ML18058A361
| ML18058A361 | |
| Person / Time | |
|---|---|
| Site: | Palisades  |
| Issue date: | 04/06/1992 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML18058A360 | List: |
| References | |
| NUDOCS 9204240293 | |
| Download: ML18058A361 (10) | |
Text
(
~
"'t,P.f' REG(/~
,,,<!.'~'
.011....
~
~
('>
Cf 0
\\ii
.* '/.:
0 r!'
o' UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20555 I')
~
~***~
SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION
1.0 INTRODUCTION
1.1 Background
CONSUMERS POWER COMPANY PALISADES PLANT DOCKET NO. 50-255 Consumers Power Company (CPCo) had performed a seismic evaluation of the Palisades Nuclear Power Plant for the USNRC Systematic Evaluation Progra~
(SEP).
As a result of this evaluation, it was determined that the Palisades fuel design would need to be evaluated for seismic adequacy. This evaluation was performed in conjunction with the development of an advanced High Thermal Performance (HTP) fuel design for Palisades Reload Cycle M.
The HTP fuel assemblies for the Palisades Reload Cycle M were designed and fabricated by Advanced Nuclear Fuels Corporation (ANF)..
In support of the fuel design seismic evaluation, CPCo contracted Impell to develop seismic input loadings for the reactor vessel internals and reactor core. These loads were provided to ANF for evaluation of seismic integrity of the fuel.
The ANF performed an analysis to predict the seismic response, determine component loadings, and evaluate the seismic adequacy of the HTP fuel assemblies in the Palisades reactor. The methods, models, assumptions, and results of the analysis and evaluations were documented in an ANF report that was submitted to the NRC in November 1989.
1.2 Review Scope and Methodology The Mechanical Engineering Branch (EMEB) and its consultants at Brookhaven National Laboratories (BNL) performed a detailed review and evaluation of the methods, models, assumptions, and results of the ANF seismic analysis and structural evaluation of the Palisades HTF fuel.
The development of the seismic input forces used in the analysis was beyond the staff's scope of review.
The evaluation of seismic integrity was limited to the HTP fuel assemblies.
Seismic effects on the integrity of reactor vessel internals were not addressed.
Review of the original submittal (Ref. la) resulted in the need for additional information. Selected references from th~ report (Refs. 2 thrriugh 9) were requested as background information or for more detailed review.
Based on the staff's review of the report and the additional reference material, the staff found the overall approach acceptable but identified a few items which needed 9204240293 920406 PDR ADOCK 05000255 P
PDR further verification and review.
A.request for additional information (RAI) was prepared and submitted in December 1990.
As part of the request, CPCo was asked to provide some of the detailed component stress calculations for review.
The staff received a partial set of RAI responses by letter dated April 16, 1991 (Ref. 10). However, in the process of verifying the analysis to address concerns raised over questionable results, ANF identified a mathematical error in the analysis.
As a result, ANF had to correct the error and perform a complete seismic reanalysis and evaluation.
The ANF report was revised (Ref.
lb) and resubmitted.
The final RAI re~ponses and the revised report were submitted to the staff by letter dated June 30, 1991 (Ref. 11).
The ANF would not release its proprietary calculations but made them available for review at their offices. A final meeting between ANF and the staff was held at the ANF Richland, Washington facility in September 1991 to review calculations and resolve remaining issues.
2.0 FUEL ASSEMBLY DESCRIPTION The HTP fuel assembly is made up of a cage assembly (skeleton), fuel rods, a lower tie plate, and an upper tie plate. Its cross-section is approximately 8.25 in. by 8.25 in. square and its length is 149 inches.
The skeleton consists of eight guide bars which cage ten egg crate-like spacer grids, attached by welds at the two guide bar locations on each spacer grid face.
The fuel rods are friction supported in a 15xl5 array with eight of the peripheral fuel rod locations filled by the eight guide bars.
The assembly is held together by the guide bar attachments to the upper and lower tie plates.
The lower tie plate h~s alignment pins to match alignment holes in the lower core support plate. The upper tie plate has alignment pins in the upper core plate.
