The following is a description of some specific verification runs that have been made.
These benchmark problems used to verify DYNAHIS solutions fall into three classes:
(1) comparison with classi~ cal closed form linear solutions which exercise the features of.DYNAHIS (e.g..-decoupling of mass matrix and input force history from external files); (2) ronlinear problems which utilize the gap e'lements and friction elements and compare results wit
.I textbook numerical analyses; and (3) problems which compare DYNAHIS results',
corresponding results obtained from general-purpose finite element codes.
Problem 1 2 Degree of Freedom (DOF) Linear System with Mass Coupling 1 + mE + KX1 = X sin nt 2mW 2
mi)+2mi2 + KX2 = 0 The table below shows results obtained for x<(t) and x2(t) from the a = 750 lb-sec'g/in., K = 100 analytically d rived solutign and from DYNAHES fo lb/in., n = 6.28 rad /sec, and K = 386 lb.
SOLUTION COMPARISON x1 x 10+3 x2 x 10+3 Time Analytical (DYNAHIS)
Analytical (DYNAHIS)
(sec)
Solution Numerical Solution Numerical
.1
.204
.208
.063
.064 2.1
.188
.189
.022
.021 3.6
.252
.256
.054
.054 4.1
.330
.332
.040
.041 5.7
.457
.460
.033
.033 8.7
.433
.433
.084
.088 10.8
.423
.423
.109
.109 11.2
.398
.398
.061
.063 13.3
.407
.408
.090
.091 14.4
.230
.229
.081
.081 As can be seen, there is good agreement between the analytical solution and the DYNAHIS results.
07195/0042K
~
Problem 2 Mas 5 Dropped on a-Beam ("The Component Element Method in Dynamies,P S. Levy,
~
McGraw Hill, 1976, pp. 151-155). See Figure 1.for details of the model.
One half of the beam is modeled by a 7 degree of freedom system (including rotary inertia in the beam, and one DOF.for the dropped mass).
The DYNAHIS.
results are plotted and compared with the' text results..'The con. tact force time-history maximum values are predicted with good agreement.
The' additional higher frequency oscillations in.the gap element forces are due to the' fact-that DYNAHIS includes the efft:t of higher. frequency rotary inertia degrees of' 3j freedom.
The plots in Figure 2.show the textbook results, and Figures 3 and-4 show the DYNAHIS result.s.
The agreement is found to be excellent.
I Problem 3 Dropped Mass with Friction Elements (From Levy text, pp. 55-56).
As shown in Figures 5 and 6, the DYNAHIS output plot is in good agreement with the curve (F f 100 lb) from the text.
This problem demonstrates that the DYNAHIS code correctly accounts for friction effects.
Problem 4 Comparison of ANSYS with DYNAHIS.
The comparison of results was carried out using ANSYS example problem E6.2 which is a pendulum system as shown in Figure 7.
The same numerical values for masses, springs, gaps, etc. were used in the DYNAHIS model.
Since the time-history algorithms are different in the two codes, the. excellent i
agreement obtained verified the suitability of the integration algorithm'in DYNAHIS.
Figures 8 and 10 are reproductions from the ANSYS examples' manual-showing the mass displacements and contact forces.
Figure 9 shows the DYNAHIS mass displacements.
The DYNAHIS results for contact' forces at selected time.
l intervals are superimposed on the ANSYS output in Figure 10.
The two. analyses produce results that are in close agreement.
i Problem 2
)
G Model I
gl 1
ebq l
3 m
9 9
O O
I I
{
h
- M q6 $ q4 V 92 h
q7 q5 '
q3 s
L
/.
(
Data for the Problem:
b = 1.32",'
h $$.316" L = 80a g = 386 in/sec
~
y =.3 0/cu.in.
t = 13.333" we obtain the total beam mass as m =.0259 i sec /in p
Figure 1
Problem 2 i
TMt coumNew? ciruchi strTitoo IM p{
I f
I
{
l 3
-(
En i
i P
h O 10 - - -
b
/
o
.-h3 0 10 0 15 t;u Time (sc:)
&d5 0**
ny n'y
._ A e.
f J
1 N
4.5 negli:ttle sii
% I.1 K
o
'1i
~2
\\
F1 CURE 3.26 Centa61 force of b!! va besm and defiertions in generalind coordinates i
l i
Figure 2
'-~~- '
j d
7.:
4 3 '11 i
1 A*se
- r 3
m s
= v O k
- o
- a 5.'cn 4,vs t e
- i. g m
c c E
m -
w k
.O
.m c.
gp) o x
,,S k
en: G f
n.
=.
.v 4,
s-m-c o
v x
m.
e 4o9
- n
.t y.
oll
+c' e
?
Figure 3' j
a
. c c
3
. 3
../
- Ilt t.
