ML19323A106

From kanterella
Jump to navigation Jump to search
Forwards Response to Questions Contained in NRC 790827 & 0921 Ltrs Re Proposed Spent Fuel Pool Mods.Requests That NRC Allow Storage of 1830 Fuel Assemblies Instead of 1760 Assemblies as Originally Requested
ML19323A106
Person / Time
Site: Calvert Cliffs  Constellation icon.png
Issue date: 04/14/1980
From: Lundvall A
BALTIMORE GAS & ELECTRIC CO.
To: Reid R
Office of Nuclear Reactor Regulation
References
NUDOCS 8004170238
Download: ML19323A106 (38)
v • d • e


Text

{{#Wiki_filter:_ _ _ _ a BALTIMORE GAS AND ELECTRIC COMPANY P. O. BOX 1475 BALTIMORE, M ARYLAN D 21203 April 14, 1980 ARTHUR E. LUNOVALL.JR. vice passiormt Su ppsv Office of Nuclear Reactor Regulation U. S. Nuclear Regulatory Commission Washington, D. C. 20555 Attn: Mr. Robert W. Reid, Chief Operating Reactors Branch #h Division of Operating Reactors

Subject:

Calvert Cliffs Nuclear Power Plant Units Nos. 1 & 2, Dockets Nos. 50-317.& 50-318 Spent Fuel Pool Modification, Sunplementary Information

References:

a) BG&E letter dated 7/3/79 from J. W. Gore, Jr., to H. R. Denton, request for amendment. b) BG&E letter dated 1/15/80 from A. E. Lundvall, Jr. to R. W. Reid, same subject. c) NRC letter dated 8/27/79 from R. W. Reid to A. E. Lundvall, Jr., same subject. d) NRC letter dated 9/21/79 from R. W. Reid to A. E. Lundvall, Jr., same subject, i Gentlemen: Reference (a) requested approval of our proposed modifications for the Calvert Cliffs Units Nos. 1 and 2 spent fuel pools to install high density poison storage racks. As a result of questions asked in reference (c), BG&E submitted a revised criticality analysis (Reference (b)). At the same time, we asked that the licenses be a= ended to allow for storage of 1760 spent fuel assemblies (total both pools). to this letter contains the ansvers to the questions in reference (c), and Attachment 2 contains the answers to the questions forwarded by another NRC letter, Reference (d). In addition, during your review of the enclosed information we request that you treat this as a change to allow storage of 1830 fuel assemblies instead of 1760 as originally requested. This should not affect the analyses previously done. In addition, because of supply problems, we i plan to use as a poison material either the Carborundum Boroncarbide, as described in reference (a), or the Brand Industry's material known as Boroflex, which vill have the same boron loading as the previously described Carborundum product. 1 Very truly ours, fY f 4 ciqu ?ydd ( l 8804170 23,9

Mr. R. W. R;id 2-April 14, 1980 y cc: J. A. Biddison, Esquire G. F. Trowbridge, Esquire Mr. E. L. Conner, Jr. O

I 4 y ATTACHIENT 1 . 7 A. CRITICALITY QUESTIONS NRC Question No. 1 Provide the change in your calculated keff for the proposed storage lattice when the fuel assembly loading is changed from.48.5 to h3.8 grams of uranium-235 per axial centimeter of fuel assembly. BG&E Response The calculated keff for the Calvert Cliffs high density fuel storage racks decreases 0.021 when the fuel assembly loading is changed from h8.5 to h3.8 grams of uranium-235 per axial centimeter of fuel assembly. NRC Question No. 2 Provide a description of the calculations you made for the above change in fuel loading and also the one for the k,ff cf the minimum lattice pitch (i.e., 9.625"). Include a statement on whether or not you generated a neutron energy spectrum for each of the above cases to get accurate, effecitve, neutron reaction cross sections for every case. BG&E Resnonse Since this set of NRC questions was received, the fuel racks have been redesigned so that the minimum lattice pitch is nov 10.0625 inches. The k rf calculations for the redesigned racks (with 10.0625 inch pitch) e vere performed with a combination of KENO-IV with 16 g oup Hansen-Roach cross sections, KENO-IV with 123 group cross sections, and HAMMER / EXTERMINATOR diffusion theory. The k rf for the reference configuration was determined with both e 16 group and 125 group KENO and the more conservative value (123 group) was chosen. Variations in keff caused by changes in pitch, poison thickness and concentration, fuel enrichment, etc., were investigated with HAMMER /EXTER-MINATOR. In all calculations, KENO and HAMMER / EXTERMINATOR, separate calculations of the neutron energy spectrum were performed for each individual case. In particular, for the HAMMER / EXTERMINATOR problems, separate calculations with blackness theory were performed for each variation in borated slab thickness, concentration, temperature and surrounding fuel. NRC Question No. 3 Since it is our understanding that the 16 group Hansen-Roach cross section set was developed for fast spectrum reactors, provide justification and verification of its use in a primarily thermal neutron energy spectrum. m

s BG&E Resnonse Although developed for fast spectrum reactors, the 16 group cross section set does have thermal groups and is capable of treating primarily thermal spectra. Its accuracy for such spectra has been checked by bench-marking against a series of five critical experiments performed at Oak Ridge la.tional Laboratory. The five critical experiments chosen had pin pitch, diameter, and fuel loadings closely matching PWR fuel assemblies. Some of the experiments contained borated poison plates similar to spent fuel racks. The benchmarking calculations were performed at Babcock and Wilcox (B&W), and a copy of the B&W report describing these calculations and the five critical experiments 1.s enclosed as Attachment A. The benchmarking calculations show that KENO vith 16 group cross sections consistently predicted keft values about 2.4 percent higher than the experimental values. KENO vith 16 groups was then used in the design of spent fuel stroage racks. The B&W report states: "This cross section set has been highly qualified and is adequate for this system." Since this set of NRC questions was received, the criticality calculations for the Calvert Cliffs racks have been redone using 123 group cross sections. This group of cross sections gave higher keff values (about 1.4 percent higher than 16 group), and the higher value was used as the reference k,ff value for conservatism. NRC Question No. 4 Provide a proof that the effective boron region cross sections you used in the KENO-IV verification calculation are accurate. f BG&E Restonse KENO-IV vith the 123 group cross section set was benchmarked by NES against critical experiments performed at Battelle-Pacific Northwest i Laboratories with good agreement. Some of these critical experiments contained borated slabs such as are used in spent fuel storage racks. The exteriments are described in NUREG/CR-0073-RC. KENO-IV with the 16 group cross section set was benchmarked by; NES against critical experiments performed at Oak Ridge National Laboratory with good results. Some of these experiments also contained borated slabs. Eee the answer to Question No. 3 for more information on benchmarking 16 group cross sections. NRC Question No. 5 Provide a copy of NES 81A0567, " Nuclear Design Analysis Report for the Calvert Cliffs Unit No.1 Nuclear Plant High Density Spent Fuel Storage Racks," dated March, 1979, which is referenced in your submittal.

. BGLE Resnonse Revision 2 of NES 81A0567 (dated December, 1979) has been submitted to NRC by BG&E letter of January 15, 1980 from A. E. Lundvall, Jr., to R. W. Reid. NRC Question No. 6 The NRC requires verification that the amount of boron used for the - calculations is actually in all of the storage containers which are put into the spent fuel pool. Part of this required verification is an on-site neutron attenuation test for missing plates. Provide a description of the way you vill verify that the amount of boron you used in your calculations is actually in the storage containers. Include a description of the on-site neutron attenuation test that you will perform. BG&E Resnonse The neutron attenuation tests vill be performed after the fuel racks are installed into the pool. A test fixture containing a neutron source and suitably shielded detectors will be lowered into each fuel storage location in each rack, one cell at a time. The backscattered neu-tron flux vill be measured to confirm the existence of a neutron poison material. NRC Question No. 7 The NRC requires the licensee to have a method to verify that the amount of boron used for the calculations remains in the plates throughout the life of the racks. Provide a description of this verification method. i BG&E Response Verification vill be accomplished by placing samples in the high gamma areas of the spent fuel pool and then periodically removing them throughout the life of the fuel racks for various mechanical tests. I ? ~ i

.r ~, .p.. 0 } ATTACmfEdT A l lpSAFETY CALCULATIONS dND l m_ a _m _a l UBENCHM'ARKING OF BABCOCK l-AND WILCOX DESIGNED.CLOSE 2 " U '; g"' d d :: " ' SPACED FUEL STORAGE RACKS

