ML20214T110

Proposed Findings of Fact & Conclusions of Law Submitted by Pg&E.*Director of NRR Authorized to Make Immediately Effective,Amends 8 & 6 to Licenses DPR-80 & DPR-82, Respectively.Certificate of Svc Encl

ML20214T110
Person / Time
Site:	Diablo Canyon
Issue date:	06/04/1987
From:	Norton B PACIFIC GAS & ELECTRIC CO.
To:
References
CON-#287-3681 OLA, NUDOCS 8706100128
Download: ML20214T110 (32)

Text

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

.' 3HI L'

C%fP 1 UNITED STATES OF AMERICA '87 JtN -8 P3 :25 NUCLEAR REGULATORY COMMISSION BEFORETHEATOMICSAFETYANDLICENSIN$bAROL:i 0 3

• 4

) Docket Nos. 50-275_ O .B 5 In the Matter of ) 50-323

)

6 PACIFIC GAS AND ELECTRIC COMPANY ) (Reracking of Spent Fuel Pools)

)

7 (Diablo Canyon Nuclear Power )

Plant Units 1 and 2) ) June 4, 1987 8 )

9 PROPOSED FINDINGS OF FACT AND CONCLUSIONS OF 10 LAN SUBMITTED BY PACIFIC GAS AND ELECTRIC COMPANY 11 -

12 I. INTRODUCTION 13 By memorandum and order dated April 9, 1987, the Atomic Safety and 14 Licensing Board ordered that the parties file with the Board and serve on each 15 other proposed findings of fact and conclusions of law on or before 16 June 5, 1987. The Board also stated that the parties will be given an 17 opportunity to revise the proposed findings and conclusions after the hearing, 18 scheduled for June 15, 1987, at the San Luis Bay Inn, Avila Beach, California.

19 Since PGandE has not received the prefiled testimony of the other-20 parties, the proposed findings and conclusions presented below are based only 21 on PGandE's prefiled testimony, and will be revised as necessary following the 22 hearing.

23 There were a number of contentions admitted by the Board for hearing; 24 however, several were deleted by prior stipulation or withdrawal of the 25 petitioning parties. The contentions to be litigated at the hearing were 26 raised by the Sierra Club and relate primarily to the structural integrity of 8706100120 G70604 5 ,

PDR ADOCK 0500 c, g a

I the high density racks under a postulated Hosgri event. The specific 2 contentions to be litigated are set forth in Appendix A attached hereto.

3 Both the NRC Staff and PGandE will offer testimony and sponsor witnesses 4 for each of the contentions. The Sierra Club is expected to offer testimony 5 on all of the contentions to be litigated except for Contention I B(7), which 6 relates to evaluation of alternatives to reracking, since the sponsoring 7 witness (Dr. R. B. Ferguson) for the Sierra Club indicated that he is not an 8 expert on this subject and he does not intend to file prefiled testimony on 9 this contention. PGandE's prefiled testimony is being submitted 10 contemporaneously herewith, and a list of the exhibits which PGandE intends to 11 introduce into evidence is contained in Appendix B attached hereto.-

12 The proposed findings presented below are cross-referenced at the end of 13 each finding to the corresponding paragraphs of PGandE's prefiled testimony.

14 II. PROPOSED FINDINGS OF FACT 15 A. Desian and Analysis of Hiah Density Racks 16 1 Diablo Canyon Units 1 and 2 have separate spent fuel handling and 17 storage facilities. Each unit has a spent fuel pool with storage capacity for 18 270 spent fuel assemblies. Each pool currently contains spent fuel from the 19 first refueling outage, which occurred in late 1986 for Unit I and mid-1987 20 for Unit 2. Based upon operating schedules and the desirability of 21 maintaining full core discharge capability, it is necessary that the spent 22 fuel storage capacity for both units be increased. (A4) 23 2 Reracking with high density racks was chosen because it is a safe 24 method of increasing onsite stcrage. Further, high density reracking is the 25 most prudent, reasonable, economical, and timely method of the various l 26 techniques available to provide increased storage. (A6) 1 L

I 3 The high density spent fuel racks to be installed in the Diablo 2

Canyon fuel pools consist of a total of 16 racks of varying sizes for each 3

pool, with a total of 1324 fuel assembly storage cells plus 10 miscellaneous 4 storage locations. The number of cells range from 24 to 110 per rack and the 5

individual storage cells have an 8.85-inch (nominal) square cross section.

6 Each cell is sized to contain and protect a single Westinghouse-type PWR 17x17 7

fuel assembly and the cells are arranged with a 10.93-inch center-to-center 8

spacing in the rack modules. Stainless steel gap channels are welded between 9

the cells to provide a " honeycomb" type structure which provides considerable 10 rigidity and resistance to impact as well as to seismic loads. (A10, All)

II 4 The rack modules are freestanding, with no connection to the pool 12 floor, walls, or adjacent rack modules. The rack support feet rest on bearing 13 plates on the pool floor. Each module is equipped with a girdle bar on each 14 side near the top. The girdle bars protect the rack module from potential 15 impact loads during a seismic event and maintain a specified minimum gap under 16 all loading conditions between the cell walls of adjacent rack modules. (All) 17 5 The rack modules are specifically designed for storage of spent fuel 18 with varying amounts of burnup. There are three modules (290 cells) 19 designated as Region 1; these utilize a neutron-absorbing material, Boraflex, 20 on all sides of individual storage cells. These cells are designed for 21 storage of new fuel assemblies with enrichments up to 4.5 weight percent U-235 22 and spent fuel that has not achieved a specified burnup. There are 13 modules 23 (1034 cells) designated as Region 2; spent fuel stored in this region will be 24 required (by Technical Specifications) to have a specified minimum burnup and, 25 thus, no Boraflex is used in this region. (A12) 26 fi. The spent fuel pools at Diablo Canyon are located at each end of the l

