ML20148J631
| ML20148J631 | |
| Person / Time | |
|---|---|
| Site: | North Anna  |
| Issue date: | 03/16/1988 |
| From: | WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP. |
| To: | |
| Shared Package | |
| ML20148J628 | List: |
| References | |
| SG-88-03-016, SG-88-3-16, STD-7.2.2.1-807, STD-7.2.2.1-8079, NUDOCS 8803300330 | |
| Download: ML20148J631 (23) | |
Text
{{#Wiki_filter:s' .i-STD-7.2.2.1-8079 SG 88-03-016 NORTH AhWA 2 STEAM GENERATOR
'ITJBE FATIGUE EVALUATION AND REMEDIAL ACTIONS REPORT MARCH 16,1988 PREPARED BY:
WESTINGHOUSE ELECTRIC CORPORATION SERVICE TECHNOLOGY DIVISION P. O. BOX 3377 PITTSBURGH, PA.15230
~
8803300330 880324 PDR ADOCK 05000333 Q DCD
TABLE OF CONTENTS Page Number List of Figures 2
1.0 INTRODUCTION
AND
SUMMARY
1.1 Introduction 3 1.2 Summary 3 2.0 EVALUATION CRITERIA 2.1 Tube Fatigue Mechanism 3 l 2.2 Prerequisite Conditions 4 l 3.0 EDDY CURRENT REVIEW 3.1 Tube Denting 4 3.2 Identification of AVBs from Eddy Current Testing 5 i 3.3 0,1 or 2 AVBs Present Eddy Current Calls 7 l 4.0 UNIT 2 VIBRATION AN ALYSIS 4.1 Similitude with North Anna Unit #1 7 4.2 Modeling at Tube Bundle Periphery 16 5.0 AVB MAPPINO AND SUPPORTEDTUBES 5.1 Consistency Checks in AVB Insertion Mapping 17 5.2 Criteria for Considering Tubes as Supported 17 6.0 CORRECITVE ACTIONS 6.1 Preventive Plugging 18 6.2 DowncomerModification 18
7.0 CONCLUSION
S 20
8.0 REFERENCES
22 1
, , .r- r,m ,- -e v- r , - ,- - - -,-,- - . - - - -su.----py - r--r r---,----,, -- -- -
9 s 4 LIST OF FIGURES Figure 3-1 Typical Eddy Current Signal of an AVB Figures 3-2 to 3-4 "Visible" and "Invisible" AVBs of SG A, -B, and -C Figures 3-5 to 3-7 AVB Insertion Maps for VGB SG A, B, and -C Figure 41 Tube Stability Ratios Figure 5-1 Detailed View of VGB SG B R9C32 Figure 6-1 Typical Downcomer Flow Resistance Plate l I l t 2
1.0 INTRODUCTION
AND
SUMMARY
1.1 INTRODUCTION
North Anna Unit #1 and Unit #2 are each three loop plants with Westinghouse Series 51 steam generators and essentially identical design and operating conditions. A tube rupture event occurred at North Anna Unit #1 on July 15,1987. The tube was located in Row 9 Column 51 in steam generator "C', and the rupture location was at the top support plate on the cold leg side of the tube. Extensive analytical and test investigations were performed to determine the cause of the tube rupture in Unit #1 to be high cycle fatigue (Reference 1). 1.2
SUMMARY
A series of tube fatigue evaluations performed for North Anna Unit #2 are described. The ( prerequisite conditions for tube high cycle fatigue mechanism, and U bend tube vibration analytical results are described. Eddy current techniques used to characterize AVB positions, and AVB location maps specific to North Anna Unit #2 are provided. A description is provided of the implemented preventive measures, which include preventive plugging and the installation of a downcomer Gow resistance plate. As a result of the modifications, the potential for large leakage l occurrences due to tube fatigue is reduced to negligible levels, based on multiple criteria which significantly reduce the probability of occurrence of tube fatigue. This report documents the tube ( fatigue evaluation conducted for North Anna Unit #2 in October,1987. The evaluation was conducted using essentially the same criteria and methods as the North Anna Unit #1 evaluation documented in Reference 1. 2.0 EVALUATION CRrmRIA l l 2.1 Tube Fatigue Mechanism As previously stated, the North Anna Unit #1 tube rupture was determined to be due to high cycle fatigue. The source of the loads is a combination of a mean stress level in the tube and a superimposed alternating stress. De mean stress was produced by denting at the top tube support plate and the alternating stress was due to out of plane denection of the tube above the top tube l support caused by flow-induced vibration. The loads were sufficient to produce fatigue and are consistent with a lower bound fatigue curve for the Alloy 600 tube material. He magnitude of the 3 l
