ML20196G292

Technical Evaluation Rept,Evaluation of High Density Spent Fuel Rack Structural Analysis for St Lucie Plant - Unit 1

ML20196G292
Person / Time
Site:	Saint Lucie
Issue date:	02/29/1988
From:	Degrassi G BROOKHAVEN NATIONAL LABORATORY
To:	Office of Nuclear Reactor Regulation
Shared Package
ML17221A675	List: ML20196G292
References
CON-FIN-A-3841 NUDOCS 8803080060
Download: ML20196G292 (57)

Text

.-

APPEf0IX A TECHNICAL EVALUATION REPORT EVALUATION OF THE HIGH DENSITY SPEh7 FUEL RACK STRUCTURAL ANALYSIS FOR FLORIDA POWER AND LIGHT COMPANY ST. LUCIE PLAhT - UNIT NO. 1 By G. DeGrassi STRUCTURAL ANALYSIS DIVISION DEPARTMEhT OF NUCLEAR ENERGY BROOKHAVEN NATIONAL LABORATORY UPTON, NEW YORK February 1988 Prepared for U.S. Nucelar Regulatory Commission Office of Nuclear Reactor Regulation Fin A-3841 b h

1 Executive Summary This report describes and presents the results of the BNL technical evaluation of the structural analysis submitted by Florida Power and Light Company in support of their licensing submittal on the use of high density fuel racks at St. Lucie Unit No. 1. The review was conducted to ensure that the racks meet all structural requirements as defined in the NRC Standard Review Plan and the NRC OT Position for Review and Acceptance of Spent Fuel Pool Storage and Handling applications.

The proposed reracking of the spent fuel pool involves the installation of seventeen f ree-standing, self-supporting modules of varying sizes arranged within close proximity to each other and the pool walls. Each rack module consists of individual cells of square cross-sectica, each designed to accommodate one fuel assembly. Since the racks are neither anchored to the pool floor or walls nor connected to each other, during an earthquake, the '

racks would be free to slide and tilt. For an earthquake of sufficient intensity, the racks could impact each other and the pool walls. Because of the nonlinear nature of this design, a time history analysis was required to characterize the seismic response of the fuel racks.

The BNL review focused primarily on the seism.'c analysis of the fuel rack modules because of the complexity of the analycis method and the number of simplifying assumptions that were required in developing the dynamic models.

BNL also reviewed other analyses performed by the Licensee including fuel handling accident analyses, thermal analyses, and spent fuel pool and liner analyses.  ;

During the course of the review, a number of questions were raised regarding the adequacy of the fuel rack dynamic models. Concerns were raised that single rack models may underpredict seismic forces and displacements that would occur in the real multiple rack f uel pool environment (Section 4.1.1).

Concerns were also raised regarding the adequacy of fluid coupling assumptions used in the models (Section 4.1.2). In responce to these questions, the Licensee provided additional information and performed additional studies, including multiple fuel rack seismic analyses, to demonstrate the adequacy of j the design basis results. The additional studies indicated that the design i basis models predict conservative seismic loads and displacements. It was l noted, however, that the most significant factor contributing to the '

conservatism was the use of twice the fuel assembly weight in the design basis models. Nevertheless, the results of these studies coupled with the significant safety factors in the results provided a high level of confidence to conclude that there is sufficient conservatism in the results to compensate for analytical uncertainties.

Based on the BNL review of the Licensee's analyses, it was concluded that the proposed St. Lucie Unit 1 high density fuel racks and spent fuel pool are designed with sufficient capacity to withstand the effects of the required environmental and abnormal loads.

iii l

l l

l i

l TABLE OF CONTENTS Page '

1.0 INTRODUCTION

1.1 Purpose 7

1.2 Background

1.3 Scope of Review 1 2.0 ACCEPTANCE CRITERIA 2 3.0 FUEL RACK DESCRIPTION 3

4.0 TECHNICAL EVALUATION

4 4.1 Fuel Rack Seismic Analysis 4-4.1.1 Dynamic Model 5 4.1.2 Fluid Coupling Effects 7

4.1.3 Friction Effects 10 4.1.4 Damping 11 L

4.1.5 Seismic Loads 11 4.1.6 Load Cases 11 i 4.1.7 Analysis Method 12 4.1.8 Analysis Results 13 4.1.9 Evaluation of Results 14 4.2 Additional Fuel Rack Seismic Studies 14 f 4.2.1 Single Rack Studies 14 i 4.2.2 Multiple Rack Studies 15 l 4 . 2 . -) Overall Evaluation of Seismic Analysis Results 16 4.3 Thermal Ar.alysis 16 4.4 Fuel Handling Accident Analysis 17 4.$ Spent Fuel Ptol Analysis 19 4.5.1 Load and Load Combinations 19 4.5.2 Spent Fuel Pool Structure Analysis 20 4.5.3 Pool Liner and Anchorage analysis 20,

5.0 CONCLUSION

S 21

6.0 REFERENCES

23 l

l v

l

LIST OF TABLES TABLE TITLE PAGE 1 MODULE DATA 25 2 MODULE DIKENSIONS AND WEIGHTS 26 3 RACK MODEL PARAMETERS 27 4 RACK SEISMIC ANALYSIS RESULTS IMPACT LOADS AND STRESS FACTORS 28

, 5 RACK SEISMIC ANALYSIS RESULTS

SUMMARY

DISPLACEMENTS AND FLOOR LOADS 29 6 SUKMARY OF SAFETY FACTORS IN CRITICAL 30 FUEL RACK LOCATIONS 7 RESULTS OF SINGLE RACK STUDIES-FULLY LOADED G1 RACK WITH C0F = 0.8 31 8 RESULTS OF MULTIPLE RACK STUDIES-FULLY LOADED 1A , A2 , B ,1 B2 RACKS WITH C0F = 0.2 32 9 RESULTS OF MULTIPLE RACK STUDIES-FULLY LOADED iA , A2 , B ,1 B2 RACKS WITH

. C0F = 0.8 33 10 RESULTS OF MULTIPLE RACK STUDTES-FULLY LOADED 1A , A2 , B 1, B2 RACKS 34

11 SPENT FUEL POOL STRUCTURE MAXIMUM STRESS

SUMMARY

35 1

l l

1 vi

e ,

LIST OF FIGURES FIGURE TITLE PAGE 1 SPENT FUEL POOL LAYOUT 36 2 TYPICAL RACK ELEVATION-REGION 1 37 3 TYPICAL RACK ELEVATION-REGION 2 38 4 TYPICAL CELL ELEVATION-REGION 1 39-5 3 x 3 TYPICAL ARRAY-REGION 1 40 6 TYPICAL CELL ELEVATION-REGION 2 41 7 3 x 3 TYPICAL ARRAY-REGION 2 42 8 ADJUSTABLE SUPPORT LEG 43 9 SCHEMATIC MODEL OF FUEL RACK 44 10 FUEL RACK MODEL SHOWING RACK-TO-RACK IMPACT SPRINGS 45 l 11 IMPACT SPRING ARRANGEMENT AT NODE i 46 12 SPRING HASS SIMULATION FOR TWO-DIMENSIONAL MOTION 47 13 NORTH-SOUTH SSE 48 i 14 EAST-WEST SSE 49 15 VERTICAL SSE 50 16 SPENT FUEL POOL MAT F',AN AND SZCTION 51 17 SPENT FUEL POOL MODEL OVERALL VIEW 52 vii

e .

1.0 INTRODUCTION

1.1 Purpose This technical evaluation report (TER) describes and presents the results of the BNL review of Florida Power and Light Company's licensing submittal on the use of high density fuel racks at St. Lucie Unit No. I with respect to 4

their structural adequacy.

1.2 Background

The existing racks in the spent fuel pool have 728 total storage cells.

With the presently available storage cells, St. Lucie, Unit No. 1 lost the full-core reserve storage capacity af ter the seventh refueling which was completed in the spring of 1987. To correct this situation and provide sufficient capacity to store discharged fuel assemblies, the Licensee has requested NRC to issue a License Amendment to replace the existing storage racks with new high density speat fuel storage racks. The new racks will allow for more dense storage of spent fuel, thus enabling the existing pool to store more fuel. The new high density racks have a usable storage capacity of 1706 cells, extending the full-core reserve storage capability until the year 2009.

The proposed racks consist of individual cells of square cross-section, each of which accomodates a single PWR fuel assembly. The cells are assembled into distinct modules of varying sizes which are to be arranged within the existing spent fuel pool. Each module is free-standing and self-supporting.

The Licensee provided a summary of his safety analysis and evaluation of the proposed racks in a Safety Analysis Report (Ref. 1). The report described

' the structural analysis of the new fuel racks and the existing fuel pool. It also gave a description of postulated dropped fuel and jammed fuel accident analyses. ,

BNL reviewed the Safety Analysis Report and generated a list of additional information needed to complete the review (Ref. 2). The Licensee provided the additional information in later submittals (Ref. 3a, b, c). In addition, BNL participated in a limited audit of the fuel rack analysis and fabrication in the offices of Holtec International, the fuel rack designer, and at Joseph Oat Corporation, the fuel rack fabricator.

1.3 Scope of Review

The objective of the BNL technical review was to evaluate the adequacy of the Licensee's structural analysis and design of the proposed high density spent fuel racks and spent fuel pool. Due to the complex nature of the fuel rack seismic analysis, the primary focus of the review was on the adequacy of the non-linear fuel rack models and their dynamic analysis. The structural evaluation of fuel racks subjected to the dropped fuel and jammed fuel 1

I handling accidents described in the Licensee's report (Ref. 1) were included in this review. However, the definition of these postulateit accidents and their parameters (drop height, uplift force, etc.) were beyond the scope of this review. A limited review of the spent fuel pool was conducted to insure ,

that appropriate loads, methodology and acceptance criteria uere applied.

2.0 ACCEPTANCE CRITERIA The acceptance criteria for the evaluation of the spent fuel rack appli-cations are provided in the NRC OT Position for Review and Acceptance of Spent Fuel Storage and Handling Applications (Ref. 4). Structural requirements and criteria given in this position paper were updated and included as Appendix D to Standard Review Plan 3.8.4, "Technical Position on Spent Fuel Pool Racks,"

(Ref. 5). These documents state that the main safety function of the spent fuel pool and fuel racks is to maintain the spent fuel assemblies in a safe configuration through all environmental and abnormal loadings, such as earth-quakes, and impact due to spent fuel cask drop, drop of a spent fuel assembly, or drop of any other heavy object during routine spent fuel handling.

Section 2 of SRP 3.8.4, Appendix D gives the applicable Codes, Standards and Specifications. Construction materials should conform to Section III, Subsection NF of the ASME Codes. Design, fabrication and installation of stainless steel spent fuel racks may be performed based upon the ASME Code Subsection NF requirements for Class 3 component supports.

Requirements for seismic and impact loads are discussed in Section 3 of Appendix D. It states that seismic excitation along three orthogonal directions should be imposed simultaneously for the design of the new rack system. Submergenco in water may be taken into account. The effects of sub-mergence are considered on a case-by-case basis. Impact Loads generated by the closing of fuel assembly to fuel rack gaps during a seismic excitation ,

should be considered for local as well as overall effects. It should also be l demonstrated that the consequent loads on the fuel assemblies do not lead to fuel damage. Loads generated from other postulated events may be acceptable if suf ficient analytical parameters are provided fcr review.