l

I east side of the auxiliary building. Each pool is approximately 35 feet wide, 2 37 feet long, and 40 feet deep. The normal water level in the pool provides a 3 minimum of 23 feet of water above the top of the stored fuel. The concrete 4 pool walls are 6 feet thick except around the fuel transfer canal where the 5 wall is 5 feet thick. The concrete foundation of the pool has a minimum 6 thickness of 5 feet and is founded on approximately 5 additional feet of lean 7 concrete placed directly on rock. The pool walls and floor are lined with 8 stainless steel plate with a thickness of 0.25-inch on the floor and 9 approximately 0.125-inch on the walls. (A13) 10 2 The high density spent fuel racks, when fully loaded with spent fuel, 11 will increase the overall mass of the auxiliary building by less than one 12 percent. The liner plate and pool structures were evaluated for these new 13 loading conditions and found to be adequate to support and transfer the high 14 density rack reaction loads. (A14) 15 B. The NRC has established acceptance criteria and design guidance for 16 safe storage of spent fuel. The seismic design criteria and guidance are 17 primarily contained in Section 9.1.2 and Section 3.8.4, Appendix D of the 18 Standard Review Plan (SRP), and in the NRC Position Paper, "0T Position for 19 Review and Acceptance of Spent Fuel Storage and Handling Applications," dated 20 April 14, 1978. (AIS) 21 2 SRP Section 9.1.2, Paragraph III.3.a. requires that spent fuel 22 storage racks be classified and designed to Seismic Category I requirements.

23 The criteria for seismic design and fuel assembly impact loads are provided in 24 Section IV (3) of the Position Paper.Section IV (5) of the Position Paper 2S states that SRP Section 3.8.4 provides acceptable procedures for modeling and 26 analyzing the seismic responses of the spent fuel racks. Further,Section IV

1 (2) of the Position Paper identifies either of two industry codes,Section III 2 of the ASME Code or the AISC Specification, as being acceptable for deriving 3 the allowable stress criteria for the racks. Other codes are acceptable based 4 on a case-by-case review. Structural acceptance criteria are provided in 5 Section IV (6) of the Fosition Paper. This section permits rack impacts and 6 provides specific guidance on how such impacts are to be incorporated in the 7 rack design. (A16, A17) 8 The Diablo Canyon high density racks comply with the applicable 10 9 seismic design criteria in that:

10 . The racks were designed as Seismic Category I components in accordance II with SRP Section 9.1.2, Paragraph III.3.a.

12 .

The allowable stress criteria for the racks were derived from the 13 Subsection NF requirements of the ASME Code for Class 3 component 14 supports. Construction materials conform to Section III, Subsection NF 15 of the ASME Code and were selected to be compatible with the fuel pool 16 environment.

17

• The seismic excitation was simultaneously applied in three orthogonal

, 18 directions. Increased damping of fuel racks due to submergence in the 19 spent fuel pool was not considered for conservatism. Local impact of the 20 fuel assemblies within the spent fuel rack cells was considered in a 21 manner which maximized forces acting on a rack module.

22

• The procedures used for modeling and analyzing the seismic responses of 23 the Diablo Canyon spent fuel racks were consistent with the requirements 24 of the Position Paper. The models were developed based on current 25 engineering practices.

26

• The possibility of gross sliding, tilting, and rack impacts under the 1

1 postulated Hosgri event were evaluated in accordance with the acceptance 2 criteria specified in Section IV (6) of the Position Paper. (A18) 3 11.. No exceptions to acceptance criteria were taken for the design of the 4 Diablo Canyon high density spent fuel racks. The racks were designed and s

5 constructed using the approved acceptance criteria to maintain the spent fuel 6 assemblies in a safe configuration for normal and abnormal loads, including 7 potential impacts between racks and between the racks and the fuel pool walls, 8 which may occur during a Hosgri event. (A18) 9 12 The analytical model developed by PGandE for high density rack 10 analysis was a nonlinear dynamic model, and appropriately considered 'the 11 potential effects of the following possibilities: movement of the fuel 12 assemblies, frictional resistance at the base of the rack, rack sliding and 13 rocking behavior, rack uplift and subsequent impact on the bearing plate, and 14 rack impacts with adjacent racks and pool walls. In addition to the potential 15 rack movements addressed in the analysis, fluid effects, known as hydrodynamic 16 coupling, were also considered. (A19) 17 13 Potential impact of the fuel assemblies within storage cells was 18 simulated by impact springs (designated as Kg ). These impact spring 19 constants were selected based on a series of parametric studies to assess the 20 irrpact of a fuel assembly on the cell wall of the rack. The cumulative impact 21 loads of the fuel assemblies on the rack module were calculated by assuming 22 that all fuel assemblies in the rack move in unison; 1.e., they will impact 23 the cell walls of the rack at the same instant and in the same direction. (A20) 24 14 Rack sliding behavior was addressed in the model. A set of springs, 25 Kg , was included in the model near the base to simulate the sliding friction 26 of the rack, whereas other springs, K ,6 combined with gap elements.

- - - - - - - _ - - - - - - - - - - - - - - )

I simulate lift-off resulting from rocking. Another set of springs, K R, was 2

included to capture the rotation of the leg in the vertical plane. A further 3

set of springs, K ,g was included in the model to determine rack-to-rack'and -

4 rack-to-wall impact forces. (A21) 5 15 Friction coefficients of 0.8 and 0.2 which bound the experimental 6

data were used in the analysis to maximize the inertial force and horizontal 7

displacement of the racks. This wide range of friction values is typically 8

used in the industry for rack design. (A21) 9 16 Fluid inertial effects, produced by rack motion, were also addressed 10 in the model. In particular, the accelerating fluid mass results in two types II of inertial effects. As a rack starts to slide, the water inside and 12 surrounding the rack is set in motion. This produces an additional inertial I3 force on the rack, which was addressed in the analysis by adding an 14 appropriate amount of water mass, known as " virtual mass," to the mass of the 15 rack and fuel assemblies. The second effect of the accelerating fluid mass is 16 hydrodynamic coupling. As the space between moving racks or between the racks I7 and adjacent walls is reduced, the fluid between the bodies is expelled from 18 that space. This causes fluid pressures to develop on the surfaces bounding I9 the fluid mass, which retards the seismic motion of the racks. The effects of 20 the fluid motion on rack displacements are determined by the kinetic energy of 21 the fluid. By underestimating the kinetic energy of the fluid, one 22 necessarily overestimates the rack displacements. If the kinetic energy of 23 the fluid is ignored completely (e.g., assuming the absence of fluid), one 24 will grossly overestimate the rack displacements. The calculation method used 25 for rack analysis includes fluid motion but underestimates the fluid kinetic 26 energy and, accordingly, overestimates rack displacements; 1.e., the e

I calculation method is conservative. PGandE's use of virtual mass and 2 hydrodynamic coupling in the analysis is based on the fundamental principles 3 of fluid dynamics. (A22) 4 11 The analytical process used in the design of the racks consisted of:

5

• Development of a nonlinear dynamic model of a rack module consisting of 6 inertial mass elements, hydrodynamic coupling, and gap and friction 7 elements.

