ML20006B474

Nonproprietary Indian Point Unit 2 Evaluation for Tube Vibration Induced Fatigue

ML20006B474
Person / Time
Site:	Indian Point
Issue date:	05/31/1988
From:	Connors H, Frick T, Pitterle T WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
To:
Shared Package
ML100331190	List: ML100331189 ML20006B474 ML20006B489
References
WCAP-11812, NUDOCS 9002020264
Download: ML20006B474 (138)

Text

{{#Wiki_filter:mm 3 I .i WESTINGHOUSE NON-PROPRIETARY CLASS 3 j'c2

WCAP-11812--

p 1 l s

• t h

INDIAN POINT' UNIT 2 EVALUATION FOR TUBE l; VIBRATION INDUCED FATIGUE r [ MAY, 1988 AUTHORS t H.J. CONNORS H.M. HU T.M. FRICK H.O. LAGALLY J.M. HALL-A.Y. LEE G.W. HOPKINS P.J. PRABHU J.L. HOUTMAN R.E. SMITH l., M.J SREDZIENSKI o APPROVED J'- T.A. PITTERLE, MANAGER STEAM GENERATOR ENGINEERING -WORK PERFORMED UNDER SHOP ORDERS GN7D-22972 AND GN7D-2402F b y 3 1 1 WESTINGHOUSE ELECTRIC CORPORATION POWER SYSTEMS BUSINESS UNIT SERVICE TECHNOLOGY DIVISION P. O. BOX 3377' PITTSBURGH,PA 15230-0355 i 900202O264 900119 PDR ADOCK 05000247 q. Q' 3;_ - -PDC

l 1 ABSTRACT On' July 15, 1987, a steam generator tube rupture event occurred at the North-Anna Unit 1 plant. The cause of the tube rupture has been determined to be high cycle fatigue. The source of the loads associated with the fatigue mechanism is a combination of a mean stress level in the tube with a superimposed alternating stress. The mean stress is the result of applied - loading, the manufacturing process and denting of the tube at the top tube support plate, while the alternating stress is due to out-of plane deflection of the tube U-bend attributed to flow induced vibration. -This report documents the evaluation of steam generator tubing at Indian Point Unit 2 for susceptibility to fatigue-induced cracking of the type experienced at North Anna Unit 1.. The evaluation utilizes operating conditions specific to Indian Point Unit 2 to account for the plant specific nature of the tube loading and response. The evaluation also includes field measurements taken for Indian Point Unit 2 to establish AVB locations. This report provides background of the event-which occurred at North Anna, it presents a criterion . for fatigue assessment, a summary of test. data which support the analytical approach, field measurement results-showing AVB' positions, thermal hydraulic analysis results, and calculations to determine tube mean stress, stability ratio and tube stress distributions, and accumulated fatigue usage. The results of the analysis indicate that none of the tubes remaining in service are expected to be susceptible to fatigue rupture-at the top support plate. Therefore, no modification, preventive tube plugging (other than the two tubes that were plugged in Steam Generator 21), or other corrective actions are required. o 0158M:49/041188 2 i

{O fn

SUMMARY

OF ABBREVIATIONS. o ASME - American -Society of Mechanical. Engineers ATHOS -' Analysis of the Thermal Hydraulics. of Steam Generators 3' AVB - Anti-Vibration Bar AVT - All= Volatile Treatment ECT - Eddy. Current Test EPRI - Electric Power Research Institute FFT - Fast Fourier Transform FLOVIB - FL0w induced Vibrations MEVF - Modal Effective Void Fraction 00 - Outside Diameter ~ RMS - Root Mean Square SR - Stability Ratio TSP - Tube. Support Plate- 'F --degrees Fahrenheit hr --hour ksi - measure of stress - 1000 pounds per square inch lb - pound mils - 0.001 inch MW - mega watt psi - measure of stress - pounds per square inch psia - measure of pressure - pounds per square inch, absolute-

b 11

-0158M:49/032988-3

e 7j

n WESTINGHOUSE PROPRIETARY CLASS 2 TABLE OF CONTENTS SECTION PAGE l.

1 1.0 Introduction.......................... 1-I j

2.0 Summary and Conclusions.....................

2-l-3.0 ' Background............................- 3-1 l n' 3.1 ' North Anna Unit 1 Tube Rupture Event........... 3-1 j 3.2 Tube Examination Results....-..........-... 3 i 3.3 Mechanism Assessment................... 3-3 4.0 Criteria for Fatigue Assessment....._........... 4-1 4.1 Stability Ratio Reduction Criteria............ 4-2 4.2 Local Flow Peaking Considerations............. 4 y 4.3. Stress Ratio Considerations................ 4-8 j g' i 5.0 Supporting Test Data...................... 5-1 5.1 Stability Ratio Parameters-................ 5-1 l;- 5.2 Tube Damping Data..................... 5 5.3. Tube Vibration Amplitudes withSupport............... gle-Sided-AVB j) Sin 5-8 L 5.4 Tests to Determine the Effects on Fluidelastic H Instability of Columnwise Variations in AVB j L Insertion Depths..................... 5-9 l L 5.5 References........................ 5-12 l 6.0 Eddy Current-Data and AVB Positions............... 6-1 6.1 Tube Denting at Top Support Plate............. 6-1 6.2 Tube Wall Thinning at the AVB Supports.......... 6-1 6.3 Eddy Current Data for AVB Positions'............ 6-1 6.3.1 Steam Generators 22 and 24............. 6-2 '6.3.2 Steam Generators'21 and 23............. 6-2 6.4 AVB Insertion Depths................... 6-3= '6.4.1 AVB Assembly Desi 6-3 bh'.............. 6-4 6.4.2 [ '[ 6.4.3 AVB Ja, - 5 6.4.4 AVB Map Interpretations.............. 6-6 6.5 Unsupported Tube Summary................. 6-7 7.0 Thermal Hydraulic Analysis................... 7-1 7.1 IndianLPoint Unit 2 Steam Generator O Conditions.............. perating 7-1 7.2 ATHOS Analysis Model................... 7-2 7.3 ATHOS Results....................... 7-3 7.4 Relative Stability Ratio Over Operating History...... 7-5 7.5 References........................ 7-8 1 0158M:49/041188-4 111

n TABLE OF CONTENTS (Continued) SECTION f&ff =* ~ 8.0 Peaking Factor Evaluation................... 8-1 ' 8.1 North Anna Unit 1 Configuration............ -.. 81 8.2' Test Measurement Uncertainties.............. 8-5 8.3 Test Repeatability.................... 8-6 8.4 Cantilever vs U Tube-....................- 86 a 8.5 Air vs Steam-Water Mixture................ 88 f 8.6 AV8 Insertion Depth Uncertainty.'.........- 8-11 8.7 Overall Peaking Factor with Uncertainty........... -8 12 4 9.0 Structural and Tube Vibration Assessments........... 9-1 0 9.1 Tube Mean Stress.-............. 9I ij 9.2 Stability Ratio Distributions Based Upon ATH0S......- 9-2 9.3 -Stress Ratio Distribution with Peaking Factor....... 9-3 6." 9.4 ' Cumulative Fatigue Usage................. 9-4 i P i 1 iv 0158M:49/041388-5

m g i LIST OF FIGURES FIGURE = PAGE-3-1~ Approximate Mapping of Fracture Surface of Tube R9C51 C, ' S/G "C" Cold Leg, North Anna Unit 1......... 3- ~ 5 3-2 Schematic Representation of Features Observed-During-4' TEM Fractographic Examination of Fracture Surface of Tube R9C51, S/G "C" Cold Leg, North Anna Unit 1... 3 y ) 3-3 Calculated and Observed Leak Rates Versus Time . 3-7 4-1. Vibration Displacement vs. Stability Ratio . 4-12 42 Fatigue Strength of Inconel 600 in AVT Water at 600*F,,4-13 j 4-3 Fatigue Curve for Inconel 600 in AVT Water l Comparison of Mean Stress Correction Models..... 4-14 4-4 Modified Fatigue with 10% Reduction in Stability j Ratio'for Maximum Stress Condition .. 4-15 'e 4-5 Modified Fatigue with 5% Reduction in Stability Ratio 1for Minimum Stress Condition .. 4-16 1 5-1 -Fluidelastic Instability Uncertainty Assessment... 5-15 y 7 2 Instability Constant - A........-........ 5-16 5-3 Instability Constants, A, Obtained for Curved Tubes l hi from Wind Tunnel Tests on the 0.214 Scale U-Bend s e Model........................ 5-17 s m 5-4 ' Damping vs. Slip Void Fraction 5-18 5-5 Overall View of Cantilever Tube Wind. Tunnel Model.. 5-19 o '5-6 Top' View of the Cantilever Tube Wind Tunnel Model.. 5-20 1 5-7 Fluidelastic Vibration Amplitude with Non-Uniform l Gaps........................ 5-21 l 5-8 Typical Vibration Amplitude and Tube /AVB Impact Force Signals for Fluidelastic Vibration with Unequal Tube /AVB Gaps.................... 5-22 J, 5-9 Conceptual Design of the Apparatus for Determining the i L Effects on Fluidelastic Instability of Columnwise L: Variations in AVB Insertion Depths 5-23 1 V '0158M:49/040488-6 ~. 'i

) y +, 't ) l LIST OF. FIGURES (Continued) 3-FIGURE PAGE f 5-10 Overall View of Wind-Tunnel Test Apparatus....... 5-24 j '5-11 Side View of Wind Tunnel Apparatus with Cover Plates i Removed to Show Simulated-AVBs and Top Flow Screen. -. 5-25 5 12 AVB Configurations Tested for Indian Point.2..... 5-26 5-13 Typical Variation of RMS Vibration Amplitude with Flow. Velocity for Configuration la'in Figure 5-12..... 5-27 6-1 AVB Insertion Depth Confirmation ...... 6-12 4 6-2 Indian Point 2: S/G-21 AVB Positions........ 6-13 4 j '6-3 Indian Point-2:- S/G-22 AVB Positions........ 6-14 6-4 Indian Point-2: S/G-23 AVB Positions........ 6 15

s u<

6-5.I6dian Point-2: S/G-24 ABB. Positions..... 6-16 It 7-1 Plan View of ATHOS Model for Indian Point Unit 2 7-12 j pn d 7-2 Elevation View of ATHOS Model for Indian Point Unit 2. 7-13 j. 7 Hot Leg Side Plan View of ATHOS Model........ 7-14 7-4' Flow Pattern on Vertical Plane of Symmetry..... 7 l 7-5 Flow Pattern on Horizontal Plane (Z-22) at U-Bend Region 7-16' L 7-6 Flow Pattern on Top of Tubesheet 7-17 ln l 7-7 Tube Gap Velocity and: Density Distributions for Tube ( at Rowl0/Co13.................... 7-18 -l 1 l 7-8 Tube Gap Velocity and Density Distributions for Tube 4 at Rowl0/Co120................... 7-19 7-9 Tube Gap Velocity and Density Distributions for Tube at-Rowl0/Co140................... 7-20 m 7-10 Average Velocity and Density in the Plane of the U-Bends Normal to Row 10.............. 7-21 5 vi 0158M:49/040488-7

p l y t s LIST 0F FIGURES:(Continued)- Xi 'l 1 FIGURE PAGE. t A./ 7-11 Radial Distribution:of Axial Mixture Velocity Above the Top TSP in. Indian' Point Unit 2 Steam Generators. 7-22 I 7-12 Radial Distribution of Axial Mixture Velocity Above the Top TSP in a 51 Series Steam Generator...... 7-23J ~ 7-13 Radial Distribution of the Ratio of Axial' Mixture e Velocity in Indian Point 2 to that in a.51 Series Steam Generator Above the top TSP.......-....... 7-24 I 7-14 Indian Point 2 Nomalized Stability Ratio Based on High. Power (>90%) Operation ........'. 7-25' 8-1 Original North Anna AVB Configuration.........- 8 i .8-2 Schematic of Staggered AVBs............. 8-24 8-3 AVB " Pair" in ECT Trace................ 8-25 a 8-4 -North Anna 1,-Steam Generator C, AVB Positions Critical Review I " AVB - Vi s i bl e" Call s................. 8-2 6 - 5 North Anna 1, Steam Generator C, i R9C51 AVB [ ]a,c Matrix......-....... 8-27 y 86 North Anna RSC51 AVB Final ( Ja.c Positions.. 8 1 ii l-H '8-7 Final Peaking Factor for Indian Point........ 8-29 9-1 ' Axisymetric Tube. Finite Element Model 9-9 2 Dented Tube Stress Distributions t 4 I Pressure Load on' Tube...._............ 9. 9-3 Dented Tube Stress Distributions Interference Load on-Tube.............. 9-11 9-4 Dented Tube Stress Distributions Combined Stress Results............... 9-12 1 i 9-5 Relative Stability Ratio Using MEVF Dependent Damping. 9-13 9 Stress Ratio Vs. Column Number 9-14 4. Vii 0158M:49/041388-8 ~ 4 i

' ~ 1 .a. 4 i 4 L-LIST 0F TABLES iTABLE: PAGE i t 4-1 Fatigue Usage per Year Resulting'From Stability 1 Rati o Reducti on. :.................. 4 11 51 Wind Tunnel Tests on Cantilever Tube Model 5-13 5-2 Fluidelastic Instability Peaking Factors for Columnwise Variations in AVB Insertion Depths.... 5-14 1 ~6-1 Utilization of I Ja.c for Determining. AVB Positions..................-.. 6-8 4 6 2 Resolution of Support Conditions for Row 10,11 g L 'and 12 Tubes with Single AVB. Indications ......6-9 6-3. Unsupported Tubes Summary............... 6-10 7-1 Indian Point Unit-2 Steam Generator Operating Conditions 7-9 7-2 Steam Generator Operating Conditions Used for ATHOS o An al y s i s '...................... 7 - 10 7-3 Indian Point. Unit 2. Operating History Data 7-11 -81 Stability Peaking-Factor Due to Local Velocity Perturbation............. 8-16 [ 82 Comparison of Air and Stream Water Peaking Factor Ratios..... 8-17 e 83 Effect of Local Variation-of AVB Insertion....... 8-18 .8-4 Uncertainties in Test Data and Extrapolation..... 8-19 i 8-5 Extrapolation of Test Results to Steam Generator Conditions....................... 8-20 l L 86 Final Peaking Factors................ 8-21 8-7 Stability Peaking Factors of Specific Tubes..... 8-22 p L' 9-1 1007. Power Operating Conditions Indian Point Unit 2................. 9-6 L L* 9-2 Indian Point Unit 2 Evaluation of the More Salient Unsupported U-Bends................. 9-7 9-3 Duty Cycle Description for Indian Point Unit 2.... 9-8 L 1 p: . 0158M:49/041188-9 viii b

W: i J y

1.0 INTRODUCTION

{= m This report documents the evaluation of steam generator tubing at Indian Point Unit 2 for susceptibility to fatigue-induced cracking of the type experienced at North Anna Unit I in July, 1987. The evaluation includes three-dimensional flow analysis of the tube bundle, air-tests performed to support the vibration - analytical. procedure, field measurements to establish AVB locations, structural and vibration analysis of selected tubes, and fatigue usage calculations to predict cumulative usage for critical tubes. The evaluation utilizes operating conditions specific to Indian' Point Unit 2 in order to account for plant specific features of the tube loading and response. Section 2-of the report provides a summary of the Indian Point Unit 2 evaluation results and ove all conclusions. Section 3 provides background for th'e tube rupture event which occurred at North Anna Unit 1 including results of P the examinstion of the ruptured tube and a discussion of the rupture event mechanism. The criteria for predicting the fatigue usage for tubes having an environment conducive to this type of failure are discussed in Section 4. Section 5 provides a summary of test data which supports the analytical vibration evaluation of the candidate tubes. A summary of field measurements j used to determine AVB locations and ultimately to identify unsupported tubes is j provided in Section 6. Section 7 provides the results of a thermal hydraulic analysis:to establish flow field characteristics at the top suppoit plate which are subsequently used to assist in identifying tubes which may be dynamically unstable.- Section 8 describes the overall peak'irig factor evaluation to define . test based peaking factors for use in the tube vibration and fatigue assessment. The final section, Section 9, presents results of the structural and vibration asssessment. This section determines tube mean stress, stability-ratio and tube stress ratio distributions, and accumulated fatigue usage, forming the basis for the conclusions for Indian Point Unit 2. l;- t l. l' l 0158M:49/041188-10 ,-) l

? q 'b~ } q 2.0 SUMARY AND'CONCl.USIONS l The Indian, Point Unit 2 steam generators are evaluated for the_ potential of q c + unsupported U bend tubing with denting at the top tube support plate.to a fatigue rupture of;the type experienced at Row 9 Column 51 (R9C51) of Steam Generator C, North Anna Unit 1. 2.1 Background- .s The initiation of the circumferential crack in the tube.at the top of the. top r tube support plate was due to limited displacement, fluid elastic' instability. The unstable condition prevailed in the R9C51 tube from the time when the tube P first experienced' denting at the support plate. A combination of conditions b were present:that led to the rupture. The tube is not supported by an -anti-vibration bar (AVB), has a higher flow field due to the uneven insertion t . depths of AVBs, has reduced damping because of denting at the top support plate, and has reduced fatigue properties because of the additional mean stress-p with the denting and in the presence of the all-volatile treatment (AVT) L:, . chemistry of the secondary water, { i L -2.2 Evaluation Criteria-q U The criteria established to provide a fatigue usage less than 1.0 for a finite period of time (i.e., 40 years) is a 10% reduction in stability ratio that is expected to provide a 58% reduction in stress amplitude (to <4.0 ksi) for a Row- .9 tube in the North Anna Unit I steam generators. This reduction is required 4 to produce a fatigue ' usage'of <0.021 per year for the Row 9 tube. This same-L. criteria is used as one of several criteria in the evaluation of Indian Point Unit 2 tubing.. k'ith additional effects accounted for through a stress ratio criteria that permits the calculation of fatigue usage to demonstrate tube' acceptability,.the final criteria of cumulative fatigue usage to date plus l future operation with current operating parameters can be satisfied. 2-1 4 0158M:49/041188-11 l,