  • .~3;fW"""5"J~"a tions l

WILLIAM D.' BROMLEY* and JAMES S. OLSZEWSKI... Babcock and Wilcox Company, P.O. Box 1260 Lynchburg, Virginia 24505 Received January 17,1978 Accepted for Publication August 8,1978 MUN 6-17"ITyb.& MMWwW Wf hsEvaRfr$4 MgheMM@ e structure. The' KENO calculations were then bench-Responding to the existing industry concern over spent fuel storage. capacity, fuel storage racks have marked against crithnl experiment data. been redesigned to give closer spaced stcrage capabil-Systematic errors caused by biases in the calcula-identified by comparing. ities. In this redesign, the Babcock and Wilcox (B&W) tional techniq'ues were objective was to offer an optimum balance between (benchmarking) calculated values with experimental maximum storage and minimum cost, with 'an ade-values. Random errors caused by statistical errors in quare safety margin. Design conditions were: Monte Carlo calculations, and material and fabrica-tion tolerances were combined before being added to-

1. B&W Mark "B" and Mark "C" pressuri:ed applicable systematic errors.

Final results indicate that with a lattice spacing of water reactorfuelassemblies 33.02 cm (13 in.) and a 0.340-cm (0.134-in.)-thick

2. UOz at 3.S wt7o enriched or MO at 4.5 wt%

unborated stainless-steel can, all nuclear safety, 2 thermalhydraulic, and seismic design requirements enriched are met. litis basic design is a conservative approach to the fuel storage problem that utilizes current

3. k,tt 4 0.95..

technology, and offers an optimum balance of the To achieve good geometric representations of the design variables at the minimum cost with an ade-quate safety margin. For some plants, where variables fuel assemblies and storage racks, KENO Monte Carlo such as fuel enrichment are different,. closer spaced computer. codes were used. (Parametric studies were made of kett versus can spacing, thicknen, and baron racks using nonpoisoned stainless-steel cans can be concentration for an infinite storage pool.) The UOz utili:ed. The same basic design can also be taed to retro-fit operating plants and increase storage capabil-studies used KENO 11 with a 16-energy-group neutron structure and the standard Hansen-Roach library. The ity from the conventional if-) cores (1 yr discha MO studies used KENO IV with a 123-energy-group over 3T cores (8-yr discharge. 2 ~ X- ??E151EhdhcSkutwin%OMM ;-QLv w ~ INTRODUCTION placed a greater emphasis on alternate approaches to storage and disposal of spent fuel. Until recently, Spent fuel storage is an item of increasing indus-these approaches have been considered only tem-try concern due to the present and foreseeable future parary with the anticipation of fuel reprocessing. lack of operating fuel reprocessing facilities. The In response to the existingindustry concern over fuel licensing process and the political situation have de-storage capacity, fuel storage racks have, been re- !ayed the startup of reprocessing plants. This has designed"' to give a closer spaced lattice. These <= h fuel storage racks are intended to provide a safe, "Present address: Bechtel ' Corporation,- P.O. Box 3965, San ' effective means for temporary underwater storage of new fuel assemb. lies and for short-or long-term Francisco, California 94119. NUCLEAR TECHNOLOGY VO!_41 MID DEC.1978 0029 5450/78/0015 0341502.00/0 o1978 ANS 341

\\- .Bromley and Olszewski SAFETY OF FUEL' STORAGE RACKS i s i. storage cf irraditted assemblies. The new designs in-racks are also design-d so that fuel assemblies can be Ol5 crease present storage capability by two to four' inserted only into the stainless-steel cans. ' ME: times, depending on the design limitations of the Ti.55 individual plant site and the initial enrichment of the 7 CONSTRtJCTION fuel. The individual fuel cells are constructed by but-DESCRIPTION ting two "C" channels together to form a box assem-i bly (see Fig. 2): A 10-gauge plate is positioned along ' De Babcocit.and Wilcox (B&W) fuel storage the butt joint in the interior of the b'ox cell and racks are a lattice network of full-length 10-gauge.- intermittently plug welded to the box. This welded 0.340-cm-thick (0.134-in.) unborated stainless-steel plate ensures cell mechanical integrity, increases cans positioned on a 33.02-cm (13-in.) square pitch. nuclear fuel isolation, and provides square interior - This pitch is possible considering 3.5 wt% UO fuel corners. The square interior corners are compatible 2 and 4.5 wt% MOs fuel. Such a network is shown de-with the square spacer grid corners of the fuel tailed in Fig.1. The can is fabricated using two 10-assembly and protect the spacer grids during handling gauge stainless-steel doubler plates plug welded to operations as well as seisniic events. opposite internal sides of the can. The racks are A " funnel" top and the clearance between the -designed with the doubler plates of the cans oriented fuel assembly and the can ensure compatibility with~ in the same direction throughout the array. These the fuel handling equipment even with distorted fuel I .7 SPACES @ 0.330 m. : (13 in.) : O.229 m ++ <r 0.'330-0.002 m SPACING .. = L BETWEEN RACKS - :rr. A I ,1 L_ 0 I 4 p F k. .) S, ~.[ I-M i i{ !v

f

.i ' 5 ., ) ,y [y,)

t...

.I - a. z.

c. - t u..

-.f' f'*f i 3: ?L

k..

.i .? ?.1 y .$,f; e n.- ..), r:. a.- ,.y ud y. 1; G. 4.209 m MA?C I.h h N d I b Y ,:/ I O t .h V ~ [h ',j ~ - y _, yJ; y p, ?.,

.m p

G r DOUBLE PLATE .9

.4

.i

. (.3r.'

3: .% f 4- "j; g. .s e. r

v.

4 c s- !.<y, ' l l M / FUEL ASSEMBLY CELL ry \\ ', F 3 -.E hrw-J.%,bp::._:.:%=1 o Y 0.048 0.003 m _.3 N~

  • FUEL ASSEMBLY SUPPORT PLATE Fig.1. 6 X8 spent fuel storage racks.

342 l NUCLEAR TECl!NOLOGY VOL41 !.!!D.DEC.1978 - w

i Bromley and Olszewski i SAFETY OF FUEL STORAGE RICKS - [ FUEL ASSEMBLY d / g n /EMNsN-( / !?EE N

y

=2.g; ~ ( l

...i h

-l ' 'l'; CELL. CORNER k COMPATIBLE WITH SQUARE FA Y-- GAP = 0.635 cm .l.- (0.25 in.) -l.

    • FUNNEL" LIPS

.-{ L / \\ \\... : - / \\ ..x. (;(h;::u:?". .. :.:: :, ;: : ::...ll i ..-v .;. t : -:... ~.. : - , 33.02 cm (13 In-)l.. '. ; :.': l*.;.. ' Y. %:G,:{w::l..... '. :...~ f.n. x . / . -: - y '/ m.Q \\ g"EAM

  • f.Y

} [ PLUG.WELOS' 22.86 cm' SUPPORT

6 (9 in.)

[ OPENING IN ~~~ A i BOTTOM-SUPPORT PLATE 12.7,0 cm (5 in.)I 10 GAUGE ..,1, l, + ~ 22.86 cm v t / Nyj{.[/ e T '10 GAUGE Fig. 2. Fuel storage cell structures. assemblies. Individual cells are welded to supporting 4.The pool was assumed flooded with nonbo-beams at the top and bottom to form standard 6 X 8 rated water at the most reactive temperature. modules. The rnodule array size may vary, depending 5.He neutron poisoning effect of the stainless-on specific fuel pool requirements, such as the num-steel cans was accounted for, but no credit was ber of failed fuellocations desired. Type 304 stainless taken for the double plates nor the structural steel is used throughout to ensure a 40-yr life in material that supports the cans. borated water and to be compatible with the stain-less-steel poolliner material. 6.No poison credit was taken for control com-ponents (centrol rods or lumped burnable Poison rods) that may be in the fuel assemblies. S ASIS i:0 R THE CALCULATIONS Tne new rack was designed to keep the calculated ker < 0.95. A computer model of the rack system CRITICALITY DESIGN TECHNIQUES represented the following conditions: The criticality analysis of the pool is the deter-1.Tne pressurized water reactor (PWR) fuel mining factor on the closeness of the fuel assembly assemblies used were the B&W Mark "B" and spacing.- The Monte Carlo computer code KENO Mark "C" assemblies. (Refs. 5 and 6) was esed to perform the k r 21cula-d c t

s. The KENO code permits an accurate descrip-
2. Both UO (3.5 wt% 238U) and mixed-oxide tion of complex geometnes. To achieve a good (4.5 wtG plutonium) fuel were analyzed.

mathematical representation of the storage array, the ~.. _ 3.The fuel storage array was modeled to be fuel assemblies anli cans, including fuel, gap, clad-infinite in both horizontal directions. ding, and moderator,.were described exactly. The 343 F1.CI. EAR TECHNot.OGY VOL41 MID-DEC.1978 4

