ML20106A827
| ML20106A827 | |
| Person / Time | |
|---|---|
| Site: | Catawba  |
| Issue date: | 09/30/1992 |
| From: | Wootten M WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP. |
| To: | |
| Shared Package | |
| ML20106A824 | List: |
| References | |
| WCAP-13495, NUDOCS 9209300042 | |
| Download: ML20106A827 (279) | |
Text
,
WESTINGHOUSE PROPRLCTARY CLASS 3 WCAP 13495 SG-92 09-015 -
Catawba Unit-1 Technical Support for Steam Generator Interim Tube Plugging Criteria for Indications at Tube Support Plates September 1992 i
, J Approved b * "I^
- 61. J. Wootten, ManageN Steam Generator Technology & Engineering This document contains informaaon proprietary to Westinghouse Electric Corporation; it is submitted in confidence and is to be used solely for the purpose for which it is furnished and returned upon request. This document and such information is not to be {
- reproduced, transmitted, disclosed or used otherwise in whole or in part without the prior written authorization of Westinghouse Electric Corporation, Nuclear Services Division, P. O. Box 355, Pittsburgn, PA 15230-0355.
l
@ 1992 Westinghouse Electric Corporation
- All Rights Reserved 9209300042 920918 3 DR ADOCK 0500,
... - .. - ...- .. - .. ~--~ ~ - --
Table of Contents SECliCD .Ilfe lbZt
1.0 INTRODUCTION
1-1 2.0 CONCLUSICNS 2-1 3.0 SUPPORT PLATE REGION PULLED TUBE DATABASE (3/4" TUBING) 31 3.1 Introduction & Definitions 3-1 3.2 Catawba-1 Corrosion Degradation 3-2 3.3 Plant E-4 Corrosion Degradation 3-3 3.4 Plant B Corrosion Degradation 3-4 4.0 MODEL BOILER SPECIMEN PREPARATION & TESTING 4-1
~
4.1 Preparation of Specimens 4-1 4.2 Leak Rate Testing 4-3 4.3 Burst Testing 4-4 4.4 Destructive Examinations 4-4 4.5 Model Boi!er Data Base Summary 4-7 5.0 NDE EXAMINATION 51 5.1 Voltage Normalization for APC 5-1 5.2 Eddy Data Analysis Guidelines 5-2
- 5.3 NDE Results for Model Boiler Specimens 52 5.4 Voltage Trends for EDM Slots 5-3 5.5 Frequency Renormalization Based on ASfdE Stds. 5-3 5.6 Renormalization of Catawba-1 Pulled Tube Data 5-4 5.7 Renormalization of Belgian Pulled Tube Data 5-5 5.8 NDE Uncertainties for Catawba-1 56 5.8.1 General Approach for APC 5-6 5.8.2 Catawba 1 NDE Uncartainties 5-7 .
6.0 PULLED TUBE AND FIELD DATA EVALUATION 6-1 6.1 Utilization of Field Data in Tube Repair Limits 6-1 6.2 Summary of Pulled Tube Database 6-1 6.3 Operating Plant Leakage Data for ODSCC at TSPs 6-2 6=4 Voltage Renormalization for Attemate Calibrations 6-2 6.5 Tensite Property Considerations 6-3 6.6 Evaluation of Catawba-1 Pulled Tubes 6-3 6.7 Evaluation of Plant E-4 Pulled Tubes 6-7 6.8 Evaluation of Plant B-1 Pulled Tubes 6-8 6.9 Growth Rate Trends 6-8 6.10 Summary of Pu!!ed Tube Test Results 6-9 7.0 LEAK RATE AND BURST CORRELATIONS 7-1 7.1 Introduction 7-1 7.2 Summary of Data Base for 3/4 Inch Tubing 7-1 7.3 Burst Pressure vs Voltage Correlation 7-1 7.4 SLB Leak Rate vs Voltage Correlation 7-4 7.5 Bounding SLB Leak Rate vs Voltage 7-6 i
Tablo of Contents (Continued)
Secti2D Ill'c fbac 8.0 ACCIDENT CONDITICN CONSIDERATlONS 81 8.1 Tubo Deformation Under Combined LOCA + SSE 81 8.2 Tube Deformation Under Combined SLB 4 SSE 8-6 9.0 CATAWBA UNIT-1 INSPECTION RESULTS 9-1 9.1 Inspection Scope 91 9.2 Summary of Inspection Results 91 9.3 Cross Calibration of ASME Standards 9-2 9.4 1992 Inspection Results at TSP Elevations 9-2 9.5 Voltage Growth Rates 9-3 10.0 CATAWBA UNIT 1 IPC EVALUATION 10-1 10t introduction 10-1 10.2 Catawba 1 Interim Plugging Criteria (IPC) 10-1 10.3 Equivalent Catawba 1 APC Repair Limit 10-2 10.4 Monte Carlo Methodology 10-3 1 10.5 Projected EOC 7 Voltage Distributions 10-4 10.6 Tube Burst Margin Assessment 10-7 10.7 SLB Leak Rate Assessment 10-8 10.8 Operating Leakage Limit 10 9 10.9 Conclusions 10-10 -
APPENDIX A Catawba Unit 1 Resizing NDE Analysis Guidehnes A-1 ,
e
- . .- -- - - - - - - . - . - - - - - - - ~-- - - - .
i Section 1 NTRODUCTiON k
This report provides the technical basis for interim tube plugging criteria (IPC) for outside diameter stress corrosion cracking (ODSCC) at tube support plate (TSP) intersections in the Catawba Unit.1 steam generators (S/G). The recommended repair limits are based upon bobbin coil Inspection voltage amplitude which is correlated with tube burst capability rand leakage _
potential, The recommended critoria are demonstrated to provide conservative margins relative .
to the guidelines of Regulatory Guide (R.G.) 1.121.
The tube repair limits are based upon the conservative assumptions that the tube to TSP crevices are open (negligible crevice deposits or TSP corrosion) and that the TSPs are displaced under accident conditions. The ODSC" existing within the TSPs is thus assumed to be free span degradation under accident conditions and the principal requirement for tube plugging considerations is to provide margins against tube burst per R.G.1.121. The open crevice assumption leads to maximum leak rates compared' packed crevicos and also maximizes the potential for TSP displacements under accident conditions. However, a TSP displacement analysis has not been performed for Mode! D steam generators and hence the potonilal for TSP displacements under SLB conditicns cannot be ruled out. Prevention of tube rupture by the reinforcement of the support cannot be assured without further analysis. Therefore the requirements for tube burst margins assuming free span degradation have been applied to 3
develop the tube repair limits for Catawba-1 S/Gs.
The repair limits were developed from testing of laboratory induced ODSCC specimens, extensive examination of pulled tubes from operating S/Gs including Catawba-1 and field experience for leakage due to indications at TSPs. The recommended IPC represent very conservative limits based upon Electric Power Research Institute (EPRI) and industry supported development >
programs that are continuing toward further refinement of the repair limits. The currently available data base is used to define burst pressures at the lower 95% confidence bound, The IPC repair limits provide significant margins against the currently recommended U rrelations and satisfy R.G.1.121 guidelines for attemate correlations reflecting current uncertainties in the burst pressure correlation.
implernentation of the tube repair limits is supplemented by 100% bobbin coil inspection at TSP elevations having ODSCC indications, reduced operating leakage requirements, inspection ,
guidelines to provide consistency in the voltage normalization and rotating pancake coil (RPC) inspection requirements for the larger indications left in service to characterize the principal degradatica mechanism as ODSCC In addition, potential SLB leakage has been assessed by both Monte Carlo and deterministic analyses for tubes with TSP indications left in service to demonstrate that the cumulative leakage is less than allowable limits.
Nine tube to TSP intersections from five tubes pulled from Catawba Unit-1 in 1991 were burst tested and destructively examined to provide direct support for the IPC. Five of these burst tests
,- were either incomplete bursts or burst at laboratory prepared grindmarks outside the TSP and hence could not be incorporated into the tube burst database. However, three intersections (one had no flaw indication from either NDE or destructive examination) could be combined with data from other plants and model boiler specimens. This results in a significant database supporting the IPC repair limits.
1-1
.. .. .. . --.. ...-.- . . - . . . .- . - - - - . . - . . . - ..-... - . ..- ~ .- ..- -. ....-
[
4 To provide the technical bases for tube plugging due to ODSCC at TSPs, the following activities. -
have been performed as docum6ated in this report: ..
o . Preparation of cracked test specimens, their non destructive examination (NDE), leak:
- l rate testing, burst testing, and destructive examination Section 4 L!
o NDE data analysis guidelines, NDE inspection results for the test specimens, voltage trends ;
for EDM (electrodischarge machining) slots, voltage normalizations and overall NDE ' +
uncertainties - Section 5 ,
i o Review of Catawba-1 and other plant pulled tube examinations - Section 6 !
o Leak rate and burst correlations to refrN the NDE parameters to burst strength and leak rate under SLB conditions Section 7 l
o Evaluations of combined accident conditions (LOCA + SSE) Section 8 o Review of Catawba-1 eddy current inspection results including historical growth rate data
- Section 9 o Integration of the inspection, leak rate and burst test results to develop the interim tube repair limits - Section 10.
t The overall summary and conclusions for this report are described in Section 2.
I t
'I 9 .
J 1-2 1
. _ - , . . . - - . .. ,....-,---.m. .e -, - . .,, y ,,- e
- . _ _ _ _ _ . . ~ . _ . - _ ._ _._ _ _ . . _ . _ _ _ _ _ . . _.m e
Section 2 CCNCLUSIONS This report documents the technical support for a Catawba.1 Intbrim plugging criteria (IPC) of '
1.0 valt for ODSCC indications at TSPs. The database of pullod tubo and model boiler specimens used in the evaluation of the IPC are described in this report. This database is used to develop correlations relating burst pressure to bobbin voltago and SLB leak rate to bobbin voltage.- Thoso-correlations, including conservative variations allowing for data uncertaintios at this timo, are used in the tubo integrity assessment to demonstrate Catawba 1 IPC margins against R.G.1.121 criteria for tube plugging limits.
The overall conclusions of this report are: '
o R.G.1.121 criteria for tube integrity are conservatively satisfied at EOC 7 for an IPC repair limit of 1.0 bobbin volt P
o At EOC 7, burd pressure capability (expressed as margir ratios relativo to 3APNO and APSLB) is expected to have ratios of about 1.25 relativo to 3aPNO at 90% cumulative probability levels and about 1.35 rotative to APSLB at 99% cumulativo probability levels. A burst pressure margin ratio of 1.4 relativo to 3AP NO for Catawba-1 at BOC conditions is comparable to typical values for plants with 7/8 inch diameter tubing with an IPC repair limit cl 1.0 volt. Thus the two tubing sizes can be considorod to have equivalent margins for IPC repair limit of 1.0 volt.
o Potential SLB loakage at EOC 7 is expected to be negligible (-0.01 opm) as supported by both Monte Carlo and deterministic ovaluations including sensitivity analyses.
o R.G.1.121 criteria for tube burst aro satisfied and no0ligibio St.B leakage is expected even under conservativo assumptions for ino voltago/ burst and voltage / teak rr'o correlations.
o The maximum EOC 7 bobbin voltage resulting from indications left in servico below the l
repair limit is expected to be about 2.53 volts, o The operating leak rato limit of 150 Opd imp.amented with the IPC satisfies R.G.1.121 guidelines for leak before break. This limit providos for plant shutdown prior to reaching critical crack longths for SLB conditions at a 95% confidence level on leak ,
l rates and for 3aP conditions at less than nominalleak ratos.
o inspection requirements for application of IPC repair limits were implemented in the 1992 inspection following Cycle 6 operation. The inspection included 100% bobbin coi!
inspection of TSP intersections,. RPC inspection of all bobbin flaw indications >1.0 volt and an RP_C samplo inspection of dented TSP intersections. Tubo repairs have been -
implemented at EOC 6 consistent with the IPC repair limit of 1.0 volt.
o The Catawba-1 pulled tubes (1991 outage) show that the crack morphology for indications at TSPs can be described as multiple ODSCC axial cracks within the TSP 21
length and with negligible volumstric IGA involvement. Burst data for 3 intersections from the Catawba 1 pulled tubes are included in the burst pressure vs bobbin voltage correlation.
o Recommended correlations of bobbin voltage to burst pressure and to SLB leakage, as well as alternate correlations for sensitivity analyses, are developed in this report. ,
These correlations form the basis for determining margins for burst and leanage as summarized above.
o Pulled tube and model boiler data are used to define a 2.0 volt threshold for leakage at SLB conditions. Between 2.0 and 3.5 volts (3.5 volts exceeds inaximum expected EOC indication of 2.53 volts, an SLB leak rate of 1.0 liter /hr can be conservatively applied for " deterministic' SLB leak rate analyses.
M S
e 2-2
- ~ - - .- ~_- - ~ -, _ . - - - - _ - .- - -. -. - - -.- -.--
i Section 3 1
SUPPORT PLATE REGION PULLED TUBE DATABASE (3/4 INCH TUBING)
. 3.1 Introduction & Nfinitions The following provides summary infornation regarding OD or@nated corrosion at support plate {
crevice regions of Alloy 600 tubing pulled from OtST n gcuerators at various plants including i Catawba Units 1. The data is presented in supput of the development of tube plugging criteria 1 for Catawba-1. First, pulled tube data from the Catawba 1 are reviewed followed by data from other plants.
The typc of intergranutar corrosion with regard to crack morphology and density (number, length, depth) oi cracks can influence the structural integrity of the tube and the eddy current
, response of the indicalic s. To support the tube repair criteria, the emphasi; for destructive examination is placed upon characterizing the morphology (SCC, IGA involvement), the number of cracks, and characterization of the largest crack networks with regard to length, depth and remaining ligaments between cracks. These crack details support laterpretation of structural .,
parameters such as leak rates and burst pressu.e, crack length and depth, and of eddy current
. parameters such as measured voltage w4h the goal of enMncing structural and eddy current 4 l evaluations of tube degradation. In selective cases, such as the 1990 Plant A 2 pulled tubes, the pulled tube evaluations included leak rate measurements, lo addition to the more standard burst ;
pressure measurements, for further support of the integrity and plugging limit evaluations.
Before the support plate region corrosion degradation can be adequately described, some key corrosion morphology terros need to be defined. Intergranular corrosion morphology can vary 2
from IGA to SCC to combinations of the two. IGA (Intergranular Attack) is defined as a throo dimensional corrosion degradation which occurs along grain boundaries. The radial dimension has a relatively constant value when viewed from different axial and circumferential coordinater. IGA can occur in isolated patches or as extensive networks which may encompass the entire circumferential dimension within the concentrating crevice. Figure 3-1 provides a sketd1 of these IGA morphologies. As defined by Westinghouse, the width of the corrosion should be equal to or greater than the depth of the corrosion for th6 degradation to be Classified as IGA.
4 The growth of IGA is relatively stress independent. IGSCC (Intergranular Stress Corrosion Cracking) is defiaod as a two-dimensional corrosion degradation of grain boundaries that is -
strongly stress dependent. IGSCC is typically observed in the axial-radial plane in steam generator tubing, but can occur in the circumferential-radial plane or in combinations of the i two planes. The IGSCC can occur as a single two dimensional crack, or it can occur with branches coming off the main plane. Figure 3-2 provides a sketch of these IGSCC morphologies. Both of the IGSCC variations can occur with minor to major components of IGA. The IGA componant can 4
occur simply as an IGA base with SCC protruding through the IGA base or the SCC plane may have -
a semi-three dimensional characteristic. Figure 3 3 provides a sketch of some of the morphologies possible with combinations of IGSCC and IGA. Based on laboratory corrosion tests,'
It is believed that the latter, SCC protrusions with significant IGA aspects, grow at rates similar to that of SCC, as opposed to the slower rates usually associated with IGA. When IGSCC and IGA i are both present, the IGSCC will penetrate throughwall first and provide the leak path.
~
To provide a semi-quantitative way of characterizing the amount of IGA associated with a given crack, the depth of the crack is divided by the width of the IGA as measured at the mid-depth of ,
the crack, creating a ratio D/W. Three arbitrary D/W categories were created: minor (D/W 3-1 i
s
. - , - ~ , , . -- . . .. -
greater than 20) (all or most PWSCC would be included in this category if it were being considered in this analysis); moderate (D/W between 3 and 20); and significant (D/W less than
- 3) where for a given crack with a D/W of 1 or less, the morphology is that of patch IGA.
The density of cracking can vary from one single large crack (usually a inacrocrack composed of many microcracks which nucleated along a line that has only a very small width and which then
~
grew together by intergranular corrosion) to hundreds of very short microcracks that may have_
partially linked together to form dozens of larger macrocracks. Note that in cases where a very high density of cracks are present (usually axial cra;ks) and where these cracks also have significant IGA components, then the outer surface of the tube (crack origin surface) can form regions with effective three dimensionalIGA. Axial deformations of the tube may then cause circumferential openings on the outer surface of the tube within the three dimensional network of IGA: these networks are sometimes mistakenly referred to as circumferential cracks. The axial cracks, however, will still be the deeper and the dominant degradation, as compared to IGA.
Recognizing all of the gradations between IGA and IGSCC can be difficuli. In addition to observing patch IGA, cellular IGA / SCC has been recently recognized, in cellular IGA / SCC, the cell walls have IGSCC to IGA characteristics while the interiors of the cells have nonNraded me'al. The cells are usually equiaxial and are typically 25 to 50 mils in diameter. The cell wa!is (with intergranular corrosion) are typically 3 to 10 grains (1 to 4 mils) thick. The thickness and shape of the cell walls do not change substantially with radial depth. Visual examinations or limited combinations of axial and transverse metallography will not readily distinguish cellular IGA / SCC from extensive and closely spaced axial IGSCC with circumferentialledges linking axial microcracks, especially if moderate to significant IGA components exist in association with the cracking. Radial metallography is required to definir 31y recognize cellular IGA / SCC, Cellular -
IGA / SCC can cover relatively large regions of a support plate crevice (a large fraction of a tube quadrant within the crevice region). Figure 3-4 shows an example of cellular IGA / SCC from Plant L. '
A given rupport plate region can have intergranular corrosion that ranges from IGA tMough individual IGSCC without IGA components.
3.2 Catawba.1 Corrosion Degradation Five tubes have been removed from Catawba-1 for which nine intarsections have been destructively examined. Data collected in the post-pull laboratory examination that supports APC applications included NDE, burst :ests, leak tests at room temperature (one at prototypic conditions) and meta!!ography including sequential grinds to characterize crack patterns / depths and morphology. The NDE data for the Catawba-1 pulled tubes are discussed in Section 5 and the burst tests, leak tests and crack maps are presented in Section 6.6. This section focuses on the crack morphology for the Catawba-1 pulled tubes.
Degradation at the Catawba-1 TSP intersections has been found primarily in the form of axially oriented, intergranular stress corrosion cracks. Uniform IGA attack up to 5% throughwall was also found. in the tube examination resuits, isolated IGA patches are reported up to 50% .
throughwall. These regions called IGA patches would not be classified as significant IGA by the definitions given in Section 3.1 as the width of the IGA or volumctric involvement is small.
Four TSP intersections were incrementally ground and polished and metallography taken at each grind (maps given in Section 6.6). Examples of the metallography are shown in Figures 3-5 to 3-2
3 8. These Figures are typical of the variation from minimal degradation to the . ,uest and most closely spaced cracks. Typical areas identified as IGA patches are shown in Figures 3-5 to 3 8.
The volumetric involvement of IGA at these patches is too small to be significant for tube
. structural integrity considerations. Figures 3-7 and 3 8 show regions of cinely spaced axial cracks, in some cases, the closely spaced cracks have similar wall penetration as seen in Figure 3 7.
Overall, the Catawba 1 tubes show multiple, axially oriented ODSCC cracks with minor volumetric IGA involvement. This morphology is very similar to other domestic pulled tubes with no or minimal cellular SCC.
3.3 Plant E-4 Corrosion Degradation Steam generator tubes at support plate crevice regions in the European Plant E-4 have developed cellular IGA / SCC. The cellular IGA / SCC is localized in the crevice region such that most of the crevice region is free of corrosion. The crevice regions had moderate crack densities, moderate IGA components associated with individual major cracks, ar.d no significant IGA independent cracking. Burst tests conducted produced the expected axial opening through complex mixtures of axial, circumferential and oblique cracks. For the more strongly affected areas, while the ,
cracking remained multi-directional, there was a predominance of axial cracking. Figures 3-9 and 310 provide radial section photomicrographs through two of the more strongly affected areas showing cellular IGA / SCC at Plant E-4, 3.4 Plant B 1 Corrosion Degradation A description of the corrosion found at TSP 5 of Plant B-1 is provided below. This region is singled-out for two reasons. First of all, it has through wall corrosion. Secondly, the tube had a small region believed to have cellular IGA / SCC.
OD origin, axially orientated, intergranular stress conosion cracks were observed confined _
entirely within the fifth $Upport plate crevice region on the hot leg side of tube R4 C61 from Steam Generator C of Plant B-1. Six axial macrocracks were observed around the circumference. The largest of these was examined by SEM fractography without any metallography. The macrocrack was 0.4 inch long and through wall for 0.01 inch. However, the crack was nearly (onectively) through wall for 0.1 inch. The macrocrack was composed of seven individual microcracks that had mostly grown topther by intergranular corrosion (the separating ledges had intergranular features that ranged from 40 to 90% of the length of the ledges). Since no metallography was perforrred on the axial crccks,it is not possible to definitively describe the axial crack morphology at this location. At the eighth support plate region of the same tube, metallography showed that the morphology was that of SCC with a crack depth to IGA width ratio (D/W) of 15. Figure 3-11 summarizes the crack distribution and morphology data for the fihh support plate crevice region.
, in aodition to the OD origin axial macrocracks observed at the fifth support plate region, one location adjacent to the burst crack had five intergranular circumferential cracks. The maximum penetration observed for the circumferentia' cracking was 46% through wall. The morphology of the circumferential cracking was more that of IGA patches than of SCC. In addition to the 5 main circumferential cracks, the region had numerous smaller cracks aligned in both the axial and circumferential directions providing a crazed appearance. See Figure 4
3-3
.-. . . . . . . . . . - -_ . . _ _ - ~ ~ . - . - - . . . . -_- ... . . .._. -, .
3-12.-- This crazed degradation is now recognized as probably being cellular IGA /SCO.
- Previously the crazed pattern was thought to represent only shallow IGA type degradation that
. completely disappeared a short distance below the surface. Figure 3-13 provides micrographs
. of relevant cracks showing the morphology of axial and circumferentialcracks. As stated above.
- the axial cracks had a mo phology of IGSCC with a moderate DM ratio of 15 while the circumferential cracking had a morphology more like that of IGA, with a DM ratio of 1.
Field eody current bobbin probe inspection (in June 1989, just prior to the tube pull) of the fihh support plate crevice region produced a 1.9 volt,74% deep indication in the 550/100 kHz differential mix.
1 O
e i
Y W
O 9..
34 4
m,, . . . .. -. - ._ .,. . .-. ...
,- tube 00 l -
P A
tube 10 Patch IGA tube 0D
\
h K.
, s, ,
Ah, g;- " '
bb ,
) \ tube 10
) f'
/
/
/
Unifom IGA Figure 3-1 Patch and unifom IGA morphology as observed in a transverse tube section. (A similar observation would be made from a longitudinal section.)
3-5
-. .. . ~ . . . . . . .-
k
~
- tube 00 tube ID Branch SCC Simple SCC transverse section schematic ,
tube 00 tube ID longitudinal section schematic Figure 3-2 Schematic of simple IGSCC and branch IGSCC. Note that branch and simple IGSCC are not distinguishable from a longitudinal metallographic section. From a longitudinal section, they also look similar to IGA (See Figure 4.3).
3-6
. . . . . . - _ . . . . . . ., .. .~. . - - - .. .. .. .. - .... . -..~ ... -. .-.
e
.- tube 00.
t IGA with IGA IGA with IGSCC/ IGA fingers i IGA with IGSCC ( ) (0/W . 1).
fingers ,
(0/W > 10) f .
/ tube 10 transverse section schematic t
l -
l l-tube 00 -
1 tube 10 longitudinal section schematic
.- Figure 3-3 Schematic of IGA with IGSCC fingers and IGA with IGA fingers. Note that neither of the above variations can'be distinguished from a longitudinal section.
l I
3-7
! I
'l
-l 1
r , j1( .!! ! ,illt i, ' l .- ;l il ; :Il" Fr il ' !! ;
- X 5 .
I I
^
y .
1'~ % e. ag .
[l - .$
- M 4 .
r . O .* ,,' -
(~ -
' ,\ . 5 (-
. f . .
Jg -
p e
. 1 - ;
t g- 3a) ' e D
. 7 7.,/a)f' ', f
, . , s l h
, . i 1 M t w i
o
' 1 w l
' , 2 l e ntlb
. ore .
iocs s
> gp epel l l n.
. ruha.l st dc e
~ o a 1 e i ,n2ne
. ' t . n4i eh
.\. l, -
- c. o7 dtt e Cent d - gaaf ir. e6n l o D m1 a1f
. rRh1 l 1-e, v l .
- a. .
o f e c,aa sl Q ;
t x .
rbe4 ww eul
'A . A. pt tf nl
. (.,. tooa e ' ,
- yni ii .
, . - p h ol stt
'a (
[.. ;
e e
p anht ee hcn D
rot psr
- gi i e e
'1,/-
s -
\g s ot wdtf .
. - /!.
- ? i l a lddt cu um
<\ ;
.y( ~ )
N -1 t M aana c t ru egoshi er f- '
1 mef st c
. 1 d a l s n.t e
,?
A. 5. - .-
- 4. . . ^
4 di al ak ah iawcht atAh it g
rnGt en h'. *%~
- fr eI l on ti
- s- oerlNe
.A A - f aa p
- t sml huu pcll ey w .o arll cl
. rieeal gcC cf a o ri rd cn ia1 a
.dut nsn e
[ - '
S 3
~
y 3 L - -
m ol ee0e t at p oiaaee h xlh h r Papst p 0r f
T'- :. ' p
..~
' y ,F +
p '-
, D e
e 4 f ._ ~ .
.- s 3
}/h Y. ) '
1f.,),Q t .R l
%\
s tf =,
- r. ,*.. -
~ .-
i H e r
u h..
4 g .
i
,~ .. . {- e F
. ^ . .
~
~
t , % , E" . y$l:-
'.hvf.
'ay..
A.
u.c o q, , ; ,I
ZGA Paich
- 23' ,7.5%
17' .
) !
i l
l
, e,, .%.. . ;g ,.3.l t. .
, r
' a' , . ;* . l
$ '.... 5;'t.
\
- h ,
t-
- l. ly0l(
-V
%'.? ~
' yR ,'i
! hih,y,.v.&ik.m(/
', ,f,
, fst r,, 1, h'- ',g
/; ,.q,.,. Yqp.
~4
- ., y . ,
' g.'d.h
- ,r v.2 , j i
is _ - w WS M,., -U?-
! 1 F92-024 116A I
3rd Grind 20' I
' _B A
. 98'/39% 93*/395 e
1 I
s
.r s ,- 3 c .* .
+
4
) , h j, F
. 4,42 9- ~
T '4* .- ..
~..
) . -
t F92-036 60X.
i - 4th Grind. 95*
1 Specimen 5 10-69-48-2E Figure 3-5 Metallography Resul ts: R10C69. TSP 3 i ;
3-9 1
.174* 177' 180*
-12% 24%
37%
I F92-087 50X Grind #1d 175* '
?%
- n , -
ti D
a I
F92-095 50X Grind fil '
10-6-5B-2E, Mount #2 Figure 3-6 Metallography Results: R10C6, TSP 2 3-10
l rGA Pdch K I I J 67% 69f8T'. 60.9%
- 6 .3 60.9% 58.1%
275' 260*
h I 't - -
'p
, ,, i l- . : 8 .
\ *
~
\l l
l l .. i I
- T K L m 7 l 920218-06 COX 16th Erind l l
l i l F I J B 3 7 . 8*. 50. 9Y. 58.1% 20.4%
290* i 275'
.e
' ' l i'
y '
hf p p - -
8 l
.L
- j. . -
F .'y
, 2 I
. g ;.; . ;
^
N
. f.
l $ f;'#" . . ' g ; ~. f, * . L . , . . .
..z..
.) ;
920218-09 50X 16th Grind SPECIMEN 7-47-4B-2E Figure 3-7 Metallography Results: R7C47, TSP 3 4
3-11
329' 44.3% 44.5% 44.5% 62.75 315' s' \
t F92-163 60X 9th Grind 347' 60.5% W .9%
74.4% 330' F92-164 60X 9th Grind ~
METALLOGRAPHY RESULTS ON 5-112-88-2B Figure 3-8 Metallography Results: RSC112 TSP 3 3-12
____..._._ _ _ ___ ~ _ . - . . _ . _ . _ -._ - . _ _ . _ _ _ _ _ . _ _ _ ._ _ -.. _ _. _ _. _ .. _
i f
-:;.-- -.y . .
.[,.
,.2.
v , .u.
, , u..
i
~l " ~ . ,. -
t'
?. '. * '- q .
. l 4 , I
.' h, y
.f. .(, .
3, .k . . .,
- r. -
. :.s. . .
- t. . .
.T,x ..-
e
.. .u-
~ ' ~
~ *
' '. 3. , ' . ' .
g
, ~ .. . . ,
7
.c : . '. M.
(" .i .' ".. % _..
- Jk . :.x LW ..
, ' ,4-- ...
~
n
~.
^.
$r fg... 4
. . '. I '. . , ,
i . . .. "
ms- P c
r . o i
. f. ' -
qp. . -
3v N
m.
~
,') ', ,
[
. .- 3 s a
' =
\ . ;..
- p. ,
q N
..- U ;;
. o .s
. Q1 , ,
, ,', 4 <t:
. .,9 g _'.
.~. .
'N . . .
.. y; n
i * '~ -
4 x .
Q
- - s .
?*
x .
' .,s f) .-
I' o
i 9
,s-
'r .
.* ,j ~ .
I
. , .[
l T. . ,.
50X Figure 3-9 Radial metallographic section through a portion of the third
. support plate crevice region of tube R19-C35 from Plant E-4. A cellular IGA / SCC structure is observed. The depth of the section was not specified.
3-13 l
( . --* M I '. " ,,' 1 7 ,
.. C .' . h,h.%
hh , A ' C . ." ; ? '? ' 'j '; 'i hp e
5 m, . Q' ,.: 3 jr*
^
. 4 . ., .
4 . *i. ..- .
.5
/- ."
e e-lp,,, . g _ .. . g . .. ;
'l ,,, \ L,i,. f 3 e
- g
.. . ; &.. . < ~ 3 . ,
y ,. ,, ,:s' s M' .- m '
, -v o
-r'-J, R ),
s Ag ,
4,u . .,, *. - , a u
\.t.
- 4 '\ -
v+
[ fY .-
o 4
.- s
- -Q 3 ' $.
.&g "r~)4$
j eK
. : .2
~
f[. , % f h'.' E
~ .. ., h ,. y .
N 'g Q. g . -
.'1,e , n[, ~ f . \, . , ,
_ x. j m. . p.
f .- %. . . g, ;f .
y :' ,-d.D
- .. . a .. ., ..
ea. . ..:.t.
50X l
Figure 3-10 Radial metallographic section through a portion of the l fourth support plate crevice region of tube R19-C35 from Plant E-4. A cellular IGA / SCC structure is observed. The depth of the section was not specified.
3-14 "W '*- ------r
. .- . = - _ - - - _ _ . . . _ . . -. .. . - . . . _ . . . . - - . . . - _ . . . . - . . . . . -.
, 00
) .
i (/I 10
/ .
Sketch of Burst Crack Macrocrack Length = 0.4 inch Throughwall Length = 0.01 inch Number of Microcracks = 7 (all ligaments have predominantly intergranular features)
Morpholcgy = IGSCC with some !GA aspects (circumferential cracking has more IGA characteristics) 0.75 inches - - SP top q
/J 0.0 inches - - SP bottom 0
180 0
270 0 0 0 90 0 180 Sketch of Crack Distribution-Figure 3-11 Description of OD origin corrosion at the fifth support plate crevice region of tube R4-C61. Plant B 1 3-15
- _ - ~ _ _._ _ - - - - - . _ . - . - . - . - . - . . - _ _ - - . . - . - - . . . - . - .
'i 1
a I
AxialDirecUon .
- b . ,
Fracts 4 '
y ,
{
v 's / Face o;, f l Axial Crack i
I, 4
I t
I l
I t
e i
i s ".
. r i
r Figure 3-12 - ,
Network of small, mostly circu=ferentially oriented OD Tube R4-C61. surface cracks observed in the 5th support plate regien, t
1 3-16 l
1 i .__.......__,_,.._,._,,_....,,.__m.,_,,_-,._,_. . . . . . . _ . _ , _ . . . ,
4 ; >' ,1.
c ~~ e,p * ' e . ,'. s ;
- v. . .
. . .r-.r.
,h. <: .g~ .
~\.g y* - i ,k .
