ML19340A124

Criticality Evaluation for Dry Storage of Fresh Fuel Assemblies in Oconee Unit 3 Spent Fuel Pool.

ML19340A124
Person / Time
Site:	Oconee
Issue date:	04/16/1976
From:	DUKE POWER CO.
To:
Shared Package
ML19340A120	List: ML19340A118 ML19340A124 ML19340A125
References
NUDOCS 7911250044
Download: ML19340A124 (13)

Text

N O

e ATTACHMENT 1 CRITICALITY EVALUATION FOR DRY STORAGE OF FRESH FUEL ASSEMBLIES IN OCONEE UNIT 3 SPENT FUEL POOL i

9 911250 NT

~.

,i -

4

, CRITICALITY EVALUATION FOR DRY STORAGE -

0F_ FRESH FUEL ASSEMBLIES IN OCONEE UNIT 3 SPENT FUEL POOL I. INTRODUCTION AND'

SUMMARY

A criticality analysis is provided herein .to support the proposed dry

-storage of fresh fuel assemblies of' enrichment up to 2.9 weight percent U-235 in the high capacity fuel storage-rack modules located in Oconee Unit,3 spent fuel pool by demonstrating that the multiplication factor of the array is less than or equal to the design limit multiplication factors of 0.95 under flooded conditions and 0.98 under optimum moderation conditions.

Placement of1 fuel assemblies in each storage location and'a checkerboard loading scheme-(only diagonally adjacent storage locations being occupied by fuel assemblies) are examined. Results of the analysis show that under fully flooded or uniformly dispersed aqueous form conditions either fuel

~

loading scheme results in' multiplication factors less than 0.88. As-expected,-the checkerboard loading pattern results in lower calculated multiplication factors _ than .the fully loaded pattern and provides margins of 0.12 and 0.36 in-units of k at the fully flooded and mist conditions, respectively. The analyses for the fully loaded rack conditions demonstrate'that-the inadvertent misloading of a single fuel assembly in an otherwise checkerboard array does not lead to a violation of suberiticality margins.

-Potential non-uniform-flooding conditions have also been examined.

Specifically, it is hypothesized that the storage cells were filled with

. full density water and that the space between storage cells was filled with' lower density water. Under this hypothetical condition also, the checkerboard loading pattern was_found to be acceptable.

II. DESIGN BASES AND INPUT PARAMETERS A criticality, analysis is performed to demonstrate that the effective

' multiplication factor of the~ normal 1y dry' fuel storage rack, when loaded with fresh fuel of the' highest anticipated _ enrichment, will not exceed:

+

t -

.(1) 0.95 shen flooded with pure water, and .(2) 0.98 assuming optimum moderation (aqueous foam. condition). -These criteria are consistent with the applicable. industrial standard, ANSI N18.2( ).

The maximum enrichment of the' fuel assemblies to be stored in the normally dry - rack.is ' assumed to be less than or equal to 2.9 w/o U235 ,

' Relevant physical parameters of the-fuel assemblies employed in the analysis are the' nominal' design values. Where ranges of parameters are shown, extremum values were chosen such that the predicted multiplication factor of the storage rack is a maximum. The inherent neutron-absorbing effect of.the stainless steel ~ storage box wall structure is explicitly treated in the analysis. Credit has not been taken for neutron absorption of the assembly grid spacers and end-fittings, nor for neutron absorption by structural steel components of the storage rack otba- than the-individual storage box wall structure.

Storage cells and modules are shown in Reference 2. The analysis is performed assuming that the storage cells consist of rectangular boxes, with a nominal inside dimension of 9.375 inches, constructed of 0.25 inch thick type 304 stainless steel (the guide' funnel and end casting are l- neglected). These boxes are welded to structual components to form storage modules with a nominal center-to-center distance between adjacent boxes of 14.09. inches. Two modules of 6 by 8 storage cells each are l welded together to form'a regular 8 by 12 array of-storage cells. To account for manufacturing tolerances, it is assumed that the dimensions j (within tolerance) of the storage cells and modules are such that the l

j. predicted multiplication factor is a maximum. Hence, results of the analysis presented in this report are based on an assumed: (1) trinimum i i

cell center-to-center spacing of 13.965 in., (2) minimum box wall thickness of 0.24 in. , and (3) maximum box inside dimension of 9.9375 in.