A total of 204 fuel assemblies are arranged in a symmetrical pattern throughout the reactor core.
The fuel assemblies are aligned laterally such that the shortest row at the periphery is made up of six assemblies while the longest row in the center consists of sixteen assemblies.
The fuel assemblies are separated by gaps of several sizes.
The gaps between peripheral assemblies and the core shroud is 0.14 in. The gaps between fuel assemblies alternate between 0.37 in. and 0.11 in. to allow for insertion of cruciform control blades between assemblies.
During a seismic event of sufficient magnitude, contact between adjacent assemblies and between peripheral assemblies and the core shroud can potentially occur.
The fuel assemblies are held in position by alignment pins at the lower core support plate and the upper core plate.
In the vertical direction, there is a gap between the* upper core plate and the fuel assembly.
During a seismic event of sufficient magnitude, the fuel assembly can lift off the lower core plate and potentially contact the upper core plate. If fuel uplift occurs, potential loss of lateral constraint at the lower plate must also be considered.
3.0
SUMMARY
OF SEISMIC ANALYSIS The ANF performed a fuel system mechanical response analysis to assess the structural integrity of the Palisades HTP fuel assembly.
To perform the analysis, ANF developed several lateral and vertical analytical models of single and multiple fuel assemblies.
The lateral models included the following types:
- 1.
A detailed single fuel assembly linear model (detailed finite element model) was used for the development of component loads and stress analysis.
- 2.
A single fuel assembly simplified model (simplified modal model) was derived from the detailed model for use in critical parameter test verification and applied transient selection.
- 3.
Two single fuel assembly non-linear models were used for model test verification and applied transient selection.
- 4.
A six fuel assembly (shortest row) core region, non-linear model (six fuel assembly model) was used to evaluate the fuel system mechanical response.
- 5.
A sixteen assembly (longest row) core region, non-linear model (sixteen fuel assembly model) was used to evaluate the fuel system mechanical response.
The vertical models included the following two types:
- 1.
A single fuel assembly non-linear model (no-slip vertical model) was used to evaluate assembly response under the fuel rod no-slip condition. This model allows. for vertical lift-off of the assembly.
- 2.
A second single assembly non-linear model (fuel slip vertical model) was used to evaluate assembly response in the event that the vertical loads were of sufficient magnitude to cause fuel rod slippage within the spacer grid. This model also allows for assembly lift-off from the lower support plate.
CPCo supplied a number of seismic transients (developed by Impell from a system model subjected to several different natural earthquakes *in three specific soil conditions) for use in the fuel assembly system seismic response analysis.
The ANF performed studies to select the most severe seismic event and applied that event to determine the worst case fuel assembly ~esponse.
The ANF performed a stress analysis by applying the results to the detailed finite element model.
Acceptance criteria of the various fuel assembly components were developed based on the guidelines of the ASME Boiler and Pressure Vessel Code.
The ANF also performed additional assessments to assure that the core region would maintain mechanical function and coolability for the blade insertion and assembly lift-off conditions.
Based on the results, ANF concluded that all acceptance criteria were met.
All stress limits were met and minimal assembly lift-off occurs.
The fuel assemblies will return to their original positions in the core region, following the postulated seismic events, guaranteeing control blade insertability and a coolable geometry.
4.0 TECHNICAL EVALUATION
The staff's technical evaluation of the ANF seismic analysis concentrated on the adequacy of the analytical methods used to predict fuel assembly maximum responses and component stresses, the basis and experimental verification of modeling parameters, the adequacy of the acceptance criteria, and the reasonableness and acceptability of the results. Based on a review of the original summary report (Ref. la) and selected references (Ref. 2 through 9),
the staff prepared a request for additional information which included seven questions.
The questions and CPCo responses are included in this safety evaluation report.
In the process of responding to questions regarding reasonableness of results, ANF identified a mathematical error in their analysis.
As a result, a complete reanalysis and reevaluation was carried out.
The results were documented in a revised summary report (Ref. lb).
