ML20043E804

Nonproprietary Criticality Safety Analysis for Grand Gulf Fuel Storage Racks W/ANF-1.4 Fuel Assemblies.

ML20043E804
Person / Time
Site:	Grand Gulf
Issue date:	06/01/1990
From:	Gerrald L SIEMENS POWER CORP. (FORMERLY SIEMENS NUCLEAR POWER
To:
Shared Package
ML20043E802	List: ML20043E801 ML20043E804
References
ANF-90-060(NP), ANF-90-60(NP), NUDOCS 9006130403
Download: ML20043E804 (36)

Text

. . , . . . . .~ nnn. . - - . - . .. .

. . . . . . . - - . n

(, *-# .

V ,

1 h  !

ANF-90-060(NP)

[}$

N Jn.. . ADVANCEDNUCLEARFUELSCORP RATION CRITICALITY SAFETY ANALYSIS FOR THE GRAND GULF FUEL STORAGE RACKS L

WITH ANF-1.4 FUEL ASSEMBLIES-l

• j

)

L ,

1 I

l.

JUNE 1990 l

1.

o l

l r 61" [2f3 6 E P

+

i ADEANCED NtJCLEAR FlJELS CORPORATION ANF 90-060(NP)

Issue Date: 6/1/90 ,

.l CRITICAUTY SAFETY ANALYSIS FOR THE GRAND GULF SPENT FUEL STORAGE RACKS 1 WITH ANF 1.4 FUEL ASSEMBUES

-I 1

'l l'

l

(.

i.

l

'l l

June 1990 l

i4

.m ..a.a a a wd. , 4-in.4J'a A e of .6ma *4& 4- .2 -.a d.-a3-e4+44e enaA-.2 8 A.A- e.4e-Ie 'a e

u ,

1( ,

l l

l CUS7000ER DISCLAltfER l

1 N ANT N M N W M Olt M 00CUest?

8LE485 81540 CAREPULLY esmmg the siegeet meer of Ine coeument are tnoes est form m tne Agreement between Advangell Nueleer Puste Corporesen and the Cuommer pursuant to women mee assument a neues. Accorengsy, esseet as omermee omroesy pro-i-

wenn in suon AqPoement, nemer Asvenood Nuelser Puole Corporeson nor any pereen eeong on sie tienelf males any wervenly or representemen, egressed or

l. wnsees. wan resseet a me seawney, comonenses, or unmunese e me wwor. l meson annemes m tne eneument. or ene me use a any wwormanen. aooersus.

memed or plassee desseed m thee desument wel not minnge prwassey owned regno; er sesumes any lleudlines won respect a me use of any informaten. ap-peress, memell or psesses 880100ed m Inse soeument.

The Nilanneen eeniesned nerem a for the solo use of Cuesomer, in oreer e aveel imessment of rignes of Asvenese Nueeser Fusse Corporeson m '

potenes or woveneens unen may se ineoused a the moormamon contamen m me meument. Ste resseent, try she aggestance of tne document, agrees not to pussen er mese puses use on ene pesem use of me terms of euen wwormenon unot so auenersed in wrong my Asveness Numeer Pues Corgerenon or untd aner six (S) rnOnWIS % N or eEpersion of tne omreened Agreement and any essensen Uternet, unsees einermes seerasedy provices e me Agreement. No none or doenses m or e any assene are wnones my me fumenmg of me one#

rnent.

i t

l l

1; ANF-3145.472A (12/87) 1 L.e_ .

ANF-90-060(NP)

Issue Date: 6/1/90 CRmCAUTY SAFETY ANALYSIS FOR THE GRAND OULD SPENT FUEL STORAGE RACKS ,

WITH ANF 1.4 FUEL ASSEMBUES

.i Prepared By: k Date:

L D3errald, Criticality Safety Specialist '

Safety, Security, and Ucensing Reviewed By: e. / Date:' [f// 90 J. FAper, Second Party Reviewer '

Safety, Security, and Ucensing l

Reviewed By: '

Date: 6!/,!f > -

t-L J. Feddried, Manager I BWR I Engineering Approved By: ~ Date: / b T. C. Probasco, Safety Supervisor Safety, Security, and Ucensing i i

1; .c ,

K ANF 90-060(NP) li Pagei l:

I 1 TABLE OF CONTENTS r

Section Eggg 1- q

, 1.0 I NTR O D U CTI O N . .. .... . .. ........... . . .. . .. .... ..... . . . . . . . . . . . . .. . . . . . . . . . . .. . . 1  !