8

• Generation of the equations of motion and inertial coupling and solution 9 of the equations using a computer program, DYNAHIS, to determine rack 10 forces, moments, and displacements. -

11

• Computation of the detailed stress field in the rack (at the critical 12 locations) and in the support legs using the forces, moments, and 13 displacements calculated in the previous step. (A23) 14 18 Using the methodology described above, PGandE calculated the 15 potential loads on the racks. These calculations were performed in conformity 16 with the loading combinations and acceptance criteria specified in the NRC 17 Staff's Position Paper and Section 3.8.4,. Appendix D, of the Standard Review 18 Plan. The loading combinations included the combined effects of dead load, 19 live load, thermal interaction within the pool, and seismic inertia loads due 20 to seismic events. A series of rack loading cases (fully loaded, partially 21 full) was considered in order to establish the design loads. The resulting 22 stresses in the racks were determined to be lower than the allowable stress 23 values permitted by acceptance criteria. These allowable values provide a 24 sufficient factor of safety when compared with the ultimate capacity of the 25 racks. (A24) 26 12 The design basis analysis was performed with a single-rack model.

1 Conservatisms were built into the evaluations performed for the high density 2 racks in terms of modeling assumptions, postulated loadings, and safety 3 margins on stress allowables. Several of the conservatisms inherent in the 4 design basis analysis are:

5

• Adjacent racks were assumed to move in a manner equal and opposite (out l

6 of phase) to the rack module being analyzed, thereby maximizing the 7 potential for rack-to-rack impact.

8

• A value of 4 percent damping was used between the fuel assemblies and 9 racks, between adjacent racks and between racks and walls. The analyses 10 neglected fluid damping. A value of 10 percent for impact damp 16g (in 11 addition to structural damping) has been used at other plants licensed by 12 the NRC.

13

• The impacts between cell walls and the fuel assemblies were assumed to 14 occur in phase. In reality, the fuel assemblies exhibit complex and 15 random behavior. However, they were all assumed to move in unison so 16 that the maximum response could be obtained.

17

• The form drag opposing the motion of the racks within the pool water was 18 conservatively neglected.

19

• The fluid coupilng coefficients were calculated based on the conservative 20 assumption that the adjacent rows of racks are an infinite distance away 21 (the distance measured perpendicular to the horizontal ground motion).

22 This reduces the " cross-coupilng effect" of the adjacent rows of racks 23 and yields conservative displacements and impact forces.

24

• The calculation of fluid inertial effects included an underestimate of 25 the fluid kinetic energy and resulted in a conservative overestimate of 26 rack displacement.

3 1

• Hydrodynamic coupling coefficients used in the analysis neglected.certain 2 nonlinear 1 ties of the motion. Studies in the 11terature show tnat 3 incorporation of these nonlinear effects would significantly lower rack 4 response. (A25) 5 20 Industry practice with regard to high density rack design and 6 analysis was reviewed by PGandE for 10 other U.S. nuclear plants that have 7 freestanding spent fuel racks. The rack suppliers for these plants included 8 Joseph Oat Corporation; Exxon Nuclear Company; GCA Corporation, par Systems; 9 Nuclear Energy Services, Inc.; Westinghouse Electric Corporation; and General 10 Electric Company. The analytical techniques used for designing those~ racks 11 were similar to the techniques used for the Olablo Canyon rack analysis. (A27) 12 B. Contention 1 13 Contention I(A)3 14 21 Data regarding the velocity and displacement of the fuel pools as a 15 function of time in three dimensions for the postulated Hosgri earthquake is 16 not necessary for review by the NRC Staff in evaluating the technical adequacy 17 of the rack design since the acceleration time-histories are used for that 18 purpose. Consequently, the velocity and displacement time-history data for 19 the fuel pools were not included in the Roracking Report because a record of 20 such data was not required during the design process. (A28) 21 22 The design process for the racks utilized the postulated Hosgri 22 earthquake acceleration time-histories for the base of the spent fuel pool.

23 Velocity and displacement information can be derived from the acceleration 24 time-histories used in the design, which are contained in the Reracking 25 Report, Figures 6.1.1, 6.1.2, and 6.1.3. (A28) 26 ///

1 Contention f(A)4 2 23 The maximum velocity of the racks is not documented in the Reports 3 since it is not a value needed for design of the racks. However, the maximum 4 displacement for a loaded rack module is included in the Reracking Report in 5 Table 6.8.2. (A29) 6 24. In the design basis analysis, the maximum displacement relative to 7 the pool structure of a loaded rack module is approximately 2.8 inches. The 8 maximum displacement relative to the pool structure of a nearly empty module 9 is approximately 4.2 inches. (A30) 10 Contention I(B)2 11 25 The rack analysis performed by PGandE considered potential resonant 12 behavior of fuel assemblies. The design basis analysis performed to evaluate 13 the fuel racks utilized a mathematical representation of the various 14 components and their response behavior. Since resonant behavior is a 15 fundamental condition described by the equations of motion, and since the 16 equations of motion were appropriately represented, the analysis considered 17 the possibility of resonant behavior. (A32) 18 26 The design basis analysis demonstrated that, due to the specific 19 conditions present, the fuel assemblies do not experience resonant behavior.