Load and load combination requirements are provided in Section 4.

Specific loads and load combinations are acceptable if they are in conformance with Section 3.8.4-11.3 and Table 1, Appendix D of the Standard Review Plan. i Changes in temperature distribution should be considered in the design of the pool structure. Temperature gradients across the rack structure due to differential heating effect between a full and an empty cell should be incorp-orated in the rack design. Maximum uplift forces from the crane should be  ;

considered in the design. ,

)

i Section 5 diacusses design and analysis procedures. It states that design and analysis procedures in accordance with Section 3.8.4-11.4 of the Standard ,

Review Plan are acceptable. The effects of gaps, sloshing water, and increase  ;

of effective mass and damping due to submergence in water should be quantified. Details of the mathematical model including a description of how ,

2

the important parameters are obtained should be provided.

Structural acceptance criteria are provided in Section f The acceptance criteria are given in Table 1 of Appendix D. For impact loe ing, the ductility ratios utilized to absorb kinetic energy should be 'uantified. When considering seismic loads, factors of safety against gross sliding and over-turning of the racks shall be in accordance with Section 3.8.5-II.5 of the Standard Review Plan unless it can be shown that either (a) sliding motions are minimal, impacts between adjacent racks and ietween racks and walls are prevented and the factors of safety against tilting are met, or (b) sliding and tilting motions will be contained within geometric constraints and any impact due to the clearances is incorporated.

3.0 FUEL RACK DESCRIPTION The new high density spent fuel storage racks consist of individual cells with 8.65 inch by 8.65 inch nominal square cross-section, each of which "ccom-modates a single Combustion Engineering or Exxon PkR fuel assembly or equiva-lent, from either St. Lucie Unit 1 or Unit 2. A total of 1706 cells are arranged in 17 distinct modules of varying sizes in two regions. Region 1 10 designed for storage of new fuel assemblies with enrichments up to 4.5 weight percent U-235 or for fuel assemblies with the same maximum enrichment which have not achieved adequate burnup for Region 2. The Region 2 cells are capable of accomodating fuel assemblies with various initial enrichments which have accumulated minimum burnups within an acceptable bound as discussed in the Licensee's Safety Analysis Keport (Ref. 1). The arrangement of the rack modules in the spent fuel pool is shown in Ff 3ure 1. Typical Region 1 and Region 2 racks are shown in Figure 2 and 3. Each rack module is equipped with girdle bars at the upper end, 3/4-inch thick by 31/2 inches high. The modules make surface contact between their contiguous walls at the girdle bar locatione and thus maintain a nominal 1 1/2 inch gap between adjacent module cell walls. The modules in the two regions are of eight different types.

Tables 1 and 2 summarize the physical data for each module type.

The rack modules are fabricated from ASME SA-240-304L austenitic stainless steel sheet and plate material, and SA-351-CF3 casting material and "A 564-630 precipitation hardened stainless steel for supports. The weld fi.ler material utilized in body welds is ASME SFA-5.9, Classification ER 308L. Boraflex serves as the neutron absorber material. Boraflex is a silicone-baeed polymer containing fine particles or boron carbide in a homogeneous, stable sstrix.

Each rack module consier.s of the following components:

Internal square cube Neutron absorber eaterial (Baraflex)

Poison sheathing (Region 1 only)

Gap element (Region 1 only)

Baseplate 3

Support assembly  ;

Top lead-in (Region 1 only)

Figures 4 and 5 show a typical Region 1 cell elevation and a typical 3x3 array horizontal cross-section. Figures 6 and 7 show the same views of a typical Region 2 rack module. The figures show that the major difference between the Region 1 and Region 2 module designs is the larger pitch between cells in Region 1. Channel shaped gap elements are welded between the Region 1 cell tubes to maintain the minimum flux trap required between adjacent internal cells. Region 1 modules use poison sheathing (cover sheets) to position and retain the Boraflex absorber material around each cell wall. In Region 2 modules, the Boraflex absorber material is placed between the walls of interior cells and kept in place by stainless steel connecting strips. The Region 1 modules also provide lead-ins at the top of each cell wall to facilitate fuel assembly insertion.

The adjacent cells of each module are welded together either through gap elements (Region 1) or side connecting strips (Region 2) to form a honeycomb structure. The honeycomb is welded to a 3/4 inch thick baseplate with 3/32 inch fillet velds. The baseplate has 6-inch diameter holes concentrically located with respect to each square tube, except at support leg locations, where the hole size is 5 inches in diameter. These holes provide the primary path for coolant flow.

Each module has at least four support legs. All supports are adjustable I

in length to enable leveling of the rack. The varisble height support assembly consists of a flat-footed spindle which rides into an internally-threaded cylindrical member. The cylindrical member is attached to the under-siac of ice baseplate through fillet and partial penetration welds. Figure 8 shows a vertical cross-section of the adjustable support assembly.

The support legs will rest on 1-1/4 inch thick plates on the spent fuel pool floor. Additional plates will be provided for those areas of the pool >

floor where the rack support legs are located which do not already have ,

plates. The new plates will not be attached to the pool floor. Aside from I the addition of these plates, the Licensee has indicated that no other spent fuel pool modifications are needed to accomodate the new racks.

j

4.0 TECHNICAL EVALUATION

4.1 Fuel Rack Seismic Analysis The spent fuel storage racks are seismic Category 1 equipment required to 3 remain functional during and af ter a safe shutdown earthquake. As described <

in Section 3.0, the proposed racks consist of 17 distinct free-standing '

modules which are neither anchored to the pool floor, attached to the side walls, nor connected to each other. Any rack may be completely loaded with  ;

fuel assemblies, partially loaded, or completely empty. The fuel assemblies i are free to rattle within their storage cells.

I l

4

. o Seismic forces are transmitted to the racks through friction at the support leg to pool floor interface. If seismic displacements are large enough, the racks can impact against each other or the pool walls and the support less can lift off and impact the pool floor. Because of these non-linearities, a time history analysis of nonlinear rack models was required to characterize the seismic response of the fuel racks. BNL's review of the details of the modeling technique and analysis method is described in the following sections.

4.1.1 Dynamic Model The Licensee's mathematical model of a spent fuel racks module is shown in Figures 9, 10, 11, and 12. The Licensee indicated that the rack structure is very rigid and that its motion can be characterized in terms of six degrees of freedom at the rack base. Figure 9 shows the rack as a rigid stick on a rigid base with three translational (qi, q2 93) and three rotational (qc, q$, q6) degrees of freedom. The fuel assemblies are treated as five lumped masses located at different elevaticas (nodes 1* to 5*). All fuel assemblies in a rack module are assumed to vibrate in phase. Each lumped mass is assumed to rattle independently within the rack cell gaps and is repre-sented by two horizontal translational degrees of freedom. Impacts between i the rack and fuel assembly lumped masses are accounted for by the use of compression-only gap elements as shown in Figure 11. The support legs are modeled as compression-only springs (S1 to S4 in Figure 9) which consider the local vertical flexibility of the rack-support interface. Friction elements are used at the bottom of the support legs. Figure 10 show che impact springs acting through gap elements to simulate the interfe e with adjacent rack modules or pool valls. Five impact springs per side at t used at both the girdle bar and baseplate elevations. Figure 12 shows a two dimensional repre-sentation of tha model with only one "rattling" tuel mass to clarify the j overall model cor:ept.

Fluid coupling betsaen rack and fuel asemblies, a.d between rack and r

{

adjacent racks or walls is simulated by including inettial cotpling terms in I the equations of motion. This is discussed in detail below. Fluid damping between rack and fuel assemblies, and between rack and adjactat racks is neglected in the model.

In order to simulate the motion of adjacent fuel racks, tha model assumes a symmetry plane midway between adjacent racks. Thus the at/.el assumes that each adjacent rack moves completely out of phase with the t ck being ,

analyzed. This assumption is intended to predict consero.rtve rack to rack I impact forces.

To complete the review of the adequacy of the model, the Licensee was requested to provide typical fuel rack and tuel assembly design drawings and a list of key modeling parameters. The Licensee provided typical drawings (Ref. 6-9) and a list of model parameters shown in Table 3. Impact spring values were based on local stiffness of the rack at the support foot to pool a

t 5

g y- - , -

. o L

liner interface and at the fuel assembly to rack cell interface. Rack to rack impact spring values were set at 1 x 100 lb/in. This value is reasonable for base plate locations but significantly larger than would be expected at the girdle bar locations. However, the use of a high spring constant at this location should be conservative and overestimate peak impact loads at the girdle bars.

The Licensee indicated that the new racks will rest on 1-1/4 inch base  !

I plates on the pool floor. The plate material is 304 stainless steel which is the same material as the pool liner. Most of the baseplates were added to the pool at the time at the last rerack. Additional baseplates will be added to ,

accomodate the support configuration of the new racks. The existing. plates were welded to the pool liner. The new plates will not be attached to the pool floor. The baseplates were not included in the rack model but were assumed to move in the same manner as the rack floor. The Licensee indicated that this assumption is reasonable because the friction coefficient between the baseplates and liner should be greater then the friction coefficient between the baseplate and rack support feet because of the differences in materials. A review of the drawings indicates that the baseplates are large enough to accomodate a reasonable amount of slippage of the fuel racks during an earthquake. Overall, the use of baseplates is a desirable design feature since they will serve to distribute fuel rack loads over a large area and will protect the pool liner from local punching or tearing at the rack leg interfaces.

The weight of the fuel included in the model was based on 2500 pounds per fuel assembly which is about twice the design weight. The Licensee used the higher weight to account for possible use of consolidated fuel in the future.

For this application, the Licensee stated that the higher weight should pro- r vide conservative results. The results of additional analytical studies were l provided to support this position as discussed in Section 4.2.1. Since the t Licensee's proposed Licensee amendment did not involve the use of consolidated fuel, the higher weight was considered a conservative modeling assumption in this review. If the Licensee intends to use consolidated fuel at a later i date, further evaluation would be required to reassess the safety margins and ,

to consider other factors which may affect the seismic design.

The Licensee was asked to provide justification for treating the fuel assemblies as five independent rattling masses. The Licensee stated that the fuel assemblies have a natural frequency much lower than the rack and sub-l mitted additional studies to demonstrate that the effects of coupling the  :

masses are not significant when compared to the overall conservatism of the i model. This is discussed in Section 4.2.1. The fuel was modeled as five [

lumped masses at equally spaced elevations above the baseplate. In reality, fuel-rack impacts would be expected to occur at the nine spacer grid locations and at the upper and lower end fittings. The selection of only five impact locations combined with the assumption that all fuel assemblies move in-phase should result in conservative fuel-to-rack impact loads. ]

l 4

g i

i 1

1

e O The Licensee was asked to provide justification for the assumption that the motion of the rack can be represented by a rigid six degree of freedom structure. The Liceasee indicated that for a typical rack, the lowest natural frequency of the race gridwork vibrating in water is 32 Hz. For seismic analysis, it is appropriate to consider this as a rigid body whose motion can be described by a six degree of freedom model.