! 2.0

SUMMARY

l .......................'.............................

. l L. 3.0 0 E S I G N BAbi S. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

r 3.1 Fuel Design Parameters ............... .. .. .... ........... .............. ...... 4 3.2 Spent Fuel Rack Desig n.................................... ............... .. 4 i 4.0 M ETH O D O LOGY.. . . .. . . . . .. . .. ... .. ... ........... . .. .. . .. . .. . . . . .. . . . .. . . . . . . . . 12 5.0 ~ R E F E R E N C E S . . . . . . . . . . . . . . . . . ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . !.

APPENDIX A GADOLINIA-BURNUP EFFECTS.......................... A 1 '

t APPENDIX B - BORAFLEX GAP EFFECTS............................... .. B-1 APPENDIX C - CASMO-KENO COMPARISONS.......................... C.1 APP EN DIX D - U NC ERTAINTIES..................................... ............. D-1 APPENDIX E - METHODS VAllDATION....................................... E 1 h

i I

f i

b w ~-e,

ANF 90-060(NP)

Page li LIST OF TABLES 1

1 81211 f.agg 2.1 Major Conservatisms Used in Analysis.................................................. 3 3.1 5

3.2 GGNS Spent Fuel Storage Rack Design Parameters................. ............ 6

, - n-- w~-

ANF 90-060(icP)

Page ill UST OF FIGURES f]2Et -

Eggg 3.1 Grand Gulf Unit 1 Cycle 5 ANF 1.4 Enrichment Distribution for Eight Gadolinia Rods................................................ .... .... 7 3.2 Grand Gulf Unit 1 Cycle 5 ANF 1.4 Enrichment Distribution for Nine Gadolinia Rods.......... ..... ................................... ... .. .. 8 3.3 Grand Gulf Unit 1 Cycle 5 ANF 1.4 Enrichment Distribution for Ten Gadolinia Rods................. ... ........ ... ......... ...... ..... . .... . 9 3.4 10 3.5 11

s; , a ye-i: i*

4 l s

l. ANF 90-060(NP)

Page 1 FINAL REPORT CRITICALITY SAFETY ANALYSIS L

FOR THE GRAND GULF SPENT FUEL STORAGE RACKS WITH ANF-1.4 FUEL ASSEMBUES '

L L

t.

1.0 INTRODUCTION

The spent fuel racks at the Grand Gulf Nuclear Station (GGNS) were designed for 8x8 fuel assemblies with an average enrichment up to 3.5 Wt% U-235. Previous criticality safety analyses (CSAs) have demonstrated the acceptability of enrichments of at least 3.61% for 8x8 assemblies (Reference 1) and at least 3.47% for 9x9 Lead Test Assemblies (LTAs) (Reference 2).

This analysis addresses the storage of two additional 9x9 fuel assembly types (ANF 1.4L L low gadolinia fuel and ANF-1.4H high gadolinia fuel) in the spent fuel storage pool. These- .

assemblies are scheduled to be loaded in Grand Gulf 1 Cycle 5 and both types contain natural l-uranium axial blankets, axially-zoned gadolinia, and an enriched zone of 3.79%. The new types are described in detal! in Section 4.0.

l' l 2.0

SUMMARY

This analysis demonstrates that the GGNS spent fuel storage racks remain adequately suberitical (k eff .<_ .9452) for the storage of ANF 1.4L and ANF 1.4H fuel under worst credible conditions. The analysis was performed using a conservative medel of the racks and ANF Cycle 5 fuel. . Model conservatisms applied to the fuel assembly and the racks are listed in Table 2.1.

Uncertainties in the calculation model and those due to system component tolerances were also included in the evaluation. This analysis includes the reactivity effects of shrinkage of Boraflex due to irradiation.

L 1 1

I 1

i

r ANF 90-060(NP)

Page 2 The methodology used included CASMO 3G (Reference 3) and KENO Va (Reference 4).

Both codes share wide acceptance for fuel analysis througnout the nuclear industry. This .

analysis takes into account the effects of both gadolinia and burnup on storage array reactivity.

The fuel was depleted with CASMO 3G to determine the maximum reactivity of the assembly over its lifetime. Storage array reactivity k eff was determined assuming all fuel is at its peak reactivity during its ilfetime. Storage of ANF 1.4L and ANF 1.4H fuelin the GGNS spent fuel storage racks meets the requirements of NUREG 0800, i.e., k-eff of the array under worst credible conditions -

including all uncertainties will not exceed a k-eff of 0 95 . .

e

1 1[ .

[

i

(.