20 These conditions include the nonlinearities of the system (including the 21 presence of water, the movement of the fuel assemblies within the fuel racks, 22 and the presence of friction at the fuel rack base). The analysis 23 appropriately represented these physical conditions and demonstrated that the 24 integrity of the racks is maintained. As a practical matter, resonance will 25 not occur since the amplitude cannot increase beyond the 0.302 inch clearance 26 between the fuel assembly and cell wall. (A34)

- II -

I Contention I(B)7 2

22 PGandE evaluated the two methods of onsite storage facilities 3

mentioned in the contention. The evaluation was brief since these two ' -

4 specific methods, additional storage facilities and acquisition of modular 5

storage equipment, do not offer any increase in safety over high density 6

racks. (A37) 7

23. A discussion of alternatives is documented in PGandE's Reracking 8

Report, Chapter 9. In particular, Olablo Canyon was designed to store spent I

fuel' for a nominal period of one year and then ship the fuct offsite for 10 reprocessing or disposal. Due to the unavailability of fuel reprocessing II fact 11 ties and of permanent disposal sites, the spent fuel must now be stored 12 for an extended period of time at Olablo Canyon. Therefore, the alternatives I3 that must be considered, in addition to onsite storage, consist of various I4 methods of storing the spent fuel offsite or shutting down the reactor. The 15 consideration of alternatives, including offsite shipment of spent fuel and I6 shutdown of the reactor, was documented in the Reracking Report, Chapter 9 I7 While the onsite storage alternative was chosen, there are no regulations 18 which specify the nature of onsite storage methods that must be considered or II documented. The discussion included in the Roracking Report was sufficient to 20 comply with NRC requirements. However, the Reracking Report did not 21 specifically address PGandE's evaluation of other onsite storage methods 22 because the expansion of the spent fuel pool capacity through reracking was 23 clearly superior. (A38) 24 An additional storage pool was considered less attractive because it 22.

25 would not provide any added safety for spent fuel storage than with properly 26 designed high density racks in the existing pools. Moreover, the costs of I

constructing a new seismically qualified structure and auxiliary support 2

systems would obviously be very high compared with rcracking. Finally, this 3

would involve increased handling of the spent fuel. (A39) 4 3D. Acquiring modular storage equipment was considered less attractive 5

because such equipment would not provide any added safety over and above 6

properly designed high density racks. Further, modular equipment such as dry 7

cask storage was not a licensed concept at the time the reracking decision was 8

made by PGandE, and casks were still being tested. In any event, dry cask 9

storage is not a viable option for Diablo Canyon based upon the design of the 10 dry casks currently available. The dry casks are designed to store only fuel II that has been discharged from the reactor at least five years prior to cask 12 storage. Thus, this storage method could not be used for at least five years I3 following the first refueling outage.

14 The existing low density racks at Diablo Canyon were originally designed, 15 in accordance with early NRC guidelines, to accommodate spent fuel discharged 16 from one refueling (roughly 70 assemblies), plus a reserve capacity of a full I7 core offload (193 assemblies) in the event a full core discharge is necessary.

18 The storage space associated with one refueling discharge is currently occupied I9 at Diablo Canyon Units I and 2 after the first refueling outages. Based upon 20 operating schedules and the desirability of maintaining full core discharge 21 capability, it is necessary that the spent fuel storage capacity for both 22 units be increased. Further, the cost of the casks, assuming their avail-23 ability, which would be required for the needed capacity at Diablo Canyon 24 would be high compared with the reracking alternative. At the time that PGandE ,

25 made the reracking decision, there were no plants in the United States using 26 modular storage facilities for spent fuel storage. Subsequently, two plants I

were licensed to use modular storage facilities such as dry casks, but these 2

plants did so only when all of the storage space in existing pools had been 3

filled after they had previously reracked with high density racks. (A40) 4 Contention f(B)8 5

31 The use of anchors, braces, or other structural members to prevent 6

rack motion is not discussed in the Reports since freestanding racks meet 7

safety requirements. (A41) 8 32 Structural anchors, braces, or other structural members are not 9

required to prevent rack motion and potential subsequent rack damage. The 10 freestanding racks satisfy NRC criteria and guidance applicable to spent fuel II storage racks. The design accommodates the calculated rack motion during the 12 postulated Hosgri earthquake and shows that the racks have sufficient safety 13 margins. In addition, freestanding racks have several advantages over 14 anchored or braced racks. Particularly, freestanding racks reduce the stress 15 on the liner caused by thermal loads from the heat generated by the spent 16 fuel. Further, sliding provides a very effective means to dissipate energy.

17 A freestanding rack is, therefore, considered a better design to absorb 18 seismic energy and, thus, has a distinct advantage over anchored or braced 19 racks. Further, no welding is required to install the freestanding racks.

20 Finally, inspection and/or replacement of racks, if necessary, is simpIlfled 21 by the use of freestanding racks. (A42, A43) 22 Contention _ EMS 23 The high density racks do not uttilze Boraflex neutron-absorber in 33 24 all modules. These racks utilize the two-region concept in which one region 25 (Region 1) is designed for new fuel or low burnup fuel, and a second larger 26 region (Region 2) is designed for fuel of a specified minimum burnup. The f

t b

I spent fuel racks were arranged in a safe and cost-effective design utilizing ,

a 2 the space available in the storage pool. (A44) 3 33 The use of neutron-absorbing material between cells in the storage l

4 rack reduces the spacing required between fuel assemblies while still assuring 1

5 an acceptable margin to criticality. Both water-spacing and neutron-absorbing (

j 6 material (or a combination) serve the purpose of reducing the neutron 7 multiplication factor te a safe leval, and the use of Boraflex, as in

! 8 Region 1, allows more fuel to be stored in a given space. In Region 2 j 9 Boraflex is not needea since the spent fuel to be stored is restricted to I

i 10 lower reactivity fuel of a specified minimum burnup. (A45) ~

11 35 The use of Boraflex in Region 2 would not have afforded any advantage i

12 in operation or safety. The racks were designed to provide the spent fuel

13 storage capab111ty needed by PGandE. Further reduction in spacing by using 14 Boraflex in Region 2 was unnecessary. Furthermore, both regions were designed j 15 to provide an adequate safety margin to criticality and to satisfy NRC 16 acceptance criteria for criticality safety. (A46) i  !

