ML20198B499

Non-proprietary Replacement Pages for Attachment F to Which Proposed to Change TS 5.5, Storage of Unirradiated & Sf

ML20198B499
Person / Time
Site:	Nine Mile Point
Issue date:	05/15/1998
From:	HOLTEC INTERNATIONAL
To:
Shared Package
ML20138M115	List: ML20198B493 ML20198B499
References
HI-91738, NUDOCS 9812180178
Download: ML20198B499 (31)

Text

,

. l 0

ATTACHMENT 3 l 2

REPLACEMENT PAGES FOR ATTACHMENT F (NON-PROPRIETARY) TO OUR MAY 15,1998 SUBMITTAL l

1 i

I

Table 2.1 DESIGN DATA FOR NEW RACKS I.D. (inside dimension) 5.9" Cell Nominal Pitch 6.06" Boral Loading (min.) 0.015 gm per sq.cm. ( 'B)

Boral Plate (nom.) Width 5" Boral Picture Frame (Bounding) Size 5.125" x 146-1/2" )

i Boral Length 146" '

Boral Thickness 0.070" Cell Height 165" Baseplate Thickness 3/4" Bottom Plenum Height 5" (nom.)

Number of Supports Per Module Four (min.)

Support Type Remotely Adjustable 111-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION 2-8

Table 2.2 MODULE DATA FOR RERACK CAMPAIGN I MODULE NUMBER OF CELLS I.D. QTY. North-South East West Total Per Total No. of Cells for Direction Direction Rack this Rack Type A 1 18 15 254 254 (18x15-4x4)

U 1 18 15 250 250 (18x15-5x4)

C 1 18 13 234 234 D 1 18 15 270 270 E 1 18 20 276 276 (18x14+24)

F 2 18 17 301 301 (18x17-5)

G 1 7 17 118 118 (7x17-1)

M 1 12 16 137 137 (8x16-S+15)

TOTAL: 8 - - - 1,840 HI-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION 2-9

Table 2.3 MODULE DIMENSIONS AND WEIGHTS FOR RERACK CAMPAIGN I Dimension (inches)t Module I.D. Shipping Weight in Pounds North-South East-West A 109.44 91.2 17,900 B 109.44 91.2 17,600 C 109.44 79.04 16,400 j D 109.44 91.2 19,000 E 109.44 121.6 19,500 F 109.44 103.36 21,200 G 42.56 103.36 8,400 M 72.96 97.28 10,000 l J

l

~

' Nominal rectangular planform dimensions, 111-91738 SHADED REGIONS ARE IIOLTEC PROPRIETARY INFOILMATION 2-10

Table 2.4 MODULE DATA FOR RERACK CAMPAIGN II Total No. of MODULE NUMBER OF CELLS Cells For This Rack Type N-S E-W Total Per Rack )

I.D. Qty. Direction Direction Hl-H4 4 18 15 270 1,080 1

l J1-J2 2 18 17 306 612 K 1 18 16 259 259 (18x14+7)

L 1 18 17 295 295 (18x17-1x11)

TOTAL: 8 - - --

2,246 l

l I

I l

l l

t it

111-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION 2-11

l .

l-

! Table 2.5 l

MODULE DIMENSIONS AND WEIGHT FOR RERACK CAMPAIGN II DIMENSIONS (inches)t Module I.D. Shipping Weight in Pounds North-South East-West l H1-H4 109.44 91.2 19,000 J1-J2 109.44 103.36 21,500 K 109.44 97.28 18,200 L 109.44 103.36 20,700 i

l l

' Nominal rectangular planform dimensions.

i 111-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION 2-12 i

1

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

1 1

(

v. . The baseplate provides a conformal contact surface for the " nose" of the fuel i i

.. assembly. j i

)

l vi. The module design affords built-in flexibility in the fabrication process so as to '

maintain the desired cell pitch even if certain fabricated " boxes" are slightly oversize.

The foregoing objectives are fully realized in the module design for the NMP-1 racks 'as described in the following.

3.2 Anatamv of the Rack Module The new high density rack mcdule design employs storage cell locations with a single poison j

panel sandwiched between adjacent austenitic steel surfaces. ,

1 A complete description of the rack geometry is best presented by first introducing its constituent parts. The principal parts are denoted as: (1) the storage box subassembly, (2) the baseplate, (3)

- the neutron absorber material, (4) picture frame sheathing, and (5) support legs.

i l

Ech part is briefly described below with the aid of sketches.

I

1. Storage cell box subassembly: The so-called " boxes" are fabricated from two j precision fonned channels by seam welding them together in a seam welding l machine equipped with copper chill bars, and pneumatic clamps to minimize distortion due to welding heat input. Figure 3.1 shows the storage cell box assembly.