ANF 90-060(NP)

Page 3 l

n L l TABLE 2.1 - MAJOR CONSERVATISM USED IN ANALYSIS I l

l l

FUEL ASSEMBLY PARAMETERS i The most reactive region in either of the Cycle 5 assemblies'is the "9G3.0" lattice (nine h

rods with 3.0 Wt.% Gd,0) 3 which is present in the upper six inches of the 132 inch' enriched zone of the ANF 1.4L design. The remaining 126 inches are considerably less reactive due to poison rods with a higher Gd,03 content. The calculations conservatively

l. assumed that the entire enriched zone was the 9G3.0 lattice.

lJ STORAGE RACK PARAMETERS The Boralex panels were modeled using very conservative assumptions on its properties:

- The panels were assumed to shrink 4% in width.

The length shrinkage was assumed to result in a 6-inch gap. A conservative I

probability distribution of the number of gaps per cell was assumed.

The gaps were assumed to be randomly distributed within the central 50% of the panel length.

The panels were modeled at the minimum (nominal minus tolerance) length and minimum width.

The B C content was assumed to be 90% of the 95/95 lower limit of as fabricated i value.

l

. The rack was modeled as an infinite planar array. The actual system is less reactive due to radial leskage and due to intermodule spacings. l l

OTHER PARAMETERS 1 i

The system temperature was modeled as 20*C, The higher temperatures which exist in the actual system will result in a less reactive system.

The pool water was modeled as pure water; it contained no boron or any other '

absorbers.

I 1

l 1

j s

ANF 90-060(NP)

Page 4 3.0 DESIGN BASES

l. i 3.1 Fuel Desion Parameters g The design description of the ANF-1.4L and the ANF 1.4H assemblies are provided in Table 3.1 and Figures 3.1 through 3.5. Nominal values were modeled unless noted otherwise.

The effect of component tolerances on the fuel assembly and rack designs on the system

- reactivity is described in Appendix D.

3.2 Soent Fuel Rack Desian The GGNS spent fuel racks are a high density array of storage cells with fixed neutron absorbers. The wall between adjacent cells is 0.070" nominal thickness Boraflex with 0.063" nominal thickness stainless steel cladding on both sides. The dimensions for the racks were based on Drawing D-7373 (Rev. 6) of Joseph Oat Corp, for the GGNS spent fuel storage racks.

Key parameters are listed in Table 3.2.

The fixed neutron absorber (Boraflex) was conservatively modeled in dimensions and composition:

Two widths of Boraflex were used in the rack construction: 11.5 1 0.0625" (spanning two cells) and 5.625 0.0625"(spanning one cell). Ninety six percent of the minimum values were used in the KENO model.

l The minimum Boraflex length (143.75") was modeled.

Boraflex was modeled as pure B,C with B 10, B-11 and C densities at only 90%

of the 95/95 lower limit of the vendor's assay data. The modeled densities for B-10, B-11, and C are 0.107,0.473, and 0.161 gm/ce, respectively.

L Since the Boraflex was modeled with the minimum length, minimum width, and minimum B 10 density, no uncertainty adjustments are needed for these parameters. The strips of stainless steel at the edges of the Boraflex regions wore included in the model.

L L;

l l

"c T' '

Wtgr:34 7*6 4 "V ' ,'2WF7N 3

)

.-1 e s

= k i

a e kL

p y - ANF 90-060(NP);-

Page 5 ]-

g: ,

h

>f

..e

] 4 m,

] ,

I e

.V

=,

- s

~

l

(

i I

.r

-l

)- (

~

- !j 4

I 0

[-

. ].

^'

,3 t 3

?

[- (

! 1

l. I L t p d i,

(

j. h f $

V ]

v  :

i

.c

-  ;)

[. -

a-+u- .~ . -

E

\

ANF 90-060(NP)

Page 6 E TABLE 3.2 GONS SPENT FUEL STORAGE RACK DESIGN PARAMETERS g

i l

[ .

Value Assumed in Parameter Soecification - Model lI- '

Cell Pitch (inch) 6.2585 1 0.062 6.2585 Boraflex:

Thickness (inch) 0.070 0.007 0.070 Length (inch) 144 1 0.25 143.75 Minimum gm B 10/sq.cm. 0.0190 0.0171 l

Stainless Steel:

Thickness (inch) 0.063 i 0.006 0.063 l

1:

6 l

l .'