' r * ., ' ,-
.Y,
? .-- A q ?, ,,,. . - N, . . r% ~ .-
, h ;'z
- 7
. w]g* i.~. cf.**~ . r' p .. j'N, R ,, ** - " 4'.3 c
G.hy,q%qf,r.:.: ,
a L<
%$ ..Wy ,'f Q, u.g.g
- ]-
.;u .
h , ' d ;.':
Iv '*%Wl.%- i' kJ t . h A ij'4.=
f '"in ~u V,s;y:;r~
. y;,'. -
. J. . . ,% p *:. r,: .t. ~fd-K-K A P~ ..'. ':. . 'c LwW_. % .~~:,g , .gr- ::.. r.,e. , 5WI.; u...i .*/ x;;7
- ' ; W .6g. % ,. ,. . A .: g , , '. F. M: .,L .;, Q
.s ,
g.,4, r t.y t~ f. *%
3* a' e .& c .u.w.m Ed hg, 100X
- i. . e x
- s. .
~
- 1 P #-
- %h :* % -
sc . -
?
&Mu = fplt- :}Yk*'yh >V
? * '
sk
- - g&. ... .i -
... . , .-- .e p a.4 g . ~%e- . ~~
s-
- ~ -
W,.
,- ' $.m .
. g. - t.
W%
. .- "al , .
.~.
hg. IcoI Figure 3-13 Photomicrographs of tube R4-C61 corrosion degradationws .
Top photo sho
, axial crack morphology (transverse section) at the eighth support plate beation (no transverse metallography was performed at the fifth support plate region). Bottom photo shows circumferential crack morphology (axial section) at fifth support plate region.
3-17
r- ----,,,n-,,,
W' G
w-4 m
O e
4 h
%--. % - _ -~~ __ % __
l Section 4
! ABORATORY SPECIMEN PREPARATION
. 4.1 Preparation of Specimas l Cracked tube specimens were produced in the Forest Hills Single Tube Model Boiler trt facility.
1 -
The f acility consisted of thirteen pressure vessels in which a forced flow primary system transfered heat to a natural circulation recondary system. Appropriate test specimens were olaced around a single heat transfer tube to simulate steam generator tube support plates The tesk ;/ete conducted in two boiler configurations, shown schematically in Figures 4-1 and 4 2.
The majority of the tests were conducted in the vertical!y oriented boilers shown in Figure 4-1, in which four nupport plates were typically mounted on the tube. A few tests were conducted in horizontally moun'ed boilers, shown in Figure 4 2. Because there was no steam space in the horizontal boilers, seven support plates could be mounted on the heat transk. tube. Since
~'
capillary forces, rather than gravity forces, dictate the flow pattern in packed tube support plate crevices, the tube orientation should have little effect on the kinetics of the corrosion processes.
The thermal-hydraulic specifications utilized in the test are presented in Table 4-1. As indicated, the temperatures are representative of those found in PWR steam generators, and the heat flux is typical of that found on the hot leg side of the steam generator. The tests utilized 3/4 inch (1.9 cm) O.D. mill annealed alloy 600 tubing from heat NX7368. The tubing was manuactured by the Plymouth Tubing Co. to Westinghouse specifications. The chemical and physical properties of the tubing are presented in Table 4 2.
The cracks were produced in what is termed the reference cracking chemistry, consisting of either 600 ppb (1X) or 6 pp' (10X) sodium as sodium carbonate in the makeup tank.
Typical!y a test was initiated with the 1X chemistry, and if a through wallleak was not identified after 30 days of operation, the 10X chemistry was applied. The occurrence of primary to secondary leakage was determined by monitoring the boilers for lithium, which would ordinarily only be present in the primary system. Because of hideout in the crevices, the boiler sodium concentration was typically between 50 and 75% of the makeup tank concentration. Hydrazine and ammonia were also added to the makeup tanks for oxygen and pH control, respectively.
A summary of the test pieces which were subsequently leak and burst tested is presented in Table 4-3. Two groups of tests are listed; the EPRI test pieces were prepared under (funded by) this program, while the Spanish test pieces were fabricated for a group of Spanish utilities.
Permisrion frorn the utilities hcs been obtained to use the results of these tests in other applicatior- 59 only difference between the two groups of tests is that the crevices were f acked with ' *nt sludge formulations. As in most previous model boiler test programs, the EPRI tests tu .. at is termed simulated plant sludge while the Spanish tests used a formulation more rephrnative of that typica!!y found in steam generators in Spanish plants. As indicated in Table 4-4, the only difference between the two formulations is that magnetite has replaced the metallic copper content in the simu!sted plant s!udge.
. As outlined in Table 4-3, three means of packing the tube support plate crevices were utilized.
in the fritted configuration, loose sludge was vibratorily packed into the crevice and then held in place with alloy 600 porous frits placed over both ends of the crevice. In this configuration, cracks were typically produced near the interface between the sludge and the frits, in some cases, multiple cracks were produced at both ends of the crevice.
4-1
The dual consolidated configuration consisted of two sludge regions,in which the outer teglon contained chromic oxido, while the inner region contained oither simulated plant or Spanish sludge. The regions had the following dPnensions, with the distances given in millimeters:
)
I A Chromic oxide I E A 19.0 Simulated Plant Sludge A 6.4 V
n 3.8 k V 4- 4.1-> 3.2 +- 4.1 ->
1 11.4 >
The two-region sludge configuration was specified in order to limit cracking to the small inner region, containing an oxidizing sh. he. Chromic oxide is nonoxidizing, and previous testing had found that accelerated corrosion is less likely to occur in its presence. The outer region provided thetmal insulation fcr the 1"",i r region, so that the temperaturo in the inner region was sufficle'itly high to produce ccolorated corrosion. The two sludge regions were baked onto the tube using a - xture consisting of 5% sodium hydroxide,2.5% sodium sulfate, and 0.8% sodium silicate. The support plates were then mounted on the tube over the sludge and hold in placo with externally mounted set screws. Sirico corrosic, should be contined to the inner region, this configuration was W anded to produce short, individual cracks.
The mechanically coasniioatec sivdge configuration was fabricated by mechanically compacting sludge within a tubo rouport plate simulant, drilling a hole in the sludge for the tube, and then sliding tho tube through the holo until positioned properly. Tsis configuration was used because relatively low voltage indications had been produced in previous tests usieg this configuration.
As indicated in Table 4 3, thero was considerablo variation in the time taken for a crack to be produced in a given test picco in general, cracking was produced in shorter time spans with this heat of material (NX7368) than for the heats used in similar tests performed with 7/8 inch .
diameter tubes. Cracks were typically produced most rapidly with the fritted configuration and most filowly with the dual consolidated configuration although a few cracks were produced very quick'y with the dual consolidated configuration. Det ,ils of crack networks produced in the model boiler specimens are preser'ted in subsection 4.4.
42
l 4.2 Leak Rato Testing l
The objective of the leak rate tests is to determino the relationship between oddy currcnt characteristics and the leak rates of tubes with stress cerrosion crucks. Leak rates at normal operatMg pressure differentials and under steam line broek conditions are both of interest, since loakago limits are imposed under both circumstancos. Tho SLB toak rato data aro used to dovolop a formutation between leak rate and bobbin coil voltago.
Crovien condition is an important factor, Tightly packed or dented crovices are expected to significantly impedo leakago throug5 cracked tubes. Sinco denting is readily detectable by non dest'uctive means while crevice gaps cannot be readily assessed, the emphasis is placed upon open crovices and dented crevices as the limiting cases.
Loak testing of cracked tubes is accoraplished as follows. The ends of the tubo are plug wolded. ~
One ond has a fitting for a supply of lithlated (2 ppm LI), borated (1200 ppm B) and hydrogenated (1 psia water to the tubo inner diameter. The specimen is placed in en autoclave and brought to a temperature of 616'F and a pressure of 2250 psl. The pressure on the outer diameter is brought to 1000 psi. A back pressure regulator on tho secondary side maintains the 1000 psi pressure. Any leakage from the primary side of the tubo tends to increase the secondary pressure because of the superhoated conditions. The back pressure regulator then opens, the fluid is released, condensod, collected and measured as a function of timo. This provides the measured leak rato. The cooling coilis located prior to the back pressure regulator to provent overheating and to provido good pressure control. Typicalleakago duration !s one hour unless leak rate is excessive and overheating of the back pressuro regulator occurs.
Pressure is controlled on the primary side of the tubo by continuous pumping against another back pressure regulator set at 2250 psl. The bypass fluid from this regulator is returned to tho makeup tank.
To simulato steam imo break conditions the primary [ ,ssure is increased to 3000 psi by a simple adjustment of the back pressure regulator ar;d secondary side is vented within one to three minutes to a pressure of 350 psl. The pressure differential across the tubo is thus 2650 psi. Temperature fluctuations settle out in several minutes and the leakage test period lasts for _
approximately 30 minutes.
A summary of leak test results is provided in Tablo 4-5. Leak rates at normal operating pressure differential and at steam line break conditions were obtained for all specimens. The .
steam line break conditior.s increased the leak rates by about a factor of three compared to normal operatin0 conditions. Moro variation in this factor can be expected. Prolonged loak ri..e testing under operating conditions is expected to lead to lower rates. The increase in the leak rato upon transition to accident conditions then becomes rnore variable.
4.3 Burst Testing Given the assumption that significant support plato displacements cannot be excluded under
, accident conditions, burst tests of tubes with stress corrosion cracks are conducted in the free span condition and burst pressure is correlated with bobbin coil voltage. This burst pressure correlation is then applied to determine the voltage amplitude that satisfies the guidelines of Reg.
Guido 1.121 for tube burst margins.
Burst tests v,ero conducted using an air driven differential piston water pump at room 43
temperature. Pressure was recorded as a function of time on an X.Y plotter, Sealing was accomplished by use of a soft plastic bladder. Burst tests of tubos with stress corrosion cracks were done in the free span condition. No foil reenforcement of the sealing bladders was used e,ince the crack location which was to dominato the burst behavior was not always readily apparent. Some of the maximum openings developed during burst testing were not sufficient to
- causo extensive crack tearing and thus represent lower bounds to the burst pressures. The openings were largo enough in all cases to lead to largo leakago. Burst test results are ,
summarized in Table 4 5.
4.4 Destnictive Examination 4.4.1 Objectives The objectrve of this task is to characterize the slzo, shape, and morphology of the laboratory created corrosion in alloy 600 tubo specimens which have been leak rato and burst losted. The crack morphology is also to be compared generally to the corrosion morphology observed in tubes pulled from operating power plant staam generators. A summary of the available results for T4 inch OD specimens is presented in this section.
4.4.2 Examination Methods Examination methods includo visual examinatbns, macrophotography, light microscopy an:i'or SEM (scanning electron microscopy) examinations, SEM fractography, and metallography. A number of model boiler test specimens were selected for destructivo examinations. Most of these were leak and burst tested. '
The specimens were initially examined visually and with a low power microscope. The burtt .,
opening and visible cracks around the circumference of the tube within the tube support plato intersection were visually examined and their location in relatien to the burst crack noted.
(When the crack networks woro particularly complex, such as when circumferential components were strongly present, photographs of the crack networks woro taken and included in this report for more complMo documentation of the data.) The major burst crack was then opened for fractographic observations including crack surfaco morphology, crack longth, and crack depth using SEM. One metallographic cross section of each tube specimen was selected containing the majority of seccridary cracks within the tubo support plato region. The location of the cracks within this metal!ographic cross section was noted, the cracks measured as to their depth and a crack was photographed to show the typical crack morphology. Note that the one metallographic section through each specimon will provide the secondary crack distribution at that location. Secondary cracks at other olevations would not be recorded unless the burst test happened tc open the secondary cracks sufficiently for visual examination to record their location.
4.4.3 Destructivo Examination Results 11Lb21&1 The crevice region of tube 5901 rhowed only three axial cracks, two of which wero through wall. The longest of the two through-wall cracks caused the burst opening. At the tube ,
burst opening, the macrocrack (composed of ono microcrack) was 0.275 inch long at the OD and 0.21 inch long at the ID. The crack morphology was !GSCC. A metallographic cross section capturing the three axial cracks is shown by a sketch in Figure 4-3. The crack morphology is 44 l
i h
shown in a photomicrograph in this figuro. The shapo of the burst crack and 11s morphology is !
described in Figure 4 4 together with the 00 crack distribution found in this tubo.
- I 3 lubfL5912 l
Tao crovico region of tuoo 500 2 had largo numbers of axial and circumfotontial cracks. Tho '
cracking was concentrated on ono quadrant of tho tubo's circumferenco. Photogre9hs of the tubo ;
following burst testing ato shown in Figuros 4 5 and 4 6. The burst fracture occurrod in a :
highly irropular fashion dictated by tho axlal and circumforential tubo degrada:lon. The burst opening was formed by at least five small cracks which joined partial drcumforential cracks to i form the irregular overall crack pattorn. Tho macrocrack longth duo to corrosion measured !
0.38 inch at the OD surfaco and it was through wall for 0.30 inch. The microcraAs and their t ligaments had intorgranular ligaments and the morphology of tho burst crack was that of IGSCC l 1 (Figure 4 7). A metallographic cross section through the region with the highest crack donsity ,
showed a crack distribution as sketchod in Figuro 4 8. A photomicrograph of two typical -
r socondary cracks is also shown in this figuro. They suggest that the cracking is primarily IGSCC with somo IGA contnoutions. Figure 4 9 providos a summary of tho overall crack distribution and summary information rogarding the burst crack.
Tube 500-3 l l
Rupturo in tube 590 3 occurrod from a single axial OD origin crack confined to tho erovico '
region. The macrocrack was 0.31 inch long and was through wall for a longth of 0.27 inch.
Only one microcrack could obviously be observed on the macrocrack Its morphology was that of IGSCC. Figuro 410 providos summary data regarding the corrosion observed on tubo 590-3. ,
Iube 511:.1 -
Burst in tubo 591 1 occurrod from a singlo, relatively small axial crack which was 0.24 inch long on the OD and 0.18 inch long on the ID. While two small axial secondary cracks woro e observed away from tho burst near the bottom of the crevico region, no secondary cracks wero -
observed near the burst opaning. However, a metallographic cross section through the contor of ;
the burst opening revealed two additional axial secondary cracks which woro located away from >
the burst. The location of thoso cracks in relationship to tho burst opening is indicated by a ,
sketch in Figure 4-11 A photomicrograph of one of the secondary cracl,s is also shown. All cracks had a morphology of IGSCC. Tho shapo of the main crack and tho distribution of cracks aro -
depicted in Figuro 412. ;
Tube 591-2 I Tho berst fracturo in tubo 591-2 occurred in an area of the crevico region whoro many sma3 !
! but doop axial cracks woro concentrated. The burst croated a macrocrack which was 0.21 inch !
I long on the OD and it was formed by four smaller microcracks. The crack was through-wall for a
- I longth of 0.03 inch. The ligaments forming the macrocrack all had ductilo foaluros and tho i morphology of the cracking was that of IGSCC. Figure 413 shows the crack distribution ,
observed by metallography in a circumferentialcut through the lower region of the crevico- !
.. where the crack density was hl0 host. A photomicrograph of one of the cracks is also shown. A :
sketch describing the shapo of the burst macrocrack, as well as the overall distribution of secondary cracks within the crovico region as observo6 by visual examination, is shown in .
Figure 414.
45 ,
we-+---+--,- , .-em6e,s--v.-
_ _ _ _ _ .. . m. _ _ _ _ _ _ _ _ _ . . _ . - _ _ _ _ _ . _ _ . _ _ _ _ - - . .
i Tubo 591-4
, A group of small, doop, OD origin, axlat cracks, conc ( atrated in one region of tube 591-4 within !
the crevice region, caused the burst fracture. The irregular shape of the burst opening (Figure !
4-15) was formed by five small microcracks which grow together by intergranular corrosion
- to form the macrocrack. The morphology of the macrocrack was that of !GSCC. The macrocrack !
crack was 0.45 inch long and through wall for 0.35 Inch. A metallographic cross section ,
through the conter of the burst crack revealed many secondary cracks of considerable depth. The cracks are depicted by a sketch in Figure 416 together with a photomicrograph of the cracking. ,
Summary data regarding the burst crack and the overall crack distribution are shown in Figure t 4 17. '
Tubo $98 3 The burst fracture in tube 596-3 occurred from a group of small axlat cracks of OD origin.
Four of the deep microcracks jolned together during the burst test to form the burst opening l macrocrack. The ligaments betwoon the microcracks had only intergranular features and the crack morphoiogy was that of IGSCC. The macrocrack caused by IGSCC was 0.45 inch long on the OD and was through wall for a length of 0.44 inch. A metallographic cross section through the region with the highest density of cracking revealed a crack distribution showr' by a sketch in Figure 418. A photomicrograph in this figure shows the crack morphology of two of the secondary cracks. A summary of the burst crack data and of the overall crack distr!bution within the crevice region is shcwn in Figuro 419, 4.4.4 Comparison with Pulled Tube Crack Morphology i
Most of the support plate cracking on pulled steam generator tubes was OD origin, intergranular stress corrosion cracking that was axially orientated. Large macrccracks were frequent'y j present and were composed of numerous short microcracks (typically < 0.1 inchos long) '
separated by lodges or ligaments. The ledges could have either intergranular or dimple rupture i features depending on whether or not the microcracks had grown together during plant ocoration. l Most cracks had minimal to moderate IGA teatures (minor to modarate DM ratios) in addition to the overall stress corrosion features. Even when the IGA was present in association with the cracks in significant amounts, it did not dominate over the overall SCC morphology. The numbers of cracks distributed around the circumforence at a given elevation within the crevice region varied from a few cracks to typically less than 100. In a few cases, the number of cracks was.significantly larger than this,in one caso possibly approaching 500. For this situation, "
patches of IGA formed where the cracks were particularly close and the individual cracks had some IGA characteristics. Even for this situation, the axial SCC was still the dominant corrosion morphology Ps the IGA was typically one third to one half the depth of the IGSCC, in addition, cellular IGA / SCC was occasionslly observed confined to small areas within the crevice region.
Finally, IGA, separate and independent of SCC, has been observed, it is usually prosent as small isolated patches of IGA. In the few cases where more uniform IGA has been observed, it is typically shallow and intermittently distributed within support plate crevice regions.
The model bollor corrosion observed in this investigation was similar to that observed within typical pulled tube support plate crevice locations. Most corrosion was axially orientated IGSCC -
with negligible to modarate IGA aspects (minor to moderate DM ratios) in association with the cracking. Some of the model boiler specimens had cracking with almost pure IGSCC.'i.e., with no obvious lGA aspects (D/W ratios of 50 or higher), more similar to PWSCC than to the typical OD !
IGSCC obterved within support plate crevice corrosion on pulled tubes. IGA independent of the cracking was not observed in the model boiler specimens. The numbers of cracks at a given 46
i elevation was typleally loss than 20, similar to that observed in many of the pul!ed tubes. l However, only one model boiler specimen had a moderato crack density and none had high crack .;
densities as have boon occasionally observed in plants. A number of the model boiler specimens i from the second set of tests conducted in 1991, howesor, did have very complex crack networks :
that frequently had circumfoecntial cracking in associal on with the predominant axial cracking, . I 4
Some of the complex crack networks may have had cellular IGA / SCC components similar to that occasionally observed in pulled tubos, e
4.4.5 Conclusions from Specimen Destructivo Examinations It is concluded that the laboratory generated corrosion cracks have the same basic features as :
support plate erevice corrosion from pulled tubes, The laboratory created specimens frequently had somewhat lower crack densitics, but individual cracks usually had similar IGA aspects (minor to moderato D/W ratios), IGA independent of IGSCC was not observed in the model boiler .
specimens as was sometimos observed in pulled tubos The observed ditforences in corrosion i morphology between the model boiler specimens and the pulled tubos is not believed to be significant.
-l 4.5 Model Boiler Databaso Summary i
As described in the above subsections, model boiler specimens have boon fabricated and tested to augment the pulled tubo database at support plato intersections. 53 laboratory specimens have ;
been prepared using 3/4 inch OD tubing. Tho specimens woro subjected to oddy currors' 1 examination. Degradation at simulated tube support plato intersections have ranged from NDD to 65 volts in bobbin coil amplitudo. All of theso specimens have been burst tested, with the reLits displayed in Tablo 4 5. Specimens with significant degradation (41) have also been leak
, testod. Further, several of the samplos were destructively examined to determino degradation characterists and crack morphology The currently available maximum and through wall crack longth data obtained for many of those specimens from the destructivo examinations are listed in Table 4 5 The model bolMr databaso is combined with the pulled tubo databbse and the total used for dolormining leak rato and burst correlations. l
)
l v
g I 47 [
Tablo 41 Model Doller Thermal and Hydraulle Specifications Primary loop temperature 327*C (620*F)
Primary loop pressure 13.8 MPa (2000 psi)
Primary boiler inlet temperature 324'C i 3*C (615*F 15'F)
Primary boiler outlet temperaturo 313*C 3*C (595'F 15'F)
Secondary Tsat at 6.1 MPa (900 psi) 278'C 13*C (532*F 2 n**
~
Steam biced 0.1 0.2 t' day Blowdown 8cm3/ min (continuous)
Nominal heat flux 16.28 x 104 kcal/m2 hr (60,000 Blu/ft2.hr) 8 h
4 48
-l
Tablo 4 2 Chemical and Physical Proportios of TubinD Material (NX7368)
Chemical Composition (Weight %)
Ni 76.21 Cr 14.87 Fo 7.98 C 0.04 Mn 0.41 Si 0.30 Cb 0.15 Q) 0.04 Physical Proporties Ultimato Strongth (KSI) 109.4 (744 mPa)
Yield Sirongth (KSI) 54.2 (300 mPa)
% Elongation 37.0 Hardness 83. (Rockwell D) l l
I 4-9
\
\
. - .- . .... - - - - . _ - . . - . . - . _ - . - , ~ -- .-- .- ., . _ - .
1 l
Tablo 4 3 Test Speelmen Summary Specimen Crovice Days in
& ,. Groun Confiouration Test ,
590-1 EPRI Frit 8 500-2 EPRI Frit 15 590 3 EPRI Frit 15 590-4 EPRI Frit 19 591 1 EPRI Frit 8 591 2 EPRI Frit 10 501 3 EPRI Frit 21 591-4 EPRI Frit 10 592 1 EPRI Moch. Cons. 138 :
592 2 EPRI Moch. Cons. 138 592-3 EPRI Moch. Cons. 138 592-4 EPFl Moch. Cons. 138 i 592 5 EPFl Moch. Cons. 138 592 6 EPFl Moch, Cons. 138 592-7 EPRI Moch. Cons. 138 593 1 EPRI Dual Cons. 133 593 2 EPRI Dual Cons. 133 593 3 EPRI Dual Cons. 133 **
593 4 EPRI Dual Cons. 133 594 1 EPRI Dual Cons. 85 .
595 1 EPRI Dual Cons. 34 595 2- EPRI Dual Cons. 84 595 3 EPRI Dual Cons. 84 505 4 EPRI Dual Cons. 113 :
596 1 EPRI Dual Cos. 46 596 2 EPRI Dual Cons. 10 596-3 EPRI Dual Cons. 5 596-4 EPRI Dual Cons. 48 '
597 1 EPRI Dual Cons. 133 597 2 EPRI Dual Cons. 133 597 3 EPRI Dual Cons. 133 597-4 EPRI Dual Cons. 133 593 1 EPRI Moch, Cons. 27 598 2 EPRI Moch Cons. 27 6 598-3 EPRI Moch. Cons. 27 598-4 EPRI Moch. Cons. 43 603 1 EPRI Frit 34 603 2 EPRI Frit 34 603 3 EPRI Fril 34 -
F 603-4 EPRI Frit 34 f'
(Continued on next page) 4 10 l
l Tablo 4 3 (Continued)
Test Specimen Summary Specimen Crevice Days in
. A. G!TLl0 C2.n11',t!2d!2n Tett 004 1 EPRI Frit 14 604 2 EPRI Frit 7 604 3 EPRI Fril 22 604 4 EPRI Frit 22 600 1 Spanish Dual Cons. 10 000 2 Spanish Dual Cona. 14 000-3 Spanish Dual Cons. 3ti ;
601 1 Spanish Frit 12 001-2 Spanish Frit i2 601 3 Spanish Frit 17 001 4 Spanish Frit 17 601-5 Spanish Frit 17 601 6 Spanish Frit 17 e
i P
e 4 11
.I i
a Table 4-4 Composition of Sludge Used for Crevice Packing i
Weight %
Simulated .: '
Plant Spanish Conttituent Studg2_ E!ud2A -
Magns :10 59.7 92.2 ;
Copper 32.5 Cupnc Oxido 4.5 4.5 Nick 01 Oxido 2,1 2.1 i Chromic Oxido 1.2 1.2 Q
O
[
t i
I 4 12
____ _____- ~ __...__._.-________.,.__.__..m.m__ . - _ _ _ _ _ _ _
Tablo 4 5 :
1 Leak Flato & Burst Test ilosults for 3/4 Inch OD Laboratory Specimens l Preliminary Bobbin Durst Dostructivo Exam, 1 Amplitudo __Lf 9hJlate it'hrL. Pfossuro . . . . . Longilljnch)
No. SpecitucIl (vnitsL N.O.AP SLDAP _.jpidL Maximum JJuur J
. 9 l 1
i i
i 4
l t
0 l
f i
t i
a (Continued on next pago)
. }
When crack is not throuG)I wall, inaximum depth of penoiration is shown. _
L 4-13 s
y-* <r+-- e, ,yg,=er-,ex.-- ---w-r--,,-g .m.w,,m,,,,m.,..--,,r-,,,,-c-,v -y,e--e .r ., -rwe :,. w, ow. _- my+ , , :.,e,e- + - . . ,+.r-..-.,-- e. - - , , , , r . . - ~ . .e---.---+---cM
Table 4 5 (Cortinued)
Leak Rate & Burst Test Results for 3/4 inch OD Laboratory Specimens Preliminary Bobbin - Burst Destructive Exam. .
Amplitudo Lenk Rate (l'hri Pressuto Lenoth (Inch) th. Soecimen _f.YQltil_ N.O.AP SLB AP fosi) Maximum Thruwall' 0
t When crack is not throughwall, maximum depth of penetration is shown. -
l l
l l.
I-4-14 i
L._--___.--------------- -- --
~~' - " ~ ~
CONDENSER n
REFERENCE COLUMN l
l PRES $URE I
switch l I
l CONDENSATE l TANK
~~
rnOM "A" LOOP !
TO A" LOOP T U
~~
I l
/1J l '
- l p - _ J lll l 1 L _ __ _ _I -
f e .
l W- PRES 5URE r-- -
l SWITCH +
l*
v0io ! I MAKEUP oETECTORS TANK i SAMPLE \
l l
L. _ _ _ I l
@\ -
I p AP I I
l j
CELL .___J 1 i i l - j l
L_JI ,
I I h Y '
l j
l D I t s MPa ,________._._________a b
U l
Figure 4-1. Schematic of Vertically Mounted Single Tube Model Boiler l
l 4 15 1
l l
W c0NDtNstM ntFtAENCE IL T co'U"" .
sauptt enessunt r- --
switch l T I e! I
, l k l l P l
l
/.1_J l coNotNsatt
[ OnlGINAL l
l T^**
MODEL N 8 BOILER " Y l
I r-- ' entasunt l switch VOID i otttcions 8 Wawtur 1 --
-g TANA I ,
l I N i
L__ _ l N l l
- I l
i Oar I l
3 7 g ,, _
ctLL - - -- - J -
uNt lg ii j
i q L_Jj g
I .
I l
L. ...$^"" - - - - . - - - - - - - - - - - - - -
Dl 3
HORIZONTAL
/ MODEL DOILER
_1 1 /
,[, y RECIRCuLAtlON a y uNe f- _ _ d
'l' I M
L__. J ,,,,,, nn.,u PRIMARY sf sit M Figuro 4 2. Schematic of Horizontally Mounted Single Tubo Model Boiter 4 16
100%
100%
60%(A) i 00 l.
,i ---
a kk , s. , ,
,~ 4 , z,
^
, 1 .
92 N 4 r . l Ii a ,
i
.)
~
n%
Crack A Figure 4-3. Sketch of a metallographic cross section through the crevice region of tube l' ' 590-1. The burst crack and two secondary cracks were observed. A photomicrograph of a secondary crtck is also shown. The crack morphology is that of IGSCC. Mag.100X 4 17
00 ,
x\ s -
i s x 10 - A i
Sketch of Burst Craqi Macrocrack Length . 0.275 ;nch Throughwall Length . 0.21 inch Number of Microcracks - 3 Morphology = IGSCC 0.75 inches - - SP top 0.6 inches -
i 0.2 inches -
0.0 inches - - SP bottom 180 0 2700 00 90 0 1R00
~
Sketch of Crack Distribution Figure 4-4. Summary of burst crack observation and the overali arack distribution at the crevice region of tube 590-1.
4 18
)
1, 1
)
p.
i
' n
.. > . k i.
~
4 j ,,
- l -
4 , s (A k ,
-- ~
- s , .
- .1 s ' .
b e -- d" es U b
. . > - m,%
I
{ 44 t
' .. L a . a u ' , +, ' W'TT~r 4 ,,
a ~r ,
% . s i
- , %, ,% j
- n. .. .
> . s ' ;, .
, ...s - , , ,,,,a _ . .$ .
s s, s . , . .
( ,3
. " ;'. 8, . ' , !
. r..
Fi 8 W
..Y . . . $ * $O .. ,
a '% st % . . I
- . " ' M' - j .
4 ., l ,
)
. 3 c. ,1 ~
s' 45 l
F
[i Fig are 45. Photographs of the burst opening in tutio 590-2 showin0 arial and circumferential cracxing. I
?
4-19 f
?
t
& mm m mmw+em.w ww~w enm-e w m . w mne. ., _wwmem.=e-~~~~---we-- .e,- . - . ~ . - , ~ ~ ,
' ~
~.
- =a,,
o # 4t" '
s , .
' ~
.. % . \e y e e,
. ~
ca \,
,.i .J.,.. s v +. i
, .. . .. ..,,s.,. , < . ..,, . -
. , - .-),, .
~
, . .~ ' ' * .. C ; 't,1* I c.. y-s ..
., g'-;.4.tv , e. - -
e 4x W
., .. ,- /, .
4 * - ~p ~4
/.dr.
- r. -
,( - ',
c r . - . ,; ', -
g P', . '
^
e -
4 w
'^ -
!. . . = %. r s- ' :s J
- ~ ~ f.. , ed ,, . .. .; ...
- s. . .
- r. 3, h '
.~
a,.
.] _
.j *
..i, e,,
,r gy9 a d . . J ' .,
- y. Q e.-t,
-~4
. 3- ,
4,. f' ' ' , ,
o ,dMf
- k',g'g) t
. *;. , .. .is
- .. . , .. , ,}. 4, .-- y.. o
.n. ,
h' ,, '
'* r., . * , m' .
r
- 4
+r,.-,. 4 ,-, a s .
J . % ', , f , ' ' ~
- - 1 .,** ,
- *,..r *
- f. . . ... ' - s, n- .t A ,. ... .
- p. g y. 1 *a z , . s . V . . . y . . .&.
10X o .
7
< Figure 4 6. Photographs away from the burst opening in tube 590 2 showing axial and circumferential cracking.
a 4 20 l
l
)
i J
l l !
1 l 4
4 ;
i l
- k i
i 4 '
i
! I e I 1
4 d
i t
o 1
I ,
} i 1
,l .
I
! +
l l
il i
\s' l r
i i
- o i
I
' i I
I i
i00F I
i l
i i
l i
[
l Figure 4-7. Photomicrograph of the burst erack in tubo 590 2 showing IGSCC morphology.
i 4 21 t L._,__.__,. ,, . - - - - - - - , . - - - - - - ~ ~ . - - - - - - - - - - - - - - - - - - '- --~~"--~ ~ -~~~* ~ ~'~ ~ "'~ """
l 607. 505 1007.
707 p 107 60Y. w (A) .c 00
.... . .w
.: . . . . . w. ; - .c:. c . ~v, u -
.c .. .
..,.i-- . ..
.. a.
r.-
, .s a , j
- +
. . .4 .,s.:
> + * -
- . , . s,
.3 . ; :. . .
~ .n . .
. '. .. . ,: . t r
~,' - ' . .
.,..g.- , , .%..
3 ,. . . . .. -
. . . . . .m.s w.c . ,
y,. ..~...,.a.
3 . .
.~. . . .:. :.,.
mn....
.~ ..
. ,/m:.. :. . . e. <. , n.~~m.
- y. c. . g
~
~ : ..- ,,;.. .;. ;-
s: . .....
. a y u 2 c :
. ,5 . p . ..
,~ q~: .. .: ..m: . ; > ,..
r .. n. .u. =: . v..
x.
- +. ,.. ...:,,.s; a.. , ..-.,.
.- .. ...v. ..
,: ..q.,,.
h5f'gy k[aMkbl' k Nb. , .=.. s .db -
~
- 8bI D Crack A Figure 4 8. Sketch of a metallographic cross section through the crevice regbn of tube 590 2. The burst crack and a number of secondary cracks were observed. .
A photomicrograph of two secondary cracks is also shown. The crack morphology is that of IGSCC with some IGA contribution. Mag.100X 4 22
. _ - _ __ _ _ _ _ . _ _. _ .._____.. _ _ _ _ _ _ _ _ m _ . _ . _ _ . . . _~ _. . _ . .