The adequacy of analyzing only a single rack model in the seismic analysis was questioned. The seismic motion of a single rack is coupled to the motion of adjacent racks through impact forces and fluid coupling forces. The single rack model conntrains the motion of a rack within an imaginary boundary.

Maximum displacements cannot exceed one-half the gap to the adjacent racks.

For sufficiently strong seismic motion, sliding and tilting motions of the racks could be larger than those predicted by a constrained single rack model resulting in higher impact velocities than would be predicted by a single rack model. Under worst conditions, rows of racks could slide together in one direction and pile up against a pool wall. The additional mass of racks involved in the impact could generate larger loads on the racks and the pool walls. This concern may be more critical for the pool valls, since they are not designed to accomodate seismic impact loads from the fuel racts. In response to these concerns, the Licensee committed to perform a two dimensional multiple rack analysis of a single row of fuel racks to determine the extent of rack displacement under an SSE. The results and evaluation of the multiple rack analysis is discussed in Section 4.2 2.

4.1.2 Fluid Coupling Effects The effect of submergence of the fuel racks in a pool of water has a significant effect on their seismic response. The dynamic rack model incorporated inertial coupling (fluid coupling) terms in the equations of motion to account for this effect. For two bodies (mass mi and m2) adjacent to each other in a frictionless fluid medium, Newtons equations of motion have the form:

.. .. i (mi + M11) X1-M12 X2 = applied forces on mass mi

-M21 X'1 + (m2 + M22)'X'2 = applied forces on mass m2 X1 , 'X'2 denote absolute accelerations of mass mi and m2 respectively. M11, M12 M21 and M22 are fluid coupling coefficients j which depend on the shape of the bodies and their relative disposition. The '

basic theory is summarized in a paper by Fritz (Ref. 10). The equations indicate that the effect of the fluid is to add a certain amount of mass to the body (Mil to body 1), and an external force which is proportional to the acceleration of the adjacent body. Thus the acceleration of one body affects the force on the adjacent body. The force is a strong function of the interbody gap, reaching large values for very small gaps. It should be noted that fluid coupling is based on fluid inertial effects and does not constitute damping. Fluid damping was not included in the model.

7

0 0 Fluid coupling terms were included in the equations of motion for fuel masses vibrating within the racks and for racks vibrating adjacent to other racks or the pool wall. The coupling terms modeling the effects of fluid flowing between adjacent racks were computed by assuming that all adjacent racks are vibrating 180 degrees out of phase with the rack being analyzed.

Therefore, only one rack was considered surrounded by a hydrodynamic mass com-puted as if there were a plane of symmetry in the niddle of the gap region.

Fluid virtual mass was included in the vertical direction vibration equations of the rack. 71rtual inertia was also added to the governing equations corresponding to the rotational degrees of freedom. The effect of sloshing of water in the pool was neglected. This effect was shown to be negligible at the bottom of the pool.

Several questions were raised regarding the conservatism of the fluid coupling parameters used in the analysis: (1) The Fritz approach makes the assumption that the vibratory deflections are small relative to the size of the gaps. This assumption does not correspond to the conditions that would prevail during an earthquake where the rack-to-fuel and rack-to-rack displace-ments would be as large as the gaps. Fluid coupling coef ficients are cal-culated on the basis of a constant gap assumption. As fuel racks move away from each other, the coupling coefficients should decrease, resulting in lower fluid coupling forces and possibly higher velocities. (2) The assumption that adjacent racks are vibrating 180 degrees out of phase seems to maximize the retarding effect of fluid forces and reduce the maximum impact velocities of the racks. This can result in unconservative rack-to-rack impact forces. (3)

The rack-to-fuel fluid coupling terms were calculated based on the assumption that the fuel assemblies are solid square cross-sectional bodies, and that all of the surrounding water flows in the fuel assembly / cell wall space around the periphery of the fuel. In reality, the fuel assemblies are arrays of fuel rods with gaps between rods. As a fuel assembly vibrates within a cell, water can flow both around and through the fuel. The resulting fluid coupling forces can then be much lower than predicted by this model.

The Licensee provided additional information to justify the conservatism of the fluid coupling assumptions. Previous studies b'; Singh and Soler (Ref. 11) have shown that for large deflections, the contribution of the fluid leads to terms in the mass matrix and to terms which can be considered as non-linear springs. For the small deflection assumption, the non-linear spring terms disappear and only the mass matrix terms are included as shown in the equations above. The referenced paper provided the results of a study which considered the effects of the non-linear spring terms in a fuel assembly / cell model. The Licensee stated that the study demonstrated that the inclusion of these terms leads to lowering cf the structural response. In  !'

response to the question regarding the consideration of flow area through the fuel assemblies, the Licensee indicated that the flow of water through a fuel assembly array of rods involves repeated changes in the flow cross-sectional area which would result in significant hydraulic pressure losses. The hydraulic pressure loss due to flow through the narrow convergent / divergent j l

8 l

l l

l

channels is an important mechanism for energy loss from the vibrating rack system.

The referenced paper was reviewed for applicability. The study involved the non-linear seismic analysis of a simplified two degree of freedom model of a single fuel assembly / rack cell system. The fuel assembly was modeled as an unperforated square cross-section to simulate a channeled BWR fuel assembly.

Equations of motion were written to incorporate large deflection inertial coupling and fluid damping due to frictional losses. A time history analysis was performed by applying a sinusoidal ground acceleration to the model.

Cases which were analyzed included the following considerations: 1) No fluid effects, 2) Small deflection fluid coupling, 3) Large deflection fluid coupling, no fluid damping, 4) Large deflection fluid coupling with damping,

5) Large deflection fluid coupling with reduced damping. In case 5 the fluid damping was taken as 1% of the values used in case 4 in an attempt to simulate the possible damping effect of unchannelled fuel assemblies. The authors recognized the differences in fluid effects between unchannelled fuel such as the St. Lucie PWR fuel assemblies and channelled fuel assemblies used in BWR's. Channelled fuel assemblies can be appropriately represented as solid square cross-sectional bodies. The authors stated, "It is clear that the damping and virtual mass effects from an unchannelled fuel assembly should be substantially less since the confined fluid has more unobstructed area in which to flow as the fuel assembly moves relative to the cell wall. In addition, there are substantial dif ferences in the flow field which should be considered in any analysis of unchannelled fuel. Nevertheless, case 5 may give some indication of what might be expected if only unchannelled fuel assemblies are in the rack".

The results of the study were presented in terms of fuel-to-rack impact forces and rack spring forces. The latter forces are a measure of rack stress level and pool floor loads.

The results showed that the forces predicted by the small deflection model (case 2) exceeded the forces predicted by the large displacement models with damping (cases 4 and 5). A comparison between the results of the small deflection (case 2) model and the large deflection model ,

with no damping (case 3) showed that the small deflection model predicted higher rack spring forces but lower fuel to rack impact forces.

The referenced study does not resolve all of the concerns related to the fluid coupling model assumptions. It provides evidence that large deflection inertial effects combined with damping tend to predict lower forces then a small deflection model. However, none of the models ptoperly modeled fluid inertial effects for unchannelled fuel as is used in St. Lucie. The reduced damping used in case 5 was only meant to give an indication of trends which might be seen for unchannelled versus channelled fuel rssponse. There was no I analytical or experimental evidence to demonstrate the equivalence of case 5 parameters to unchannelled fuel parameters.

1 l

i 9

l

On the other hand, the case i results (vibration in air, no fluid effects) may be interpreted as an upper bound case. When compared with the case 2  ;

small deflection model results, esse 1 predicted 25% higher fuel to rack  :

impact forces and 20% higher rack spring forces than case 2. When viewed '

together, the results of the 5 cases provide a measure of the sensitivity of ,

variations of fluid coupling parameters in p edicting seismic response forces. This together with safety margins can be used to assess the adequacy ,

of the design.

4.1.3 Friction Effects L

Friction elements were used at the bottom of rack support leg elements of i i the model. The value of the coefficient of friction was based on documented '

test results given in Reference 12. The results of 199 tests performed on austenitic stainless steel plates submerged in water showed a mean value of coefficient of friction to be 0.503 with a standard deviation of 0.125. Based on twice the standard deviation, the upper and lower bounds are 0.753 and 0.253, respectively. Two separate analyses were performed for each load case with values of coefficient of friction equal to 0.2 (lovet limit) and 0 8 )

(upper limit), respectively.

The Licensee was asked to provide justification for using the same friction coefficient for both static and sliding rack conditions. He indi- ,

l' cated that there is only a small difference between the static and sliding values. The use of both an upper and lower bounding value is judged to be i appropriate. Previous studies have indicated that low friction results in i maximum sliding response of the racks while high friction results in maximum ,

] rocking or tilting response. Consideration of both cases should provide worst j,

case displacements, stresses and impact loads.

4 1.4 Damping Since the model assumed that the fuel rack gridwork and baseplate are  ;

rigid, and the fuel assemblies can be treated as independent lumped masses, no j damping resulting from structural deformations of the components was assuecd.

i Structural damping was included in all of the impact spring elements. For SSE  ;

, load conditions, 2% structural damping was used. This value is in accordance  ;

l with the FSAR and represents an acceptable, conservative value for impact '

damping.

l i

4 1 5 Seismic Loads l

Seismic floor response spectra for the spent fuel pool floor were  !

, developed using the methods described in the FSAR. The parameters of the l original lumped mass model of the Fuel Handling Building were adjusted to l 4

reflect the increased mass corresponding to the new high density spent fuel I storage racks. New response spectra curves were generated using the same 4 method which was used in the original dynamic analysis. Minimum and maximum fuel rack weights were considered in the analysis, corresponding to the empty I

1 10 4

4 d

,m --- -- - -m. _ -, , - - , , - , , - - , - . ,m ,~,-m. , - - - - - - - - , , , -----,,--,-~.--.-,.er,--

I and full conditions of the racks. Three ground motion acceleration records (as used in the original plant design) were used as input. These six combi-nations of parameters resulted in six response spectre curves, which were then l broadened + 20% and enveloped into one curve which envelopes the full spectrum of rack loading conditions. Six such curves were developed, two (OBE and SSE) for each direction (NS, EW, vertical). The response spectrum curves are included in the Licensee's Safcty Analysis Report (Reference 1)

The revised response spectra were used to generate statistically inde-pendent acceleration time histories, one for each of the three orthogonal directions. A computer program was used to generate the artificial time histories as a sum of sinusoidal waves. The program used a iteration approach whereby the calculated response provided by the simulated seismic excitation was compared with the "target" design response spectrum. The amplitudes of the sine waves were modified at each iteration step so as to obtain the best agreement at certain control frequencies specified by the user. The resulting time histories used in the fuel rack analysis are shown in Figures 13 to 15.

The Licensee was asked to provide a comparison of the design response spectra with the artificial time history response spectra. This comparison was provided in terms of velocity response spectra plots in Reference 3a. The t l plots showed reasonable agreement between the calculated curves and the design l curves. l Based on the Licensee's description, the methodology used to develop the j seismic input for fuel rack analysis is acceptable and consistent with -

industry practice.

4.1.6 Load Cases Rack modules 32, G1 and H1 (see Figure 1) were analyzed to show that structural integrity is saintained during a seismic event. The Licensee was ,

i j

asked to provide the basis for selection of thewe specific racks and gave the following information:

Module B2 is representative of a region 1 rack. It is the largest region 1 rack and is located in a corner of the pool.