1 i

l l

lC ,

ANF 90 060(NP)

. Page 7 L1 MLI. M1 MH1 MH1 MH1 M1 ML1 L1 ML1 LLl* MH1 H2 LL2* H2 MH1 LLl* ML1

. M1 MH1 MH1 H2- H2 H2 MH1~ MH1 M1 MH1 H2 H2 LL2 W H2 H2 H2 MH1 MH1 LL2* H2 W W W H2 LL2* 'MH1 . ,

1

= MH1 H2 H2 H2 W LL2 H2 H2 MH1 1

MI MH1 MH1 H2 H2 H2 MH1 MH1 M1 ML1 LLl* MH1 H2 LL2* H2 MH1 LLl* ML1 L1 ML1 M1 MH1 MH1 MH1 M1 ML1 L1 Ll. Rods ( 4) ---

2.67 w/o U235 ML1 , Rods ( 8) ---

3.33 w/o U235 M1 Rods ( 8) ---

3.66 w/o U235 MH1 Rods (24) ---

3.98 w/o U235 H2. Rods (22). ---

4.73 w/o U235 ,

LL2 Rods ( 2) ---

2.27 w/o U235 LLl* Rods ( 4) ---

2.27 w/o U235 + 5.5 OR 7.0 w/o Gd203 LL2* Rods ( 4) ---

2.27 w/o U235 + 5.5 OR 7.0 w/o Gd203 W Rods ( 5) ---

Inert Water Rod FIGURE 3.1 GRAND GULF UNIT 1 CY5 ANF-1.4 ENRICHMENT DISTRIBUTION FOR EIGHT GADOLINIA RODS

s ANF 90-060(NP)

Page 8 t

L1 ML1 M1 MH1 MH1 MH1 M1 ML1 L1 L

• ML1 LLl* MH1 H2 LL2* H2 MH1 ~LLl* - ML1 M1 MH1 MH1 H2 H2 H2 MH1 MH1 M1 MH1 H2 H2 LL2 W H2 H2 H2 MH1 I MH1 LL2* H2 W W W H2 LL2* MH1 MH1 H2 H2 H2 W LL2* H2 H2 MH1 M1 MH1 MH1 H2 H2 H2 MH1 MH1 M1 ML1 LLl* MH1 H2 LL2* H2 MH1 LLl* ML1 l

ll ML1- M1 MH1 MH1 MH1 M1 ML1 L1 l l

1 l-l L1 Rods ( 4) ---

2.67 w/o U235 ML1 Rods ( 8) ---

3.33 w/o U235 M1 Rods ( 8) ---

3.66 w/o U235 MH1 Rods (24) ---

3.98 w/o U235 i H2 Rods (22) ---

4.73 w/o U235

LL2 Rods ( 1) ---

2.27 w/o U235 l: LLl* Rods ( 4) ---

2.27 w/o U235 + 3.0, 4.5, 5.5 OR 7.0 w/o Gd203 I LL2* Rods ( 5) ---

2.27 w/o U235 + 3.0, 4.5, 5.5 OR 7.0 w/o Gd203 L

W Rods ( 5) ---

Inert Water Rod l FIGURE 3.2 GRAND GULF UNIT 1 CYS ANF-1.4 i ENRICHMENT DISTRIBUTION FOR NINE GADOLINIA RODS 1

l l

l I

ANF 90-060(NP)  ;

Page 9

• L1 ML1 M1 MH1 MH1 MH1 M1 ML1 L1 L

L

• ML1 LLl* MH1 H2 LL2*- H2 MH1 LLl* ML1

• M1 MH1 MH1 H2 H2 H2 MH1 MH1 M1

• MH1 H2 H2 LL2* W H2 H2 H2 MH1 4

• MH1 LL2* H2 W' W W H2 .LL2* MH1 MH14 H2 H2 H2 W LL2* H2 H2 MH1-

• M1 MH1 MH1 H2 H2 H2 MH1 MH1 M1

1. ML1 LLl* MH1 H2 LL2* H2 MH1 LLl* ML1 L1 ML1 M1 MH1 MH1 MH1 M1 ML1 L1 l

l

.L1 Rods ( 4) ---

2.67 w/o U235 MLI. Rods ( 8) ---

3.33 w/o U235 l MI' Rods ( 8) ---

3.66 w/o U235 l MH1 Rods (24) ---

3.98 w/o U235 1 H2 Rods (22) ---

4.73 w/o U235 L LLl* Rods ( 4) --- 2.27 w/o U235 + 4.5, 5.5 OR 7.0 w/o Gd203 l LL2* Rods ( 6) --- . 27 w/o U235 + 4.5, 5.5 OR 7.0 w/o Gd203 2

l W . Rods ( 5) ---

Inert Water Rod L

FIGURE 3.3 GRAND GULF UNIT 1 CYS ANF-1.4 ENRICHMENT DISTRIBUTION FOR TEN GADOLINIA RODS

p r-

' h'N .?' m

.l

'l ANF 90-060(NP)'

. Page 10 ' -(

i s

L-

^ f, I'

7

,t

!I .i k

1 r ,

l l

A w

f 1

p.