I y \ )g 3 10 -
Sketch of Burst Crack Macrocrack length = 0.38 inch ,
Throughwall Length = 0.30 inch Number of Microcracks - 5 (ligaments have intergranular '
features)
Morphology = IGSCC 0.75 inches - -
SP top 0.6 inches - 4) !}
P 0.2 inches -
0.0 inches - - SP bottom l 180 0
2700 00 90 0 180 0 Sketch of Crack Distribution F! cure 4 9. Surnmary of burst crack observations and the overall crack distribution at l the crevice region of tube 590 2, l
4 23 l
's \
\ ,,
ID Sketch of Burst Crack Macrocrack Length . 0.31 inch Throughwall Length - 0.27 inch Number of Microcracks - 1 Morphology - IGSCC 0.75 inches - - SP top 0.6 inches - ,
f 0.2 inches -
0.0 inches - - SP bottom 180 0
270 0 0 0
90 0 1800 Sketch of Crack Distribution Figure 410. Summary of burst crack observations and the overall crack distribution at the crovice region of tube 590 3.
44 24
100?,
787,(A) 607, ~
r .
00 t_ ,
2
'4 D ,by ,..
u . g; a 3 -
C, 4 s'
? ?
/.. ..
F. -
, , t . H:
? ..
~!.
Crack A Figure 4-11. Sketch of a metallographic cross section through the crevice region of tube
. 591-1. The burst crack and two secoiidary cracks on one quarter of the circumference were observed. A photomicrograph of a secondary crack is also shown. The crack morphology is that of IGSCC Mag.100X 4 25
CD 10 Sketch of Burst Cract Macrocrack length = 0'.24 inch Throughwall Length = 0.18 inch Number of Microcracks - 1 Morphology - IGSCC 0.75 inches - -
SP top I
i'i 0.6 inches - '!
0.2 inches -
.\
t 0.0 inches - -
SP bottom 0 0 00 900 180 270 1800 Sketch of Crack Distribu110D Figure 412. Summary of burst crack observations and the overall crack distribution at the crevice region of tube 591-1.
4 26 l _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .. . _ _ ___
~ . - _ - - . . _ - . . .- -. . . -. - - . - - ._ . . -
100%
I t}%.. .
100%
- 1
\ l I
l s
Ph% %
/ti 80% (A) 85*.
.- t* .
r' l .
4' q.
, h, a v
. .q ,,
. L. .
3 i a . -
\,7 $ f ,
d5'l t. , % '. '
. .QW;;;. t,,: v>... .i. ,:
. '.'a %....y,.?: ..
Crack A Figure 413. Sketch of a motatlographic cross section through the crevice region of tube 5912. The burst crack and a number of secondary cracks around the circumference wore observed. A photomicrograph of two secondary cracks is also shown. The crack morphology is that of !GSCC. Mag.100X 4 27
00 10 ,
Sketch of Burst Crack Macrocrack Length - 0.21 inch Throughwall Length . 0.03 inch Number of Microcracks - 4 (ligaments have ductile features)
Morphology - IGSCC 0.75 inches .. - SP top .
\
)
0.6 inches -
1 0.2 inches -
t
)
0.0 inches - li 4 - SP bottom 0
180 2700 00 90 0 1800 Sketch of Crack Distribution Figure 414. Summary of burst crack observations and the overall crack distribution at the crevice region of tubo 5912.
4 28
b i.
(
l l
i..g4k$
. .. .c .. :g
- ,.c .: . .
l
- i I
I q
! i l
~', . , . ,
~ ... , .,.,...
. s
.,e
~
n,.. _.
g** ....giy. s;
. .s .. .
, . . , [
\
- 2. . . .s. . ,r -
s s
~
i 1
1 t
t l
r.
1 i, . ~
l E.:3? ?L_; . . .h dfMiiGS$
i 2
' o
- L
, e
- - r i
4 .
l l Figure 415. Photographs of the burst opening in tube 5914.
1 3
4
! 4-29 I
i c n
W-.* --w. - -W'egvt= < e w $* e -w,-----,-w--e-e--,- - - - - - , - - , , - - - - - , --
. - _- - _ _ = - . . .. - . . - . - _ _- . - _ -
100%
5 85% (A) 95%
95%
i 90%
00 .
e<
, an. . ,, r ,
s -
, j7
, ,, .y :. ~ -,-
~
- q. . . , . .
., t _. , . .) .
- ,. . , ,-u
a
.,3,y
~- , .,
.,, ,, * . ; s. ;.
- - [~ /, s r -
,,.t,*} - k+.T7.h
..- y .
Crack A Figure 4-16. Sketch of a metallographic cross section through the crevice region of tube 591-4. The burst crack and a number of secondary cracks around the circumference were observed. A photomicrograph of the burst crack and a ,
secondary crack is also =hown. The crack morphology is that of IGSCC.
Mag.100X 4 30
00 -
10 Sketch of Burst Crack Macrocrack length = 0.45 inch Throughwall length - 0.35 inch Number of Microcracks - 5 (ligaments.have intergranular features)
Morphology - IGSCC 0.75 inches - -
SP top 0.6 inches - l , 0.2 inches - 7 0.0 inches -
} kI l - SP bottom 270 0 00 0
180 90 0 1800 Sketch of Crack Distribution Figure 4-17. Summary of burst crack observations and the overall crack distribution at the crevice region of tube 591-4. 4-31 I l
100% 30% 30% (A) . 100% 9
* ' q 3. .,.yp: .s .. CD
.. n. .1 .
u,,.,?3..,~.;
.. a .,.,..: c ..- .~
. .m - '.
. .-.>',;, . . % nw .t ~~
- b4 u s -Qa c .. . , . .
G
, ? qc 'q. . ' C ;r7,.k- iRQ
?. t. . {-
.. p ._ r .
jf; . 2
,7;..' ) ,.
., -s . ,.t.-
- a. ~ :-
* :.r.~.,).. *
,'A.'7J.',.
,, ,, 7, ,.'g.* .,
u . r-.
, * ,s,-
.,s . . g.
.. 9
., .;t -.~
-,1._. x.p. . ., e. . ;g w : mg.;.
.y,.p.
, y J. _ r
-. . - . .. i . . .1, W,h.e
.Y.;;'
, . . s ..,
+
s sc . &m.?- s. 5: ,, .
. < ..: s .c.;,.t w
. . . , ...a .s. s.. W :>,
t
;. c
-s ., . .. ;.. . ,
.a - . .. 44
.w c., 4 _ a 5, n : ;
'[. .., Y. 4
- l. s r . q*%' h y.- ..Q 3 *.,'
y * '. -
. . .v ,
.e: .r e C'h$il* f', _. .f
~
. ', . , N.*. i'"Rk.A$5ffh
.s o v-~.x%
- - e: <
..tt..v..=..
Mg..AWf
, m 'i'<
S -. g . n ..h a -
.. 3[7,' 4 ~S 0-x , ,S : +w dcw Crack A Figure 4-18. Sketch of a metallographic cross section through the crevice region of tube 596-3. The burst crack and a number of secondary cracks in one quadrant ,
of the circumference were obser,ad. A photomicrograph of two secondary cracks is also shown. The crack morphology is that of IGSCC. Mag.100X 4-32
\' \ \' \ '
x\s x
\
N ID '\ Ske*ch of Burst Crac.k Macrocrack Length . 0.45 inch Throughwall length - 0.44 inch Number of Microcracks = 4-(ligaments have intergranular features) Morphology = IGSCC
, 0.75 inches - -
SP top 0.6 inches - i g i 0.2 inches - 't I l 0.0 inches - - SP bottom 180 0 270 0 00 90 0 1800 i Sketch of Crack Distribution ' ~ Figure 4-19. Summary of burst crack observations and the overall crack distribution at the crevice region of tube S96-3. 4-33 l
E Section 5 NDE EXAMINATION 5.1 Eddy Current Voltago Normalization for APC Normalization of observed support plate ODSCC signal amplitudes is performed to permit direct comparison of the voltage levels associated with field measurements with the laboratory calibration used to join pulled tube and model boiler signal amplitudes, in cases where field data is collected using different voltage calibrations or different frequencies, conversion factors are developed which permit the fleid data for tubes and ASME standards taken at different voltage normalizations to be integrated into the overali database for voltage, burst pressures and leakage correlation;, The existing data base for pulled tube and model boiler samples includes amplitude measurements which are referenced te a common voltage calibration for both 3/4 inch diameter 0.043 inch wall thickness and 7/8 inch diameter 0.050 inch wall thickness tubing. Specifically the bobbin EC coil reference calibrations for 3/4 inch x 0.043 inch tubing are: 4.00 volts at 550 kHz for 4 X 20% ASME noles, and 2.75 volts in the 550/130 kHz support plate suppression mix output, also for 4 x 20% ASME holes l l The frequency and voltage normalizations applied for the 3/4 inch diameter APC data base, j including all model boiler specimens, were developed based on scaling from the 7/8 inch j diameter practices (400/100 kHz frequency mix). The 3/4 inch and 7/8 inch tube diameters l are geometrically similar, that is, alllinear dimensions are scaled by the same factor. Eddy l current probe dimensions are scaled by the same factor applied to the 0.720 inch,7/8 inch probe diameter to obtain a 0.620 inch,3/4 inch probe diameter. The probe excitation frequency is inversely proportional to the square cf the tubing thickness. This applies to the 550/130 kHz and 400/100 kHz frequencies used for 3/4 inch and 7/8 inch tubing, respectively. Thus the bobbin coil dimensions and test frequencies are intended to yield similar responses for the 3/4 inch and 7/8 inch tubing.- However, the ASME calibration standard holes are not scaled and other probo characteristics such as coil size are not scaled so that the resulting 3/4 inch and 7/8 inch tubing eddy current measurements are not directly comparable. For the above t obbin coil voltage normalization, the probes tested (Echoram, Zetec) in the laboratory yielded both 4.0 volts at 550 kHz and 2.75 volts at 650/130 kHz. However, testing l of other prober with different frequency sensitivity can yield a different ratio between 550 and 550/130 kHz normalizations. For probes yielding the laboratory ratio between 550 and 550/130 kHz, the voltage normalization at 550 kHz to 4.0 volts is preferred as it is less sunsitive to small analyst variations in setting up the mix. However, the voltage normalization to 2.75 volts is more generally applicable to different probes and should be used for probes
, differing from the laboratory ratio by more than about SE That is, if the voltage is normalizaed to 4.0 volts for the 20% ASME hole at 550 kHz and is outside the range (5%) of 2.6 to 2.9 volts when carried over to the 550/130 kHz mix. the bobbin voltage normalization to 2.75 volts for the mix should be used for the data evaluation. For Catawba-1 voltage measurements the 2.75v mix normalization was the basis employed for sizing 'he support plate indications.
5-1 I
. - , l .. E , ~# , , - - , , - , ...,.,r m .,-.m-.
_ . .m ~ _ _ __ __ _ _ Thus the 1992 Catawba-1 support plate amplitude measurements used for repair limits disposition were taken in a fashion consistent with
- Appendix A* guidelines presented in the Farley and Cook APC submittals.
5.2 Eddy Cunent Data Analysis Guidelines The general inspection protocol for bobbin probe EC testing spac;fied it at data be collected at 4 frequencies - 550 kHz,400 kHz,130 kHz and 35 kHz. For Cetawba-1 t Tis represented a change in that prior inspection data did not include 550 kHz and that 100 kHz was used as the quarter-frequency for support plate suppression. As explained in Section 6.5 renormalization factors were calculated to facilitate determination of cycle growth rates . 5.2.1 Ouke Power Company Analysis Process for Support Plate ODSCC , The Catawba Unit 1 EOC 6 bobbin coil data was analyzed in accordance with the Eddy Current Anatvsis Guidelines. Catawba Nuclear Station Unit 1. Rev. 2, dated 7/9/92. All signals indicative of degradation were reported regardless of depth and with no minimum voltage threshold. All data were analyzed with 2 independent reviews (primary and secondary analysts). The results of the primary and secondary analysts were compared and resolved by a team of 2 resolution analysts. Every HL (hot leg) TSP call was then remeasured to obtain a voltage value consistent with Westinghouse recommendation for measuring bobbin voltages for ODSCC degradation at HL TSP's. This process was simply a measurement exercise, to obtain a voltage value related to a specific - normalization, channel (550/130 kHz MIX 5), and signalisolation. This was not a reanalysis as the presence of the degradation at each reported TSP had already been determined and was not changed. The remeasu'ement was performed in accordance with the Hotloo Tube Succort . Re-sizino Analysis Guidelines. Rev. O, dated 8/12/92. An analyst remeasured each HL TSP call and generated bobbin coil graphics depicting the call. Each call was reviewed by a team of 2 resolution analysts who concurred dh the accuracy of the measurement, and assured all HL TSP calls were resized. Every HL TSP call was then remeasured two more times: Once again on the current EOC 6 bobbin data with a 400/130 kHz mix and also from the EOC 51991 data with a 400/100 kHz mix. These remeasuremer.:s were performed to obtain ODSCC growth trending information. The remeasurements were performed in accordance with the Hotleo Tube Succort Re-sizino Anaivsis Guidelines for Growth Trendina, Rev. O, dated 8/12/91 These measurements were also performed by an ana!yst with two resolution analysts reviewing them for accuracy. The guidelines are attached as Appendix A. To obtain growth trending information over 2 cycles, a set of HL TSP calls in tubes plugged after the EOC 5 analysis in 1991 were remeasured using the EOC 51991 data and the EOC 41990 data using a 400/100 kHz mix. Resizing and resolution were performed in identical fashion to the growth trending described in the paragraph above. e 5-2
i 5.3 NDE Results for Model Boiler Specimens The range of signal amplitude vs burst pressure data available from pulled tubes is supplemented
. by data produced from laboratory specimens prepared in model boilers. Table 4-5 presents a summary of NDE data (bobbin voltage) corresponding burst pressures and leak rates measured in the EPRI-sponsored laboratory study. The bobbin voltages were measured in accordance with the analysis guidelines used for the EPRI and consistent with the Appendix A guidelines provided for 7/8 inch tubing as typified by the Farley alternate plugging critoria (WCAP 12872, Rev.
2, February 1992). The 3/4 inch tubing analysis guidelines used in the laboratory as well as the Catawba 1 re sizing analysis guidelines are consistent in the manner and techniques used for amplitude measurements with the Farley submittal guidelines. RPC testing and analysis done to assist in the characterization of the model boiler specimens was also performed in a fashion which assures consistency with the data obtained from 7/8 inch tubes.
=
5.4 Vol. age Trends for EDM Slots In order to anticipate the behavior (bohbin amplitude response) of cracks, EDM slots of varying depth and length were prepared for 3/4 inch tubing. As with the 7/8 inch data for EDM slots, the NDE measurements were made according to the EPRI study guidelines on which the Appendix A guidelines from the Farley submittal were based. For 3/4 inch tubing, the support plate mix (550 kHz/130 kHz) data obtained using a 610 mil bobbin probe were evaluated to determine the peak-to-peak voltage values for each notch. These data are displayed in Figure 5-1. The trends apparent in these data are virtually identical to those collected with a 720 mil probe from the 400 kHz!100 kHz mix channel for 7/8 inch tubing (Figure 5-2). A comparison of similar configurations for 3/4 inch and 7/8 inch tubing is given in Table 5-1 to illustrate the equivalence of the readings for the corresponding support plate mix measurements. 5.5 Frequency Renormalization Based on Calibration Standards Past inspections at Catawba 1 did not employ the 550 kHz and 130 kHz channels for bobbin probe EC inspection. For 1992 the bobbin frequencies used included 550 kHz,400 kHz and 130 kHz; the corresponding frequencies used in prior inspections were 300 kHz,400 kHz and 100 kHz. To permit application of the Catawba-1 pulled tubes (400/100 kHz) and of growth estimates for prior cycles, it was necessary to develop conversion factors to translate 400/100 kHz and 400/130 kHz mix channel amplitudes for prior cycles into voltages comparable to the 550/130 kHz calibration basis for 1992. For voltage differences between alternate normalizations and frequenc:es, bobbin voltage conversica factors can be obtained using machined calibration standards such as ASME Standards. Table 5-2 provides voltage values for the various frequency mixes based on an ASME standard, including the values for the APC voltage normalization at the 20% ASME hole for 550/130 kHz. Figure 5-3 demonstrates the excellent correlation obtained between the 550/130 kHz mix and the 400/100 kHz mix as obtained from the Catawba-1 pulled tube in the post-pull laboratory
. NDE.
Figure 5-4 presents a similar correlation as obtained from field measurements between 550/130 kHz and the 400/130 mix based on 1992 Catawba-1 inspection results. It is seen that 550/130 kHz mix correlations with both the 400/100 kHz mix used in prior inspe:tions and the 400/130 kHz mix used in 1992 are very good. Therefore compensation for the 5-3 . l
- .- , .~ - - - - - ~ . -- - ~ - - - - - - -. -
differences between 400/100 kHz and 400/130 kHz calibrations is obtained by multiplying the . voltage value from the solution of the simultaneous equations relating each of them to the 550/130 kHz mix: Volts (400/100 kHz) = 0.94 x Volts (400/130 kHz) . 0.17 (51)- The compensation developed in this fashion permits determination of per cycle growth estimates for the support plate ODSCC indications identified at the end of each cycle. 5.6 Renormalization of Catawba-1 Pulled Tube Data Tubes which have been pulled from Catawba 1 steam generators in prior inspections were field examined using the 400/100 kHz mix, with the bobbin probe, as previously stated, calibrated on the basis of a carbon steel support simulator (ring) on an ASME standard tube yielding 5.0 volts at 400 kHz. To include this information in the 3/4 inch tubing database, the field EC data was recalibrated to the 2.75 volt APC normalization for the 4 X 20% hole on the ASME standard, in addition, the post-pull data provided by B&W on Catawba 1 tubes was used to develop renormalization ratios from the field to the APC normalization. The field and post-pull laturatory data were taken on a 400/100 kHz mix calibrated to a support plate ring, as described above. Post pulllaboratory data were also obtained for 550/130 kHz mix with the APC normalization at 2.75 volts. The post-pull voltages are much higher than pre-pull voltages and thus are not -- used to support the APC development. However, the post pull data are used to develop the conversion factors for renormalizing the 400/100 kHz field data to the 550/130 kHz normalization. Using the correlation of the 550/130 kHz mix to the 400/100 kHz mix when both evaluations are independently normalized to 2.75 volts for the 20% ASME hole (Figure 5-3), one obtains: APC volts (550/130 kHz) - 1.094*(400/100 kHz volts) + 0.143 (52) The pre-pull Catawba 1 voltages were converted to the APC normalization using this equation. For this voltage normalization, the standard TSP volts at 400 kHz were also obtained to permit adjustment of the field data to a normalization of 2.75 volts for the 400/100 kHz mix. The measured TSP volts for the 2.75 volt normalization are given in Table 5-3. Division of these TSP voltage measurements by the field normalization of 5.0 volts yields the voltage adjustment factor given in Table 5 3 for obtaining the 20% ASME hole normalization (2.75 volts for 400/100 kHz mix).- This adjustment factor is applied to the field evaluation with TSP normalization as shown in the field evaluation columns of Table 5-4 to obtain the field voltages for the 400/100 kHz mix normalized to 2.75 volts for the 20% ASME hole. The Westinghouse - evaluation for the 400/100 kHz mix is also shown in Table 5-4. The agreement is generally better than 15% between the field and Westinghouse evaluations. Table 5-3 also shows B&W post pull bobbin voltage evaluations for the 400/100 kHz and for the APC 550/130 kHz mix normalized to 2.75 volts for the 20% ASME hole. These voltages were used in Figure 5-3 to obtain voltage renormalization factors as given by the above equation. The voltage renormalization factors were then applied to the pre-pJil 400/100 kHz voltages of Table 5 4 to obtain the APC normalization voltages also given in Table 5-4. The Westinghouse 5-4
evaluated voltages are used for the APC development although differences from the field : evaluation are small. Also shown in Table 5-4 are the Westinghouse evaluated RPC voltages based on evaluatior; of the available field data at 300 kHz with normalization to 20 volts for a 0.5 . inch long EDM notch, The field RPC voltages were normalized to 10 volts for the ASME holes and are not directly comparable to the APC voltage normalization. Comparison of the pre pull voltages of Table 5-4 with the post-pull voltages oi Table 5 3 shows that the bobbin voltages for the larger voltage indications increased by factors of 1.5 to 4 as a result of the tube pull operations. The largest four voltage indications, which show increases of factors of - 2.4 to - 4.4, are associated with the lowest four burst pressures for the Catawba-1 pulled tubes. 5.7 Renormalization of Belgian Pulled Tube Data The 3/4" tube database is significantly expanded by inclusion of tube pull and burst test data produced by Laboroloc from Plant E in Belgium. In support of the iridustry effort to develop alternate plugging criteria for support plate ODSCC, Laborelee has collected field data using both Belgian and APC voltage calibrations on U.S. testing equipment (MlZ-18) as well as Belgian equipment; this data has included several pulled tubcs among ~57 indications evaluated. The pulled tube data are summarized in Tab!e 5 5; Figure 5-5 presents the relationship as reported by Laborelec between the 300 kHz calibration used in Belgium (4 x 100% 49 mil holes - 2.00 volts) and the APC calibration for 3/4" tubing using Belgian test equipment for 300 kHz data and U.S. oquipment for the 550/130 kHz (APC) data. The correlation between Westinghouse and Laborelec evaluations at the APC normalization is shown in Figure 5-61 excellent agrooment is shown for the 550/130 kHz data evaluation with U.S. test equipment (MlZ 18). Similarly an excellent correlation is obtained (see Figure 5-7) for the Laborelec evaluation of data obtained at 300 kHz and 550/130 kHz using the MlZ-18 equipmcnt. These data permitted development of the voltage renormalization factor (~4.93) based on Laborelec probes and calibration standards. To merge the Belgian data into a consistent iadastry population for 3/4" tubing, a cross-calibration of the Belgian analyses w;;h the U.S. la'ooratory ASME standard for 3/4" tubing (AS-009-91) was performad. The cross-calibration process began with the testing of an ASME transfer standard (ASR 002-92) together wi".h the laboratory standard. For the four - 20% holes used as the reference basis, the following results were obtained: Lab. Std. Transfer Std. U.S. Frecuenev AS 009 91 ASR-002-92 Lab / Transfer 550 kHz 4.00 volts 3.96 volts 1.010 550/130 Mix 2.85 volts 2.78 volts 1.026 The U.S. trantfor standard was provided to Laborelec to obtain the cross-calibration ratio for the Belgian ASME standard (ft62952). These two standards were tested with Belgian probes with a MlZ-18 (Zetec) EC tester to obtain 550 kHz and 130 kHz data; the following results were
, obtained for the four 20% holes:
U.S. Std. Belgian Std. Frncuency AS 002-92 #62952 Belaian/U.S. Transfer 550 kHz 2.51 volts 4.00 volts 1.594 550/130 Mix 1.56 volts 2.75 volts 1.763 55
l l I Using this data,it is possible to complete the cross-calibration of the U.S. and Belgian data sets. Amplitudes measured from the 550 kHz/130 kHz mix channelin Belgium must be multiplied by 1.763 (Belglan/U.S. Transfer) x 1.026 (U.S. Lab /U.S. Transfer) to convert them to the ampli'ud9 scale of the U.S. data base: - Belgian Volts x 1.763 x 1.026 - U.S. Volts U.S. Volts - Belgian Volts x 1.809 This cross-calibration, obtained using Belgian probes with a U.S. EC tester, yields a ratio quite consistent with the ratio predicted from the U.S. work, summarized in Table 5-6. Laborelec is continuing a study of differences between U.S. and Belgian practices; preliminary results of their work is summarized in Table 5-7. These data suggest that part of the difference observed is found in the calibration bases; there appears to be little difference h responses between the U.S. and Belgian probes and test equipment. For conservatism in this report, the Belgian cross calibration factor has been applied as 1.5 (rather than 1.809) pending completion of the Laborelec study. 5.8 NDE Uncertainties for Catawba Unit 1 5.8.1 General Approach for APC The usual industry practice with respect to NDE uncertainty is based on the adequacy of a sizing - model which relates the measured NDE parameter (e.g. depth from phi.0 angle or amplitude for EC testing) to the true value as determined from metallographic examination of representntive specimens, actual or simulated it has been shown that unique interpretations of amplitude from bcbbin signals are not to be expected and that depth as measured from phase angle is not an adequate predictor of the structural capability of a tube. The need to relate measured NDE paiameters to structural adequacy has resu;ted in the subject amplitude (voltage)-based relationship with burst pressure as a predictor of structural adequacy. This approach is based on the relationship between amplitude and volume of tubing affected by degradation, a well-founded dependency which predicts that as the tube condition becomes more extensively degraded the EC signal response in volts becomes larger; concurrently the more extensively degraded the tube becomes the less capable is the tube with respect to the intemal pressure it can withstand before burst. Thus for NDE uncertainty the focus is placed on standards and measurement repeatability. Since all the measurements must be taferenced to a known condition, the industry practice of using ASME standards is the comerstone of the APC practice. To minimize effects of the variability of standards, each particular ASME tubing standard used to calibrate the field NDE responses is cross-calibrated to the ASME standard used in the EPRI laboratory study. Thus each standard is constrained to produce measurements which are directly comparable to those produced from each plant using the same size tubing. To assess the effects of probe construction differences on amplitude measurements, the EPRI study compared bobbin probes manufactured by Zetec and . Echoram, finding them essentially equivalent for the purpose. Additionally Westinghouse has compared the responses of a number (12) of production probes built by Echoram on the same standard. It was found that the variability of the resoonses in the support plate mix cnannel was - less than 5%. Eddy current system - cabling, instrumentation, etc. - variab:lity arising from noise is of the order of 0.1 volt at the calibration used for field measurements; this is essentially 5-6 l
negligible compared to other sources of error for applica'.lons to plugging limits of the order of ! one or more volts. '
, .Speclat concern attends measurement variability arising from wear of the probes' centering '
devices. Excessiv9 play may result in off center positioning of the probe relative to the flaws which affect the EC response. Thus a new probe with design centering produces the proper response, while the same probe with worn centoring devices may lean away frem the flaw cr toward 11 producing smaller or largot amplitude responses. To reduce this variability, limits t are placed on the usage of an otherwise electricalh sound probe; each probe is required to give amplituoes no greater than 115% different at any time from the responses when new to four identical,100% deep holes staggered axially on a standard tube (" probe wear standard") (This - device was not availtble at Catawba-1; estimates or probe wear uncertainty are described in Section 5.8.2.) Periodic rneasurement of the probe wear standard identifies when the probe centering is inadequate and replacement la required. An allowance of 15% is provided in the
- plugging lim!! calculations, though somewhat lower variability is expected.
Data analysis guidelines for voltage measurements are provided in FC sizing guidelines,in Appendix A. It has been found through experienc9 at Plants a J and L that when given a common ' orientation to specific measurement guidelines that the variability arising from analysts' differences are reduced to less than 110E As expected, this uncertainly is larger at low voltages; this results from the lower signa! to noise ratio. As the S/N value increases,
. measurement variability diminishes with the result that for plugging limit voltages the overall average is a conservative correction.
To reduce the spread of possible respor:ses to a given morphological condition, the measurements used for the voltage / burst pressure correlation are taken from the field, pre pullinspection data, it has been observed on many occasions that there are unpredictable differences in " amplitudes of flaw signals between pre pull and post-pullinspection data. This results from mechanical deformation of the tube, such as elongation, denting, scratching, etc. which occurs in the process of removal The contributions to the NDE uncertainty at the 90% cumulative probability are calculated for each of the error sources. These sources are treated as independent variables and combined as a root mean square (RMS) to obtain the not NDE uncertainty. This value is then applied in the calculation of the tube plugging voltage limit. For probabilistic SLB leak rate evaluation, the cumulative probability distribution or a normal distribution of NDE uncertainty is utilized. 5.8.2 Catawba 1 NDE Uncertainties The EC uncertainty consists of the EC analyst variability and the probe wear contribution. For the 1992 Catawba-1 inspection, these are developed as described below: ' EC Analvst VariWity The most extensive evaluation for the EC analyst uncertainty was performed at Plant L. Figures 5-8 and 5-9 show the indications and analyst uncertainty from the Plant L study. At 90% cumulative probability, the EC enalyst uncertainty is 10% The uncertainty in percent represents the voltage difference from Figure 5-9 divided by the mean voltage of 1.41 volts. An upper limit on the analyst variability uncertainty results from plant specific guidelines for resolving voltage differences between analysts as described below. 5-7
,, , - _ , , , .._,___~._._,___m- - - - , - - -
For Catawba 1, the 1992 Iidications and the associated 1991 Indications for developing growth were reanalyzed by Duke Power using guidelines consistent with APC requirsrvnts as described ' in Section 5.2. A sampir of 18 indications were independently reviewed by Westinghouse. This partial assessment for ti e largest indications in S/G C support consistency with the APC . . guidelines for the 1992.nspection evaluation. Thus it is reasonable to apply the Plant L EC anafyst uncertainty foi the 1992 Catawba 1 indications. For the 1992 inspection, a resolution process was implemented to obtain the final voltage amplitudes for application of the IPC criteria. This resolution process required that each analyst call on resizing the indications to the APC guidelines was reviewed by a teu n of 2 resolution analysts who concurred with the accuracy of the measurement. This rnsclution process can be expected to limit the EC variability described above for Plant L (based on differences Detween analysts with no resolution process) at upper / lower bound cutoff values. To estimate the cutoff or upper bound values for analyst variability,123 indications were given a repeat analysis which can then be compared with the primary analysis (final outage voltages) for each indication. Emphasis on the primary voltage analysis was placed on conservative peak te peak voltages. The primary analysis was performed utilizing the voltage resolution process described above. The repeat analysis was a single analyst evaluation (no resolution process) of the voltages. Thus the voltage differences between the primary and repeat analyses can be expected to bound the voltage variability that would be expected with two independent analyses carried through the resolution process. The resulting voltage differences between the resolution process (resolved volts) anc' the single analyst are shown in Figure 5-10 as a fur.ction of the primary voltage call (resolved volts) for each indication, in most cases, the primary calls are higher amplitudes particularly above about 0.5 volt. At low (<0.5 volt) - amplitudes, the voltage variability is higher with both plus and minut values. This result is generally exoected for flaw signals of comparabie magnitude to noise and residual signals. Above , about 0.8 volts, the maximum voltage difference is <20% of the resolution process voltages. Since this range of voltages is of most significance for IPC applications, an EC analyst variability cutoff at 20% can be applied. Thus the EC analyst variability can be represented as the distribution of Figure 5-9 with a cutoff or maximum uncertainty at 20% This uncertainty has been applied for final Catawba-1 analyses while preliminary analyses were performed with no cutoff on the distribution, Probe Wear Uncertainty Figures 5-11 and 5-12 show the database on voltage sensitivity to probe wear, For plants implementing the probe wear standard, the voltage variability of Figure 5-12 is obtained from the Figure 5-11 data by including all data to 20 mil radial wear for the Echoram probe and to 6 mils for the Zetec probe. The resulting probe wear uncertainty has a standard deviation of 7% The probe wear standard was not implemented in the 1992 Catawba-1 inspection. Zetec probes have been used for the 1992 Catawba-1 inspection. Thus the data for the Zetec probe (bottom figure) of Figure 512 are applicable to Catawba-1. Mockup tests with the probe wear standard have shown that at 0.0075 inch wear, the wear standard requires probe replacement for 90% of the tests and only the data up to 5 mils wear was used for the EC unceitainty of 7% With the - probe wear standard, the probe wear uncertainty is cut off at 15% by the probe wear replacement requirement. It is reasonable for estimating the Catawba 1 probe wear uncertainty to include the 7.5 mil data in determining the standard deviation and apply this uncertainty with - no cut off. 58 s _ ~ ._,
~ . _ _. _ - _ . _ , _ . _ _ _ _ . _ _ . .. . _ _.-
r-Figure 5-13 shows the probe wear standard amplitudes and voltage differences from tho'mean- ; for all Zetec probe maasurements between no wear and 7.5 mils wear, This simulates measurements between a new and well worn probe. -The uncertainty of Figure 5-13 shows a .. standard deviation of 0.55 volts for an average of 6.05 volts or a standard deviation of 9%. For - > Catawba 1 the probe wear uncertainty.can then be estimated by rounding to a standard deviation of 10% with no cut off at high confidence levels, it can be noted from Figure 5-11 that the average voltage, as well as the standard deviation, tends to increase (conservative for tube repair) at large probe wear (7.5 mils). The probe wear uncertainty of 10% standard deviation was applied for preliminary Catawba 1 analyses. To further evaluate the probe wear uncertainty, data obtained during ASME standard calibration runs at the begi_nning and end of each data tape were collected for further evaluation. Table 5-8 shows an example of this data for one of the larger number of tubes inspected (1456 tubes) with a given probe. The single,100% throughwall ASME hole data can be used to estimate the probe wear uncertainty by evaluating the variab_ility in the 100% hole voltages. Data such as Table 5-8 were collected for 10 probes covering the inspection of 8947 tubes on 58 tapes for S/G's C and D. The 8947 tube inspection results spanned 95 tapes of which 100% hole voltages were collected at the beginning and end of 58 tapes. The number of tubes inspected per probe. varied from 63 to 2391 for these data. To be representative of the IPC measurements made for each tape, the 100% hole measurements were adjusted to 2.75 volts for the corresponding 4 x 20% ASME holes used for data calibration of the field measurements, The 100% hole data (116 calibration standard measurements) were then evaluated to determine 1 the mean with the standard deviation about the mean (voltage diffarences) applied as a measure . of the probe wear NDE uncertainty Figure 5-14 shows the resulting distribution for the 100% hole measurements. The standard deviation is 16% of the average or mean voltage. .This probe wear uncertainty of 16% is applied for the final Catawba-1 IPC analyses, it can be noted that the range of voltage differences from the mean is about -25% to +50%. The larger spread for overestimating voltages due to probe wear leads to cor'servatively large voltage measurements for the field applications. The negative voltage differences or underestimates of the voltage amplitude are bounded by less than two standard deviations. Combined EC Uncertaintv The probe wear and EC analyst NDE uncertainties can be considere( ~o be independent variables. For Monte Carlo analyses to obtain EOC voltages, separate distributions can be used and independently sampled for the two contributions to the NDE uncertainty. For deterministic analyses of tube integrity, the EC uncertainties at 90% and 99% cumulative probability are required. The independent uncertainties can be combined as root-mean-square (RMS) averages. The results of the preliminary and final analyses for the NDE uncertainties are summarized in Table 5-9. The preliminary values, as developed above, were a 10% standard deviation for probe wear and tube analyst variability of Figure 5-9 with no cutoff at the larger voltage differences. The final values are a 16% standard deviation for probe wear and the analyst variability of Figure 5-9 with a cutoff on the distribution at 20%. 9 5-9
Table 5-1 Comparison of Bobbin Signal Amplitudes Between 3/4 inch and 7/8 Inch Tubing for Different Flaws Flaw Tvoe & Size 3/4" - 43 Mi! Tubina 7/8" - 50 Mil Tubino (550/130 kHz mix). (400/100 kHz mix) 20% ASME Holes 2.78 volts 2.75 volts 100% ASME Hole 6.4 volts 8.2 volts 1/4" Long,100% Deep Axial Slot 42 volts 43.5 volts 1/2" Long,100% Deep Axial Slot 77 volts 75 volts 4 h r ! 5-lo l
Tablo 5-2 Voltage Normalization Trends Between Frequency Mixes (1) Sianal Amolitudes for ASME Standard Hofe S!zes Frecuency Mix 40 % 60 % B0 % 2PL% 102 % Voltages 550/130 kHz 2.75 3.40 5.12 5.80 5.83 400/130 kHz 2.75 3.30 4.80 5.26 5.15 400/100 kHz 2.75 3.26 4.60 5.01 4.88 Ratio of Voltages 550/130 1.00 1.03 1.07 1.10 1.13 400/130 550/130 1.00 1.04 1.11 1.16 1.19 400/100
- 1) Adjustments applied for Catawba-1 growth rates at '00/130 and 400/100 kHz to 550/130 kHz are based on voltage ratios for field indications. Evaluation of the ASME standard described in this table independently demonstrates larger renormalization factor for adjusting the 400/100 data of 1991 than for the 400/130 data of 1992.