Btemley and 01sanwski SAFETY OF FUEI' STORAGE RACKSE ';\\ 2. fuel assembly sp;cer grids were homogenized with 1.10 $5 the water, and the doubler plates were ignored. A 16-energy-group structure was used with the =- .t.= Knight-modified Hansen-Roach 7 cross-section set to perform all of the UOa calculations. His cross-section set has been highly qualified and is adequate'for this 1.00 0.003.m system. However, these cross sections were not ade-CANS quate for performing mixed-oxide studies without ex-J NO cab. tensive review on the correct application of the resonance 1 potential scattering. Therefore, a 123-0.006-m energy. group cross-section' set, developed from the 0.005-m CANS AMPX library with the code NITAWL, wat u. sed in CANS 8 all mixed-oxide work. ~ ' O.003-m CANS i Output from KENO includes the calculated value 1 wt% BORON of herr for the stated conditions and an estimate ofits 0.80 standard deviation. The standard devia' tion calculated 27. gg1 n.) (12 in.) (1 in ) (14 n.) (15 i ) (16 n.) is a measure of precision rather than. accuracy and is FUEL ASSEMBLY SPACING (em) Inherent in Monte Carlo techniques. Fig.3; Calculated kerr versus spacing. PARAU.ETRIC STUDIES The major portion of the criticality calculational effort was a series of parametric analyses. The parameters investigated that affect the kerr of the system include: 9-

1. the assembly pitch or center-to-center spacing 2.the thickness of the stainless-steel can around the fuel assembly [0 to 0.635 cm (0.25 in.)]
3. the use of boron in the stainless-steel can (0 to 2 wt% of the stainless steel)
4. the gap distance between the assembly and the can [0 to 0.635 cm (0.25 in.)]

5. possible movement of the assemblies within ~

6. the effect of moderator temperature on kerr.

Results of the parametric studies showing KENO-calculated kerr versus fuel assembly spacing and can i thickness are shown in Fig.3 and Table I.The signiti-cant effect of placing boron in the stainless steel can Fig.4.. Crowded assernbly configuration. be noted. Vadations in the gap between the fuel assembly and the can were also studied. Tnere was no chance was determined for an infinite um. form array. Results - in the calculated k rr for a variation in gap of up to ese calculations show that 20 C (68 F) is the e 0'635 cm (0.25 in')'

  • ** reactive case. This was expected, since the sys-tem is undermoderated at 33.02 cm (13-in.) spacing.

Each fuel assembly is designed to have an ~0.635-cm (0.25-in.) peripheral clearance from the can. The case in which every fuel assembly happens BENCHMARKING - i to lie so that it is as close as possible to the center of the array was investigated (see Fig. 4). The kerr ob-Both fuel systems were benchmarked versus { tained from this array was not statistically different available exNrimental data.The UO, fuel system was from the value of the corresponding uniform array. benchmarked using unpublished data from five crit-The dependence of k rr on moderator tempera-ical experiments performed at the Oak Ridge Na-e ture between 20,65.6, and 93*C (6a,150, and 200*F) tional Laboratory.' Experiments were chosen whose 344 NUCLEAR TECHNOLOGY VOL il MID.DEC.1978

s Bromley and Olszewski SAFETY OF FUEL STORAGE RACKS TABl.E1 h.. ' KENO Fuel Stora'ge Calculatio s* ' s:Ei (3.5 wt% UO,687, I atm) (. 2 w Stainless-Steel korg Ie for Center.to-Center Distance Between Assemblies Boron Stainless. Steel Concentration 25.40 cm 27.94 cm 30.48 cm 33.02 cm 35.56 cm 39.37 cm Can Thickness -(wt%) (10 in.) (11 in.) (12 in.) (13 in.) - (14 in.) (15.5 in.) ' ' = =

  • 1.091 1.010 0.973 0.923 -

0.898' O ~ 0.010 10.009 0.008 0.004 20.008 1 0.340 cm 0 1.254 1.104 0.996 0.942 0.909 - 0.878 (0.134 in.) 0.007 20.009 10.007 10.004 0.004 20.005 ( 3*"I') 0.5 1.070 0.947 0.886 ~ 20.010, 0.008 20.004 1.0 .1.028 0.910 0.842 0.820 20.008 20.004 0.009 t'O.011 2.0 0.821 10.009 0.455 cm 0 0.991 0.922 0.859 (0.179 in.) 10.010 ., 20.012 0.010 (7 88"8') 0.5 1.070 0.935 0.852 iO.006 0.010 20.008 1.0 1.021 0.906 0.834 'fg' 10.010 20.010 0.008 0.635 cm 0 0.983 0'.916 0.852 (0.250 in.) 0.004 . 20.008 20.010 ( E*"I') 0.5 1.077 0.944 0.860 20.011 !0.008 20.008 i 1.0 0.910 0.841 0.804 0.007 20.007 20.011 i 2.0 0.805 20.011 0.935 cm 0 0.965 (0.375 in.) 20.011

  • Errors are one standard deviation. Smaller deviations result from weighing the standard deviations of several : ens.
  • Spacing is one element completely surrounded by water and fully moderated.

pin pitch, diameter, and fuelloading closely matched The mixed-oxide cases were benchmarked against the PWR assemblies used in the calculations.The vari-the Saxton28 and Westinghouse Reactor Evaluation -ables investigated in modeling the U.0 experiments Center" (WREC) critical experiments. The Saxton 2 included: and WREC critical experiments considered the effects f varying the following factors:

1. number of fuel rods.
2. use of a depleted uranium metallilock reflector
1. pin pitch "
3. use of a stainless-steel plate placed through the
2. number of fuel rods fuel pins c.m.:
3. array c~onfiguration
4. use of a Boral plate (boron. impregnated alumi-l num) placed through the fuel pins.
4. '"Pu content of the fuel rods.

NUCLEAR TECHN01.OGY vo[.41 MID DEC.1978 - 345 I.

^ g t 'Bromley and Olsaewski SAFETY OF FUEL STORAGE RACKS i 5 Results of the benchmarking (shown in Tables Il resu!Es of Ehe studies show that a small bias must be and III) indicate that for UO2 fuel, the calculated a'dded to tlie ker calculated for mixed-oxide fuel n neutron multiplication was overpredicted for the systems, but a somewhat larger bias may be sub-(d)... critical experiments. For the mixed-oxide cases, a tracted from the array k r when considering the d '~ statistical approach indicates that at the 95% confi-UO. systems. He fuel rack design takes no credit for 2 dence level a k rr of 211ghtly less than 1.0 could be this conservative errorin the UO system. e 2 calculated for a critical system; therefore, a small . Note that moderator temperature is also a source' bias must be added to the MO: effective neutron of systematic shift in ker. However, all paramet-multipiications. ric analyses were conservatively performed at 20*C.' (68*F). Tw types f random uncertainty are considered. UNCERTAINTY AND PROP 0GATION OF UNCERTAINTY R ese are: ~ KENO is a statistical code, thus requiring a prop-

1. the statistical uncertainty that is inherent in a agation of uncertainty analysis. He first errors that Monte Carlo calculation must be accounted for are the systematic errors intro-duced by bias in the calculational technique.-Tnese
2. errors or possible increases in.h r due to d

were considered by the benchmarking studies. The material variations an'd manufacturing toler-ances. Errors of the first type can,in general, be reduced TABLE II to a relatively small value (<<l%) by increasing the KENO Results for UO Benchmarking Experimenti number of neutrons to be tracked. Errors of the 2 'second type that were investigated include: Number Pitch KENO

1. Van. tions in the gap between the fuel assem-a.

Lattice Description ofPins (cm) k.fr la bly and can. Within the statistics, this was Clean 272 1.530 1.016 2 0.005 found to have no effect. Clean 203 2.050 1.026 2 0.006

2. Van. tions m the position of the fuel assem-a.

E Intemal water gap 484 2.050 1.021 2 0.006 Intemal stainless steel bhes within the cans. Tnis was also found to poison plate 265 2.050 1.033 t 0.004 have no effect. Internal Boral

3. Van. t.a tens tn the density of the maten. ls used.

poison plate 268 2.050 1.025 ! O.003 a Tnis was accounted for by assuming a maxi-Avera'ge 1.024 2 0.003 mum theoretical density in the fuel. Average with 95% confidence 1.024 2 0.006

4. Variations in assembly center-to-centerspacing Conservatism = 0.024 -0.006 due to manufacturing tolerances [c0.16 cm

= 0.018 (0.0625 in.)]. TABLE II: KENO Results for Mixed-Oxide Benchmarking. Experiments Number of Pitch HO 2 wt%

  • Pu Experiment 2

Fuel Pins (cm) Temperature (*C) in Plutonium k rr: la Number

  • e 506 0.132 25.80 8.57 0.9974 2 0.00SI S

169 0.187 24.10 8.57 1.0042 : 0.0108 S 121 2.642 19.90 8.57 1.0157 0.0037 S 160 2.479 3.98 7.65 0.9957 0.0094 W 247 2.479 3.98 23.50 1.0174 'O.0085 W Average 1.0061 2 0.0041 Average with 95% confidence 1.0051 1 0.00S7 E :. Lowest possible value for k rr= 1.0051 - 0.0087 = 0.9974 e Correction for calculated values = 1 - 0.9974 = 0.0026

  • W stands for Westinghouse, S for Saxton critical PuOrUO experiments.