Although the analysis results changed, the conclusions regarding seismic adequacy of the fuel assemblies remained the same.
As part of the response to the remaining RAJ questions, CPCo provided the revised report.
To complete the staff review, a final meeting was held between the staff and ANF in Richland, Washington, to close out remaining issues and review selected calculations.
The following is a summary of the major areas of review and the staff's evaluation.
4.1 Analysis Method~ and Models Advanced Nuclear Fuels Corporation developed several different models to determine the lateral and vertical seismic responses of the fuel assemblies.
The lateral models incorporated the capability to simulate impacts between adjacent assemblies or between peripheral assemblies and the core shroud.
The vertical models simulated the lift-off capability of the fuel assemblies and their potential for impacting the upper core plate.
Both linear and non-linear single fuel assembly lateral models were generated.
A detailed linear NASTRAN finite element model was developed for use in component stress evaluations.
Both a fuel assembly skeleton model (guide bars and spacer grids only) and a complete fuel assembly model (guide bars, spacer grids, and fuel rods) were generated. Details of the models*and their development are given in the ANF summary report (Ref. lb). Static load deflection test data of a prototype skeleton and ass~mbly were used to adjust element properties to provide the best correlation.
In addition, random vibration testing of a prototype assembly was performed to determine mode shapes and frequencies for comparison with those predicted by the model.
Good correlation was demonstr~ted for the first three modes.
Modal damping values were determined from a pluck test.
The fuel assembly detailed finite element model was reduced to an equivalent simplified modal model using NASTRAN modal synthesis techniques.
The simplified model was subjected to static and modal analysis to verify good correlation with test results. Finally, the model properties were adjusted for reactor temperature and for the effects of submersion in water..
Two single fuel assembly non-linear lateral models were generated.
The first was a simple two-impact spring model for correlation of spacer-grid impact properties to prototype assembly pluck impact tests. The second model included eight-impact springs and was used in the applied transient truncation study. This model provided the basic non-linear fuel assembly model for the multiple assembly core region models used in the seismic analysis.
The element properties assigned to the spacer-grid impact springs were based on prototype test data. Details of the models were provided in the ANF summary report (Ref. lb).
Multiple fuel assembly models were generated to determine the lateral seismic response of the fuel assembly system.
A six fuel assembly model representing the shortest row of the core region and a sixteen fuel assembly model representing the longest row in the core region were developed.
Both models were developed from the single non-linear assembly model with eight in-grid and eight through-grid elements per assembly.
Gap elements were included to represent the appropriate lateral gaps between adjacent assemblies and between peripheral assemblies and the core shroud to determine the lateral impact forces that occur during the seismic event.
Two non-linear single fuel assembly vertical models were developed.
A no-slip vertical model was developed to evaluate the potential condition of fuel assembly lift-off from the lower core support plate and potential impact with the upper core plate. The model consisted of the upper tie plate, lower tie plate, fuel rods, guide bars, and spacer grids.
The component properties were determined from design drawings.
The vertical impact stiffness of the fuel assembly is controlled by the axial stiffness of the eight guide bars.
No provision for fuel rod slippage within the spacer grids was included in the model.
Elastic springs were incorporated into the model to monitor the forces between the fuel rods and the spacer grids. The maximum load in these springs was calculated and compared to the rod breakaway load value to determine if the fuel slip model should be used.
The fuel slip vertical model was identical to the no-slip model with the exception that the fu~l rod to spacer grid connection was repres~nted by friction elements instead of elastic springs.
Gap elements were also added to account for potential impact between the fuel rods and the upper and lower tie plates.
In both mod~ls conservative structural, impact, and fluid damping values were assumed:
y The staff reviewed the information presented in the ANF report and reference documents and found the overall modeling approach to be reasonable and consistent with analytical procedures previously accepted by the NRC.
Advanced Nuclear Fuels Corporation had conducted an extensive test program on a prototype Palisades HTP fuel assembly at their Richland facility.
The test program (Ref. 7) included static and dynamic tests on the skeleton, spacer grids, and the fuel assembly.