O 5-11
Table 5-3 Voltage Adjustment Factors to Obtain APC Normalization for 550/130 kHz Mix Post-Pull 550/130 kHz - Factor for Adjusting Field and 400/100 kHz Data (4) > JSP Norm. to 20a4 ASME Norm. 400/100 kHz 550/130 kHz Iute Ed TSP Volts (1) Adiustment Factor (2) ygits yg),tg RSC112 2 6.92 1.38 0.25(5) 0.37 3 4.44 5.06 R10C6 2 6.4 1.28 1.82 2.07 3 4.77 5.34 R10C69 2 6.4(3) 1.28 --- NDD 3 2.92 3.31 R20C46 2 6.04 1.21 0.59 0.82 3 0.75 1.04 R7C47 2 7.8 1.56 --- -- 3 3.65 4.13 J Notes-
- 1. Westinghouse measure of standard TSP volts when 20% ASME volts set at 2.75 volts.
- 2. Voltage adjustment to convert voltages normalized to 5.0 volts at standsrd TSP to normalization of 2.75 volts for 20% ASME hole.
- 3. Adequate TSP not available on standard. Assumed same as tube R10C6.
- 4. B&W evaluations of post-pull data.
- 5. The 400/100 kHz data were renormalized to 2.75 volts Dr the 20% ASME hole.
l i l l 5 12
l Table 5-4
, Fleid and Westinghouse Evaluations of Catawba-1 Pre-pull Voltages Field Evaluation ,,, Westinchouso EvMuation 400/100 kHz Mix 550/130 kHz 400/100 kHz 550/130 kHz 20% Hole 20% Hole 20% Hole 20% Hole RPC
'[de E TSP Norm. ASVE Norm. ASVE Norm. ASME Norm. (SME NortrL Y2115(1)
R5C112 2 NDD --- 0.31 0.a8(2) 3 1.15 1.59 1.88(2) 1.53 1.82 1.30 R10C6 2 0.82 1.05 1.29 1.20 1.46 0.93 3 0.77 0.99 1.23 1.07 1.31 1.20 R10C69 2 NDD - - - -- NDD --- 3 0.93 1.19 1.45 1.22 1.48 0.97 R20C46 2 0.31 0.38 0.56 0.25 0.42
, 3 0.40 0.48 0.67 0.59 0.70 R7C47 2 0.33 0.40 0.58 0.34 0.51 3 0.80 1.25 1.51 1.20 1.57 1.40 N21eS:
- 1. RPC volts at 300 kHz normalized to 20 volts for 0.5" EDM notch.
- 2. Obtained from 400/100 kHz evaluation using Equation 5-2.
O 5-13
- - -. - -. .-. - - .- .- .- ~ . . .. - . . - ..
Table 5-5 Delgian and Westinghouse Evaluations of Plant E-4 Eddy Current Data (1.2) Belalan Fle!d Eva!uation Westinahouse Evaluation .. Zetec Eculo. Belaian Eculo.(3) Iub2 ISP. P 550/130 kHz 300 kHz 300 kHz 550/130 kHz 200 kHz Volts Volts Volts Depth Volts Depth Volts R20C34 2H 4.95 1.17 1.33 71% - 5.03 70% 1.11 (9.10)(7) R16C31 1H 5.75 1.43 1.27 65% S.85 69 % 1.32 , (10.58)(7) 2H 9.30 2.02 2.25 70% 9.25 72 % 1.95 (16.73)(7) R40C47 1H 0.17 0.09 NDD(5) -- 0.17 41 % 0.09 (0.31)(7) R45054 1H 9.57 2.29 2.25 65 % 9.53 69 % 2.21 (17.24)(7) 2H 0.45 0.22 0.20(4) -- (6) 0.83 53 % 0.16 (1.50)(7) R47C66 1H 9.28 2.26 2.12 72 % 9.39 69% 2.13 - (16.99)(7) 2H 1.39 0.53 0.52 40%(4) 1.37 38% 0.51-(2.48)(7) R33C96 1H 3.57 0.96 1.07 68 % 3.54 67% 0.88 (6.40)(7) Notes:
- 1. Voltages at 550/130 kHz mix normalized to 2.75 volts on 20% ASME hole
- 2) Voltages at 300 kHz normatized to 2.0 volts on 4 throughwall,1.25 mm (0.049") holes
- 3) Amplitudes and depth measured by automated signal analysis
- 4) Manual correction for small signal to noise ratio
- 5) Signal below automated detection t'reshold
- 6) No depth measuremem for insufficient signal to noise ratio
- 7) Voltages in parentheses corrected for cross calibration factor of 1.809 between Belgian made ASME standard and the reference laboratory standard. Laborelec is conducting additional studies to determine if other adjustments are appropriate.
5-14 4
.~. . .-, - ,-
6 Table 5-6 natio of U.S. 550/130 kHz to Delg!an 300 kHz telgien Asset 4 IW A'~ 4 IW 1.N f r e<pA rr y 100% 0.0M 0.027 0,04$ 0.049 tot 601 80% Prote f yre 100* fee, S t arvtar d 20} rwr Prote 18,1 l 2.62 4.13 4.54 4.17 man-mag. No tab 2.75 15.7 l 550/130 let x 3.14 4.17 4.1T 3.50 i 3.93 2.0 550 0.65 0.69 0M 0.52 n.89 300 l 2.51 4.15 4.81 4.44 Yes tab 2.75 550/130 Echo. Mag-biss 4.43 3.72 3.So 3.01 4.16 550 0.62 0.62 0.5! 2.0 l 0.68 0.51 500 5.63 5.51 18.6 fransfer 2.68 3.18 4.23 (che. Mes-bias tes 5.22 4.69 36.1 550/130 3.82 3.82 4.31 550 0.65 0.72 0.63 2.0 0.62 0.60 300 3.63 3.09 16.32 2.75 2.5G 3.17
'es setsian 9.58 550/130 0+teten 3.57 2.91 3.31 3.49 2.90 550 0.79 0.77 9.59 2.0 0.94 0.72 300 4 g io $50/130 (APC) to 300 (selefan) 6.88 5.02 9.M tem 3.09 4.03 5.99 t e mag so letec 4.92 6.69 7.T6 S.11 Yes tab 4.04 tche. Wag-bias 6.50 T.82 8.75 9.30 yee fransfer 4.32 5.30 Ecbo. Meg-bies 4.71 5.24 5.16 Selgia, 2.93 3.47 4.01 Tes Setsfon I
l T.+#2001 :072192 5-15
-1 Table 57 NirrJr.ary Lebotelec Hes' :ts for Renormalization of Belgian to U.S. Volts 3/4" Tubina-Volts 7/8" Tubina-Volts .
Jtem 300 KHz 550/130 KH 240 KHz 4_90/100 KHz A. tiLnufac t uri na P.r_qtus : 4-TW 1.25 mm Holes V. S. Drilled 1.99 10.58 1.88 9.26 Belgian EDM 2.00 10.56 2.00 10.17 Ratio EDM/Orilled 1.005 0.998 1.064 1.098 Manufacturina Process: 4-20% A1ME Hol n V. S. Drilled 0.63 1.56 0.71 1.90 Belgian EDM 0.94 2.75 0.94 2.75 Ratio EDM/ Drilled 1.492 1.763 1.324 1.447 B. In. fluence of Probe Desian: U. StASME Std. 20% Holu (U.S. Standard) Echoram 0.61(l) 2.76(2) 0.94(I l 2.77(2) Laborelec 0.64 2.74 0.76 2.74 ~ Ratio APC to Belgian Normalization at 20% Depth . o Echoram 4.52 2.95 o Laborelec 4.28 3.61 4 TW ).25 mm Holes (U.S. Standard) Echoram 2.00 18.52 2.00 11.10 Laborelec 2.01 18.77 2.00 13.11 , Ratio APC to Belgian Normalization at 100% Depth o Echoram 9.26 5.55 o Laborelec- 9.34 6.56 o Ratio applied -4.93*l.5 7.4 in Catawba-1 Evaluation Notes: 1) Normalized to 2.0 volts for 4-TW l.25 mm holes.
- 2) Normalized to 2.75 volts for 4-20% ASME drilled holes .
(APC Calibration). 5-16
l i l ) Tablo 5-8 VOLTAGE COMPARISON _FOR THE Life OF A FROBE EfC C-INLET Prece S/N 0144619, samo probe from tape 001 to 014. Probe type .610 M/ULC ASME Std. S/N 050415 Calibration setup IAW RESIZE guidelines. Voltage normalized at 2.75 P/P on 20% FBH's on 550/130 mix at tape 001.ca101 and not changed for measurements taken for the life of the probe. Approximate number of tubec examined with this probe: 1456
=
l Voltage measurements taken peak-to-peak on 550/130 mix. i l TAPE CAL / ASME FLAT BOTTOM HOLES INITIAL-01 ASME 10% ASME 20% FINAL-02 OD ID 100% 80% 60% 40% 4 X 20% 4 x 100% GROOVE CROOVE 01-Ist pull 4.68 2.74 4.66 001 01-2nd pull 4.60 2.75 4.69 01-3rd pull 4.57 j5.13 4.29 3.02 2.75- 4.61 4.52 86.02 C01 02 4.59 5.04 4.33 3.06 2.79 4.47 { 01 4,36 4.95 4.04 3.01 2.90 4.77 005 l 02 4.32 4.84 4.12 2.90 2.87 4.61 01 4.36 4.85 4.23 L.98 2.86 009 4.72 - 02 4.56 S.14 4.36 3.10 2.87 4.81 j014 01 4.09 4.52 3.83 2.80 2.95 4.90 .
! 02-Ist pull 3.92 4.50 3.80 2.77 3.05 4.82 C2-2nd pull 3.90 4.52 84.07
!C14 3.00 4.86 02-3rd pull 3.88 2.90 4.87
! I Circumferentially Circumferentially Assymetrical Synetrical 5-17 l
___ _ _ _ _ _ _ _ - _ - . _ . - _ _ _ _ _ _ - - - - - - - - - _ - - - - - - - - - - - - - - - - - --- -~- - - -
_ ~ = . - _ . . . . _ ~ . . . - . . .. . - . . - . I l l Table 5-9 Summary of Catawba 1 EC Uncertainty-
- Analvst Variability Pc ae Wear RMS Averpgg Prelimirary Estimate Distraution for Cum. Prob. In % Norma! Distr. - Apply separate Monte Carlo Columns 2 and 3 ~ mean 0,r=10% distributions of Table 510 Value at 90% 10% 13% (1.28r) 16%
Cum. Prob. Value at 99% 34 % 23% (2.33r) 41 % Cum. Prob. Final Estimate i Distribution for Cum. Prob. in % Normal Distr. Apply separate Monte Carlo Columns 2 and 4 mean=0,r-16% distributions - of Tablo 510 Value at 90% 10% 20% (1.28r) 22 % Cum. Prob. , Value at 99% 20% 37% (2.33r) 42% Cum. Prob. 5-18 l
}
Table 5-10
.. Cumulctive Probability for EC Analyst Variability
% Uncertainty Percent Voltace Din for Vohace Bin (1) Cumutatiw Probab!!ity Preliminarv - .fjDal
-1.0 -71 0.11 1.00 to .75 -53 0.39 0.75 to -0.50 -35 0.93 20 0.00- -
0.50 to -0.25 -18 3.94 3.94
-0.25 to -0.20 -14 5.49 5.49
-0.20 to -0.15 -11 8.56 8.56 c 0.15 to -0.10 -7.1 12.6 12.6
-0.10 to -0.05 -3.5 21.6 21.6
-0.05 to 0.00 0.0 48.3 48.3 0.00 to 0.05 3.5 75.3 75.3 0.05 to 0.10 - 7.1 86.9 86.9 0.10 to 0.15 11 91.8 91.8 0.15 to 0.20 14 95.1 95.1 0.20 to 0.25 18 97.2 97.2 20 100.0
.25 to 0.50 35 99.2 0.50 to 0.75 53 99.8 0.75 to 1.00 71 99.97 1.00 to 1.30 92 100.0 Note 1: % Uncertainty obtained as mid-voltage value for each b!n divided by mean voltage of 1.41 volts from Figure 5-2.
5-19
Figure 5-1 Bobbin Coil Voltage Dependence on Slot Length and Depth 0.75" Tubing , 0.75 INCH OD 43 MIL THICK ALLOYSCO TUBE 100 0 _- a "
~
u , to m . H O' a
> 10:
ui :' O g A b - A
$ ~
z 3 r g C O a 1; " U U c: - a a o ; a w w - w U l l l O $10 INCH DA' METER BOBBIN CO!L AT 550/130 kH:- 0 . , , 0 0.1 0.2 0.3 04 0.5 0. 6 0. 7 0.8 0.9 1 AXIAL SLOT LENGTH, INCH l5 TH:iu wAL. A sc% OgE= 2 ecgggg= l i lD 50% 055:
- 4cn CE5:
l I l l l I l 5-20 I
Figure 5-2 Bobbin Coil Voltage Dependence on Slot Length and Depth - 0.875" Tubing 100 4 a 2 E a X 5 Rectangular Axial Sict _ t
. X E g ,
<r M Xg F-J O
> 10; wt-a 1
A - l E / / Tapered Axial Slot Tube Wall C. 3 E _Z_ E . C b - C C n C o a
~
u - C .! J n i~ J- ; 1 l l Om:NC:=t C'.AMt.rM BC* SIN COL AT AT/1C:et i { 0.1; . . i . i . . , - 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 i SLOT LENGTH, INCH E TriRU-WALL A 80% DEE? Z 60% DEE? O 50% DEE? X TAPERED 5-21
~
l Fgure 5 3 CATAWBA 1 Pulled Tubes - Correlation 550/130 to 400/100 KHz mix 6 5- Unear Regression
-T ,
Slope 1.094
.M Intercept 0,143 k 4- R Squared 0.999 I Std Dev h N 0.012] .
O 3-m
$o 2- ,
1- ' 0 . . , 0 1 2 3 d 5 -6 Vohs 400/100 KHz mix ( O 5-22
Figure 5-4 Signal Arnplitudes at 550/130 and 400/130 kHz for Catawba 1 TSP Indications (1992) CORRELATION OF 550/130 TO 400/130 kHz RESULTS FROM 1992 CATAWBA-1 INSPECTION 4.5 4-3.5-3- o C 2.5-6 g 2-
$ 1.5-d a
> 1-0.5-0-W---------~----------
-0.5 , , , , , , ,
0 0.5 1 1.5 2 2.5 3 3.5 4 VOLTAGE AT 400/130 kHz u DATA y = 1.038 x - 0.047 O 5-23
Figure 5 5
+
DOEL UNIT 4, S/G: B . Evaluation of 1992 Voltage ind. at TSPs , 16 m f 14- ( ) No. Data Points 45 m
!5 En 12- EC equipment Tu 550/130 - MlZ18 "
cn 300 - Belgian E O ( ) E Z 8- m , y "" m ( Unear Regression o 6- " g
- Slope 4.93 g , m, Intercept -0.75 - -
g 4- , Std Dev. 0.21 E -
*
- R Squared 0.927]
o 2- N
}o o. f.=..............................................................................................
-2 . . . . i 0 0.5 1 1.5 . 2 2.5 3 .
300 KHz, Belgian Norm., Belgian Eval. ? . 5-24
Figuro 5 6 DOEL UNIT 4, S/G: B Evaluation of 1992 Voltage Ind, at TSPs 18 r 9 16- No. Data Points 53 , U, a EC Equipment - MlZ18 g 14- ' > 3
*/
g' 12-a-
,3 10- .-
2 Linear Regression g 8~ / Slope 1.00
<C Intercept 0.07 y 6- -
Std Dev. 0.004 x g 4- \ .. R Squared 0.999) - b to 2-A
/
0, - > > > > > > 0 2 4 6 8 10 12 14 16 550/130 KHz, APC Mix Norm, Belgian Eval e 5 25 i
. . - . _ _ - _ . - _ - . . - - . . _ - - . . - - . . . . - - . _ - _ . - ~ . - - . - _ , . _
Figura 5 7 DOEL UNIT 4, S/G: B 16- - Evaluation of 1992 Voltage Ind. at TSPs
*iB s 14- ( 3 i g No. Data Points 53 5> 12- EC equipment 4 Tu 550/130 - MlZ18 m " ,
- 300 - MlZ18 O
10- ( ) m # z
.x 8-g,p . ( Linear Regress.
ion T - 6- Sl Pe 4.67 '
- a. A Intercept -0.64 4 * "
Std Dev. 0.06 ' d~ x # JF ( R Squared 0.991 o 2-n e R g o.- .......................................................................................... m
-2 i i i i i i 0 0.5 1 1.5 2 2.5 3 3.5 300 KHz, Belgian Norm., Belgian Eval.
I i 1 . 1 I 5 26
Figuro 5 8 Distribution of Voltage Indications Used for EC AnaYst Variability Evaluation DLANT L MEAN OF ALL SIX ANALYSTS 120 110 -- . : : : : 100 101 / -90 d m 100- - 94 %- . va. .. ' 1.., -80 $ _ 8t4 Dev o.4s d, 85 ' j
- t >
tt 80- b -70 g a g w
/ ,(
i -60 9 60-1 / 1 e0 b { c =
-50
. b Ir 't k x . y a
-40 h y 40- 1 , f I i 35 20 I/ $' s :': l* 7 -30 $j 3 t t' 1 i. '
i 26 H z 20-f 4; g :q 4 i 20 -20 $ j g i y 16 m../ 3,. a w {s,:w +a a v v 3
,c: ; i! % $ # :g: A -10
,4 il h i
*E 4:
.a f'j 4 2
0 Mr y
;s; a
4 a$N i i i i A i m "0""O i . > 1
.i 0
0.60 1.00 1.40 1.80 2.20 2.60 3.00 3.40 3.80 0.80 1.20 1.60 2.00 2.40 2.80 3.20 3.60 VOLTAGES 4 5-27
- ..- ~. . - _ ~ - - _ . - - . - - . . - . . - . _ - . . . - . - - _ - . . . - . - .
t i l Figure S 9 i i Distribution of Voltage Differences Between Individual Analysts and Mean Values f PLANT L I 1000 DISTRIBUTION OF ALL DIFFERENCES ! g-2 7
- : -100 7*
T - 900- y -
-90 6 -
Differences in volts Y )' 800- === o.co Average Voltages co t W std Dev o.146 '
- n. . 1,42 -80 _ .
' # std Dev 0.48 i !
!$ 700- $ + ' J w 3, w -70
, cc "
@ 600- <
[ ! -60 500-
-50 fg ) 415 !
400-C
@H4{ -40
- i'*
3 ) h300-o [ '$ fl -30 $ - z ! )' 200- 171 -20 107 109g j ' l 120 100-4 j_0 h{ Sig j j j 74 72 -10
] [
0., - . . . . . . g g l ll37
.it . . .
1 0 1.00 0.50 -0.20 -0.10 0.00 0.10 0.20 0.50 1.00 0.75 -0.25 0.15 -0.05 0.05 0.15 0.25 0.75 1.30 VOLTAGE DIFFERENCES P W e 5 28
, vvn - = s, ww a.s e g rvw . ,g e w w - , m w ea cm.r e-,-en- a,e- ,,-,noswme.- -wwa.vw ,,m,-vsr-a,<c... ,
i Figure 510 i Catawba 1 EC A..alyst Variability: Comparison of Voltage Differences Between Resolution Process and Single Analyst - 1 r CATAWBA-1 EC ANALYST VARIABILITY (Resolved Single)/ Resolved Volts 100 3 ON 80-w s g 3 - - u 40- = , f a> ~ "" 3 20 -
@ "Da -
Et 15 -. - ==- -
.9 0--
.'*far ..' - - " .
v.= =- E" -20 w- .
==
g .40- . 8 e .. cc m. "
- .80- .
=
100 i i i . . . i 0 0.5 1 1.5 2 2.5- 3 3.5 4 Resolved Volts ; i e h e f 5 20 5 Fi---9--m-'-d' ---wSUI-'6a -4: g gg w wt +W,eF'sme-e-Q w$ +gveJT N. p ra-~r+w-
Figure 511 Bobbin Coll Amphtude Dependence on Probe Wear Ectuam Pm
- 14 Wear Stardard n Vevtcal TWe Locaten i3 ............................,. ............. ............. ..............,..............g....g.......
n............... . ...... .
,.,,,,,,,n,,, ....... ........... . ............. ....
e .< uu ;........ ,a n. ..... .. ........... , ....... ......... e..e .
. , ,, ....... 3 ..
c, $3 ......... .n
- g. ,. .
. ..... n.
7.........
.. p ...
Ju ............,............i..............f......-r..{.............g...........4,. ............{....1P........ p ,............ ............ . y1 . . . . . . . . . . . . . . . . . . . . . .............. 1
............ ..............g....
................i... . ....... .............
. 4.............. ..... ......,
L as b ou ok o oi oois om oos e.m. *.n ano Ecraam Prtbe ie Waar Stardard Horumtahn Channdhead 33 ............................. .............
- q. ..............p . .
. . ............t............. 4...............'.............g............ 4...............
$ u-........+.............**.* g g io.,.............p , .,. n .............,..............t............ 4.............. n se sw.un (r p. . . . . . . . . . . . . . ' . 2 e.,uno" ...e i
,u.'
e.n Y s s.n "*""i*""*""""I**"*""""4"*"""""" 7 ......... - en
..a J ..u u- ..n : : :
f -.... ..............
....4..
4............... S am o oss ooi o ots os om ons e,e. w.= m Zetoc Prtbe i4 Waar Statuard Horuontal n CtannWheed 13 .............. y.............;............
. . 3.............. . . .
t ., ...............,.............. y 3 o- ,......... n,.................. ............. .............;ouooon. gono.ooo & 10-,........ su am u.w. . . . 1 4...............;............. 4............. 4,.............. l .... 8.19 .. *............. l.94 6. I4 ' ..............;.............d............... i . . . .
%,.... tu s.es e.u j. .....;............
4..............;.............;.... . 7.. . . . . . . . . . . . . . ... s.a . .... ..... y ..............
. .............*t.*******'
- e. ............. .... .... .... ....
4............. 4. ............ 4............. 4....... .......
%........................... .... .... 4............. .........n.. .............. ..............
h h n.m b-30
i Figure 512 j VoMage VariabHity Due to Bobbin Probe Wear a Pieene West Vefuge Vesuengir
- teasern Psees a
m huruswa e Vertsof Tm usuem (0 a a 0 00' mour)
. > " **='s ,u wa te M
?
- - . i I'# g')
' F,
- ,p 93 }d
- ' 93
.g
*?O 2 li 1P M 4 :
g '" M io :'/p:
/ .,1 f3 3 A y, f! :- .
- ryg o g g : ,
T] [ +$.. f 4 O Ll d ((; O d 80 Lj $ ,
.. ,, e ,
u ,, I ,, Dnesanoce n Veen a P'edre Wear Vcmage VanessWy . (chorum Petee Woat $eruses e Hortecriam fee uncaem 10 0 9 001l' muer) 20C - 400 ! t 80- ,,,,,,,,,,,,,,e 40
,e :- :::: ;; !-
1e i' ~ , N ,j 70 137 $ f' /' 40 10t>' !k, f
.f 40
[ r 47 ,N
; I _.' !; il 30 e m ['gG : . : i; :
L, -ao i o '- ) . L.i . 2_ 1 40 42 02 QA 1 1. Oftsserene m %ds 4 P'tte Wear Vasmes Venahery Zaeme Poem t Wear Senasse m Mosesrvist Tee Laumann (00 e 4005' meer)
- t a- -
100 t44 je passee-o to was
,,, ~- . . . .
M i, *j 7e .E
- L) ,. ,a
!9 _,,
{ ej c .> 3 !: L
,,. L'" y a ;1 ,l .
- o 4 i :: !
3, ,
/ '
9 34 [ i : ' . :
.h 0
h ,] , u j d5 8,, 4 o
,, . . , , m , ,, ,, ,,
, CdimerenemVOUB 5-31
. - - . - _ = . - _-. .- _ . _ - - _ . _ . --.--- -_- - -. Figure 513 Probe Wear Vohage Variability
- Zetec Probe (0.0 to 0.0075" wear) 220 Wear Standard Honzontalin Channehead _ _ _
100 20> 90 1D 16'
$ 100 r , -00 f voltages
~5 g: n.an s.es -70 140- 'J / '. sto o.v o.ss c-3'/ 3',g t ; .go 120- 112 fs y 10&
[ h ,D ' j p ' y .ff x:
$,j -40 e 65 3p
~ 7 ')
{x '$' q
-30 k: 7 '; ' f 40- 35 S V y ,' ',' : , ' . % 29 4, . o> -20 20-x.
J/g; 9 w X '6 m _26 21 pg z: f: R:
^
~- - - n.
,44 ,
a
- n. g. g. 8 5
-10 -
2
.m C , ,
a S . 5.5. 6 05 7 7.5 8 O 8.5 5 25 5 75 6 25 6.75 7.25 7.75 8 25 Segnal Amphtude(Vohs) Probe Wear Vohage Variability Zetec Probe (0.0 to 0.0075" wear)
.20 Wear Standard Honzonta!in Channelhead 100 200- *
-90 l#
gico. r
, -80 6 151
.4 m 140-1 mean o '.'co***** -70 b
g ~'41
. 1.30 '
120- % : *
> -60 5
.i;jr;' A1 7, 100- + '
a: 2 -50 - 80- 73 E @': 81 '
-40
- '+: 7 k 60-
- 53 $,h,.,,& h ~~%
+. s 29 -
f 32 y -20 s 20- m w W %'/?4 . * #x. ~D B 20 y m $ R
,W4 N. $.- ;n $.. n R 13 -10 0 . .
% J :d , bi ??,.
QL 0 1 -0.6 02 02 06 1 2
-0 8 -04 0 0.4 0.8 1.5 2.5 -
Ottfererces in Voks 5 32
. - - - - - - ~ - . - - , - . - . - . . . - - - -
Figure 514 CATAWBA 1 PROBE WEAR DATA Volts for ASME Hole over Probe Life 30 110 20
~
-100 25- .
-90 b ,
*h r ,
-80
*' 20- #
V lt*S**
@ / Avg 5.12 -70 C g f/ sta oev o.s3 j d 15- S 3- f.76[
'/ :
11 -50 ].