+ Theistability ratios for. Indian Point Unit 2 tubing, the corresponding stress 1 [ amplitude, and the:resulting cumulative fatigue usage cannot be calculated 'directly but;must be evaluated relative to the ruptured tube at Row'9 Column 51, North. Anna Unit 1, Steam Gencrator C, for two reasons. The local. effect on. the flow field due to various AVB insertion depths is not within the capability of available evaluation techniques and must be determined by test. In addition, an analysis and examination of the ruptured tube provides a range of' -initiating stress amplitudes but can only bound the possible stability ratios -that correspond to these stress amplitudes. Therefore, the evaluation of Indian Point Unit 2 tubing is based on relative stability ratios, relative flow j peaking factors and stress ratios. The criteria for establishing that a tube has support from an AVB and-therefore eliminate it from further considerations is that.at least one sided support is present to at least the tube centerline. This is established by analysis of i eddy current (EC) measurements. AVB support of the tube conterline may be established by an EC indication of both legs of the AVB. or may be established by projecting the depth of insertion knowing the geometry of the AVB. The AVB insertion depths are a key factor in.the assessment of the potential for a fatigue induced tube rupture since the AVB positions' determine the local flow peaking factors. The local flow peaking factor is a direct factor on the apparent stability ratio for uniform flow conditions where a small percentage change causes a significant change in stress amplitude. The relative flow L peaking factors for Indian Point Unit 2 tubes without direct AVB support have l been determined by instability tests. 'These factors are applied to relative i. stability ratios determined by 3-D tube bundle flow analysis and the combined relative stability ratio is used in the stress ratio determination. [ l The analysis of eddy current measurements shows that virtually all of the tubes L in Rows 8'to 12 are dented at the sixth (top) support plate. It was assumed L for this evaluation that all tubes are dented at the top support plate. It also shows that none of the tubes in Rows 8 through 12 have wall thinning-c indications at the AVB locations. Therefore, it is unlikely that those tubes have been unstable. Additionally, the eddy current data are used to define AVB 0158M:49/051288-12

..,3g,, y g insertion depths and the extent of tube support for each column of tubes within 1 "the region. -Virtually all row 12 tubes are supported except for a few y- - peripheral columns. Most row 10 and row 11 tubes' are supported. Some row 8 L and row 9 tubes are supported. 0;- Twenty-eight tubes are identified as the more susceptible of the dented and l! unsupported tubes. These are:. STEAM STEAM STEAM STEAM J . GENERATOR TUBE GENERATOR TUBE GENERATOR TUBE GENERATOR TUBE c 21 R10C3 22 R10C24 23 R1003 24 R10C71 R10C16 R10C20 R10C24 R11C3 R11C3 - R10C62 R11C37 R1102 R11C46 R11C3 R11C51 R11C3 R12C2 R1104 R11C4 R11C47-R12C2 R11C86 R11C87 R11C68 R11C89 R12C2 R12C89 R12C90 The local flow peaking corresponding to-AVB insertion depths are factored into - the evaluation for each of the above tubes. This results in the relative stability ratios and stress ratios shown below. (All ratios are in comparison to R9C51,-North Anna Unit 1, Steam Generator C.) 0158M:49/041388-13 2-3 9

fL a, f ' .m. '/k ' r I t RELATIVE STRESS STEAM h GENERATOR TUBE. STABILITY RATIO RATIO 21 R10C3. 0.64 0.13 [

R10C16 0.64-0.13 R10C24-0.64

.0.13 q -R10C62 0.67-0.17 ,7 ( f R11C3 0.77 0.33: n R11C4 0.76 0.30 4 h R11C86 0.71 0.20 0.73 0.24-1 R11C87 R11C88 0.74 0.27: R11C89 0.76 0.30 _( e R12C2 0.73 0.21 l i. R12C89 -1.01 >1. 0 - R12C90 1.05 >1.0 22 R10C24 0.64 0.13 q.. p4 R11C3 - .0.77 0.33 R11C37 0.67 0.15 h R12C2 0.73

0. 21..

'23 R10C3 0.64' O 13 R10C20 0.64 0.13 K R11C2 0.53 0.04-1 R11C3 0.77 0.33 4 L R11C4 0.76 0.30 R11C47 0.79 0.40 1 n [;, H 24 .R10C71 0.89 0.94 1, R11C3 0.77 0.33 R11C46 0.79 0.40 R11C51 0.68 0.16 R12C2 0.73 0.21 h h; 0158M:49/032588-14 2-4 U w

q l t. .With the exception of two tubes in Steam Generator 21, all tubes have relative stability ratios-less than 0.9 and stress. ratios less'than 1.0. The two-tubes J exceeding these criteria are R12C89 and R12C90 in Steam Generator 21. These [ 'two tubes.have been plugged as a preventive measure to eliminate them from further consideration. Sentinel, or tell-tale, plugsLwere used to permit y detection of tube degradation if it should' occur in. the future. ~ 0f the remaining tubes that meet the criterion, the tube at R10C71, Steam Generator 24, has the highest stress ratio, 0.94.*- The expected maximum stress. n . amplitude for this tube is 3.8 ksi. Based on the operating history of Indian i Point 2', which is presented in Figure 7-14 (see Section 7)'in the form of normalized stability ratio versus time in days, the cumulative fatigue usage to a e date is expected to be less than 0.161. Projecting this for a 40 year design L basis, the cumulative fatigue usage is expected to be less than 0.70. This is-i .-less than the ASME Code limit of 1.0. The fatigue usage calculation utilizes a .i s l -lower bound fatigue curve that is consistent with the fatigue properties of the I North Anna ruptured tube. k. J. 2.7 Conclusion LIn conclusion, Indian Point 2 tubes remaining in service are not expected to be susceptible to fatigue rupture at the top support plate in a manner that is [ similar to that which occurred at North Anna 1. Therefore, no modification, preventive tube plugging other than the two tubes that were plugged in Steam Generator 21 or other measure to preclude such an event is judged to be necessary. L Earlier work included in a Consolidated Edison submittal to the NRC showed L a stress ratio of 0.29 for R10C71. However, based on the results of. recent L flow tests (see Section 8 of this report) the stress ratio has increased. L Figure 9 6 shows that Tube R10C71 (R10C22 is its mirror image) in Steam Generator 24 has.the highest stress ratio. This result is due to the I specific AVB arrangement adjacent to R10C71. This AVB arrangement produces local flow peaking. 0158M:49/041888-15 2-5 s

_ _ _ _ _.. _ ~ _ _ _ _ _ 4

3.0 BACKGROUND

~ On July 15, 1987, a steam generator tube rupture occurred at the North Anna i Unit l'. The ruptured tube was determined to be at Row 9 Column.51 in Steam-Generator."C". The location of the opening was found to be at the top tube support plate on the cold leg side of the tube and was circumferential in - orientation with a 360 degree extent. 3.1 North Anna Unit 1 Tube Rupture Event The cause of the tube rupture has been determined to be high cycle fatigue. The stresses associated with the fatigue mechanism have been determined to be a L combination of a mean stress level in the tube and a superimposed' alternating - stress. The mean stress has been determined to have been increased to a L maximum level as the result of denting of the tube at the top tube support L plate' and the alternating stress has been determined to be due to out-of plane ] deflection of the tube U-bend above the top tube support caused by flow induced L. vibration. These stresses are consistent with a lower bound fatigue curve for the tube material in an AVT water chemistry environment. The vibration mechanism has-been determined to be fluidelastic, based.on the magnitude of the alternating stress. A significant contributor to the occurrence of excessive vibration is the reduction in damping at the tube-to-tube support plate interface caused by the denting. Also, the absence of antivibration bar (AVB) support has been concluded to be required for requisite vibration to occur. The pNsence of an-AVB support restricts tube motion and thus precludes the deflection amplitude required f:,r fatigue. Inspection data shows that an AVB is not present for the Row 9 Column 51 tube but that the actual AVB installation depth exceeded the minimum requirements in all cases with data for AVBs at many other Row 9 tubes. Also contributing-significantly to the level of vibration, and thus loading, is the local flow field associated with the detailed geometry of the steam generator, i.e., AVB insertion depths. In addition, the fatigue properties of the tube reflect the lower range of properties expected for an I b' AVT environment. In summary, the prerequisite conditions derived from the evaluations were concluded to be: s-t 3-1 7 0158M:49/041388-16 y

yy+ )! ib ll i Fatiaue Reauirements Er.ar.gggisite' conditions - %,J

Alternating stress-Tube vibration

~ - Danted support - Flow excitation - Absence of AVB q q Mean~ stress ' Denting in addition to applied stress' 7 ~ Naterial fatigue! properties AVT environment - Lower range of properties } s 3.2 ' Tube Examination Results Fatigue was found to have initiated on the cold leg outside surface'of Tube R9C51 immediately above the top tube support plate. No indications of ( significant accompanying intergranular corrosion was observed on the fracture j . face or'on the'immediately adjacent 00 surfaces. Multiple fatigue initiation ' sites were found with major sites located at 110', 120'., 135' and 150', Figure 3-1. The plane of the U bend is located at 45' with the j orientation system used, or approximately 90' from the geometric center of ~ thelinitiation zone.at Section D-D. High cycle fatigue striation spacings approached 1 micro-inch near the origin sites, Figure 3-2. The early crack i: front. is believed to have broken through-wall from approximately.100' to 140'..From this point on,- crack growth is believed.(as determined by .m striation spacing, striation direction, and later observations of parabolic dimples followed by equiaxed dimples) to have accelerated'and to have changed . direction with the resulting crack front running perpendicular to the

circumferential direction.

Q 3-2 7 0158M:49/032588-17

d - 3.3 Mechanism Assessment 1 ~ d To address a fatigue mechanism and.to identify the cause cf the loading, any-loading condition that would cause cyclic stress or steady mean stress had to j c be considered. The. analysis of Normal, Upset and Test conditions indicated 'a ? relatively low total number of cycles involved and a corresponding low fatigue 1 usage, even when accounting for the dented tube condition at the plats. This analysis Liso showed an _ axial tensile stress cor.tribution at the tube 00 a short distance above the plate from operating; pressure and temperature, thus t providing-a contribution to mean stress. Combining.these effects with denting deflection on the tube demonstrated a high mean stress at the failure -location. Vibration analysis for the tube developed the characteristics of first mode, cantilever resporse of the dented tube to flow induced vibration for the uncracked tube and for the tube with an increasing crack angle, i beginning at 90' to.the plane of the tube and progressing around on both sides to complete separation of the tube. ^ Crack propagation analysis matched cyclic deformation with the stress intensities and striation' spacings indicated by the fracture inspection and analys'is. Leaka'ge data and crack opening analysis provided the relationship I between leak rate and circumferential crack length'. Leakage versus time was then predicted from the crack growth analysis and the leakage analysis.with initial stress amplitudes of 5, 7, and 9 ksi. The comparison to the best estimate of plant leakage (performed after the event) showed good agreement, Figure 3-3. - Based on these results, it followed that the predominant loading mechanism responsible is a flow-induced, tube vibration loading mechanism. It was shown that of the two possible flow-induced vibration mechanisms, turbulence and fluidelastic instability, that fluidelastic instability was the most probable cause. Due to the range of expected initiation stress amplitudes (4 to 10ksi), the fluidelastic instability would be limited in displacement to a range of approximately [ ]a,c This is less than the distance between tubes 6t the apex, [ ]a,e It was further confirmed that displacement prior to the rupture was limited since no indication of tube U-bend (apex region) damage was evident in the eddy-current signals for adjacent tubes. 3-3 0158M:49/041188-18 =

IP'.v: 1 I i[ Given the probability. of limited displacement, fluidelastic instability, a L ' means of establishing the change in displacement, and corresponding change in stress amplitude, was developed for a given red 0ction in stability ratio-(SR). [ Since the rupture was a fatigue mechanism, the change in stress amplitude resulting from a reduction in stability ratio was converted to a fatigue usage LI benefit through the use of the fatigue curve developed. _Mean stress effects were included due to the presence of denting and applied loadings. The results q indicated that a 10% reduction in stability ratio is needed (considering the range of possible initiation stress amplitudes) to reduce the fatigue usage per j year to less than 0.02 for a tube ~ similar to Row 9 Column 51 at North Anna j Unit 1. 1 f I t .i l .q i 3-4 0158M:49/032588-19

1 .l l:- -t i w, L h t L ). D=0' Q L c-CC i f C j g. l' Region of y Herring 6ene g . r Pattern t J s .==== 90' 270*==== ~ g -3 8-B 0* F g.g 3 t.~ - A A TAB : 'o,- - A-A p,, Coarse.Testure.- and Displed Rupture i ( Indicates Origins 1 i j r 4,. Figure 3-1 Approximate Mapping of Fracture surface of Tube R9C51, S/G "C" Cold Leg, North Anna Unit 1 3-5 O i ,-r, r e- - -,

p-t, ) l 4 hl :. I heavy i-estee Atteek 3 g,g3,g,, i 34 l /s 3 e 1.8/1.88 e is. 13p* I $4 \\ ~ \\ m h Perebelts l

90' 270*=

14 01seles Vl/ ene Inte m 1~ Necking i 3*8 \\ Y Sedir 3 e 1.8/4.0 e in. I M 11 Peened 34 4 Nearl t idsed 5 = 4.1/4.9 e in. tgyse bete: Arrous I M icate Direction of Fracture Propagation Figure 3-2 Schematic Representation of Features observed During TEN Fractograhic Examination of Fracture Surface of Tube R9C51, S/G "C" Cold Leg, North Anna Unit 1 1, 3 3-4 i =

p. 1 1 + .I 1 i 5 1 1 j 100000 i Calmleted and essened lent estas verses tlas. etsense values bones es gasemos ogsslessensaaser air aeoster i s I > 4. 51914 &

• O KS!

- SISta &

• 7 MSI

......... ggata a. g gag l l O AP=48 O Ne=tM ^j a a = 1000< g g ? w >= E y g 3 100 = a. = / O O i 30 O a a a a a A A A A A k h i w w w w w 3 y i V 0 SCO 1000 1500 3000 2500 3000 3600 4000 4500 5000 ll30 6000 TIME (MINUTES) n Figure 3 3 Calculated and Observed Leak Rates Versus Time I 3-7 --- t... ::....-. s.. p.

i 4.0 CRITERIA F0P. FATIGUE ASSESSMENT Evaluation against criteria to show that Indian Point Unit 2 steam generator tubing will not rupture by fatigue in the manner of (iorth Anna Unit I can only be done by an assessment relative to the Row 9 Column 51 tube of Steam j ~ Generator C, North Anna Unit 1, since,1) methods for direct analytical ) prediction of actual stability ratios incorporate greater uncertainties than a relative ratio, and 2) the stress amplitude (or displacement) associated with a specific value of stability ratio can only be estimated by the tube rupture l analysis of North Anna Unit 1. For these reasons, the North Anna Unit I tubing l evaluation was done on a relative basis to Row 9 Col 51 and a 10% reduction in i stability retic criteria.was established to demonstrate that tubes left in l service are expected to have sufficiently low vibration stress to preclude future fatigue rupture events. To accomplish the necessary relative assessment of Indian Point Unit 2 tubing to Row 9 Column $1 of North Anna Unit 1, several criteria are utilized.

First, stability ratios are calculated for Indian Point Unit 2 steam generators based on flow fields predicted by 3 0 thermal hydraulic models and ratioed to the stability ratio for Row 9 Column 51 at North Anna, Unit 1 based on a flow field obtained with a 3 D thermal hydraulic model with the same degree of L

refinement. These ratios of stability ratio (called relative stability ratios) for each potentially unsupported U bend in the Indian Point Unit 2 steam generators should be equivalent to < 0.9 of R9C51, North Anna (meeting the 10% l reduction in stability ratio criteria). This provides the first level of screening of susceptible tubes incorporating all tube geometry and flow field differences in the tube dynamic evaluation. It has the inherent assumption, however, that each tube has the same local, high flow condition present at Row 9 Column 51, North Anna Unit 1, and does not account for tube geometry differences in the stress calculation. To account for these differences, flow paaking factors can be incorporated in the relative stability ratios or, as noted below, in the relative stress ratios, o.. l l M 0158M:49/051288 23 1 i } -n

_q L I l J e The second criteria is to obtain stress ratios, the ratio of stress in the Indian Point Unit 2 tube of interest to the stress in Row 9 tolumn 51, North j Anna Unit 1, and after incorporating the requirement that the relative i stability ratio to Row.9 Column 51 (R9C51) for the tube of interest is equivalent to < 0.9, requim the simss ratio to be < 1.0. The stress ratio incorporates the [ Ja,c with R9C31 in relation to the stress calculation and also incorporates the ratio of flow peaking factor for the Indian Point Unit 2 tube of interest to the flow peaking factor for R9C51 -(flow peaking factor is defined in Section 4.2). This should provide that all tubes meeting this criteria have stress amplitudes < 4.0 ksi. Finally, the cumulative fatigue usage for plant operation to date and for l continued operation with the same operating parameters is evaluated. A fatigue usage of < 1.0 is not necessarily satisfied by meeting the stress ratio criteria since Indian Point Unit 2 has a different duty cycle than North Anna Unit I and the tube at issue may have a different (higher) frequency of = vibration. The Indian Point Unit 2 duty cycle could be more demanding and cycles could accumulate at a more rapid rate. Therefore, the time history of l operation is evaluated on a normalized basis and used together with the stress ratio to obtain a stress amplitude history. This should permit the calculation of current and future fatigue usage for comparison to 1.0. l 4.1 Stability Ratio Reduction Criteria L For fluidelastic evaluation, stability ratios are determined for specific l configurations of a tube. These stability ratios represent a measure of the potential for flow. induced tube vibration during service. Values greater than unity (1.0) indicate instability (see Section.5.1). Motions developed by a tube in the fluidelastically unstable mode are quite large in comparison to the other known mechanisms. The maximum modal displacement (at the apex of the tube) is linearly reisted to the bending stress in the tube just above the cold leg top tube support plate. This relationship applies to any vibration in that mode. Thus, it is possible for [ an unstable, ( Ja,c boundary condition tube to deflect an amount in the U bend which will produce fatigue inducing stresses. 0158M:49/041388-24 42 l'

N, The major features of the fluidelastic mechanism are illustrated in Figure f' 4-1. This figure shows the displacement response (LOG D) of a tube as a 4 function of stability ratio (LOG SR). A straight line plot displayed on log log coordinates implies a relation of the form y= A(x)n, where A is a constant, x i is the independent variable, n is the exponent (or power to which x is raised), [ and y is the dependent variable. Taking logs of both sides of this equation f leads to the slope intercept fors of a straight-line equation in log form, log y = c + n log x, where c = log A and represents the intercept and n is the slope. In our case the independent variable x is the stability ratio SR, and the dependent variable y is tube (fluidelastic instability induced) displacement response, and the slope n is renamed s. From experimental results, it is known that the turbulence response curve (on l log log coordinates) has a slope of approximately ( Ja,b,c Test results also show that the slope for the fluidelastic response depends somewhat on the t instability displacement (response amplitude). It has been shown by tests that a slope of ( Ja,b,c is a range of values corresponding to displacement amplitudes in the range of [ Ja,c,whereasbelow ( Ja.c are conservative values. The reduction in response obtained from a stability ratio reduction can be expressed by the following equation: ( Ja c [3) where D1 and SRI are the known values at the point corresponding to point 1 of Figure 4-1 and D2 and SR2 are values corresponding to any point lower on this curve. Therefore, this equation can be used to determine the reduction in displacement response for any given reduction in stability ratio. This equation shows that there is benefit derived from even a very small percentage change in the stability ratio. It is this reduction in displacement for a quite small reduction in stability ratio that formed the basis for - demonstrating that a 10% reduction in stability ratio would be sufficient to prevent Row 9 Column 51 from rupturing by fatigue. 0158M:49/041188 25

i l I 'he fatigue curve developed for the North Anna Unit 1 tube at R9C51 is from j [' s j i i I ja.c Thus,' f e,c [2] ] where, a'a is the equivalent stress amplitude to 'a that accounts for a maximum stress of a, the yield strength'. The 3 sigma curve with y mean stress effects is shown in Figure 4-2 and is compared to the ASME Code Design Fatigue Curve for Inconel 600 with the maximum effect.cf mean stress. ~ The curve utilized in this evaluation is clearly well below the code curve- - reflecting the, effect of an AVT environment on fatigue and [. Ja,c for accounting for mean stress that applies to materials e in a corrosive environment. Two other mean stress mod &is were investigated for the appropriateness of their use in providing a reasonable agreement with the expected range of initiating l stresc amplitudes. These were the [ Ja c shown in Figure 4-3. With a [- ja,C ~ 0158M:49/041888 26 4 v = v -,-------se-sv,-e-s -w+-,w-ew--- w---ww-ve--e -+v-w =<--- -m~ w