alternating stress was consistent with a fluidelastic tube vibration mechanism. The reduction in damping at the tube to tube support plate interface caused by denting was a significant contributor to excessive vibration. De absence of , itivibration bar (AVB) support was also required for vibration to occur. He pn:sence of AVB support limits tube motion and therefore the deflection amplitude required for fatigue. Also contributing to the level of vibration, and thus the loading, was the local flow field associated with the detailed geometry of AVB insertion depths in the steam generator. The ruptured tube was considered to have a worst case combination of loadings and fatigue properties. 2.2 Prerequisite Conditions Summarizing the information described above, and described in detail in Reference 1, the following are the prerequisite conditions for the tube fatigue mechanism: 1 ! Fatigue Requirements Prerequisite Conditions l 1 Mean Stress . Denting l Altemating Stress . Tube Vibration
- Denting at top TSP
- High local fluid forces
- Absence of AVB support l
Material Fatigue Propenies . IAwer range of properties 3.0 EDDY CURRENT REVIEW 3.1 Tube Denting As noted, tube denting was determined to be a prerequisite for the North Anna #1 type tube fatigue mechanism, and therefore of interest in the North Anna #2 tube evaluation. Tube support plate crevice corrosion products and tube dent sizes as small as 1-2 mils (0.001" to 0.002") are detectable by eddy current testing. Although the difference between tube denting and the presence of crevice corrosion products is significant in terms of the tube fatigue mechanism, it was judged 4
conservative to consider a tube to be dented if either tube denting or top tube support plate crevice corrosion products were detected. Based upon the preliminary eddy current informaoun available at the time, all North Anna Unit #2 tubes were assumed to be dented at the top tube support plate in the October,1987, tube preventive plugging evaluation. Subsequent eddy current evaluations of top tube support plate corrosion conftrmed that most of the tubes had crevice corrosion product buildups at the top tube support plate, but were not dented. Of the 118 tubes recommended for preventive plugging,104 tubes were evaluated as having top tube support plate corrosion, 2 showed denting, 6 showed no detectable denting, and 6 were unreadable. Tube wall indications were also found in two tubes - R10C22 of Steam Generator B, and R9C3 of Steam Generator C. These tubes were plugged. 3.2 Identification of AVBs from Eddy Current Testing Since tube supports limit tube vibration potential, an evaluation of the presence or absence of AVB support was needed for each North Anna #2 tube. He minimum depth of AVB insertion in the Series 51 steam generator is to tube Row 11, and in most cases, the actual depth of insertion went to Row 10,9,8, and sometimes Row 7. Direct visual or fiberoptic observation of AVB positions would have been difficult and potentially unreliable, therefore, eddy current test information was l used to evaluate presence of AVBs. A typical ECT data trace indicating the intersection of an AVB with the tube is shown in Figure 3-1. The basis for the use of eddy current information to evaluate the presence of AVBs was developed l using data from laboratory tests with a standard bobbin coil. Considering a number of factors, including noise levels, lateral and vertical offset of the AVB from the tube, and the effect of tube deposits, the tests showed that under limiting conditions AVBs could be detected at a distance greater than the tube to-AVB gaps. As discussed later, AVBs provide adequate support for tubes even at the maximum possible Series 51 tube to AVB gap distance. Derefore, the presence of a verified AVB eddy current signal that provides support at the tube centerline indicates at least one-sided support. The term "visible" was used to indicate that a measurable, above noise level AVB eddy current signal was interpreted for the tube of interest, and conversely, "invisible" was used to indicate that l 5
, , .. . . _ . . . _ . . . _ . , . . . . . . ;.: :: .:.. n : . 4 1, I t I t l l xY O!!*,Cv g I le==sL. l .' CHa>= - ." l l T l rnE3 toe tx i 4 i sPow 75 1 l ROTATICH = 273 OG i
LErt stPtP eweer
, CW.no -
sv Ts FPEG 100 kM: l o
\U SPAN ROTATION - 273- CG 75 H
"" R!owt siete canse
~5ANe C - "Y s t FREC 109 k>=
, i ,
$PGH 84 ROTATION - 132 CE3
= f
.f ,. . at - tr.te fYSTE" CO#!O'.ipc? :N
~
j , mang - CEFET e of CHAH- 8 5:42:17 PM
. . . OCTCEER 26 !?ff i G L-
- n t.- .u ,..? e.- .