1

~

h I

I i

l l

1.a a a - ..a - - - - ..a.- * . - - - ~ --x a-~- a , a

'1l

-'?

.e" >4 f' t

(:

I ANF 90-060(NP)

Page 11-  ;

en I. -

).

1 1

1 e

i l

l i

l 1

l l

1 1

i h

't, l

?

s l

t

~

> )

, c ,

. yl,

.g,' '

4 I

ANF 90-060(NP)

Page 12 ,

4.0 METHODOLOGY The system k eff was calculated as follows:

As shown in Appendix A, CASMO 3G was used to calculate the peak reactivity of the racks when filled with 9G3.0 fuel. A "burnup credit" was calculated as the difference between the peak reactivity at any burnup and the reactivity of the racks l-(

filled with new fuel (9G3.0) without the gadolinia. The calculated burnup credit is O.0751, but with an additional conservatism of 0.005, CASMO-calculated burnup u

g credit is set to 0.0701. As shown Appendix C, the KENO-equivalent burnup credit i

' is set to 0.0680. .

l

• Appendix B contains KENO VA calculation results using 16 group cross sections I L i l

for racks containing gaps in the Boraflex neutron absorber panels between cells.  ;

The methods used to calculate the 95/95 upper limit on the system k-eff are detailed in Appendix B. This upper limit includes the effects of uncertainties due to tolerances on the fuel and rack components (Appendix D) and those due to:

l- calculation uncertainties (Appendix E).

This final result is 0.9452. I 1

l-I.

ANF 90-060(NP)

F ge 13 ,

5.0 REFERENCES

i

1. '

"High Density Spent Fuel Racks", AECM-85/0143, GGNS-1 Docket No. 50-416, May 6, 1985.

2. " Criticality Safety Analysis For Tne Grand Gulf Spent Fuel Storage Racks with Cycle 4 9x9-5 LTA", ANF-88-170, Advanced Nuclear Fuels Corporation, November 1988.

1

' 3. "CASMO-3: A Fuel Assembly Burnup Program (Methdology)", Studsvik/NFA-86/8, Studsvik Energiteknik AB, Nykoping, Sweden, November 1986. I 4.

" SCALE: A Modular Cod? System for Performing Standardized Computer Analyses for Ucensing Evaluation," NUREG/CR 0200.

l S. " Advanced Nuclear Fuels Methodology For Bolling Water Reactors - Benchmark Results For The CASMO-3G/MICROBURN B Calculation Methodology," XN-NF-8019(P), Volume 1, Supplement 3, Advanced Nuclear Fuels Corporation, February 1989.

6.

Baldwin, M.N., et.al., " Critical Experiments supporting Close Proximity Water Storage of Power Reactor Fuel", BAW 1484-7,' July 1979.

7. Bierman, S.R., Durst, B.M., and Clayton, E.D., " Critical Separation between Suberitical Clusters of 4.31% Enriched UO Rods in Water with Fixed Neutron Poisons", NUREG/CR-0073, May 1978.

• i 8. W. Marshall, P.D. Clemson, G. Walker, " Criticality Safety Criteria", ANS Trans, 35, 278 i (1980).

9.

" Criticality Safety Analysis of the GGNS, Unit 1 Spent Fuel Storage Racks with Gaps in the ,

Neutron Absorbing Panels", ESI Report RPAS-SR-89/007, February 1989, Attachment to AECM-89-0037.

l' 1

l o

h '

?,

l ANF 90 060(NP)-

Page A 1-'

APPENDlX A 1.

l. .

c,--r l

l r

ANF 90-060(NP)

Page A 2 l BURNUP CREDU l

A credit for fuel burnup is calculated in this Appendix.L This "Burnup Credit" is defined as the difference between the in-rack reactivities of new fuel without gadolinia and that of exposed l 9G3.0 fuel at the most reactive state in its lifetime;. This credit is used in Appendix B to convert in rack KENO calculation resuit's for new fuel (no gadolinia) to a conservative equivalent KENO result for exposed fuel at peak reactivity.