4

Module G1 is a large region 2 rack located in a corner of the pool wall 1

and the cask area vall. This rack has six feet, two of which have an initial

]

gap and are designed to come into contact with the floor only when rocking is sufficient to close the gap. The eccentric placement of its main support legs 4

causes this rack to be relatively more prone to rocking, thus resulting in potentially higher displacements and stresses than a more conventional region 2 rack.

Module H1 is a region 2 rack with a cut-out and one additional support foot. For conservatism, the rack was considered to have 104 cells loaded with fuel but used a planform for analysis that was less stable than the planform actually present.

11 1

o ,

For each rack module, several analyses were performed to investigate the variations in friction coefficient (C0F = 0.2 and 0.8) and fuel load condition (fully loaded, half full and empty).

The Licensee's choice of modules does not cover every configuration but the selection was based on reasonable conservative considerations such as large weight and tendency to rock. All of the rack modules analyzed are located next to a pool wall or corner. Modules in this area would have less significant fluid coupling forces. The variation in friction coefficient and fuel load cover a reasonable range of conditions.

4.1.7 Analysis Method Once the rack seismic models were assembled, equations of motion of the system were written and solved using the DYNARACK computer program. The analysis method is based on the component element method of analysis described in Reference 13. The solution of the problem involves the following steps:

1. Development of a mathematical model of the rack structure in terms of lumped masses, non-linear springs, fluid, coupling elements, and provisions for three dimensional kinetic degrees of freedom.

2. Development of equations for the kinetic, energies of the rack, the fuel assemblies, and the entrained and.coupli'ng fluid energies.

3. Application of Lagrange's formulation to, assemble the displacement coupled second order differential equations in t,he prescribed generalized coordinates. The set of equations are then" numerically solved by the DYNARACK computer program.

The Licensee was asked to provide additional information on the DYNARACK program and its verification. This program'is a refine version of the DYNAHIS program which has been used and accepted by*NRC in previous fuel rack s analyses. Both programs provide the numerical solution for non-linear models l

of structures under time history inputs. Th6 Licensee stated that verifi- .

cation of the DYNARACK program was carried out in accordance with Quality AssuranceProceduresfollowing10CFR50, App (ndixB. Validation of DYNARACK  ;

results involves: (1) comparison 'sith analytical solutions and with numerical i solutions obtained from other computer codes, and (2) manual calculations of mass atrix terms and comparison with results #

determined internally by 4 DYNARACK.

l Based on the information provided, the application of the component element method and use of the DYNARACK progfam to analyze the non-linear lumped mass models of the fuel racks is acceptable.

1 12

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

?

$ s 4.1.8 Analysis Results i

The DYNARACK program computed displacements and element forces at each instant of time during the earthquake. Stresses at critical rack locations 1 were computed based on the nodal forces. These stresses were checked against l the design limits. Stresses were presented in terms of highest stress factors for each load case. Stress factors R1 through R6 were defined as the max- i i

iaus computed stress to its allowable value. The stress limits were derived i j

from the ASME Code,Section III, Subsection NF, in conjunction with material j properties from the Section III appendices and supplier's catalog. The faulted condition (Level D) limits from Section III, Appendix F, were used for the SSE allowables.

The stress factors were defined as follows:

R1- Ratio of direct tensile or compressive stress on a net section to -

its allowable value (note support feet only support compression) ,

R2- Ratio of gross shear on a net section to its allowable value R3- Ratio of maximum bending stress due to bending about the x-axis

to its allowable value for the section 1

R4- Ratio of maximum bending stress due to bending about the y-axis to ,

its allowable value i R5- Combined flexure and compressive factor (as defined in ASME Code

Section III, Append *x XVII) l f

• i R6- Combined flexure and tension (or compression) factor (as defined  !

l in ASME Code,Section III, Appendix XVII)

The limiting value of each stress factor is 1.0 for OBE conditions. For SSE conditions, the limit is 2.0 for the rack material and upper part of the l support feet, and 1.53 for the lower support feet.

l l

q Maximum stress factors for the rack base and support feet for each load I 1 case are presented in Table 4. The Licensee stated that the critical stress i factors reported for the support feet were all for the upper segment of the

feet and should be compared to a limiting value of 2.0. Table 4 also presents maximum fuel assembly-to-cell impact loads, rack-to-rack impact loads and l rack-to-vall impact loads. Table 5 presents maximum rack displacements and i floor loads.

1 In addition to determining stress factors, the Licensee performed j additional calculations to evaluate the adequacy of welds, the effects' of

rack-to-rack and rack-to-fuel impact loads, and other local ef fects. These j

caleviations were not included in the Safety Analysis Report (Ref. 1). During the audit at Holtec International, sample calculations were reviewed. Table 6 j summarises the safety factors in critical rack locations.

13 1

1

e i r

. 4.1.9 Evaluation of Results The results of the Licensee's seismic analysis indicated that all stresses in the racks would meet their allowables, impact loads on fuel assemblies would not damage the fuel, and rack displacements would not be large enough to

result in impacts with the pool wall. However, considering the potentially unconservative modeling assumptions discussed in Sections 4.1 1 and 4.1.2 regarding multiple rack effects, rattling fuel mass representation, and fluid coupling considerations, it was judged prudent to have the Licensee perform additional studies to address the questions raised. An issue of particular concern was the possibility that the single rack models would underpredict l displacements of racks adjacent to pool valls and that rack-to-wall impacts would occur. The walls were not designed to accommodate seismic impact loads from the fuel racks and damage to the walls or liner could result in unacceptabic leakage of water from the pool.

The additional studies performed by the Licensee are discussed in Section 4.2. The evaluation of their results and the overall assessment of the seismic analysis results is given in Section 4.2.3.

4.2 Additional Puel Rack Seismic Studies J As a result of questions raised during the review of the fuel rack dynamic analysis model (Section 4.1.1), the Licensee performed additional analyses.

Single rack model studies were carried out to address questions regarding the adequacy of treating the fuel assemblies as five independent rattling masses and using twice the fuel weight in the models. Multiple rack studies ware performed in response to questions regarding the adequacy of a single rack <

model in predicting forces and displacements that would occur if multiple rack '

effects were considered. A description of these additional analyses and their l results is discussed below.

1 4.2 1 Single Rack Studies i

l Two additional seismic analyses of single rack models were per' formed for a fully loaded G1 rack with coelficient of friction equal to 0.8 and a fuel ,

weight per cell equal to 1300 lbs. The design basis analysis had indicated that this load case was the most critical case which predicted the highest overall response. In the first additional analysis (Case 1), the fuel was modeled in the same manner as the design basis analysis, i.e. as five inde- ,

I pendent rattling masses. In the second run (Case 2), the five fuel masses

); were connected by springs, thus providing a beam representation of the fuel assemblies. The springs did not represent the actual flexural rigidity of a fuel assembly but were based on the properties of a fictitious channel around the assembly. This flexural rigidity appears to be of the same order of mag-1 nitude as the actual flexural rigidity and is judged to be reasonable for this study.

\

i 1

i 14

)

O I o

1 The results of the single rack model studies are presented in Table 7.

Key forces, stresses and displacements are compared. The Case 1 versus Case 2 comparison indicates generally comparable results. The elastically coupled mass model (Case 2) results do not exceed the independent mass model (Case 1) results by more than 15%. However, the results of both cases are clearly enveloped by the design basis case by significant margins. Therefore, this study demonstrates that modeling the fuel with twice its actual weight provides a significant level of conservatism which adequately compensates for

, the smaller potential unconservatism of modeling the fuel as independent rattling masses.

4.2.2 Multiple Rack Studies The Licensee performed additional seismic analyses of a row of four racks to investigate the adequacy of the design basis single rack models in pre-dieting the response of fuel racks in the actual multiple rack fuel pool environment. An issue of particular concern was the possibility that in a multiple rack environment, the peripheral racks may hit and damage the walls of the pool.

The four racks studied were racks A1, A2, B1 and B2, next to the south pool vall shown in Figure 1. A simplified planar two dimensional model of the row of racks was developed. Each rack was represented by a four degree of

freedom model representing horizontal and vertical translations of the rack, planar rotation (rocking) of the rack, and horizontal translation (rattling) of the fuel assemblies. The racks were assumed to be fully loaded with fuel

, using the nominal fuel weight (1250 lb/ assembly) which is half the weight used j in the single rack design basis models. Support spring constants, impact r spring constants and gaps were consistent with the design basis models. Fuel to cell fluid coupling coefficients were reduced to 50% of the "blunt body" r

,~ value in an attempt to compensate for the potential overprediction of fluid coupling forces predicted by the design basis models as discussed in Section 4.1.2. Runs were made for both the 0.2 and 0.8 coefficients of friction. The i seismie leading of the E-W and vertical SSE were applied simultaneously to the model.

t l The key responses were compared with the corresponding responses from the single rack design basis analysis of the 52 rack. These results are presented in Tables 8 and 9. The Licensee stated that the results support the conserva-tism of the design basis model. Both displacements and impact loads were pre-dicted to be lower by the multiple rack model. The smaller displacements supported the conclusion that the peripheral racks would not hit the pool walls.

During the course of the analysis, the Licensee decided to modify the side gap spacing between the pool wall and the peripneral racks from 4.5 inches to 5.5 inches. The multiple rack analysis was rerun to reflect the revised spacing. A comparison of responses between the two multiple rack studies is ,

presanted in Table 10. The results showed slight increases in responses but 1 1

2 15 I

_ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _l

the loads and displacements were still enveloped by the single rack design basis results by significant margins.

In evaluating the results of this study, several factors had to be weighed. The row of racks that was selected for analysis was a representative row and was not expected to have the highest response. The Licensee therefore made an appropriate comparison when he compared the results of this study to the results of the B2 rack design basis model. It was important to evaluate the results on a comparative basis and recognize that they are not worst case results.

It should also be recognized that the results may be somewhat unconserva+

tive because the model assumed planar two-dimensional motion. In this type of model, only one horizontal component of the earthquake could be applied.

Three dimensional cross-coupling effects could not be accounted for. Never-theless, it is reasonable to expect this 2-D multiple rack model to capture the primary response and potential interaction effects of a row of fuel racks in one direction.

4.2.3 Overall Evaluation of Seismic Analysis Results The results of the additional studies presented in Tables 7 through 10 support the adequacy of the design basis (single rack model) results. Both the single and multiple rack models used in these studies utilized actual fuel weight instead of twice the fuel weight as used in the design basis models.

It appears that the high fuel weight was the most significant contributor to the conservatism of the design basis model results. Further studies would be required to prove that single rack models using actual fuel weights would always give conservative results. However, for this application, these studies have provided a reasonably high level of confidence in the adequacy of the results. A review of the safety factors predicted by the design basis models (Tables 4 through 6) provide further assurance that the racks are designed with sufficient safety margin to compensate for uncertainties in the seismic analysis.

Based on the results of the Licensee's seismic analyses, it is concluded that during an SSE, the fuel racks will maintain their structural integrity, fuel assemblies will not sustain dsmage, and rack displacements will not be large enough to result in pool vall impacts.