.a+ . . - 2
_N . 4-
/l 100' ' s H f 10? -i ! '
PLANT UNIT $/G LEG DATE E'J !! M I ATF.C CD-tv-1 1 PCCK UP 11'12<$7 l l l l l l I I l { r l FIGURE 3-1 Typical Eddy Current Signal for AVBs e, 1 l ! 6
.._..p-. ,_ ,_. , , _ _ , _ y _, . , , . , _ _ . - - - _ , , , - , , _ _ ,,, . . _ , , , o, , _ _ , . _ ,,, ,m-_-,
no AVB eddy current signals were apparent. Initial ECT evaluations of North Anna #2 AVB positions were limited to interpreting whether or not an AVB was "visible" or "invisible". Rows 8 through 12 were examined and "visible" and "invisible" interpretations were made for all tubes. Some anomalies appeared where it is likely that the signals were masked by copper deposits. Figures 3 2 through 3-4 sMw the "visible" "invisible" map (without AVBs)in this preliminary stage of data interpretation. 3.3 0,1 or 2 AVB Present Eddy Current Calls To obtain additional information for evaluating the AVB positions, eddy current information for selected tubes was also examined to evaluate the number of AVB intersections (including zero)
"visible" along the U-bend length. Tubes for the "O,1, or 2 AVB Present" callout were selected based upon the results of preliminary AVB positioning from the "visible" "invisible" maps.
Zero indicated that no AVBs were "visible", one (1) indicated that a continuous uninternipted AVB signal (regardless of length) was observed, and two (2) indicated that two or more distinct, separate AVB leg signals were observed. Single intersections (l's) can be interpreted in a number of ways: a) as a single bar (one sided support) inserted to a depth such that the AVB centerline is approximately tangent to the tube wall centerline, b) as two bars (two-sided suppon) both inserted to essentially the same depth, and c) as two bars (two sided support) with the second bar lower than the first bar, but still close enough to provide a continuous AVB eddy current feedback signal. The identification of one AVB signal does not, by itself, mean that the AVB is located at the tube centerline or below (i.e., supporting the tube on one side), since it is possible for the eddy current probe to identify an AVB that is slightly above the tube centerline. The overlay of the "visible" -
"invisible" data and the "O,1, or 2 AVB Present", as well as top tube support plate denting results, are shown in Figures 3 5 to 3-7 (a description of AVB Insertion Map techniques and Prevendve Plugging Criteria follow in Sections 5.0 and 6.0).