It provided load-deflection characteristics, natural frequencies, mode shapes, and spacer-grid impact stiffness properties.
Analytical correlations with the test data provided a high level of confidence in the accuracy of the models.
In developing the vertical models for which test data was not available, ANF took a conservative modeling approach to maximize seismic response.
The staff found that ANF made appropriate adjustments in the modeling parameters to account for reactor conditions.
In the RAI, the staff asked for clarification on how spacer grid stiffnesses were adjusted for reactor conditions.
The ANF response was found to be acceptable.
During the September 1991 audit at the ANF facility, the staff reviewed sample calculations and inspected the test facility. Their calculation records were found to be clear and well organized. Selected calculations were audited and were found to confirm the information presented in the summary report.
However, during the course of the calculation review, an apparent modeling discrepancy was identified. It was noted that the detailed finite element stress model would always predict a zero axial guide bar stress for lateral bending.
Upon further* investigatioh, it became apparent that this is due to the models lack of representation of the vertical coupling provided by the welds between the guide bars and the spacer grids.
As a result, the model could only predict bending stresses in the guide bars.
The significance of the modeling assumption was discussed in detail.
The ANF pointed out that this would not affect the dynamic analysis which was based on the simplified modal model that had been synthesized from modal test data.
The detailed finite element model was only used to determine stresses.
Furthermore, ANF believed that test data could show that the zero axial stress.
assumption is reasonably accurate.
In the prototype tests, ANF had installed strain gages at guide bar locations and measured strains during the lateral load deflection test. Following the audit, ANF provided a comparison between guide bar stresses predicted by the analytical model versus those determined from test. The test data demonstrated the presence of an axial stress in the guide bar.
However, the magnitude of the axial stress was small when compared to the magnitude of the bending stress {approximately 15%).
For equivalent loads, the analytical model predicted approximately 10% higher bending stress and zero axial stress. The ANF also noted that the imposed test deflection correlated to an in-reactor deflection of 1.1 inches which is 30% greater than the maximum deflection predicted for Palisades.
In the test, none of the fuel assembly compo~ents were damaged.
Based on the test data and the stress margins, the staff agreed that the modeling discrepancy did not introduce a significant error in the evaluation.
For Palisades fuel, this modeling assumption was acceptable.
.. 4.2 Seismic Loading The seismic input forces applied to the non-linear lateral and vertical fuel system models were in the form of displacement time histories at the upper core plate, the lower core support plate and the core shroud locations.
The time histories were derived from the loads that were developed by Impell (Ref. 2).
Impell had performed soil-structure interaction analyses to generate the seismic input.
The soil-structure models included a representation of the Nuclear Steam Supply System (NSSS) and the reactor vessel with its internals.
Three soil conditions (soft, medium, and stiff) were evaluated for each of
- five natural earthquake records considered in the analysis.
The earthquake records included 1954 Taft S69E, 1984 Coyote Lake Sl5W, 1971 Castaic N67W, 1979 El Centro Array #4 NOOE, and 1979 El Centro S40E.
The earthquake records were scaled to 0.2g peak horizontal ground acceleration corresponding to a safe shutdown earthquake.
Each analysis was performed for a duration of 28 seconds to assure that the strong motion of each record was completely captured.
The results of these analyses were provided in the form of 28-second displacement time histories at the upper and lower core plates for the fifteen records considered.
The ANF performed -a study to determine if the number of input cases could be reduced and if an analysis time of less than 28 seconds could be justified.
The non-linear single fuel assembly lateral model including eight impact springs was used in this study.
The study model included a representation of the core shroud and the adjacent fuel assembly to determine impact fo~ces.
The model was subjected to the earthquake strong motions for the fifteen records.
Significant impacts were obtained for the Coyote Lake, Taft, and Castaic soft soil condition transients.
Based on the results of the study, the Coyote Lake soft soil record from 0 to 6.25 seconds was used for the final fuel assembly analysis. Vertical input motions were obtained from the system analysis consistent with the horizontal core plate motions and applied at the upper and lower core plate nodes in the*vertical models.