- ~ The assessment of the benefit of a reduction in stability ratio begins with the 1, relationship between stability ratio and deflection. For a specific tube geometry, the displacement change is directly proportional to change in stress L.. so that stress has the same relationship with stability ratio, a,e [3] 34 slope in this equation can range from ( Ja.c n a log scale depending on the amplitude of displacement. Knowing the stress resulting from i to SR, the cycles to failure at the a change in stability ratio from SR3 2 stress amplitude was obtained from the fatigue curve. A fatigue usage per year j was then determined assuming continuous cycling at the natural frequency of the tube. The initial stress was determined to be in the range of 4.0 to 10.0 ksi by the fractography analysis. i L lt was further developed that the maximum initiating stress amplitude was not more than 9.5 ksi. Thiswasbasedonthe( Ja,c The corresponding stress level is 5.6 ksi. The maximum stress, 9.5 ksi, would be reduced to ( Ja c with a 10% l reduction in stability ratio and would have a future fatigue usage of (- ')a,c per year at 75% availability, Figure 4-4. The minimum stress, 5.6 ksi, would be reduced to ( Ja,c with a 5% reduction in stability ratio and would have future fatigue usage of ( ']a,c per year, Figure 4 5. In addition, if a tube were already cracked, it could be as large as [ Ja c in length and thru wall and would not propagate if the stress amplitudes are rediiced to < 4.0 ksi. 0158M:49/041188 27

7..__ i s Subsequent to the return to power evaluation for North Anna Unit 1, the time history of operation was evaluated on a normalized basis to the last cycle. ( I i i f Ja c cumulative fatigue usage may then be computed to get a i magnitude of alternating stress for the last cycle that results in a cumulative usage of 1.0 for the nine year duty cycle. The result of the iterative analysis is that the probable stress associated with this fatigue curve during the last cycle of operation was approximately ( Ja.c for Rgt$1, North _ Anna Unit 1, Steam Generator C, and that the major portion of the fatigue usage came in the second, third and fourth cycles. The first cycle was conservatively omitted, since denting is assumed for purposes of this analysis to have occurred during that first cycle. Based on this evaluation, the tube fatigue crack initiation may have occurred over most of the operating history of North Anna Unit 1. L A similar calculation can be performed for the time history of operation assuming that the [ r l Ja,c L On this basis, the effect of a 10% reduction in stability ratio is to reduce ,the stress amplitude to 4.0 ksi and results in a future fatigue usage of [ ja.c Other combinations of alternating stress and mean stress were evaluated with 3 L sigma and -2 sigma fatigue curves to demonstrate the conservatism of the 10% reduction in stability ratio. Table 41 presents the results of the cases analyzed clearly demonstrating that the 10% reduction in stability ratio combined with a -3 sigma fatigue curve and with maximum mean stress effects is conservative. Any higher fatigue curve whether through mean stress, mean stress model, or probability, results in greater benefit for the same reduction in stability ratio. And for any of these higher curves, a smaller riduction in stability ratio than 10% would result in the same benefit. In addition, there is a large benefit in terms of fatigue usage for relatively small changes in the fatigue curve. 4-6 01 ERW. AC /nd 11 RR.9 A

Id 4.2 Local Flow Peaking Considerations [ Local flow peaking is a factor on stability ratio that incorporates the effect on local flow velocity, density and void fraction due to non uniform AYB L, insertion depths. The flow peaking factor is applied directly to the stability {' ratio obtained from thermal hydraulic analysis that does not account for these i . local geometry effects. Being a direct factor on stability ratio, a small percentage increase can result in a significant change in the prediction of 1 tube response, The development of flow peaking factors for Indian Point Unit 2 is presented in detail in Section 8. The establishment of AVB positions for Indian Point Unit 2 is presented in Section 6.4. Since the evaluation of Indian Point Unit 2 is relative to R9C51, North Anna Unit 1, the flow peaking factors are also relative. The peaking factor is applied as a ratio of stability peaking factor of the specific tube to that of North Anna R9C51. The flow peaking relative instability is obtained by testing in the air test rig described in Section 5.4, where the peaking factor-is defined as the critical velocity for R9C51 AVB pattern compared to critical velocity for a uniform AVB pattern. As explained in Section 8, the minimum value of [ la,b,c is appropriate for R9C51 of North Anna 1. Applying a nominal uncertainty to the Md0S analysis result of 15%, the predicted stability ratio for R9C51 is [ Ja,c using the nominal damping that is a function of the modal effective void fraction (slip). Applying the ( la,b,c factor gives a corrected stability ratio of [ ,)a,c Therefore, using the minimum flow peaking factor and nominal damping that reflects reduced damping compared to undented tubes, the tube at R9C51, North Anna Unit 1, Steam Generator C is shown to be unstable by comparison to the same tube with a uniform AVB pattern. It is therefore now considered to be sufficient to address the relative susceptibility of tubes to fatigue rupture on the basis of modal effective void fraction (slip) dependent damping since nominal damping is sufficiently low. The case of assuming constant damping versus void fraction in two-phase flow that had to be used in the North Anna Unit I evaluation is not necessary. 0158M:49/051288 29

m L) t i i 4.3 Stress Ratio Considerations In Section 4.1, a 107,reduct',on in stability ratio was established to reduce L the stress amplitude on the Row 9 Column 51 tube of North Anna Unit 1 to a ~ level that would not have ruptured, 4.0 ksi. To apply this same criteria to another tube in the same or another steam generator, the differences in tube geometry on stress must be accounted for in addition to the geometry effects i considered in calculating their respective relative stability ratios. l "i a,c Also, bending stress is calculated with a M (r/I) (5) where (r/I) is the section modulus for the tube. Since the Indian Point Unit 2 tubing has the same cro'ss section as the North Anna Unit 1 tubing, the Indian Point Unit 2 stress agp is ir, proportion to the North Anna Unit I stress aNA as: a,e s 9 6 _J 6 9 . 4 4'8 33 q:9 -

O,c l i i By establishing their equivalent effect on the stress amplitude that produced t n the tube rupture at North Anna 1, several other effects may be accounted for. These include a lower mean stress (such as for non dented tubes), different I frequency tubes from the [ Ja,c hertz frequency of R9C51, North Anna 1, and shorter design basis service (such as 30 years). l t In the= case of-lower mean stress, the stress amplitude that would have caused the failure of R9C51, North Anna 1, would have been higher. (' t ja.c 5 L

1..

I 4-9 0158M:49/041388-31 1 l. . 1_

m l 8: P L A lower or' higher frequency tube would not reach a usage of 1.0 in the same-n L length of time as the R9C51 tube due to the different frequency of cycling. ,j [* The usage accumulated is proportional-to the frequency and, therefore. the U allowable number of cycles to reach a-usage of 1.0 is inversely proportional to }* frequency. Yhe equivalent number of cycles'to give the usage of 1.0 for a differentfrequencytubeis(- 1 Ja.C i For a different service life of an unsupported tube, the ( i l t r t ja.c.e l L, 3 Knowing the magnitude of the stress ratio allows 1) the determination of tubes l that do not meet a value of 11, and 2) the calculation of maximum stress in the acceptable tubes, b [ 3a,c .(l2] Having this maximum stress permits the evaluation of the maximum fatigue usage for Indian Point Unit 2 based on the time history expressed by normalized j stability ratios for the duty cycle (see Section 9.3 g.4). [ I4 0158M:49/041888 32 4-10

? I ?> I I I 2 Table 4-1 Fatigue Usage per Year Resulting l Fr9m Stabiltty Ratio Reduction LINEA" FATIGV USAGE YEARS TO CURVET [) STRESS) SR, 5. MODEL PER YEAR 1.0 USAGE BAS 15 l REDUCTION 5. 9 yrs to 0.0207 48 fail (5.6) rs to 0.0107 9f1(7.0) 5. fa 5. 9 yrs to 0.0014 fail (8.0) l 0.0209 10. max.streg) amplitude (9.5) 0.0053 l 10. max. stres ) amplitudel (9.5) 0.0004 10. max.stresk) i amplitudel i L (10.3) 0.0142 10.

nax. strer-)

amplitudeN L (11.6) 10. max stress 0.0020 ? based on l' duty cycle (5) l (9.5) (1) This gives the basis for selection of the initiating stress amplitude and its value in ksi, if North Anna Unit 1 failed in 9 years with an initial a of 5.6 ksi. (2) S, is the maximum stress applied with S, = Smean + S

• a (3)

( Ja c (4) Cycles to failure implied by this combination of stress and fatigua i properties is notably less than implied by the operating history. 1 Consequently this combination is a conservative, bounding estimate. -(5) Cycles to failure implied by the operating history requires 1.3 sigma fatigue curve at the maximum stress of 9.5 ksi. 0158M:49/041888-33 4-11 i .^. ._.._,._,_._.m..._,,

7 l 1 I- ) .s j \\ .l j P (1!' j 1 i l 's 1 i i I A,B,C 1 t 1 1 j THIS FIGURE IS CONSIDERED PROPRIETARY i n IN ITS ENTIRETY l :,. 3 b e l-L- i e i I f. k W,, 4 o l: m 6: i I,.. ,s ir n: l' lp. p e hpure 44 VibraMon Displacement vs. Stability Ratio n: } '.l. l f' il i, a 4-12 1 -+

.i 4, -i t ?:t p A. C 't t i THIS FIGURE IS CONSIDERED PROPRIETARY .i IN ITS ENTIRETY i i i i l l 6 ^ i l j i .3 g 4 -h Figure 4-2 Fatigue Strength of Inconel 600 in AVT Water at 600'F E G - + 4-13 t ~< ? s

. _. ~._._. i ' I i l i 1 A, = C 1 i THIS-FIGURE IS CONSIDERED PROPRIETARY IN ITS ENTIRETY l ~ e \\ e s ) 1 l: i 1 Figure 4-3 Fatigue curve for Inconel 600 in AVT Water Comparison of Mean Stress Correction Models 4-14 e .ww. m.en-e u-%m-w m- =-----w--u

• e-e.

d ---r.---- - m.. --iew --m m.-

I o t i 0 (,. l t V l i l A..C f + ii t ) THIS FIGURE IS CONSIDERED PROPRIETARY l IN ITS ENTIRETY. =. l l l l l i' o k i i (- Figure 4-4 Modified Cumulative Fatigue Factor with 105 Reduction e in stability Ratio for Maxima stress condition 4-15 4.. ~.

U l f; 5.0 SUPPORTING TL5) DATA This section provides a mathematical description of the fluid elastic l mechanism, which was determined to be the most likely causative mechanism for the North Anna tube rupture, as discussed in Section 3.3, to highlight the physical conditions and corresponding parameters directly related to the event j and associated preventative measures. The basis for establishing the appropriate values and implications associated with these parameters are provided. Where appropriate, test results are presented. 5.1 Stability Ratio Parameters Fluidelastic stability ratios are obtained by evaluations for specific configurations, in terms of active tube supports, of a specific tube. These stability ratios represent a measure of the potential for tube vibration due to instability during service. Fluidelastic stability evaluations are performed I i with a computer program which provides for the generation of a finite element model of the tube and tube support system. The finite element model provides the vehicle to define the mass and stiffness matrices for the tube and its ~~ support system. This information is used to determine the modal frequencies D* (eigenvalues) and mode shapes (eigenvectors) for the linearly supported tube being considered. i The methodology is comprised of the evaluation of the following equations: for mode n, Fluidelastic stability ratio SR = Uen/Uc where Uc (critical velocity) and Uen (effective velocity) are determined by: D))W (3) 2 f-U,=$f D ((m, 6 ) / (#o n n and; N 2 I (sj/p ) U3 jn Zj 4 o o I23 U en 2 2 (m /m,) (3n J z 3 3} l 0158M:49/041888-39 5-1

E 1 t ( C-

where, i

tube outside diameter, inches l D = effective velocity for mode n, inches /sec U en number of nedal points of the finite element model N = I sj, Uj, pj = mass per unit length, crossflow velocity and fluid density at node j, respectively l p,, m, reference density and a reference mass per amit = length, respectively (any representative valses) logaritWe decmumt (damping) i = n normalized displacement at node j &n the selb mode of vibration 4jn = zj average of distances between node j to j-1, and j to J+1 = an experimentally correlated stability constant A = Substitution of Iquations [1] and [2] into the expression which defines stability ratio, and cancellation of like terms, leads to an expression in fundamental terms (without the arbitrary reference mass and density parameters). From this resulting expression, the stabihty ratio is seen to be directly related to the flow field in terms of the secondary fluid velocity L times square-root-density distribution (over the tube mode shape), and inversely related to the square root of the mass distribution, square root of modal damping, tube modal frequency, and the stability constant (beta). The uncertainty in each of these parameters is addressed in a conceptual manner in Figure 5-1. The remainder of this section (Section 5.0) provides a e discussion, and, where appropriate, the experimental bases to quantitatively 5-2 0158M:49/032988-40 L -

tl establish the uncertainty associated with each of these parameters. In addition, Section 5.3 provides the experimental basis to demonstrate that tubes with single-sided AVB support respond under instability conditions with { displacements which are on the order of the tube-AVB gap size. This implies that those tubes with single-sided support would not have to be modified I because their instability response amplitude (and stress) would be small. The very high degree of sensitivity of tube response (displacements and stresses) to changes in the velocity times square-root-density distribution is addressed in Section 4.0. This is important in determining the degree of change that can l be attained through modifications. Frecuency 4 It has been demonstrated by investigators that analytically determined frequencies are quite close to their physical counterparts obtained from measurements on real structures. Thus, the uncertainty in frequencies has been shown to be quite small. This is particularly appropriate in the case of dented (fixed boundary condition) tubes. Therefore, uncertainty levels introduced by the frequency parameter are expected to be insignificant (see j . also " Average Flow Field" subsection below). L Instability Constant (Beta) l The beta (stability constant) values used for stability ratic vid critical t velocity evaluations (see above equations) are based on an extensive data base comprised of both Westinghouse and other experimental results. In addition, previous field experiences are considered. Values have been measured for full length U bend tubes in prototypical steam / water environments. In addition, measurements in U-bend air models have been made with both no AVB and variable AVB supports (Figure 5-3). To help establish the uncertainties associated with ATH0S flow velocity and L, density distribution predictions on stability analyses, the Model Boiler (MB 3) tests performed at Mitsubishi Heavy Industries (MHI) in Japan were modeled 0158M:49/051288-41 53

I using ATHOS. A beta value consistent with the ATH0S predicted flow conditions j and the MB 3 measured critical velocity was determined. These analyses ) supportedabetavalueof[ la,b,c A summary of the test bases and qualifications of the beta values used for these assessments is provided by Figure 5-2. The lowest measured beta for i tubeswithoutAVBswasavalueof[ ]a,b,c This value is used for the f beta parameter in all stability ratio evaluations addressed in this Report (see also " Average Flow Field' subsection below). l Mass Distribucion The mass distribution parameter is based on known information on the tube and primary and secondary fluid physical properties. The total mass per unit length is comprised of that due to the tube, the internal (primary) fluid, and I the external (secondary) fluid (hydrodynamic mass). Data in Reference 5-2 suggests that at operating void fractions the [ l. 1*>c l Tube Damoina Test data'are available to define tube damping for clamped (f.ixed) tube supports, appropriate to dented tube conditions, in steam / water flow conditions. Prototypic U bend testing has been performed under conditions leading to pinned supports. The data of Axisa in Figure 5 4 provides the principal data for clamped tube conditions in steam / water. This data was obtained for cross flow over straight tuber. Uncertainties are not defined for the data from these tests. Detailed tube or.tiping data used in support of the stability ratio evaluations addressed in this report are provided in Section 5.2, below. $-4 l 0158M:49/051288-42

I Flow Field - Velocity Times Souare-Root-Density Distribution Average and U bend-local flow field uncertainties are addressed independently in the following. Averace Flow Field Uncertainties in the average flow field parameters, obtained from ATHOS j analyses, coupled with stability constant and frequency, are essentially the i same for units with dented or non-dented top support plates. If the errors associated with these uncertainties were large, similar instabilities would be expected in the non-dented units with resulting wear at either.the top support i plate or inner row AVBs. Significant tube wear has not been observed in inner row tubes in operating steam generators without denting. Thus, an uncertainty estimate of about [ Ja,e for the combined eff'ects of average flow field, stability constant and frequency appears to be reasonable. To further minimite the impact of these uncertainties, the Indian Point Unit 2 tubes are evaluated on a relative basis,.so that constant error factors are essentially eliminated. Thus, the uncertainties associated with the average velocity times square-root-density (combined) parameter are not expected to be significant. U bend local Flow Field Non-uniform AVB insertion depths have been shown to have effects on stability ratios. Flow peaking, brought about by the " channeling" effects of non-uniform AVBs, leads to a local perturbation in the velocity times square root-density parameter at the apex of the tube where it will have the largest effect (because the apex is where the largest vibration displacements occur). Detailed local flow field data useci in support of the stabil'ity ratio evaluations addressed in this report are provided in Section 5.2, below, and Section 8.0. 015BM:49/051288 43 --.m

f Overall Uncertainties Assessment Based on the above discussions, and the data provided in the following sections, it is concluded that local flow peaking is likely to have contributed significantly to the instability and associated increased vibration amplitude for the failed North Anna tube. Ratios of stresses and stability ratios relative to the North Anna tube, R9C51, are utilized in this report to minimize uncertainties in the evaluations associated with instability constants, local flow field effects and tube damping. 5.2 Tube Damping Data Thi dnping' ratio depends on several aspects of the physical system. Two primary determinants of damping are the support conditions and the flow field. It has been shown that tube support conditions (pinned vs clamped) affect the damping ratio significantly. Further, it is affected by the flow conditions, i.e., single-phase or two phase flow. These effects are discussed below in t more detail, i Reference (5-1) indicates that the damping ratio in two phase flow is a sum of contributions from structural, viscous, flow dependent, and two phase damping, i The structural damping will be equal to the measured damping in air.