3 346 NUCLEAR TECHNOL.OGY VOL41 MID.DEC.1978 t

s ~ Br:mley and Olszewski{ SAFETY OF FUEL STORAGE RACKS. 1 S. Variations in can thickness due to manufac-TABLE IV js turing tolerances [H).015 cm (6 mils)). 8 Uncertainties in k,rr jim 7 In this study, the worst cor.dition is assumed to 33.02.cm "..Voccur for each of these effects for every assembly. (13.in.) Spacing. This is an extremely conservative assumption, since 0.134.In. Can the actual enors have a distribution about some mean Source ofUncertainty with No Boron

  • near zero and certainty less than the tolerance limit.

~The assembly spacing error in particular cannot be Calculations 0.0067 accumulative and must be negative as often as it is C*P width, ^ ,' positive. However, the calculations were done in a Cm ess 0 0032 more conservative manner by treat, g the mduced m Assembly spacing 0.0023 increases in k rr as standard deviations. A total stan-Total (square root of the sum e dard deviation caused by random errors was obtained of the squares) 0.0078 by adding these values to the calculational standard deviation by the root-sum-square method. The resul-

  • At 1.Nn. center.to-center spacing with a 0.134-in. can, the k rt equals that calculated by KENO plus the total uncer.

tant standard deviation at the 95% confidence level o tainty = 0.9420 + 0.0078 = 0.9498. was then obtam.ed by multiplying by the appropriate statistical number dictated by'the number of neu- ~ trons tracked in KENO. These results are reported in Table IV. The total k rt reported for this anange. THERMAL..HYDRAUI.!C DESIGN e ment (0.9498) is that calculated by KENO plus the resultant standard deviation.There is 95% confidence Physically, the racks are elevated above the pol-floor. The pool water flows by natural circuladon that'the actual value of k rr will be less than the value reported. through an operiing in the bottom of each can and up e through the can and fuel assembly. Natural circula-tion pressure loss calculations have been performed ACCIDENT ANAL.YSIS to verify that sufficient natural circulation exists in the can to prevent bulk boiling within the can for the "Eih. The postulated accident of a fuel assembly drop-maximum predicted bulk pool water temperature. (' ' as the "T-bone" accident) was investigated in depth, assumptions: ping across the top of the fuel storage array (known These calculations -were based on the following In a KENO model, the horizontal fuel asseinbly was 1.The maximum pool temperature is 96*C placed m the most reactive configuration (i.e., so as to completely cover as many vertical assemblies as (20S*F)* possible. Results show that the k rr calculated for the

2. The average pool depth is 12.192 m (40 ft).

e dropped assembly 0 cm above the array (0.949 0.010) is well within the statistical uncertainty of the 3.The analysis was performed for a can pitch calculated kert for any array without the fallen fuel f both 31.75-cm (12.5-m.) and 33.02-cm assembly (0.947 0.011). ( m.) centers.

4. Water exiting tlie hottest fuel assembly is at the saturation point; the pressure at the can exit is SEISMIC DESIGN 172.69 kPa (25 psia).

~ S. Decay heating rates for the hot assembly The se.ismic analysis was performed on the as-sembled lattice contammg a full loadm, g of fuel were calculated using ANSI Standard ANS-5.1 N18*6* assemblies. Both static and seismic loads are trans-mitted to the pool floor and to the walls by the rac.k system. The seismic design basis includes two strin-CONCLUSIONS. gent requirements for maintenance of fuel assembly integrity during a seismic event. The first requires We conclude that with a lattice spacing of 33.02 that stored fuel be reusable in a reactor after a one-em (13 in.) and a 10-gauge (0.34036-cm (0.134-in.)} half Safe Shutdown Earthquake (SSE). The second can of unborated stainless steel, all criticality safety, requirement is that the stored fuel assembly maintain thermal. hydraulic, and* seismic design requirements its structural integrity and be readily removable from are met. This basic design is a conservative ap- .... the can following a full SSE. The can design rigidity proach to the fuel storage problems. It utilizes , is ~ and construction considers the impacting loads during current technology 2nd offers an optimum balance of the seismic events analyzed, and the proposed design the design variables at the minimum cost.The design meets the above two requirements, presented is described in the B&W Standard Plant NUCLEAR TECHNOLOGY VOL41 MID.DEC.1978 .347

~. _, ...e Brrmley and 01saewski SAFETY OF FUEL' STORAGE RACKS Safety Analysis Report (B-SAR-205) as being stan-

2. E. A' GRIMM, L L ZAHN, and R. C. KARZMAR' i

Mi dard for future B&W plants. ' Die same basic design "High. Density Fuel Storage for Boiling Water Reactors," l l4 can also be used to retrofit operating plants and. Trans. Am. Nuel Soc.,26,256 (j977). increase storage capability from the conventional

3. E. M. SPIER. D. J. KAMINSK!, C. J. RAWLEY, and lj cores (1-yr discharge plus a full-core emergency J. A. NOWICKI," Westinghouse High Density Spent Fuel Rack unloading) capability to over 3y cores (S-yr dis.

Design," Trant. Am. Nuct Soc.,26,257 (1977). charge). l, For some plants. where variables such as fuel

4. R. S. HARDING R. J. KLOTZ, L C. NODERER, and !'

ennchment are different, closer spaced racks using J. E. ROSENTHAL, " Analytical Techniques Employed in nonpoisoned stainless steel cans can be utilized (e.g., - Reactor Fuel Handling and Storage Related Nuclear Criticality r 31.75.cm (12.5-m.) spacing for a maximum of 3.3 Analysis,* Trans. Am. #uel Soc,23,583 (1976). wt% UOa enrichmentl. - As discussed previously, cell spacing can be con-

5. G. E. WHITESIDES and N. F. CROSS, " KENO-A' siderably reduced by using a neutron-poisoned rack Multigroup Monte Carlo Criticality Program" CTC.5, Union material. Industry has inves igated the use of borated Carbide Corporation Nuclear Division (Sep.1969).

l 2 t ' stainless steel and Boral in a poisoned. rack design. - Borated stainless steel may be utilized but is eco-

6. L M. PETRIE and' N. F. CROSS, " KENO IV-An !

nomically unattractive. Current technical progress in. Improved Monte Carlo Criticality Program," ORNIA938, Oak l the manufacturing of Boral, coupled with its more Ridge National Laboratory (Nov.1975); attractive economics, has shown it to be an accept-able material if a tighter spaced lattice (<27.94 cm

7. G. E. HANSEN and W. H. ROACH, "6-and.16. Group t

(11 in.)] is desired. Tne same codes, benchmarking, Cross Sections for Fast and Intermediate Critical Assemblies," and analysis techniques used in this study would LAMS-2543, Los Alamos Scientific taboratory (1961). apply.when designing a fuel storage rack utilizing a poisoned can.