4.0 UNIT 2 VIBRATION ANALYSIS 4.1 Similitude with North Anna Unit #1 Since the design and operating parameters of North Anna Unit #1 and Unit #2 are essentially the same, the same vibration analyses are applicable to both. Test results documented in Reference I showed that single-sided AVB support limits the vibration amplitude for fluidelastic vibration. 7
999000 999900: 999900: 999900~ 999990: 999900; 999900: 999900^j 999900: 999900 999990-999990: 999990- 999990-999990: 999990: 999990-999900: 999990: 999990-S99900 999900: 999990-999900 999000: 899900-9 9 9 9 0 0= 9 9 9 0 0 0= 999900s 999990: 999000 999900: 999900: 999000: 999900: 999990 999000: 996900: 4 999990* 999000: 999900=
, 8 ;. 989000: 999000: 999900e d s2 899000: 999000: 999900*
5 af 999000: 999000: 999990: E 30 984000: 999900* 999990-
) 999000: 999900: 999990:
999909: 9999002 999990: 999000: 9999002 999990: 999000: 999000 999990: 999000: 999900: 999990 999900: 999900 999990: 999990: 999900: 999900: 999990: 9999002 999900: 2 l 999900* 999809- 999990: 51 899990: 999990: 999990- i l ! 999990: 999000s 999990*gg 499990: 9990002 999990:00 999990: 999000 999990:
- - - - - - 2 - - -
- 2 2 : 2 2 :
5 3 eeeeee 999900= eeeeeO~ esecee; eteeOO eteeeea 999900" 009000* 099900~ eeeeee* eeOOOO eeeeOO" seeste: 999000* 94e900- ! 300e08: 609900; eeeeee* eseeOO-e00000" ee0000-990000: 999900 e90900-e00000; 999000 eseeOOs 0e9990: 000000; e00000: 3e9990: 900000: eggeOO: eeeeeO OOOOOO; 090000;
- 3e9990 999000; 999000:
f 990400: SOOOOOc eeeeOO* m a; 999900: 490000* 090900:
$4 0e0900: 499000; etteOO:
s af ee9900: 999900: 8li eseeOO-090900: ee9900: e000C0* 099900 9999003 eeeeeeg I 999900: 699000; eteetO 999000: 90e0002 4e4000 099000: 9999002 esseeO 090900 999000; eteeOO: 690900* 9999008 9800000 eseeeO: 99990G* ee0000: ee9990: 999999 eteeOO 699000* Seesse: 900000" 0e4800: 999999 .i 999900: 969990: s3 999900 eteeGOR si e99000 499900 e00000 $$ l i 999000:
- s*--
Ge##ee: :: e--- 30
- e---
999900: 999900: 909'900: 999900: 990090~ !* 099000: 499900: 999990-999000: 999000: 999990-999900 999900: 999990-999900: 999900: 990990-999000* 999900: 999990-999900: 999900: 999990-999900: 999000: 999990 G 9 9 9 0 0: 999000: 999990: 999990: 999000~ 999990: 999990: 999000; SOSSOOu 999990 909000: 999900= 9 999900 999000: 999900: L 999900; 999000: 999900:
% ni S99900: 999000: 999900:
$d 999000: 999000: 999900:
8 aj 999000: 499000* #96900: i 999900: 499000: 999900: 999900: 9990002 999900: 499000: 9990002 999990: 499900: 999000: 899990: 999900: 9900002 999900: 999900: 999000 999000: 999990: 999000- 999600: l 999990: 990000: 999900: 999990* 999900: 999900: 3 i 999990: 699900: 999800: ij 999990: 999000: 499900: ",; 999990: 999000: 999900: aa 999990: 999900 999900: 00
--- a n s - -- ---
- 2 :
l
M SOOs @@@600s il
@@@@OOs @@@6008 @@GOOO"2 3 j
@@@@@Os @@@eOOx @@G@eO"ggl ;
@@@eOO8 @@@ GOOS @@G@eO'j3?j
@@@@@Om @@@@@Os @@@@GO* a 2 ! j
@@@@@Os @@@@@On @@@@GO-
@@@800s @@@@@O @@G@GO-{ !!]~1
@@@eOOR @@66OOn @@G@GO h..a* ..