4.3 Thermal Analysis Weld stresses due to heating of an isolated hot cell were computed. The analysis assumed that a single cell is heated over its entire length to a I temperature above the value associated with all surrounding cells. No thermal  !

gradient van assumed in the vertical direction. Using the temperatures associated with this unit, weld stresses along the entire cell length were found to be below the allowable value with a safety factor of 2.2 as indicated in Table 6. i l

16

)

1

4.4 Fuel Handling Accident Analyses The Licensee performed structural analyses and evaluations for three post-ulated fuel handling accidents. The accidents and the analysis summaries were described in the Safety Analysis Report as follows:

1. Dropped Fuel Accident I A fuel assembly is dropped from 36 inches above the module, falls into a cell, and impacts the base. The final velocity and total energy at impact was calculated. To study the baseplate integrity, it was assumed that the energy was directed toward punching of the baseplate in shear and thus was transformed into work done by the shear stresses. It was determined that shearing deformation of the baseplate was less than the thickness of the baseplate so it was concluded that local piercing of the baseplate will not occur.

Direct impact with the pool liner would not occur. The suberiticality of the adjacent fuel assemblies would not be violated.

2. Dropped Fuel Accident II One fuel assembly drops from 36 inches above the rack and hits the top of the rack. By applying an energy balance approach, it was determined that permanent deformation of the rack would be limited to the top region such that the rack cross-sectional geometry at the level of the top of the active fuel and below is not altered. The region of local permanent deformation was shown not to extend below six inches from the rack top.

3. Jammed Fuel Randling Equipment A 4000 pound uplift force was applied at the top of the rack at the weakest storage location. The force was applied on one wall of a storage cell as an upward shear force. The plastic deformation was found to be limited to the region well above the top of the active fuel.

The Licensee concluded that these analyses proved that the rack modules are engineered to provide maximum safety against all postulated abnormal and accident conditions. During the audit, the Licensee was asked to provide the calculations for the dropped fuel accident I for a detailed review. The review and evaluation of this calculation is discussed below.

A key element of the dropped fuel accident analysis was the calulation of impact velocity. The model for predicting velocity treated the fuel assembly as a free body falling through a channel. The model considered gravity and fluid forces and accounted for virtual mass effects. Fluid forces were determined by applying basic fluid mechanics laws of continuity and energy.

17 f

However, the model was found to be unconservative in calculating the pressure

, build-up within a cell. The model assumed that as a fuel assembly falls through a rack cell, all of the water in the cell is forced out through the baseplate holes at the bottom. Flow through the fuel assembly was neglected.

Since that flow area is significant, the model may have overpredicted the fluid retarding force and underpredicted the impact velocity and kinetic energy of the fuel assembly as it hits the baseplate.

In evaluating the potential penetration of the baseplate, the kinetic energy of the fuel assembly was set equal to the work performed as a slug is punched out of the baseplate. The calculation showed that the depth of pene-tration is less than the plate thickness and concluded that penetration would not occur. Furthermore, the Licensee stated that the purpose of the calculation was only to show that there is no danger to the pool liner. In the event of a dropped fuel assembly, the base plate could be expected to plastAcally deform and separate from the rack cells but this would not af fect center-to-center spacing. The Licensee stated that the baseplate, even with plastic bending occuring, would not touch the liner floor.

A number of weaknesses were noted in the evaluation: 1) The equation for work required to penetrate the plate lacked sufficient experimental verification. The use of er -enetration formulas would have been more appropriate. 2) The shear terpredicted. This area was based on the solid square cross-sect ' the fuel assembly. In reality, the fuel assembly rests on four 1 % s.ch smaller contact area. 3) The conclusion that the baseplate wc tact the floor was not substantiated by any calculation to  : deformation or ductility ratio.

Although a number of weaknesses were identified, there were also a number of conservatisms which must be considered. The fuel weight used in the velocity and energy calculations was 2500 pounds which is nearly twice the actual weight. The fuel was assumed rigid for impact stress calculations.

All of the impact kinetic energy was assumed to be directed toward punching of the baseplate in shear. None of this energy was directed toward bending of the plate or compressing the fuel.

To evaluate the final conclusions of this analysis, BNL performed some simplified calculations using bounding assumptions. Kinetic energy at impact was :alculated on the basis of the actual fuel weight and velocity of a f ree-f alling body in air. The resulting kinetic energy was approximately twice that used in the Licensee's calculations. Baseplate penetration was evaluated by empirical penetration formulas for steel targets commonly used for missile penetration analysis in the nuclear industry. Both the Ballistic Research Laboratory (BRL) formula and the Stanford Equations were applied (Ref. 14). The missile contact area was based on the fuel assembly leg area rather than the total cross-sectional area. The results of this analysis concurred with the Licensee's conclusion that baseplate penetration would not occur.

1 18

The Licensee's conclusion that the baseplate would not contact the pool floor could not be verified by a simplified analysis due to the complex nature of the structure. However, it should be noted that since the pool floor had been shown capable of withstanding a fuel cask drop, it would be reasonable to conclude that the floor has sufficient strength to withstand the impact load resulting from the drop of a single fuel assembly.

4.5 Spent Fuel Pool Analysis 4.5 1 Loads and Load Combinations The reanalysis of the spent fuel pool considered the following design loads:

Structural Dead Load (D)

Live Load (L)

Seismic Loads (SSE and OBE)

Normal Operating Thermal Loads (T)

Accident (Loss of Fuel Pool Cooling) Thermal Load (T ) A Fuel Cask Drop Load (M)

The following load combinations, from the St. Lucie, Unit No.1 Updated FSAR, Section 3.8.1.5, were considered:

a) Normal Operation 1.5 (D+T) + 1.8L b) OBE Condition 1.25 (D+T+0BE+0.2L) c) SSE Condition 1.05 (D+T+0.2L) + 1.0 SSE d) Accident and Cask Drop 1.05 (D+TA+0.2L)

For the evaluation of the liner and liner anchors, the above load combinations were applied except that load factore for all cases were equal to 1.0.

Linear analyses were performed initially to determine the critical load combinations. As a result, the following loading cases were selected for the non-linear concrete cracking analysis:

1. 1.5D + 1.8L

2. 1.05 (D + Twinter + 0 2L) + 1.0 SSE

3. 1.05 (D + Tsummer + 0.2L) + 1.0 SSE 19

4. 1.05 (D + 0.2L) + 1.0 SSE

5. 1.05 (D + TA + 0.2L)

6. 1.05 (D + Twinter + 0.2L) + 1.0M

7. 1.05 (D + 0.2L) + 1.0M 4.5.2 Spent Fuel Pool Structure Analysis A finite element model of the lower portion of the spent fuel pool structure was developed. Since the effect of the additional fuel rack load on the pool floor is limited to the mat in the pool area, the upper portion of the pool walls was not reevaluated. The model included the lower portion of the walls up to elevation 45.25 f t, the pool floor and the underlying soil.

The structural components included in the model are shown in Figure 16. A computer plot of the finite element model is shown in Figure 17 which shows the overall view of the model indicating the composite of the four exterior and one interior wr.11s.

In this analysis, the EBS/NASTRAN program was used. The Licensee was asked to provide additional information on this computer program. This was provided in Reference 3c. EBS/NASTRAN is an enhanced NASTRAN program developed by Ebasco. It has all of the NASTRAN capabilities plus additional features. One of the additional features is the ability to pertorm concrete cracking analysis. This feature incorporates a special plate element which consists of a user-specified number of layers, each having a different proportion of steel to concrete area, representing the presence of reinforcing steel. Each layer will crack or re-close according to the stress-strain relationships of the concrete and steel. Thus, a cracking pattern and stress redisttibution can be determined. A verification problem was submitted which demonstrated good agreement of analytical results with experimental data.

The maximum stress results in the concrete and rebars from the nonlinear i

analysis of the seven load cases are presented in Table 11. The design stress limits described in the St. Lucie Unit 1 FSAR were used in the evaluation.

The capacity of all sections was computed in accordance with ACI 318-63 Part l IV-B, Ultimate Strength Design. Table 11 indicates minimum safety factors fot each loading case. Safety factor is defined as allowable stress divided by maximum actual stress including load factors. The smallest safety factors are 1.10 for reinforcement bar tension, 2.65 for concrete compres9 ion, and 1.05 ,

for concrete shear. Based on these results, it can be concluded that the i spent fuel pool structure can accomodate the revised loads. I 4.5.3 Pool Liner and Anchorage Analysis The l!ner and its anchors were evaluated for the temperature load, the strain induced load due to the deformation of the floor, and the horizontal I seismic load. The POSBUKF computer program was used for the liner buckling '

20 1

i i

analysis due to the temperature and strain induced loads. The Licensee was asked to provide additional information on this computer program. This was provided in Reference 3c. POSBUKF is a program developed by Ebasco to examine the elastic post-buckling behavior of a flat plate subjected to thermal and lateral loading using an energy method approach. The program determines the deflected shape of a buckled plate by minimizatica of potential energy, and from this calculates plate stresses utilizing strain-displacement and stress-strain relationships for the particular case under study. The program was verified by comparison of test problem results to hand calculation results.

The liner anchors were evaluated for the unbalanced liner in-plane force due to the temperature and strain induced loads, as well as horizontal seismic in-plane shear force.

The acceptance criteria for the liner and anchors was in accordance with the requirements of ACI-ASMI Section III, Divisisn 2. Subsection CC for containment liners. The critical loading case for the liner was the case which included accident thermal load. The analysis showed that the maximum calculated strain was below the Code Strain allowable with a safety factor of 5.2. The buckling analysis indicated that the liner plate would not buckle.

Two loading conditions were considered in the liner anchor evaluation; one was the strain-induced load which produced the unbalanced in-plane force at the edge of the pool area, and the other was the horizontal seismic load transmitted through friction between the rack support and the liner. The analysis indicated that Code allowables were met with minimum safety factors of 2.5 for the strain-induced load case, and 1.33 for the seismic load case.

Based on these results, it can be concluded that the fuel pool liner and anchorage can accommodate the revised loads.

5.0 CONCLUSION

S Based on the review and evaluation of the Licensee's Safety Evaluation Report and additional information provided by the Licensee during the course of this review, it is concluded that the proposed St. Lucie Unit 1 fuel racks have sufficient structural capacity to withstand the effects of all required environmental and abnormal loadings J1scussed in this report. Impact loads generated by the closing of fuel assembly to fuel rack cell gaps during the SSE vould not lead to damage. Furthermore, the existing spent fuel pool should have adequate capacity to acco=modate the increased loads resulting from the stor2Ee of more fuel assemblies in the pool.

All concerns related to the adequacy of the dynamic single rack design basis models including multiple rack effects (Section 4.1.1), rattling fuel mass representation (Section 4.1.1), and fluid coupling considerations (Section 4.1.2) were resolved by additional studies performed by the Licensee. These studies (Section 4.2.) investigated multiple rack effects and the sensitivity of model variations. They demonstrated that the single rack design basis models predict conservative seismic loads and displacements.