8. N. M. GREENE et al., "AMPX: A Modular Code Benchmarking has been done against critical System for Generating Coupled Multigroup Neutron.-Camma :

I experiments that have not dealt with the geometry of Ubraries from ENDF/B," ORNL/TM 3706, Oak Ridge Nation- . g* a spent fuel storage p'ool. Currently, B&W is perform-al Laboratory (Mar.1976). j T. ing several critical experiments under contract to the U.S. Department of Energy (DOE) for the array

9. E. B. JOHNSON, Oak Ridge National Laboratory, Private geometry of such a pool. Progres$ reports are being Communication.

issued quarterly during the measurements phase of, the contract and are numbered sequentially beginning

10. E. G. TAYLOR, "Saxton Plutonium Program Critical with the designation BAW-1484. (They are avail--

4through the DOE Technical information Center.) _ Experiments for 3385 54, Westinghouse Electric Corporation (Dec.1965). 4 Tbs.gm, wy,cuiet wim worx recesp" should Battelle-racitic norw% Law.wues,

11. R. D. LEAMER et al., " Westinghouse Reactor Evaluation '

allow some of the conservattsm mherent m, these Center PuO.UO Fueled Critical Experiments," WCAP.3726 2 types of calculations to be relaxed in future designs. 1 Westinghouse Electric Corporation (July 1967). REFERENCES

12. S. R. BIERMAN, B. M. DURST, and E. D. CLAYTON, '

" Criticality Separation Between Suberitical Clusters of 4.29

1. H. J. RUBINSTEIN, P. M. CLARK, and J. D. GILCREST wt% *U Enriched UO: Rods in Water with Fixed Neutron

" Spent Fuel Storage for New Nuclear Pour Plants," Trens. Poisons," NUREG/CR.0073/RC, U.S. Nuclear Regulatory ; Am. NucL Soc., 26,255 (1977). ~ Commission (May 1978). 4 5 s N q $ + -e t i 348' NUCt. EAR TECHNOt.OGY. yOL 41 MIDDEC.10'M +

B. Radiation Exposure, Naste Handling, and Cask Drop NRC Question No. 1 A collective dose of 3.75 man-rem has been estimated for modification of the North pool. Provide a breakdown of this estimation in accordance with a) the number of workers involved in each phase of the operation, b) the exposure rate (Mr/hr) to the workers during each phase of the operation and c) the time the workers vill spend in this radiation field. Include in your breakdown the exposure rate from the contaminated racks and commensurate expsoure to workers during removal, decontamination, crating and shipping of the racks. BG&E Restonse Since the original submittal, BG&E has modified the license report to include poison racks for both the Unit 1 (north half) and Unit 2 (south half) pools. A collective dose of 7 5 man-t'em is now estimated, 3.75 per pool. Dose rates on the racks are expected to be 10-15 mr/hr in situ. The pool itself vill be decontaminated to less than 1.5 mr/hr before modification work begins in the pool. The removal of one rack is expected to take three men three hours working in a field conservatively estimated to be 15 mr/hr for a total of.270 man rem for the 20 racks to be removed. Modification of the poel floor (after decon of the liner plate) is conservatively estimated to take up to six men a total of 1250 man hours per pool in a 1.5 mr/hr environment for both halves of the pool for a total of 3.75 man rem in this phase. The racks will not be crated and shipped (see the response to question B.3) but vill be decontaminated on site. The radiation fields during the decontamination phase vill range up to 15 mr/hr. Estimating 20 man hours per rack in a 7.5 mr/hr (average) field decontamination vill yeild a dose of 3 man rem. Another 0.h8 me.n rem was assumed for miscellaneous handling for a total of 7.5 man rem for the modification of both halves of the spent fuel pool. NRC Question No. 2 Provide the additional occupational exposure (man-rem /yr.) that vill be received by workers in the SFP area during normal operations, including refueling, as a result of the modification.

BG&E Response As a result of the spent fuel pool modification, the increased occupational exposure received by workers in the SFP area vill be negligible. The dose to workers in the area is on the order of 10-5 mr per hour which is negligible compared to the dose normally seen in that area from other causes. Due to the nature of radioactive decay, almost all of the 10-5 mr/hr dose is from the freshest fuel. The modification only increases the number of long discharged spent fuel assemblies in the pool which contribute almost nothing to the already negligible 10-5 mr/hr. t NRC Question No. 3 Provide the estimated volume of contaminated material (e.g., spent fuel racks, seismic restraints) expected to be shipped from the plant i because of the pool modification to a licensed burial site. BG&E Restonse l The modification of the spent fuel pool vill not generate additional low-level vaste to be shipped off-site. Barnwell has informed BG&E that they will not accept spent fuel' racks for burial, therefore, the rack vill be kept on site, decontaminated, and sold as scrap. r NRC Question No. h Discuss in some detail the impact of the proposed pool modification on radioactive liquid and gaseous effluents and solid radvaste shipments from the plant. Include a discussion of the pool leak collection system and history of leakage from the pool. BG&E Response The -impact of the proposed spent fuel pool modification on radioactive liquid and gaseous effluents and solid radvaste shipments from the plant vill not be significant. The two reasons for this are; 1) almost all the radiation comes from the most recently discharged fuel, and the modification vill only increase the amounts of older fuel in the pool.

2) from this older fuel the amount of gases, the volume of additional resin and filters used is small when compared with that of the rest of the plant.

The pool leakage collection system consists of channels running under the liner plate which collect to a common drain. Normally, we see no leakage from the pool. NRC Question No. 5 Provide the estimated failed fuel fraction for each fuel cycle at Calvert Cliffs Units 1 ad 2. ~.

BG&E Restonse The below table illustrates the nunber of leaking fuel pins estimated per cycle for each plant. This estimate is based on the I-131 concentration during each cycle. Estimated Number Unit Cycle of Leaking Pins 1 1 7 2 4 3 3 2 1 3 2 2 Average: 19/5 = h pins / cycle Each cycle has between 38,192 to 37,000 fuel pins depending on the number of poison pins. This results in an average failure rate of 1 pin / 10,000 fuel pins. This can be equated to 1 - 2 failed fuel pins per discharge batch. Since there are presently five discharge batches in the pool, there are probably 5 to 10 tailed fuel pins in the Spent Fuel Pool. NRC Question No. 6 Identify any heavy load or cask drop analyses performed to date for your facility. Provide a copy of all such analyses not previously submitted to the NRC Staff. BG&E Resnonse A new cask drop analyses has been performed for the spent fuel shipping casks and the Unit No.1 spent fuel pool. No new analysis has been done for the Unit No. 2 half of the pool because we vill be putting fuel racks in the Unit No. 2 cash area. The new analysis was made for both a truck shipping cask veighing 25 tons and a rail cask veighing 100 tons. The analysis determined that the pool and supporting structure are able to safely withstand postulated cask drop of the truck shipping cask. However, the supporting structure can not withstand postulated cask drop of the rail shipping cask. The analysis. included perforation and structural response effects. Table B1 is a sumury of the cask drop analysis. NRC Question No. 7 Provide a list of all objects that are required to be moved over or near the spent fuel storage pool. For each object listed, provide its approximate weight and size, a diagram or description of the transfer path utilized, and the frequency of movement. t-

h-BG&E Response o No Item Weight 1 CE Superstand Uper 8,000 lbs Lover 13,000 lbs h Work Platform 850 lbs (Moved every 2 refueling outages) UNtr 2 2 Shipping Casks (fuel) 50,000 lbs (Moved once every 2 years) 3# Sleeving Platform l 780 lbs Jib i 800 lbs (Moved every outage) l h Vacuum Cleaner 580 lbs (Moved once a year) i 5 Misc. Small Casks i 5 - 10 ton (Moved once a year) 6 Spent Fuel Pool Gate i 3,300 lbs 7 Spent Fuel Handling Machine 800 lbs (For Mast Repairs) (Moved once every two years) g 8# Personnel Man Basket 300 lbs e (Moved once a year maximum) W 9 Fuel Transfer.mtine 1,500 lbs Jib Hoist (Has not yet been moved) 10 CEA Training Device 900 lbs (Will be moved when we rerack) 11" Pool Lights Fixtures 100 lbs (Moved once a year)

  1. Items which may be moved over fuel assemblies i

TABLE B-1 Sumary of Cask Drop Analysis Unit #1 Spent Fuel Pool Weight of Cask 50 KIPS Length of Cask 16.1 Feet Diameter of Cask 2.0 Feet Free Fall in Air 3.5 Feet t Fall in Water 39 Feet Water Density-62.h lb/ft3 l Strength of Concrete 3,000 psi Perforation Formula Used Ballistic Research Formula Structural Response Force-Time Response Analysis Minimum Perforation Thickness Required h8 Inches Slab Thickness Provided 72 Inches Actual Allovable Punching Shear Stress 303 psi 367 psi Beam Shear Stress 207 psi 367 psi f Ductility Ratio for cending 0.9 10 i Ductility Ratio for Compression 0.8 1.3 t O__