1To_ conservatively represent neutron scattering by materials surrounding the storage rack, it has been assumed that the array is bounded by a

, three foot thick concrete wall spaced one foot from the edge of the sto. rage array on all six sides.

l l

t.

L

f j_g

' III .1 ANALYTICAL METHODS

~ A. General In order to more accurately predict the~ multiplication factor of the 1 . storage arr4.s, reliable calculation of the spatial flux distribution, especially in'the neutron absorbing steel-regions, is essential.~ For this reason, one and two dimensional transport calculations for the storage rack are employed. In the two dimensional-transport calculations, each componentaof the fuel storage location " cell" is explicitly represented.

Thus, in the normal' storage' cell calculation, the fuel assembly, the water channel between the fuel assembly and the box walls, the steel box, and one half of the~ water gap between adjacent storage locations are represented as separate regions. The fuel assembly itself is represented as a 15 x 15 array.of cells containing moderator and either fuel pins, guide tubes or-instrument tubes. Four neutron group- cross sections are generated for each fuel assembly cell and for each component of the storage cell with special attention given to the effect of adjoining regions on the spatial thermal neutrcn spectrum and hence broad group thermal cross sections of each~separato region.of the storage cell. Flux-volume weighted cross sections, extracted from the two dimensional transport calculations, are used in the one dimensional transport calculations as described below.

B. Cross Section Generation.

The CEPAK lattice program (Version 2.2 Mod 10) is used to calculate four neutron group cross sections. This program is the synthesis of a number of computer codes, many of which were developed at other laboratories, e.g., FORM (3), THERMOS (4) , and CINDER (5) . These programs.are interlinked in a consistent-way with an extensive library of differential neutron groups between 0 and 10 Mev. Neutron leakage in a single Fourier mode is -

represented by-a B-1 approximation to transport theory throughout this entire range. Resonance shielding is determined analytically. The effectiveLfuel temperature is incorporated into the calculational model Eby meansfof the Hellstrand correlation renormalized to a gold resonance sintegral of 1565 barns. This correlation'is a semi-empirical fit to

.  ; experimental data-for both-metal and oxide uranium rods. The Hellstrand

, 3

m. ~

~

correlation is employed for'U-238, with appropriate adjustments.

' guided by Monte Carlo calculations of.resonanc'e capture in U-238 so.as to provide agreement with selected measur ements of the conversion-ratio.

Plutonium resonance integrals are determined from an intermediate resonance formulation using equivalence relationships for the lattice representation ( ). .The-Dancoff factor D, which is a measure of the shielding.of a fuel rod resulting from the presence of neighboring-fuel rods, is calculated by the Fukai method (0) for a uniform lattice. This method carries out the numerical integrations necessary for the computation of the moderator and clad transmission probabilities. Vacancies in the lattice are treated by an approximation used successfully by Hicks (9} which apportions the uniform lattice Dancoff correction,C, (C = 1 - D), equally among the nearest neighbors.

~

The. data base for both fast and thermal neutron cross sections for this version of ~ the. CEPAK program is derived from several sources, mainly ENDF/B-II, BNL-235, and early Bettis libraries. This data base gives good-agreement with measured data from critical experiments and operating reactors. The standard multigroup cross section library employed in the CEPAK lattice program for SS-304 has a macroscopic 2200 m/s absorption cross section of 0.2597 cm . The use of ENDF/B-4 2200 m/s absorption cross sections would_ yield a larger 2200 m/s macroscopic cross section by approximately 3.7% with a variation of approximately 1% due to typical variations in nuclide composition and density of the type 304 allo'. ;

Thus the.2200 m/s vas e of the absorption cross section derived from 1CEPAK should yield a more1censervative thermal' absorption rate in SS-304 than one derived from.ENDF/B-4 data sources.