17 Contention II(A)1. 2. 3 18 .16 During the postulated Hosgri event, there will be no collisions l

! 19 between the racks and the pool walls that will result in impact forces on the 1

l 20 racks significantly larger than those estimated in the Reports. The Reports l j 21 describe the analyses performed for a wide range of collision scenarios and

, t 22 the resulting impact forces are expected to bound those that could reasonbly

) 23 occur. Further, there will be no collisions between the racks and pool walls l

l 24 that will result in impact forces that could cause significant permanent i l

l l

25 deformation and other damage to the racks and pool walls. The calculated l 26 forces and stresses are less than those which would cause significant I

I 1

1

I permanent deformation or other damage to the racks and pool walls. Hinor 2 local damage to the liner, concrete pool wall, or racks may occur, but such 3 damage does not adversely affect the integrity of the fuel pool, racks, or 4 fuel. (A47, A48) 5 32 The design basis analysis and the techniques used to determine 6 potential impact forces were based on a conservative mathematical 7 representation of the racks to ensure an upper-bound impact force for 8 rack-to-rack and rack-to-wall impact. Some of the conservative assumptions 9 include:

10 e Adjacent racks were assumed to move in a manner equal and opposite (out 11 of phase) to the rack module being analyzed.

12 . A value of 4 percent damping was used between the fuel assemblies and 13 racks, between adjacent racks, and between racks and walls. The analyses 14 neglected fluid damping. A value of 10 percent for impact damping (in 15 addition to structural damping) has been used at other plants licensed by 16 the NRC.

17 e The impacts between cell walls and the fuel assemblies were assumed to 18 occur in phase. In reality, the fuel assemblies exhibit complex and 19 random behavior. However, they were all assumed to move in unison so 20 that the maximum response could be obtained.

21

• The form drag opposing the motion of the racks within the pool water was 22 conservatively neglected.

23

• The fluid coupling coefficients were calculated based on the conservative 24 assumption that the adjacent rows of racks are an infinite distance away 25 (the distance measured perpendicular to the horizontal ground motion).

26 This reduces the " cross-coupling effect" of the adjacent rows of racks i

4 i

I '

and yields conservative displacements and impact forces.

2 *

The calculation of fluid inertial effects included an underestimate of 3

the fluid kinetic energy and resulted in a conservative overestimate of 4

l rack displacement.

5

• Hydrodynamic coupling coefficients used in the analysis neglected certain j 6

j nonlinearltles of the motion. Studies in the literature show that f incorporation of these nonlinear effects would significantly lower rack

8 response.

l 9 '

In addition, the dynamic interaction between the pool wall and the 10 peripheral racks was considered in the rack-to-wall interaction. The' II f evaluation showed that the rocking frequencies of the rack (approximately 10 t 12 cycles per sqcond) are significantly different from the natural frequencies of f I3 the pool walls (greater than 30 cycles per second). Such a large differenco I4 l in the frequancies precludes the possibility for any significant amplification 15 of the impact force.

16 Several parametric studies were performed by PGandE that included both I7 simplified and complex two-dimensional, single- and multi-rack analytical 18 l models, as well as enhancements to the original design basis, I 19 three-dimensional, single-rack model. The results of these studies confirm in 20 all cases that rack impact loads and stresses due to the postulated Hosgri

[

t ,

! 21 earthquake are below allowable values. Therefore, the design basis evaluation i 1 t i 22 was conservative and the high density spent fuel racks satisfy acceptance I 23 criteria and will maintain their integrity for the postulated Hosgri event.  !

24 (A49)

I j 25 33 While impact forces are important to the design process, of moro

) 26 significance are stress ratios in that they better reflect the effect of l

! l 1

h i

t I  !

1 impacts on the racks. The controlling stress ratios for the racks have an 2 allowable value of 2.0. The highest stress ratto for the impacts determined 3 from the design basis anaysis was 1.436. For the impacts determined from the 4 parametric studies, the highest stress ratio was 0.743. Thus, the design 5 basis evaluations were shown to be conservative and bounding. (A50) 6 32 In evaluating the walls and the rack components, impact loads were 7 conservatively assumed to be static. No credit was taken for the short 8 duration of the loading. Stresses derived from these calculated forces were 9 significantly smaller.than the stresses the racks and walls are capable of 10 withstanding without any adverse effect. (ASI) 11 10 Decause of the conservative assumptions and methods used to analyze 12 rack-to-rack and rack-to-wall impact forces, the resulting impact forces on 13 the racks bound those that might occur during the postulated Hosgri event.

14 (A52) 15 41 If a rack should impact an adjacent rack or the wall, the impact 16 force would occur at the girdle bar or at the baseplate. The fuel rack 17 strength at the girdle bar level is significantly greater than that required 18 to resist the design loads. As the rack impacts the wall, the rack girdle 19 bars perpendicular to the wall would be loaded in compression by direct 20 bearing. These bars can sustain a direct impact load greater than 175,000 21 pounds each before the onset of yielding, and incipient failure is at least 22 twice the yloid force. The impact resistance along the girdle bar which 23 impacts flat against the wall is greater than 20,000 pounds per storage cell.

24 Hith regard to the baseplate, its resistance is substantially greater than 25 that for the girdle bars. (A53) 26 42 Impact forces result in loads being applied to the rack girdle bars

1 and baseplates. These forces were compared with the calculated capacities of 2 the girdle bars and baseplates. For all cases, including straight-on impacts 3 and those where the corner of one rack impacts an adjacent rack away from the 4 corner, the calculated capacities are greater than the maximum expected impact 5 forces. In addition, impact forces generate stresses in the rack body and 6 support feet. For all important structural members, stress levels were 7 determined to be within acceptance criteria. Finally, the effects of impacts 8 between racks and the pool walls were also evaluated. Local and overall 9 stresses were evaluated for both the pool liner and concrete pool walls.