~. J i ) ATTACEMENT 2 CALVERT CLIFFS REQUEST FOR ADDITIONAL INFORMATION' NRC-Question: 1. Discuss the provisions employed to limit the :aaximum height of a fuel assembly passing over the rack assembly to 24 inches. BG&E Resnonse: l A mechanical limit switch on the spent fuel handling machine prevents the fuel assembly from being raised, while over the storage rack area, more than 2h inches above the rack. NRC Question: 2. Provide sufficient details of the rack base supporting structure, sliding surfaces, all gaps (clearance and expansion) of the rack structure, and fuel handling system. BG&E Response: The racks will sit in the spent fuel upon the 3/16 inch liner plate. When the old racks are removed, the pool floor vill be modified with shims if need be, to provide a smooth sliding surface for the rack. The gap between racks vill be two inches minimum. The basic structure of the spent fuel handling' machine is a traveling bridge which spans the spent fuel pool and moves north and south on rails. A hoist is mounted on a carriage structure which travels along the bridge in an east-west direction. The hoist supports the handling tool (operated manually) for grappling fuel assemblies. An upper limit of hoist travel is provided. Mechanical stops are also provided to limit north-south-east-and vest movement. NRC Ouestion: 3. Discuss the effects of the increased loads due to the new rack structures on the fuel pool liner and structures. BG&E Resnonse: Bechtel has reanalyzed the spent fuel pool for the installation of' poison racks using a three-dimensional plate element model of the pool.. Bechtel concluded that the pool can support the poison racks without any modification to the structure. Bechtel is currently analyzing the effect on the liner plate. If the analysis indicates a problem, the liner vill be modified. l

. i NRC Question: k. Discuss the effects of postulating inclined fuel assembly drop on top of the rack. BG&E Response: I j The structural. analysis of the s' pent fuel storage racks indicates that { maximum external kinetic energy for the inclined drop cf a fuel assembly on top of the storage rack (20.71 in-k) is less than the kinetic energy (33.17 in-k). Consequently, the effects on the storage rack 'and the pool floor liner produced by the inclined drop vill be less ses are than the j effects produced by a straight drop which are described in the response to Question 10. NRC Ouestion: i 5 Describe the provisions employed to prevent movemer i naavy objects over the spent fuel asserablies. Include a descrip+i,_ of all items which l may be moved over the spent fuel assemblies. State W aner the consequences of dropping any of these items onto the rack tre morJ 9evere than the [ fuel drop accident. ~ BG&E Response: Crane limits vill be modified at Calvert Cliffs to prevent cask movement over the new spent fuel rack area. Nothing which vill be moved over the spent fuel assemblies will, if dropped, have consequences more severe than the fuel drop accident. f For a description of all items which may be moved over the spent fuel 5 assemblies see the answer to Question B.7 in attachment 1. NRC Ouestion: r e 6. Discuss the inservice surveillance plans, if any, that you have developed to assure long-term corrosion protections for the fuel rack system in the pool environment. BG&E Resnonse: I Other than neutron absorbing material, all fuel rack components are ( fabricated of type 304L stainless steel and 17-4 PH stainless steel. The 17-4 PH stainless steel being utilized as support leg components vill 0 be heat treated at 1100 F, the surface film removed and correct heat j treatment verified by hardness testing. l The tyne 304L and 17 h PH stainless steels used in the fuel storage. racks are compatible with the storage pool environment, which is oxygen-saturated, borated water (1720 pp) and a maximum temperature of 150 F. In this pool i

3-vater environment, the corrosive deterioration of rack components is negligible relative to initial thickness. Dissimilar alloy interaction between 30hL and 17 h PH stainless steels of-the storage racks, the Zirealoy cladding of the spent fuel assemblies, and the stainless steel pool liner vill be of no significance due to the similar electrical potentials of the respective metals. Based on the above, no inservice surveillance plans are deemed necessary for the metallic rack components. An inservice surveillance program for the neutron absorbing material is discussed in a response to another NRC question. NRC Onestion: t 7. Provide discussion on the material, fabrication, installation and quality control of the spent fuel racks. Indicate whether these i requirements are in conformance with Subsection RF of the ASME Code. BG&E Resnonse: All materials, fabrication, installation, and quality control of the spent fuel racks are controlled in accordance with an effective quality assurance program meeting the requirements of 10CFR50, Appendix D and Subsection NF of the ASME Code. NRC Question: 8. Provide the damping values used in the non-linear sliding analysis sud include any justification for any values higher than those specified in the FSAR. i BG&E Resnonse: The effects of damping have not been considered in the non-linear sliding analysis. Excluding the effects of damping provides conservative analysis results becauce the portion o* the external energy that would normally be absorbed in the damping elemen is available to increase the flexural deformation and the sliding of, e fuel storage rack. P e w~.

k-NRC Question: 9. Discuss the effect of the temperature gradient across the rack structure due to differential thermal effect between a full and empty cell. BG&E Response: Differential heatihg produced by full cells being adjacent to empty cells has negligible effect on.the fuel storage racks. The rack base is totally unaffected by such differential heating since the base remains at the temperature of the pool water entering the rack. Differential heating between the full and adjacent empty storage cell will tend to produce differential axial (vertical) expansions of the fuel storage cells. The construction of the storage rack is such that*the peripheral walls of the 2x2 or 2x3 module will resist the free thermal expansion of the full storage cell. For the condition of differential heating, peripheral walls will be subjected at most to a small sheer stress of c.65 ksi. Even if it is assumed that the storage cell heating produces a thermal gradient solely across the two walls of the storage cell (inner cell wall at raised temperature and outer cell wall at pool temperature), the resulting axial stresses in the cell members are less than 6.3 ksi _with the resulting axial load easily accommodated by the storage cell welds. NRC Question: 10. For the accident fuel assembly drop condition, describe in detail the assumptions, type of analysis, ductility rati., and allowable stresses used in the analysis to insure that the acceptance criteria for this case are satisfied. provide, also, your basis for concluding that the leak tightness of the fuel pool is saintained. BG&E Response: For the accidental fuel asse=bly drop condition, 1300 pound weight (fuel assembly) was postulated to drop on the rack from a height of 24 inches above the. top of the rack. Three cases were considered: 1) a direct drop on the top of a 2x2 module, 2) a subsequent tipping of the assembly onto the surrounding storage cans, 3)-a straight ~ drop through the storage . cell with impact to.the rack base structure. Linear and non-linear _ analysis techniques using energy balance methods were used to evaluate the constructural' damage resulting from a pool assembly drop'onto the rack. The' acceptance criteria for the accidental fuel' assembly drop on the rack are 1) theresulting. impact will not adversely affect the overall structural integrity of the rack,the leak-tightness integrity of the fuel pool floor-

and liner plate and 2) the deformation'of the impacted storage cells

e l i will not adversely affect the value of keff or the ability to cool adjacent fuel elements. The results of the fuel assembly drop analysis using energy balance methods are summarized in Table 1. From Table 1 it can be seen that for a straight drop of fuel assembly on top of the storage cell, the maximum stress in the storage cell is slightly greater than the dynamic yield stress for stainless steel, tLas indicating that the storage cell and its flare will undergo local permanent deformation but the overall storage rack structure will not yield. It can also be seen that the maximum stresses in the rack base structure, rack support legs and the bearing stress on the concrete floor under leg are within the allowable values. 'The maximum punching shear stress in the liner plate is greater than the allowable shear stress value but_less than the yield stress value, thus indicating small deformation of p the liner plate under the impacted leg. The external kinetic energy of p the dropped fuel will be absorbed in the local deformation of the storage i cell flare at the top of the storgae cell, and in the small deformation ll of the liner plate under the impacted leg. However, the liner plate j will not be perforated insuring the leak-tight integrity of the fuel pool liner plate will be maintained. The free fall of a fuel assembly through the storage' cell from a height of 24 inches above the top of a storage rack and its' impact on top of the cell base plate and rack base structure was analyzed using. empirical missile-equations (the Ballistic Research !.nboratory). The results indicate that the maximum thickness of steel plate that could be perforated by such a-missile is slightly less than the thickness of the cell base plate. There-fore, during a fuel assembly drop accident of this type, the fuel assembly lower end fitting feet will perforate the cell base plate, however the lower end fitting support plate will prevent further peretration of the fuel assembly and subsequent impact to the pool floor liner plate. The kinetic energy developed during the free fall will.be absorbed by both the bending and shearing of the cell base plate. Since for this fuel assembly drop case, the external energy is absorbed-in the flexural deformation of the flexible cell base plate and rack base structure, the reaction load transmitted to the rack base structure, rack feet and pool floor is less than that for fuel. assembly drop on top'of-the storage cell..Therefore, the damage to the pool floor will be less severe for the fuel assembly drop through the storage cell than that for the fuel assembly. drop on top of the storage rack. The fuel assembly drop analyses have been performed by conservatively assuming that no . energy will be absorbed by'the fuel assembly itself. The energy absorbed L S 4 +