I Hi-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION 3-2 i

b

The minimum weld penetration is 80% of the box metal gage which is 16 gage.

The boxes are manufactured to 5.9" nominal I.D. (inside dimension).

As shown in Figure 3.1, each box has two lateral holes punched near its bottom edge to provide auxiliary flow holes. In the next step, a picture frame sheathing is press formed in a precision die. The sheathing is shown attached to the box in Figure 3.2. The sheathing is made to an offset of 0.077" to ensure an unconstrained installation of Boral* plates. The " picture frame sheathing" is attached to each side of the box with the poison material (Boral ) installed in the sheathing cavity. The top of the sheathing is connected using a smooth continuous fillet weld near the top of the bo::. The edges of the sheathing and the box are welded together to form a smooth lead-in edge. The box with integrally

- connected sheathing is referred to as the " composite box".

The " composite boxes" are arranged in a checkerboard array to form en assemblage of storage cell locations (Figure 3.3). The inter-box welding and pitch adjustment is accomplished by small longitudinal austenitic stainless steel connectors shown as small circles in Figure 3.3.

An assemblage of box assemblies thus prepared is welded edge to edge as shown in Figure 3.3 resulting in a honeycomb structure with axial, flexural and torsional rigidity depending on the extent ofintercell welding provided. Figure 3.3 shows that two edges of each int:rior bor are connected to the contiguous boxes resulting in a well defined peth for " shear !!ow".

i 111-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION 3-3 I

l l

I

2. -Baseplate: The baseplate,3/4 inch thick, provides a continuous horizontal surface for supporting the fuel assemblies. The baseplate has a concentric hole with a 45 taper in each cell location to provide a seating surface conforming to the fuel assembly. Refer to Figure 3.4.

The baseplate is attached to the cell assemblage by fillet welds. The baseplate projects beyond the cellular region of the rack modules by 0.125 inch (nominal).

These baseplate projections serve as the designated impact locations for the racks in the event that the modules undergo kinematic movements during a seismic event.

3. The neutron absorber material: Boral" is used as the neutron absorber material.

Refer to Section 3.3.3 for more details on this material.

4. Picture Frame Sheathing: The sheathing is a part of the " composite box  ;

assembly" described earlier. The sheathing serves as the locater and retainer of the  !

poison material. As such, it is made in repeatable precise dimensions. This is i accomplished by press-forming stainless sheet stock in a specially high tolerance die. The schematic of the sheathing is shown in Figure 3.2.

Figure 3.4 shows three storage cells in elevation with the fuel assembly shown in phantom in one cell. The poison screen extends vertically over a 146" distance, straddling the active fuel length of all fuel assemblies to be used in the NMP-1 reactor.

5. Support Legs: As stated earlier, all support legs are the adjustable type (Figure 3.5). The top (female) portion is made of austenitic steel material. The bottom HI-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION 3-4 4

r w -11 ,s,- c --- <

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

d

,e .

2: part is made of 17:4 Ph series stainless steel to avoid galling.

P t

The support leg is equipped with a socket to enable remote leveling of the rack ,

after its placement in the pool using a specially designed long-handled tool.

3.3 Material Considerations 3.3.1 Ingoduction i

' Safe storage of nuclear fuel in the NMP-l' spent fuel pool requires that the materials utilized in

the fabrication of racks be of proven durability and be compatible with the pool water environment. This section provides the necessary information on this subject.

3.3.2 Structural Ma'ariale The following structural materials are utilized in the fabrication of the spent fuel racks:

a. ASME SA240-304 for all sheet metal stock.

b. Intemally threaded support legs: ASME SA240-304.

c. Extemally threaded support spindle: ASME SA564-630 precipitation hardened stainless steel (heat treated to 1100*F).

d. Weld material - per the following ASME specification: SFA 5.9 R308L.

3.3.3 - Poison Material In addition to the structural and non-structural stainless material, the racks employ Boral*, a patented product of AAR Brooks and Perkins, as the neutron absorber material.

HI-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFOP.MATION 3-5

s 1

Boral is a thermal neutron poison material composed of boron carbide and 1100 alloy aluminum.

. Boron carbide is a compound having a high boron content in a physically stable and chemically inert form. The 1100 alloy aluminum is a light-weight metal with high tensile strength which is protected from corrosion by a highly resistant oxide film. The two materials, boron carbide and aluminum, are chemically compatible and ideally suited for long-term use in the radiation.

thermal and chemical environment of a nuclear reactor or the spent fuel pool.

Boral's use in spent fuel pools as the neutron absorbing material can be attributed to the following reasons:

i. The content and placement of boron citrbide provides a very high removal cross section for thermal neutrons.

ii. Boron carbide, in the form of fine particles, is homogeneously dispersed throughout the central layer of the Boral panels.

iii. The boron carbide and aluminum materials in Boral are totally unaffected by long-term exposure to radiation.