10 While minor local liner or concrete deformations may occur, overall stresses 11 were found to be within allowables. Thus, based on conformance with NRC 12 acceptance criteria, the structural integrity of the spent fuel pool and racks 13 is assured. (AS4) 14 43 The racks are rugged, stainless steel, honeycomb-type structures.

15 Although minor deviations from manufacturing tolerances may exist for such 16 components, the effects of such minor deviations are not significant and are 17 accommodated by conservatisms in the analysis methodology. In addition, the 10 racks are fabricated from a very ductile steel which makes minor deviations 19 insignificant. (A55) 20 M. While there may be minor local deformation to the racks or pool 21 walls, there would be no permanent deformation or other damage that would lead 22 to criticality, radiological releases, damage to the fuel, increases in heat 23 generation, or otherwise adversely affect the public health and safety. (A56) 24 Contention _II(AL4 25 45 Each Olablo Canyon fuel assembly consists of a 17 x 17 array of l 26 cylindrical rods of which 264 rods contain fuel pellets. The assembly is

I approximately 8.4 inches square and 13.3 feet in length. Each fuel rod is a 2 Zircaloy tube containing uranium dioxide fuel pellets. Grids are positioned 3 at vertical intervals along the length of the fuel assembly to maintain the 4 rod spacing. (A57) 5 M. The active fuel region is the region within the fuel assembly which 6 contains fuel pellets. This region extends 144 inches, from approximately 7 3 inches above the bottom of the fuel assembly nozzle, which rests on the rack 8 baseplate, to approximately 10 inches below the rack girdle bars. (A58) 9 d2 The maximum forces generated by the postulated Hosgri earthquake will 10 not result in a reduction of the design spacing of 10.93 inches between the 11 active fuel region within any rack module. This spacing, with its tolerances, 12 was used in the criticality analysis. (A59 A60) 13 Contentloa_1HA15 14 48 Criticality analyses were performed for the Olablo Canyon high 15 density spent fuel storage racks to assure that a k,77 equal to or less than 16 0.95 is maintained when the racks are fully loaded with fuel of the highest 17 anticipated reactivity in each of two regions and when the pool is flooded 18 with unborated water at a temperature corresponding to the highest 19 reactivity. The maximum calculated rtactivity includes a margin for 20 uncertainty in reactivity calculations and in mechanical tolerances, 21 statistically combined, such that the k,gg will be equal to or less than 22 0.95 with a 95 percent probability at a 95 percent confidence level. (A61) 23 49. The Olablo Canyon spent fuel pools will be continually maintained at 24 a boron concentration of at least 2000 ppm as require by the plant Technical 25 Specifications. This soluble boron not only provides an additional and very 26 large subertticality margin under normal storage conditions, but precludes the 1

possibility of exceeding a k,77 of 0.95 under credible abnormal conditions.

2 including the postulated Hosgri event. (A62) 3 50 The spacing requirement to maintain k,gg less than 0.95 without 4

borated water is essentially the fuel assembly spacing in the rack design 5 (10.93 inches), based upon the criticality analysis described in Section 4.0 6

of PGandE's Reracking Report. With borated water normally present in the 7

spent fuel pool, the k,77 would not reach 0.95 untti the water gap between 8

storage cells in Region 1 (nominally 1.786 inches) has been reduced to less 9

than 0.1 inch uniformly everywhere, an extremely implausable condition. While 10 analyses have demonstrated that significant rack deformation would not occur, 11 even if it were assumed that there was zero gap between storage cells, the 12 resulting configuration would still not be critical. In Region 2, reducing 13 the gap between storage cells to zero from the nominal 1.9 inches would not 14 result in k,gg exceeding 0.95. (A63) 15 51 With unborated water in the spent fuel pool, the highest k,77, 16 including an allowance for uncertainties and manufacturing tolerances, was 17 calculated to be 0.920 in Region I and 0.938 in Region 2. Both calculations 18 are based upon conservative specifications of fuel enrichment and burnups and 19 provide subtriticality margins greater than that required by NRC regulations.

20 Hith the normal concentration of soluble boron present (2000 ppm), the safety 21 margin below criticality is much larger, with the maximum k,gg being less 22 than 0.75 in both regions. There are no postulated collisions or plausible 23 reductions in spacing that could result in k,gg exceeding the limit of 0.95.

24 (A64, A65) 25 Contention _Il W 6 26 52 Any postulated condition that would cause the release of radiation I would require the fuel cladding to rupture; however, fuel cladding rupture 2 cannot occur unless the fuel assembly grids are crushed. In this case, the 3 calculated impact forces are not large enough to cause crushing of the grid 4 and rupture of the cladding. (A66, A67) 5 53 During the postulated Hosgri event at Diablo Canyon Units 1 and 2, 6 due to the motion of the rack module relative to the motion of the fuel 7 assemb11es, the fuel assemblies in the spent fuel pool storage racks could 8 contact the stainless steel walls of the storage cells. However, the maximum 9 impact force on a fuel assembly grid has been calculated to be only 10 approximately 1700 pounds and the maximum fuel rod bending stress has~been 11 calculated to be only approximately 800 pst. (A68) 12 54 The structural integrity of the fuel assembly was evaluated by 13 comparing the calculated forces against capacity determined from analytical 14 and experimental data. Specifically, the maximum 1.mpact force on the grid, 15 the fuel rod bending stresses due to flexure, and the fuel rod local contact 16 forces at the grid supports woro evaluated. (A69) 17 55 Dynamic impact tests have been performed by Hestinghouse on fuel 18 assembly grids for all Westinghouse 17 x 17 fuel assembly designs to determine 19 their ultimato strength, which is the load at which incipient plastic 20 deformation of the grid cells occurs. The evaluation showed that the safety 21 factor for the grids, which is defined as the ratto of the ultimate grid 22 strength (i.e., greator than 3400 pounds) divided by the maximum impact force 23 applied to the grid, was greator than 2. The evaluation of fuel rod bending 24 stresses showed that the ratto of the fuel rod allowable stress limit 25 (i.e., greater than 16000 psi) to the maximum calculated stress during the 26 Hosgri event is greater than 20 for all Hostinghouse 17 x 17 fuel assembly 1

l.' '

I designs.