The staff reviewed the results of the ANF study for selecting and truncating the seismic input motion.
The overall methodology was found to be reasonable.
The results of the ANF study support the selection of the Coyote Lake soft soil earthquake as the most severe in terms of fuel assembly response.
By inspection of the time history record and the response spectra for the 6.25 second and full 28-second record, the truncation of the time history to 6.25 seconds appears acceptable for developing maximum fuel assembly response.
The use of a vertical input motion consistent with the selected horizontal earthquake record is acceptable.
4.3 Dynamic Analysis Results The seismic displacement time histories were applied at the upper and lower core plate and core shroud (horizontal model) nodes of the non~linear models to determine the dynamic response of the fuel assemblies.
The NASTRAN program utilized the Newmark-Beta integration method to solve the equations of motion.
. The six assembly and sixteen assembly non-linear lateral models were used to determine spacer-grid impact forces and fuel assembly deflected shapes due to the horizontal component of the safe shutdown earthquake (SSE).
The results of the analyses were presented in the ANF summary report as tables of maximum spacer-grid impact loads and maximum fuel assembly deflections.
ANF applied the maximum deflected shape to the detailed fuel assembly finite element model to calculate component loads. This provided maximum bending moments and shear forces in the fuel rods, guide bars, tie plates, and cap screws for use in the component stress evaluation.
The no-slip vertical fuel assembly model was used to predict the vertical seismic response.
Maximum forces between the fuel rods and spacer grids were determined to be low enough to validate the use of the no-slip model.
The vertjcal analysis provided the maximum fuel assembly uplift deflection and the axial loads in the guide bars, fuel rods, and tie plates for use in the component stress evaluation.
The staff reviewed the methodology and results and raised additional questions regarding the spacer-grid impact load results and the methodology for determining component loads from deflected shapes.
The spacer-grid impact loads presented in Tables 8.1 and 8.2 of the original summary report (Ref. la) appeared questionable.
Maximum impact forces occurred at the highest and lowest grid elevations.* These forces are generally expected to be highest at the center of the fuel assemblies where maximum velocities develop.
In responding to this question, ANF reviewed the analysis and discovered a mathematical sign error in a transformation matrix.
The source of the error was discussed during the September 1991 meeting with ANF.
As a result, ANF had to correct the error and perform a reanalysis.
The reanalysis results were presented in a revised report (Ref. lb).
Impact loads were substantially lower with maximum loads occurring near the center of the fuel assembly as expected.
The staff questioned the adequacy of the method used to calculate maximum component loads.
The. original analysis applied the maximum fuel assembly deflected shape from the multiple assembly lateral seismic analysis to the
.detailed finite element model to determine the maximum component loads.
The maximum deflected shape was essentially a first mode shape with maximum deflection near the center.
The staff pointed out that while the first mode shape provides the highest defl~ctions, it may not necessarily provide the highest moments end shear forces. Higher mode shapes may induce larger relative deflections with higher internal forces.
In the reanalysis, ANF investigated maximum relative deflection between adjacent node points along each fuel assembly to identify the time and location of maximum stress.
Deflected shapes were plotted and six shapes were selected to determine maximum stress. The deflected shape which yielded the largest stress was not in the form of a first bending mode since it occurred at the time of maximum spacer-grid impact.
The staff found the revised methodology for calculating maximum component loads acceptable.
The staff also reviewed the ANF methodology for predicting vertical seismic loads and found it acceptable.
- 4.4 Evaluation of Structural Integrity and Functionality The ANF developed component design criteria to evaluate the structural integrity and functionality of the fuel assembly.
The ultimate criterion was to maintain a post earthquake core geometry such that the control blades could be inserted and the core cooled. Control blade insertion is assured if the fuel assemblies maintain their original position in the core and no component plastic deformation or geometric instabilities occur.
Coolable geometry is maintained if insertability occurs and the spacer grids maintain the original fuel rod spacing.
The component stress criteria were based on the ASME Boiler and Pressure Vessel Code, Section Ill, Appendix F, for component supports and on the USNRC Standard Review Plan.