@@@SOOK @@6900s @@GOOO-) } } )
@ @ @ G '7 0 R @@6900; Q=,Q= B,G O O s E t i i
====@OR @@60004 O= @ G S O O n 3 3 3
@@@ GOON @@6SOOc @@GOOOu j j j j{
5.@@@@@On @@6800* @@G900: 0000 ei@@@@@On @@6000s @@GOOO:
$j@@e900m @@6900: @@GSOOe _
m $u @@6800 @@G900s 3 S EE @@GOOg @@ @e G O O e @@Q@GO~ $
@@6900*
l 8 08 SOOs @@6SOO @@@@GOe gl
" il @@ @@ e@ @@@eOOs 900s == = Oe 5 ll@@@eO@s @@@SOOs M O a ! m}
*t@@@ GOOS @@8900s s @@G@GOaO@
@@@SOOs @@6@OOs ==G@@On
@@@900s @@8900s == =
On rg
@@e800s @@@SOOs
@@@@@Om @@@600s @@G800n]js 6 0: H
@@@@@Os @@@600m @@GOOOg P
@@@ GOOS @@@@O@s @@G@GOn1 3
@@@@@Os ==6660s
= = = =00a = = eeOOs @@@@GO g
@@@@GO g jj j
OOa @@6900s @@@@@Og hy-
@@@@60a @@SOOOu M@Om j
- --- 2: --- 2: 2 ---
y bl
~
FIGURE 3-6 North Anna #2 (VGB): S/G -B i AVB insertion and PlugginD ) 22 g gg g g g g g g g g 'gggggggggggggggggggg i ii
@ @@@@@@@@@@@@@@@@@O@@@@@@@@@@@@
l 10 @@@ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ 9 @@@@@@ @ @ @ @l@ @ @ O 0 0.@ O 0 0 @ @@@@ @ @ @ @ @ @ 8 OOOOO@@@OOOOOOO@O@O@@@OOO @ @ @ @ @ @g I 7 0000000000000000000000000@@@@@'4 93 92 91 90 69 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 l 22 ge@@@@@gegog@@@ @ @ @@ @ @ @ o g ggg @@@ ! ii @@@@@@@@@@@@@@@ @@@@@@@ @@@@@@@@ ! 10 @@@@@@@ @@O@@@@@ @@@@@ @@@ @ @ @@@@ @ 9
@@e@ @ @@@e99999999999999@@@@e94 @
j 8 000@ @ @O00000000000000000000000@ { 7
@O0@@@@O00000000000000000000000 62 61 60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 1
22 g]gggggggggg@gg@@@gggggggggggge ii 9@@@ @@OO@@ @@@ @@@ @ @ 3G@ @@ @@@@@ @ @ iO 9@ @ @ @ @ O O @ e @ @G@ @ @ @ @3 G @ @ @ @ @ G @ @ @ @ } 9 9@ @@@@@O@@@ @e@@@eeGG@@@ @@ G @@@ @ l 8 O@@OOOOO@O@@OOOOOOOOOO@@OO@69@ l 7 00000000000@O00000000000000000 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2
@ Preventively Plugged; Tube Denting Deforrnation Detected
,
- Nurnbers in circles represent the nurnber of distinct AVB leg @ AVB Visbie rnents ted
- frorn the tube. No entry or zer $ Preventively Plugged; Top Support Plate Corrosion Detected
_ @ Plugged (Wall Indication)
@ Preventively Plugged; No Detectable Donting siiseas
~
i @ Preventively Plugged; TSP Corrosio.1 Data Unreadable I
FIGURE 3-7 i North Anna #2 (VGB): S/G-C ] AVB insertion and Plugging j 22 ggggggge ggggggggggggggggggggg
" @@@@@@@@ @@ @ @@@@@@@@@@@@@@@O@@
) l io @@@@@@@@ @@@@@@ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ l 9 @@@@ @ @ @ @e999999@@@ @ @ @ @@@@@@@@@@ l 8
@e9@@@@O0000000000@@@O000000000
! 7 0000000000000000000000000000000 93 92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 12 h U U U h o U hh . n @ o o o g o o lo D @e@ De e@@e e@eeee@ @ @e@ @ 9 ete@ esteeeee@e@ee@eeeeO@@@e@@ j 8 7 0000000000000000000000000000000 i 0000000000000000000000000000000 62 61 60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 l
'2
@ @ b@@@@%@@@@@@@@@@@@@@@OulO=@@@@@
i @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @Ou @ @ @ @ @ l
@@@@@@@@@@@@@@@@@@@@@@@@@O@@@@
9
@@@@@@@@@ @ @ @@@@@@@@@ @ @a @ @ @ @ e @
000000000@@O00000000@===@@@@@@ J 8 7 l 000000000000000000000000000000 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 ] 2
- Numbers in drcies represent the number of distind AVB leg @ AVB Visable i segments indicated "Wsible" from the tube. No entry or zero p, g Preventively Plugged; Top Support Plate Corrosion Detected means no AVBs are "visable".