~
S 10- -9 -40 ,g 2 89 h T/f ,
-30 3 5- 4, 4 '
4 A 3 2 / - 10 0 ' < > > > - 0 3.75 4.25 4.75 5.25 5.75 6.25 6.75 7.25 4 4.5 5 5.5 6 6.5 7 7.5 Bobbin Volts Voltage Differences from Mean for 100% ASME Hole 25 200
- * -180 21 8
m 20~ 19 ;
-160p
.6 ? r '
-140 if 5 -
Voltage Differences 2 # - Avg 0.00 h g 15- 1,4,
; sto oe, o.83 -120 g b ,
31 j/
-100f o -
a 10
-9 -
4 -
-80 6
.8 ' ,
8 3
~
5 e h i Y 5 I 7
' 6 -60 33 5- -
W
& : s 4 f 3
-40 E 7 6 j '-
3 1
-20 0
- ,i-
- > - - - - '.n '~
0n 0 1.25 -0.75 -0.25 0.25 0.75 1.25 1.75 2.25
-1 -0.5 0 0.5 1 1.5 2 2.5 Voltage Differences 5 33 i
{ i
Section 6 PULLED TUBE AND FIELD DATA EVALUATION This sectioriidentihos tho fiolif oxperienco data from operating stoam gonorators that aro utilized in the development of tubo plugging critoria for ODSCC at TSPs. The field data utilized includo pullod tubo examination results neluding tubes pulled from Catawba Un:t 1 during 1991 and occurrences of tubo loakage for ODSCC indications at support platos. Emphasis for the pulled tubo data are placed on bobbin coli voltages, burst pressures and leak rato monsutomonts. 6.1 Utilization of Fiold Data in Tubo Repalt Limits Cporating steam Generator opperienco represents the proferred source of data for the plugging - critoria. Sinco the available operating data are insufficient to fully defino plugging critoria, data developed from laboratory induced ODSCC specimens are used to supplomont the field data base. Tl.o hold data utilized for the plugging critoria aro identified in this section. The overall approach to the tubo plugging critoria is based upon establishing that R.G.1.121 guidelines ato satisfied. It is conservatively assumed that the tubo to TSP crovices are open and that the TSPs are displaced under accident conditions such that the ODSCC generated within the TSPs becomes froo rpan degradation under accident conditions. Under those assumptions, proventing excessivo leakage and tubo burst under SLB conditions is toquired for plant operability. Tubo rupturo under normal operating conditions is provented by tho constraint provided by the d'Illed holo TSPs with small tubo to TSP clearances (typicatly ~16 mil diametral clearanco for open crovicos). For the piogging critoria, however, the R.G.1.121
~
guidolinos for burst margins of 3 timos normal operating e.assuto differentiats are applied to defino tho structural requirements against tubo rupturo, in addition t. providing margins against tubo burst,it is necessary to hmit SLB leakago to acceptablo lovels based on FSAR ovaluations for radiological consequences under accident _ conditions. Thus SLB leakage models are required for the plug 0 i n0 critoria in additior, to tubo burst data. Based on the above considerations and the plugging critoria objectivo of relating tube intogrity to NDE measutomonts, the primary data requirements for the pluggin0 critoria are the correlation of burst pressuro capab!!ity and SLB leak rates with bobbin coil vnitago. For plant operational considerations, it is desirable to minimizo the potential for operating leakage to avoid forced outages. The field data of this section indicato very low leakago potential for ODSCC at TSPs even at voltago amplitudos much higher than the plugging limits. C.2 Summary of Pulled Tube Data Base
. The availablo pulled tubo data base for ODSCC at TSPs in Westinghouse steam generators is summarized in Tablo 6-1 for both 3/4 and 7/8 inch d;amotor tubin0. The number of 7/8 inch pulled tubos is provided as a general comparison with the 3/4 inch data and is not utilized in the 3/4 inch ovaluation of this report. Both tubing sizes have a comparablo number of pulled tube intersections although the 7/8 inch tubing has more tube burst data. This group includes hvo tubos from Catawba-1 with nino intersections destructively examined. None of the pullod tubos 0-1 l
- - _ . - - _ - - -=
. ~ . - - _ - - ---
t t i have been reported as leakors during plant oporation. The field oddy current data for all pulled ! tubes woro reviewed for voltage normalization consistent with the standard adopted (see Section 5.1) for the plugging critnria development. l Operating plant leakago orporience for ODSCC at TSPs is summarlzod in Section 6.3. Evaluations of tho 3/4 inch diametor, pulled tube burst and !vak rato data are given in Sections 6.6 to 6.8. i The Catawba-1 burst and leak rate data are ovaluated in Section 6.6. It is shown that three of the i nine Catawba 1 burst tests did not result in sufficient cponing of the cracks and the burst ! prossurcs are not considered reliable. In addition, two intersoctions burst outsido the TSP at ! hand ground location marks and thus ai:>o are not rollablo data. The romalning four Catawba 1
.{'
burst losts are found to require some adjustments to the measured burst pressures as described in Section 6.6. It is also found that the one Catawba-1 intersoction found to leak at normal operating and SLB conditions in the laboratory most likely resullod in throughwall penetration of tho 97% doop corroslon crack during the tube pulling operation and thus does not provido a reprosentativo leak rato for including in the data base. 6.3 Operating Plant Leakago Data for ODSCO at TSPs I Tablo 6-2 summarizes the availsfe information on throo suspected tuoo loaks (3/4 inch ' tubing) attributable to ODSCO at TSPs in operating steam generators. Those leakers occurred in Europoan plants with two of tho suspected loakers occurring at one plant in the samo operating cycle, in the latter case, five tubes including the two with indications at TSPs woro suspected of ; contributing to the operating loakage. Loakage for the two indications at TSPs was obtained by a fluorescoln leak test as no dripping was detected at 500 psi secondary sido pressuro. ' '- For the Plant B 1 leakage indication, other tubos also contributed to the approximatof f 63 opd , totalleak rate. Hellum leak tests identified other tubes leaking due to PWSCC indications. Using relative hollum leak ratos as a guide, it was judged that the loak rato for tho ODSCC indication 1 was loss than 10 opd. These loakage events indicate that limited leakago can occur for Indications above about 6.2 volts. No leakago at Catawba-1 has been found that could be attributable to ODSCC at TSPs. 6.4 Voltage Renormalization for Attornate Calibrations To increase the supporting data base, it is necessary to renormalize available data to the calibration values used in this report. For data on 3/4 inch diameter tubing, voltage renormalization has been obtained by applying a normalization for the ASME 20% holes of 4.0 ' volts in the 550 kHz channel and 2.75 volts in the 550/100 kHz mix. The APC voltago normalization and data analysis guidelines have been discussed in Sections 5.1 and 5.2. The 1 voltage renormalization for the Catawba 1 and Belgian pulled tubes aro described in Sections 5.6 and 5.7. The Catawba 1 renormalization from 400/100 Jiz mix to the 550/130 kHz mix was - obtained from post pulllaboratory measurements and is shown in Figure 5 3. The Bolglan renorma!!zation was obtainod by direct measuromonts of field indications as shown in Figure 5-4. As discussed in Section 5.7 the Belgian voltages have boon increased by a 1.5 factor . pending completion of further Laboroloc studies of the voltago renormalization to the APC guidelinos. 1 62
, _._ u ..-. _ . _ , - , . - - .,u. _a _. . . _ _ - - - _ _ _ a.,. _ _ . . = = _ . _ . . = .
l i 6.5 TonsHa Property Considerations Tho 3/4 inch diamotor model boiler specirnens have above averago tensito proportlos while the l pulled tubo daia have both highor and lower tonsile proporties thsn average values. The tensilo property differences botwoon rnodol boiler and pulled tubo data are greator for 3/4 inch tubing than found for 7/8 inch tubing. Tho 3/4 inch model boilor tubing had abovo averago (6%) material proportlos v hlle the 7/8 inch model bollor tubing had proporties cilghtly bolow average For tho 0/4 inch tubing APC development, all model boiler and pu!'od tubo burst prossure data are renormalized to averago tensito proportlos for 3/4 inch tubing as described in this section. Tubing manufacturing data havo boon utilized to develop mean tensilo proporties together with t.4 standard dovlation and lower 95% tolerance limit at room temperature and 650 0F. Those ; data are given in Tablo 6 3. Also given in the table are the values for (Sy + Su). The mean (Sy l
- Ou) value of 154 ksi(twice the flow stress) at room temperature (WCAP 12522) is used to normalize the measured burst pressures for the model bollor and pulled tubo data. The ratio of the 95/95 Lower Tolerance Limit (LTL) flow stress at 650 0F to thu m:sn flow stress at room temperature is utilized to adjust the voltage / burst oorrelation obtained at room temperature to obtain the operating temperatuto LTL correlation. This ratio is 0.848. i Tablo 6-3 also includes the tonsite proporties for tho 3/4 inch model bollor spoclmens and for -
each of the availablo pullod tubes. Sinco burst pressures are propurtional to the flow stress, the measured burst pressures are normalized to mean proporties by the ratio of the tubing mean (Sy
+
+ Su) of 154 ksi(flow stress of 77 ksi) at rvom temperature to the tube specific (Sy + Su) g? von in Table 6 3.
6.6 Evaluation of Catawba 1 Pulled Tubes This section describes an ovaluation of the Catawba 1 pulled tubo burst test anu leak rate measuromonts. The burst test data are ovaluated to assess whether the measured pressuros are representative of a burst or a more limited crack opening causing leakage. A completed burst test is charactorized by fis!. mouth opening of the crack, bulging of the tubo and/or tearing at the - edges of the corrosion crack. In general, the burst test opens up the entire macrocrack length or a very large fraction of the crack length. It is shown below that three of the Catawba.1 burst ' tests resulted in minor crack opening such that the resulting pressures do not represent a burst , test. The bobbin voltages for the Catawba 1 tubes are discussed in Section 5.6. The first four columns of Table 6 4 provide the tubo locations, bobbin volts and measured burst pressures. Column fivo provides thS !!ow stress adjustment factor for adjusting the burst pressures to , average material proporties as developed in Tablo 6 3. The following evaluations form the basis , for the adjusted burst pressure column of Table 6 4. Tube RSC112. TSP 3 i-j
. Catawba 1 pulled tube RSC112, TSP-3 had a field bobbin voltage indication of 1.82 volts and a post pullbobbin voltage of 5.06 volts. By destructivo exam, the maximum corrosion depth (at 1 of 23 axial Orinds) was 97% in the laboratory, the indication was found to loak at about 500 1 psi while pressurizing for a burst test. The tube section was then leak tested at prototypic
! conditions before further burst testing and found to have leak ratos of 0.078 and 0.56 t/hr. at normal operating and SLB pressure differential, respectively. A bladder was then inserted to 63 l
,J.-..._ , ,. _..- ._. _ . ,_. - . ...--.. _ ,..L..,m.,., , . _ . . . _ . .,~,-.m , . _ , _ . . , .. _ . . , _ _ . .. _ _ _ _ . . _ . _ _ ._ _n _2. u.-
b continue burst losting. The ' burst" pressure measured was 4,150 psi A burst prossure of this magnitude would be associated with a uniformly throughwall crack size of about 0.45 inch longth. This section ovaluates the destructivo examination data to assess the Ukolihood of the low i burst pressure and whether the leakage was duo to tearite of the crack to throughwall during * ' tubo removal operations or during leak testing. Figures 61,6 2 and 6 3 show, respectively, the post burst tost crack, a map of OD crack indications and the crack depth vs longth of the macrocracks that oponod during leak and burst - testing. Figure 6-3 also shows the locativn and length (-0.19 inch) of throughwall crack . opening following the burst test. From Figuro 61,it is soon that two post burst crack openings are separated by a ligament. The longths of tho two throughwall ponotrations are about ! 0.11 inch and 0.08 inch. These longths are typleal of individual crack initiation sites ' (comotimos called microcracks). From observations of tho metallography obtained at the various grind depths of Figure 6 2, the two oponod cracks are parts of separate macrocracks with an untorn ligament soon betwoon the opened cracks. Even the end to end opened crack length of about 0.19 inch is rnuch less than the 0.45 Inch throughwall crack expected for a burst i pressure of 4,150 psl. A completed burst test is charactorized by fishmouth opening of the crack and/or tearing at the edges of the corrosion crack. This burst test shows neither of these burst features and did not opon up olther of the macrocracks. The overall macrocrack longth from Figuro 6 2 is about 0,43 inch which is loss than the 0.45 inch throughwall crack length expected for a 4,150 psi burst pressuto, it is concluded that the
- burst" test is an incomplete lost. It is postulated that a slow pressurization rato permitted the bladdor to enter the microcracks as they opened and caused the bladder to tear which terminated the test. The burst ,
lost was not repeated at a higher pressurization rate. As a consequence, the " burst" lost is not considored rollablo and is not included in the voltago/ burst corrolation data base. # A threefold increase in oddy current bobbin voltage and th. appearance of leakage at a pressure of , 500 psi f i a post-pull tast raises questions of damage to tubo RSC112 prior to leak rato testing and the suitability of it'cluding this leakago data in leak rate . bobbin voltago correlations. The measured crack depth of Figuro 6 3, as obtained from the metallography of the successive ar.lal grinds of Figure 6 2, was used to estimate the pressure at wtuch fracture of the romaining crack depth ligaments would be expected. As discussed below, the ostimated pressure at which ligament fracture and thus leakage would 60 orpocted is many timos highor than the observed pressure of 300 psl. This indicatos the tube was damaged prior to leak rate testing and should not be included in the generalleak rato database. The measured SLB leak rato for this indication of 0.56 litor/hr would be orpocted to have a throughwall crack longth of about 0.1 inch or larger. Typically, the bu,st pressure of a cracked steam generator tube is expected to reach the full < plastic collapso vatuo. However, very deep cracks with consequently very small remaining ligaments in the depth direction can exhibit ductilo tearing prior to reaching the full plastic collapso pressure. For full plastic collapse or burst, the remaining wall thickness ligament spanning the crack facot, must be able to stretch without tearing until full plastic collapse of the l tube occurs. A ligament which is only a few mits in dtipth cannot stretch more than a few rnils in l height. Obviously this situation is exacerbated by long cracks which, if they were throughwall, I would open much more than shorter cracks. Thus a small wall thickness ligament may open to
- leakago prior to bursting of the tubo. <
A partial throughwall crack can bo viewod as a throughwall crack whose opening is opposed by springs, that is, ligaments. For a liganont to survive until full plastic collapse of the tubo, the ligament must be large and strong eno"Oh to limit the crack opening to some acceptable value. To 6-4 '
-- . - . . - - - . . - - - . - - - ~ . - - - -- - . - - . - . - , - . -
e t t a first approximation, the maximum amount a ligament can be expocted to stretch is about equal ,
- to !!s width. Honco, the critical value of the crack tip opening displacement, CTODc ritical 18 equal to the size of the remaining ligament, As ono limiting caso, CTODcritical(br a throughwall crack is then expocted to be equal to tho tube wall thicknoss. This agroos with the fracture !
appearance of past test specimons and a privato communication from Belgatorn. As a general I
, critorion for the fracturo of the remaining ligamord of a partial throughwall crack, CTODcritical is taken equal to the romaining ligament depth.
Elastic plastic calculations of CTOD for partial throughwall cracks in cylindors ato avalfablo in l the literaturo. However, the crack depth is assumed to be uniform as opposed to the trapezoldal ; shape of the crack in R5C112. An equivalent uniform shapo was thorofore assumod but - CTODcritical was based on the doopost part of the crack. On an area basis the equivalent crack i depth in tubo R5C112 is 60% With a maximum crack depth of 97% nnd a romalning ligament of 3% the CTODeriticalvalue is 0.0013 inch, From the calculations of Erdogan,Irwin and Ratwani(0, tubo RSC112 is expoclod to tear throughwall at a pressure of about 3300 psl. A : modiflod Dugdalo approach to the samo probl6m lod to a calculated pressuro_of about 3700 ps! for throughwall tearinp. This pressuto wov'd be associated with loakage and would be expected to be below the burst prossure for plastic collapse of the resulting throughwall crack. Even for dooper equivalont crack dopms, for examplo 80% the computed pressures for crack tearing aro - many timos greator than the first leakago pressure of 500 psi. Honco,it is concludod that l damago to tubo R5C112 occurred prior to leak rato losting and the leak rato measutomont should ! not bo include in the databaso. H10C09. TSP 3 This indicatiot, had a bobbin voltage of 1.48 volts which increased to 3.31 volts in the post pull inspection. Figuro 6-4 shows tho burst crack opening after the burst lost and attor cueng and bonding of the tubo to open the macrocrack. It is soon that only minor crack opening has occurred and the opened length is very short (0.1 inch). After bonding to open the macrocrack, a largo part of tho macrocrack is not throughwall. Figuro 6 5 shows that the burst opening - roprosonts only part of the approximately 0.37 inch macrocrack longth_ which had a maximum ; depth of 75% The measured burst pressure of 5000 psiis approxiniately the expected burst ' pressuro for a 0.37 inch throughwall crack (soo Figuro 712) and rnuch less than expocted for an average depth equal to the maximum 75% depth, it is concluded that the burst lost did not rosult ic a comploto burst. Thorofore, this indication is not included in the burst prossure data base. R7047. TSP 3 The 3rd TSP Intorsection of R7C47 had a 1.57 volt indication that increased to 4.13 volts in the t post-pullinspection. Figuro 6 6 shows the burst crack oponing for this indication and Figuro 1 o 6 7 shows the crack map. The macrocrack associated with the burst opening is about 0.43 inch ! long with a maximum depth of about 87% Figure 6 8 shows the crack depth vs longth which ! Indicatos an avorage depth of about 60 to 651 The crack longth having depths greater than
- 1) Erdogan, F., Irwin, G. R., and Ratwani, M., *Ductilo Fracture of Cylindrical Vossols Containing a largo Flaw," Cracks and Fracture, ASTM STP 601, American Society for Testing and Materials,1976.
6-5 i
~-,,,ew-+--. , , , ~ , , r,n,,a -rw- ~-m-, nn-m- nm,-,~,,,un,-,_,,,.,,,n. .,e-+ ,,m -,,--,~mn -oe,< >--- - , - , -- . , -----,,--,,,r.
L 70% is <0.1 inch. The burst pressure for a 0.43 inch long crack with an averago depth of 65% wdd bo expected to be at least 6900 psi cornpared to tho measured 5800 psl. From Figuro 6 6,it is soon that the burst lost secultod in only a minor crack opening of about 0.05 inch which again indicates an incompleto burst lost. The burst pressuro for this Indication would bo
- oxpected to be more than 15% higher than the measured vatuo. Thus the burst pressure should bo increased by at toast 15% or tho data point excluded from the database. The latter option was selected for this report.
- R20C46. TSPs 2 and 3 Both intersoctions of tubo R20C46 burst just above tho TSP olovation at hand hold grinding tool mrAs. These marks woro applied in the laboratory for location purposos. Sinco the burst pressures are associated with the grinding marks outsido the TSP rathor than the degradation within the TSP, thoso indications are not includod in the databaso.
R10CG9. TSP 2 No detectable bobbin indication was found in olther the field or post pull inspection for the 2nd TSP Intersection of R10069, The destructivo exam also shows no monsurablo degradation at this TSP intersection. Figuro 6 9 shows the burst opening which is contored at the TSP Indication. The burst shows a ductilo, fishmouth rupturo typical of bursts for indications with modest degradation. The measured burst pressuro was 9.400 psi. To evaluato the potential nood to adjust the measured burst pressure for this type of indication, an undo 0radod freospan picco of tubo R7C47 was burst by Wostinghouso for comparison with a burst of the Catawba.1 freespan tubing as part of the destructivo examination program. The Westinghouse test yloided a burst pressure of 11,100 psi which is simur to that found for undograded model boiler tubing. The froospan burst pressures during the destructive exam program woro in the range of 9,400 to 9,900 psi or about 12% lower than the Westinghouse tests. Historically, burst possures for undogradod 3/4 inch tubing have been in the range of 10,000 to 12,000 psi. The low burst pressures obtained during the destructivo exam tests tend to indicato a potontial systematic problem in the time framo of those tests. Based on the comparisons of the Westinghouse and destructivo exam burst test results, a minimum increase of 10% to the destructivo exam results should bo applied.- Applying the 10% increase to R10C69 TSP 2 yloids an adjusted burst pressure of 10,340 psl for this intersection as shown in Table 6-4. RSC112. TSP 2 The 2nd TSP of R5C112 was called NDD in the field ovaluation,0.48 volts by roovaluation of the field data and 0.25 volt for the post pull evaluation, The burst opening is shown in Figuro 6 9 for a burst pressuro of 9,700 psi. It is soon in the figure that the tubo burst above the TSP location and thus should correspond to an undegraded tubo burst pressure. The expected range of but.t pressures for undograded tubing is 10,60012,000 psi. For the near mean mntorial proporties for this tubo (800 Tablo 6 3), the undegraded burst pressure would be expected to be ' about 11,000 psl. As described abovo, an adjustment factor of 1.12 was found betwoon tho . Westinghouse and destructivo exam burst test on an undegraded tube section. Thus the burst pressure for the 2nd TSP of R5C112 should be increased by at least 10%. The flow stress and burst pressure adjustmonts load to a burst pressure of 10,880 psi as shown in Tablo 6-4. ' R10CG. TSP 2 6-6
. . . . . . . . . _ . _ . . . _ . . . _ . _ . _ _ . . _i
i The 2nd TSP of R1006 had a 1.46 volt bobbin Indication which increased to 2,07 volts la the post pullinspection. Tho burst crack opening is shown in the upper part of Figure 6 " Li two
, magnifications. The crack oponhig is about 0.33 inch long with minor bulging or tearing. Figuro 611 shows the OD crack map and associated depths. The macrocrack that opened in the burst lost is about 0.33 inch long with a maximum depth of 72% The expected burst pressure for a r +
0.33 Inch long crack conservatively assuming an averago depth of 72% would be about 7260 psi or significantly in excess of the measured 6,000 psi, it is concluded that the reported burst . pressure underestimatos a comploto burst by at least 15% The 15% adjustment factor is typical of that found upon repoat burst tests following crack opening with no significant crack I tearing and onvolopos tho abovo discussion for a 10% factor. As shown in Table 6-4, the adjusted burst pressure for the 2nd TSP of R10C6 is 7,100 psl for use in the voltage / burst correlation. ' R10Cet TSP 3 The 3rd TSP of R10C6 had a 1.31 volt bobbin indication which increased to 5.34 volts in the post pull examin1 tion. Figure 610 shows the burs! crack opening for this indication. The burst opening longth is about 0.38 inch with a maximum depth of ES% Similar to the 2nd TSP , for this tubo, a minimum increase of 15% in the measured burst pressure of 4,850 psiis appropriato for this indication. Tho adjusted burst pressum for the burst data baso is then 5,740 psi as shown in Tablo 6-4. Cinck Merchology Figuros 6 2,6 5,6 7 and 611 show availablo OD crack maps and associated maximum depths found in the tubo examination. Those figuros also show regions on the tube which wore
~
charactorized in the destructivo examination as lGA. Tno IGA depth was generally nogligiblo (<5%). However, the 3rd TSP of R7C47 was identified as having very local IGA depths up to the ' 5175% rango as shown in Figuro 6 7. The IGA charactorization used to defino the OD crack maps is not known, A review of the metallography data indicatos negligib!o volumetric IGA Involvement. As described in Soction 3.2, the Catawba 1 pullod tubo crack morphology can be classified as multiple ODSCC with minor IGA. Summarv of Catawba 1 Pulted Tube Results i Based on the above assessment of the Catawba 1 pulled tubo data, the ovaluated results are given in Tablo 6 5. The NDE data evaluation supporting Table 6 5 is given in Section 5.6. P 6.7 Evaluation of Fiant E-4 Data Rocont (1992) tubo pulls from Plant E-4 provide a major contribution to the 3/4 inch tubing burst pressure and loak rato data base. In addition, the oddy current data were obtained to the Belgian and APC voltage normalizations to provide the basis, as described in Section 5.7, to
, convert prior and future Belgian data to the APC data base. In Section 5.7, the results of cross -
calibration of Belgian (EDM holes) and domestic (drilled holes) ASME calibration standards are discussed. The cross calibration factor of 1.8 can be applied to the Belgian data for APC applications. However, Laboroloc is continuing further studios on voltage normalization to APC guidelinos. Ponding finalization of the Laborotoc study, a factor of 1.5 increaso is conservatively applied to the Plant E-4 voltages. The Plant E-4 data are described in this 67 e e,---.m.+w-..- +-,---,-,-+-y,*.w+,-,-- ,ry.r.. --,.,yv3+-,v - , &.e w - r r ,-w-,---w.n.- - . , , ,. ,--a.w , e----- -. ~.--- - .+ + - ~- - , . .-= = ee- +- =
1 section. I Leak rato and burst test measutoments woro performed on the Plant E 4 pulled tubes as , summarized in Table 6-6. Theso data includo bobbin voltages up to - 10 volts which are highor . than obtalnod for other 3/4 inch pulled tubes. The Plant E-4 burst tests woro performed with a plastic bladder and no foil reinforcement. The - ! burst lost results showed tearing and are considered to require no adjustmonts to burst plossures other than the adjustment for material proporties. Free span burst pressures were obtained for six intersections with bobbin Indications and one NDD intersection. The loak rate measurements are also given in Tablo 6 6. This table includes tubo R19C35 which had been previously (1991) pulled and examined. Loak rates were measured in free span at tonm temperature. The Plant E-4 leak rate measutomonts woro mado at room temperature at 1450 to 1525 psig and 2400 to 2750 psig for normal operating and SLB differential prossures, respectivo!v. Laboroloc has applied an analytical procedure using measured leak rato dependence on pressure differentials to adjust the room temperatt ro test results to prototypic temperatures and pressure difforentials. Only the leak rato data are used for R19C35 and R45C54 TSP 2 in the APC data baso. Ponding review of the analytical proceduto for the leak rato ad;ustment, the 3 leak rate data are not used in the referenco correlation of this report. l: 6.8 Evaluation of Plant B.1 Pullod Tubes Bobbin and destructivo examination data are available for 16 intersections from Plant B Units 1 - and 2 pulled tubos. However, only the 5th TSP intersection of R4 C61, Un:t 1 was burst tosted and this data point is described in this section. The bobbin data was obtained at a 550/100 kHz , mix normalized to 2.75 volts for tho mix at the 20% ASME hole. The 550/100 kHz mix la sufficiently close to the 550/130 kHz mix of the APC normalization such that no voltage adjustment is necessary. The pre pull field bobbin voltage for thls indicat!on was 1.91 volts and the depth was 74%. The post. pull bobbin data was 2.33 volts and 80% depth, Tube R4C61 at the 5th TSP was burst testod with no bladder and inside a TSP simulant (0.75 inch long,0.016 inch diamotral gap). No leakage was detected (by less at pressure) until the crack opened to a largo leak rate and loss of pressure at 6750 psi. The ir'itial crack was found by destructive exam to be 0.40 inch long with a 0.01 inch long throughwall penetration. Given the throughwall penetration and that leak rates woro not measured with significant accuracy, this indication is not used in the APC leak rate database. The post-burst erack had minor opening of tha crack faces with negligible tearing at tho edges of the crack. The max l mum change in tube diamotor as a result of the burst test was 1.3% OD or about 0.010 inch which is loss than the 0.016 inch diametral c!carance in the simulated TSP. Thus there is no apparent influence of the TSP on the leak / burst lost such that the data point can be used as a lower bound to the imrst
- pressure.
No metallography was performed on the axialindications at tho 5th TSP. A mapping of the OD
' indications was obtalnod visually following the burst test. The axialindications are typical .
ODSCC with negligibte IGA involvement. Short circumforential branch Indications show more IGA involvement at the faces of the cracks. The largest axial macrocrack was examined by SEM fractography and found to bo 0.4 inch long with 0.01 inch throughwall penetration. The crack - - l was nearly throughwall for a 0.1 inch length. Savon individual microcracks comprising the l l 68 -
i macrocrack had mostly grown together by corrosion w!th only par 1; ally uncorroded ligaments tomalning. The maximum depth found in the circumforentici branching cracks was 46% throughwall. ; 6.9 Growth Rato Tronds !
; For implementation of attornato plugging criteria (APC)in the range of 2.5 4.0 volt repair !
limits, growth rato dependonce on BOC voltago amplitude becomes important to establish the repair limits. This results as current domestic plugging limits result in little data in the higher range of voltage amplitudos noar the APC repair limits. For the Catawba 1 Interim plugging critoria (IPC) limit of 1.0 volt, the growth rato data developed in Section 9.5 do not requito any extrapolation to higher BOC voltagos. It may be noted that the BOC voltages for Catawba 1 cvor . the last two cyclos exccoded the 0.0 to 1.0 vo!! range, in soveral cases. It is desirable to comparo j the Catawba.1 average growth rato trends with other domestic plants and with European plants, i This comparison is provided to show that the Catawba 1 growth data are comparable to other domestic plants for percontage growth with a trend for percentago growth to decrease with increasing 800 voltages. French and Belgian plants, which operate with highor voltage indications in service duo to differencos in plugging critoria, tend to show loss dependence of percentage growth on 800 amplitudes. Availabio French data (Plant H 1) indicato porcent growth rates nearly indopondent of inillal i amplitudo (WCAP 12871). Belgian growth data from Plant E 4 have not boon ovaluated for percentage growth although tho trends appear similar to the French units. For Plant E 4 BOC - : amplitudos in the rango of about 0.5 (0.1 volt Belgian) to about 3 volts (0.6 volt Belglan), the ; average growth increase in amplitude was about 2.8 volts (0.57 volt Belgian). No strong trend
- of growth dopondonco on initial amplitudo was found although a linear fit to the broad scattor of growth data indicate a trend for the change in voltage to increase with amplitude. Overall, the European plants operate with higher voltage amplitudes in service and with trends toward higher growth ratos than domestic plants. "
The Catawba 1 percentage growth trends (doveloped in Section 9.5) are compared with other domestic plants and French Plant H 1 in Figuro 612. This figure shows that the Catawba 1 i growth rates are comparable in magnitudo and dependence on 800 amplitude with other domestic , plants. The other domestic plants shown in Figure 612 have 7/8 inch diameter tubing. i 6.10 Summary of Pulled Tubo Test Results , , Based on tho above evaluations, the 3/4 inch tubo diamotor, pulled tubo data for app 9 cation in tuce burst and leak rate correlations is summarized in Tablo 6-7. Catawba 1 tubo R5C112, ; TSP 3 is not included as both tho burst and leak rate measuromonts are not considered reliable i
= as noted previously in Section 6.6. Three Catawba 1 voltage and burst points are applicable to -
the voltage / burst correlation. 'ho Belgian Plant E 4 data providos 7 burst data points and 9 l SLB leak rate data points (8 points with >0 leakage). The bobbin coil voltages are shown without
, and with the 1.5 factor on voltages (Soction 5.7) applied as a preliminary adjustment in this !
report. As noted in Section 6.7, the Belgia.110ak rate data have not been included in the - : reference leak rato correlation of this report, ilis shown in Section 7.4 that the Belglan data with the adjusted voltages and leak ratos are in generally good agreement with the model boiler data. i j 6-9 i i: 1-
The overall pulled lubo database having bobbin voltages and destructivo exaf nination depths for 3/4 inch tubing is shown in Figura 613. For comparisons, the equivalent 7/8 Inch data is i shown in Figure 614 and the 3/4 inch and 7/8 inch data ate combined in Figuro G 15 A voltags reduction f actor of 1.36 (Soction 5.5) was appliori to the 7/8 inch data for commrison * - with the J/4 inch data. Overall, the data sets are comr .f able in size and gonoral tror oward i higher voltagos at incroating depth. The European puhed tubo data show a number of pullod tubes in the 10 to 30 volt rango , a h t 6 [ i 4 1 0
'I l
i e b 6-10 t i
,-,,m,----w.,,,w,---~. ,.--,-,-,-~..w,.. .- - - , - - . - ~ - . .~u,. ..,--w._-
t i Tablo 61 . Number of Pulled Tubes with NDE and Destructive Exam Data f Number of Intersections ! Number Burst Leak Destructive fSn1 of Tubes Terted Ierted ham 3/4 inch Pulled Tubo Data Dano Summary Catawba 1 5 4 (5)(1) O(8)(2) 9 E4 9 7 9 13 B1 1 1 0(i)(2) 5 B2 3 0 0 11 0-2 2 0 0 4 i Totals 20 12 9 42 7/8 inch Pulled Tube Data Base Summary A1 1 0 0 1 A2 4 3 3 4 D2 5 3 0(3)(2) 11 L 8 21 0(22)(2) 23 P1 2 2 0(3)(2) 3 J1 9 1 5 13 , Totals 29 30 8 55 Notes:
- 1) Number in parentheses represents number of additional pressurization tests performed without completo burst. Data not included in data base.
- 2) Number in parentheses represents room temperature pressure tests performed with no identliled leakago at pressuro difforentials excooding SLB conditions.
One additional Catawba-1 tube was loak tested but throughwall penetration is likely the result of tube pulling and results are not included in data base. I 1 6- 11
~- . . - . . , ., _. - - - , - . - - _ . _ . . . . _ _ . . _ - - - . . _ - _ . . - - .
, . _ . _ _ . ~ . _ . . - _ . _ _ _ . _ . _ . - _ _ . - _ _ _ . _ . _ -
Tab!o 6 2 Field Experienco: Suspected Tubo 1.eakago for ODSCC at TSPs(1) Bobbin Coif PJaD1 intcect!cn Volts (3/4" Tubinoi pepth Comments
- B.1: Outago following 7.7 92 % Total plant leak rato varied from R22C58 suspected leak -19 near beginning of cycle to
-63 ppd at end of cycle prior to inspection. Other tubos with PWSCC contributed to the total leakago. Based on hollum leak test results,it was estimated that tho ODSCC leak rate was <10 gpd.
E-4: Outage following - 6.2(1.4)(2) 75 % - Totalleak rato from_SG was -149 R11087 suspected leak gpd. Fivo leaking tubos woro detected by fluorescolno leak test . at -500 psl. EC data obtained I after loa 6, test indicated that 3 R17C50 Outage following 20.0(4.2)(2) 80% leakers woro affected by roll suspected leak 'Jaqsition PWSCC and 2 by TSP ; ODSCC. From visualobservation. - the largost leaker showod slight drlpping and was associated whh , PWSCC, UT Inspection indicated one axial crack in each tube of " longth 12 and 11 mm (0.47 and 0.43 inch), respectively. Notes-
- 1) Field experience noted is for nominal 0.75 inch OD tubing with 0.043 incii wall thickness No data are known to be available for tubos with 0.875 inch OD.
- 2) Field vohagos of 1.4 and 4.2 volts, as given in parentheses, were obtained at 300 kHz with Bolglan normalization. Voltages converted to 3/4 inch tubing i normalization of this report utilizing Figure 5-4.
S i 6+12
- _ . _ __ __ _._.~._e__ .
i i f i i 1 Tcblo63 ; i
, Tensite Strength Proporties for 3/4 inch Diameter Tubing l i
Source of Tubina Sy.Yiold Strenoth.Krl Su-U! tim. Stronath fsj Sy4 Su.Ksl Room Temo. fd20E Room Temo- f52 E Room Temo, cssoE , i Tubing Manufacturing Data: Mean 53.05 45.78 101.29 97.35 154.34 143.13 Standard Deviation 4.86 3.91 4.22 3.97 8.28 7,13 Lower 95% Tolerance 44.55 38.95 93.92 90.40 139.85 130.65 Modo! Bollor Samplos 54.2 - 109.4 - 163.6 - Catawba 1 Pulled Tubes - RSC112 52.3 -- 98.9 - 151.2 - R10C6 49.7 -- 99.5 - 149.2 - R10C69 53.7 -- 101.5 - 154.2 - l R20C46 54.2 - 103.4 - 157.6 - R7C47 52.4 - 103.4 -- 155.8 - i Plant E 4 Pulled Tubes R26C34 53 49 100 93 153 142 R16C31 60 59 112 108 172 167 R40C47 46 46 101 101 147 147 -! R45C54' 54 44 97 22 151 136 R47C66 . 51 40 97 91 148 131 R33C96' 54 44 97 92 151 136 i, Plant D 1 Pulled Tubes t R4C61 52.0 - 101.0 -- 153,0 -- i i l Same Heat e J 6 13
.w---- --",,-,=-,,n,,m. .,,.,a,,,,,,-w.,e~-,-,,www,,,n-, . , , . . . .- m ,,v,.,,o,--w.,,,,,,,w.. w,,,.-, , , ,,...,,,,4 -,.,n,-.--- . - ~ ~ -,.nw-.,- , . ~,,,,-v-- ,
. . - - . .. - - - . _. - - - - . . . - - . _ - - - - . ~ - - - -
Tabte 6.4 Durst Pressures for Catawoa.1 Pulled Tubes Flow Crack Stress Opening - I Bobbin Measured Adjust. Adjust. Mjusted J.ubc LSP. Y2!!1 Burttetl Factor Fador- Buret osi Comments I RSC112 2 0.48 9,700 1.02 1.10 10,880 Ductile, fishmouth rupture just outside TSP. 3 1.02 4,150 1.02 > 1.25 Unreliable Crack opened (largest -0.1", others <0.05"), no apparent bulging or tearing. Max. , corrosion depth 97% Mar. single macrocrack length -0.43 . (<0.2" TW by burst). R10C6 2 1.46 0,000 1.03 1.15 7.100 Crack opened, minor bulging l or tearing. Maximum corrosion depth -72% Burst crack - macrocrack length -0.33". 3 1.31 4,850 1.03 1,15 5,740 Crack opened, minor bulging or tearing. Maximum corrosion , depth -85% Burst crack length
-0.43". -
H10069 2 NDD 9,400 1.0 1.10 10,340 Ductile, fishmouth rupture TSP region. 3 1.48 5,000 1.0 >1.25 Unreliable Minor crack opening, no bulging or teating. Maximum corrosion depth ~75% Maximum singte macrocrack length -0.37" (not completely or < 2" opened by burst), R20C40 2 0.42 8,600 0.98 - Unreliable Both R20C46 intersections burst just above TSP at a hand 3 0.79 7,200 0.98 - Unreliable held grinding tool mark applied for location purposes. R7C47 3 1.57 5,800 0.99 >1.25 Unreliable - Minor crack opening, no apparent bulging or tearing. Maximum corrosion depth
-87% Maximum single microcrack length (-0.05" opened by burst) ~0.44". ,
l i 6- 14 i
- . _ . . . _ _ . _ - . . - . _ . . _ . . _ . . _ _ . . _ . - - . . - . . - . . , _ . . - , - . . . . . - - . . - _ . ~ , - . . . . . . ,.._ - . . . _ . > . , _ , - , . . _ _ _ . . . _ , - - , _ ,
Table 6 5 Summary of Catawba 1 Pulled Tubo Results
. MLettinecure fwd Dd1Evat Lab Dertrucive Eram Bobb'n RPC Post Puti Max. Durt,t Leat Rate fi'hrt
.Ide. ISE Yds h2db Wdi D.C.Ve"s bd Lereth m Uprm.Oo. SLD R5c112 2 0 48 -0% 0 37 Super 1cial 10.880 0.0(3) 00(3) 3 1.82 BG% 1.30 5.06 07%(2) 0.50' N.R.(1) N R.(1) tJR.(1)
R10C0 2 1.4 G 83% 0 08 2.07 72 % 0.40' 7,100 0.0(3,5 g.o(3) 3 1.31 70% 1.20 5.34 85% 0.43* 5.740 0.0(3) ;3.o(3) R10C09 2 tJDD PEO Nano 10,340 0.0(3) 0.0(3) 3 1.48 72% 0.97 3.31 75 % 0.45* tJ R (1) 0.0(3) 0O(3) R20C46 2 0 42 30% 0 82 12% 0.05* N R.(1) 0,0(3) 0.0(3)
, 3 0 79 28 % 1.04 -0% fJ R.(1) 0.0(3, o.o(3)
. R7C47 3 1.57 78 % 1.40 4.13 85% 0 42' N R.(' O.0(3) oo(3)
Notos: 1. N.R. . Not Rel:abio. _
- 2. Evaluation indications crack opening for leakage may have resulted during tube pull.