. in the deformation of the flexible fuel assembly will result in reducet damage to the storage rack and the pool liner plate than that predicted by the conservative analysis. It has, therefore, been concluded that neither the straight drop of the fuel assembly on top of the storage cell or.the straight drop of-the fuel assembly through the storage cell with impact on top.of the rack base structure will damage the storage rack and the pool liner plate sufficiently to adversely affect the value of keff or the leak-tight integrity of the pool. TABLE 1 RESULTS OF AN ACCIDEh"rAL FUEL ASSEMBLY DROP Straight Drop on_'/op of Storagejell Weight of Fuel Jssembly (kip) 1.380 Maximwn Drop He: ght (in) 24.0 Kinetic Energy of Drop to be Absorbed (ir.-k). 33.12 Maximum Strain in Storage 1 Cell (in/in) 0.001723 0.485 ' Maximum Cell Axial Deformation (in) 0.290 Maximum Stress in Cell (ksi) 32.74 30.0 i Maximum Transmitted Reaction Load (kips) 137.0 Maximum Stress'in Rack Base Structure -(ksi) 22.9. 30.0 3 I Maximum Stress in the Weld Between -10.73 23.3 the Beans and Suppert-Legs (ksi) 4 Maximum Stress in Rack Support 17.73 74.7 Leg (ksi) Maximum Local Bearing Stress on 1.74' 3.57 - Concrete Floor (ksi). Maximum Punching ~ Shear Stress in-24.99 16.02 the Liner Plate -(ksi) .30.0 1 -t e

_7-Maximum Drop Height (in.) 192.5 Maximum Free Fall Impact Velocity (ft/sec) 32.13 Maximum Unsupported Plate Thickness that may be perforated by missile free fall velocity, (in.) BRL Formula 0.454 0.50-Maximum External Kinetic Energy '(in.k) 265.7 Maximum Transmitted Reaction Load ~ (kips) 48.1 1. Ultimate strain for stainless steel 2. The allowable stress value represents dynamic yield stress for stainless steel. 3. Allowable stress in the weld - 1.6 x 21 x 25 = 23.3 ksi. 36

4. Buckling Stress for 17-4 PH stainless steel at design temperature.

NRC Question:_ 11. Indicate whether tilting motion of the racks and rack modules under seismic effects (OBS and SSE) has been considered in the analysis. If so, provide the factor of safety under all loading combinations. BG&E Response: The tilting motion of the fuel storage rack due to Operating Basis Earthquake (OBE) and Design Basis Earthquake (DBE) has been evaluated using energy balance method. The results of the storage rack stability analysis indicate the following: 1. The 10x10 storage rack will remain stable during the, Operating-Basis Earthquake (OBE).and Design-' Basis Earthquake- (DBE) events. During the' OBE event, the storage rack will no.t lift up. During the DBE event, .one edge of-the rack will lift up no more than 0.1437 inches. For the DBE event, the m'ximum. impact load that will be generated 2. a during recontact with the pool floor has been calculated to-be 289 kips. This i= pact load.will act as an impulse load on the pool-floor. It

should be notad that all the racks will not recontact the. pool floor

. simultaneously.

. 3. During recontact, a maximum reaction load of 60.8 kips will be developed in any single foot of the rack. fThe reaction load of 60.8 kips is smaller than the maximum reaction load of 120.0 kips resulting from dead plus live pig s DBE Geismic loadings. 4. The maximum' stress generated in the rack base structure'during recentact has been calculated to be 20.4 ksi, which is less than the allowable stress value of 24.43 ksi. The spent fuel storage rack stability analysis has been performed by con- ~ servatively assuming that no energy will be absorbed in the local deform-ation of the rack base structure, pool fl6er. liner plate and in the local crumbling of the concrete under the rack feet. It has, therefore, been concluded that the storage rack will maintain its structural integrity during liftup and recontact to the pool floor. NRC Question: 12. Provide sufficient details (discussion and sketches) regarding the mathematical models used in the seismic analysis. Indicate how the shear forces in each module is calculated and discuss the effect of sliding on the response. BG5E Response: The Calvert Cliffs Nuclear Plant Unit 1 high density spent fuel' storage racks have been designed to meet the requirements for Seismic Category I structures. Detailed linear seismic analyses have been performed to verify the adequacy of the design.to withstand the loadings ence2ntered during the severe and extreme environmental conditions of the Operating Basis and Design Basis Earthquakes. Detailed non-linear time history seismic analyses have been performed to evaluate the maximum sliding of the storage racks and to determine the maximum frictional' resistance i load transmitted by the storage racks to the pool floor liner plate ~ l during the Design Basis Earthquake. The details of the linear seismic analysis and the non-linear time history seismic analysis are given below: Linear Seismic Analysis For the horizontal and vertical seismic analyses,Jthe following mathe-matical models'were developed: l l 2 x 2 Modular Cell Unit-i -The first model, is a detailed three-dimensional finite element model of an equivalent 2.x 2 module on the storage rack _ base structure. This 2 x-2 module / rack model with_its distributed lumped masses and boundary conditions is shown in Figure _l..

. L This model is used in determining the natural frequency and seitmic response (displacement, velocity, acceleration, member forces and strestes) cf the 2 x 2 module. 10 x 10 Rack Model The second rodel consists of twenty five single mass cantiles er beams (representing twenty five 2 x 2 modules) rigidly attached to the rack base structure tnd attached to each other at the top by spacer bars. Each single mass cantilever beam has the same dynamic (frequency) characteristics as a 2 x 2 module. This model is used in calculating the maximum stresses in the rack base structure and the reaction loads and stresses in tFs rack suppvrt feet. The distributed masses corresponding to the fuel a,embly storage cells, poison elements and contained plus hydrodyns ac mass are lumped at appropriate no'dal points. The hydrodynamic mass calculations are based on recommendations given in Reference 3 and 4. The horizontal and vertical weights are distributed such tnat the resulting lumped mass multi-degree-of-freedom model best represent', the dynamic characteristics of the fuel storage rack. The seismic analyses are performed for the fully loaded 10 x 10 racks only since this lo. ding condition { results in the lower frequency, high seismic accele-ations, higher stresses and reaction loads. In order to determine the ma*.imum seismic response of the storage racks, the r'acks are conservatively assumed to be pinned (not sliding) to the pool floor at the support feet locations. The boundary conditions and lumped mass locations for the horizontal and vertical seismic-analyses are shown in Figure 2 through 4. The eigenvalues (natural frequencies) and the eigenvectors (mode shapes) for each of the natural modes of vibration are calculated by using the Lanczos Modal Extraction Methods. The Seismic Response Analyses are. performed by. the Response Spectrum Modal Superposition Methods of dynamic analysis, using the applicable Response Spectra Curves.. Individual modal response of the system are combined in accordance with Section 1.2.1 of Regulatory Guide l.92. The maximum response (deflection, acceleration, velocity, shear forces, moments, stresses, reaction loads) of the system for-each of the three orthogonal spatial components (two horizontal and one vertical) of an earthquake are combined on a' square root of_the sums of the squares (SRSS) bases (Regulatory Guide 1.92). The rattling" effects of the' fuel assembly made the storage cell'are con-servatively accounted for by imposing the.following assumptions: ~ 1.- All storage cells contain the fuel assembly. 2. All fuel assemblies simultaneously impact the storage cells. 3. The effect of fuel assembly impactEis' a two-fold increase'in the seismic, ' inertia loadings produced by the i=pacting fuel assemblies-mass. The1 use of'an impact factor of 2.0 is conservative as: verified by. corr ring-

10 - l I I the base shear response results.of the linear response spectrum modal super position analysis (Refer tnce 1) and the non-linear time history analysis (Reference 2). 4. The impact and seismic inertia loads of the impacting masses are added to the seismic inertia loads of the non-impacting masses. The linear seismic ~ analysis of the fuel storage racks are performed utilizing the STARDYNE computer code. Non-Linear Time History Sei$mic Analysis For the non-linear time history seismic analysis of tbf spent fuel assembly / storage cell structure, a 10 x 10 storage radk,ar." the stored fuel assemblies have been represented by a two dimensional lumped mass finite element model (Figure 3). The model consists basically of two coincident finite element cantilever beams, one representing the 100 storage cells and the other representing the 100 stored fuel assemblies attached to a " floor" mass by means of a non-linear sliding element. I The fuel element cantilever beam consists of masses lumped at the model nodal points interconnented by discrete beam elements. Each lumped mass represents the tributary weight of the fuel element mass. The stiffness characteristics of the beam elements are related to the effective flexural rigidity of the fuel assemblies. The storage rack cantilever beam similarly consists of lumped masses-interconnected by discrete elastic beam elements. Each lumped mass represents the tributory weight of the storage cells, water trapped inside the cells and the virtual water mass to account for the hydro-dynamic effects. The stiffness characteristics of the storage rack beam elements are related to the dynamic characteristics (fundamental fre-quency of vibration) of the storage rack. In order to account for fuel assembly impact, adjacent masses of the fuel assembly beam and the storage rack beam are. laterally couple 0 by means of non-linear spring gap elements. The non-linear spring / gap elements permit the adjacent masses to impact.each other whenever the gap closes during a seismic event. The. stiffness of_the non-linear spring is taken as-the stiffness value for each spacer grid. An initial gap of 0.21625 inch, reflecting the lateral gap between the fuel assembly and the storage cell wallLis provided. The non-linear spring /ga'p. elements are effective for fuel assembly impact on either side of the-storage cell. e