2 The maximum local contact force that a fuel rod can sustain without 3 cladding failure was calculated by Westinghouse employing finite element l 4 analysis methods. A finite element model of the fuel rod was formulated which 5 consists of discrete elements, each of which has stress and deflection

! 6 characteristics defined by stress-strain theory. The calculated local stress 7 levels caused by the reaction force were well below the allowable stress 8 levels in the fuel rods, ensuring that the integrity of the fuel cladding will 9 be maintained during.the Hosgri event. Thus, the integrity of fuel assemblies 10 stored in the high density spent fuel racks at Diablo Canyon will be ~

11 maintained, and there can be no resulting release of large quantitles of heat 12 and radioactive material. (A70) 13 14 Coatention IICA)7. 8. 9 l 15 M. The racks have been quallfled to withstand the impact loads which may 16 result from collisions between racks and pool walls during the postulated 17 Hosgri earthquake. Therefore, no damage to the fuel would occur, and there l

18 can be no resulting releases of large quantitles of heat and radioactive 19 material. Additionally, the racks will maintain the fuel assemblies in a 20 subcritical configuration even during any such collisions, and releases due to 21 criticality in the pools cannot occur. Consequently, no radioactive 22 contamination of humans and other living things in the vicinity of the plant 23 above the levels permitted by federal regulations could result from collisions l 24 between the racks and the pool walls during the postulated Hosgri earthquake.

25 (A71) 26 ///

I Contention II(n 2

52 Because of the dissimilarity of the racks (in terms of geometry.

3 tolerances, and gap spacings) it is highly unlikely that groups of racks would 4

move as a unit under a random seismic motion. (A73) 5 53 PGandE did, however, conduct several multi-rack parametric studies, 6

which confirmed that the design basis analysis had resulted in a conservative I

rack design. (A74) 8 52 The parametric studies on multi-rack interactions utilized realistic I

modeling assumptions and evaluated variations of all key parameters that might 0

affect the qualification of the racks. Some of these parameters include II loading of the racks, hydrodynamic coupling coefficients as they apply to the 12 specific location of the rack, manufacturing tolerances, and friction I3 coefficients. These studies show that the loads on the racks are comparable I4 to those predicted by the design basis analysis, and, in all cases, these 15 loads are signficantly lower than the allowables. Thus, the parametric I0 studies confirm that PGandE's modeling assumptions in the design basis 37 analysis adequately represent potential group behavior of the racks. All I8 potential collision conditions under the postulated Hosgri event are bounded II by the loads for which the racks have been qualified. (A75, A76) 20 21

!!!. CONCLUSIONS OF LAH 22 The Board has considered all the evidence submitted by the parties 23 and the entire record of this proceeding. That record consists of the 24 Commission's Notice of Hearing, the pleadings and testimony flied by the 25 parties, the transcript of the hearing, and the exhibits received into 26 evidence. All issues, arguments, or proposed findings presented by the

I parties, but not addressed in this decision, have been found to be without 2 merit or unnecessary for this decision. Based upon the foregoing findings 3 which are supported by reliable, probative, and substantial evidence as 4 required by the Administrative Procedure Act and the Commission's Rules of 5 Practice, and upon consideration of the entire evidentiary record in this 6

proceeding, the Board, with respect to the issues in controversy before us, 7 concludes that:

8 1 Pacific Gas and Electric Company has fully met its burden of proof on 9 each of the contentions decided in this Decision.

10 2 There is no reasonable alternative to the proposed reracking of the 11 Diablo Canyon Units I and 2 spent fuel pools if shutdown of Diablo Canyon 12 Power Plant is to be avoided in the near future. Further, there is no 13 assurance that any of the alternatives can or will become available in a time 14 frame such that shutdown could be avoided.

15 3 There is reasonable assurance that the analysis and design of the 16 proposed high density spent fuel racks for Diablo Canyon Units 1 and 2 were 17 performed in accordance with regulatory requirements and Ccmmission 18 guidelines, and that the racks and pool structures meet applicable licensing 19 criteria.

20 4 There is reasonable assurance that the Diablo Canyon Power Plant 21 spent fuel pools can be reracked without endangering the health and safety of 22 the public, as authorized by License Amendment Nos. 8 and 6 to License Nos.

23 DPR-80 and DPR-82, respectively, issued by the NRC Office of Nuclear Reactor 24 Regulation on May 30, 1986. Accordingly, the Board affirms the issuance of 1

1 25 said Amendment Nos. 8 and 6, and additionally concludes that no modifications l

l 26 thereof or additional conditions are required.

1 5 The activities authorized by License Amendment Nos. 8 and 6 are not 2 inimical to the common defense and security or to the health and safety of the 3 public.

4 It Is Ordered, in accordance with the Atomic Energy Act of 1954, as 5 amended, and the Commission's regulations, and based on the findings and 6 conclusions set forth herein, that the Director of Nuclear Reactor Regulation 7 is authorized to make immediately effective License Amendment Nos. 8 and 6 to 8 License Nos. DPR-80 and DPR-82, respectively, consistent with the Board's 9 decision in this case.

10 -

11 -

12 13 Respectfully submitted, 14 H0HARD V. GOLUB RICHARD F. LOCKE 15 Pacific Gas and Electric Company P. O. Box 7442 16 San Francisco, California 94120 (415) 781-4211 17 BRUCE NORTON 18 c/o R. F. Locke P. O. Box 7442 19 San Francisco, California 94120 (415) 781-4211 20 Attorneys for 21 Pacific Gas and Electric Company gg (

• KW .

By G hu.u 4t 23 Bruce Norton 24 DATED: June 4, 1987 25 26 APPENDIX A l Sierra Club Contentions Contention I I(A). It is the contention of the Sierra Club, Santa Lucia Chapter (Sierra Club), that the report submitted to the Nuclear Regulatory Commission (NRC) entitled Reracking of Spent Fuel Pools Diablo Canyon Units 1 and 2 and other communications

. between Pacific Gas and Electric Company (PGandE) and the NRC d

which are available to the public on the same subject (the Reports) fail to contain certain relevant data necessary for l independent verification of the claims made in the Reports regarding consistency of the proposed reracking with the protection of the public health and safety, and the environment.