- 3. No leak identifed during room torrporature pressurization tests.
s e 6 15
l 1 Table 6 6 ! Plant E 4 Pulled Tube Burst Pressures and Leak Rates ; Bobbin Actual Leak Ratos flhr) Burst Prorsures
. Tube. ISE h Max Decih No'm Oc (1305 oti) SLB fM10 os!) Measured Adi.for Mat Preo. . .
L
.9 R26C34 3 R16C31 2 l
3 i R40047 2 R45C54 2 3 R47C66 2 3 , l 4 R33C06 2 R19C35 2 R26C47 2 Notes: 1. Belgian voltage converted to APC volts using Figure 5 4.
- 2. Burst test conducted inside TSP TSP constraint judged to have influenced the burst pressure and thus the burst value is not included in APC data base.
- 3. Room temperature test results given in parentheses were adjusted to operating temperatures (without parentheses) by Laborelec.
- 4. Bobbin voltages do not iriclude any adjustments to Belgian voltages. ;
- s 6-16
Table 0 7
. Pulled Tube Leak Rate and Burst Test Measurements for 3/4 inch Diameter Tubing 4
- uit / Borbin 0011 RPC Destructive Eram Lenk Rate thr- Burst (7)
./ ECnj ISE Mchs(1) DeDth ygts Max. Death Length (2) N.Ocer. SLQ Pressure l
Catawba 1: R5C112 2 0.48 -0%
~
Superfcial - - - 10,880 Rt006 2 1.46 83% 0.98 - 72% 0.40 0.09) 0.00) 7.100 3 1.31 76% 1.20 85 % 0.43 0.0 0.0 5,740 Plant E 4: p
*9 R26C34 3 R16C31 2 3
R45C54 2 3 R47C66 2 R33C96 2 ; R19C35 2 . R26C47 2 Plant B 1: R4C61 5l . L - Notes
- 1. Voltage normalization for 550/130KHz to 2.75 volts on 20% ASME holes. For the Be8;ian date, 1 votteges include 1.809 f actor for cross calibration of Belgian made ASME standard (EDM 20%
holes) to the reference laboratory standard (drilled 20% holes). Voltages with cross calibratbn-factor reduced to 1.5, as used in this repor1. pending completion of Laborelec study are shown in parenthesis.
- 2. Maximum burst crack corrosion length in inches with throughwall length in parentheses.
. 3. Tested at room temperatute,
- 4. Not measured at 550/130 KHz. Voltage renormalized from 300 KHz data.
- 5. Leak rates measured at room temperature conditions and analytically adjusted to operating cond6tions.
- 6. Observed during burst test at room temperatura.
- 7. Burst pressures adjusted to mean flow stress of 77 ksi.
6 17 i
, - - . . - ~ - ~,~....,.-..--.a >sa---n--.e~-,-,-,- -,,.r~, - - , - - , - - ~ x. 2 rs r~- - a, ., r-. . - . - - . - - ---
y - - ,- - - - - - - - - mum.a w -
- m.-.,.um.,-w_w--m--a---n.-~-_,--.,-mm t
l i
; .s / , .. . * ... em3a..
h ba q .. "[ f -> > 1, .
}.3 v .C-l e A ;. - $- 9 (f) .
., 'lr * * ,. i v if"i f-l{h.,'J &$'t '*
'wr.or. sg
. l N.11
,a l
I i L. i l ' 6 ' (
'I' T.
v y; g e i l r/. .
,, v M
, ,, $ J . td I
~
s ef Om">WNk s 6'g,f* U l :%AM.-*r o..ja h .c,o .rc -
,t..l;I %' y 1 :
q :s . 3:
, _ C :.. :, t '
,5 ~
, r) * ' ',?;,.,,*g . $l - h Y i k tt it ,y- c 'd.
1 o g-g w. y;,'
<t -
i '- t; o E a @' : P hk * . U * . .
* *9 l CD ,
- Q i .
3 : sm ~ ..t.* , . s .. - %w.. e, , , . , u:,e &. l l 91 l ee' . , 7 l .
,s y
\
i 5 * - o i
* ^ *
+
'b . ..
~ ,
' p '- .
, l -
st !
. . y t; I F~{ %} R;?L_. .
- n. .
g:
- i' i)5".
l ,
. ., '4 g ., , .,
l af qW. * ., , ..h. ~'
' *.; ) - p p , W~s!f,T
- f. . . >
l , . . .: - - l %; ' b . b - ? t. I;, O.)}?&j^~.,9,y;s h , g ,
-c. A, ...
{ :. 4 -
,[. i . ,
, .. . _'? '
, , .' 1-[. in \h,# v !
.:8
-=... ;
. . 4. g !
, l ' ' , " -
g, 1. . . . '\h / i U . , ; g !
. - V
<.. g,-
e
- . g4
( l b [ j I Q (r ) , C]!]//h3 l PQr,*/d Of O &,) 12, T S ' T hQ.[k BI: fore and Attor Burst Test i )
' ~ _ ..-- .-----.-.- -- - _.- -- ... .-.. ,.- - - -. - __.-. . _ . _ ..-.,.. . _ n _ _ , _ - _ -., _ _ ,~.., _
g, ,- -m--..* -m ..- - - - c- ..--.---.. - - ___m - - -- --.- - y.- --. -
- j% ga g g_4 g. a g.pmmuggg, 1-(
I 0 a P c 1 = 1 h d a f i h I 1 : i 8 [ a i 1 , 4 J f: 4 e { 4 i i a t 4 4 1 ; e 6 a
?
f
- I e
, 4 h
e a t a. e 3 1 ! i ? 1 1 l
..l l
= !
f 5 i l l
'l 1
1
-. . . . - .-. -----_ . . .. .._. '"#NMWWWW-e c ep g, , _
""*MPewww m gm ,
"Nrew,,n.ewg._,,, ._ ,
I 1 lia l i s i l' s I s gj . j 2
. I YYMMYTYYMY MYYYYYY F -
__ ,mn:1 :- m... . r=.; i c : : ~$ 1 74_ _w. . . . . . . . m. .=_
. m . t i ,
m..,y,,._._.._.___.
._ ._. - _ _ /_,._ .
,._iL, , , . _.._ _.._s_ _ , _ . . . _ _ _ . . . . _ . . -. .
e _y=,...-,.,~ .. , ._ _....- =. . .
.. .t. .. ..i n ..
i
+ { }.-{ .i_ - .t.._....... I
.IL 7.rT. i .*
..L.t,.._.i ,.. 7e, _._
!. ~ .- i , .
_e. i e tw a g.=f. !* , L_.. . f.i
= _ _ _ . _ . . .
o ,l A.i.! . _.4. ( ,1. . . . _c.... .- i ; ; L4_. i p_- .
, i.1 i.s.1.! ...._i_w.t
. ! m;J.. .
... ne i . _ . . =
j.. '. b.-e: Meh%.{.4.gL'Qg
._ _ ___ _ ___y - ..._
_R (D .____ i.4 ..j_.
/..! t....H H.2..L.
i 47.t,..wjg.yj.
,-. . i. i 6..
aZ^ U 8
. 7.
f [d + j f f l f, .hn
@ _h,/..; s ..,f-M..h.f .. j im_. di ; i .14g 4 h, _. ..
s .
. . _ . i
. y.
[t ; O g ; '....p..g . .n t_{.a,!.y: s' > > I_l,._1; .. i 1 . i .. cy ___ ,/_.q...................._
.t i : 1 1 i q_ n_m . . .,i -
i i i e t ,
.3 i3
,O o
1== ; _p.4.q . p j .j 4.7.._..
; 7.q _.m-- -M-n --
N -& 3 q.. _
- 4. 4 )_ _..<
T'"' 1 (...... .i.I.d..k). d ' 4, L; 4 , 3 4..d . .. - 5 i i ......
.i .t . ;: ;._t- '...A'.p' . 7l_.p 1 i J
,o 3
; . ,i .p , .i t I . . ,,
.c en i .' t . ...,_ ._ .i.
i_.
,1_ _.i.i9 4 I
} w cn c - _
s--
- r. ..i ,:.. 1. ...e
,i i i g~! l { j i ,
Q oum
,- i , ,-
y L mx L;;..L ..._
...1..;
'i 4 : .;.;
- i. .._I ,,
.g i
y - _. r c>
- t. cp 8 i
- p. . ,i. 7..
j , L,
; y_ i i .i 4_ ._q...l
- j. 2 L L L..q ..' . .i.p.....L . .;, - _ . . _
_ aye ;: LL._L! 1 - Q .. g;o;c m, a o [_LL;...i..L.W i
- .i _.! .- _. .> . t ;.
i, l!;.LLI.p._I
, , i - i i.
l
..k-54 .___
u, H ,t_4'
! iil l i i I I I .I i 1 _. _ .
v o i i
! .! ! i 4 .I. ir-i +-+:: !.b' !l I"- -4i it-4!i' i--+.i .
fm m. - 4 L. .J f i L.p3....' . _. i_..l ;. !..!.a l i !
'#.._---- 'l
~
d..!_._.
._. ._. ,I~ a '~ j _
[~I ~ ~ . _ _ _ _ __ *g _._I.~I_ii
.. - .t -
.!. ., * ,! --.i i .- l"..~. l ! ' I! ;
i 1. i
;i. i
~r- .
3 i:
~ ~ ~
i i .ee
-h : ' . . .
j.
.k A A A..! .;e ;_ .;,_1 # i t _L..
!
- A.p.j_,;i i 1 i!! . i iii t ; i l
; h-jl--.' .r;....i. r.t N
^
.~
! !l - . . i i e !
' I'
*! t--*n i i 1 ie i' ii #
-b W i i i . . i i
$ [ $ $ $ $ $ $V W
6 i (sliW) NOI.l.lSOd 7VlXV Figure 6 2. Catawba-1 Pci!ad Tube RSC112, TSP-3: Incremental Grind ar.d Polish Results J LT
.1 a
i Tube RSC112, TSP 3: Crack Depth Prof'te A 0*/, - - Tube 0.D. S . i ' o e 1 Ch ' ? * * .. 6 10% w , ia Ei n 20% i g 6 - g - !! ,. .
. c o
Q 30% ~ j
~
y ~ ~ ^ ~ '
'b ~ Azimuthal Ligament (8*-10*) *
,== 2: . . . . B
'.# etween Opened Microcrads
~
O. E 40% ,
~
-4 e ,
Q-E * - + - . i . .. : o g 50%
~ ' ' '
"A
^ I. . .
3 3: '
. :S '; :. .I : VV../.1 .
O e - - - a . 2. .
.l. .
.o p" s 60%
y . y * (n O _ 7 , m e 70% 0 m Length Opened by Burst Test @4,150 psi; '
;f e P Length - 0.19' Separated by Ligament. :,' -
. f .
h $ 80% -. Longest Single Open Crack -0.11 inch Q..-..
- g. .
- N.:. , . . , . . . ..
o 90%
- i. : ..i. N~ n Q - ' '
3 . . . . a..
\ h .
. T h
m a a 1To m o m Tube 1.0. o e h ,
' ~
o e o m o m o N N "
- o. o. o. .
- u u 9 5 9 9 9 o' o o 6 6 6 Axial Distance Referenced to Approximate Max. Penetration Location (In.)
_ _ . . . -2.--.a. --_a._iaxe,_m.--.a..w..Ah.4,.-m_m4,ow.m, .w.e-.-- 4-64 .-mi.-s. dea. - - - - m-W._--mea w a Aw3_uAmamme .mias 2._ _ _ a_ , _ . _,a.___..-w _aJ _
.y'
)., . > *~ ' v
s :c; c. e c. n s .f}/h,.,lI,,jy;4 ., e.,
' 7 A
- G.&: n..,~tk(..'
iJ.)'y, .. . fg.? . sc-
' ~.' 't s * ;
l t , l e i
- 4. . E ;
p'y j ' ,y{ / , , I
- &y
. :l. I ?
v 4 yi 9@- i i g 1 o"y , : s . .r1 l n 'i9 \ , Qj l* . fd l
\, -
a g
. 1, /o is b W k . l l S --
h '@ . s' f y .--- _ g.,' E, ( a- -----
. m.. ,~ s
. i y r. - - - - - - -
.+ , o . <
! W b ' A f Q '
.d4 -
g,[/ [ sx v 8 -4 3 g! yhi hG ' ($ 5 ' lI,l h ge.w ~$ w _k$,,h,'fr' W; i i . ?e,. m ,1 5.; 1 .
\,
. G u WW 6.:v- I,c , ; 4^ i' Q* R.%Qk,44>g1._ __ \ f -O tl ? l 2 .
, s.;., ~ . c,,.a, . . . ..
- .,Mr,
. a
- p- .. .:.
i ,
, s. : i, <<. - :
4l ,, Z s%%. V^ \ n ., w) jfQ f '[ \ .* l
-), V ' h , lf )
j j , "gkf*) 'e N. ~ - ggj : am \f , a C v .w.. m g ; 1,) .4 bW. i
,,ng - $n.r. . ~Q:fQ-c,"l _
n m s.,<j'l no 9 b.
- l. ?.o 7p}[~
w Tyhi ."Sj ,%D$ ib i) h, 4 !l (
f (
~
0 .%,'L % 6 sh '- ^ 6 1 ^ . I
+
21 SPECIMEN 10-69-4B-2E
+ 375 ; TOP.
m ,
=
j . o +300 - 5 l .
& i-
? i
+#~
=
o _. g g/ 1
- onno ,20
- onno , .
- 8. -
, O -* Onno ais y ^
yf [ g f /
- onne as7 a
[ l. g I + 100 - /
- Oetnd ats m 3 / [ /? ..
- Gdad 815 8
w
/. /3 J
-* Onno a14
- onnd ats 7 .
m
.e O
<f D -* Ortws #12 C- i ..
l
'-* o,wus ati a E TSPQ y4 - I
- o,.ws ,,o m m -
/ ; -
4i - ow s. 6 C -
/ [
/./ /
- Gdndas E
/f I
5 V /
- ontws as 2 -im - .
q g( wo w, ,, y [
- ~
%*g' -43g E2
- o,ews ,4
( (h i
%a 17% -e o,,ws ,2 ,
E g 200 - ( l l-i
- onews at 5 i !j Max. Crack Opening RUPT'UFE Moa from 180* so E I
- Mad. 7'5% T # wascMect W
{g
~
' Crack,pettoms are ocer g 375 l ,
II *I l , , I am E 0" 15* 30' 45* 60* 75* 90* 105* 120* 135* 150* 165* 180
" ANGULAR POSmON IGA DEPTH: { < ss { sw -2ss me%-so% ] s1%-7s% > 7ss
l l l l l * .- , . , . . _ ~ _ . . . . . s $ 660 al ' l $ d. ~ . . - l l . W ,, ? u l l d afj l [ *a* ,' %;Le,s
. .h. . . %p l
W,u'e y , . '( (. a (,.Q ga . n ,ge., -.(-
,ge , ,
., ~ [' 1N'/ -Q) '
l
. s . ,p s , g ., [h $k[fD. j . ,
; phe ' < ~ . ,
t A: . tX ey l
---. e. >
~
l .
.- .e
- . 4 ~- -
~
# w= .
..,x o', ..s - ,'. t
.' gg
,8j+. - -
- ., , l' y
! . . s . A., . ..
- 4. .
,, tb
', s -
4' . , g" - +' h b, "Sj. ~ ,x .,
,' s<'.,
,u - _
N . - ' *N l _ g ,: . J ,' ,
%k I i
g , h l Photoaraphs of Crack Opening After Burst Test - fl7C47,3rd TSP l P
. Fiaure 6-6. Catav.ba 1 Puned 'ube R7CC. TSP 3: Crack Opening After Burst Test i
F k
n SPECIMEN 7-47-4B-2E ' o +375 TOP E m : e j i =a( Gnnd #32 i t + ,
- i i-i ? . ....jt . . t . ,..-.......-
,Y i
....; 4 p
+300 - __ ..;.;
.1
.. i _. j i
. ; . .i . ;! .L. -; . ; ..;..
-;.... ;.4.;_;....4 4 . j . [ ; ; ; . . . 134.g
' ~
4g Q w ,, g ._ _. _ . ._ l ._ ,... . p4.a j. p.; j i ;j ..q. ~ j..;..! . . .,.i. .i.. j . 4 ..g 2 4
.._.._............}._.
. +. . . . -. 7 . p . .
4 g ^ Burst Opening Y' t t~ -
- i~ i +
m yl G +200 - TW #0.05 sQ gf ,. 4. g, 4 --.-
-;, ) .
J I //j w g ,3 ; l u ._3 86% Max. Depth F e.i p c - , I g 7 / , 4 a
.-4 _. i 1._ .
. _..p.
. . . + . ....? 3.7. 7 s s ,l lj t 4
, 4
. p __ L l
+100 - (
l } g __ _.. .-
-}_. _
j ... .. , ;
? * ~ '2a n O - -
/ r 4 y i,f.r---~_._t,7._....._t,_.L.i,
. .. g . y . . .. .. . .. } .. h ' ,
I '
?
l4 o_ _s g 73p , h ,,3
-1
- 7. 3. ..........n pp. ; 7. 3 . , . s g O .. ;..j t p; _;.1. j 3 f /' ,
e y p/ 9 n. . . , ,_ -;. . g. .p.. L;g - r
#,, , , po(.-
, .._..n.- ,.. .7 $p \ ' N\'M t' - - ' - i~'
i 4)' +;g. Q L - - I - l' '- ' - + g_ a f -100 - 3
.j.
j _-. .[_
' -]' , Q - yy 2
g anna ,,o g _ 49... 7
,p 7 ,
.p j ... . . . .-:- . .. +. . . .
g .,._ 4 .. . 9.._.. 4 3 .. , .. ._ 3 . .+._ A+.- 4.. , .4 9. . . + . + . . . + _ - . . . . .L-+_
._.. ;. .... g .j_g p._ ._ .q.
g .... g s7Aj 56,1 q..q. j 4. q ..q
- 1. .T. }.. .._.3
... i_ _
.. p ..rp .g... . . 7. _. ,. - .7. 7 .g. ... .-
w
; w m ,3
. .,4_..)7.,.__ _,3, .- . ,o ....g.. ..r; _ .;.. }
g
- . 7. . . .. , -. L t ._4 ._. . .. ; .- ; 4 t j . , _+ . 7 jg.
- }.
a, _._ j . _ . 4. . } _ , . . _ I. LI
. } .3 ..]..._. . ! ...-. m. _L -.
- 1_ .
- .!.i..
q !...i. +..1 _,. g_ _p 9_.._5. 9,. 3, ...1p I_!..!..I. Li
& .y 6
.. ;_ . p_. l . g m ,,
_ _ ___a ,_
...-...,.......p_,I g i i, ,
E, sorrou , 5 ' ^ ! . ,
?
37s , 27o- aoc- 330- seo-s_ o- so- 230 24o-
" ANGULAR POSITION IGA DEPTH: 3 < 5% 3 5%-25% @=%-so% @ ass-7s* E > 7s* -
Catawba #1 Tube R7C47, TSP 3: Crack Depth Profile A 0% Tube 0.D. E - o :
~
y
~
10% o -
/
R _ 20 % I
- /
y cr *
$ $ 30% , s -
E i: . . g E 40% , .x e a 2 .
.: T .. ,
E o
=
a L- - 50 % 5 D
- g . _ : _ .. .
p g 60% . p .~ _ . g y % - P E 70% W 8 6 Q *
- n. 80 %
~ Length Opened h Burst Test @5.800 psi; E m g , Post-Burst ThroughwaB Tearing -0.05 inch. ,
h i NY'i 5 - N ' eg 90 %
...u.. .
>.. .2. .
m . . .. 8 100 % Tubei.D. - o O $ $ k N o N o o o N o N o b o 9 9 9 9 Axial Distance Referenced to Approximate Centerline of TSP (in.)
_ _ _ . _ . . ______. __ __. . _ _ .. . _ _ _ _ _ _ . _ _ _ . _ _ . _ . . _ _ _ _ _ _ _ _ _ _ . . . . _ . . . ~ , - . _ . . _ _ . _ _ . . _ l 9 I f R5C112 - 2nd TSP
$ . I l 2. = 5 8
'
- A W E A B M S T . f h f',- -
N
' i- ],gggy6 97oo9s dss Q
@j{nkpgh;Q% '
{t N. 3
)
.W (r i
} fJ"i= = '
. A ,,,, , N - .r. .
- . ', s J .
~ ..,,.
t . . .
-a -
*.*f._,,.e 1
s.a d y g,y :,* S m. . i i
. e sq . . -
i
..e- , ,v . .
- ., s ' *f ~[ ~ ${j .-; k'L fs!
' - g- . ',
'w. * . . ' . . -
- u t, . ,,, s..
q %. ,% .8 ,9 \ y_ . . . - IE y,gr-, y ,n
- # ,.A ^ i .4, ,
. . . . , ~ +
, _2 . i i l R10C69 - 2nd TSP j . k5 0 ,W. 3 b Af hr barsi . E., 6 g
'h?;hfh'fg, f(WNdi/: #2 Gl* a' ~
& t,3
+ %J,,
, .y.:.y : ;qe g., ..t.;'*g -; m < {' i .g ',<~
?...,
=
i ,s,......
-rv y Q.,,_ ...--
v .1. .. i ey ' - g3p,. , _ . , ,. l l < l I
' ha . ,,Y,
. v% , Pia , ' C $* N *d' E b'
- l l
Photographs of Durst Openings for 2nd TSP- R5C112 and R10CG9 l 1 I i Faure C 9. Cat:1..b:1-1 Pu! 'd Tube R5C112 TSP 2 and R10069, TSP-2: Cra% Cp:ning M!ct Burst Test
aw,a_- a - - - - - - --.- ---~e.n.---~a -----~e***- --saa- - a m- n ======**r.~~w+.+=-o mam""**s--.-- 2hanWM*~m>-^ ----'s~L4"-G~*--~=-^~^'S g , ,, ' ' ?g, .l. ,- f
=
. b.. tip '
d ' ;4g$er
; s , n u .: ~ f,- :
ua 9 .. , , N
, m' > f'.&~; .
'& - w.
~
" -',..p%', -
L
. y ;, L. .? * ' '
l
,, . . .f, . ::(
; u g ' '
,pv'
,A >
p ,;'. .
,r ' .; ,. ,
j ,f , . fjf gg2- l 1 ?' ', t, . (\ .,.:. , . . . ~ Ir a. ] 4 ~@- i , 3. a% "" , ' . g . . . p j #g OL (e'; _,[,, e. '; -
},,
l l 1 (.
}. ,6 ,y % ,d S.
/
4 j, u at, y :. .v g 1 y p;. r \ .s : yg %. . i
\~'. .. .. , , ,
e -
.4- 'f . .
#E a ' . g i., : .,
N b , 4 '* , ' ?,.4 $% #.2.f[ .$. ' * ,. N -
'Yy
? ,A .'.. ,
R > ,. . y.; .. '
.ya 7
- _. e . - :
3
; ,' oo
}
"n
, >r ' E ~E
$ o
- u;i -
- .s
' x i
I T .' O l L 5 ' I g' ,
-- '] ,
n < O j h
* 'Ta.:. q, .
s , 4, i w s
,Y'- ,
, I ((A
- u. m -
, yu m: ;. . :.. .
#.h ;% '
d n 3. y .. i s l.g 1
%:n . ;c J -- , jj . .
3 )i s .,._ , -
- ,1*
V"' * - lNl
.:c
% :c' .-' yn;:.;' -. 4'f ;:g-u<
r v#V)e (D ': O s g_l
~'
~.,,,., 4 w y y[G.,pQ,6p%C - '
y
. W,-i,.:
Q Q'd.' _'hj[ ( 7 .-QAA; U
.L QA se
$! L4l N. c l$2';,f .
kj .l. $ c'q. 4[',7 . .* ' po X tr 3~ ? - N ,.fL ' 12 .I i
;. 3 kIOith . 'Il']~ y ., ~15 .( ;
s F - 7' {j
~
>"y , 4 (
- h- ( ) { j _] f ' ,' J , r -- f - 4
]f} _ ,
! l l ll \i i' Fe yI*
~
p o T v<, . M06 3 ji fr/ i
)
s a
\o i'
k\ 0 r
- 3 i
I 3 > r / 1 i ){ ) E
/! e
!I lt i -: I g
/ w' "0 0
W. 3 i l 4liI
!ll eTWyI %
I 5 ss 7 0 - E 2-lri y e2mf p7 p x. p uaeY 7 2 s i s lM p" r l 4 8 2 N[ 5 O
- I 6- rT so 1 I 2S s .
0 - O - P %s 1 ~ 0 8
? ,(
R 2 N 4 , 1 A E . rU@ L M pCFb:h* 5 G 1 N C i 1 s 7. 3 A ss . 3 r 2 E dni$u i 1 3 i J s 7 s
- 4 11 2
1 P .*O O S m]d e- Tn** 3 *+ 5 S I 9 0 M 0
' 6
_~
~ / %
5
/ "0 <
.'_. 3 l
J V H 5 0 0 0 s T 7 3 0 3 g 0 1 0 0 0 0 0 0 y P
+ + , + 1 2- 3- s. E Q - D P
G A T GI
$b 2hDOO i .
mW8 & ok j $ ? & rf b g j ?"5Sa8E ya"g d 5 $ $
~. . . - . _ _ . _ . _ - . - _ _ ._. - .. - -. ._ . . .
80
- : : U.S. VOLTAGE NORMALEATION Lu : : :
a 7a : : : . . O : : : : : O i ! ! ! !
- 6& i i !" i i H : * : : : :
Z : : : : : w : : : : : O 50- D-t i ! ! i i cc : : . tu : : . u) :
- u -
u - - : : : Q. 4g ----> . n . . . . I =' : : : : : p- : : : : : O m 3& C ! " ! ! ! ! g : : : : : O 2} ! ! i i i
<C F-M i
i i J . O ,} i - 8 : : :
> 1 : :
0 ? i
?
i i I i
?
i 0 1 2 3 4 5 6 INITIAL BOBBIN AMPLITUDE, VOLTS A CATAWBA 1 O PLANT A 1 m PLANT A 2 X PLANT H 1 Figure 6-12. Average Percent Voltage Grow 1h Rates for Catawta-1, Plant A and Plant H-1
t 3/4" Pulled Tube Osts: Bobbin Cell Voltage Vs. Maximum Oepth from-Destrucilve Exam *
.9 2
o Maximum Depth from Destructive Exam " a Plant B 1 0 Plant B 2 ' ' Plant C-2 0 Catawba 1 -. .A Belgian Pulled
. Tubes Figure 6-13. 3/4 Inch Pulled Tube Data: Bobbin Coil Voltege versus -
Maximum Depth from Destructive Examination
l l l 7/8" Pulled Tube Dats: Bobbin Coll Voltsge and Depth from l Destrucilve Exam 9 l l i l
= '
O I l l Maximum Depth From Destructive Exam -- O Plant A o Plant D 0 Plant L e Plant M A Plant N
- Plant P 8 French Pulled Tubes Figure 6-14. 7/8 inch Pulled Tube Data: Bobbin Coil Voltage versus Maximum Depth from Destructive Examination
_ . . . . _ . _ . . _ _ _ . . . _ . _ . . _ . ~ - - _ - _ _ . _ - . _ _ . _ _ - . . . l 3/4" and 7/6" Pulled Tube Osts: . Bobbin 0011 Voltage Vs. Maximum .- Depth from Destructive Exam ;
~
.9 2
ii a d Maximum Depth from Destructive Exam-8 3/4' Pulled Tubes O 7/8" Pulled Tubes
-Note: 7/8" voltages decreased by factor of 1.36 to correspond to 3/4" tubing.
1 Figure 6-15. 3/4 inch and 7/8 inch Pulled Tube Data: Bobbin Coil Voltage ' versus Maximum Depth from Destructive Examination i
?
Section 7 LFJX RATE AND BURST CORRELATIONS 7.1 Introduction This section utilizes the model boller (Section 4) and pulled tube (Section 6) data to develop correlations of burst pressure vs voltage and SLB leak rate vs voltage. The correlations are considered to be preliminary pending resolution of the Belgian voltage normalization (Section 5.7) and review / concurrence on the 3/4 inch tubing database and correlation methods by the EPRI APC Task Team. The EPRI Task Team provides an additionalindependent review of the available database and an industry concurrence on the APC methodology. This process leads to a consistent methodology, particularly fnr the correlations,in developing plant specific APC submittals. The 3/4 inch database and methodalogy have not yet been reviewed by the Task Team. The methods for the SLB leak rate correlation are in the process of being revised to enhance the statistical methodology together with the physics of the correlation. The SLB leak rate correlations presented in this section are a preliminary step in evolving the enhanced correlations. 7.2 Summary of Data Base for 3/4 Inch Tubing The data base developed in Sections 4 to 6 is summarized in Table 7-1 for both the model boiler specimens and the pulled tubes. Data points not used in the correlations, such as zero leakage points with <90% actual crack deptn, are specifically addressed by footnotes. Destructive examination crack lengths in thb table are preliminary pending more detailed examination of the data. 7.3 Burst Pressure vs. Voltage Correlation The data set of 3/4 X 0.043 inch tubing contains 57 possible burst pressure versus bobbin voltage data points from mode! boiler (MB) samples and pulled tube intersections. Of the pulled tube data, the voltage amplitudes of samples from the Belgian Plant E-4 are adjusted by a factor of 1.5 to put them on a comparable basis with U.S. voltage normalization. The conservatisrn of the 1.5 factor is discussed further in Section 5 of this report. In addition, some of the pulled tube burst samples from Catawba-1 (not listed in Table 7-1) do not in our judgment reflect the characteristics of a tube that has experienced burst or in fact had burst in regions outside of the degraded region (see discussion in Section 6.6). For these reasons, not all of the Catawba 1 data are atilized in the recommended correlation. However, to provide a basis for judging the son: Jvity of the correlation to these data, correlations are provided in the following that include these data (see Table 6-4 for values). The burst pressures of all the room temperature data are normalized to the mean of the flow
, stress,77 ksi, for the mill annealed Alloy 600 tubing at room temperature (WCAP-12522). A second order regression analysis is performed on the data to obtain a best fit and the associated error of the estimate. Based on the standard error, prediction interval methodology is applied to establish the lower 95% and 99% probability intervals (one-sided). These lines are '. hen ratioed to reflect the effect of operating temperature and lower tolerance limit (LTL) strength properties (WCAP-12522).