1 m.N
.I f
t..
l.
=..
e e
M
)
) n 1 o
- i I.'
.tI
.t.t t.
s.
t Fu Ol Do
( s o
es l 'y N
c O
i v I
t e T
rl 2
U a
L P o m
O t
e S
s se l
b S
at Mi o
i s
r l
P A
f o N
op Y
p D
to nen mo ei ct ac l e pr si i d D(
~
v &C1 b t.a
Problem 3
~
w ame w unuu mAna sisnan M
?
r
/
Atenen fncuan
\\
artide
/
l l23,, I 90 lb/in L
F10UILE 2.13 o
Packsand arode arikes powment after 2 U n./sec 8Mr*8 h!!ans 5 A from tadgate of trwk. Fncoco veJuss o(100 lb aad 400lb
///////////////////
w mi ecwo.mc n.nawr um.co tw ovwoocs
^
10 i
r.100 ib, e,-xi0 ib/in.
g I
o2
/N E O o 5 A H Tinw (soo
~
'j' U
L/
L i
- 10 Figure 5
=, :..
e Problem 3 s
DYNAHIS SOLUTION J
problam wlth feletton from t e n t, '
f
- 10 0 15.00 l
10.00
/
5.ee I
f' t\\
/
/
/ \\ *, /
s
/
\\
/
\\
0.00 Ni
\\
)
L 1
f
-5.00 o
u l
0 l
-10.00 l
f 1
(
1
-15.00 -
.'l 0.00 0.10
- 0. DP) 0.30 0.40 0.50
- 0. 60 ' l 20.00 I
t tm (
.e. )
i Figure 6 l
l I
~~
E6.8 Q
Problem 4
///
/
/
/
/
/ //
/-
/
a
/
4
+ --
/
/
/
h31.k 6~
/
/
rd j
V i
\\
U
\\
/.
j~
hl l, V
aA n
b i
2 3
4 K
3 PH e a t, i
i Figure 1 System of Impac':ing'Fendulums'/
1 r
l Ag A.
As
. da
-WW-- @4 -'/M- @5 x
8g1
-vwv-
@ a
-'AW--d j@ 3 i j g --
ij
-+
jr-=
q
=r-
- 2--
4 4
--c 4
--c:
a.
2*
i a
c c
c e
Q - Element Numbers Figure 2 Finite Element Model of Pendulum system.
Figure 7
1 edra i
E6.19
{
Problem 4 sw _
I
^
i we j
m amm r
tw f
I as h f:
18 Pact g
3
- .1
/
i~
\\
i A,
if
/ N r
F A
4 i
-5
. r :.,.
l a
1 u
wr I?i gi.3 id'
/
b Ok f
/
I l
&ar
- @.,W w
cener M F a vm m a praens Pom u s mars t
Figure 8 Figure 5 Linear (X) Displaecment of Nodes (or Spheres) 1.'4, and 5 versus Time.
s l
l
._______-____N--N.
a Problem 4 DYNAHIS SOLUTION Mass Displacements d isp. of nodes vs time 2.50 I
i 2.0@
$5 l
/
1 l
1.50 l
1.00 -
/
b 0.50 j
1 p
i d
C
='
0.00 m
.N-j
-~
-~,
~
i' l
ae o.e ge
)
1.00
)
l
-1.50
,/
/I.
-2.00
/
/
L 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
-0.50 ttm. (see.)
I 1
l Figure 9
E6.23 e
Problem 4 3.4 II 3
_2 Y
1 J) i iso
-q A
-Y
(
i 1
_e#
.c 4
J
- It I
\\
\\
)
\\
- s-x.
M.s
_-.N-1 L
JA
. to q.g
~
6 I'?i.,
- aJ 1
-2Y F '.
k[i. 3 e..i (P r--
m
<n m
u en
=
-ZF y & prggy W OPAC3R @DRAp5 MEU It e.
byu A H is AE 5 0' T 5 1-)
W v-5
- 5-v o
Figure 8 Contact Forces Between Spheres 1 and 2, 2 and 3, 3 and 4, and 4 anc! 5.
Figure 10
I l
7590 01 1
UNITED STATES WUCLEAR REGULATORY COMMISSION 1
PACIFIC GAS AND ELECTRIC COMPANY
'{
DOCKET NOS. 50-275 AND 50-323 NOTICE OF ISSUANCE OF AMENDMEN1' TO FACILITY OPERATING LICENSES AND FINAL DETERMINATION OF N0 $1GNIFICANT HAZARDS CONSIDERATION The U.S. Nucleer Regulatory Comission (the Comission) has issued Amendment No. 8 to Facility Operating License Nos. DPR-80, and Amendment No. 6 to Facility Operating License No. DPR-82, issued to Pacific Gas and Electric Company (the licensee), which revise the combined Technical Specifications for the Diablo Canyon Nuclear Power Plant, Unit Nos. I and 2 (the facilit'ies) located in San Luis Obispo County, California.