However, in two' phase flow, the damping ratio increases significantly and is dependent on the void fraction or quality.

It can be shown that the damping contribution from viscous effects are very small. l Damping ratios for tubes in air and in air water flows have been measured and reported by various authors. However, the results from air water flow are poor L representations of the actual conditions in a steam generator (steam water flow athighpressure). Therefore, where available, results from prototypic steam water flow conditions should be used. Fortunately, within the past few years test data on tube vibration under steam water flow has been developed for both pinned and clamped tube support conditions. 1 \\ 5-6 0158M:49/032988 44 l

1 i I } Two sources of data are particularly noteworthy and are used here. The first is a large body of recent, as yet unpublished data from high pressure steam water tests conducted by Mitsubishi Heavy Industries (MHI). These data I were gathered under pinned tube support conditions. The second is comprised of 1 . the results from tests sponsored by the Electric Power Research Institute J (EPRI) and reported in References (5-2) and (5-3). The damping ratio results from the above tests are plotted in Figure 5-4 as a function of void fraction. It is important to note that the void fraction is determined on the basis of [ Jac(Reference (5-4)). The upper curve in the figure is for pinned support conditions. This curve represents a fit to a large number of data points not shown in the figure. The points on the curve are only plotting aids, rather than specific test results. The lower curve pertains to the clamped support condition, obtained from Reference (5-3). Void fraction has been recalculated on the basis of slip i flow. It may be noted that there is a significant difference in the damping l ratios under the pinned and the clamped support conditions. Damping is much - 7 j larger for pinned supports at all void fractions. Denting of the tubes at the top support plate effectively clamps the tubes at that location. Therefore, L the clamped tube support curve is used in the current evaluation to include the effect of denting at the top tube support plate. The Reference 5-3 data as reported show a damping value of *0.5% at 100% void fraction. The 100% void fraction condition has no two phase damping and is considered to be affected principally by mechanical or structural damping. Westinghouse tests of clamped tube vibration in air has shown that the mechanical damping is only [ Ja,c:rather than the 0.5% reported in Reference (5-3). Therefore the lower curve in Figure 5-4 is the Reference (5-3) data with all damping values reduced by ( Ja,c 1 i 0158M:49/051288 45 a ' w.

• w a.- -

I 3 I J f 1 l . Tube Vibration Amplitudes With Single Sided AVB Support 5.3 A series of wind tunnel tests were conducted to investigate the effects of tube /AVB eccentricity on the vibration amplitudes caused by fluidelastic vibration. [ Ja.c. Prior test results obtained during the past year using this apparatus have demonstrated that the i fluidelastic vibration characteristics observed in the tests performed with the a-cantilever tube apparatus are in good agreement with corresponding characteristics observed in wind tunnel and steam flow tests using U bend tube arrays. A sunnary of these prior results is given in Table 5-1. An overall view of the apparatus is shown in Figure 5 5. Figure 5 6 is a top view of the apparatus. [ I f e Ja.c, H 0153M:49/040488 52

W j As shown in Figure 5-7, the tube vibration amplitude below a critical velocity is caused by ( .)a,C . Figure 5 7 shows the manner in which the zero-to peak vibration amplitude, expressed as a ratio normalized to [. Ja,c varies when one gap remains at (' Ja,c For increasing velocities, up to that corresponding to a stability ratio of (. l .)a,c Figure 5 8 shows typical vibration amplitude and tube /AVB impact force signals corresponding to those ~ ' obtained from the tests which provided the results shown in Figure 5-7. As expected, impacting is only. observed in the [ Ja,e It is concluded frcm the above test results that, for a ( ja.c-5.4 Tests to Determine the Effects on Fluidelastic Instability of Columnwise Variations in AVB Insertion Depths L This section summarizes a series of wind tunnel tests that were conducted to investigate the effects of variations in AVB configurations on the initiation of fluidelastic vibration. Each configuration is defined as a specific set of insertion degiths for the individuni AVBs in the vicinity of an unsupported U-bend tube. L The tests were conducted in the wind tunnel using a modified version of the' cantilever tube apparatus described in Section 5.3 Figure 5-9 shows the conceptual design of the apparatus. The straight cantilever tube, ( 5-9 0158M:49/041188 47 i L l

p q o i i

i. c l

a x.. I l P' j i b j ja,c 1 L The [ Ja.c Figure 5 11 shows the AVBs, corresponding to field modified units, when the side panel of the test section is removed. Also shown is the top flow screen which is [ t 4 Ja,c The AVB configurations tested are shown in Figure 5-12. Configuration la corresponds to the final AVB configuration for Tube h R9C51, the failed tube at North Anna following a critical review of original 1 AVB Configuration Ib. Configuration 2a corresponds to one of the cases in which the AVBs are inserted to a uniform depth and no local velocity peaking - . effects are expected. [ As shown in Figure 5-9, the [ i ja,C 1. t t 0158M:49/041888-48 5-10

rp .l I i All the tubes except the instrumented tube (corresponoing to Row 10) are ( f Ja.c As discussed in Section 5.3, prior testing 't i indicates that this situation provides a valid model. The instrumented tube [ ja,c as shown in Figure 5-10. Its l [ Ja.c direction vibrational motion is measured using a non contacting transducer. (, n ja,c The instrumented tube corresponds to a Row 10 tube as shown in Figure 5 9. However, depending on the particular AVB configuration, it can reasonably represent a tube in Rows 8 through 11. The AVB profile in the straight tobe model is the average of Rows 8 and 11. The difference in profile is gyite small for these bounding rows. The (- Ja.c using a hot-film anemometer located as shown in Figure 5 9. i Figure 5-13 shows the ras vibration amplitude, as determined from PSD (power spectral density) measurements made using an FFT spectrum analyzer, versus flow velocity for Configuration Ib (which corresponds to tube R9C51 in North Anna). Data for three repeat tests are shown and the critical velocity is identified. The typical rapid increase in vibration amplitude when the critical velocity for fluidelastir. vibration is exceeded is 6vident. [ The main conclusions from the tests are: t 1. Tube vibration below the critical velocity is relatively small, typical of turbulence induced vibration, and increases rapidly when the critical velocity for the initiation of fluidelastic vibration is

exceeded, t

2. Configuration Ib (the original AVB configuration near R9C51 in North Anna prior to the critical review) has the lowest critical velocity of all the configurations tested. 3. Configuration Ib is repeatable and the configuration was rerun periodically to verify the consistency of the test apparatus. 0158M:49/041188-49 5-11

7 7 3------ e, p, The initial test results obtained in support of the Indian Point Unit 2 T evaluation are summarized in Table 5-2. The peaking ratio istdefined as the ratio of the critical velocity for Configuraticn la divided by the critical' velocity for any other configutation. - u 5.5 References n 5-IL Carlucci L. N., and J. D. Brown, " Experimental Studies of Damping and Hydrodynamic Mass of a Cylinder in Confined Two-Phase Flow," ASME J

Journal-of Vibration, Acoustics, Stress, and Ra11 ability. in Design, 4

January 1983. y iq 5-2~ Axisa F. - et: al " Flow Induced Vibration of Steam Generator Tubes," EPR1-NP-4559, May, 1906. 5 3 Axisa F., et'al, " Vibration of Tube' Bundles Subjected to Steam Water Cross. Flow: A Comparative Study of Square and Triangular Pitch Arrays," Eighth SMIRT Conference in Brussels,. August 1985. L5 4 Le11ouche G. S. and B. A. Zolotar, "A Mechanistic Model for Predicting ~ ' Two-Phase Void Fraction for Water in Vertical Tubes, Channels and Rod Bundles," EPRI-NP-2246-SR,1982. 5-12 -0158M:49/022988-50 'k m MJ

a: 7 3 1 4 Table 5-1 [ Wind Tunnel Tests on Cantilever Tube Model' 7 n k t OBJECTIVE: . Investigate the effects of tube /AVB fitup on flow-induced tube vibration. \\ APPARATUS: -Array of cantilevered tubes with end support's that [' -f f i Ja,C i t MEASUREMENTS:-Tube vibration amplitude and tube /AVB impact forces or preload -[ forces. p RESULTS: a,b,c 1 .[ l- ? 1; i l [' l f. j. 5-13 0158M:49/041188-51 6

.v. x-. 7. a Table 5-2 h" Fluidilastic Instability' Velocity Peaking Ratios-for Columnwise Variation in AV8 Insertion Depths b (Indian Point.2). .,c, p i Type of Peaking Ratio i U,/Un Insertion Configuration 3 t n a,b,e p la. } 1<" lb 2a 3 p 4a 4 4b 4C l.;... Sb Sa

i. -

u Sc ( 5e' 6a 1 '; 6b 1jo 6c l .e Note: U is. instability velocity at inlet for AVB insertion n type n. + 't, I1.. .e l-5-14 g 0158M:49/032988 <-

er p. i a. v. x ,F.. , di i

g.,

7 i, j -s. .1 ~ e,c c l Li ' i li-l t s .-t m ^.' 4 ? \\; ,.e 11-3

(

f 1-f t-1 [ + f.- I 3 h L 1: j.e-t 1 1 ll' l' l r s L t l 1 +: 4.- Figure 5-1 Fluidelastic !nstability Uncertainty Assessment 1 1 x r,' 0158M:49/041388-53 5 15 t

l o i K 'U-Band Test Data' i c= - 1)- MB 3 Tests-K

1. values of'[

Ja,b,c 2) MB 2 Tests ) of [ Ja,b,c 3) Air Model Tests. . A of, [ Ja,b,c. without AVBs Tendency.for A to increase in range of ( '*)a,b,c-withinactiveAVBs(gapsatAVBs) .) Tendency for # to decrease toward a lower bound of -l ((.]a,b,c with active AVBs 'l Verification of Instability Conditions A t 1) Flow conditions at critical velocity from MB 3 2) Measured damping for the specific tube (,. 3) Calculated velocities ~ from ATHOS 3D analysis 4)

1. determined from calculated critical. values J,

Good agreement with reported -A values 5) ATHOS velocity data with A of-[ 'Ja,b,c and known damping 1.: 11 should not significantly underestimate instability for regions of-uniform U band flow-j L I, i l* k i 5 Figure 5-2 Instability Constant - p 0158M:49/041188-54 5-16

s .~ 7 2 .a. i ~g Y a,b,cx 4 x' t t; c t,.p 4 1 c..,, _ a c fj f e a-m: a9: f i t >D p b t .le i ? (f-3

f

.-c : ~.

1..

s S 1' ) l.p 1 o 1' l' f.- i. s If; I. a l' I. 4 M M '{ t Figure 5-3 Instability Constants, 4, Obtained for Curved Tubes frc-w. Wind Tunnel Tests on the 0.214 Scale U-Bend Model 1 ? 5-17

n. %... 4.:-. m...e.

o. i >4, i b I 4 i ( .) 'b 'l e

l i.- e 3 4 ( 4 t.'f s') I j ' i t i l',~.. . 7 t; El t a. n I,I. f 4 g ' s t % 3 a,b,'> !.s J ;c.p s i., t .[

-iY i !

. t. ? e 1 s. 1 .v a ' 1 ? ? s .'h b-Jl r -4 b i .r t c', i I i .f - Figure 5-4 Darf a ve-911p Void Fraction t k A. w +3 4 9 f,- I m

,, i' N15. Aj _ s 3 h-- m' r I 4 1 Ul0 r x q -. l f v s,' h'. 6.-; ) a a,b,c - .l 2'. m' .m a 4 i' I... t c _, _\\ '? / ~+ ', 't t 1 Y t v THIS FIGURE IS CONSIDERED PROPRIETARY 4 IN ITS ENTIRETY

7..

J.. I J '.. t 4';; 'i 1 l> I p-lc. f s 1 k F Figure 5 5 ' Overall View'of Cantilever Tube Wind Tunnel Model -9..- .er ,e, '..? 253641 N, E Y. lhf'f,w;t ; ~ i

T.. n "i. - -m u Jt 3 ,s. o

.

i'.

i' I

) .e, s; ,:y y a,b,c.'

s. g.

~; 1,, I h' r +'. r s, v t;;- THIS FIGURE IS CONSIDERED PROPRIETARY 'f-IN ITS ENTIRETY r i o 3 .w. . c i' .ij i t t ( .1 4 1 4 + i. Figure 5-6 Top View of the Cantilever Tube Wind Tunnel Model W 1 .i. 1,. 'g ,'l j., 35HS2 4' s

g, p ,;tij-. H: .) =v D 4 'i !$@ <, e y u c. 4 ~ , a, b', c ' 4:L g.j. y

b E >

1 ) _i-e,: 1 E 4 1 e r 't. '.4.sc 1 iltw I-n- \\e { s .i .i k t i; s t

c J

[e'. -t s i 4 f. 4 t - t, h i ^! o .c Figure 5-7 Fluidelastic Vibration Amplitude With Non-Uniform Gaps -4 f-t 0158M:49/041388-59 5-21 Y p ^.

.+ A=: J l QR c,6,c, ;

b

, i n: .i

1 is.

i .g... i 2'; IQ a. 3 f ', t,. a [ I ,s r s'$ s x l.' g. s_ l v i-i ii i .s. I j ..-( k; 7 {'. t h o ? i c l. [-t Figure 5-8 Typical Vibration Amplitude and Tube /AVB Impact Force signals for Fluidelastic Vibution with Unequal Tube /AVB Gaps 1.i y. 0158Mt49/041388-60 5-22 u, l' id l;l,(,( s r

,,yg, .s .e

• -x t

s r: o. q j 'y

= t '.

5

j H

_.,. s e, b,'., s .) d i ( A. ar, 'l .t t P I s d 'f ) -( n i. l .i 4 I L f N' Figure 5-9 Conceptual Design of the Apparatus for Determining.the Effects on Fluidelastic Instability of Columnwise Variations in AVB Intertion Depths I0158M:49/041188-61 5 - I

p,_my..: _.. m o A e, h E 1 3

7 s

.m 7. ~ r p j 4W h. e.,. 1.,.- t

,a a,b, a?

e ~ D i l L'. ' ' '; r.., l '{ m i .,,r. 5 E ~ 4 THIS F'!GURE 15 CONSIDERED PROPRIETARY / '!N ITS-ENTIRETY \\ 6 i di o e i :i E ? ~. o.:, ? l. g.:. 'M' g '1 ) a l !5g 3. - Figure 5-10 Overall View of Wind Tunnel Test Apparatus

e

i n .253M 3 .(., ' f .i k m.- .l

~ g g,. og ,-;i-p~, . s it g._. -t 4 !U s R% sT '}; r 5 t a,b,c. ;' ~ t . r !. [ k.I' -t .. c d f,1' ,j P /t e 1 [!. .{ 7 i

t THIS FIGURE:IS' CONSIDERED. PROPRIETARY d

.IN~ITS ENTIRETY r 9- , k

p..

[ lli. >r e e" ' 4 :' a. l-s l. t E, L e M' g 1 ,< 1I. .o L Figure 5-11. Side View of Wind Tunnel Apparatus with s'A' ver Plates i Removed to Show Simulated AVBs and Top Flow Screen + Ni i l, i 'Y ? ,c_

mw

.y y. N-

f.,p; J

f yx ,.s., l' h E! ', ' .r,> t 4 TYPE OF AVB ' ME OF AVB - i WSERTION ' NSERTION - e,b,c e,b,e 4 1a 5a s t b.. - 1b- .5b 2a 5c \\- u.K,, 3 5e

• ).

da .6a a i - 4b-6b 4c 6c V \\ lo r \\ 1 1 i

f. _

Figure 512 AVB CONFIGURATIONS TESTED FOR INDIAN POINT 2 R I..

4..

l y-n 5-26 r;, 9; l ' zll

• j

g y:y,x t. } ..u,

e e,b,e 4

i 1.. ls .c i i + h 'l a a L ti I s t- '4 !. ) e!- 3 s ~" Figure 5-13 Typical-Variation of RMS Vibration Amplitude with Flow ,y Velocity for Configuration la in Figure 5-12 r s' t' i 5-27 t -r ,I 4. 't

i s 6.0. EDDY CURRENT DATA AND AVB POSITIONS n > 1 ~ 6.1 Tube Denting at Top Tube Support Plate Eddy current tr.$t data from the 1984 and 1987 outages were examined to assess the. incidence vi denting at the top tube support plate for the tubes in Rows 8 through 12 of the bundle. This examination' indicated that the occurence of s denting was~so frequent for these tube rows that all the tubes were assumed to be dented at the top support plate. 6.2 Tube Wall Thinning at the AVB Supports No tube wall thinning was observed at the AVB/ tube intersections in Rows 8 through 12 of the bundle. 6.3 Eddy Current Data for AVB Positions p-The AVB insertion depths were determined principally on the basis of direct observation from the eddy current. data. To directly locate the AVBs, the ECT data traces were searched for the characteristic signals-which' indicate-the 1' I intersection of an AVB with the tube (Figure 6-1). The' number of these h intersections, including zero, were reported for each tube to indicate the-presence or absence of AVBs. Where only a single intersection was-indicated by the data, the length of this intersection was also reported to provide additional information to assess the ad?quacy of support for the tube. The AVB visible / invisible data and single contact are lengths are summarized on-the AVB 1 insertion maps, Figures 6-2, 6-3,~6-4, and 6-5. 6-1 '0159M:49/041388-1

,t J Since ambiguity can occur in the interpretation of the ECT data, due to inability.of ECT to' differentiate at which side of a tube.a " visible" AVB is .focated, other information was used to assist in establishing the location of the AVBs. Consistency' with the design of the AVB assembly, consistency of data for adjacent columns and verification by projection were utilized to determine ths. final depth of insertion of the AVBs. 'For the cases of single ~ AVB contacts, : verification by projection was used in many instances' to confirm support of the tube with'a single contact. 2 - 6.3.1 Steam' Generators 22 and 24 Data from the 1984 outage were reviewed to determine the positions of the AVBs. The Eddy Current Data Analysis was performed principally by the EPRI NDE Center j and is summarized in Reference 1. Westinghouse performed data analysis for a ~ few tubes to determine the number of AVBs visible and the length of single AVB contacts. The EPRI data interpretations were the principal bases for ( establishing the AVB positions, and the Westinghouse reviews were used to provide additional confidence in the resulting AVB positions. L' For a number of tube column's, the AVB positions could not be confidently - determined, either because the ECT signals were masked by copper deposits on the tubes, or'because the. data indicated single contacts or contacts less than 2-inches separated. The position of the AVBs in the tube columns identified in Table' 6-1 were established by ( ja,c 6.3.2 Steam Generators 21 and 23 'P The'ECT data from the 1987 outage were reviewed to (a) determine the presence of AVBs in-Tube Rows 8 through 12 directly from the data, and (b) to provide-data for ( C Ja c This review was performed on-site by Westinghouse. (1) Letter from EPRI NDE-Center. K. Krzywose to con Ed, B. Greene, dated ~ 10/30/87, " Locating Anti-Vibration Bars from 1984 ECT Data". 62 0159M:49/051288-2