@ Preventively Plugged; No Detectable Denting an s re a Q Preventively Plugged; TSP Corrosion Data Urveadable
Reference 1 (Section 4.2) also describes the method used to determine the tube stability ratios for Unit #1. Figure 4-1 provides calculated results developed from Table 4.2.3.3-1 of Reference 1, and shows the tube stability ratios for tube Rows 8 through 11 pre- and post modification relative to the pre modification rado for R9CSI. "I'he modifications (principally the Downcomer Flow Resistance Plate or DFRP) are the same in both North Anna Unit #1 and Unit #2. The pre modification stability analyses represent the North Anna #1 operating conditions at the time of the R9C51 tube rupture. The expected relative stability ratios are based on the nominal or test based damping values for clamped tube supports and no local flow effects. Results for a low damping assumption with reduced dependence on void fraction and for constant damping (independent of void fraction) are also included in Figure 4-1. The relative stability ratio results with low and constant damping values are provided to assess the sensitivity of the downcomer modification to the damping dependence on void fraction. A summary of the stability ratio reductions can be obtained from the Row 9 results of Figure 4-1 as followv Reduction in Stability Ratios Due to Downcomer Flow Resistance Plane (DFRP) Modifications Stability Tube Dampine Assumption Ratio Reduction Low tube damping independent of void fraction 24 % Low tube damping with associated linear dependence 15 % on void fraction Nominal tube damping with test based dependence 9% on void fraction 14
FIGURE 4-1 Pre- and Post-Modification Stability Ratios Relative to R9CE Stability Ratio Relative to R9C51 Tube Row Nominal Damping Low Damping Constant Damoing
.. .... . ...... ... .. .. ..... .. ... Pre -Modifi cation -- --------------------------------
Row 8 0.84 0.85 0.86 Row 9 1.0 1.0 1.0 Row 10 1.16 1.15 1.16 Row 11 1.45 1.44 1.45
.. . ..... .. .Pos t-Modifica tion--------------------------------
Row 8 0.76 0.72 0.66 Row 9 0.91 0.85 0.76 Row 10 1.08 1.00 0.90 Row 11 1.34 1.25 1.12 i Notes: 1.) Ratios relative to Pre-Modification R9C51 2.) Calculations performed for Columns 51 in all rows 3.) Nominal Damping - Expected tube damping based on test results Low Damping - Reduced tube damping dependence on void fraction Constant Damping - Tube damping assumed independent of void fraction l l 15 1 4
If low tube damping (substantially lower than clamped tube test data) rather than local flow peaking effects were a contributor to the R9C51 tube rupture, the results of Figure 4-1 show that the downcomer modification is more effective in reducing stability ratios with the low damping assumption. Test data available since this evaluation show that local flow peaking effects are the dominant contributor to the R9C51 tube instability and that low tube damping assumptions m not necessary to explain the tube rupture. From Figure 4-1, it is seen that the Row 9 post-modification (with DFRP) relative stability ratios of 0.91 are close to meeting the Reference 1 criteria for a 10% stability ratio reduction. His is the case for other Row 9 tubes having the same local flow peaking factor as R9C51. For unsupported Row 9 tubes with lower now peaking than R9C51, as is expected to be the case, the 10% lower stability ratio criteria is satisfied. Unsupported Row 8 tubes with a relative stability ratio of 0.76 compared to R9C51 are acceptable even with peaking factors comparable to R9C51. He Row 8 tubes have negligible potential for flow peaking and thus can be expected to have relative stability ratios much less than 0.76. At North Anna #2, unsupported tubes in Row 9 or higher rows h ve been preventively plugged. Consequently, the highest relative stability ratio for a remaining active tube relative to R9C51 is less than 0.76. 4.2 Modeling at Tube Bundle Periphery The use of the ATHOS code to model the U bend region of North Anna Unit #1 was discussed in Section 4.2 of Reference 1. In addition to this basic information, certain limitations to ATHOS model were exammed. The ATHOS code does not include the capabdity to adequately model the presence of the AVBs in the U-bend region. However, Westinghouse has modified the code to include the capability to model the AVBs at a uniform depth ofinsertion via flow cell bounciary resistance factoni. Practical lower limits of cell size in the ATHOS code, however, prevent a fine grid representation of the AVB V-bar shape which, in turn, limits the accuracy of the AVB representation. ATHOS calculations have been performed with and without AVBs in the model. In addition, calculations show that the relative stability ratios for tubes near the center of the steam generator are essentially the same with or without AVBs in the models. The ATHOS AVB modeling sensitivity studies with uniform insertion show some tendency for the AVB resistance effect to lower tube gap velocities near the central regions and to increase velocities near the peripheral tubes. However, the magnitude of this effect is uncertain due to the limitations in ATHOS for modeling of the 16