7-1 l l
1 i Pressure ratios at beginnir g c,f cycle (BOC) and end of cycle (EOC) relative to 3APNO and APSLB are used in Section 10 to define pressure margin ratios and to provide comparisons with 7/8 inch tubing. This section devebps burst pressure /vc;tage values to support analyses describec'
'in Section 10. Interim plugging criteria comparable with that approved previously for 7/8 x - -
0.050 inch tubing is accomplished with a burst capability at the lower 95% probability of , 5290 psi for Catawba at BOC and verification of 3750 psi capability (3aPNO) at EOC (1.4 AP SLB is enveloped), using 90% cumulative probability values of EC error and degradation-growth. In addition verification of end of cycle APSLB capability (2650 psi) is required with the lower 99% probability (LTL) curve and with maximum EOC voltage. Recommended Correlation The burst data of Section 7.2 provide a total of 57 data points that have been scaled per the above discussion and used to develop a correlation between bobbin coil voltage and burst pressure. The scaled data used in the correlation for 3/4 inch tubing are listed in Table 7.2 A higher order regression analysis of this data has been performed providing an equation for the
- mean curve using a second order polynomial equation. The equation for burst pressure (BP) as a function of volts (v) obtained is:
[ ]S The coefficient of correlation for this regression fit is 0.87 and the error of the BP estimate is 1.13. A -95% prediction intervalis established using the expression: [ ]S where, [
]S l The recommended burst pressure versus bobbin voltage correlation is shown in Figures 71 through 7-3. In Figure 71, the voltages (4.1 volts and 10.95 volts, respectively) corresponding to 3 APNO (3750 psi) and APSLB (2650 psi) are presented. These voltages represent the values that would form the basis for an attemate plugging criteria. The EOC voltage value of 4.1 volts results in an EOC burst pressure capability that meets R.G.1.121.
Margin to burst of cracked tubing (throughwall cracking in the limit) is a direct function of crack length, applied pressure, flow stress and radius to thickness ratio. .With the others remaining equal, the applied pressure then dictates margin. Consequently, the most limiting - case will envelope all other cases and demonstrating that the most limiting case is satisfied ! envelopes allothercases. 3 AP NO si limiting and enveloping since it is greater than 1.4 APSLB-i I l 7- 2 i
l Figure 7 2 illustrates the burst strength of 5290 psi corresponding to the BOC amplitude of 1.0 volt, and 4740 psi corresponding to the 90% cumulative probability EOC voltage of 1.66 volts. The 4740 psi EOC capability exceeds 3750 psi (limliing and enveloping case) by a wide margin.
. At the maximum EOC voltage,2.53 volts, a burst capability of 3580 psi is illustrated for a lower 99% probability in Figure 7 3. Therefore, the probability of rupture due to 2650 psiis (much)less than 1X10-2. The computed EOC probability of rupture at 2650 psi is 1.1X10 5 11 can be noted that the 2.53 volt maximum EOC voltage corresponds to 99.94% cumulative probnbility of the Monte Carlo projected EOC voltage distribution.
Sensitivity to Catawba Data The first assessment of serisitivity with regard to treatment of the burst pressures of the Catawba pulled tube intersections is to eliminate the 10 to 15% increase applied to the three data points incorporated in the recommended correlation described above.The equation of BP as a function of V for this data set is: [ ]" The coefficient of correlation for this fit is 0.86 with a standard error of 1.15. Figures 7 4 through 7-6 provide results that correspond to Figures 71 through 7-3, respectively. The EOC allowable voltage at 3750 psi is 3.76 volts versus 4.1 volts and the 2650 psi EOC voltage allowable is 10.6 volts versus 10.95 volts. The cenclusions regarding EOC burst capability at 1.66 volts and 2.53 volts remain unchanged. However, the BOC voltage providing 5290 psi burst strength drops to 0.88 volt. The secono assessment of sensitivity to Catawba data is performed including what are judged to be
, unreliable data. The Catawba-1 pulled tubo data added to the above data set are:
E!cbbin votts Burst oressure (ksi) 1.82 4.23 0.10 9.40 1,48 5.00 1.57 5.74 The resulting equation for BP is: [ ] The coefficient of correlation for this fit is 0.83 with a standard error of 1.25. This is shown in Figures 7-7 through 7 9. The result of including these "unburst" data is to unrealistically reduce the BOC voltage corresponding to 5290 psi to 0.6 volt. Other requirements regarding EOC voltage are still satisfied although with smaller margins. The judgment that the correlation resulting from including these data is unrealistic is further reinforced by the trend analysis in the following subsection wherein it is shown that very small changes in throughwall crack length (less than 0.03 inch) are associated with the voltage range of 0.1 to 1.0 volt. Consequently, a
, likewise very small change in burst capability would be expected (!ess than 500 pst) for the same crack morphology over the same voltage range.
7-3 l 4
Trends Esotween 34 Inch md 7/8 inch Correlations
- To obtain additional insight into the difference between the 3/4 inch (lower structural voltage limit) and 7/8 inch voltage / burst correlations, a very preliminary assessment has been -
performed of expected differences based on crack length correlations. Relations between voltage and throughwall crack length were estimated using preliminary crack length data (See Figures 710 and 711). Although average crack lengths are more relevant to burst capability than ' throughwall lengths, only the throughwall data were immediately available for this trend analysis. For the present application, the use of throughwall crack lengths leads to an overestimate of burst capability but the trends between 3/4 inch and 7/8 inch tubing should be representative of the expected trends. Existing (Belgian and Westinghouse) correlations for burst pressure vs uniform throughwa!! crack length (Figure 712) were then combined with the voltage / crack length relations to obtain voltage / burst pressure correlation estimates for both 3/4 inch and 7/8 inch tubing. The resulting, preliminary correlations, Figures 713 and 714, indicate a steepcr slope (decreasing burst pressure) with increasing voltage for 3/4 inch tubing and lower burst pressures at a given voltage (>2.0 volts) for 3/4 inch tubing. These-trends are the same as found for the direct correlation of bobbin voltage measurements to burst pressure as presented at the August 28 meeting with the NRC staff. The principal contributor to the trend 3 is the lower burst pressures of 3/4 inch tubing at a given crack length as augmented by somewhat higher voltages (factor not considered reliable from current data) at a given crack length for 7/8 inch tubing. These prelim l nary results indicate that the general differences found between the 3/4 inch and 7/8 inch vo!! age / burst correlations should be expected based on crack length considerations. 7.4 SLB Leak Rate vs. Voltage Correlation Recommended Correbtion The regression analysis techniques for the 3/4 inch tubo data are consistent with those performed for the 7/8 inch tubing and described in WCAP-12871, Rev. 2. The resulting mean and upper and lower 95% probability lines (one-sided) are shown in each of the following plots. Also presented is a curve representing the arithmetic (numerical) average of an assumed log-normal distribution of leak rate at each value of voltage. Figure 7-15 utilizes the non-leaker data at 0.0001 1/hr. as done in WCAP-12871, Rev. 2, for 7/8 inch tubing. However, based on the oliscussion presented below, the recommended regression solution, presented in Figure 7-16, utilizes the non leaker data at 0.001 1/hr. The correlation is estabiished utilizing linear regression analysis of the logarithms of the corresponding leak rates and voltages thereby establishing a leakage rate model of the form: [ ]o where, [
)O Prediction intervals for leakage rate at a given voltage are then established to statistically define the range of potentialleakage rates.
7-4
~,
i The SLB leakage rate data from 32 model boiler specimens used to establish the recommended-correlation for 3/4 inch tubing are fisted -in Table 7.3. Linear regression analysis of the logarithms of this data results in the following mean leakage rate correlation: [ tS The coefficient of correlation for this regression fit is 0.65 and the error c .. e estimate is 1.31. A prediction interval is established using the expression: 1 [ ]a where, [ f
)g-Sensitivity to Plant E-4 Data As should be noticed, the set of data evaluated in Figures 715 and 7-16 does not include the Plant E 4 data. They are excluded since both leak rate (taken at room temperature) and voltage must be adjusted and consensus has not been reached on either adjustment. To assess the potential sensitivity of the correlation, the regression analysis of Figure 716 was repeated -
with the Plant E-4 data added. The voltage and leak rates were included at values recommended by the supplier of the data, Laborelec, and the voltage values further factored by 1.5 as was the case with the recommended Nurst correlation. As seen in Figure 7-17, the mean regression equation and the 95% predic,.on intervals are not changed significantly. If, however, the Plant E-4 data are included in the regressic n analysis without the additional 1.5 factor on voltage, leak rates at a given vcitage are approxir6,ly doubled (Figure 7-18). The factor of 1.5 is expected to be conservative as discussed in Section 5. SLB Leak Rate vs. Volface Trends An evaluation has been completed that provides results supporting the recommended-correlation methodology employed for SLB leak rate versus bobbin voltage. Of concern to some reviewers has been the use of non-leaking, degraded tubes in the correlation. _ Our reasoning for including - them is to establish a log-log slope on the order 4, similar to leak rate versus crack length curves, even though the selection of leak rate at 0.0001 was arbitrary (yet consistent with the - accuracy of the leak rate measuring instrumentation / method). The evaluation establishes leak rate versus voltage from:
- 1) Formulation of throughwall crack length (L) versus voltage (v) correlation from available data using regression analysis (L - tf (v)}.
- 2) Calculation of leak rate (O) as a fu,1ction of L using the CRACKFLO computer code.
Formulate relationship (simplified) F. : ' a function of L through regression analysis of 7-5
. - -. - -- ~ .
.. . - - .- _ _ - . - - . - - - . . - . - ~ - - - - - .~ -. t CRACKFLO predletions (O - 2f (L}}. '
- 3) Development of correlation bbtween O and V from the above. Substitite the formulation L - f t(v) into 0 - f 2(L) to get O - f 3(v). Compare 0 - f 3(v) to test data.
The result of the first step is shown in Figure 710 which illustrates the mean of the regression . formulation provided in the title of the figure. This is the same correlation described in Section 7.3. This crack length correlation is considered preliminary as additional crack lengths for other specimens are being obtained and the existing data is being verified. The CRACKFLO code for calculating leak rate from throughwall axial cracks (WCAP-12871, Rev. 2) is utilized to provide predictions of SLB leak rate as a function of crack length. For simplicity, these solutions are then fit via regression analysis by two equations to improve the accuracy of the correlation and to eliminate further CRACKFLO calculations. Figure 719 provides the fit utilized below Log (0.25) and Figure 7-20 the fit utilized at Log (0.25) and above. Combining the crack length versus voltage and SLB leak rate versus crack length equations results in an equation that estimates SLB leak rate as a function of voltage. A comparison of CRACKFLO predictions to measured magnitudes w shown in Figure 7-21, The mean of predicted values is greater than the measured values over a range of 0.003 gpm and higher. Thus, applying CRACKFLO directly results in an assumption that predicted equals measured and is conservative in that CRACKFLO overpredicts measured leak rates. Thus, as Figure 7-22 indicater., the calculated mean is higher than the bulk of the data as reflected by the regression fit mean also plotted. Note the neariy equal magnitude of slope. This slope is obtained by including non leakers in the correlation but at 0.0011/hr rather than 0.00011/hr as had . previously been done. Use of 0.001 t/hr. Is recommended in lieu of 0.00011/hr. since the - slope agrees with the alternate method and since regression analysis of the data excluding the non-leakers would result in a lower slope. - 7.5 Bounding SLB Leak Rate vs. Voltage This section documents the development of bounding leakage under steam line break conditions. The leak rate database includes both the 7/8 inch and the 3/4 inch diameter tubes. The database consists of 74 model boiler specimens and 93 pulled tube intersections. The voltage amplitudes from Plant E-4 pulled tubes were no1 adjusted to account for the differences in voltage normalization used for *. Jata acquisition in Belgium and for the APC database (in the U.S). The - data from 7/8 inch tubing is at 400/100 kHz differential mix with the 20% holes in the ASME standard normalized to 2.75 volts in the mix. Similarly, the 3/4 inch tubing data is at 550/130 kHz differential mix with the 20% holes in the ASME standard normalized to 2.75 volts, in order to combine the data from the 7/8 inch and 3/4 inch specimens, a conversion factor equal to the square of the diameter ratios is applied. This factor results from the fact that the ASME standard hole size is the same for 3/4 and 7/8 inch tubing. Thus, to convert the 7/8 inch data to the same basis as the 3/4 inch data, the voltage amplitudes from the 7/8 inch data (at 400/100 kHz) were divided by the factor 1.36. The results were then combined with the 3/4 inch database.
~
As per the detailed discussion in Section 6.6, the Catawba-1 pulled tube R5C112 (TSP-3) would not leak at SLB condition had it not been damaged during the tube pull operations. Therefore, for 7-6
this spocimen (and only for this specimen) the post pull amplitude of 5.06 volts rather than the prepull value of 1.82 volts is used in this context. Thh tube had an SLB leak rute of 0.55 Phr. , The leak rates were,in most (131 of 167) cases, the direct result of measurements in the laboratory under SLB conditions. In other (36) cases, laboratory data on leak rate measurement was not available and the likelihood of leek rate was inferred from crack morphology (throughwall depth and length) obtained from destructive examination. The data was classified into leaking and nonleaking specimens. Frequency distribution of voltage amplitudes (corresponding to the 3/4 inch data normalization) in each classification was determined. This is shown in Figure 7-23 as a stacked bar chart. The number of leaking specimens in each voltage range out of the total numbor in that range is also shown listed at the top of each bar in the figure. The ratio of the number of leaking spec l mens in a voltage range (bin) to the total number of specimens in the bin was calculated from the above frequency distribution of voltage amplitudes. This result, probability of leakage, within each voltage range is plotted as a bar chart in Figure 7-?4. Instead of classifying the data into leaFers and nonieakers, a second classification was made with a leakage threshold of 1.0 l'hr (specimens leaking at rates less than 1.0 Vhr and those leaking at a higher rate). For each of these classifications, frequency distribution of voltage amplitudes was determined. Probsoility of leak rate above 1.0 l'hr was calculated for each voltage range as was done for the 0.0 leakage threshold. The frequency diwibutions and the probability distribution are displayed in Figures 7-25 and 7-26, respectively. The above results (Figuras 7-23 through 7-26) were developed using data from both the 7/8 and 3/4 inch diameter tubing. If the 7/8 inch data is excluded and only the SLB leak rate data from 3/4 inch tubing is used, the resu!ts are not changed significantly. This is displayed in Figures 7 2 through 7-30. For the 3/4 inch tubes, the data supports no leak under SLB conditio .s for indicstions up to 2.0 volts (bobbin data for 550/130 kHz mix wnh 20% hole ASME standard normalized to 2.75 volts) and smallleak rate (< 1 Phr,if any) in the voltage range of 2.0 to 3.5 volts. Table 7-4 summarizes the bounding SLB leak rates determined from the pulled tube and model boiler database for voltage ranges up to 3.5 volts for 3/4 inch diameter tubes. Recommendations for bobbin signals above 3=5 volts are also included in the tcble. The table also shows the low voltage indications which contribute to the threshold voltage values of 2.0 volts and 3.5 volts. It may be noted that if the voltage amplitudes for the Plant E-4 data were factored up to account for the differences in European and U.S. voltage normalizations, then the threshold voltage for low (< 1.0 t/hr) leak rate would increase to 4.2 volts. Table 7-4 summarizes the lowest voltage indications having leakage <1.01/hr. and the lowest having >1.0 L'hr. The lowest voltages with <1.0 t/hr. help to define a threshold for leakage. The smallest voltage having a measurable leak rate is >2.0 volts (R46C73) based on reducing the
, 7/8 inch tubing voltage of 2.81 volts by a factor of 1.36 for the approximate 3/4 inch tubing voltage. As discussed above, Catawba-1 tube R5C112 has not been applied for the leakage threshold as the post-pull voltage of >5 volts should be associated with the leakage due to
. expected damage to the indication during the tube pull. The lowest voltage of 1.9 volts (R5C61 from Plant B-1) found for a throughwall crack with no leakage (pressurization during burst test) is consistent with about a 2.0 volt threshold for leakage. EPRI report TR-100407 also 7-7
-- -- - - = - - .- -- -. . -.- . - - - - . - . uses the 7/8 inch R46C73 to defino a leakage threshold for SLB conditions. The EPRI report scalcs the 2.8 volts to a BOC estimate of 1.9 volts and suggests 1.5 solts as a BOC estimate. For this Catawba 1 assessment, the threshold applied is for EOC volts and applies the same basis as the EPRI report. The threshold for leakage >1.0 litor/ hour is about 3.5 volts from Table 7-4, -
. However, the Belgian tube R33C96 bobbin voltage of 3.54 volts should be increased by a factor of at least 1.5 to >5 volts. Thus no indications below about 4 volts have been identified with a leak rato >1.0 liter /hr. In model boiler or pulled tubes. For Catawba.1, a threshold >3.5 volts for leakage >1.0 liter /hr does not significantly influence the SLB leakage as the largest projected EOC indication is about 2.53 volts.
The threshold for SLB leakage can be assessed by evaluating the lowest bobbin voltages resulting. In leakage at SLB conditions and by evaluating the throughwall crack length generally required for rneasurable leakage. If the throughwall crack length associated with measurable leakago can be defined, the voltage vs crack longth relationship of Section 7.3 above can be used to assess the voltage threshold for leakage. The crack length method for estimating a voltage threshold for leakage providos a more physical insight into the threshold estimate. I can be noted that significant efforts were applied in the 3/4 inch tubing model boiler specimen preparation to obtain the lowest voltago associated with leakage. In the model boilers, leakage is monitored by sensing for lithium _which provides a leakage sensitivity of about 3X10 3 1/hr. Upon detection of any leakage, the model boilers wore shutdown and the tube (typically 4 to 6 TSPs) was removed for NDE inspection. TSP intersections with bobbin indications above about 1 volt were removed from the tube for further NDE, leak and burst testing, The smallest bobbin voltage from this program having a measurable leak rate in the leak test facility (capability to measure down to 10'3 to 10-4 /hr) was 4.24 volts (No. 601 1). It h possible that a specimen
- at 2,79 volts (No. 595 2, throughwa!! crack length - 0.17 inch) was detected in the model boilers with no measurable leakage in the eak test facitity. .,
As discussed above, the leakage threshold can be assessed by exam %g crack length data. Table 7-5 shows specimen throughwall crack longths for no leakage, leakage <1.0 liter /hr. and between >1.0 and 6.0 liter /hr. No leakage has been found for throughwall cracks up to 0.17 inch. With the exception of 601-1, leakage <1 liter /hr occurs for crack lengths of 0.11 to 0.27 inch. Specimen 601-1 had a 0.051.,ch throughwall corrosion crack with a thin ligament suspected to have par 1ially opened at SLB conditions. This specimen had no leakage at operating condition pressure differential and 0.33 liter /hr. at SLB conditions. For this type cf indication (above APC repair limit), voltage is a better Indicator of leakage potential than even throughwall crack length or total crack length (0.29 inch). Overal!, the data of Table 7 5 indicate that throughwall crack lengths >0.1 inch are generally required for SLB leakage. Small leakers, such
'as 601 1 could result at small crack lengths but voltages significantly exceeding EOC voltages for an IPC with a 1.0 volt repair limit. The Table 7-5 data indicate a throughwall crack length - of about 0.14 inch for leakage >1.0 liter /hr.
From Figure 7-10, it is seen that a crack length of 0.1 inch, as an estimated no leakage threshold corresponds to a bobbin voltage of 3.0 volts. A crack length of about 0.14 inch for . leakage >1.0 liter /hr. corresponds to about 4 volts. Thus, the leakage thresholds from crack length trends support the leakago thresholds obtained above from voltage consMerations only. Overall, the dPla support no leakage below about 2 volts, leak rate of less than 11/hr below - about 4 voits, and leakage greater than 11/hr above about 4 volts at EOC. 7- 8
= . _ - . . - . - . . - . _ _ . . .- = -. .. . -- .-
l 1 Table 71 Leak Rate & Burst Strength Database for 3/4 inch Tubing
. Preliminary Bobbin Burst Destructive Exam.
Amplitude Lenk Rate (1/hri Pressure Lenath (inch)
& Soecimen (voht) N.O.AP SLB .iP fosh Maximum - Thruwall*
_Q (Continued on next page) Maximum depth of penetration is shown when crack is not throughwall, or when throughwalllength is not available. 79
Table 7-1 (Continued) Leak Rate & Burst Strength Database for 3/4 inch Tubing Preliminary , Bobbin Burst DestructNe Exam. Amplitude Leak Rate (l'hr) Pressure Lenath (inch) th. Soncimen (vohs) N O.AP SLB AP (Dsh Maximum Thruwall' _. g Notes. Maximum depth of penetration is shown when crack is not throughwall, or when throughwall length is not available.
- 1. Tested at room temperature.
- 2. Belgian voltages wib the cross calibration factor of 1.5 (rather than 1.809).
- 3. Leak rates measured at room temperature conditions and analytically adjusted to operating conditions by Laborelec.
- 4. Not measured at 550/130 kHz. Voltage renormatized from 300 kHz data.
- 5. Observed during burst test at room temperature. ,
- 6. Burst pressures from Table 6-7 including adjustments to mean flow stress, and, for Catawba 1 data, ter burst testing methods.
7-10 i
Table 7-2 Burst Pressure Data Used .. .Jevelop the Recommended Correlation Bobbin amolitude Burst oressure ~ (volts) (ksi) _o (Continued on next page) 7- 11
. - ~. . -. - - - - ~ .. . - -- - -
i ITable 7 2 (Continued) Burst Pressure Data Used to Develop the Recommended Correlation . I Bobbin amolitude Burst cressure (volts) (ksi)
~ Q Plant E-4 data l
l. i 7-12 i
~ . - . . - - . - _ . .
Table 7-3 SLB Leak Rate Data Used to Develop the Recommended Correlation Bobbin amolitude Leak Rate
~
(volts) (Vhr) _. 9 usan i e 7-13
- . . . ~ . . _ . _ _ _ . __. . - _ . . . _
Table 7-4 Summary of Bounding SLB Leak Rates for 3/4 Inch Tubing Voltage Ranges Bounding VoNace Rance SLB Leak Rate Lkpitino Indications for Leakaoe s_2.0 volts 0.0 L'hr Tube R46C73 from Plant A 2: 2.06 volts (2.81 volts for 7/8 inch tube) 0.17 thr 2.0 to 3.5 volts - 1.0 L'hr Tube R33C96 from Plant E-4: 3.54 volts (1) 1.5 L'hr , Model boiler specimens: 558 1 (7/8"): 4.79 volts,4.08 L'hr 600-3 (3/4"): 4.25 volts,44.4 t/hr 601 1 (3/4"): 4.24 volts,0.33 t/hr 6041 (3/4"): 4.93 volts,5.00 t/hr 3.5 to 4.2 volts 10 t/hr Based on NRC recommendation for IPC implementation (D. C. Cook)
> 4.2 volts li projected EOC amplitude for indications in 3/4 inch tube exceeds 4.2 volts, additional evaluation on bounding SLB leak rate will be required
- 1) Utilizes minimum voltage increase in renorma;; zing Belgian data to APC normalization.
Final renormalization factor is expected to significantly increase voltage. s i e 1 a 4 7-14
Table 7 5 : Dependence of SLB Leakage on Throughwall(TW) Crack Length
'M . No Lenkace TW <1 litt air . TW >1 liter /hr. A < 6 Phr th, Bobb!n V.011s TW Lenoth h Bobbin Volts TW Lenclb h Bobb!n Volts TW Lenoth
_Q Notet:
- 1. Bathtub flaw (thin OD ligament) reasonably expected to open crack at SLB AP.
l l l O P 7 15
. . . . . _ _ _ _ _ _ . _ . _ . ~ _ . . - . _ _ _ _ . _ _ . _ _ . . . . - _ . . _ . . . _ . . - . . _ - . . _ . _ _ _ _ _ _ _ . _ ._._
I i i e 4 f S Page us Intentionally latt b!ank, ? s
)
e
- f i
I I r n 5
+
h t i i e i B
* )
1 e i h
+
t a t f~ f
. t I
l 7*16-I l' _ _
*r" e DNM-vu7&4T- rnM-mw-g mw ww w eerWGw-'-TtW4W"-M-'9--@W-W-Tre 7ev-+w"T-eVW %W p-evvawwwpwo-Trm-- w1 Mr W -O == W *MWn Ts-'*rq W'y '*+&wfM -
m--Wg-"- y F9 4-f- e -49tW
i :
! +
i r I s f i i a,g
- - i i
l 1 Y t [ 4 t i 3 h Figure 71. Recommended Correlation for Burst Pressure versus Bobbin Voltage 4 i 7 17 y.-,.--, . . . , , - - - --w..g.,--. .f,,-..-,79.,,9myy,--,,.c-.. y c,
...,3 ,y_. . . --w. ri- .. e r - mwim..+s3.mv- -s -
t-e,- . ,e , e --m ..m,...w4 -
.+=w --+='-
l
)
1 i a,g , ; P e ! i
\
i i i t r
-i P
l - l t l ' I l l , l Figure 7 2. Recommended Correlation for Burst Pressure versus Bobbin Voltago 7 18 , l l l;_ . - - . - - - . - . - _ _ - - . . - . - - _ . . -.- - . - - - - - - , . - - . - - .
I k i
?
. -B,9 1
')
k
'C P
k I i, i k r t t t E l l
?
Figure 7-3. Recommended Correlation for Burst Pressure versus Bobbin Voltage l I 7 19 _.;_:-._.;_._-.___~._____._.---_.._,_... _ . . ~ , _ _ _ ,
.m___..__._ . . _ _ _ . _ . _ _ . _ _ _. - _ _ _ . . -_ . . _ . . _ _ - . _ . _ .
t t l i a,g * ,
?
- t u
l e
. i I
e G i 1 6 F T i Figure 7 4 SensitMty Correlation for Burst Pressure versus Bobbin Voltage, includes Unadjusted Catawba 1 Pulled Tube Data , 7 20 t t
-,.r-,..b, m,.. ..,_.-~~,A.,;r,,.-.~.,,,,m .: m,,,y, ..,, . , ,.y._- . . , - . . , . . , ;,,ce. ....,.y,.,.-~,-r...,.-.-.,.~,,,n,,..,_,--ze.--.-,,-,mm.,,,m -w
_ _ _ _ _ ..__ __ _ _- ---._-.-.--.-._.____..___.__.___._._._.___.____m A i i ag [ i i i I i i i i e 5 1 i i k
?
i i
?
-i
+
i a 4 9 i t Figure 7 5. SensitMty Correlation for Sarst Pressuro versus Bobbin Voltage. Includes Unadjusted Catawba 1 Pulled Tube Data. 7 21 g- <r- g+e,-- rT ** W' F M W
- NF**--'Tr*'Wr#'m'pr ws-w tgww e-
<w---o-,g-~, , W m,e e, wy m rw,, m.,-w.-
r
. [
k i i i 1 0
< l r
.r L f i I I t l I 1 i 1
., 1 4 !
Figure 7 6. Sensitivity Correlation for Burst Pressure versus Bobbin Voltage' ' ] includes Unadjustod Catawba 1 Pulled Tube Data. 7 22
. . . . . . . . _ _ , , . , - . . . _ - . . _ ~ , - , . ~ _ . . _ . - , - , - - . _ . - _ . . . _ . - . , , _ _ , , , _ . - - . _ - . , . . . . - . _ . . , _ - , . . . . _ . . _ . _ _ _ . . . _ . _ - - _ - - - _ _ _ - -
. - . _ - . - _ - . . . . - . . - - . - . . . . - . - ~ . - - - . . - _ _ . . . . - . . . - ~ _ . - . - . .
t
- a.g t
[ f A t i.
. Figure 7 7, SensitMty Correlation for Burst Pressure versus BotMn Voltage.
Includes Both Unadjusted and Unreliable Catawba 1 Pulled Tube Data. t 7 23
,. r s
?
4 1 i I t a,g l t F- . i i I I i t f
.1
!I i
I i i
, i i
r i h w L 4 Figure 7 B, SensitMty Correlation for Burst Pressure versus Bobbin Voltago, includes Both Unadjusted and Unreliable Catawba 1 Pulled Tube Data. 7 24
.l
'FF-9*N 9'M' VP *T- 1*ia' ti r ' % 7 9 P eF v M W-f
- v y'WW67 "' W T P C- ?W'-l'WM*""W W 7'MV N'D'F 'y77'WFY**'tf'ENT
_ _ . _ _ _ _ . _--. __ _._.__ _ _ . . _ . .._m_._____.. l I i a,g 4 I L i I-L i i t i, i
+
Figure 7 9 Sens!!Mty Correlation for Burst Pressure versus Bobbin Voltage. Includes Both Unadjusted and Unreliable Catawba 1 Pulled Tube Data. 7 25
-r - --c -+ w , . . - - , . - . . - , - . ,e --
-4 e ew w., +s v'a w ,,e ,,---s-.-s b wyer sw-, -r--,w-,.<vr e w--t. w w r-,---+w-sm- av---mm.,---m-=-+ww==t, <- w,- vi- --4,w c-
ag . Figure 710. Crack Length versus Bobbin Voltage for 3/4 inch Tubing , 7 26
_. _ . . _ _ _ _ _ _ _ . _ . _ . _ _ _ . _ _ - _ _ _ . - . . _ . . . - _ _ _ _ _ __-__m._._._.._._.-_ . __-_ _._ t r
.g -
l i 4 s 4
-i l
p I a 8
.i r
Figuie 7-11. Crack Leregth versus Bobbin Voltage for 7/8 Inch Tubing 7- 27 L 2..._., .,.~. . -.m..,- n,.... ..,. ..,--.w....,.m,e,n.we...,-...w_.e.,. ~ , ,, , - - - , . , _ ,,,+.p%w.,-,,-w ,, , - ~ . e,,,.,.,ww., r,.,,,,r,,,,r ...m,y-e s.,u-,. 7m..r.,.,_
i t 1 a . ; t l ; E S 4 i I-i i e Figure 712. Burst Pressure versus Crack Length . 7 28 L y ...,y.., - - , , ,,m- - ... - -.-- .,,-+- y .-we, . . , , -,y,., -,y,~ , , . - - . . , , . ~ , . , _ ,-,,.-w.,,,m,, y,, -
~w v%w... ,.-...-m.. .,w_,-. . ~ ~ . _ , .--c..,,w,-
ag Figure 7-13. Trend Analysis for Predicting Burst Pressure versus Bobbin Voltage for 3/4 inch Tubing. 7 29
i
-i a,g .
I f r
-I P
P
?
L
~f T
i I F Figure 714. Trond Analysis for P.adicting Burst Pressure versus Bobbin Voltag6 , for 7/8 Inch Tubing. 7 30 _,e ---,e-m., e
I .- a.g P P r
- )
. Figure 715. SLB Leak Rate versus Bobbin Vohage with Non leakers at 0.00011/hr for 3/4 Inch Tubing, 7 31 1
..1... ,,.-..a.. . _ . . , , ,_ ..
ag
~
f Figure 716. Recommended Correlation for SLB Leak Rate versus Bobbin Voltage - for 3/4 Inch Tubing (with Non leakers at 0.001 thr) 7 32 d
a.g 4 l e i l t i 4 D' l 1 I i Figure 717. SensitMty Correlation for SLB Leak Rate versus Bobbin Voltage for 3/4 inch Tubing (with Non leakers at 0.001 Ltr). . Includes Plant E 4 Data with Voltages Factored by 1.5. , I I 7 33 l l
-s., e .w-se--.gm,y , v. ,. --my----e--,.e--kpqws,q eg.,-.g< - - - . -r wuv w' ---r---tw-wsi.w.w+em,y- g gi v w- e um e---t-r'-yy *-'y , bp g It h -T-4T- W T-;--$N-te '-Wa= e
.. .. - -.. - - - - ... . . . - . - - . . . - . _ . . - . . . . . ~ . - - . - . - . ~ . - - . -
i i
?
I i t
,r
-a,g !
- i i
1 i
?
t F L e
'f f
f i r Figuro 7-18. Sensitkity Correlation for SLB Leak Rate versus Bobbin Voltage _, i for 3/4 Inch Tubing (with Non leakers at 0,001 t/hr). Includes Plant E 4 Data without the 1.5 Factor on Voltages. i 7 34
. - . . _ , . . _ _ _ . - - , _ . _ . . , _ - _ . . , _ _ _ . . . . . . . . _ . - - _ . _ - _ , . - _ . . - - . . , _ . . , _ - . _ . _ - , _ _ . . - . . . ~ _ _
a
'b d
Figure 719. SLB Leak Rate versus Crack Length from CRACKFLO Code 7 35 im .
a . 4
- m Figure 7 20. SLB Leak Rate versus Crack Length from CRACKFLO Cr4< .
d 7 36 r gW# y 7 r m yiet g v-e 4f---- v -*?w - w
- ' * - #ve g- * --me w wg-ww'- -',w-y-vg=--4-'ev-v--w- - ---ow
. a.g W
I I 1 I l* l
.- Figure 7-21. Measured versus Predicted SLB Leak Rate Using CRACKFLO Code 7 37 '
. . - . ._. _.. _ . _ _ _ . . _ , . - _ _ . _ _ _ _ . __. .1 i
i 1 1 1 a.g-
- lJ i
I d t 5 f 4 i
- - w l
l i l Figure 7 22.. Trond Analysis for Predicting SLB Leak Rate versus Bobbin Voltage -
-7 38 v - =*e- -
--w~.- _--e, y- .. m,, , , - -,. . ...-r..,. ,,,w , . . + ..- c r - - - - - - sr. - - - . , , ,,,,e e- w,%rw, , m rtr
i i a.g
?