,jy o l 6.4L AVB Insertion Depths l lf Figures.6-2'through 6 5 are the AVB position maps for Steam Generators 21 a through 24, respectively. Previously plugged tubes, AVB visible indications,- L single contact are lengths and projection results are also shown on these figures. The direct observation data are the principal basis for determining the AVB l positions in Steam Generators 22'and 24. Where the direct observations were ambiguous and there was a conflict between observations and projections, the more conservative data were used to determine the AVB positions. Greater i conservatism is generally interpreted as the AVB being less inserted although consideration must also be given to the resulting flow peaking factors. No T attempt'was made to inflate the flow peaking factors through conceptually possible, but unreasonable, interpretation of the data. For example, in Steam Generator 22 the presence of a single AVB contact in Tubes R11C46 and R11C47 could provide a basis for further insertion of the AVB between these two columns, thereby raising the peaking factors for the Row 11 tubes in Columns 45 and 48. This was considered an unreasonable interpretation in light of the projected AVB positions for Columns 46 and 47. 3 The AVB positions for Steam Generators 21 and 23 were determined entirely by projection. The judgement was made during the inspection that the data for the I tubes-in Rows 8 through 12 would not reliably locate the AVBs, and that the most efficient method for locating them was by, projectic.. 6.4.1 AVB Assembly Design The design of the AVB assembly for the Model.44 Steam Generator includes a [ 3a,c.e ^ 6-3 0159M:49/051288-3 r

m l s T - {- ja,c.e f ~ 6.4.2 [1 ja,c 1

The adequacy of support provided by [1 Ja c indicatioris must be 4

resolved in the cases where a potentially susceptible tube is concerned, since plugging the tube may be required if adequate support cannot be shown. Preliminary analysis indicated that' Row.10 tubes with peaking factors greater - than the North Anna failed' tube and any tube in Rows 11 and 12 were potentially susceptible. Consequently, the primary focus regarding (: Ja.c was for the tubes-in Rows 10 through 12. [ ~Ja,c were utilized as a basis for determining-the support conditions of the tubes only when [' Ja,c data were not available. l [1-r- )**C data were utilized to determine the AVB positions l '. in Steam Generator 24, Columns 85, 87, 88 and 90. b Table 6-2 summarizes the potentially susceptible tubes with [ la,c - indications in Steam Generators 22 and 24, and the resolution of the support conditions for these tubes. Steam Generators 21 and 23 were evaluated exclusively by the [ Ja,c method and, therefore,_do not have any [ Ja,c to resolve. l' lt L 6-4 3 0159M:49/041388 4 a m m

7 un q; II. : ..1 1.I q J 2

6~ 4.3 L AVB-[

'Ja,c- 'i x [. Ja,c is useful where noisy ECT! signals prevent direct observation- 'l 4y of, the AVBs, where: testing is impossible due.to plugged tubes, and in some' j instancesto'resolveambiguitiesintheECTdata..The[-- ] j it .i, ja,C: t - For Steam Generators.21:and 23; and for some of the columns in Steam Generators-y 22 and 24, the AVB-characteristic signals could not be confidently determined. 1 "~ J due to:a noisy signal or pre-existing plugged tubes. In these instances, [ -s j l 1 t 'g a i I ja,C ~ V 0159M:49/041388-5 c-WW v s

1 ) 6.4.4 AVB Map. Interpretations i E Steam Generator - 21 t The AVB positions-in Columns 3-4 and 89 90 are based on data from a single j + tube. The best [ Ja c is based on the tube most near the actual pointofAVB[ Ja,c For these columns, there is no basis-for checking the single tube projection. However, tho' data upon which these : [ Ja,c are based are consistent with the [' Jac-thus there is also no basis..to doubt these [- Ja,c Based on the s projections, the Row 12 tubes in Columns 89 and 90 are unsupported and the s R12C3 tube is supported. i Steam Generator -22 The indications in Row 12 in Columns 2 and 91 are interpreted to be the retainer ring-as noted above. These tubes are assumed to be unsupported. The Row 12 tubes in Columns 4 and 90 must be supported because the adjacent p column of tubes is supported and.no AVB exists between Columns 2-3 and 90-91. I i; Steam Generator - 23 The AVB map for' this steam generator has no anomalies'and is self explanatory. Steam Generator - 24 - Data were unavailable for the Row 12 tubes in Columns 45 through 49. The [- Ja.c in the Row 11 tubes in Columns 45, 46,'49 and 50 - indicated support of the Row 12 tubes. Data available for [ Ja,cin Column 50 verified support in Column 50. Since there is no evidence to the contrary, the Row 12 tubes in these columns are considered supported. 4 6-6 0159M:49/041388-6

( b \\ l-p !'-(r, s ,4 t.' ~.p ^,_ 'i p;

z ;

'f t ) .j. O> T6-5'UnsupportedTube, Summary-e Table 6-3 suussarizes the tubes-forlwhich. support cannot confidently be assured! iJ . based on the depth of insertion evaluations above. ! Tubes are adequately-l, j)g

supported if they!are supported by at least'one AV8,.

w ' e i i .p h' y\\ e. t, t i f'.. i t 'L s w I -? I t r i k '( L 'j 4 k-o L, j. t h 1,. i ? ) l. ie 1 ..i b 3 ? 6-7 0159M:49/032988-7' L.: s > -o

9; } n. ,f.- TABLE 6-1 6 3, UTILIZATIONOF[ Ja,cMETHOD j FOR DETERMINING AVB POSITIONS i Steam Generator 21 'All Columns Steam Generator 22 Columns 2, 3, 5, 11, 12, 22, 25,-31, 32, 37, 46, 47, 52, 67,_70, 73,-80, 85, 88, 90 p Steam Generator 23 All Columns ) Steam Generator-24 Columns 4, 5, 6, 12, 13, 19, 21, 27, 28, 29, 30, 32, 33, 36, 38, 44, 45, 46, 49, 50, 51, 61, 64, 65, 69,- 70, 71, 72, 73, 74, 75, 76, 81, 82, 83, 85,-9' b h k l i t s a + i9 s % t b. 6-8 , 0159M:49/041380 8 7 1 i

(. T U TABLE 6-2 RESOLUTION OF SUPPORT CONDITIONS FOR-R0W lo, 11=AND 12 TUBES WITH SINGLE AVB INDICATIONS -l TUBE LOCATION RESOLUTION OF SUPPORT CONDITION Steam Generator 22 i l-R12C2 Assumed unsupported. Indication is retainer ring. R12C3,4 Support verified by (projection]a,c. R11C5 Assused unsupported R10011,12 Non-critical peaking. factor; not susceptible R10C24,25 Non-critical peaki,ng factor; not susceptible R11C46,47 Non-critical peaking factor; not susceptible j R11C52 Non-critical. peaking factor; not susceptible < R10C67 Non-critical. peaking factor; not susceptible R10C70 Non-critical peaking factor;, not susceptibles j R10C73-Non-critical peaking factor; not susceptible' i R11088 Non-critical. peaking factor; not susceptible R12C90 Supportverified.by'(projection]a,c. R12C91-Assumed unsupported. Indication is retainer ring.. Steam Generator 24 R11C4,S Assumed unsupported-R1006 Non-critical peaking factor; not. susceptible R11C45,46 Assumed unsupported. L R11C49,50 Assumed unsupportet i L R10C84,86 Non-critical peaking. factor;l not. susceptible; + R1'1C87,88 - Non-critical peaking factors; not' susceptible. l R11C89,90 Tubes. supported; (arc length]a,c > 3. inches; _ no AVB. ~ in column 90, 91 - R11C91 . Assumed unsupported R12C91' Assumed unsupported-t i ' 0~9 [ 6 0159M:49/041388-9 s t------ a

i _ 'i D Table 6-3 Unsupported Tubes Summary-l:.. L (. j ". i 1: Steam Generator 21 1 l. L Row 8 all Columns Row 9 all Columns l t l 6-13,'11,: 17, 20 11, 28, 33-57,= 1 Row 10' . 2 1, 4, 5, 11,.69, 70,~ 73-79, 82-87,'88 91; l Row 11 2,1,1,37-53,ti,AZ,38,'31,;90,91~- . Row 12' 1,31,39,91~ t Steam Generator 22 [ t Row 8 all Columns ' I i Row 9 all Columns except 80 s Row 10 2-5, 11-14,-11, 25, 34-61, 64'-70, 73'-78,!88-91 b . Row 11 2,1,4,5,1Z-52,:88. Row 12 1, 91 4 ' Note: Tubes show in Boldface are candidate tubes-for. further evaluation Q viz the N. Anna tube rupture issues. h . 6-10, I 0159M:49/041188-10 4

L. L- .g y Table 6-3 Unsupported Tubes Summary. j .(Continued) Steam Generator 23 Row 8 all Columns. A Row 9 all Columns o Row 10 2,1,~4,15,1 16,;17,'10,35-54,81~,82,86,87-91 1 Row'11 1, 1, 1,.32, 52 54, 91' J l l Row 12 2, 91 1 -l Steam Generator 24 Row 8 1-20,'27-3'3, 36-61,-70-92 1{ 4 Row 9 2-7, 38-61, 70-74, 78 Row 10-2-6, 40-55, II,'86, 87, 88-91. Row 11 2',1,4,5,45,16,49-11,91-1 Row 12' 1, 91-t Note: Tubes shown in Boldface are candidate. tubes for further e. valuation' viz the N. Anna tube rupture issues. i h-6-11 [ 0159M:49/040488-11 1.

f WESTINGMOUSE FROPRIETARY CIAss:2~, 1 .) ,4 '. l I +.. - mer. r-em - e. = ammmm. m en l f. C/ - aus - a.sem m -me l ; me WW } W t, -. e..e am assum 1

5..U.sm.5. "

) 5 o 4 -( i eenen e g l f ] Na ^ ( SO:: .{ Clear AVB Signal m... o l i~ W **m" E M so - - e ammuus se a i. 1 mass summme ammme W W 1 E=u H, numan. me an W l [ q -

== - aan - s. am g m ammme-T e-sumum-e i g i P-3 e w m> e. t l N

i

== *

I s

.ma.ne.m -- m $8F - ~ .c -L Multiple AVB Signals As Clear Call ( Figure 6-1. AVB Insertion-Depth Confirmation 6-12 L i I -.m - m - - - =. - - c-. .me. r-e-g ee.

1. a A

,i i l: -6 h t aim.n U I# S se ar me m.emens q .oooooo$izi;h;h?GiGi(h?(h;i;$;i?Gh;hzhzh?DYO4l OO 3 r--mui r> <i r-2 <=---i t----i u v 0 5 I ! 0 0 0 0 0 0 6 6' ' 'U U 6 6 E O O O Ox< t 7 00000000000000000000000000000-7 000000000000000000000000000000- [ ............:......=... C CiTO'6'6'66'66 o 6406"O'4W6'666D 6 o o( 1i [XXXXXXXXXX XXXXXXXXXX] [] "7 KXl [] [] [] E a=a5 5 5 5 1 .'6 0 0000000000000000000000000000 3 0000000000099900909000000000040: 0000000000000000900000000000000 2 .a l sms anna acna:sturns nemi at's Isi:starituraracaruracIsras as usuiu na ns:monali *4,C ~ 7 oo ci, o o o ci 2 m ci, n o ci-4, o c4-4-4-A-A-A-A-a-a-A i o 1 ? =45 8 5 5 ?A5 R 7A=> 5 5 <=A=4=2 & Fa=A=A=A=A=&=A=1666o = I 5 d U d d d6dld6dU d 66'6 d'6' ' '66 ' OOOOf l 000000000000000000000000000000' J L { 0000000000000000000000000000000; .s . ( i i.,i.i.i.i.i.i.i i...,i.i i i..i.i. i.,i.i.i. i.i.i...i n,5.C' T.,,,.. -... f ei j l i( l AI L. 4 l. i C Figure 6-2. Indian Point 2 Steam' Generator 21:AVB' Positions' ,7 6-13 .1 _ _ _1_ _ ; __

q t q q a ] ammanMsYeo as u 'a_a_a_a_a_a_a_ _a_a_a_a_a_a n_a_ O w_a_a_a_ _ My U y y y y y y L" " 7 =2727 X ' 2,;,1727 7;7;7 7} [ OOOeT U U U u u u u t= = = = =i= =J = = = = = = = = =v a .. C O S OCiO O 9 0 00 0 000000@0090090000 09000090G00900000000000000000 .OO9OOOO00OOOOOOOOOO00OOOOOOOOO-1 j _a _ a_ a_a_ e lale_a_ a_ a_ a _a _ el a_ a_ R_ a _ a_ a_a_a_ e_a_a_E _._a_ a_ a_ q.' o n ). = .===_=======_=_===========_===J 90000990'OO'6600'OOOgqus0 Os. L L 000000000000000000000000000000 0 i 0000000000000000000000000000000; i 9090000000000000000000000000000' r-i.i..i.i... i.i.i.i.i..i.i.i.i.,i.i.ioi.. 1.i.i.i

i..

i.i. g. q [ q n o cA-A, ci 2 o o o o o cai -A-A-A 2 m n 5 c4-Arkae: ' 12 m n n l i a a i=> 5 5 5 5 5 <=L=A=> <=4=>5 5 5 5 5 ?A=&=, A=4=4=> #2 8 5 d?g'66c i =5655 5 6G'6'O ~C6 66 6 6 6" ' '@~9T 3 4? 6OOOO I ~ .000000000000000000000000000000 l L .0000000000000000000000000000000 L ,, & i..i m i...i.i..i.i..... ............i.,i .. i,......... i. ,u I .,ma g. .m .....m. u. .g ju. J S O - a .I ) 1 -i i q rigure 6-3. Indian Point 2 Steam Generator 22'AVB Positiens! l 5 6-14 j: ..s... ... O

u_,b; y I" e 4 l e-me M e"se.m - m e reemene i "7 o,-A, ' *-A,=io o c A,=-A,=-A,=2 o o o o o oo oo o o o o oo oo O w= =,=.,_=.J = = = = = = = = = = = = = = = =

= =

u v u= = =J u v u u u u u u u u v u v u v u v q h l :a: 000000006000000 0 0000000000 0 7 00000000000000000000000000000 7 000000000000000000000000000000-i..........,.......,....... ; "E O O O O O O O FE-A-E^k' k $1OOOOOOOOOO'vuuvlv;lv;$$$$$$J($$$$1Q v v v=_= =J t=_= = =J u u u I ; 0000 0 0 0000000000000000'00000'00 00 7 0000000000000000000000000000000 0000000000000000000000000000000' t r-a uns staa as na annens ararmenuaracantururucuracanarsa.ur~nsins:na.nrnrum!n ' 4,C ' a ) - [.. i 7 m o n o m o n o n o m o n n o ra-1, o n o n o n n n 5gg>o- - t r j l-5 5 A 5 5 5 5 5 5 5 5 5 5 5 5 :=4=A=2 6 5 5 5 5 5 5 5 coo. 1 2 6666666666666666'6'666666 6666000 i 000000000000000000000000000000-l L .I-0000000000000000000000000000000. 5 _i - i.,i-i i i.,i-i-i-i-i.i i-i-i i.,i-i .................... g ,e-L l. 1 =- \\ T* l * O **"

• r l.

P l-1; ~b . Figure 6-4. Indian Point 2 Steam Generator-23 AVB Positions 1 7 6-15 2 Y

4 n i i l' l' t, a .o aims M Y s o.es 'AWS Pesmens 1 o._ _a_a_a_I_a_'_'_ _i_ ca_ sis _a_ _i o [i [, i o [ i [ i [ i [ [ X - X. =X. =2 . M.M.X i [i [ i [i [ : ri o [ i [, [: _.=-= = = = = = =

.=.= = = = = = = = = = = m al ra ] [ ] [ ] [ ; [] [J [ ] [ ] [] [ EMXX] [; [ ] [] [ ] [ ] [ ] [ ] [ ] M p_3 = = = = = = = =

.=.=-=.

= =

=

= O O O O O v=.= =www w w w w w w u v utw v c : < " ' ' " > < ' i 00000000608000C S @OOOOv 'iOC C ..0000000000000000000000000@SO00 j _ a _ a _ a _ a_ s i s _ si n_ a _ a _ a _ a _ s'_ a _ I_ s i ale _ _ ( d [_i ti [ ] [] [i [] [ XXXXXXXXXXXXXXXXXXi [] [ i [] [ i [] A3 6 6 6 6 6 0 6 5 6 6 6 6 CMEMMi9'sWO5'WFW'00 a 0 0@@000000000000000 0 0 0 0 0 0 0 000000000000000000000000 O O 0 0 0 800.0000000000000000000000000@@@: 4 l -a,c - ~ ' '-4-h e - - o r i o o o o o o o o o o o o o o o o o o o o [ i 6 6 6 6 C =i 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 (y'eeOOO. gg6o = = 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 e 0 0 0 6-J O O O O O E' O @sessO O@@OSOOO900: . 699996SGS890GOOOOO9000000000000 I e.m I M 1 WI WI WI 'FI "I WI "I "1 "I l *! "1 "I l "I WI "I "I WI " I W l ' I

• I 7 } ' I 'I 8'I ' I ; r j a,,.

{ L ..".2.1:"".:'".":':. G. === J,. t , l. Figure 6-5. Indian Point 2 Steam Generator 24:AVB Positions 'I 6-16 i i k

.=.

I 21

m ll .I 7.0- THERMAL AND HYDRAULIC ANALYSIS This section presents the results of a' thermal and hydraulic analysis of the flow field on the secondary side of the steam generator using the 3 D ATHOS computer code, Reference (7-1). The major results'of the analysis are-the water / steam velocity components, density, void fraction,' and primary and secondary side tube wall temperatures. The distributions of the tube gap ~ velocity and density along a given tube were obtained by reducing tha ATHOS. results. In the following-subsections, the ATHOS model and some sample results of analysis are described. Because of the staggered anti-vibration b'ar insertion-configurations, loca' flow peaking occurs at certain tubes in the' U-bend. Its-effect on tube gap velocity perturbation was obtained using _ test data and-applied to the indian j Point Unit 2 steam generators.- Normalized stability > ratios over the operating history of the plant were determined based on the reported plant operating history. The results of these investigations are also presented in this 1 section. ) 7.1 Indian Point Unit 2-Steam Generator Operating Conditions p The recent steam generator operating' conditions for Indian Point Unit 2 - provided by Con Edison are showr. in Table 7-1. The' number of active tubes-which were plugged were reported to be 177,;249, 149, and 201 in:. Steam { Generators 21, 22, 23, and 24, respectively.. In addition,- it.was reported'that l the downcomer flow resistance plate was. located in the'." uppermost". position; its location is pertinent to the secondary side hydraulics calculations. With these data the Westinghouse _GENF computer code calculations'were performed i to verify the plant data and to establish a complete list of operating conditions' required for. the ATH0S analysis. _ The GENF code determined <the. t L primary side temperatures and feedwater flow-rate required to obtainzthe l-specified steam-pressure at the given power rating.' The GENF computed steam = generator Thot is 579'F, and the computed T old is 525'F. The. calculated feedwater flow rate is 2.93-x:10 lb/hr. Of primary importance ~ i 0159M:49/041388-17 7-1 .l.