AVBs. Further, the global flow resistance of staggered AVBs would be less than that for uniform insertion. Based on the sensitivity studies using ATHOS with and without uniformly inserted AVBs, the most reliable relative stabilit/ ratios (for actual steam generators with non-uniform AVB insertion depths) are expected using ATHQS models excluding AVBs. Results of ATHOS models available at the time of the North Anna #2 evaluation indicated potential peaks in the flow velocities of a tube row at tube locations 2 and 3 columns from the periphery. These peaks were included in the relative stability ratio calculations for North Anna #2. Later ATHOS modeling results have shown that the peaks at peripheral tubes are a consequence of inadequate mesh in the model. Consequently, the use of the stability ratio peaks at peripheral tubes for the North Anna #2 evaluation provides further conservatism in this assessment. 5.0 AVB MAPPING AND SUPPORTED TUBES 5.1 Consistency Checks in AVB Insertion Mapping From the AVB eddy current signal callouts, the AVB positions were evaluated. Since ambiguity can occur in the interpretation of the ECT data due to the inability of ECT to differentiate which side of a tube a "visible" AVB is located, and due to possible masking of signals due to copper deposits, other factors were considered to establish the location of the AVBs. These factors include consistency with the design of the AVB assembly, consistency of data for adjacent columns, and consistency of data within a tube column. In each instance, a conservative l judgement was made, locating the AVB insertion depth to an evaluated minimum possible insertion i depth. 5.2 Use of Eddy Current Data to Determine Supported Tubes To conservatively accommodate these uncertainties, the following guidelines were adopted for defining tubes with AVB support: 1.) The eddy current data should indicate the presence of 2 AVBs such that the bottom apex of the AVB would be below the tube centerline, OR 2.) When only AVB "visible" data is available for a tube column, only the larger row tube of two consecutive rows with AVB "visible" data is assumed to be supported by an AVB. 17 l l 1
The first guideline was adopted since eddy current can detect AVBs in the vicinity of the tube, even though the AVB may not support the tube, as discussed in Section 3.3. An example the second condition be see'n in SG B, Column 32 (See Figure 5-1). R8C32 and R9C32 are reported as "visible". By the second guideline, R9C32 is assumed to be supported, and is shown with single-sided AVB support. Although support is reasonably assumed for R9C32, both adjacent tubes, R9C31 and R9C33 were plugged, since it could not be determined on which side support was provided to R9C32. Figures 3-5 through 3-7 show AVB insertion maps developed according to the above guidelines of the AVB "visible" analysis, "O,1, or 2 AVB Present" analysis, as r ell as top tube support plate denting results of preventively plugged tubes. 6.0 CORRECTIVE ACTIONS 6.1 Pmventive Plugging After evaluating the tube support conditions (not supported, one sided support, two-sided support) from the AVB mapping, the tube vibration analysis results were used to set the criteria for preventive plugging. The criteria for acceptable tubes were as follows:
- Tubes with stability ratios less than 90% of North Anna #1 R9CSI are acceptable. This l criterion was developed in Reference 1.
l
. All Row 8 tubes, with the exception of Columns 3,4,91, and 92 were judged to be acceptable without AVB support. This conclusion is based on the results of Figure 4-1 which show Row 8 stability ratios, except for the noted peripheral columns, to be less than 90% of North Anna #1 R9CSI. Row 8, Columns 3,4,91, and 92 were judged to be acceptable provided that single sided AVB support is assured based on the criteria for AVB support given in Section 5.1.