The pea'K in rack reactivity of the 9G3.0 lattice was determined using a set of in-series CASMO calculations:

1. Fuel depletion calculations were made using several coolant void fractions.(O to:

90%) using in-core geometry.

2. Partially depleted fuel at several burnups from each of the coolant void cases were

=

l then modeled in the storage racks. The storage rack model contained no xenon  ;

in the fuel, no voids in the water, and the temperature was 20*C. These in rack reactivities are shown as a function of burnup and depletion coolant -void percentage on the attached Figure A1.

The maximum reactivity for all of the. partially depleted fuel cases in the rack is 0.8577.

The CASMO-calculated reactivity for new fuel (9G3.0) without gadolinia in the racks is 0.9328. "

Therefore, the burnup credit as calculated by CASMO is 0.0751.

Cold critical core follow calculations using CASMO(5) indicate maximum uncertainties of 0.0015 to 0.0025 and no effect of burnup. However, as an added conservatism, the burnup credit was reduced by 0.005 to 0.0701.

l As stated earlier, the burnup credit is used to adjust KENO calculation results. Therefore, the 0.0701 result from CASMO was converted to an equivalent KENO result using the conservative 0.9702 factor established in Appendix C. Therefore, the KENO equivalent burnup credit is 0.0680. .

The burnup credit is conservative for allin core conditions. The conservatism in the model due to the omission of Xenon is greater than any increase in reactivity with cooling time.

l Therefore, the burnup credit will always be conservative.

l l

~ ,

1 *:

2. . t .- ..

I .

L. .

n I 1

, l i t.

l l',~ ANF 90-060(NP)' j i

l Page B-1 .l:

l l 1; ,

l i

1 1

l 1

i l

l' l l: 1 1

1' l APPENDIX B 'l l

l

l. l r

Ie I

.1 l

p.

l l

l'.

l l

I l.

l y --w,,

v+.. 1 1 ,

ANF 90-060(NP)

- Page B-2 i

KENO CALCULATIONS INCLUDING BORAFLEX GAP EFFECTS The methodology used here is similar to that used in a previous submittal'(9).  !

= KENO calculations were made for an infinite planar array of storage rack cells with seven Botaflex gap configurations. A one sided 95/95 upper limit k-eff was calculated for each -

configuration. This upper limit include; tolerance uncertaintles (Appendix D) and calculational i uncertainties (KENO and bias uncerta!ntles combined per Reference 8). It also includes the '

"bumup credit" (Appendix A) and the bias correction (Appendix E). The system k-eff was calculated using the seven responses (95/95 upper limit k-eff values) using Monte Carlo sampling

. with conservative probabilities for each of the seven configurations.

The KENO-Va calculation results presented in this Appendix are based on the conditions L ' listed in Table 2.1 except that gadolinia was omitted from the model. The KENO calculations l

were for new fuel, but as stated before, the results have been converted to 95/95 upper limits for g peak reactivity fuel, l

All possible configurations for cells with up to four gapped panels have been conservatively represented using the following seven configurations:

1. One of the four panels per cell has a six-inch gap at the vertical centerline; the other three panets have no gaps.

2. Two of the four panels have a gap at the vertical centerline; the other two panels i have no gaps. The two gapped panels are adjacent to each other.

3. Three of the four panels have a gap at the vertical centerline; the other panel has no gap.

4. All four panels per cell have a six inch gap. The vertical center of the gaps in the north, east, south, and we'st panels are displaced from the center of the panel by 36",18", 0", and -18", respectively.

,. w ,

t ANF 90-060(NP)

Page B 3 KENO CALCULATIONS INCLUDING BORAFLEX GAP EFFECTS (Cont'd) .

' 5. As for Configuration 4, except that the chplacements are 18',0",0", and -18".

6.- As for Configuration 4,'except that the dispiccoments are 18",0",0", and 0".

7. As for Configuration 4, except that all gaps are at the vertical centerline of the panel.

The probability of each configuration in the rack was calculated as follows:

All cells were assumed to have at least one panel with a six inch gap. The conservative probabilities of 1,2,3, and 4 gapped panels per cell were set at 0.22, 0.28,0.25, and 0.25, respectively. These probabilities were established by ESI in a manner consistent with that used previously (9).