In particular, the Reports fail to contain data regarding

3) the expected velocity and displacement of the spent fuel pools (pools) as a function of time in three dimensions during the postulated Hosgri earthquake (PHE); .

! 4) the expected maximum velocity and displacement of the L racks obtained from the computer modeling of rack behavior during the PHE; l

I(B). It is the contention of the Sierra Club that the Reports fail to include consideration of certain relevant conditions, phenomena and alternatives necessary for independent verification of claims made in the Reports regarding consistency of the proposed reracking with public health and safety, and the environment, and with federal law.

In particular, the Reports fail to consider:

2) the resonant behavior of the spent fuel assemblies in the racks in response to the PHE and the consequences of such -

behavior;

7) alternative on-site storage facilities including:

(1) construction of new or additional storage facilities and/or; (ii) acquisition of modular or mobile spent nuclear fuel storage equipment, including spent nuclear fuel storage casks;

8) the use of anchors, braces, or other structural members to prevent rack motion and subsequent damage during the PHE;

9) the use of "boraflex" neutron absorbing material for all spent fuel racks.

A-2

Contention II II. It is the contention of the Sierra Club that the proposed reracking is inconsistent with the protection of the public health and safety, and the environment, for reasons which include the following:

A) during the PHE, collisions between the racks and the pool

~

walls are expected to occur resulting in:

1) impact forces on the racks significantly larger than those estimated in the reports;

2) impact forces on the racks significantly larger than those expected to damage the racks;

3) significant permanent deformation and other damage to the racks and pool walls;

4) reduction of the spacings between fuel assemblies;

5) increase in the nuclear criticality coefficient k(eff) above 0.95; i 6) release of large quantities of heat and radiation;

7) radioactive contamination of the nuclear power plant and its employees above the levels permitted by federal regulations;

8) radioactive contamination of the environment in the vicinity of the nuclear power plant above the levels permitted by federal regulations; 1

A-3

.. _ u. . ._ - _ _

+

4

9) radioactive contamination of humans and other living things in the vicinity of the nuclear power plant above the levels permitted by federal regulations.

B) during the PHE, collisions between groups of racks with each other and/or with the pool walls are expected to occur

, with results similar to those described in II(A) above.

i se e

t

\

I i

A-4

APPENDIX B Exhibits To-Be Introduced Into Evidence

1. PGandE Letter DCL-85-333, October 30, 1985; License Amendment Request 85-13, Reracking of Spent Fuel Pools.

2. PGandE Letter DCL-85-306, September 19, 1985; Reracking Report.

3. PGandE Letter DCL-86-019, January 28, 1986; Additional Information - Spent Fuel Pool Reracking.

4. PGandE Letter DCL-86-067, March 11, 1986; Response to Questions on Spent Fuel Racks.

5. PGandE Letter DCL-87-022, February 6, 1987; Rack Interaction Studies.

6. PGandE Letter DCL-87-072, April 9,1987; Additional Information on Rack-to-Rack Interactions (Proprietary and Nonproprietary)..

7. PGandE Letter DCL-87-082, April 23, 1987; Three-Dimensional Studies.

8. PGandE Letter DCL-87-ll5, May 18, 1987; Additional Information on Reracking Analysis.

9. Seismic Analysis Report, Rev. 3, September 3, 1986.

10. NRC Standard Review Plan, Section 9.1.2, NUREG-0800.

11. NRC Standard Review Plan, Section 3.8.4, Appendix D, NUREG-0800.

12. NRC "0T Position for Review and Acceptance of Spent Fuel Storage and Handling Applications," April 14, 1978 (Supplemented January 18, 1979).

~

UNITED STATES OF AMERICA

[ NUCLEAR REGULATORY COMMISSION c xn X UV

) Docket Nos. 50-275 g 4 In the Matter of ) 50-323 87 M 'O E3 :25

)

PACIFIC GAS AND ELECTRIC COMPANY ) (Reracking of Spent Fu,91 Pools) . .

) '

(Diablo Canyon Nuclear Power 00CKt a -

) BWA' Plant Units 1 and 2) )

)

CERTIFICATE OF SERVICE I hereby certify that on June 4, 1987, copies of the following document in the above-captioned proceeding have been served on the following by deposit in the United States mail, first class, or as indicated by an asterisk through delivery by Federal Express: PROPOSED FINDINGS OF FACT AND CONCLUSIONS OF LAH SUBMITTED BY PACIFIC GAS AND ELECTRIC COMPANY.

B. Paul Cotter, Jr., Chairman

• Docketing and Service Branch Administrative Judge Office of the Secretary Atomic Safety and Licensing U.S. Nuclear Regulatory Commission Board Panel Hashington DC 20555 U.S. Nuclear Regulatory Commission .

4350 East West Highway 4th Floor Bethesda HD 20814 Glenn 0. Bright

• Lawrence Chandler, Esq.*

Administrative Judge Benjamin H. Vogler, Esq.

Atomic Safety and Licensing Office of Executive Legal Director Board Panel U.S. Nuclear Regulatory Commission U.S. Nuclear Regulatory Commission Maryland National Bank Building 4350 East West Highway 4th Floor Room 9604 Bethesda HD 20814 7735 Old Georgetown Road Bethesda HD 20814 Dr. Jerry Harbour

• Diane M. Grueneich*

Administrative Judge Grueneich & Lowry Atomic Safety and Licensing 380 Hayes Street, Suite 4 Board Panel San Francisco CA 94102 U.S. Nuclear Regulatory Commission 4350 East West Highway 4th Floor Bethesda HD 20814 Rithard F. Locke Pacific Gas and Electric Company 77 Beale Street 31st Floor San Francisco, CA 94106 Dated at San Francisco, California, this 4th day of June,1987.

. . .