The fuel assembly region of- a storage cell is represented by a 15 x 15 array -of fuel assembly cells having a basic pitch of 0.568 inches and has an overall square dimension of 8.52 inches. Microscopic cross sections fo'r nuclides.in the fuel _ assembly cells as well as those exterior to the fuel--assembl'y but within'the outer boundary of the stainless steel box are averaged over the multigroup spectrum calculated by the FORM portion

of the CEPAK lattice program for a homogenized representation of the fuel assembly. The broad group thermal cross sections are obtained from the 4

9 one-dimensitual THERMOS. portion of CEPAK for ccch typa of fusi usembly cell;' control rod guide tube and instrument cells employ an explicit representation of moderator and structure within thecell and a homogenized fuel pin cell environment. Four broad neutron group (3 fast and 1 thermal) microscopic cross section edits are obtained from the CEPAK lattice pro-gram. Heterogeneous fast fission effects are included in the' top broad group cross sections by applying correction factors derived from an auxiliary two-dimensional' integral transport calculation that employs the collision probability technique to compute sub-region dependent reaction rates in_an explicit geometric representation of the fuel rods and associated structure of a fuel assembly. The correction factors are the relative flux ratios for the fuel, clad, and moderator within a fuel rod

- cell. The 3 fast broad group cross sections for the moderator region between storage boxes are obtained from a uniform moderator medium CEPAK calculation using water of an appropriate density as the moderating material. The thermal cross sections for the water and steel regions-are derived from slab geometry THERMOS calculations with an appropriate fuel assembly environment.

t C. Two Dimensional Transport Calculations The two dimensional, discrete ordinates transport code DOT-2W(10) (Version l.0 MOD 1 - May 7, 1973) is used to determine the spatial flux solution and multiplication factor. An S-6 order of angular quadrature is used with a 1.0001 convergence factor (the ratio of successive eigenvalues for each outer iteration). In the fully loaded storage cell calculations, one quarter of an assembly is. represented with one mesh interval for each fuel assembly cell; the surrounding water channel, steel, and water gap regions are calculated with 2, 4 and 6 intervals, respectively. Thus the X-Y representation of the fully loaded storage cell is a 20 x 20 mesh interval problem. The'same general principles are followed in the representation of the checkerboard loading scheme. In this model one quarter of each storage cell of a cluster of four storage cells is represented. The X-Y DOT representation of the checkerboard loaded storage array is a 40 x 40 mesh interval problem.

5

D. .On7 Dimen-irnal Trrn" port Calculctienc

.Non-leakage probability values of the storage rack are obtained using the' one dimensional transport code ANISN (Version'1.0, MOD. 0) as

-described below. These calculations are performed using 4 energy group modififed Po cross sections and an S8 quadrature set. Three regions are represented explicitly in these calculations: (1) a fuel assembly and storage rack region with flux-volume weighted cross sections obtained from the two dimensional transport (DOT)' calculations, (2) an aqueous foam region, and (3) a concrete region. The latter regions are represented using cross sections obtained using-the CEPAK code. The ANISN calculations are performed at several water densities of interest for: (1) an infinite storage rack, (2) a rack 8 storage cells wide, (3) a rack 12 cells long, and (4) a rack 144 inches high. The latter three calculations are performed assuming one foot of low density water and three feet of concrete surrounding the rack. The non-leakage probability is defined by the following equation.

P g = c(width) x e(length) x c(height)

(e=)

where: 1. P g is the non-leakage probability

2. c(width), c(length), e(height)-are the computed multiplication factors assuming the storage rack is of infinite extent in two directions and finite in the third dimension, i.e., length, width or height.

3. e= is the computed multiplication factor assuming the rack extends infinitely la all directions.

IV. RESULTS

_Past experience from criticality evaluations for dry fuel storage racks has shown that the multiplication factor varies with the assumed density of water' dispersed uniformly-througout an infinite array of fuel storage cells in the following fashion. As the water density is reduced below the value at 68 F, the multiplication factor decreases in a monotonic manner to a water density'in the range of 0.5 gm/ce; as the water density is reduced to zero, the multiplication factor passes through a 6-t

y .