T iv. The neutron absorbing central layer of Boral is clad with permanently bonded surfaces of aluminum.

v. Boral is stable, strong, durable, and corrosion resistant.

Holtec Intemational's QA program ensures that Boral is manufactured by AAR Brooks & Perkins HI-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION 3-6

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

, :,~ .

l l

l under the control and ' surveillance of a' Quality Assurance / Quality Control Program that-conforms to the requirements of 10CFR50 Appendix B, " Quality Assurance Criteria for Nuclear l Power Plants". I 1

l

- As indicated in Table 3.1, Boral has been licensed by the USNRC for use in numerous BWR and .

PWR spent fuel storage racks and has been extensively used in overseas nuclear installations. l l

Boral Material Characteristics )

l l

. Aluminum: Aluminum is a silvery-white, ductile metallic element that is the most abundant in the earth's crust. The 1100 alloy aluminum is used extensively in heat exchangers, pressure and storage tanks, chemical equipment, reflectors and sheet metal work. It has high resistance to corrosion in industrial and marine atmospheres. The physical, mechanical and chemical l

properties of the 1100 alloy aluminum are listed in Tables 3.2 and 3.3.

J i

The excellent corrosion resistance of the 1100 alloy aluminum is provided by the protective oxide film that develops on its surface from exposure to the atmosphere or water. This film )

prevents the loss of metal from general corrosion or pitting corrosion and the film remains stable between a pH range of 4.5 to 8.5.

Boron Carbide: The boron carbide contained in Boral is a fine granulated powder that confonns to ASTM C-750-80 nuclear grade Type III. The particles range in size between 60 and 200 mesh and the material conforms to the chemical composition and properties listed in Table 3.4.

3.3.4 Compatibility with Coolant All materials used in the construction of the NMP-1 racks have an established history ofin-pool i HI-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION l 3-7 i

l

. l usage. Their physical, chemical and radiological compatibility with the pool environnient has been established throughout the industry. As noted in Table 3.1, Boral has been used in both vented and unvented configurations in fuel pools with equal success. Austenitic stainless steel is the most widely used stainless alloy in nuclear power plants.

3.4 Codec Standards. and Practices for the Spent Fuel Pool Modification The fabrication of the rack modules is performed under a strict quality assurance program which meets 10CFR50 Appendix B requirements.

The following codes, standards and practices were used for all applicable aspects of the design, construction, and assembly of the spent fuel storage racks. Additional specific references related to detailed analyses are given in each section.

a. Design Codec

1. AISC Manual of Steel Construction, 8th Edition,1980 (provides detailed structural criteria for linear type supports).

2. ANSI N210-1976, " Design Objectives for Light Water Reactor Spent Fuel Storage Facilities at Nuclear Power Stations" (contains guidelines for fuel rack design).

3. American Society of Mechanical Engineers (ASME), Boiler and Pressure Vessel Code,Section III, Division 1,1986 Edition.

4. ANSI /AISC-N690-1984 - Nuclear Facilities - Steel Safety Related Structure for Design, Fabrication and Erection.

5. ASNT-TC-1A,1984 American Society for Nondestructive Testing (Recommended Practice for Personnel Qualifications).

6. ACI 349 Code Requirements far Nuclear Safety Related Concrete Structures.

' HI-91738 SHADED REGIONS ARE HOLTEC PROPRIETARYINFORMATION 38

, i

(. i 1

l l

Table 3.4 i BORON CARBIDE CHEMICAL COMPOSITION, l WEIGIIT %

Total boron 70.0 min.

B' isotopic content in natural 18.0 l boron Boric oxide 3.0 max. l Iron 2.0 max.

Total boron plus total carbon 94.0 min.

BORON CARBIDE PHYSICAL PROPERTIES Chemical formula B,C Boron content (weight) 78.28 %

Carbon content (weight) 21.72 %

Crystal Structure rombohedral Density 2.51 gm/cc 0.0907 lb/cu.in.

Melting Point 2450 C-4442'F Boiling Point 3500*C-6332*F Microscopic Capture cross-section 600 bam l

i HI-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION 3-19

I i

l FORMED CELL - y BOX CELL l

, ,\ , . ,

i/

i i i i

\ [

i l

i i i

l '

i i j l.. I I I I

I l I 1 1 I i  ! I i l i i e i , ,

Ll _ _ _ _ _ _ _ _ _ _ _.

, i i i i i i

l i

i l l l 1 l f I l I j h i i i i i, i t

e i  !

i i i

. , l ,

1 ii i 9 I i 6 i i l1 i i, '

l1 l

h 1 i i i u ,

i l i i 1 l h I; n .  :

j' I ,

li l d l il .