Ji 't <t are the secondary side flow conditions (including circulation ratio) and steam pressure. The operating conditions utilized in the ATHOS 3-D analysis are shown in Table 7-2. (.. The circulation ratio at the current operating conditions is-4.4. This full l load circulation ratio has not varied much over the life of the plant. It has remained in the range of.4.3 to 4.4. As the~ operating. load decreases from full load, the steam flow rate decreases, and:since the total' steam and water flow-a rate through-the bundle does not vary significantly between SM and 10M load, the circulation ratio will increase proportionately. The loading on the U bend tubes will decrease with power, and. at very low power. levels the-tubeLloads-will. be much smaller than that.at 55.- The circulation ratio at 25 load would be about 4 times greater than that at ful1~ load. t 7.2 ATHOS Analysis Model Tha ATH05 analysis model for the steam generator consists of '6000 flow cells in polar coordinates with 20 divisions in the circumferential-(x-axis), direction,. 10 divisions in the radial (y-axis) direction and 30 divisions in the axial (z-axis) direction. In the ATHOS analysis, the steam generator is considered to be symmetrical with respect to the diametral: plane of. symetry of the ' tundle. The mode 11 therefore, consists. of one-half of the hot-leg' and one-half - of the cold leg sides of steam generator. Figures 7-1 and 7-2 show the p1' n and a the elevation views of the model, respectively.- The figures show the = l distributions of the flow cells. Figure 7-3 shows the plan view of the hot leg side of the model with the tube layout arrangement superimposed. This figure illustratus the locations of the tubes in the various flow cells. It is noted that a narrow ring of flow cells was used to represent.the bypass flow area in. E the annulus between the bundle periphery and the. wrapper. -In-addition,L small I radial calls were used to represent the tubelane area. The wrapper, opening-area was represented by three axial layers of cells. In the U-bend region the model consists of 11 axial layers to provide sufficient detail for analysis.. I l l ^i 7-2 0159M:49/041888-18.

.I r 1 The ATHOS model does not include the capability _to model the presence of the l 1 AVBs in the U bend region. However, Westinghouse has modified.the'ATHOS' code t to include the capability.to model the AVBs at uniform depth of lnsertion as j [ 1 l-j o i l' -i ) L, l l l i j Ja.c This methodology has been utilized in the Indian toint Unit. 2

analyses, j

l 7.3 ATHOS Results-1 The results from the ATHOS analysis consist of the thermal-hydraulic flow 'j parameters necessary to describe the 3-D flow field on the secondary side of-J l. the steam generator plus the distributions of. the primary fluid:and mean tube wall temperatures.- Since;the velocity. components computed by ATHOS'are defined ~ on the surfaces of a flow cell, the tube gap. velocity and density distributions- 'l along a particular tube required for tube vibration evaluation are-determined -- by_a post-processor from'the ATH0S output. 'The post-processor generates a data l l file which contains this information for all the tubes in the model and'the-- file' serves as part of. the input data required for tube. vibration analyses'. d l i 0159M:49/041388-19 7,

l H 9 b Because the majority-of the flow ~ cells contain more than one tube inside a cell, the tube gap velocity and density surrounding a tube are obtained by _l

interpolation of the ATH0S calculated velocities (defined on the cell surfaces) i

.and density (defined at the center of the cell). The post-processor performs L the necessary interpolations to detemine in-plane.and out-of-plane velocities at specific intervals along the length of the tubes. l 7 l l-Figure 7-4 shows a vector-plot of the flow pattern on the vertical plane of symmetry of the steam generator (the vectors are located at the center of the flow cells shown in Figure 7-2). It.is seen that in the U-bend. region the 7 mixture turns radially outward, normal to the curvature of the bends toward the-region of least flow resistance (i.e., outside the'done formed by the U-bands). 'I Figure 7-5 shows the resultant vectors of the radial and circumferential velocity components on the horizontal plane at Z=22, the fifth plane above the< 'f top tube. support plate (see Figure 7-2). The radial outward flow.is more evident from this figure since it' ignores.the axial component. It may;be noted that the radial velocity at this. axial location is low at the center of the bundle and increases with radius. Figure 7-4 shows that'the axial component is about four times greater than the radial component. Figure.7-6 shows the flow' pattern (resultant of the radial and circumferential. components)lon top of the [ i,3esheet. Because of the thermal syphon action.(resulting from the higher 1 heat flux and vapor generation-in.the hot-leg), a portion of the cold leg side fluid flows to the hot leg side before turning upward. The relatively high in-flow velocity along the tubelane from the wrapper opening is: also^ evident. Figures 7-7, 7-8 and 7-9-show a sample of. the individual: tube gap velocity and density distributions along three tu'bes at Row 10.. In each figure'the gap: velocity and density along the. length of the tube are plotted from the hot leg. l tubesheet end on the left of the figure to the cold leg and on the.right.- L Figure 7-10'shows the plot of the average in-plane gap velocity normal.to the L tube and density profiles as a function'of the column number' along Row.10. The l average values were taken as the numerical average of the parameter over the entire ICO*' span of a U-bend at a given column location. Both velocity and i density are-higher at the bundle periphery than in the interior of the bundle. The change of the average density from periphery to interior is.slightly more than the average velocity. L 0159M:49/041388-20 7-4

,) a Figure 7-11 shows the distribution 'of axial velocity above the top tube support h plate ~(TSP) as a function of radius within the tube bundle. The radial l distribution at three different circumferential_ positions are shown: IX =.1 is near the middle of the hot leg; IX = 8 is in tho' hot leg near the tubelane and l - IX 20 is near the middle of the cold leg (see Figure 7-1). It may be noted l ~ that the velocity increases with increase in radius. Figure 7-12 shows a similar plot for a 51 Series Steam Generator like those at. North Anna. The l profile is very flat in this casa. This means that the flow approaching the p U-bend is more uniform in the 51 Series Steam Generator. .j -{ A comparative profile of the axial velocity distribution is shown in Figure 7-13. The ratio of the axial velocity at a cell in the Indian Point' Unit 2 ~ Steam Generator (shown in Figure 7-11) and the axial velocity at the corresponding cell (similar location in the tube bundle) in the 51 Series. Steam Generator (shown in Figure 7-12) is plotted along the Y-axis. Of primary j interest in this figure is the sh' ape of the distribution rather than the-absolute values. It shows that in-the Indian Point Unit 2: Steam Generator, as one moves from the center toward the wrapper, the velocity increases more than in the 51 Series Steam Generator. It.is believed that the smaller tube: pitch f in the 44 Series Steam Generator (for the same tube size)'resulting in higher cross flow resistance in the U-bend is' responsible for this difference in flow-profiles. More flow migrates towards the wrapper'as it approaches the U-bend due to the higher flow resistance in the U-bend in' the 44-Series. l t The net result, for the 44 Series, is a small' reduction in the force-2 (pV ) on the tubes in the inner rows that'are of interest in the current analysis. This effect could not be predicted from a.one-dimensional analysis. - 1 7.4 Relative Stability Ratio Over Operating History L I b One aspect of the evaluation of the Indian Point Unit 2 Steam Generators.is to examine the operating history data and use it to determine the susceptibility l }.. to fatigue from fluidelastic vibration resulting from the 14 years of l operation. 'This assessment has been completed through the.use of a parameter . termed the normalized stability ratio. The normalized stability ratio compares-4 the flui~delastic stability ratio for each period of a plant's operation (fueli I i 0159M:49/041888-21 '7-5~ s,e e s --w-- m

-l 1 i I cycle) to a reference stability ratio based on a recent operating condition. A-plot of this ratio against operating time,' therefore, provides a relative: indication of the effect of past operation on-the plant's-fluidelastic stability ratio. This normalized time-dependent ratio is subsequently combino with an absolute stability ratio for the reference operating point derived from-detailed three-dimensional thermal / hydraulic and tube vibration calculations. High values for the. net stability ratio, in particular, over a significant period of operation, coupled with other prerequisite conditions (e.g., absence: of AVB support and den, ting at the top tube support plate), could indicate an increased susceptibility to fluidelastic vibration instability and fatigue. The fluidelastic stability ratio is defined as the ratio of the~ effective fluid i velocity acting on a given tube to the critical velocity. at which large amplitude fluideltstic vibration initiates: { Fluidelastic Ueffective Stability Ratio, SR = [1] Ucritical at onset o'f: instability In this ratio, the effective velocity depends-on the spanwise distributions of flow velocity and fluid density, and on the' mode shape of vibration.f The l ~ s critical velocity is based on experimental data;and has been:shown'to be. dependent upon the tube natural frequency, damping, the geometry of the tube, the tube pattern, and the fluid ~ density, along with the appropriate correlation coefficients. The detailed calculation of this ratio using spansise velocity: and density distributions, etc., requires three-dimensional thermal / hydraulic and tube a vibration calculations which are very time consuming. : Alternately, a simplified, one-dimensional version of this ratio has!been used to provide;a [ ~! more rapid, relative assessment technique for determining the effect of pastL operation on the stability ratio'. The normalized stability ratio is defined by' the following equation: 1 i 1 0159M:49/041388-22 .7-6

j

'I i In this equation "cyc x" refers to each fuel cycle and " ROP" to the recent-operating condition.. While this simplified approach cannot account for j three-dimensional tube bundle effects, it does consider the major operational j -parameters affecting the stability ratio. Four components make up this-ratio: a loading term based on the dynamic pressure (pV ), a. tube incremental. 2 mass (m) term, the natural frequency of the tube (f ), and a damping _ n ratio () term. It should be noted that the ratio is relative, in that each component is expressed as a ratio of the value for a given. fuel cycle to that of the recent operating point. f [ i JC The R particular damping correlation which is used for all normalized stability ratio calculations is based on a dented condition at the top tube support plate (a clamped condition, as discussed in Section 5.2). The clamped condition is also assu:::ed in calculating the tube-natural frequency. ~ The reference stability ratio calculation for Indian Poir.t Unit ~ 2_wasL based on-the following operating parameters which are for a recent operating point in. 1 Cycle 8 as supplied by con Edison: Steam Flow 2.93 X-106 lbm/hr Steam Pressure 690 psia r. Circulation Ratio [. ja,c (Westinghouse calculation) i 1 I L 7-7 L '0159M:49/051288 23

J A L A series of calculations have been completed to generate a normali;:ed stability. t l ratio for each of the 8 fuel cycles s'. ace the plant became operational in May. j 1973. Data for this evaluation was also supplied by Con Edison and is l tabulated in Table 7-3. Included are cycle average values for full load steam j pressure and pr5 mary fluid average temperature. The equivalent number of plugged tubes are also listed along with the number of days that the plant has operated above 90% of full poor. Since tube vibration and possible fatigue are associated with full power operation, only these highcr power operating periods are considered important to the evaluation. The operating parameters ~ listed in Tat'le 7-3 were then input to the Westinghouse "GENF" computer code to-l determine tha overall performance of the steam generator, in particular, the = l circulation ratio for each fuel cycle. These calculated values are also listed l ~ in the table, l The resulting normalized stability ratios are shown in Figure 7-14.- In this i figure, the normalized stability ratio for each fuel cycle is plotted against cumulative operating time above 90% nower. Note that the r.atic assigned to: l each of these high power operating periods has been conservatively based on-a full power calculation. Figure 7-14 indicates that the normalized ratio has" o i be(n constant throughout most of the eight full cycles'. However, early-in'- ~ operation, during Cycles 1 and ?, lower ratios-are calculated as s result of i operation with higher steam pressures. Higher steam pressures result in higher U. bend dansity, lower U-bend velocity, and increased damping as a result of;, lower voids in the U-bend. -The higher damping, together with decreased loading on the tubes, result in the lower normalized stability ratios which are L indicated for Cycles 1 and 2. l l r 7.5 References l i 7-1 i.. W. Keeton, A. K. Siiighals, et al. "ATHOS3: A Computer Program for Thermal-Hydraulic Analysis of Steam Generators", Vol.1, ?, and 3, EPRI NP-4604-CCM, July 1986. I p. r L: 0159M:49/041388-24 7-8

e~ r t 4 Table 7-1 Indian Point Unit 2 Steam Generator Operating Conditions Power 694 MWT i Steam Pressure 690 psia Feedwat'er In19t 416*F i Temperature Water Level .40-45%of_ Narrow Range Span f i 7 =i; .P .1 L$ i 0159M:49/041388-25 7-9 l

.. ~ ~' l:o 'l .J Table 7 j Steam Generator Operating Conditions Used for

)'

L. ATH05 Analysis l :. Power 694 MWT Primary Flow Rate 3.48.x 10 Ib/hr - l 7 Primary Inlet Temperature 579'F Primary outlet Temperature 525'F i Feedwater Flow Rate 2.33 x 106 lb/hr Feedwater Inlet 416*F-1 .i Temperature-Water Level from Tubesheet 446 inches I Steam Pressure 690 psia Circulation Ratio' 4.38 i e I k j 1 ? l l b b i } e 7-10 0159M:49/032388-26

ai t-i r Full lead; llueber of lineber of. Calculated-FuelCele Full f.ead flees Fluid Avwest -futes Days Above FullI.ead h haiaatae - g P* essure losial f**oeceture (DeeF) Pluened o' et Peeer-Circulatien patic l i e,c = i !A 5173 6/75 000 569.5 '108: M H-l' i 31 II 6/75 ' 3/76 770 554 .108 1 94 u iA 9/76 i/77 770 554 100 91 et ' 30 II t/77 1/74-690 549 108 - 16 se 1 -t 3 690 549 . !!5 .380' { 4 690 549-. !Rt 169 s -5 690 .549 .177 33 0 6 660 549 137. 371 ~ 79 37 7 690 549 1 l 8 prettet 690 549-2^t 41R - g e For $ll 24

• • P'eddenn of total davs within tytin.) eM J 15 ntheated.

a 't 6

e i

9 c -Figure 7-11 Radial ~ Distribution of Axial Mixture Velocity Above ~ the Top TSP.in Incien Point Unit 2: Steam Generators-7-11' ~"'l .0159M:49/041388-38' ~- . <L --

^ ~ t a,cie I l 'I i 4 - 1 3 ,]: ..f'- / i .;l L 'l Figure 7-1 Plan View of ATH05 Model for Indian. Point Unit 2' i i 7-12 0159M:49/041388 '

k 4 I i '. ] 'l cj l l a,C,9 J . i 4 d f i k .i ) i t i 4 t, e l.. o-1 L Figure 7-2 Elevation View of ATHOS Model~for' Indian' Point Unit 2' L 7-13. -l l>

..y.;... '1 - g f;f ,4 .a,c,e ,1-1 J l ? l i 1 j Figure 7-3 Hot Leg Side Plan-View of ATH0S Model- .'i; i i 7-14 0159M:49/041388-30 l

F 4 .t - -i 1 1 1 1 e,c-J ? p M 4 + b ' I, i t Y Figure 7-4 Flow Pattern on Vertical Plane of: Symmetry.- 7-'15 i r

+ tj i s j 1 . Sj

c. :.

.) b. e,c

p..

l [' t 1 -h .E 4 ' t. 1 \\ i T f .g { .t Figure 7-5 Flow Pattern on ' Horizontal Plane (Z=22). at U-Bend: Region 7-16 .[ -0159M:49/041388-32 A 't

.__ _.._~ _ _ __. .,o f. '~ 1 i !p i 1 + .c [. (- e,c - P c k ? L 1 i l i i t -v f. i 1 1 h h j '+: l Figure 7-6. Flow Pattern on Top' of Tubesheet 1 7-17 i t. .~.

0159M:49/041388-33 y

v A

{p y,. '-i .1 l.. 4 ) I t .~. e,c - [ { t i ( l l ? 1 i i t 1 t i lp ~ b .4 Figure'7-7 Tube sap Velocity and Density Distributions for Tube" at Rowl0/Co13 7-18 0159M:49/041388-34

I L 8.0 PEAKING FACTOR EVALUATION t This section describes the overall peaking factor evaluation to define-the test based peaking factors for use in the tube fatigue evaleation. The evaluation { of the eddy current data to define the AVB configuration for North Anna Unit 1 Tube R9C51 is described. This configuration is critical to the tube fatigue assessments as the peaking factors for all other tubes are: utilized relative' to the R9C51 peaking factor. Uncertainties' associated with applying.theLair model' I L test'results to the tube fatigue assessments are also included in this-L section. Included in the uncertainty evaluation are tho'following:- o Extrapolation of air' test results to two phase steam water. o Cantilever tube simulation of U-bend tubes o Test measurements and repeatability o AVB insertion depth uncertainty 8.1 North Anna Unit 1 Configuration-s ~ 8.1.1 Background l. The AVB-configuration of the. ruptured tube in North Anna, R9C51, is the reference case'for the tube' fatigue evaluations for other plants..In; accordance with the NRC Bulletin 88-02,=the accept'bility of unsupported tubes a in steam generators at other plants -is based ~on. tube specific analysis relative to the N' orth Anna R9C51 Tube, including' the. relative flow peaking factors. Thus, the support conditions of the R9C51 tube are fundamental' to the analyses - of other tubes. Because of the. importance of the North Anna R9C51; Tube,' the support conditions of this tube,1which were originally: based on'"AVB Visible" i interpretations of the eddy. current test (ECT) data (Figure'8-1), were 1 reevaluated using the projection technique developed since the North Anna event. The projection technique is particularly valuable for establishing AVB, j positions when deposits on the tubes tend to mask AVB signals ~such.as found for E, the North Anna 1 tubes. The results of this. evaluation; are summarizedLbelow. N l 8-1 { - 0159M:49/051288-42 w v y -*-w-w.