i , l The two. cantilever beams representing the storage cells and fuel assemblies are attached to the pool floor mass by means of the non-linear sliding element to best represent the rack standing freely on the pool floor. The sliding of the rack is initiated when the lateral i force in the sliding element exceeds the frictional resistance force which is equal to the coefficient of friction times the vertical weight r of the rack. The effective vertical weight is.taken as the vertical bouyant weight of the storage rack less the uplift loads due to the vertical component of the Design Basis Earthquake. The non-linear time history seismic analyses are performed by step-by-step integration techniques (Houbolt Method) using the ANSYS computer [ program NRC Question: r 13. Indicate if this proposed modification conforms with the NRC postion on fuel pool modifications entitled "OT Position for Review and Acceptance of Spent Fuel Storage and Handling Applications", issued l on April 14, 1978, and later amended in January 18, 1979. t BG&E Response: The structural / seismic design / analysis of the fuel-storage racks are performed in accordance with the guide lines given in NRC position paper "OT Position for Review and Acceptance of Spent Fuel Storage Handling. Application", issued on April 14, 1978 and its amendment of January 18, 1979. NRC Question: 14. Discuss the possibility of swelling in the cell containing the B4C composite (inward and outward) due to offgasing generating internal pressure and discuss the provisions employed to prevent such swelling or.the provision employed such that withdrawl of the fuel assembly is insured. i BG&E Response: Swelling of the fuel-storage cell walls due to the pressure buildup of trapped gasses will not occur for the following~ reasons:

a. 'Each composite material ~ compartment contains a large vent' hole (3/4 inch diameter) placed at the top of the compartment.

(The hole also ~ perndts visual verification of the presence of. Composite material within'the compartment after cell fabrication). ~ b. The Composite material compartments are intetmittently welded.along the periphery of the ecmpartment. Therefore, venting can occur-at-any location along the length of the~ storage cell. i

c. The low rate of gas ' generation combined with the venting provisions assure that a pressure buildup within the compartment due to " trapped" gas in not feasible. Furthermore, the principal gas species generated is hydrogen which is extremely difficult to contain even in sealed systems. d. The composite material has not exhibited any significant dimensional 11 Rads. ' Con-changes during exposure to gamma radiation up to 10 aequently the gas generated by the composite material does not cause the material itself to swell. l [ e F 0

NUCLEAR ENERGY SERVICES, INC. DOCUMENT NO. PAGE OF Dynamic Degrees of Freedom g Horizontal and (Typ.) J/ Vertical Masses // O 2,000.0 lbs. a a O 4,000.0 lbs. 6,250.0 lbs. O 8,000.0 lbs. A 250.0 lbs. 2 6 500.0 lbs. p (Typ.) i (Typ) 1 [n / (TYP m A V/AfA i x1 / ffAdf/ x2 ffffdf d' Vertical Supports ( A A i FIGURE 1 i LUMPED MASSES AND SOUNDARY CONDITIONS - ICAD CASES 4 AND 5 i SEISMIC ANALYSIS OF THE FULLY LCADED 2 x 2. MODULE - E AND E' i PORM s NES 205 S/78

h NUCLEAR ENERGY SERVICES, INC. DOCUMENT NO. PAGE OF ~ X3 ES _ 35 _ sg [i 3 X1 3

  • wu

,/ p_f / / f3 Glo 1 Coordinate g3 g y /A'/*/*ZA'/" A-v / /v /"/ fl j f' 44 .Ts i 26 l FIGURE 2 10 x 10 RACK FINITE ELE'ENT MODEL DIMENSIONS AND NODE NUMBERS X3 / 1 g J.s f." y y y p. g,i $, "

  • 2 <

Globa cordinate afe' 7, w f;s '7 / *'r K M W '8, f.i6 7 4 V L fg-fexqja.,ygs.ny A i a FIGURE 2 l 10 X 10 RACK FINITE ELE ENT MODEL i DIMENSIONS AND NODE NU?GERS (Cont'd) I l - FORM

  • NES 205 9/78

NUCLEAR ENERGY SERVICES, INC. DOCUMENT NO. PAGE OF 9 6 3 ,2 g 2 ad e y ai f=. i53.,4yns y m-a - ,,w,, ,,.x ,4 3/ itsp c seS 3 5 y y% X "'.; f " f E" # ' 9 ~ y y. y. 7.,e Rigid Coluras (Typ.) gren ,6v en mAai Ae, ap.in A' FIGURE 2 10 X 10 RACK FINITE ELEMENT MODEL BEAM ELEMENT NUMBERS .2 ~)* M' t,$ c.7 da, s JL s >% K K M M y' XMM K sdAr:: Yr XK>&KKV ~ u 7 4 I i [k- 'J L JL FIGURE 2 10 X 10 RACK FINITE ELEMENT MODEL EEAM ELEMENT NUMBERS (Cont'd) PORM

  • NES 205 9/78

~ NUCLEAR ENERGY SERVICES, INC. DOCUMENT NO. M M PAGE OF Sof

f 49#

,it 2. A 8i' f f* p ,4+ ,i ,i go* C' y*S iF8 jf' t ot if t,9

  • I,S'

,lI t*\\ .19 f? if is' 6 s+.

  • f" (rv!)

.X3 A h f1 g[ X g I f ((<, (1) ,/fs j g 'j ff[ L ,-f ApN,! jVi [,, 3

  1. / '271// t/ /3 k' S // * // L//~ L // l/

Mb Mb FIGURE 3 10 X 10 RACK FINITE ELEMENT MODEL DIMENSIONS AND NODE NUIGERS. - (Cont'd) FORM *NES 205 9/78

NUCLEAR ENERGY SERVICES, INC. DOCUMENT NO. PAGE OF ' sc.s tr.t n1 1r4 7 3 2 4 "[M Ne O f lfAta k 7 Ykki( ~ 3 A / 5 }1 _ 4

4 A

,g / k / /4 fE // 5 s /E /d*I /5 f/ //n f t /E' fu fx /A fe / / n / A A // t n / g g..- __.._g FIGURE 3 10 X 10 RACK FINITE ELDENT MODEL BEAM ELDENT NUMBERS (Cont'd) I FORM s NES 205 S/78

NUCLEAR ENERGY SERVICES, INC. DOCUMENT NO. PAGE OF l X$. L / ** /M/M XfMM ~ x' '- /fMMW'? M1WMe92VX6 Rigid Plates (Typ ) /.h )4 N ~ FIGURE 3 10 X 10 RACK FINITE ELEMENT MODEL PLATE ELEMENT NUMBERS X. l1 l/ </l/6/lMl&l/4/l /2'f&fSf/b/f/5/ lSf/Gff\\t/l/kT/l/2/ lWl/>"f/1'/f/&/lMl a A A FIGURE 3 10 X 10' RACK FINITE ELEMENT MODEL PLATE ELEMENT NUMBERS (Cont'd) FORM

  • NES 206 9/18

G9 DOCUMENT NO. 4Ti NUCLEAR ENERGY SERVICES. INC.. PAGE OF Dynamic Degrees of Freedom (TYP) J / s ) ( l

  • 2 f

/ ./ s' (TY )(TY gk [ Horizontal and vertical Masses gu @ 1600.0 lbs. Vertical Q 3200.0 lbs. Support Leg Locations (TYP) Q 6400.0 lbs. FIGURE 4 s O 8000.0 lbs. LUMPED-MASSES AND BOUNDARY CONDITIONS - LOAD CASES 4-& 5 SEISMIC ANALYSIS OF FULLY LOADED 10 x 10 RACK - E AND E' ,m........,,..,,.

[...e NUCLEAR ENERGY SERVICES, INC. DOCUMENT NO. i l PAGE OF Fuel Assembly Non-Linear Springs with Gaps (Typ.) i Lumped Masses (Typ.) I 11 4 i 7 j( j j( 42.125" (Typ.) Y 10 4M l -4>67 u 168.5" 1,2 9q l p5 Global Coordinate System Storage 8(M l >4 Cell Structure Support Leg Structure >3 V 3.0" 2 c A Non-Linear Sliding 1 l /[ Element Pool Floor FIGURE 5 Non-Linear Time History Seismic Analysis Lumped Mass Finite Element Non-Linear Model comu, e NES 206 9/78}}

Retrieved from "https://kanterella.com/w/index.php?title=ML19323A106&oldid=4774298"