,i

., e i '

i i q  ; i i ,

i i  !

I i i

Il FIGURE 33; A CROSS SECTIONAL VIEW OF AN ARRAY OF STORAGE LOCATIONS IPeo r T a 11080 I u

> . - :r ,. $

r.

- L_.

r E

S E S T u c

A L

P  % >

4 E

S A

B

\

(

eSE

" L N

U 4 .

% yx aeH 4 O A

1 -

1

% N )

4E 0

4S

/S 1 A

\ (

4C L

% \

\ T R

O _

P -

%/ P -

U _

%),

S

//pl^

' ~ .

=

/ E

\

\V '/// ,' jl ; ' .  ! 't L

B 6 !i A

T S

U

/.  ! : l i\ J

/ \

/

/ - D

\ A -

Q

. A

' .\ 3. i S oWi? 5 .

L  ! d

\ " 3 L / _

6 E E /

C ', \l' .

R

\ 1

\. !i  ! ; le

/

/' U

- G I

\\

i

- F

,:i  ! ; ' ui'>

NN /

/ / / . / 3x$/ =

LN "s

u e i

" a 4

4 ,

o n

3

<aO =8 -

,3 conditions are summarized in Table 4.2.1. The double contingency principle of ANSI N16.1-1975 l

$' . (and the USNRC letter of April 1978) specifies that it shall require at least two unlikely independent -

and concurrent events to produce a criticality accident. This principle precludes the necessity of

! considering the occurrence of more than a single unlikely i independent accident condition concurrently.

Other hypothetical abnormal occurrences were considered and no credible occurrences or con-figurations have been identified that might have any adverse effect on the stcrage rack criticality l safety.

l 4.3 innut Parameters 4.3.1 Fuel Accambly Denian Snecific.ntions The design basis fuel assembly is a standard 8x8 array of BWR fuel rods containing UO2 clad in Zircaloy (62 fuel rods with 2 water rods). For the nominal design case, fuel of uniform 4.0 wt%

2"U enrichment was assumed, with eight fuel rods containing 2% gadolinia. The GE-11 fuel design, I a 9x9 array of fuel rods with 7 rods replaced by two zircaloy water channels, was also evaluated.

Design parameters for both types of fuel are summarized in Table 4.3.1.

4.3.2 Storace Rack Cell Snecifications The design basis storage rack cell consists of an egg-crate structure, illustrated in Figure 4.3.1, with

' fixed neutron absorber material (Boral) of 0.0162 g/cm' boron-10 areal density (0.015 gms B-10!cm2minimum) positioned between the fuel assembly storage cells in a 0.077 inch channel.

This arrangement provides a nominal center-to-center lattice spacing of 6.060 inches.

._ Manufacturing tolerances, used in evaluating uncertainties in reactivity, are indicated in Figure 4.3.1. The 6.060-in. stainless-steel box which defines the fuel assembly storage cell has a nominal inside dimension of 5.90 in. This allows adequate clearance for inserting / removing the fuel HI-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY 4-3

,2 l

assemblies, with or without the Zircaloy flow channel. Boral panels are not needed or used on the J exterior' walls of modules facing non-fueled regions, i.e., the pool walls.

4.4 Analysis Methodology j i

l In the fuel rack evaluation, criticality analyses of the high density spent fuel storage racks were

]

performed with the CASMO-3 code [4.1], a two-dimensional multi-group transport theory code.  ;

1 Independent verification calculations were made with the KENO-5a computer package [4.2], using the 27-group SCALEt cross-section library [4.3] with the NITAWL subroutine for U-238 resonance shielding effects (Nordheim integral treatment). ,

1 l

I Benchmark calculations are presented in Appendix A and indicate a bias of 0.0000

• 0.0024 for CASMO-3 and 0.0101

• 0.0018 (95%/95%) for NITAWL-KENO 5a. In the geometric model used in the calculations, each fuel rod and its cladding were explicitly described and reflecting boundary conditions (zero neutron current) were used in the axial direction and at the centerline of the Boral and steel plate between storage cells. These boundary conditions have the effect of creating an infinite array of storage cells in all directions.

l The CASMO-3 computer code was used as the primary method of analysis as well as a means of evaluating small reactivity increments associated with manufacturing tolerances. Bumup calculations i

were also performed with CASMO-3, using the restart option to describe spent fuel in the storage cell.

KENO-Sa was used to assess the reactivity consequences of eccentric fuel positioning and abncrmal locations of fuel assemblies.

1 i

SCALE is an acronym for Standardized Computer Analysis for Licensing Evaluation, a standard cross section set developed by the Oak Ridge National Laboratory for the USNRC.  !