1 8.1.2 Description of the Method R The basic' method utilized was the projection technique in which the AVB l position is determined based on measured AVB locations in larger row tubes in: the same column. In this study, the projection technique was utilized:in the-I " blind" mode, (AVBs called strictly based on the data) as.well as the reverse s mode (dataexaminedonthebasisofpredictedAVB; positions).- The'. objective of this application was, with the greatest confidence possible, to establish the-5 L positions of the AVBs in an 8 column range.around the R9C51 tube in North q Anna 1, steam Generator C. i 8.1.3 Data Interpretation { The ECT traces for the U-bends in Rows:8-12 (in one case, 13) were examined for Columns 48-55. The original AVB visible calls are shown in Figure 8-1. The S data were examined by an eddy current analyst experienced in~ reading thessi traces, and by an independent design engineer knowledgeable-in the geometry of f the Model 51 U-bend region. a ~ The intent of this review was to determine if the presence or absence' of' AVBs as shown in Figure 8-1 could be confirmed using the AVB projectiori t'echnique.- Preliminary projected AVB positions were based. on: geometric' data provided for ai few of the tubes near R9C51. The-features ~which were sought were evidence:of. O data " spikes" where AVBs were predicted, offset indications:(multiple spikes) where offset AVBs were predicted, single indications where_ single AVB-intersections were predicted, etc. Theldata evaluation method: used"was a-critical examination of the data, which was biased toward the presence of AVBs' L unless a confident call of "no AVB" could be made, and then checking' the, t consistency of the data among the tubes in a' column;and against the theoretical. y data for the predicted AVB positions. The data were scaled to determine the position of the signals which were considered AVBs. The scaling factor was determined based on the theoretical are length of the complete U-bend which was i - shown in,the data by clear signals at' the hot leg and cold' leg tube support; plate locations (TSP-HL and TSP-CL) in all cases. The assumption of this approach was that the velocity of the probe was constant around the U-bend,- 4 8-2 f-0159M:49/051288-43

.. ~. 'I i i I i l i l i I s i i } j 1 i ja.c u L Figure 8 4 is the "AVB visible" map for Columns 48 through 55, based on the critical review of the data. It is noted that the original data-l interpretations and the review interpretations are essentially the same, except' j for AVB position shifts in Columns 47-48 and 48-49. i I 8.1.4 Projections The arc lengths determined from the scaling of the ECT traces were utilized for projecting the position of the AVBs according to the standard format:of the projectionmethod, The results of the projections'are presented in Figure 8'5, which shows'a matrix of projections for Tube Rows 8 through-13 in Columns 48 through 55. For' ,i many of the_ tubes, more than one, and a many as three, projection values are - shown. Multiple projections are expected for a tube'if the AV8s on either t-g.3 0159M: 49/051288-44 i e - - _. _. ~,. -.

side cf the tube are not at the sama elevation, or if the upper and lower AVB { support that tube. As many as four different projections are possible if it is assumed that the tube is supported by the upper and lower AVB's, and both upper j and lower bars are staggered in elevation as shown in Figure 8-2. l'. The logic in arranging the projection data is based on the following two rules:. i Rule 1. The projections of the same AVB based on different tubes'in the same column must be consistent. If the tubes have multiple [ -f l t L i ?. i I i ja,C i. Rule 2. Two adjacent tubes in the same row ( l-la,c Consequently, the difference in the [- 34,c i The implementation of this is that if the position (either left or right) of a projected AVB is assumed for a column, then the projections in the adjacent columns are also ( Ja,c r B-4 0159M:49/051288 45 l

i 4 The arrangement of the AVBs as shown in Figure 8-5 best satisfies the rules i above and is consistent with the fatigue rupture of R9C51. The resulting AV8 arrangement based on the projection matrix of Figure 8 5, is shown in l Figure 8-6. l l 8.1.5 Conclusions i The general AVB arrangement. surrounding the ruptured tube in North Anna Unit 1, f Steam Generator C, which was the basis for the analysis, is confirmed by a detailed critical review of the-ECT data. Differences exist in the AV8 pattern 3 between Tube Columns 48 49, in which the AVBs appear to be less inserted than t previously indicated. The pattern of Figure 8 6 is the best fit to the rules l which were adopted for determining the position of the AVBs, as well as l consistent with explanation of the tube failure. The basis of the review was a projection technique which utilizes data from i tubes one or more rows removed from the actual inserted position of the AVB to l determine the position of the AVB. The intent of the review was to establish the positions of the AVBs by confirming or eliminating features of AVB I elignments such as side to side offsets, etc. of the AVBs adjacent to the tubes. Overall, the conclusions regarding the positions of the AVBs'around R9C51 in North Anna Unit 1 Steam Generator C, are based on consistency among all the available data. 8.2 Test Measurement Uncertainties j i The descriptions of the peaking factor tests and apparatus'were provided in ) Section 5.4. All practical measures were taken to reduce uncertainties. Nevertheless, some remain and should be properly accounted for.- The important parameter measured during testing that has a.significant impact on peaking-factor is the air velocity. The air velocity at test section inlet was measured using a ( la,c Based'on considerable experience 'l with the use of such instruments, it is known that the' magnitude of uncertainty j is very small. A [- la,c measurement uncertainty is used in this analysis l based on past' experience. j 85 l 0159M:49/051288-46 'I-

j 4 l 1 8.3 Test Repeatability During the peaking factor testing of AVB configuration, each. test was performed ] ~ at least two times to confirm repeatability. It has been demonstrated that the I tests are quite repeatable with the results often falling within 2 or 3% of one another for the repeat tests. An upper bound value of 5% was used in the { current uncertainty analysis. l 8.4 Cantilever vs U-Tube l A first order estimate of the validity of modeling a U bend tube by a cantilever tube in tests to determine the eftscts of AVB insertion depth on the l initiation of fluidelastic vibration has been made. The following assumptions were used: 1 8,C 2. t b 2. 3. [ 6 4. 5 6 9 'I ~ 0159M:49/041188 47 ,.,s-

_ - - -.. ~. 'i i h For the purposes of this estimate, the geometry of the cantilever measuring tube in the air test model is compared with the geometry of a prototypical Row 10 tube. [ i l ll 1 ja.c t The comparison between-a U bend tube and the model tube involves the consideration of an effective velocity associated with the flow perturbation { caused by the AVBs. It is assumed that the same [ l 't I i l t Ja,C r i 87 0159M:49/041188 48 i, a

L

i i

I Ja.c Using i [ these values, the ratio of the effective velocity for the cantilever measuring tube to that for the U bend tube is about [ Ja.c for the case treated; l A similar evaluation can be made for a Row 10 tube that lies in the { ( Ja.c or shadow of an AVB that is inserted to a depth required to support a Row 9 tube. TheAVB(, i ja c The not result is that the ratio of the effective velocity for the cantilever tube to that for the U bend tube is about [ Ja c i These results indicate that, for the particular assumptions used, the cantilever tube model appears to be a reasonable representation of the U-bend j with respect to determining relative peaking factors for different AVB configurations. This evaluation also shows-that, on the average,'the magnitude of the systematic uncertainty associated with the use of a cantilever tube to simulate the U bend is about [ ja.c 1 8.5 Air vs Steam-Water Mixture 1 L The local peaking factors from the air tests can be applied to the steam l generator steam / water conditions either as a direct factor on the mixture l I L velocity and thus a direct factor on a stability ratio, or as a far. tor on the steam velocity only with associated impacts on density, void fraction and damping. This method leads to a reduction in tube damping which enhances the peaking factor compared to the direct air test' value. For estimating an 7 absolute stability ratio, this application of the peaking factor is a best ] estimate approach. However, for the evaluation of tubes relative to stability ratio criteria, it is more conservative to minimize the peaking _ factor for the 8'" 0159M:49/041188-49

l l North Anna Unit 1 Tube R9C51 through direct application of the air test peaking j factor. This conservative approach is therefore used for evaluating tube j acceptability. i Under uniform AVB insertion (or aligned AVB insertion), there are no local open L* 4 l channels for flow to escape preferentially. Therefore, air flow is approximately the same as stean/ water flow relative to velocity perturbations. l l Under non uniform AVB insertion the steam / water flow may differ from air, as the steam and water may separate from each other when an obstruction, such as an AVB, appears downstream. The water would continue along the same channel while steam readily seeks a low resistance passage and thus turns into adjacent open channels. Two phase tests indicate a tendency for steam to preferentially follow the low pressure drop path compared to the water phase. ] Based on the above discussion, the F4 are considered to more appropriately apply to the steam phase. Thus, it follows that mixture mass velocity for the .] tube subject to flow perturbation c:n be written as follows: l l e,c j l 1 I where p is the vapor density, pf the water density,. T the g a velocity peaking factor determined from air tests, j

• the nominal g

superficial vapor velocity, and jf* the superficial water velocity. Steam quality can then be determined as follows: t,C l t l The [ Ja,c as used-in the ATHOS code, is applied to determine void, fraction. Subsequently, mixture l density, velocity and damping coefficients-for the tube which is not supported l and subject to flow perturbation is evaluated. Therefore, similar to_ the air-i velocity peaking factor, local scaling fac' tors of mixture density and velocity I 89 i 0159M:49/041388-50

,j i i and damping coefficient can be readily determined. Finally, a local stability peaking factor for fluidelastic vibration can be calculated as follows: e,c 1 l 1 d the density scaling factor,- f whereFhsthestabilitypeakingfactor,F F, the velocity scaling factor, and Fdp the damping coefficient scaling factor. If we use the air velocity peaking factor without translating to j 4 steaVwaterconditions,then e,c P l As shown in Table 8-1 stability peaking factors for the steam / water mixture are - slightly higher than air velocity peaking factors. The difference between the stea# water and air peaking factors increases as the air peaking factor increases. For application to tube fatigue evaluations,- the ratio of the peaking factor for a specific tube to that for North Anna R9C51_is the quantity of interest. l Larger values for this ratio are conservative for the tube fatigue assessment. .l The North Anna R9C51 peaking factor is one of the highest peaking factors. As_ discussed in Section 8.7, a peaking fac' or. of nearly [ la',cisdetermined f t for the R9C51 tube. The differences between [ _ l Jac Typical ' values are shown in Table 8 2. These results show that the direct application of the air test data yields the higher relative peaking factor compared to R9C51. To obtain conservatism in the peaking factor evaluation,-'the [ jac l Comparing the values in the first and last columns of Table 8-1, it may be l noted that the stability peaking factor for: steam water is-['

]a,c higher than the air velocity peaking factor.

On the average, the uncertainty associated with the conservative use of air velocity peaking factor is [ Ja,c j b 8 10 0159M:49/041188 51

1 l The conclusion that peaking factor for steam water flow would ba higher due to j h the dependency of damping ratio on void fraction was supported by an alternate j study..In this study, a section of steam generator tubes were simulated using l the ATH0S code under protypic flow conditions. The objective of this study was j to examine the magnitude of the changes in void fraction and thus stability { ratio as a consequence of non unifore AVB insertion patterns. The current j version of ATH05 has modeling limitations that prevent accurate modeling of l' local geometry effects. In addition, it is believed that an analysis using two fluid modeling procedure is mandatory to a calculation of the peaking factors for a steam generator to account for the preferential steam flow-along, f the low resistance path. Consequently, the intent of this analysis is only to help bound the uncertainty on void fraction, effects from extrapolating the air tests to steam water. i First the analysis was conducted with uniformly inserted AV3's in the ATHOS model. The ATHOS results were processed by the FLOVIB code to determine j stability ratios for the specific tubes of interest. The calculation was repeated using a non uniform AVB insertion pattern'in the model. The results show that the void fraction distribution changes as a result of flow perturbation. Further, the impact on stability ratio resulting'from the changes in void fraction profiles was about [ Ja,c This alternate calculation provices independent corroboration of the prior discussion regarding the stability peaking factors under steam-water conditions versus in air. 8.6 AVB Insertion Depth Uncertainty l The most significant uncertainty for the low peaking configurations is not in the test results, but in the determinat16n of actual AVB insertion patterns t adjacent to specific tubes. The methodology.used for obtaining the AVB insertion patterns from eddy current data can ascertain the AVB location only-to within approximately [ ja,c - The effect on peaking factor resulting from this uncertainty is addressed using 1 test results of AVB configurations that varied from one another by up to-[ Ja.c [ 0159M:49/041888-52 8-11 w -+e r---,e- ,w--'v-y+-+, 4 -,c-- ,w a--**g-- -+ 1--w--rw w*r>' N .e-s--v-- +--reeM+=* m-m-- yew--* -w

t i l l Based on maps of AVB insertion depth of various plants, several configurations J have been tested for determining fluidelastic instability flow rate by an air f f., cantilever model. Stability peaking factors were then determined from the l ratio of critical flow rate for a uniform AV8 insertion configuration to a l l. specific configuration. Figure 8 7 summarizes the AVB configurations tested. Position of AVB insertion depth is determined from Eddy Current Test (ECT) data. Positioning of AVB from ECT data reading is subject to uncertainty; its i accuracyisprobablyabout[ Ja,c pitch. A change of an AV8 [ insertion depth in a given configuration leads to a~different configuration, and thus a different peaking factor. A review'of the tested AVB type has been made and results summarized in Table 8-3. As can be'seen, a decrease in depth t of an appropriate AVB tends to decrease the peaking factor, for instance, a ] [. Ja.c Such a trend can be explained; a decrease in a specific AVB depth-l will open up more channels for incoming fluid to distribute and thus'less flow perturbation. However, this applies only to those changes without inducing the j reinforcement of flow perturbation from upstream to-downstream. l l On the average, the uncertainty in peaking factor resulting from small l' variations in AVB insertion (of the order of 1/2 tube pitch) is found to be i [ Ja.c 8.7 Overall Peaking Factor with Uncertainty [ i As discussed in the previous subsections, there are several-aspects to be l considered in applying the laboratory test data to steam generator conditions. These considerations were reviewed one at a time in those subsections. This section will integrate the pieces into one set of stability peaking factors. Looking forward to how these peaking factors are used in the analysis (Section 9), the relative stability ratio calculated for a given tube without-the consideration of flow peaking is corrected using the ratio of the peaking factor of the specific tube to that of the No;th Anna RgC51 Tube (Configuration la). It is to be noted that, of all the configurations tested, Configuration { lb produced the highest peaking factor, followed very closely by 4c, la and-8*II 0159M:49/041888 53 ___.________________.m,-__m ~~...--n, i.-.-.,...,-,,. ,,_,.,m...,_ _ _ _,, - -, _. ,,4

.i i i 5e. This is encouraging in the sense that it tends to explain why, of all the tubes in service, the R9C51 tube was the one to experience the fatigue rupture. 4 It is to be noted that the test results would be applied as ratios of a specific tube peaking factor to the R9C51 peaking factor. This will reduce the-influence of some uncertainties since the systematic ancertainties would affect [ both the numerator and the denominator in the ratio of peaking factors. The { major difference will be in those configurations whose peaking factors are significantly lower than that of R9C51. The approach employed here is intended to provide that conservative peaking factors are employed for such apparently f low peaking configurations, The uniform AVB Configuration 2a is selected as a reference configuration, and' i the peaking factors of all configurations tested are recomputed on the basis of this reference. As discussed below, some of the test uncertainties are-applied to the referenr.e case to account for its significantly low peaking relative to l the R9C51 configuration, j The uncertainties in the test results and their extrapolation-are those due to ^ test measurements, test repeatability, cantilever thbes in the test vs U-tubes' in the steam generator, and air tests vs steam water mixture. These were discussed in more detail in the previcus subsections. The magnitude of these uncertainties are listed in Table 8 4. l I Of these uncertainties, those due to measurement and repeatability of tests are [ rardom errors and can occur in any test. Therefore, these are treated l together. The total random uncertainties are calculated'by ( Ja,c The R$$ value of these is [ Ja.c Since these can occur in any test, these are to be applied to all tests. One way of doing this is to apply it to the R9C51 value, that being in' [ the denominator of the final peaking factor' ratio. Thus the peaking factor for l configuration la (R9C51) is reduced by this amount to yield a value of ( Jacinsteadofthe(

• )a,c appearing in Table 5 2.

[ l-f t 8'I3 0159M:49/041888 54

s i The next three uncertainties in Table 8 3 are systematic uncertainties. It could be argued that these appear in the peaking factors of both the specific tube under consideration and the R9C51 tube and are therefore counter balanced. However, the relative magnitude of these may be different. particularly for configurations with much lower peaking than R9C51. Therefore [- it was judged that the [ ]a,c similarly, as noted above, the iffect on peaking factor due to the uncertainty in the field AVB configuration is also included in this l reference case. Thus,the[ j Ja,c The peaking factor of I the Reference Configuration ta (Table 8-5) is raised by his amount to a value - of [ Ja,c The change ir peaking factors of Configurations la and ta resulting from the application of uncertainties as described above are shown in Column 3 of Table i 8 4. The peaking factors of all configurations are recomputed on the basis of Reference Configuration 2a. These val'ues are displayed in Column 4 o'f Table [ 8 5. L-c Some of the uncertainties were applied to Reference Configuration 2a in order to apply them to all low peaking configurations conservatively. Thus, no configuraticn should have a lower peaking factor than this reference i configuration. Therefore,whenapeakingfactorvaluelessthan[ Ja.cis calculated for any configuration, (in Column 4 of Table 8-5), it should be altered to [ ja.c. Further, for some of the configurations that are-conceptually similar, the more limiting (higher) value is used. For example, a peaking factor of ( Ja c is used for Configurations 5a and 5b based on t their similarity to Configuration 5c. The final stability ratio peaking factors calculated on this basis (with. [ Configuration ta as the reference) are shown in Table 8 6. It may be noted that the peaking factors vary in the range [ ]a,c the R9C51 peaking factor being [ ]a,bc Figure 8 7 shows the' final peaking factors with the pictorial representation of the AVB insertion patterns. ) 0159M:49/041888-55 8-14 1 .. ~ Y

~ _ - - Table 8 7 shows the result of applying the peaking factors to specific tubes in the Indian Point 2 steam generators. I The overall conclusions from the peaking factor assessment are: 1 3. As noted in Table 8-4, five elements have been included in the uncertainty l evaluation for the peaking factors. The uncertainty estimates were developed from both test and analysis results as described in Sections 8.2 to 8.6. The largest single uncerte.inty of ( Ja.c is attributable to l uncertainties of up to [ Ja.c on determination of AVB insertion depths from field oddy current data. This relatively large -l uncertainty is applicable only to low peaking' conditions where the AVB f uncertainties can contribute to small peaking factors. The definition.of "no flow peaking" was increased to encompass the small peaking effects from AVB insertion uncertainties. For the AVB patterns leading.to significant; e peaking factors, AVB's were positioned within uncertainties to maximize the j l peaking factor. For these configurations, variations of AVB insertion within these uncertair. ties are expected to reduce the peaking factor i compared to the final values of Table 8 6 and Figure 8 7. 2. Including uncertainties directed toward conservatively decreasing the. { ~ peaking factor for the North Anna Tube R9C51, the final R9C51 peaking factoris[ Ja,b,c relative to a no flow peaking condition such as with uniform AVB insertion depths.- + 3. The final peaking factors include peaking effects greater than the R9C51 tube (such as Configuration 4c) although this is believed to be a consequence of the conservative uncertainty analysis and. is not likely to t be representative of actual peaking effects.. l L e { 0159M:49/041888-56 . 3 15 i .,_m v. ...o ,m.., ...em

L l Table 8 1 t Stability Peaking Factor Due to Local Velocity Perturbation l Scaling Factors for StensVWater j l l Air r Velocity Void Stability l Peaking Fraction Density Velocity Damping Peaking

Factor, Scaling,

Scaling, Scaling,

Scaling, Factor,

[ F F, Fd y dp F, F - F a .i e,c i i NOTE: 1. Stability peaking factor for steam / water mixture is calculated as follows: e,c 2. Damping scaling factor is calculated _ using modal: effective void fraction of [ Ja.c for RgC51 tube. 5 e t 0159M:49/041888-57 8 16 b ,.y,. ,y ,..vL, ~..E.--