21 1

l l

Although the studies were limited in scope, they provided evidence which

, indicated that the most significant contributor to the conservatise of-the design basis models was the use of twice the fuel assembly design weight in the models. An additional analysis of a single rack model which used the actual fuel weight predicted displacements and impact loads which were approx-imately half of the corresponding design basis model results. Analysis of multiple rack models which also used actual fuel weights showed similar trends )

j in the results. Thus it was judged that the design basis models have suffi- l cient conservatism to compensate for potential underprediction of response due l to the modeling concerns d*,scussed in this report. '

)

l 2

i i l i l

< )

1 i

I 1

i 22 l

i l

1

6.0 REFERENCES

1. Florida Power and Light Company., St. Lucie Plant - Unit No. 1. Spent Fuel Storage Facility Modification Safety Analysis Report, Docket No. 50-335.

2. NRC letter, G. Bagchi to E. Tourigny, "Request for Additional Information

- Proposed License Amendment - Spent Fuel Rerack, St. Lucie Unit 1, Docket No. 50-335, TAC #65587", dated August 7, 1987. ,

3a. FP&L letter L-87-422, C.O. Woody to USNRC, "St. Lucie Unit 1, Docket No.

50-335, Spent Fuel Pool Rerack - Design and Analysis", dated October 20, 1987.

3b. FP&L letter L-87-535, C.O. Woody to USNRC, "St. Lucie Unit 1, Docket No.

50-335, Spent Fuel Rerack-Design and Analysis," dated December 23, 1987.

3c. FP&L letter L-37-536, C.O. Woody to USNRC, "St. Lucie Unit 1, Docket No.

50-335, Spent Fuel Rerack-Design and Analysis," dated Deceeber 23,198'i

4. USNRC letter to all power reactor licensees, from B.K. Grimes, dated April 14,1978 "0T Position for Review and Acceptance of Spent Fuel Storage and Handling Applications", as amended by the NRC letter dated January 18, 1979.

5. US Nuclear Regulatory Commission, "Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants," NUREG-0800, Revision 1, July, 1981.

6. Ebasco Drawing 8770-C-830. "Fuel Handling Bldg Spent Fuel Pit Liner",

Sheet 1, Rev. 4; Sheet 2 Rev. 2; Sheet 3 Rev. 4; Sheet 4, Rev. O.

7. Joseph Oat Drawing D-8286, Rev.1. "Details Region I, Spent Fuel Storage  ;

Racks".

8. Joseph Oat Drawing D-8288, Rev. 1, "Plan Diagram of Can to Cap Element Joints Region I, Spent Fuel Storage Racks".

9. Combustion Engineering Drawing E-13172-161-101, Rev. 5. Sheet 1 of 2 "Fuel Bundle Assembly".

10. R.J. Frits, "The Effects of Liquids of the Dynamic Motions of Immersed Solids", Journal of Engineering for Industry, Transactions of the ASME, February, 1972, pp 167-172.

11. K.P. Singh and A.I. Soler, "Dynamic Coupling in a Closely Spaced Two-Body System Vibrating in a Liquid Medium: The Case of Fuel Racks", 3rd International Conference on Nuclear Power Safety, Keswick, England, May 1982.

1 23

h d

s r i

.J 1 12. E. Rabinowicz, "Friction Coefficients for Water Lubricated Stainless  !

, Sted1s for a Spent Fuel Rack Facility", a report for Boston Edison

! Company, MIT, 1976.

l i 13. S. Levy and J.P.D. Wilkinson, "The Component Element Method in Dynamics".- ,

j McGraw-Hill, 1976.  !

i

14. R.C. Gwaltney, "Missile Generation and Protection in Light-Water-Cooled (

j Power Reactor Plants", ORNL-NSIC-22, September 1968.  ;

i .

i -,

h i

I ,

a .

1 i

1

< l 1 .

I i 4

l i

i i 1

1 j i I b \

k i

I i

l l

I i

J s

i l

i 24 i

l l

. I

--e s .l i

}

f  !

l l TABLE 1 l TABLE OF MODULZ DATA

) i I

1 >

! NO. OF NO. OF

] CELLS CELLS TOTAL NO. ,

NO. OF IN N-S IN E-V 0F CELLS '

j MODULE I.D. MODULES DIRECTION -DIRECTION PER MODULE i

i d

v y Region 1 2 9 9 81 j Al to A2  !

i e

i Region 1 2 9 10 90  :

31 to B2  ;

i  :

, Region 2 4 13 9 117  !

i C1 to C4  !

l l Region 2 3 13 8 104 I D1 to D3 i l, .

6 1

! Region 2 2 11 8 88  !

! El to E2  !

i.

j Region 2 1 12 8 96 F1

j

Region 2 2 12 9 108  !

) G1 to C2 i

i j Region 2 1 13 8 96 ,

t H1 i

i i

i n

w I

I 25 i

i i

4

o e TABLE 2 MODULE DIMENSIONS AND WEIGHTS l

l l

NOMINAL CROSS-SECTION ESTIMATED DRY DIMENSIONS WEIGHT (1bs)

MODULE I.D. N-S E-W PER MODULE Region 1 90-1/4" 90-1/4" 26.700 Al to A2 e Region 1 90-1/4" 100-7/16" 29.800 B1 to B2 l

Region 2 115-11/16" 80-1/6" 24,100 C1 to C4 Region 2 115-11/16" 71-3/16" 21,500 D1 to D3 l

Region 2 97-7 8" 71-3/16" 18,200 El to E2 Region 2 106-3/4" 71-3/16" 19,800 F1 Region 2 106-3/4" 80-1/16" 22.300 t

G1 to G2 Region 2 115-11/16" 71-3/16" 19,800 H1 26 l

l l

TABLE 3 RACK MODEL PARAMETERS Rack Module HL B2 Gl*

KI (#/in) .359 x 106 .310 x 106 . 3 72 x 106* * * ,

Kw (#/in) .1 x 107 .1 x 107 .1 x 10 7 **

Kg (#/in) .221 x 1010 .221 x 1010 .221 x 1010 Kd (#/in) .112 x 107 .109 x 107 .123 x 107 KR (fin) .567 x 108 .567 x 108 . '57 x 108 rad h (in) 6.125 6.125 6.125 H (in) 169 169 169 W4 (1b) 19800 29800 22300 Wg (1b) 260000 225000 270000 Lx (in) 71 90 80 Ly (in) 116 100 107

• 6 support feet (.1875" initial gap on 2 of 6 supports)

• • Where 2 racks are adjacent, gap between base plates =

.625"; gap between girdle bars = .375"

• • • Nominal gap between cell wall and fuel assembly = .125" KI - fuel assembly-to-cell wall impact spring rate Kw - rack-to-rack or rack-to-wall .12npact spring rate Kg - friction spring rate (active prior to sliding)

KR - spring rate representive of rotational resistance between liner and support leg  !

Kd - support leg axial spring rate h length of support leg j

H - height of rack above base plate Wr - Weight of rack without fuel Wg - Weight of fuel ,

Lx - P hnform dimension (X-direction) l Ly - Planform dimension (Y-direction) 27

TABLE 4 - RACK SEISMIC ANALYSIS RESULTS

SUMMARY

IMPACT LOADS AND STRESS FACTORS Fuel STRESS FACTORS Assembly Rack / Rack Rack / Wall to cell Impact load Impact Load (Upper Values for Rack Base -

Module / Load Case Impact CB/BP

• CB/BP Lower Values for Support Feet)

Load (#)

R1 R2 R3 R4 R5 R6 G1, p=.2,fu11 9.967x104 4.994x104/ 0/0 .094/.287 .027/.085 .219/.175 .174/.226 .331/.436 .377/.467 1.763x104 C1, U .8,fu11 9.438x104 1.022x105 /0 0/0 .133/.421 .100/.427 .392/.918 .267/.577 .495/1.137 .562/1.27 C1, U .8, 1.319x105 1.359x105 /0 0/0 .130/.427 .100/.426 .391/.917 .264/.549 .506/1.136 .576/1.27 convergence C1, p .2.1/2 fuit 9.046x104 1,951x104 /0 0/0 .052/.169 .014/.047 .120/.098 .190/.128 .262/.254 .300/.269 C1 p .8,1/2 full 9.493x104 8.264x104 /0 0/0 .088/.256 .051/.229 .759/.460 .221/.599 .369/.702 .425/.792 B2 p .2, empty 2.006x104 0/0 0/0 .014/.046 .005/.013 .038/.02s .048/.035 .061/.063 .030/.066 B2 p .8, empty 7.896x103 2.401x104 0/0 .017/.092 .025/.073 .077/.155 .059/.076 .107/.194 .124/.217 B2 p .2,fu11 8.016x104 S.312x104 /0 0/0 .087/.211 .025/.063 .176/.1;! .186/.159 .290/.311 .328/.321 B2 p .8,fu11 8.766x104 1.170x105 /0 0/0 .110/.141 .078/.851 .353/.872 .341/.631 .505/1.076 .576/1.21 1

B2 p .8,1/2 full 7.484x104 3.780x104 /0 0/0 .058/.220 .052/.242 .221/.492 .200/.247 .300/.666 .345/.740 H1 p.8,fu11 9.050x104 8.225x104 /0 0/0 .119/.346 .087/.327 .320/.726 .340/.516' .572/.905 .655/1.02 ti p .2,1/2 full 8.980x104 2.350x10 /04 0/0 .050/.164 .014/.046 .097/.096 .180/.092 .245/.244 .280/.258 HI U=.8,1/2 full 1.077x105 5.946x104 /0 0/0 .068/.218 .055/.199 .211/.439 .222/.405 .335/.582 .381/.653 i

i

• GB - CIRDLE BAR; BP= BASE PLATE

TABLE 5 - RACK SEISMIC ANALYSIS RESULTS

SUMMARY

DISPLACEMENTS AND FLOOR LOADS MAX. DISP. HAX. DISP. MAX. VERT. HAX FLOOR MAX FLOOR LOAD (i)

MODULE LOAD CASE DX (IN) DY (IN) DISPL. (IN) LOAD (i) VERTICAL /SilEAR

• 4 FEET C1 p=.2, full .3884 .6305 0 4,171x105 1.877x105 /37540.

C1 p 1.8197 .6110 .97377x10-1 5.949x105

.8, full 2.75492x105 /186325.

C1 p=.8, convergence 1.7407 .6147 .90944x10-1 5.877x105 279673./186242.

C1 p .3566 .4071 .12229x10-1 2.301x105

.2, 1/2 full 110685./22137.

C1 p=.8, 1/2 full .8427 .3744 .85291x10-1 3.903x105 167843./104113.

D$ B2 p .2, "empty' .1517 .0898 0 7.384x104 30353./6071.

B2 p=.8, "empty" .1120 .2287 .43159x10-1 8.950x104 60418./32103.

p .2464 .2088 B2 .2, full 0 3.724x105 137831./27566.

52 p .8, full .5317 .4238 .29708x10-1 4.593x105 223083./165014.

B2 p .8, 1/2 full .3802 .2786 .31333x10-1 2.543x105 144247./113093 H1 p .8. full .5092 .2548 .31422x10-1 5.181x105 226457./145819.

p .2107 .2132 .69853x10-2 H1 .2, 1/2 full 2.233x105 107217./21443 H1 p=.8, 1/2 full .2731 .2241 .44346x10-1 3.090x105 142827./93389.