111-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY 4-4

-W L

gave a k, of 0.9168 for CASMO-3 and 0.9230

• 0.0024 for KENO-5a.

i 4.5.2 Uncertainties Due to Manufacturing Tolerances l

The reactivity effects associated with manufacturing tolerances are listed in Table 4.5.1 and discussed  !

~

below.' I i

l 4.5.2.1 Boron Lnading Variation l

1 l

The Boral absorber panels used in the storage cells are nominally 0.070-inch thick, with a B-10 areal i 1

2 density of 0.0162 g/cm . The manufacturing tolerance limit is

• 0.0012 g/cm in B-10 content, 2 j

l including both thickness and composition tolerances. This assures that the minimum boron-10 areal

! density will not be less than 0.015 gram /cm'. Differential CASMO-3 calculations indicate that this l

i tolerance limit results in an incremental reactivity uncertainty of

• 0.0056 Sk.

l 4.5.2.2 Boral Width Tolerance Variation l

The reference storage cell design uses a Boral panel width of 5.00. The tolerance on the Boral width is

• 1/16 inch. Calculations (CASMO-3) showed that this tolerance corresponds to a i0.0019 Sk )

l uncertainty. i

' 4.5.2.3 Storage Cell I attice Pitch Variation l The ' design storage cell lattice spacing between fuel assemblies is 6.060 in. Decreasing the lattice pitch l

l increases the reactivity. For the manufacturing tolerance of 0.03 inches, the corresponding l

l uncertainty in reactivity is

• 0.0039 Sk as determined by differential CASMO-3 calculations.

L l'

t HI-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY f

4-6 P

m _ __._ - . _ - . . _ . _ _ _ _ . _ . _ _ _ . _ _ _ _ _ _ _ _ _ . . _ . . _ _ _ _ . _ _ _

i ' *

4.5.2.4 Stainless Steel Thickness Tolerances 1

5 4 i

~ The nominal thickness of the stainless steel box is 0.060 inches and 0.0235 inch for the steel backing  !

plate. The maximum positive reactivity effect of the expected stainless steel thickness tolerances was

!- calculated (CASMO-3) to be

• 0.0009 Sk.

I

I

! 4.5.2.5 Fuel Enrichment and Density Variation I

1 CASMO-3 calculations of the sensitivity to small enrichment variations yielded an average coefficient of 0.0061 Sk per 0.1 wt% 2"U, in the enrichment range from 4.0 to 4.25% enrichment. For an es-timated tolerance on percent2 "U enrichment of 0.06, the maximum uncertainty is *0.0037 Sk and becomes smaller for higher enriched fuel.

l 1

Calculations were also made to determine the sensitivity to the tolerance in UO 2 fuel density ( 0.14 g/cc). These calculations indicate that the storage rack k,, is increased by 0.0019 Sk for the expected ,

i maximum 10.59 gm/cc stack density. A lower fuel density results in correspondingly lower values of -

reactivity. Thus, the maximum uncertainty due to the tolerances on UO2 density is

• 0.0019 Sk.

4.5.2.6 Zirconium Flow Channel Elimination of the zirconium flow channel results in a small (less than 0.01 Sk) decrease in reactivity.

More significant is a positive reactivity effect resulting from potential bulging of the zirconium channel, which moves the channel wall outward toward the Boral absorber. For the maximum expected bulging based on estimates provided by GE and conservatively assumed to be uniform throughout all assemblies, an incremental reactivity of + 0.0039 Sk could result (determined by

~

differential CASMO-3 calculations). Fuel assembly bowing results in a negative HI-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY 4-7

Table 4.5.1 REACTIVITY UNCERTAINTIES' DUE TO MANUFACTURING TOI FRANCES Quantity Nominal Value Tolerance Sk.

Boron Loading 10B 0.0162 g/cm 2 O.0012 g/cm2 0.0056 Boral Width 5.00"

• 1/16" *0.0019 Lattice Spacing 6.06" *0.03" 0.0039 SS Thickness 0.06 and *0.004" 0.0009 l 0.0235" mean Fuel Enrichment 4.0%2 "U 0.06%2 "U 0.0037 1

Fuel Density (stack) 10.45 g/cm' *0.14 g/cm2 *0.0019 Statistical +0.0083 Combination' of l Uncertainties l I

l l

l 1

t l Square root of sum of squares in all independent tolerance effects.

! l 4

111-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY 4-16 l I

l l

l l

N

.i i

BORAL 0.0162 g 8-10/CW

. CENTERUNE TERU 5.00 t 1/1s" W1DE CENTER OF RORAL * *

. # IIIII 0.070

• 0.004" Thk SS CELL IN 0.077" CHANNEL 5.90* 0.03" 10

-m_m__.__.___.._.____.______ [ 0.060"/0.0235" T i 7 i @ wir (WEAN 0.04525")

9 x 9 Scuors Array .