For cells with at least two gapped panels, the number of " coupled" gaps for that cell is calculated using the assumption that all gaps are randomly distributed within the central 50% of the panel length. Coupled gaps are defined as those whose centers ,

are within 18" of each other. Coupled gaps are ass!cnad a k eff as if they were co-planar and all other gaps are assigned the value for an 18" separation. This is a very conservative model because the coupling effect (increased reactivity) is greatly diminished as the distance between gaps increases. Gap spacings such as 12 and 15 inches would actually have relatively small coupling effects, but in this model they are all assigned the maximum coupling effect. Separations greater than 18" may have a slightly smslier coupling effect than for 18", but we have conservatively assigned the 18" effect to all separations greater than 18". There are ten possible combinations of the numbers of coupled and uncoupled gaps per cell.

g The ten probabilities for the number of coupled gaps for 1,2,3, and 4 gapped panels per cell are in Table B1. These probabilities were calculated using 500,000 samples with the four probabilities of gapped panels per cell and with a random distribution of gaps in the central 50% of the panellength.

l

..t w .

1

~ANF 90-060(NP).

Page B-4 KENO CALCULATIONS INCLUDING BORAFLEX GAP EFFECTS (Cont'd)

L TABLE B1 L COUPLED GAP GAPPED PANEL PER CELL PROBABILITIES l'

GAPPED PANELS COUPLF.D GAPS PER CELL PROSABILITY PER CELL 0 2 3 4 ,

^

1 0.22029 2 0,15734 0.12228 '

3 0.03133 ~ 0.14852 0.07035 4 0.00096 0.09234 0.11370 0.04286 Three of the ten possibilities do not match any of the seven configurations. These three are: O coupled gaps with 2 and 3 gaps per cell and 2 coupled gaps with 3 gaps per cell. The l

first two were treated as O coupled with 4 gaps per cell (configuration 4) and the third as 2 coupled gaps with 4 gaps per cell (configuration 5). These are conservative treatments. The resulting probabilities for each of the seven configurations and the 95/95 upper limits for an infinite planar array of each cell configuration are listed in Table B2. l TABLE B2.

PROBABILITY AND RESPONSE

SUMMARY

l RESPONSE l CONFIGURATION PROBABILITY (96/96 UPPER LIMIT) 1(0/1) 0.22029 0.8948 l 2(2/2) 0.12228 0.9169 l

3(3/3) 0.07035 0.9347 4(0/4) 0.18965 0.9026 l 5 (2/4) 0.24086 0.9263 1

6 (3/4) 0.11370 0.9347 0.04286 l

7 (4/4) 0.9572 l

1 l

1 . .

y' - IJ g. L i

i ANF 90-060(NP)

Page C-1 I

?

APPENDIX C

'I t.

l l

l t

l l

l l

l t

l' l.

! 4 I'

l l

1 l

l i

1 i.

. 9 .,

i I

ANF 90-060(NP)

. Page C 2 i

CASMO-KENO COMPARISONS l .

L This Appendix contains calculation results from KENO-Va and CASMO 3G. Both codes have been extensively benchmarked elsewhere and share wide acceptance throughout the  ;

nuclear industry. This Appendix includes data and regression results to demonstrate that the '

results from the two codes are compatible and to establish the equation used to convert a CASMO calculated burnup credit (delta-k) into an equivalent KENO burnup credit.

I Based on the regression results presented, the equivalent KENO burnup credit is 1.0104 I

• 0.0207 times the CASMO burnup credit. Since the burnup credit is a negative deita-k (0.0701

!' from CASMO, Appendix A), the 95% lower limit on the regression slope will be used; the value I is 0.9702. The equivalent KENO bumup credit is 0.9702 times 0.0701 which is 0.0680.  ;

In order to establish the relationship between KENO and CASMO, the racks were modeled as an infinite array of cells containing new fuel assemblies. Each assembly was identical to the ,

9G3.0 lattice for the ANF 1.4L assembly for Cycle 5 except:

All rods contained enriched UO, over the entire 150" stack length in the KENO model. The fuel stack length wa:: infinite in the CASMO model. Cases with various '

enrichments were modeled.

All nine UO, Gd,0, rods contained the same wt.% Gd,0, for the entire stack. The Gd,0, content was varied in the CASMO and KENO models as shown in Table C1. '

The Table C1 results demonstrate that the results from the two codes are very close.

l The KENO k-eff is correlated with the CASMO k inf by the regression equation:

k-eff (KENO) = 1.0104

• k-inf (CASMO) - 0.00688 The weights of the data in this regression were proportional to the square of the reciprocal of the KENO standard deviation.