• VERTICAL - Vertical Load SHEAR - Shear Load

TABLE 6 SUhMARY OF SAFETY FACTORS IN CRITICAL FUEL RACK LOCATIONS SAFETY ITEM / LOCATION 7 ACTOR COMMENTS Support foot to baseplate 2.44 weld stress Cell to baseplate weld stress 3.15 Cell to gap channel weld 2.94 Stress due to seismic loads  ;

stress l

Cell te gap channel weld 2.20 Thermal stress due to effects I s t r e s *, of isolated hot cell Impact load on girdle bar 2.17 Girdle bar shear stress 1.70 j Cell wall stress due to 2 54 girdle bar impact load Impact load between fuel 3.58 Based on cell wall limit load assembly and cell wall Impact load between fuel 1.51 Based on plastic deformation assembly and cell wall of fue:1 sp&cer grids

• Shear load on baseplate 3.0 near a support foot Corpressive stress in cell 4.56 Based on local buckling wall considerations Rack to wall impact loads -

No Impacts with pool walls occur at any location 30 Sa ety factor on fuel rod crushing is significantly higher

TABLE 7 RESULTS OF SINGLE RACK STUDIES FULLY LOADED G1 RACK WITH C0F = 0.8 INDEPENDENT ELASTICALLY COUPLED DESIGN BASIS ITEM ?UEL MASSES FUEL MASSES MODEL (1300 lb/ fuel) (1300 lb/ fuel) (2500f/ Fuel)

Fuel / Rack 453.3 514.4 1221.3 Impact (#/ cell)

Rack / Rack 7.133x104 /0. 6.249x104 /0. 1.359x10 5 /o,o Impact (BP/GB) (#)

Rack / Wall 0./0. 0./0. 0./0.

Impact (BP/GB) (#)

R6 Stress .401/.736 .421/.795 .576/1.273 Factors (Rack Base /

Support)

Max. Disp. .5717 .5709 1.7407 DX (in.)

Max. Disp. .3230 .3479 .6147 ,

DY (in.) I Max. Vert. .0823 .0802 .0909 Disp. (in.) l Max. Floor 3.934x105 3.800x105 5.877x105 Load (4 Feet) (#)

Max. Floor 180237./ 190218/ 279673/

Load (#) 108454. 110134 186242.

Vertical /

Shear 31

. o TABLE 8 RESULTS OF MULTIPLE RACK STUDIES FULLY LOADED A 1 , A2 B 1 . B 2 , RACKS WITH C0F = 0.2 SINGLE RACK B2 ITEM MULTI-RACK MODEL Design Basis Model Rack / wall at girdle Of Of bar - impact load Rack / Rack at girdle 0# .5312 x 105 bar - impact load Rack / wall at base- Of Of plate - impact lead Rack cell wall to 613. 891.

fuel assembly (per cell - impact load)

Vertical load on .6425 x 105 lb 1.378 x 105 13 pool floor from one foot i

l Rack / Rack at Of Of baseplate - impact load Max. E-W rack .126 inch .2088 inch displacement at top of rack l

l l

32

TABLE 9 RESULTS OF MULTIPLE RACK STUDIES FULLY LOADED 1A , A2 e . B1 , B2 RACKS WITH C0F = 0.8 SINGLE RACK B2 ITEM MULTI-RACK MODEL Design Basis Model Rack / wall at girdle Of Of bar - impact load Rack / Rack at girdle Of 1.17 x 105 bar - impact load Rack / wall at base- Of Of plate - impact load Rack cell wall to 612. 974. Ib fuel assembly (per cell - impact load)

Vertical load on .715 x 105 lb 2.231 x 105 lb pool floor from one foot Rack / Rack at Of Of baseplate - impact load Max. E-W rack .091 inch .4238 inch displacement at top of rack 33

TABLE 10 RESULTS OF MULTIPLE RACK STUDIES SIDE CAPS (SG) = 4.5", 5.5" FULLY LOADED1 A , A 2 , B t, B2 RACKS C0F = .8 C0F = .2 ITEM SG = 4.5" SG = 5.5" SG = 4.5" SG = 5.5" Rack / Fuel 612. 604. 613. 618.

Impact Load (per cell)

Rack / Wall 0. O. O. O.

Impact at Girdle Bar Rack / Wall 0. O. O. O.

Impact at Baseplate Rack / Rack O. O. O. O.

Impact at Girdle Bar Rack / Rack 0. O. O. O.

Impact at

, Baseplate Max. Support 71500. 78400. 64250. 65500.

Foot Load (1 foot) l 1

Max. Horiz. .0911 .1196 .126 .1491 Disp. at l Top of Rack l

(in.)

l l

34

TABLE 11 l SPENT FUEL POOL STRUCTURE MAXIMUM STRESS

SUMMARY

Loading Maximum Stress Maximum Compressive Stress Maximum Shear Stress Case (See of Rebar (psi) of Concrete (psi) of Concrete (psi)

Section MAT SF WALL SF MAT SF WALL SF MAT SF WALL SF 4.3 1 19,937 1.81 8,610 4.18 -616 6.46 -338 11.77 83 1.48 65 1.90 2 14,979 2.40 23,549 1.53 -938 4.24 -903 4.41 115 1.07 114 1.08 3 14,333 2.51 18,646 1.93 -653 6.09 653 6.10 107 1.15 115 1.07 4 18,153 1.98 18,743 1.92 -701 5.67 -444 8.96 80 1.54 40 3.07 5 20,403 1.76 32.715 1.10 1056 3.77 -1090 3.65 66 1.86 117 1.05 6 23,375 1.54 25,486 1.41 -1049 3.79 -722. 5.51 117 1.05 78- 1.58 7 20,800 1.73 12,742 2.83 -576 6.91 -524 7.59 76 1.62 55 2.24

1. Ultimate Re* oar Stress Fa = 36,000 psi

2. Ultimate Concrete Compressive Stress Fa - 3,978 psi

3. Ultimate Concrete Shear Stress Fv = 123 psi

4. SF = Safety Factor (See Section 4.5.2)

e .

4. u a E P

m

t.s a f n r 1 h T " k I

=N " $ 0~ *ek E  :*.S g "

.W?A w ' 4 4

g g

1 I 1 1 - _

N 9 N 0N 0N "4 I' O 2 O 1 1 O

8 21 OI

/

^' G AaI 8 aI 1 E

9G E S G E 9aG E 0 9

F. R R R R E

R -

4 "8

/

1 "2 >

1 1 9:0 C 4 t. .C2 n 1Y 1 1

1 4 2 l

a "6 8 9 3 1 0 4 9 Jm /

1 - I m 2m 4m 1 7

H8 D 13 C 3 C D 3 1 3 1 5 3

( m 1 1 1 3 1 "4 1 4

4 4

j

., z -

"2 t

s 1

"s

/

8 S "6 7 1 E

" 2m 1 7 I E 1 9 9 8 /

1 9 I 1 1 m 3m 2m 1 C 3 C 3 D J 5 1 1 1 1

- F 1 E

R l

w "6 {

1

/

5 6

"2

/

1 1

"8

/

7 -

KA SE 9 8 3 4

/

6 2m 3

- AR 1 m

G2 y

m 6

• CA G- 2 F 2 0 0 1 1 1 1 1

F E

4 "R

< < / 8

/

> 1

. fF

" U n r 1

L J

f r 1 }

,ee* a- r ?g  !.C 5a j .:9N

- t, 8bA 3U ~* -

N w" 8, 3 h.

N C~ . :- b3 gn u

^ ,,

3 8

s 58W= -

mNEH 29 NoO" EMOH mm

1.05" 1012" m GIROLE BAR I PITCH )

' l I o l l F

' - -# r-9- r-'-, ---

- . ,. J f -#y M r 3 : #f * - A l '

I l

I i

i, i

i i

i I

l 1

'I oj t~-g ~-

l i ' * .

l i

i i I

I , i I i

l I

l l I

li i 1 I I l 1 l  ; i '

.- l j i i i. i I i l i

i .

i i i i , i i

! l i i i  ;

i i I ' I I I i '  ; l l 8 I l l l l  ! l l l l 1 l l l l l l l  ! l l I I I l .

1 l  ! ' l l 1 I l  !

I I I I

! e i I l t i 1 l I

I 1 I 8 I I I l l i l 1 1 1 I l 1 l 1 l l I l I

' ..I i i i ..i I

..t i . .I i . .

'I i

i i i

i

, h I I

• I ' I l I i l 1 l i I l I '

I l l l l l l 1 i I

' i I I

i i i l 1 l

' I I I l l  ; I I l I I I I l

1 1

l i I

!  : l I 1 I

I i i l l l e i . I i l '

I l l l l l 1 l l l

l l i '

l l I l

l l l l I l l 1 i l

I I i l l 1 l  ; ,

l I l I l l l I l l j l l l I I i I 8 i l i I I i l I i i I 1 I i i g. . g i '  ; l i i l i i l a i I 1 i  ; i ,

l  ! I I 1 l  !  ! I l I I I l

l l I i i i l ' l 1 I i l l l l l 1

! !kh' i  ;

! !khl! !khl!

j V l l

l l ]

i i i

FIGURE 2 ]

TYPICAL RACK ELEVATION REGION 1 37

________,__._l

.04" GiRDI.E B AR -

7 m 8RG" '

PITCH I

< i 3 .

. _ q ._

44 . -.

.g A ylhl a$

5 1 IHi ,._. Ql1' , ;Hi ii qg )

s i.il sii l inii ,e , li 6 i a i  ; l

^

, # )s 1

l .

1 l -

l i l -

I

, l

- I i

' I i

i i ,

i i 1

l i i l ,

3 l -

- I I g I l l 1 l l 1 -

- I i l l l l l I

l 1 l i i e l

- l ,

l i

l i

l , j i

i i

i i , , , ,

I e l I I i i I I i i i l i l I I 3 l i

I I

l l 3 l l l I I l 1 l l I l  :

1 I I I i i i l l l g  ;

l e i i i i i i i I i I I I 1 I I I I I I I i i I i l l I I I I I i I e .

i i ___

, I I i e i fJ ' .

i g i

' i i i

a T g N I I i 1 I

l l l l- 1 I

l I I I i

- 8 I i

' I i I 3 I '

1 l 1 1 g l l l l l 1 I I I I l  : I I I I I I i l i l 1 I l 1 I I I I I j i i l i l l

l i I I i l 1 I  !

i l 1 l

I I I 1 1 1 1

I l

i I i l

I l l l ,

l i i i i '

i l l l I l i I l  : 1 I I l 1 I 1 1 I l l 1 l 1 1 I 3 I l l l . I l 1 l l I I l 1 l l I l

l l

l l 1 i l l

, i ,

1 i i g

l

, I i g l 1

l l l l l l' i

- I 8 g i .I l l l 1 -

I i l I "

l i

I I I l l l 1 l 1

~ ~

I i i I l

• l 4 4 i i l I i 1 l l t 1 I I 'l I I I 1 I I Y r  ; i  :  :  :  :  :  ; i  : i i

. . . . l . .l.