2

. Pellet 00 0.376" s!

\, Fuel Density 10.45 g/cc

! h' Clod 10 Clod 00 0.384" 0.440" I

0.08025" om" ooog

W oler Rod 00 0.980" 0-000 s, x . ..m-Box 00 5.408" l 150 # 1 4 0 0 0 0 0 0'6. '

1. OOOOOOOh 0 0 0 0 0 0 0 0 t',

C T C-

.u. --

i I'-

5.90" CELL ID n.

a lI j 6.060" LA1TICE SPACING  ;

I .

l C

C Fig. 4.3.1 CROSS-SECTION OF TYPICAL STORAGE CELL (CALCULATIONAL WODEL) 4-19

$r /

i l

Table 5.6.1  :

i l DATA FOR LOCAL TEMPERATURE Fuel Assembly Array Size 8x8 l

Fuel cladding outer diameter, in. . 0.484 Fuel cladding inside diameter,in. 0.414 Storage cell inside dimension,in. 5.90 Active fuel length, in. 144 Number of fuel rods / assembly 62 4

Operating power per fuel assembly P x 10 , 11.873 l Btu /hr Cell pitch, in. 6.06 Cell height, in. 165 Bottom height,in. 5.75 Note: For the analysis, inertia resistance was increased by 50% and permeability was reduced 1 i

by 15% for conservatism in order to account for any slight discrepancies in the dimensions in the above table for other fuel types.

HI-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION 5-23

O.

floor is indeterminate. According to Rabinowicz [6.4.1], results of 199 tests performed on I austenitic stainless steel plates submerged in water show a mean value of p to be 0.503 with

. standard deviation of 0.125. Upper and lower bounds (based on twice standard deviation) are 0.753 and 0.253, respectively. Analyses are therefore performed for coefficient of friction values of 0.2 (lower limit) and for 0.8 (upper limit), and for random friction values clustered about a mean of 0.5. The bounding values of = 0.2 and 0.8 have been found to bracket the upper limit of module response in previous rerack projects.

Since free-standing racks are not anchored to the pool slab, not attached to the pool walls, and not interconnected, they can execute a wide variety of motions. Racks may slide on the pool floor, one or more rack support pedestals may momentarily tip and lose contact with the floor slab liner, or racks may exhibit a combination of sliding and tipping. The structural models developed permit simulation of these kinematic events with inherent built-in conservatisms. The rack models also include components for simulation of potential inter-rack and rack-to-wall impact phenomena.

t Lift-off of support pedestals and subsequent liner impacts are modeled using impact (gap) elements, and Coulomb friction between rack and pool liner is simulated by piecewise linear (friction) elements. Rack elasticity, relative to the rack base, is included in the model with linear springs representing a beam like action. These special attributes of rack dynamics require strong emphasis on modeling of linear and nonlinear springs, dampers, and compression-only gap i

! elements. The term " nonlinear spring" is a generic term to denote the mathematical element representing the case where restoring force is not linearly proportional to displacement. In the fuel rack simulations, the Coulomb friction interface between rack support pedestal and liner is typical of a nonlinear spring.

3-D dynamic analyses of single rack modules require a key modeling assumption. This relates to location and relative motion of neighboring racks. The gap between a peripheral rack and adjacent

~

pool wall is known, with motion of the wall prescribed. However, another rack, adjacent to the rack being analyzed, is also free-standing and subject to motion during a seismic event. To 111-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION 6-8

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

conduct the seismic analysis of a given rack, its physical interface with neighboring modules must be specified. The standard procedure in analysis of a single rack-module is to assume that neighboring racks move 180' out-of-phase in relation to the subject rack. Thus, the available gap before inter-rack impact occurs is 50% of the physical gap. This " opposed phase motion" assumption increases likelihood ofintra-rack impacts and is thus conservative. However, it also increases the relative contribution of fluid coupling, which depends on fluid gaps and relative movements of bodies, making overall conservatism a less certain assertion. 3-D Whole Pool Multi-Rack analyses carried out for Taiwan Power Company's Chin Shan Station, and for GPU Nuclear's Oyster Creek Nuclear Station demonstrate that single rack simulations predict smaller rack displacement during seismic responses. Nevertheless,3-D analyses of single rack modules permit detailed evaluation of stress fields, and serve as a benchmark check for the much more involved, WPMR analysis.

Particulars of modeling details and assuruptions for 3-D Single Rack analysis and for Whole Pool Multi-Rack analysis are given in the following subsections.