L

.e **

• i

)

1 l

I. i ANF 90-060(NP)

Page C-3 TABLE C1 CASMO-KENO COMPARISON NEW FUEL, ALL RODS SAME ENRICHMENT, IN-RACK (NO GAPS) l Enrichment CASMO KENO (Wt.% U 238) Wt.% Od,0, k-inf k off 3.60 0.0 0.9254 0.9328 2 0.0030 3.60 3.0 0.7406 0.7419 1 0.0028 3.79 0.0 0.9375 0.9314 2 0.0033 3.79 1.0 0.7828 0.7872 0.0038 3.79- 3.0 0.7566 0.7524 2 0.0037 )

4.00 0.0 0.9500  !

0.9562 m 0.0036 I 4.00 3.0 0.7695 l 0.7702

• 0.0029 i 4.50 3.0 0.8014 0.8043 0.0027 l

1

)

ANF 90-060(NP)

Page D 1 ,

i i

l APPENDlX D i

l-l 1

l 1

l-

, a ..

ANF 90-060(NP)

Page D-2 UNCERTAINTIES The effect of dimensional and material composition tolerances on the system k off was determined using CASMO-3G. The fuel was modeled without gadolinia and with all rods at the 3.79 Wt % U-235 average enrichment.

The effect of various tolerances are detailed below.

Enrichment When the nominal enrichments were increased by 0.05 Wt.% U-235, the k inf increased by 0.00306.

• Pellet Density The nominal smear density is based on a volume % dish and a pellet density K inf increased by 0.00215 with a pellet density and a volume % dish.

Pellet Dish With a volume % dish and a pellet, the k-inf increased by 0.00042.

• Pellet Diameter increasing all pellet diameters by produced a 0.00038 rise in k inf.

Botaflex Thickness A 0.007" decrease in Boraflex thickness produced a 0.00710 increase in k inf. To maintain a constant 6.2585" cell p;tch, the cell size was increased by 0.007".

i a Stainless Steel Thickness A 0.006" increase in SS thickness produced a 0.00157 increase in k inf. The cell size was decreased by 0.012"in this model.

• Cell Pitch A 0.0625" decrease in cell pitch produced a 0.00547 rise in k inf.

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

ANF 90-060(NP) f Page D3 UNCERTAINTIES (Cont'd) t a

Channel Bulae The channel was conservatively assumed to expand out to the nominal cell size; the channel inner / outer dimensions were 5.8474/6.0625" (0.1075" wall thickness).

The result was a 0.00490 rise in k-inf.

• Wt,% Gadolinia The peak k int with a lattice is 0.00099 greater than that for the lattice.

The RMS sum of the nine uncertainties due to tolerances is 0.0110. The uncertainty in the  :

bias is calculated in Appendix E; the va!ue is 0.00368. The tolerance uncertainty and the bias uncertainty are used with the KENO standard deviation to prepare the 95/95 upper limit k effs for each of the gap configurations described in Appendix B; more details on uncertainty treatment  !

are provided there.

l 1

i l

4 0- ~ $

S II i

ANF 90-060(NP) t~

Page E 1 APPENDIX E

.. ~ .

ANF 90 060(NP)

Page E 2 METHODS VAUDATION Supplemental benchmarking was performed using experimental data with boron poison sheets in arrays of bundles. The experiments selected are described in References 6 and 7. The benchmark data are in Table E1.

Using the methods in Reference 8, the weighted average k off and the standard deviation of the bias were calculated:

Weighted average k-off: 1,0035 Bias Standard Deviation: 0.00368 The weight of each k off value is proportional to the reciprocal of its variance.

r

['

I ANF-90-060(NP)

Page E 3 TABLE E1 BENCHMARK CALCULATION RESULTS KENO Va WITH 16 GROUP CROSS SECTIONS Case No. Calculated k off Reference 6 Experiments 2378 1.00395 1 0.00376 .

2384 1.00037 1 0.00306 2388 0.99886 0.00341 2420 1.00038 i 0.00367 2396 0.99443 i 0.00360 i 2402 1.00634 1 0.00283 2411 1.01223 0.00286 2407 1.00647 1 0.00332 2414 1.00967 i 0.00327 Reference 7 Experiments  !

9 1.00092 i 0.00487 {

10 1.00181 1 0.00412 l 11 0.99786 0.00413 12 0.99885 0.00487 31 1.00442 0.00421 1

i

O i*'*'

ANF 90-060(NP)

Issue Date: 6/1/90 CRmCALITY SAFETY ANALYSIS FOR THE GRAND GULF SPENT FUEL STORAGE RACKS WITH ANF 1.4 FUEL ASSEMBUES DISTRIBUTION N. L Garner (6)

L D. Gerrald Document Control (10) 9