. . . . - . . . ~ . . _.. . m ... u . . -

' ^

7 r

\ . t I ' I I i 1 4 . 4 FIGURE 3 TYPICAL RACK ELEVATION REGION 2 38

- n -

~, 1 _ 1 3132"

,z. .g._,,__, __ v_v_v  : r,o_.1 ._ J WELD 7 3/4' LG W _Wi ,

t

y tel SIDE S kb $ h Y i

~! s-Y f. Y [k Y h m,_s -

cc v .a n.__-.<-,, B ASE PL ATE

_A _,

, 8 650** _

~ l

' 50 (T Y P. )

H10.12" : 05"-->

PITCH

, ..z. . 2 FUEL ASSEMBLY

- 1 1/2" R E F.

If

(TY P.)

dJ., I _D SM OTH n

ua j l I

es

i a .

l

.J  :. .

.. a g .. __q ,

, I 1 .

L_ _ _)l ,

3 r_ __ .

J I ,

p___.!__l PolSON fj J, I

.__, i l

a .

b_ - - _i .

'e I '

L -

' l '

8 E -

L _ _ _4 p_. __,

-z-  : I 2-p_ i

- - _ _ _lI 7

--I l K

1" DI A,

. l . FLOW HOLE  ;

. L_ _ _ _I p_ __q v . .

l l I II l I

B ASE PLATE l

u ) _ .,. __ ,_]

\ \

)- ..

g 6" OlA. HOL E TYP.

FIGURE 4 TYPICAL CELL EI.EVATION REGION 1 39

10.12" t .05 y INNER BOX (INTERNAL SOUARE TUBE)

PITCH I

hi vr nu NXd iAsw O fxyKhd F iv>WAwMr:r d -

-d Q

I l

kU

• M I k--$ ',

Q x

Q h '

$_ _ < _ _ _ _a

}-

}x <

--y y.

A I x l d N x

I , M4  ! *-3?, s

' AvrAm oAwl -

wx wvmxxxx - '

'uYiYm1Yrrrvi -

3 g i i i i i

• K r n_1 A xyp..wi i.A -

wvvix JA^wvw -

_r yvinywarwe. = .

8 5 _ _- M > x Y Q-

, E<

M I N  : I k N d.

l ,; e

.q l l

' wr wgx xAvr x 'L r yy y Aw<yx XM - -

xxAxAvuAAXU-i i i l I e ii A INv FNy i I d - e L I ANvWY A A/N1

- m e rViYX X 'IRX XYYYE -

i -

i --4' x

g

. .m:

1. vvyyxXwvYY>u '

lYxAvrxVY m pLMYX De NEUTRON

- ABSORBER fI COVER CONNECTING SHEET GAP ELEMENT SHEET (BORAFLEX)

FIGURE 5 3 x 3 TYPICAL ARRAY REGION 1 40

% 8 86~

~...

._ y ( TYP. 7 3/32-

_. j i 1 1

WE LD 7 3/4" Lc' k @z'

. _ . .O ,

MI510ES

'I l .l. v...u i

.1 u..

l Y. l . - -r 8 65"

..Z" "Z" SQ (T Y P. ) BASE PL ATE FUEL ASSEMBLY

- 1 1/2" R E F.

y .

n T T

)\$

l l a ,

i .

p, -I PolSON

-Q {

- I l .

n _l "i

,- - l - - I .

l .

l __ q .

p .

I-~

5 La __l .

e- -y i s h- -

i ,

p__ -q i

. "Z" t- - ' "Z"

- I -

... . l . .

[ I . --  : .- h I i l

l

- 1 l .

- 3. o,(

i.., ..

r- _ - _ g FLOW HOLE _. , , ,

" I I

.l.4-  :

l B ASE PLATE

+ .

t i i

. s" ot A. HOL E TYP. -

l l

l FIGURE 6 j i

TYPICAL CELL ELEVATION REGION 2

' 41 j i

)

INNER BOX l

, 8.86" i .04" [ ,

~ ~ ~ ~ ~ ~~~

PITCH f

! I I I  !  !

I M'2O?<X x'A !

l %X'ON'<N 1 pI W2iX'<9A I

. I i . _., _ _. .

I MdWM I l kX5&2derN l _

l WA$X$WA l _

l u

l N<<692ci l i

I W M 7< M O 3 I I W@2VM9/N I I

BOUNDARY COVER SHEET

' . i l

i I j

. ~

~ - ~

i rw.w:cwN 1 i t<ux m 1 i xmeN 1 l J l i l i I t.-.. ..., 1. . . ,

NEUTRON ABSTRBER

SHE ET (BOR AFLEX) CONNECTING STRIPS

.031" OR . 06 THK.

SEE NOTE BELOW NOTE: CONNECTING STRIPS VARY IN THICKNESS FROM

.031" TO .06" RESULTING IN AVER AGE WALL TO WALL SEPARATION OF .046"IN AS. WELDED CONDITION FIGURE 7 3 x 3 TYPICAL ARRAY REGION 2 42

1 l

I l

l l

T, 4 71/2" DI A. >

BASE PLATE 3/4" THK.

i I l' < 5" DI A. > '!

1 1 1/2 Y l 1l l

• f A ,

Yg n .

'////f f

e  %)

Q\

M '

w _

g ______s (4) 1 1/2" DI A.

-- 7 /  ; ji ii FLOW HOLES j u i iii g,, '

• j h il E i \! il i n .

T /  ;

\; '\

l J 7" 4 UNC j ,l W, g l ' CL ASS 1 A 62  ;

>l

%. yi 1

t i _- 1 lf t 91/2" DI A.

x  :

. , , , , m  ;

l s 3/4" DI A HOLE THRU 4 101/2" SQ. >

mo e

• l FIGURE 8 ADJUSTABLE SUPPORT LEG l 43

)

43  !

.----- y h ,/y ,

/ / 2*

t - --- 7 og

. .. pf q6 "'

o 3*

P10 RACK GEOMETRIC - - -

pg CENTER LINE N H/4 H u4e ,

o

>P12 -

H/4

~

. P11 8

o 5 l

. 'P14 '

/ i

/

4 S4 . .

~'

p13 H/4 S 3

Y$ __

Y *2 A o X8 9

f . . = Yg =0 ,

LONG DIRECTION -

i

'? SUPP'T 4

Sg S2 44

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

TYP. FRICTION l ELEMENT 4

g gi FIGURE 9

, SCHDiATIC MODEL OF nlEL RACK 44

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

TYPICAL TOP IMPACT ELEMENT l

i 9

9 l

I 1

R ACK STRUCTURE ,

, p --

j l

)

TYP. BOTTOM IMPACT ' ' - '

ELEMENT -

h

./ ,

~v

.j w! O _ . _

o M

7 & ~h

$4 &

N3l'

$4

~  !

1 g[a.,

FIGURE 10 WEL RACK MODEL SHOWING RACK-TO-RACK IMPACT SPRINGS 45

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

Y CELL W.L F M ASS i XB i' 7

/

/ '

FUEL ASSEhlBLY/ CELL IMPACT SPR. LNG

/ 1

/

/ .

\

/

Yg -

/ \

/  ;

/ 1

. / -

/

/ u r

. - ~ ~ .. . ~ .. . .. . . . .

6 FIGURE 11 IMPACT SPRING ARRANCDfENT

. AT NODE i 46 .

, ,, . , , - , - -ww- r, --~,w,, ,-n- - - - , , , , - , - - - - , , .-e .,,e -

. - - ,,,,,---n- ----~,-,---,,-e , , , - - , , - --mn-wm,rm.--

o c l 2

1. I d

/

K K .

2 - l I

\

l GAP ELEMENTS ,

TO SIMULATE INTER RACK IMPACTS (4 FOR 2 D MOTION 20 FOR 3 0 MOTION)

R ACK CENTROID (ASSUMED AT H/2)

H RIGIO RACK BASEPLATE O

X h K6 3 Kg K KR TIGURE 12 SPRING MASS SLMULATION FOR WO-DIMENSIONAL HOTION 47

-yw w

c v o

o 8

8 N "

o m

8 o

8 g -

^

o

. - o 8

e C,

~

< o

- a e,

2 -

a s -

< -.., 8mG c .

co e Y

W o --

• ~

., ow o 2 5 - '

r-3 -

> e u . o 4

C

$ 1 s

, l

$ u g

- o w

> o w -

8 JL-

. _.. ... . . . . . .. % - - ~ ~ - '

8 n .

% o g .

~

o 8-o 00t 0 Oct 0 002 0 8 -

00 L '0 0000- 00L 0- 00Z 0- 00t~0- 00t'0 o -

(D)NOl1VW31333V FIGURE 13

• NORTH - SOUTH SSE 48

0 u O

O b

O 8

e o

~ _ g Sub 1

4 O I N

s ., o

-~-

k- I u

8  ;

- N (s -

[

E -

4 6'

5  % -

N O M ~

1_ O O o iE -

e.

O 8a o , *

. Y

• b 8

~ ~ O 2 -

em O

w -

8 8

f C

u O U -

- 8 S

l g

M -

O

.3 8'

-- :::E=====--- I h -

l j l

r. ,,

e

~

8 w

O

~

8-0L O'0- M L' - g g- gg h Mt 0-(D) NOlivW373 coy l l

I FIGURE 14 EAST - WEST SSE 49  ;

4 9

m-- . _ . y . .. , - , _,

-9 i

o e-

~

~ o

_ 8, o

g n

o N . o o

o g

~

s W o

>= -

. o 5

s C .

e o

c 0 2

5

< ~

gG h

N cc M -

c.

~

{w5

~

a U

w o .

g - o E

8 W o S

a -.

g

.o o

- h 1 l

o l N -

g o

8 8

009 0

• 00C 0 - 00 E'O 001 0 0000 gog 0 00 E'0- 00C O- 00p o. o ,

(D) Nolivv3130cv FIGURE 15 VERTICAL SSE 50 6

\ tj!>j I !l11ll Y .

L Z 4

.I ' - , .- .

s . .

E

- 0 . G K G A0

, N L ..

S A E 7. I A R R1 . LA .

• CO EN

• T AL E ,

UA F

~. S C .

.. E -

R _

a. -

0 a .

0

• _

0 5 7 .~ 0 . . .

=. . _

1 2 .

8 4 . A _

> L - 4 L . .

0 1

E E .

2 . ,

. L *

.. E * .

4 -

a . -

,cmr ,OOr .

.e.

. )

....g 4 .

. .. g - .

s.-

y '. - 1

. .. 6 -

r>4- 'e>% (

mruoio ,

mF .- Moo plII f Iw mF Noo

/

mmoa~OZ > >

mmogn e mmps4 g;nM m8r

• ga mn>4 ?e mgH~Oz e

x GST _

N P I

. RL)

PL P SAY L H LWT A

L A T LL NH T (Y UA T O W O A I UL E SW M TO ASON L -

D D 'y T T OA

- /

RM I

M .

)i( *< T P& _

TL)

L

/ s, L AAP A MC Y x %x rl b W ' - - ' I T

T HT(

CR S

A g AE L EVAN E

@LI _

SAD _

GTU N'/ hN5 yj

/

m N OI I

NT RZ G _

\/ g v PI N. .

SRO

. \ x ' LOL I H O

S

• b

//T

/

// >

R- _

/

Eg _

N 3%

L L

A _

W )

P.

H DY T < G(T I

R O

I N RR A

s B

~ , \V5 .

ll

<\ ! } ' I L

e s

x

, ., .I' L ._

A .

W _

T .

S _

E W

$Ek m@4" f ,3 goPSf $

<