.a e

maximum. Tne. maximum valus of the multiplication factor'at.both the-full and reduced. water density conditions is aLfunction of the fuel enrichment, size of the' fuel' assembly, lattice pitch'of the fuel assembly storage array, and'the amount and distribution of parasitic structural material in the storage rack. For the conditions where

~

-two 6 x 8 HI-CAP type fuel storage modules are' combined to form an

.8 x 12' array of fuel' storage locations, the lattice spacing and array size are fixed; the only remaining variable is the fuel loading configu-ration since the -limiting . enrichment- is set at .:2.9 w/o.

Figure 1 summarizes the results of analyses for an array of fuel storage locations which are of infinite extent in all directions. The data

. points at a relative water density of 1.00 correspond to complete immer-

~

~

sion of the rack in water at 680 F for three cases - all storage locations filled by fuel assemblies having enrichments of 3.5 w/o and 2.9 w.o, and a checkerboard array.of 2.9 w/o enriched fuel assemblies. The calculated multiplication factors are 0.9070, 0.8711, and 0.8233, respectively, assuming concurrent adverse dimensional tolerances as specified in s Section II. These data points are of interest for comparing the change

-in multiplication. factor with the changes in enrichment and fuel arrange-mentrelativet$thelicensedconditionswhenthefuelstoragemodules-are employed in the Oconee Nuclear Station, Unit 3 spent fuel pool.

In the event that the fuel storage array could be exposed to a suffici-ently large volume of water from fire fighting apparatus, pipe breaks, etc.,--such that the funnel at the top of each storage location would divert most of the water to the interior of the storage box, it is postulated that the most adverse condition would be a complete flooding of the interiorcof each fuel storage box,3dth a relatively low (0.02 gm/cc). water density between storage _ boxes. Analyses for this postulated condition. indicate that'for the checkerboard arrangement of fuel assem--

'blies, the multiplication factor for the infinite array increases from 10.8233-(uniformly. flooded) to-0.8404. The calculated multiplication factor of 0.8404 corresponding to this hypothetical condition shows a significant margin to the design limit of 0.98.

~7

,.s m Results are shown in Figure l~at reduced water density conditions for both the checkerboard and fully loaded _ rack conditions. For each fuel'

~

l'oading. pattern and relative water density, an upper and lower bound in the calculated _ multiplication factor is shown. The range of values

~

corresponds to changes in the four group macroscopic cross sections due to large variations in the multigroup_ spectrum. -These variations are induced, via the buckling parametar,' to examine the sensitivity of the-

'results co the slowing-down spectrum in the fuel and bulk moderator regions. Values of the energy and spatial dependent neutron leakage

, inferred from the two dimensional transport calculations lie within the

--band of assumed input bucklings.

The multiplication' factors shown in Figure 1 for reduced water density conditions show a margin to the design basis of 0.98 in excess of either 0.12 units of k for an infinite array of storage cells, each of which contains~one fuel assembly, or 0.36 units of k for an infinite array of storage cells containing one fuel assembly in every other location

- (checkerboard array) .

Figure 2 shows a plot of the non-leakage probability for the finite

-checkerboard array (8 x 12) of fuel assemblies at the reduced water density conditions. Figure 3 shows the effective multiplication factor (product of infinite multiplication factor and non-leakage probability) for the checkerboard array of 2.9 w/o fuel assemblies as a1 function of water density. For this finite checkerboard array of fuel assemblies at the ootimum moderation conditions,-a margin of 0.49 units of k exists relative to the. design basis value of 0.98.

4 8

.m. -

REFERENCES:

1. ' ANSI-N18.2-1973, "Americal Nuclear Standard, Nuclear Safety Criteria for the Design of Stationary Pressurized Water Reactor Plants, August 6, 1973.

2. "Oconee Nuclear Station Unit.3 Spent Fuel Storage Facility Modification Safety-Analysis' Report,"_ submitted to the NRC by Duke Power Company, letter of W. O. Parker, Jr. dated September 12, 1975.