Those stress limits were applied to fuel rods, guide bars, tie plates, and guide bar to tie plate cap screws.
Spacer grids were evaluated to an allowable load limit based on tests in which the grids were loaded to the point of permanent deformation.
The allowable load was calculated as the 95% confidence value of the mean of the test data factored for in-reactor temperature conditions.
Based on the results of the lateral and vertical seismic analyses, ANF calculated maximum stresses in the fuel assembly components and evaluated them in accordance with the stress criteria discussed above.
All components met the stress or load criteria with adequate safety margins.
The ANF calculated the maximum vertical relative displacement of the fuel assembly from the lower core plate and compared it to the length of the alignment pins.
It was concluded that lateral support was maintained and the assembly will return to its original position following the seismic* event.
Based on this conclusion and the stress results which demonstrated that no permanent deformations will occur, ANF concluded that control blade insertability is assured and a coolable geometry is maintained.
The staff reviewed the acceptance criteria and evaluation results.
The criteria were found to be acceptable. A number of questions were raised regarding details of the stress evaluation results.
The ANF responses were found to be generally acceptable.
The reanalysis resulted in larger safety margins than were presented in the original report. During the September 1991 meeting in Richland, the staff audited sample calculations.
No further problems were identified and all remaining issues were closed.
5.0 CONCLUSION
S Based on the review and evaluation of the information provided by CPCo and ANF as described in this report, the staff concludes that the licensee had adequately demonstrated that the Palisades High Thermal Performance Fuel Assemblies will maintain their structural integrity and functionality when subjected to a safe shutdown earthquake.
Principal Contributor: J. Rajan Date:
April 6, 1992
).
REFERENCES l(a)
ANF Report No. ANF-89-115(P), "Seismic Analysis of Palisades High Thermal Performance Fuel Design," Rev. l, February 1989.
l(b)
ANF R~port No. ANF-89-llS(P), "Seismic Analysis of Palisades High Thermal Performance Fuel Design," Rev. 3, April 1991.
- 2.
Wesley, D. A. and M. W. Salmon, Core Seismic Input Palisades Nuclear Power Plant, Impell Report 11-0540-0200, Revision 3, September 1989.
- 3.
Exxon Nu~lear Company, Inc., ENC's Solution to the NRC Sample Problems - PWR Fuel Assemblies Mechanical Response to Seismic and LOCA Events, XN-NF-696 (P)(A), April 1986.
- 4.
Letter from H. N. 8erkow, (USNRC) to J. C. Chandler (Exxon Nuclear Company, Inc.),
Subject:
Acceptance for Referencing of Licensing Topical Report XN-NF-696, ENC's Solution to the NRC Sample Problems-PWR Fuel Assemblies Mechanical Response to Seismic and LOCA Events, December 26, 1985.
- 5.
Letter from S. F. Pierce (Consumers Power Company) to R. G. Hill (ANF),
Subject:
Palisades Fuel Assembly Spacing/Gaps, June 2, 1989.
- 6.
Grubb, R. L., Review of LWR Fuel System Mechanical Response with Recommendations for Component Acceptance Criteri~, Idaho National Engineering Laboratory, NUREG/CR-1018, September 1978.
- 7.
Advance Nuclear Fuels Corporation, Palisades Reload M-Seismic Mechanical Test Description and Results, ANF-88-119-(P), Revision *1, February 1989.
- 8.
Letter from D. A. Wesley (lmpell) to R. G. Hill (ANF),
Subject:
Earthquake Reduction Factors, May 22, 1989.
- 9.
Advanced Nuclear Fuels Corporation, Mechanical Design Report for Palisades Fuel Assemblies - Reloads J, K, and L Fuel Types and High Thermal Performance Spacer Leads, ANF-88-087(P), Revision 1 February 1989.
- 10.
Letter from G. 8. Slade (CPC} to NRC, "Reload M Fuel Seismic Analysis - Additional Information," April 16, 1991.
- 11.
Letter from G. 8. Slade (CPC} to NRC, "Reload M Fuel Seismic Analysis - Additional Information," June 30, 1991.