6.4.2 The 3-D 22 DOF Model for Sinole Rack Module 6.4.2.1 Acanmntions

a. The fuel rack structure is very rigid; motion is captured by modeling the rack as a twelve degree-of-freedom structure. Movement of the rack cross-section at any height is described by six degrees-of-freedom of the rack base and six degrees-of-freedom at the rack top. Rattling fuel assemblies within the rack are modeled by five lumped masses located at H, .75H, .5H, .25H, and at the rack base (H is the rack height measured above the baseplate). Each lumped fuel mass has two horizontal displacement degrees-of-freedom. Vertical motion of the fuel assembly mass is assumed equal to rack vertical motion at the baseplate level. The centroid of each fuel assembly mass can be located off center, relative to the rack structure centroid at that level, to simulate a partially loaded rack.

111-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION 6-9

5

,e

b. Seismic motion of a fuel rack is characterized by random rattling of fuel assemblies in their individual storage locations. All fuel assemblies are assumed to move in- I

! phase within a rack. This exaggerates computed dynamic loading on the rack i l

U structure and therefore yields conservative results.

c. Fluid coupling between rack and fuel assemblies, and between rack and wall, is ,

simulated by appropriate inertial coupling in the system kinetic energy. Inclusion of these effects uses the methods of (6.4.2] and [6.4.3] for rack / assembly coupling and for rack-to-rack coupling, respectively. Fluid coupling terms for rack-to-rack coupling are based on opposed phase motion of adjacent modules.

d. Fluid damping and fornidrag is conservatively neglected.

c. Sloshing is negligible at the top of ti": rack and is neglected in the analysis of the rack.

f. Potential impacts between rack and fuel assemblies are accounted for by i

appropriate " compression only" gap elements between masses involved. The l i

possible incidence of rack-to-wall or rack-to-rack impact is simulated by gap elements at top and bottom of the rack in two horizontal directions. Bottom elements are located at the baseplate elevation.

g. Pedestals are modeled by gap elements in the vertical direction and as " rigid links"  ;

l for transferring horizontal stress. Each pedestal support is linked to the pool liner by two friction springs. Local pedestal spring stiffness accounts for floor elasticity and for local rack elasticityjust above the pedestal.

h. Rattling of fuel assemblies inside the storage locations causes the gap between fuel assemblies and cell wall to change from a maximum of twice the nominal gap to a HI-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFOILMATION 6-10

. -. < j I

theoretical zero gap. Fluid coupling coefficients are based on the nominal gap.

6.4.2.2 Model Details Figure 6.4.2..shows -a schematic of the model. Si (i = 1,...,4) represent support locations, p, represent absolute degrees-of-freedom, and qi represent degrees-of-freedom relative to the slab. H is the height of the rack above the baseplate. Not shown in Fig. 6.4.2 are gap elements used to model pedestal / liner impact locations and impact locations with adjacent racks.

1 Table 6.4.1 lists the degrees-of-freedom for the single' rack model. Translational and rotational degrees-of-freedom 1-6 and 17-22 describe the rack motion; rattling fuel masses (nodes 1*,2*,3*,

4*, 5' in Fig. 6.4.2) are described by translational degrees-of-freedom 7-16. U(t) i represents pool floor slab displacement seismic time-history.

Figures 6.4.3 and 6.4.4, respectively, show inter-rack impact springs (to track potential for impact l

between racks or between rack and wall), and fuel assembly / storage cell impact springs at one  ;

location of rattling fuel assembly mass.

I

! Figures 6.4.5,6.4.6, and 6.4.7 show the modeling technique and degrees-of-freedom associated t

l with rack elasticity. In each bending plane a shear and bending spring simulate elastic effects i

l [6.4.4]. Linear elastic springs coupling rack vertical and torsional degrees-of-freedom are also included in the model.

Additional details concerning fluid coupling and determination of stiffness elements are provided below.

l

- 6.4.2.3 Fluid Coupling Details i .

I  !

4 - HI-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION 6-11 1

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

t

,q ~*

The " fluid coupling effect" [6.4.2],[6.4.3) is described as follows: If one body (mass mi) vibrates adjacent to a second body (mass.m2), and both bodies are submerged in frictionless fluid, then Newton's equations of motion for the two bodies are:

' (mi + M ) X i + M Xi2=2applied forces on mass mi + 0 (Xi ')

ii 2

M 2: X i+ (m2 + M 22) X =2 applied forces on mass m2 + O (X2 )

X i, X denote 2

absolute accelerations of masses mi and m2, respectively, and the notation O(X') denotes nonlinear terms.

Mn, M ,i2 M , 2iand M22 are fluid coupling coefficients which depend on body shape, relative disposition, etc. Fritz [6.4.3] gives data for M for e various body shapes and arrangements. The fluid adds mass to the body (Mn to mass mi), and an extemal force proportional to acceleration of the adjacent body (mass m2). Thus, acceleration of one body affects the force field on another.