3. FORM - A fourier Transform Fast Spectrum Code for the IBM-7090, McGoff, D. J., NAA-SR-Memo 5766 (September 1960).

4. THERMOS --A Thermalization' Transport Theory Code for Reactor Lattice Calculations, Honeck, H. , BNL-5816 (July 1961).

5. CINDER - A One-Point Depletion and Fission Product Program, England, T. R.,

WAPD-TM-334 (Revised June 1964).

6. Measurement of Resonance' Integral, Hellstrand, Proceedir.gs of National Topical Meeting of the ANS, San Diego, February 7-9, 1966, The M.I.T. Press.

7. "An Equivalence Formulation for Absorption in Plutonium," S. Borrensen and R. Goldstein,'Trans. Am. Nuclear Society, 15, 296 (1972).

8. Y. Fukai, J. Nuclear Energy, 22,-355, 1968.

9. R. Alptar, METHUSELAH I., AEEW-R-135, 1964.

10. R. G. Sottesy, R. L. Disney, A Collfar, " User's Manual for the DOT-IIW Discrete Ordinates Transport Computer Code," WANL-TME-1982.

11. W. W. Engle, Jr., "A Users Manual for ANISN," K-1693, March 30, 1967.

9.

m (

Figure 1.

Infinite Multiplication Factor vs. Water Density '

I

'- 1.

,g f . ,

. Design Basis 0.98

0. 9 , A:. .(1) .

I (1) Fully Loaded 3.5 w/o.

~ ~[ (2) (2)~ Fully Loaded 2.9 w/o

.f 9

-S % (3) Checkerboard 2.9 w/o'.

0. 8 - '

/ ,

U. 2 y L- p Fully Loaded

\j/////.,: ,

h 0. 7 -

. c=g9 e

-U. ..

[/ ,

i

@5 gj 0. 6 - -> Checkerboard loading ..

.,j ' ,//

\ @ ;p Pattern 7

/ !/ / ff <

1

'& .U. O,f

~

/i.  !/,/j/ /!iiitti //

c ==3 Z .

/

). // l /- [/

l,'  ?

h

c=== Q@U. l l . /!

8

0. s _ O / ~/ '

'V y

h ,

.

;. -+--- .t l, f 1.00  :) .. O.14 0.12 0.10

-- {

'O.n8 n.06 0.04 0.02 i 0.00 RELATIVE HATER DENSITY i

~=-

I.i 1

\1, .s

~

'  ; ~-.

y' t

i s

n g

e D

r t

e 0 g :0 W -

0 s

v y

a r 2 r :0 A. .

0 d

r a

o b

r 4 e :0 k

c 0 h

C f

o y

N 6 t

i0 i O l

i '

b a

b o 8 Y r T P e0 I e n S g N E

a D k

a R e o E L t T A

n 4 0 N o

N E V

I 2

T A

2 L e  : 1. E r R u 0 g

i F

4

1 0

- - ~

9 8 7 6

'0 0 0 --0 gaA5exc wedg4.,0 39 3- 2-

?

%r6QM9)2c 9V . b :- bb y , I

1

/ i s.'

1 n

0 0

d r 2

0 o  :

b 0 r

ey i .

k t  ; !.I .i ci es I / /

h n /

C e 4 D I

! /.  : 0 f

or  !'  !* ' f 0

- e  !

rt l i oa t 1 c

as 1

'j l! \'

i ~

Q 6

F v Y.  : 0 l I ns 0 oe i i t l ab cm i e l s 8 p s . 0 i A t 0 l l ue Y M u T F I e S vf N i o 0 E t t 1 D c y .

ea 0 R f r E f r I T

EA A J

I

~ 2 E 3  : I V

. I e l*

0 T r

u A.

I g ]y E i

H F r 4

1 l .

0

- .- 1I )

}i eg f

~0 8 4 2 0 5 '4 4 4 4 4 0 0 0 0 0 0

~1, spk 6 r ~

gcu4g g~DU,g gF_ 5f~ 4:-@CQN s  ;

y meV . ' ^

7