+ All Row 9 or larger row tubes were judged to be acceptable provided that single sided AVB support is assured based on the criteria for AVB support given in Section 5.1.
18
O Figure 5-1 North Anna #2 (VGB): S/G -B Isolated View of R9C32 12 g ggg 11 @ @@@ 10 @ @@@ 9 @G99@,9@ @@ 8 OOOO@OO@O 7 000000000 36 35 34 33 32 di 30 29 28
@ AVB Visible g Preventatively Plugged Tube g Tube Plugged (WallIndication)
- Numbers in circles represent the number of distinct AVB leg segments indicated "visible" from the tube. No entry or zero means no AVBs are "visible".
3/14/88 rmwe
, . , _ _ . - . - - - - . _ - - , . _ _ _ _ . . - - - - - . . _ - - - - - . ~ . _ - - , . _ . - _ - - - , , . _ _ . _ _ _ _ _
These criteria are essentially identical with the North Anna #1 tube plugging criteria, except that the acceptability of single-sided support was extended from Row 9 to include all rows in the Unit #2 evaluation. The above criteria led to preventive plugging of all tubes without at least single sided AVB support in Rows 9 or larger, and Columns 3,4,91, and 92 in Row 8. The tubes selected for preventive plugging are also shown in Figures 3 5 through 3-7. 6.2 Downcomer Modification The mechanical, structural, and thermal hydraulic implications of the Downcomer Flow Resistance Plate modification (DFRP) are identical to those of Unit #1 (See Reference 1). Figure 6-1 shows a typical DFRP.
7.0 CONCLUSION
S Conservative criteria essentially identical to those of Unit #1 were used in the tube fatigue evaluation for Unit #2. Denting was conservatively assumed to have occurred at the top tube support plate of both the hot leg and cold leg sides of all Row 8 through 12 tubes. The installation of the Downcomer Flow Resistance Plate provided a stability ratio of all unplugged tubes which is at least 10% lower than the stability ratio of the failed R9C51 North Anna #1 tube, the criteria for acceptance. Preventive plugging associated with the tube fatigue mechanism was performed for 45 I tubes in SG A,29 tubes in SG-B, and 44 tubes in SG-C, for a total of 118 tubes. Two tubes were , also plugged for wallindications. I f With preventive tube plugging implemented for all unsupported tubes in Rows 9 or larger, the
- mimum stability ratio of any remaining, unsupported active tubes is less that 0.76 of the North l
Anna #1 R9C51 tube. This result assumes that the local flow peaking factor for the remaining unsupported tubes is the same as found for R9C51. However, the unsupported Row 8 tubes show negligible potential for local flow peaking. Based on peaking factors determined for the R9C51 tube since this North Anna #2 tube evaluation was completed, the maximum relative stability ratio of 0.76 can be reduced to less than 0.55 of R9C51. Consequently, the acceptance criteria of achieving tube stability ratios relative to R9C51 of 0.90 or less are exceeded by a large margin. l l l 20 l
/g e
/= 3/4 O -
a- . ym
\.\
FIGURE 6-1 Typical Downcomer Flow Resistance Plate 21 1 ,. - _- . _ _ _ . _ _ _ . _ _ _ _ _ . _ . _ . . _ . . _ - . - - _ _ _ _ - _ . _ _ _ _ _ _ _ _ , _ . . _ . _ _ _ _ _ _ _ _ _ _ - - _ _ _ _ _ _ _ _ _ _ _ _ _
r
's,'
8.0 REFERENCES
1.) WCAP-11601,"North Anna Unit 1. Steam Generator Tube Rupture and Remedial Actions Technical Evaluation", Westinghouse Electric Corporation NES, September,1987. 2.) Bulletin 88-02,"Rapidly Propagating Fatigue Cracks in Steam Generator Tubes", United States Nuclear Regulatory Commission (NRC), Washington, D.C. February 5,1988, 1 22}}