1 Table 8 2 Comparison of Air and Steam Water Peaking Factor Ratios I j Air Air Steam Steam j Peaking Peaking Peaking Peaking Factor Ratio Factor Ratio l ) e,c 1 i l t i 1 -i V t 0159M:49/041888 58 s.17 1 . ~ _. -. - .--.v. r + - - ~. - ~ -. _, -.

i l .? Table 8 3 j l Effect of Local Variation of AVB' Insertion ) i l l, A to B AVB Peaking Peaking Ratio Tvon A Tyne R Variation Ratio A Ratio R (B/A) e,c t l lb la -l 1 4a 5c t 5c la 1 ~ la $b t 5c 4a la 5c P r ~ Peaking Ratios Uia/Un are from Column 2 of Table 8-5. q l i i i l t i .i -is 0159M:49/032988-59 i i

O Table 8 4 Uncertainties in Test Data and Extrapolation Source of Uncertainty Iggg Maanitude. % ~ 1. Velocity measurement Random 2. Test repeatability Random 3. Cantilever vs U tube Systematic 4. Air vs steam water mixture Systematic 5. Field AVB configuration This.is not an uncertainty associated with the test data.. It results from the inaccuracy in determining the true AVB position in the field using oddy current data. ~' 0159M:49/032988 60 O

l l -l j Table 8 5 Extrapolation of Test Results to steam Generator Conditions I i Peaking Factor Test = Data with Referenced to Confinuration gg,tA Uncertainties confia. 2a [ e,b,c la lb 2a 3 i 4a 4b 4e l Sa i l' { l 5b l 5c 5e 6a t 6b 6c .? G l* l t 8-2o 0159M:49/032388 61 F

h i i Table 8 6 l FINAL PEAKING FACTORS j l. Confinuration Peaking Factor ~ r - 9 a,b.c -la 1 lb j t ta e i 3 t i 4a i (b 4c i L' la l 5b-5c 5e 6a 6b 6c 1 1 \\ l \\ 1 i i o 1 8-21 0159M:49/032388,- }

l 'l Table 8 7 1 l l i Stability Peaking Factors of specific Tubes j = I Indian Point Unit 2 steam Generator Row No. Column No. Peaking Factor i ........g.... p. 21 10 16 .i f 10 17 10 20 i 10 21 .:i 10 62 .i r 10 69 10 70 REST OF THE TUBES i l 22 10 24 lo 25 ] REST OF THE TUBES j .i 23 10 81 f 10 82 REST OF THE TUBES i 24 8 65 9 65 10 71 Il 45 11 46 REST OF THE TUBES 1 P i .i 0159M:49/041188-63 8-II-i ~

1 n o L;. ) L O 00 00000 O d@O@@OOOOO-r SGI 85 84 83 82 .81 80 49 ' 44 47 46 '45 44 .t OAVB O'AV8

• & FA! LED TUBE-L visista mvisista e,tuaaso i

[ l i 1 a .i ? Figure sel original North Anna AYB configuration i l f ~ 8-23 I IFP81.030788- .-.-..-;= .:s -...-.... J ~. -. -.

i 1 o 1. 1 l .i - l O 2 !I 8-3 I t h I I 4 [ I, i c 4 i i ? + l', E k i'r h Figure 6-2 ~5chematic of Stagpred WBs O ..n IPP81.030788 i l -l l;. .__--_.-_.u-,-- 1

1 I I l s e a. e. amme,. -, s em si m e [ I .M .l so -=== 4 so - e i esano. m ami j y ,e,m,,,,,,e,,, .i> .m _.. se - m m eene - e e m gyggp i me. emar.eer 1 = '.___ X -...- =::. J a non esa se e as ene i t i _ o s e.. e e ase. emme. e - r s om u m = [ eM l en I.= so - numm - o e d sa %ses A sum - ese em r se - e 5 m M= M e l 1> so - e eens. e am; I ? M..

m..

I I me emme.e e

er l

l i 9.' i I si w . m

== em.sr Figure 3 3 Avg

• Pair" in ECT Trace

= ..p i. 8a25 IPP81.039788

L 1o = OOO 52 L O' 000 in u 1. ooo 1 l, DOD OOO l O@ 0000000000 i f Cobmn 56 55 54 53 52 51 50 49 48 47 46 45 44 e - -- Tae e ~ ~e Nureers in circles in column range 48 55 represort readable AVB Intersection signals. I on artical review of the ECT traces Open ownie in this range means tw das is i l t i Figure 8 4 North Anna 1, Steam Generator C. AVB Positions. [ l critical Review ?AV8 Visible" calls I* s-as IPP81.030788 .t

i l i f ~ Q13-- I w. v r Ota - e.w t f M11 - - p. R1o - as .et e.te 'k l a* b L me - e. ma -- a' x C58 C84 C83 CSS C81 C80 Cet' C44 x.w assamam. P Mugges Twee

• O M A8

v. ui e

N* _.j "High Side' Pf9jeSipne ( l* Figure 8 5 North Anna 1. Steam Generator C. ~ R9C51 AVB Projection Matrix l + 8 17 IPP81.030788 _ _ _ _ _ _ _ _ _ _ _. _ _. _ _, _... -.. ~... -. - - - -.. .--4.....--v-- y - -., -... ....~....-.y..

I i. L l 1 l. a O O O O L O O O O u O O.O.O. 09 O'O O O OOOdOOOOO0: $6 35 54 $3 $2 st '- 30 49 48 47 46 45 44' [ Figure 8 6, North Anna R9C51 AVB Final Projected Positions' 8 IPP81,030788 ,-..-.,,--,-.~ms- ..=+ev<r- +==c ,e -r .e

i i ii. i type or Ave PEAKING type or Avs PEAKING i msERTON FACTOR NsERTON FACTOR e,t.c e,b.c + .l 1a 5a i 1b 5b 2a 5c i 3 5e p da 6a p. i I + 4b 6b 4e 6c ~ I l p* Figure 8 7 FINAL PEAKING FACTOR FOR INDIAN POINT 2 L. '? ? i S-29 1. ll. IPP81.030788 t l' h

l j 1 'l 9.0 STRUCTURAL AND TUBE VIBRATION ASSESSMENTS 9.1 Tube Mean Stress This section suunarizes an analysis to determine stresses in a tube which is j tight, but undented. The analysis assumes the tube to be [ l Ja,c at to W shutdown. Loads imposed on the l tube correspond to steady state pressure, differential thermal expansion i between the tube and the support plate, and a thre-wall thermal gradient. l I A summary of the temperature and pnssure parameters at 100% power in the vicinity of the top support plate are orovided-in Table 91. The tubs temperature corresponds to the average of the primary side water temperature and the plate temperature. The resulting tube / plate radial interference is i l [ Ja.c j The analysis is perforined using the finite element model shown in Figure 91. l* The model prescribes ( I l 1 l i ja.c j Two reference cases were run using the finite element model, the first fcr a primary to secondary side pressure gradient, and the second for a [- j Ja,c radia1' interference between the tube and plate. The pressure case-incorporates the axial load on the tube by applying a pressure loading along l{ the top face of the model. Plots showing the distribution of stress for the l tube outer surface for the two reference cases-are provided in Figures 9 2 and 9-3. Tube stresses due.to the thru-wall thermal gradient are calculated to be-l 7.2 ksi using conventional. analysis tech'niques. A plot showing the combined l stress distribution along the tube length, incorporating appropriate scale ( factors for the Indian Point Unit 2 operating conditions, is provided in Figure ~ 9 4. The maximum axial tensile stress is 25.3 ksi.and occurs approximately l O.134 inches above the top surface of the support plate. Adding,-for j 0159M:49/041188 64 9-1 u._-.., --a -.,,A-- .. +, ,-.,,,,.,_,n;, w.

l l ? conservatism, the surface' stress due to pressure, 0.77 ksi, gives an applied L sean stress of 26.1 ksi. In addition to' the_-applied stress, resifual stresses exist in the tube as a result of the manufacturing process.. For mill annealed L tubes with subsequent straightening and polishing certain residual stresses l' exist in the tube. The stresses are compressive at the tube surface, but 5-10 f mils below the surface, the stress levels change to 10-15 ksi tensile, Reference (9-1). Combining the applied;and residual stresses results in a i-cumulative mean stress of 41.1 ksi. Due to the presence of denting at the top: max

• 8, was.used support plate, the maximum mean stress, base of a 4

in determining stability ratios and fatigue usage. 9.2 Stability Ratio Distribution Based Upon ATH0S i An assessment of the potential for tubes to experience fluidelt.stic instability-in the U-bend region was performed for each of the tubes in Rows 8 through 12., This was performed using FASTVIB, a Westinghouse. proprietary finite element based computer code. This code was written to predict the. individual responses of an entire row of steam generator tubing exposed to a tube, location' dependent l fluid velocity and density profile. The program calculates. tubo natural j frequencies and mode shapes using a linear. finite element mMel 'of the> tube. The fluidelastic stability-ratio Ue/Uc (the' ratio of the effective' velocity to the critical velocity) and the vibration amplitudes caused by turbulence, are calcuhtd for a given velocity / density profile and tube' support condition.. The vei c N y and density distributions are determined using th'e: ATHOS computer 1 code, as described in Section 7.3..Also input,to the code are the WECAN-generated mass and stiffness matrices used to represent the, tube.- (Wecan;is also a Westinghouse proprietary computer code.) Additional input to FASTVI8 l consists of tube support conditions, fluidelastic stability constants, and LI turbulence constants. l 1 h + This process was performed for the Indian Point Unit 2 steam generator, tubes n and also for the North Anna Unit-1 Tube R9C51 using similarly appropriate ATHOS : models. Ratios of the Indian Point' Unit 2'results:to those for North Anna-Unit 1 Tube R9C51 were generated to produce a quantity that could be used to provide.an initial assessment of the Indian Point Unit 2 tubes' relative to the' j ~ ruptured tube at North Anna Unit-1. Figure 9-5 was generated using.the 5 following conditions for both Indian Point Unit 2 and North Anna Unit l1:- 0159M:49/041888-65 9-2 m - mm w w --w---e-o,- y

• w--#

w -w r, ,._c

.] i- +

1) Tube is fixed at top tube support' plate, q

q _2) Void fraction dependent damping used.- 4 ~ i 3)NoAV8supportsactive. A horizontal line is drawn at the relative stability ratio value of 0.90. This-identifies the point where a ten percent reduction in stability ratio exists l relative. to North Anna Unit 1 Tube R9C51.. (See Section 4.1 for a discussion of-q the stability ratio reduction criteria.)'-All,the tubes with ratios above thisE j line would be considered to have stability ratios larger then ninety' percent of-1 North Anna Unit 1 Tube R9C51.? This figure-indicates that all tubes in: Rows 8 1 and 9 can be considered' acceptable with some' tubes.being unacceptable ini-J Row 10. Essentially all inner tubes in Rows.11 and'12 lay' above this line. hL Tubes above this line,-that are not supported'by AVBs,-require further. I L evaluation to determine the acceptability of the tubes. Section'9.3 contains:. [ the results of the further evaluation for these tubes. 9.3 Stress Ratio Distribution With' Flow Peaking p - An. evaluation was performed to determine' the ratio of the Indian: Point' Unit 2. l tube stress over the North Anna Unit I Tube.R9C51 stress'. This ratio is determined using relative stability ratios discussed in the previous.section, flow peaking factors '(Table 8-7) and bending moment factors.- Section 4~.2 and 4.3 contain additional' information and describe tho' calculational procedure ' used to obtain the results presented'in;t' is section...The results. contained in h this section are based upon the followingl conditions:~

1) Tube is fixed in top-tube support plate.-

t t

2) Damping is void' fraction dependent.

o - f

3) AVBs do not provide support.

1j 0159M:49/041888-66 9-3 j .n

t

4) 105 criteria with frequency effects used j

-s

5) The tubes are assumed to be dented or undented.

.{ i

6) Flow peaking factors are usad.

A tube can be considered acceptable if the stress ratto is_less than 1.0 when .} calculated-using the procedure # scribed in Section 4.2 and _4.3 andincluding, 3 the conditions listed above. eng this1 criteria indicates'that the stress ] acting on a given tube will not produce a fatigue event in's manner similar:to the rupture that occurred in the R9C51: tube at North Anna Unit:1. Figure 9 6-- ] contains the stress ratio results for each of tho' Indian Point. Unit 2 tubes in - Rows 8 through 12. As can be observeu in the figure,> all of the tubss in' Rows s 8, 9,10, and 11 fall-below the acceptance line indicating that the tubis 'are acceptable with respect to U-bend fatigue, All of the tubes in Row 12,.except i L for Columns 3, 4, 5, and 6 (and the symmetrical Columns, 90, 89, 88 and'87), j also fall below the acceptance line. - The fi0ure indicates that perturbations exist in the stress ratio curves that were not(evident in the relative l stability ratio plots discussed in.the previous.'secticn (see-Figure.9' 5). These perturbations are due to,the flow. peaking effects dischsed previously. I 9.4 Cumulative Fatigue Usage h i All tubes that are unsupported and have a stresk ratio 1:1.0 have a maximum stress amplitude that is < 4.0 ksi (from 9.5 ksi) since a 10% reduction in Lthe stability ratio for the North Anna Unit 1 Tube R9C51 was the criteria basis. 3 The stability ratios for the-Indian Point Unit'2 tubing are based on the current operating parameters and with future operation on the same: basis, the tubes will'not rupture as a result o'f: fatigue if 1) they meet the stress: ratio criteria of s 1.0 and 2) their current and-future fatiguejusage will total'less - than 1.0. Table _9-2 contains a. summary of the ' combined relative. stability-ratios and the stress' ratios for the most critical, unsupported tubes in:each-of the steam generaters. : All tubes have a stress ratio less than' 1.0 with the j exception of Tubes R12C89 and R12C90::in Steam Generator 21. - These two tubes have' been plugged.as a preventNe measure;to eliminate them from further 2 consideration. Sentinel, or tell-tale, plugs.were used to permit:detaction of tube degradation should it occur in the future. y o 0159M:49/041888-67 g,4 1q I .o

~ y c?! Oj K \\ Acceptability of the Indian Point Unit.2 tubing for fatigue is accomplished by - .j demonstrating the acceptability of the tube in Table 9 2 with the highest stress ratio, 0.94 at Row 10 Column 71 of Steam Generator 24. The maximum stress with the current-operating' conditions is e.c m 1 Based on-the relative stability ratio over the operating history presented in 1 Section 7.4, the alternating stress for'each operating cycle can be determined-knowing that the stress for cycle x it, e,c Using s = 6 conservatively-(see Ser. tion 4.1) establishes the maximun: alternating stress for each~ operating cycle. The-number of cycles of vibration, a l-is obtained for each fuel cycle by multiplying the number of days times the l number of cycles per day at the frequency of the Row 10 Tube, [60]a,c hertz. l Table 9-3 sumarizes the time. history. Conservatively assuming that all fuel cycles' have been at 3.76 ksi, the cumulative fatigue usage to date:is 0.161 and the cumulative fatigue. usage.for j the. operating license period (with all future operational years at the. same operating conditions as Cycle 13) would be 0.70. T Alll of the unplugged _ Indian-Point Unit 2 tubes, therefore, meet-the fatigue usage. requirement of 1.0 with denting assumed to exist from initial startup. l t 1 4 .) 1. E! k 0159M:49/041888-68 9-5 Y

___...__._____.._..._.7. ~ -3 Table: 9-1 -j l: 100% Power Operating Parameters- "] Indian Point-Unit 2. s ll.. Primary Pressure = 2250 psi: Secondary Pressure = 768 psi Pressure Gradient = 1482 psi. i Primary Side Temperaturo = 584'F, Secondary Side Temperature = 514*F. h Tube Temperature - 549'F s t g .t o .( -[ -e -t l.. 5 s f t i ,n 9-6 l' 0159Mi49/041188-69 Ao u ..+

O I y TABLE 9-2" INDIAN POINT UNIT 2 EVALUATION'0F THE. 7 MORE SALIENT UNSUPPORTEC U-BENDS STEAM REIATIVE STRESS GENERATOR TUBE STASILITY RATIO (1) RATIO (1) 21 R10C3 0.64 .0.13 R10C16 0.64-0.13 R10C24 0.64 .0.13 i R10C62 0.67-O.17 R11C3-0.77_ 0.33L R11C4.' O.76 0.30 P R11C86. 0.71- ~0.20 [ R11C87 -0.73' O.24' R11C88 0.74. 0.27 R11C89 0.76 0.30 R12C2 0.73

0. 2 :,

'f R12C89 1.01. >1.01,2) 'R12C90 1.05 >1.0(2) s 22 R10C24 0.64-0.13 R11C3 0.77-O.33-R11C37 0.'67-0.15 i ~R12C2 0.73 0.21 l 1t 23 R10C3 0.64. 0.13: 1 R10C20 0.64-0.13-R11C2 0.53

0.04:

R11C3 0.77 0'33 R11C4 0.76 0.30- -R11C47 0.79-0.40-24 R10C71 O.89-0. 9 4 -- R11C3~ -0.77 0.33-R11C46 0.79' O.40. R11C51. 0.68 0.16 R12C2 - 0.73' -0.21 [ i 5 (1) All ratios are'in comparison to R9C51~, North: Anna 1, Steam: Generator C.. '(2 ) This tube has been plugged as a preventive measure.-- 9-7 ...h. _._~~b,'. . =,.

f; r 4 ' ;i Table 9-3. -) Duty Cycle Description for Indian Point. Units. j i Normalized. l Fuel Stability Alternating. Cycles : 1, Cycle Ratio Stress-Days' - At-[L']a,cHZ. n M s,c :- r 2B 8 1.0 1.0 2279 l IB 2A-0.916 0.591 175 IA 0.816 0.295' 234 [ .) TOTAL CYCLES. 4 A i h.; 3 s n: 1 t l .{ {,. i f i 0159M:49/041188-70. 9-8 ., f t l

F l c.. ~; e i ,'t A + l: e,c t 1 .e ' ' ' _y ,I i ? a g .h i l E + h ..i I f' i e r I I k , -. ~ $ l' -i

i

-.. i i 't l t l 1 l' 1-i } i 'a b 1 L i s. f P t Figure 9-1 'Axisymmetric Tube Finite Element:Model f i - - f. i 9-9 .i 1 k ? \\

/

l i

1 .J i e,t j A 8 l i d 4 e t 1, t i i. 4 j i

o i

.( I ~ r l l 1 Li ' j - ]w 4 La-l.J ' Figure 9-2 Tight Tube Stress; Distributions l Pressure Load on Tubei .i p (_ i ~ 9310' h, 0159M':49/,041188-72 u' = - t'

t e,c A '. s -1 I I '!- ) 1 Figure 9-3' Tight' Tube Stress Distributions-Interference Load on Tube. j 0159M:49/041188-73' 8-11 -}}