The amendments pemit the expansion of the spent fuel storage capacity for the Diablo Canyon Nucle'er Power Plant, Units 1 and 2, increasing the storage capet.ities from 270 to 1324 spaces for each of the two spent fuel pools. This expansion would be accomplished by reracking the existing spent fuel storage
. pools with new spent fuel racks composed of individual cells made of stainless steel. The new fuel storage racks will be arranged in two discrete regions I
within each pool. Region 1 wi11' provide 290 poisoned cell locations which will nomally be used for new fuel with an enrichment equal to or less than 4.5%
U-235. Region 2 will provide 1034 unpoisoned cell locations which will provide normal storage for spent fuel assemblies with an initial enrichment equal to or less than 4.5% U-235 and meeting required burnup considerations in accordance with the amended Technical Specifications. The existing fuel storage racks 3
l
7 1
l
,o :
have o nominal center-to-center spacing of 21 inch. The new racks will have a nominal 11 inch center-to-center spacing for both regions. The major components of the fuel racks are the fuel assembly cells, gap channels, rack base plate assembly including support legs, a girdle bar around the upper part of the rack assembly, and, in Region 1 only, neutron absorbing (poison) material Bor4 flex, including cover sheets. The fuel racks are designert to maintain the required subtriticality of K,ff equal to or less than 0.95 for Regions 1 and 2.
The letter requesting the license amendments, dated October 30, 1985 (LAR-85-13), includes the requested Technical Specification changes and the I
licensee's detennination on significant hazards consideration. The supporting j
report on "Reracking of Spent Fuel Pools for Diablo Canyon Units 1 and 2" had been submitted to the staff by letter, dated September 19, 1985.
'The application for these amendments complies with the standards and requirements of the Atomic Energy Act of 1954, as amended (the Act), and the Comission's rules and regulations. The Comission has made appropriate findings as required by the Act and the Comission's rules and regulations in 10 CFR Chapter I, which are set forth in these license amendments.
Notice of Consideration of Issuance of Amendments and Proposed No Signifi-cant Hazards Consideration Determination and Opportunity for Hearing in I
connection with this action was initially published in the FEDERAL REGISTER l
on January 13, 1986, (51 FR 1451) and in the bi-weekly publication on May 21, 1986 (51 FR 18699). A request for hearing was filed by (1) Mothers for Peace on February 7,1986,(2) Sierra Club-Santa Lucia Chapter on February 10, 1986, i
and (3) Consumers Organized for Defense of Environmental Safety (C.O.D.E.S.)
{
on February 12, 1986.
l i
I l
l l
U
l.
k;.
tl l
Under its regulations, the Comission may issue and make an amendment imediately effective, notwithstanding the pendency before it of a request for a hearing from persons, in advance of the holding and completion of any required hearing, where it has determined that no significant hazards consideration is involved.
The Comission has applied the standards of 10 CFR 50.92 and has made a final determination that these amendments involve no significant hazards consideration.
The basis for this determination is contained'in the Safety Evaluation related to this action. Accordingly, as described above, these amendments have been issued and made imediately effective and any hearing will be held after l
issuance.
A separate Environmental Assessment has been prepared pursuant to 10 CFR Part 51. The Notice of Issuance of Environmental Assessment and Finding of No Significant Impact was published in the FEDERAL REGISTER (51 FR 19430) on May 29, 1986.
For further details with respect to the action see (1) the application for the amendments dated October 30, 1985 and additional information provided by the licensee in letters dated September 19 December 20, and December 24, 1985 and January 28 (2 letters), March 11 and April 24 (2 letters),1986, (2) Amendment Nos. 8 and 6 to Facility Operating License Nos. DPR-80 and
- l DPR-82, (3) the Comission's related Safety Evaluation and (4) Environmental Assessment and Notice of Issuance of Environmental Assessment and Finding of No Significant Impact. All of these items are available for public inspection at the Comission's Public Document Room,1717 H Street, N.W., Washington, D.C.,
and at California Polytechnic State University Library, Documents and Maps
4-
/
' Department, San Luis Obispo, California 93407. A copy of items (2), (3) and 9
(4) may be obtained upon request addressed to the U.S. Nuclear Regulatory Commission, Washington, D.C.
20555, Attention:
Director, Division of PVR Licensing-A.
Dated at Bethesda, Maryland, this 30th day of May 1986.
FORTHENUCLEA3REGULATORYCOMMISSION l
ra i.
PWR Project Directo a e No. 3 Division of PWR Lice ing-A, NRR I
i
{