This force field is a function ofinterbody gap, reaching large values for small gaps. Lateral motion

. of a fuel assembly inside a storage location encounters this effect. For example, fluid coupling is between nodes 2 and 2* in Figure 6.4.2. The rack analysis also contains inertial fluid coupling terms which model the effect of fluid in the gaps between adjacent racks. Terms modeling effects of fluid flowing between adjacent racks are computed assuming that all racks adjacent to the rack being analyzed are vibrating 180 out of phase from the rack being analyzed. Thus, the modeled ,

rack is enclosed by a hydrodynamic mass computed as if there were a plane of symmetry located in the middle of the gap region. Rack-to-rack gap elements (Figure 6.4.3) have initial gaps set to 50% of the physical gap to reflect this symmetry.

- HI-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION 6-12

_. _ . . _ _ _ _ _ . _ _ _ . _ . . . _ - _ _ .__.m. ._ _ _ _ . .

i l

• 7 .~,c 6.4.3.4 Modeline Details l

l Figures 6.4.9 and 6.4.10.show planform views of the spent fuel pool which includes rack and l l l pedestal numbering scheme and the global coordinate system used for the WPMR analysis. In l L

j. Figure 6.4.9,8 new' high density racks are installed in the north half of the pool and 8 existing

j. racks remain in the south half of the pool. In Figure 6.4.10, the whole pool is filled with 16 new l

Holtec racks. Table 6.4.3 gives details on number of cells per rack, and on rack and fuel weights for Holtec new racks. In Whole Pool Multi-Rack analysis, a reduced degree-of-freedom (RDOF) set is used to model each rack plus contained fuel. The rack structure is modeled by six degrees-of-l L freedom. A portion of contained fuel assemblies is assumed to rattle at the top of the rack, while the remainder of the contained fuel is assumed as a distributed mass attached to the rack. The l rattling portion of the contained fuel is modeled by two horizontal degrees-of-freedom.

t i- Thus, the WPMR model involves all racks in the spent fuel pool with each individual rack modeled as an 8 degree of freedom structure. The rattling portion of fuel mass, within each rack, is chosen to insure reasonable agreement between displacement predictions from single rack analysis l

using a 22 DOF model and predictions from 8 DOF analysis under the same conditions.  !

L  ;

i l The Whole Pool Multi-Rack model includes gap elements representing compression-only springs, representing impact potential at fuel assembly-fuel rack interfaces, and at rack-to-rack or rack-to-wall locations at top and bottom corners of each rack module. Each pedestal has two friction )

l elements associated with force in the vertical compression element. Values used for spring

- constants for the various stiffness elements reflect the values used in the 22 DOF model.

6.5 Accentance Criteria Stress T.imits. and Material Propedies i

l 6.5.1 Acceptance Cnteria L There are two sets of criteria to be satisfied by the rack modules:

I L

111 91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION ,

6-16 l

l

so l

Table 6.4.3 SPENT FUEL POOL LOADING l Cell Configuration (No. of Cells in Fuel Weight (Ibs)

Rack N-S x E-W Direction) Rack We'ight (lbs) (Channelled Fuel Assembly) j A 18x15 - 4x4 = 254 17,820 680 B 18x15-5x4 = 250 17,540 680 )

16,380 I C 18x13 = 234 680 D 18x15 = 270 18,940 680 l E 18x14+26 =276 19,430 680 F 18x17-5 = 301 21,120 680 G 7x17-1 = 118 8,280 680 M 8x16-6+15 = 137 9,620 680 l

HI 18x15 = 270 18,940 680 j H2 18x15 = 270 18,940 680 l H3 18x15 = 270 18,940 680 H4 18x15 = 270 18,940 680 Jl 18x17 = 306 21,500 680 J2 18x17 = 306 21,500 680 K 18x14+7 = 259 18,170 680 L 18x17-1x11 = 295 20,700 680 HI-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION 6-41

oe Table 6.5.1 RACK MATERIAL DATA (200 'F)

Young's Modulus, E Yield Strength, S y Ultimate Strength, S, Material (psi) (psi) (Psi) 304 S.S. 27.6 x 10' 25,000 71,000 Section III Reference Table I-6.0 Table I-2.2 Table I-3.2

1. ASTM-240, Type 27.6 x 10' 25,000 71,000

(~

304 (upper part of support feet)

2. ASTM '564-630 27.6 x 10' 106,300 140,000 (lower part of support feet; age hardened at 1100'F)

)

) .

HI-91738 SHADED REGIONS ARE HOLTEC PROPRIETARY INFORMATION 6-42

. . -