A ) t h 4 e e
+
L 3 - 1 Figure 7-23. Frequsney J mnbution of Leakers and Nonleakers (at SLB Conditions) versus - Ocwn Voltage for 3/4 inch Voltage Normalization 7 39 Ft 7 3- t+_W ie v
- w r-N p$--y p myg y.r-mvW9.<ww 9y r e w-, :p p Wg *Wie-.w1e. my P 'eg g+r- g &g&y,d w86M e ty-g ' gi+r mbryte e 7 Twe - g yt- eg T g e her
_ . . . . _ ~ . _ _ . - . - - _ _ . _ _ _ . . _ . _ . _ _ , _ . _ . _ __ . _ _ . _ . . . _ . . . _ . . _ . _ _ _ _ _ . _ _ _ _ . l i !
?
a.g ; i 1 t i
=
i t F Figuro 7 24. Probability Distribution of SLB Loakage versus Bobbin Voltage 7 40 i
- -- .. .. - - . . . . . . ~ . . . - _ . . . _ , . . - - - . . , . , . , , - - , , , , , ..~,-----,-,.._,,...,-,.,,,.--..--.._u, , .- . . - - - - ., , , , , . , . - , . , -
a.g
+
7 Figure 7 25. Frequency Distribution of Specimens for SLB Leak Rate Threshold of 1 t/br / versus Bobbin Voltage for 3/4 Inch Voltage Normalization 7 41
e i a,g L ._ i Figure 7 26. Prot Ibility of SLB Leak Rate > 1 thr versus Bobbin Voltage 7- 42 i. i:- - l
, . - . , - . - . . ,.iw
a,g t 0 th un Figure 7-27. Frequency Distribution of Leakers and Nonleakers (at SLB Conditions) versus Bobbin Voltage for 3/4.nch Data 7-43
~- -. ~ - . ... . . . . . - ~ . - - . . .. . .~ . .
i l a,g
+
b
}
f T F Figure 7-28. Probability Distribution of SLB Leakage versus Bobbin Voltage for 3/4 Inen Tubing 7 44
-a,g
+
4 y a l l Figure 7-29. Frequency Distribution of Specimens for SLB Leak Rate Threshold of 11/hr l versus Bobbin Voltage for 3/4 inch Tubing 7-45
....- . - ..._.- . ~... . - .. .... .
. - - . - . . - . . .-- .. .. - - - - -.. - - _ ..... ..~- . . . -
S a,g-3 I J-Faere 7 30. Probability of SLB Leak Rate > 1 t/hr versus Bobbin Voltage for 3/4 inch Tubing
- 7-46
+ , .. r , 'n--.- - . , ~ e.-%..- - - e.- ,- . . . - ,-.e,., , . .~ -.-
Section 8 ACCIDENTCONDITION CONSIDERATIONS 4 This section deals with accident condition loadings in terms of their effects on tube deformation . and on tube burst pressure. The most limiting ac0ldent conditions relative to these concerns are selsmic (SSE) plus loss of coolant accident (LOCA) for tube deformation, and seismic plus steamline/feedline break (SLB/FLB) for tube burst capability. For the combined SSE + LOCA loading condition, the potential exists for deformation of tubes and subsequent loss of flow area and postulated in-leakage, in-leakage is a potential concern, as secondary to primary leakage may affect the Catawba Unt 1 emergency core cooling system ; (ECCS) analysis. Relative to tube burst strength, tube bending u ess is induced at the TSP intersections during postulated accident conditions. This bending stress is tenslie on one side and compressive on the other, and is oriented in the axial direction. The compressive stress has the potential to open axial cracks and to reduce the burst capability of the cracked tube due to the crack opening. 8.1 Tube Deformation Under Combined LOCA + SSE For the combined SSE + LOCA loading condition, the potential exists for yielding of the tube support plate in the vicinity of the wedge groups, accompanied by deformation of tubes and subsequent loss of flow area and a postulated in-leakage. Tube deformation alone, although it . , ' impacts the steam generator cooling capability following a LOCA, is small and the increase in PCT is acceptable. Consequent in-leakage, however, may occur if axial cracks are present and propagate throughwall as tube deformation occurs. This deformation may also lead to opening of pre-existing tight through wall cracks with consequent in leakage following the event. In-leakage is a potential concem, as secondary to primary leakage may affect the ECCS analysis as a result of the potential for steam binding in the tubes during the core reflood. 8.1.1 SSE Analysis Seismic loads result from motion of the ground during an earthquake. The SSE excitation of the - steam generators is defined in the form of acceleration response spectra at the steam generator supports. To perform the non-linear time history analysis,it is necessary to convert the response spectrum input into acceleration time history input. Acceleration time histories for the nonlinear analysis are synthesized from El Centro Earthquake motions, using a frequency suppression / raising technique, such that the resulting spectrum in each of the orthogonal axes closely envelopes the original specified spectrum in the corresponding axis. The resulting three orthogonal time histories of the earthquake are then applied simultaneously at each steam generator support to perform the analysis. The analysis is performed using the WECAN finite element computer program. The mathematical model consists of thrsa-dimensional lumped mass, beam, and pipe elements as well as general matrix input to represent the specific steam generator piping stiffnesses. The TSP-shell interaction is represented by a rotating, concentric gap-spring dynamic element, using impact damping to account for energy dissipation at these locations. The mathematical model with selected node numbers is shown in Figure 8-1. The primary loop piping and the lower column 8-1 f
support stiffnossos are input as 6x6 matricos. The upper and lower lateral support restraints are represented by compression-only (single-acting) spring oloments, with the shell flexibility includod for the upper support stiffnosses. At the lower Olovation, the support structuro is connected to a relatively rigid channel head-loot combination. In modoling the tubo
- bundio internals.to-shell interface, the TSP local shell stiffness, obtained from detailed finito olomont analysos, is also includod. The local shell stiffness at the top TSP location is highor than at lowor TSP locations because of its proximity to the upper lateral supports.
The tubo bundle geometry is shown in Figuro B.2, with tho tubo battles and support p!atos identified. The flow bafflos (E, F, J, K, N P) are typically not Included la the seismic model, both due to the difficulty in representing them accurately in the model, and also because it is conservative in terms of tube stressos to exclude them from the model. If the baffles were included in the model, it is anticipated that contact impact loads for the lower plates would be distributed among the various plates and baffles resulting in reduced loads. However, it is difficult to estimato those loads, duo to the flexiblo naturo of i o partition plate which forms ono - of the support members for theso platos. For the flow bafflos, it is mnCluded to be conservativo to uso the loads developed for the support plates (C/D, G/H, UM). Soismic loads for this analysis are taken from an analysis of a similar model steam generator, and have boon determined to be conservative by comparing the spectra and TSP loads from three different plants, all having similar steam generator geometries. Results from the bounding analysis show the flow distribution bafflo (plates A/B in Figure 8 2) to not experienco seismic impact loads. Thus,it is judged that thoro will not be any tubes at the flow distribution baffic location that are potentially susceptible to in-teakage. For reasons that will be discussed later, tubo deformation calculations are performed for throo TSP groupings. Discussions of the groupings along with the ro,Jtting TSP loads is contained in , Section 8.1.3. 8.1.2 LOCA Analysis LOCA loads are developed as a result of transient flow, and temperaturo and pressure fluctuations _ following a postulated primary coolant pipe break. Based on the prior qualification of the Catawba steam Generators for leak before break requirements for the primary piping, the limiting LOCA event is either the accumulator line break or the pressurizer surgo line break. However, bounding LOCA load calculations for Catawba for the accumulator or pressurizer surgo lines are not available. As a conservativo approximation, the availabio LOCA loads for the primary piping breaks are used to bound the smaller pipe bruaks. The lar00 pipe break loads have been shown for other model steam generators to be several times larger than the smaller pipe breaks, and thus, it is judged that these loEds form a conservative basis for the small pipe breaks for Catawba. As a result of a LOCA event, the steam generator tubing is subjected to the following loads:
- 1) Primary fluid rarefaction wave loads.
- 2) Steam generator shaking loads due to the coolant loop motion.
- 3) External by Postatic pressure loads as the primary sido blows down to atmospheric ,
pressure. 82
- - - - - . . - - -. . - - .- - -. - . . - . - = .
- 4) Bonding stresses resulting from bow of the tubeshoot due to the secondary to primary pressure drop.
, 5) Bonding of the tube due to differential thermal expansion between the tubesheet and first tube support plate following the drop in primary fluid temperature.
Loading mechanisms (3) through (5) above are not an issue since they are a non cyclic loading condition and will not result in crack growth, and/or result in a compressive membrane loading on the tube that is beneficial in terms of negating cyclic bonding stresses that could result in
- crack growth.
8.1.2.1 LOCA Rarefaction Wavo Analysis The principal tube loading during a LOCA is caused by the rarefaction wave in the primary fluid. This wave initiates at the postulated break location and travels around the tubo U bends. A differential pressure is created across the two legs of the tube which causes an in-plane horizontal motion of the U bend. This differential pressure, in tum, induces significant lateral loads on the tubos. The pressure-time histories input to the structural analysis are obtained from transient thermal-hydraulic (T/H) analyses, using the MULTIFLEX computer codo. A break opening time - of 1.0 msec of full flow area, simulating an instantaneous double ended rupture is assumed to obtain conservative hydraulic loads. The fluid-structure interaction effect due to the flexibility I of the divider plate between the inlet and outlet plenums of the primary chamber is included in the analysis. Pressure time histories are calculated for two tube radil, identified as the average and maximum radius tubes. A plot showing the tube representatiot,in the T/H modelis provided in Figure 8 3. Typical primary pressure time-histories following a LOCA are shown in Figure 8-4 for nodes 8-15 (in Figure 8-3) on the cold leg of the largest radius U-bend. For the - structural evaluation, the tube loads result from the hot to-cold log &P. Plots showing the l hot-to-cold leg &P for the maximum and average radius tubes are provided in Figures 8 5 and l 8-6, respectively. For the rarefaction wave induced loadings, the predominant motion of the U bonds is in the plane of the U-Bend. Thus, the individual tube motions are not coupled by the anti-vibration bars. Also, oniy the U-bend region is subjected to high bending stresses. Therefore, the structural , analysis is performed using single tube models limited to the U-bend and the straight-leg region l' over the top two TSP's. The node and element numbering for a typical single tube model is shown in Figure 8-7. The tube structural model consists of three-dimensional straight and curved pipe eternents. The mass inertia is input as effective material density and includes the weight of the tube, weight of the primary fluid inside the tube and the hydrodynamic mass effects of the secondary fluid, Damping coefficients are defined to realize a maximum damping of 4% at the lowest and highest significant frequencies of the structure.
, To account for the varying nature of tne tube / TSP interface with increasing tube deflection, three sets of boundary conditions are considered. For the first case, the tube is assumed to be laterally supported at the TSP, but is free to rotate. This is designated as the " continuous"
, mndition, as the finite element model for this case models the tube down to the second TSP. As the tube is loaded,it moves laterally and rotates within the TSP. After a finite amount of rotation, the tube will become wedged within the TSP and is no longer able to rotate. The second set of 8-3 l
boundary conditions, therefore, considers the tube to be fixed at the top TSP location, and is
- referred to as the " fixed" case. Continued tube loading causes the tubo to yleid in bending at the top TSP and eventually a plastic hinge develops. This represents the third set of boundary ;
conditions, and is referred to as the " pinned" caso, - For the averago radius tube, only the continuous caso is analyzed. Results for the continuous caso analysis indicate that both the tubo rotations and moments at the TSP nodes are small compared to - . thoso required to causo the locking in or plastic hinge, respectively, at the support locations. Since the main objective in analyzing the average radius tube is to determine the maximum roaction load on the TSP due to the overall responso of the tube bundle, the continuous , configuration is the most appropriate for the average radius tube analysis, in addition to the pressure induced bonding loads, the rarefaction wave analysis also includes the membrane stresses due to the primary to-secondary &P. Each of the dynamic solutions results in a force timo history acting on the TSP. Those time histories show that the peak responses do not occur at the same timo during the transient. However, it is assumed for this analysis that the maximum reaction forces occur simultaneously. Using those results, a TSP load corresponding to the overall bundle is then calculated. A summary of the resulting TSP forces is provided in Section 8.1.3. 8.1.2.2 LOCA Shaking Loads Concurrent with the rarefaction wave loading during a LOCA, the tubo bundio is subjected to additional bending loads due to the shaking of the steam generator caused by the break hydraulics and reactor coolant loop motion. However, the resulting TSP loads from this motion are small compared to those due to the rarefaction wave induced motion. - To obtain the LOCA induced hydraulic forcing functions, a dynamic blowdown analysis is , performed to obtain the system hydraulic forcing functions assuming an instantaneous (1.0 msec break opening time) doublo-ended guillotino break. The hydraulic forcing functions are then applied, along with the displacement time-history of the reactor pressu e vessel (obtained from a separate reactor vessel blowdown analysis), to a system structural model, which includes the steam generator, the reactor coolant pump and the primary piping. This analysis yields the time history displacements of the steam generator at its upper lateral and lower support nodes. These time-history displacements formulate the forcing functions for obtaining the TSP loads due to LOCA shaking of the steam generator. To evaluate the steam generator response to LOCA shaking loads, the computer code WECAN is used. The model used is similar to the one used for the seismic analysis, discussed previously. The steam generator support elements are removed, however, as the LOCA system model accounts for their influence on the steam generator response, input to the WECAN model is in the form of acceleration time histories at the tube /tubesheet interface. These accelerations are obtained by differentiation of the system model displacement tirne histories at this location. Acceleration time histories for all six degrees of freedom are used. Past experience has shown that LOCA shaking loads are small when compared to LOCA rarefaction loads. For this analysis, these loads are obtained from the results of a prior analysis for a Model D steam generator. 8.1.3 Combined Plate Loads A summary of the resulting LOCA and seismic loads is provided in Table 8-1. In combining loads, , the LOCA shaking and LOCA rarefaction loads are combined algebraically, while LOCA and SSE loads are combined using the square root of the sum of the squares. The TSP loads, which are .,
+
8-4 _, o-, .--4, v-- --'-c r
reacted by wedge groups located at their periphery, are divided into three groups based on the wedge group arrangements for the plates. The number of wedge groups varies in number, size, and orlontation among the various platos. The wedge group orlontation for plates C/D and G/H is shown in Figure 8 8, and for plates O T in Figure 8 9. The wedge group sizes and locations for the romalning plates are combinations of platos C/D and G'H, Typically, the wedge groups are symmetrical about the centerline of the bundio, and also hot-leg to cold-leg. TSP C/D and G'H i are two cases where hot-to-cold log symmotry does not exist. Relative to wedge group size for the Catawba steam generators, the wodge groups are comprisod of olther two or three wedges, each two inches wide. Thus, the overall wedge group size is either - 4 or 6 inchos. The wodge group size is important, because it affects the local distribution of load into the nolghboring tubes. In reacting the load among the various wedge groups, a cosino distribution is assumed among the wodges that are loaded. Typically, only half of the wedge groups are loaded at any given time. In determining the distribution of load for seismic and LOCA loads, tho directionality of the load is considorod. LOCA loads are uni-directional, in that they only act in the plano of the U-bend. Solsmic loads on the other hand are random, and can act in any direction. Calculations are performed to determine load factors for tho various plates, grouping the TSP by commonality of their wedge group locations. The load factors are not a function of the wedge group size, only of location. Applying these load factors to the overall TSP loads in Table 8-1, loads for each of the wedge groups are determined. A summary of the individual wedge toads is provided in Table 8-2, 8.1.4 Tube Doformation in estimating the number of deformed tubes, the results of TSP crush tests for Model D steam generators are used. The deformation criteria for establishing a tube as being susceptible to in-loakage has boon defined to be [- ]a c. In reporting the crush test results, tube deformations wore reported for various deformation magnitudes. This is the smallest deformation reported. Although test data is not available for leak rate as a function of tube deformation, it is judged that deformation levels of this magnitude will not result in significant in-leakage. Using the crush test data, a correlation is developed between elastic plate load and the number of tubes that would have a deformation of ( -}a,c or greater, it is this correlation,- summarized in Table 8 3, that is used to approximato the number of affected tubes. Summarios of the number of potontially affected tubes for each of the wedge groups are provided in Tables 8-4 through 8 7,8.1.5 Tube Maps / Summary Tables for Potentially Affected Tubes Catawba is a four-loop plant. The numbering scheme used by Duke Power in identifying tube rows and columns is the same for all foul 'ocps. The referenco configuration used in identifying tube locations is shown in Figuro 8-10. As shown in the figure, the nozzle and tubo column 1 are located al 0.
. Maps showing the location of the potentially susceptible tubes are provided in Figures 811 through 8 21. Tabular summaries of the tubes that are potentially susceptible to collapse and subsequent in-loakage are summarized in Tables 8-8 through 8-13.
Identification of the potentially susceptible tubes is based on the crush test results. The wedge / tube configurations considered in the tests are not identical to those for the Catawba steam 85 l
- - - - - . - - - - . - . . - - ~ - . - - . . - - - -. .- - . . -.
generators. As such, l_t is not possible to identify exactly the tubes that might be limiting at each
= wedge group. Thus, riue to the uncertainties involved, there are more tubes identified at each wedge group as being limiting than estimated in the f alculations.
Finally, Table 8-14 provides an index of the applicable tables and figures identifying the potentlafly susceptible tubes for each TSP. 8.2 Tube Deformatbn Under Combined SLB + SSE Since the tube support plates provide lateral support to tube deformation that may occur during
- postulated accident conditions, tube bending stress is induced at the TSP intersections. This -
bending stress is distributed around the circumference of the tube cross section, tension on one side and compression on the other side, and is oriented in the axial (along the tube axis) direction. Axlal cracks distributed around the circumference will therefore either experience tension stress that tends to close the crack or compressive stress that tends to open the crack. The compressive stress has the potential then to reduce the burst capability of the cracked tube due to the crack opening. Test results are summarized in Tabte 8-15 that demonstrates that an outer diameter bending > stress on the order of the yield strength of the tube materialis required before any significant effect is realized in the burst pressure capability of a cracked tube (WCAP-7832-A). A tube with through wall slots (Figure 8 22) was tested under combined beam bending and internal pressure to achieve the burst pressures listed. The 0.8 inch long through wall slots were oriented on the compressive and tensile sides and on the bending neutral axis. The neutral axis - and tensile side burst results are almost identical and within normal data scatter of the burst pressure without bending stress. Burst capability during accident conditions is required to be at least 2650 psi, the maximum primary to secondary pressure differential following the SLB or FLB. Regulatory Guide 1.121 requires that SSE be combined with these events. Therefore,- tube bending stresses from SSE at the TSP elevations must be combined with the SLB/FLB pressure differential to establish that burst capability meets SLB/FLB requirements. Rather than retest burst capability _with combined bending stress,it is sufficient to establish that the SSE maximum tube bending stress at any TSP elevation is less than the tube material yield strength at operating temperature based - on the test results of Table 8-15. Based o_n the results of the seismic analysis judged to be conservative for Catawba, discussed in Section 8.1.1, the maximum tube bending stress occurs at the top TSP and has a magnitude of 34.07 ksi for tubes with end of life wall thinning due to general erosion and corrosion. For the loser TSPs, the tube bending stress is 6.59 ksi. The yield strength of 3/4 x 0.043 inch mill annealad tubing at 650 0F is 39 ksi using 95% confidence / 95% probability, lower tolerance limit properties (WCAP-12522). Burst capability is therefore not affected by the SSE bending stress. 9 8-6
T Tablo 81 Summary of LOCA Plus Seismic TSP Loads Catawba Unit 1 Steam Generators Steam Generator inlet Break W t 4 e I 8-7
Table B-2 Summary of TSP Wedge Loads Catawba Unit 1 Steam Generators 8 ' O e mes M e 8- 8
i 1 Table 8-3 Number of Tubes with AD > 0.030 inch Versus Losd Catawba Unit 1. Model D Steam Generator
~
_a. a
'W en W
8-9
Tabio 8-4 Number of Tubes with AC > 0.030 inch Catawba Unit 1 Steam Generators Steam Generator inlet Break , TSP C/D, L/M 1 a l i I 9 e W e O M 8-10
Table B Number of Tubes with AD > 0.030 inch Catawba Unit 1 Steam Generators
, Steam Generator inlet Break TSP E, F, G/H, J, K, N, P 9
uses " s l i
'l 4
e Q W 8 11
Table 8 6 Number of Tubes with AD > 0.030 inch Catawba Unit 1 Stearn Generators Steam Generator inlet Break - TSP 0, R, S _ a e D' Esp pr 8-12
Table 8 7. Number of Tubes with AD > 0.030 inch Catawba Unit 1 Steam Generators
- Steam Generator inlet Break TSP T
~
a 9 l l 1 r 9 I l l 3 13 j l
Table 8-8 Summary of Tubes Excluded fromlPC Catawba Unit 1 Steam Generators TSP C/D - 8 1 e e ma M 4 6 9 8 14
Table 8-9 Summary of Tubes Excluded from lPC Catawba Unit 1 Steam Generators . TSP G/H
- 8 N
m M 4 e 8-15
Table 810 Summary of Tubts Excluded from IPC Catawba Unit 1 S'eam Generators TSP UM . O MWee m 49 l f v w 8-16 I
Table 811 Summary of Tubes Excluded from lPC Catawba Unit 1 Steam Generators , TSP E, F, J, K, N, P 6 a I e tm e w 4 8-17
Table 812 Summary of Tubes Excluded from IPC Catawba Unit 1 Steam Generators TSP 0, R, S . a r e M M l 1 l . 8 18
Table 813 Summary of Tubes Excluded from IPC Catawba Unit 1 Steam Generators
. TSP T O
- a o
N 9 l = a 8-19
.=-
Table 814 Sumrnary of Tables and Figures for TSP Row / Column Identification
SUMMARY
TUBE MAP TSP TABLES FioURES C 8-8 8-11 D 8-8 8 12 E 8-11 8-17 F 8-11 8-17 G 8-9 8-13,8 14 H 8-9 8-15,8-16 I J 8 11 8-17 K 8-11 8-17 L 8-10 8-11 M 8-10 8-17 N P 11 8-17 P 8-11 8 17 0 8-12 8-18,8-19 R 8-12 8-18,8-19 S 8-12 B 18,8-19 T _ 8-13 8-20, 8-21 a. 6 8 20
Table 815 Combined Bending and interrul Pressure Burst Tests on Tubes with Through Wall Slots e
- a D
e i lume m h I l l I l-I 8- 21 l
., - .-- . . . . . . . - . . - . ~ - . . - . . . . . - - . . . . . - - - - . . - . _ . . .
t a 4 . i e i Figure 8-1. Seismic Model Representation of Steam Generator - 8 22 i l-l _ . _ _ _ . . _. .._ _.-..._ _ . .- _ _ . . _ _ _ _ - _ _ _ . . . ,. . , _ .- __ - . _. . _ _ , =
a 0 Figure 8-2. Tube Bundle Geometry 8- 23 l l
- a Fic .re 8 3. Thermal / Hydraufc LOCA Tube Model -
8- 24 _ m
. a t
\. Figure 8 4. LOCA Pressure Time Histories, Maximum Radius Tube Nodes 815 S- 25
a Figure 8 5. LOCA Pressure Time History, Hot toCcki Leg Pressuro Differential, Maximum Radius Tube . 8 26
_a 4 Figure 8 6. LOCA Pressure Time History, Hot to-Cold Leg Pressure Differential, Average Radius Tube 8 27
i l l i 4 l l a
- n. 1 I
L t i
-h I
L t i t L i r l Figure 8 7. Structural LOCA Tube Model . 1 . r l l 8 26 i. e I-
. , . . . - , . . . - , . . . . _ , _ . - - - , . . , _ , . . . . , , . . - _ . . . - _ ~..._...--,. ~ .- ., .. . .-,s , . . , . . . , , . - , - , . , - , , . . . . - - - . , . - . . , - . - , . _ . . . - - ,
_a e Figure 8-8. Wedge Group Orientation, Plates C/D, G/H 8-29
d v . Figure 8 9. Wedge Group Orientation, Plates O T 8 30
-__a-u--_.a_ - --n--- ---
P 90* i Quadrant 2 l
- Quadrant 1
, Hot leg, i
DMder Plate i i
/ 0 o
%: 0 180. -
g , i 360 k Colurm 1 f Cob m 114 . l l i Quadrant 3 Cold Leg Quadrant 4 o 270 Figure 810. Reference Configuration for Tube Identification Looking Down on Steam Generator. Catawba Site Specific Convention 8 31
, a t
F {
- a ;
t i j .- b b t r I P h i l k 3 Figure B 11. Tubes Excluded from IPC, TSP C.L - Quadrant 1 8 32 , 9sh->e.4-
.&y - g ,av se .-a---, y ,-e-- .q, .r.,.y2., .,n.yrm . e %>w we-.,mp. .s >my ,e_ ,s--,.e,i. e gy&-*r r.+r'unr+anywwy-.ee'3:.,. ewe.w--m--w .',e.msaa.-w g.e- --w 4 .w 44 vu-te
t 4 k b
-- a !
. r 1
i p m 4 i I-
?
I Figure B 12 Tubes Exduded from IPC TSP O - Quadrant 4 s i 8 33 . 1 l c-.-1., . , . -,-n ,r . , ~ , . , . , , .-n, - . , . - , , - , , - , . . - - , - - . . . ~ - , ~ ~ . - , . . ~ , . . , , - ~ - - - , .
N a - Figure 8-13. Tubes Excludod From IPC, TSP G - Quadrant 1
- 8 34 i
t i 1
. 8 6
i r i i
\
i I
?
f P 1 F t s 1
.3 i.
.L
?
J Figure 814. Tubes Excluded from IPC, TSP G - Quadrant 2 8 35 . i
.,.._c , .. ., . . -. . , _ _ , . _ , . - . _ _ . . , _ . , , . . . . , _ . . . . , _ , . . , _ _ , . , _ . . . , _ _ . . . . _ . . _ , . . , _ . , . . _ . , , . . . . . . . _ . . , , _ . , ..,, __.___... ._...m., -
. - . . . . . = . . . _ . - _ .- . - - . - - - - - - - - . - . - - . . . . _ . = . . . _ . . .
a I l i I i t e f 4 I I I l' - - _ t , 4 t l Figure 815. Tubes Excluded from IPC TSP H - Quadrant 3 l-l l 8 36 e i I 1
-i
- --.,.,_,-,~,,.,... . , - - , - - + ,_,.,,,,,.,,.,,,,,,,_,,_n,,,,,.,,,.-,,_._.._,..._-.,_,..,,...,,,,,_.,w n, ,, . . . , . . . . . ~ , . . . ,
a C Figure 816. Tubes Excluded from IPC, TSP H - Quadrent 4 8 37
mm a a i a 4 4 Figure 817. Tubes Excluded from IPC, TSP E, F, J, K, M, N, P - Quadrant 4
- 8 38
.. _ _ _ _ . _ . _ _ - . _ . _ . . _ . . . . . . _ _ _ _ _ _ . _ _ _ _ . . .. _ _ - _ . . . - .~___ _ _ _ _. _ . _ . ._. _ _ _ _ _ _ _ _ .
1
?
i
)
i
. a- ,
l i I a k I t 6 t 1 P l
- l. -
1 1 1 Figure 818. Tubes Excluded from IPC, TSP O. R. S - Quadrant 1 l 8 39 I [. j i avurc men-..r--.-- , -e--, . . e.,- -,,,%.. . - . - - . -- , ..-+e4.---er e.+...-- .,e,*,-.e--re,ewr--- . , - - ,..%w-.,-,-esw -% ww m, yev.-_- m x , .m . m.,- -% , m. - p.+ y. .
i
?
r i a . . h 4 I I L Figure 819. Tubes Excluded from IPC, TSP O. R S - Quadrant 2 8 40
.,_my,y, m.,
. .-~ , y v - 4.-f.- , e .m s.- .,m.. . ., ., - ,,,.,,,,,. . ,, c .e., .
..,.3...,,.w.,..y...- .,.cyn..,,,o7, .,w #,
9..
)
1 i l' , a l I i l E t i e f i c Figure 8-20. Tubes Excluded From IPC, TSP T - Quadrant 1 8 41
i 1 i a a ,, 4 t t d Figure B 21. Tubes Excluded from IPC, TSP T - Quadrant 2 - 8 42 ;
, , - . . . . . ~ , . . . . . _ , _ . . . . - - - . . , . . - - , . , _ - . . . -, , , . . . . . . , _ . . . , . . - . . ,,..,,, ,..,.,. - , .._.-.,-..,- . .
l l 9 t I P/2 P/2 4- 2" > < 3" y 4- 2" % Y y I r o, _ y
~
l J d [ J k 1 2 P/2 P/2 i Figure 8-22 Extemally Applied Bending Load ai d Locations of Through Wall Slots 8 43
- - _ - - ~__...- - - - - -- - - - - .- -- - ._-_-
F i Section 9 ; F CATAWBA UNIT 1 INSPECTION RESULTS [ 9.1 Inspection Scope For the end of cycle 6 inspection (IEOC6 '92) 100% of the support plate intersections in each stoam generator, both hot log and cold log, woro inspoctod with bobbin probos . 0.610 inch diamotor for the het log and 0.630 inch diamotor for the cold leg. Allintersections exhibiting possiblo flaw indications wore recorded notwithstanding the amplitudo of the signal Eddy current (EC) frequenclos used for analysis of the bobbin data were as follows: 550 kHz . prime inspection frequency; 400 kHz for continuity with prior inspections; 130 kHz for support plate t suppression in mixing; and 35 kHz for support plato location. The Duke Power EC guidelines for Catawba Unit 1 used each of the flaw sensitive frequencies (550 kHz,400 kHz,130 kHz) along ; with several mix channels for flaw detection. Measurements of possible ODSCC voltages were performed in the 550 kHz/130 kHz differential mix channel to provide consistoney with the industry data base, as described in Section 5. RPC (rotating pancake coll) inspection was performed for all possible support plato bobbin Indications equal to or greater than 1.0 volt, as wo!! as on a substantial fraction of the possible indications below 1.0 volt. All bobbin indications 1.0 volt or groater in amplitude, which were confirmed as flaws with the RPC probo, were designated for repalrt those which were , unconfirmed and those less than 1.0 volt whether tested or not are the subject of this report. 9.2 Summary of Inspection Results As a result of the hot log bobbin inspection program,6941 support plato indications possibly reflecting ODSCC wore identified. Table 91 shows the distribution of the TSP indications among - the four steam generators. The amplitude distribution of these Indications was such that loss than 5% exceeded 1.0 volt. Prior outage tube removal operations (IEOCS) had estabilshed the probable nature of theso Indications as ODSCC, similar in character to the degradation documented at Plants A 1. A 2, L, D.1 and C 2. The dictribution of the indications relative to elevation in the tube bundle, shown in Table 9 2, confirmed the pattern of greatest incidence at the lower hot leg support plates which has been typical of support plate ODSCC observed in other plants. Figuro 91 illustratos the data in histogram format. The 1 H level is a flow distribution baffle with enlarged tube holes which directs increased flow past the tube surface; the concentration of aggressive species is limited compared to the nominal support plato creWce geometry. This explains why the number of indications are low at the 1H location. RPC testing was performed on all indication: greater than 1.0 volt and a substantial fraction of those 1.0 volt or less. The results of this program are tabulated in Table 9-3. To satisY the . interim plugging criteria commitmonts, an augmented RPC program was conducted over and above the confirmatory testing of all support plate bobbin flaw indications 2,1.0 volt. This program contained two elements, as follows:
- 1) Support plate dent indications 25.0 volts,
- 2) Support plato non flaw residuals (artifacts) with amplitudos greater than 1.0 volt.
91 -. . - - - - - .- - - _- - - - - - - . ~ - - - - , - ,,_ - -,, -, ---_
l 7hoso signal categorios were tested to provido confidenco that axial cracking indications 21.0
- volt woro not being mashed by interforences. Tablo 9 4 provides for each steam generator the number of such locations /tubos which woro examinod; none of theso intersections 136 dents, 2 artif acts - exhibited ODSCC indications consistent with bobbin Indications 12 0 volt, nor was any circumforontial cracking notod. Small axial indications observed wtro plaled on the tubo repair list.
Except for steam generator C, very low fractions of the bobbin Indications examined with RPC probos oxhibited the patterns associated with axial ODSCC. The Zoton 3 coll MRPC probes used incorporato e pancake coil, an axially wound coil proferentially sensitive to circumferontlal cracking and a circumferentially-wound coil preferentially sensitivo to axial cracking. The observed distribution of the bobbin indication relativo to support plato elevation coupled with the generally low RPC confirmation rato supports the evaluation tnat significant support plato ODSCC is found only on hot leg side of the tubo bundle. Cold log indications woro considered independently of the IPC criteria; disposition was mado in conformance with tho 9sual Toch. Spec. plugging critoria. 9.3 Cro s Calibration of ASME Standards in order to reduce the uncertainty associated with the bobbin voltage amptitudes from the current (1992) inspection results, the cahbration standards used during the inspection were cross calibrated against the reference laboratory standard used in the APC data acquisition. This was accomplished as described below. - A new ASME standard (AS-015 91) was first eddy current tested in the laboratory along with ' the referenco laboratory standard (AS-009-91) used for the APC data acquisition. The former is called the transfer standard. The test was performed usnig a Zotec 610 ULC probo and was repeated ton times (ten separato passes through the standards). The test results indicated that for the 550/130 kHz mix, when the 20% holos in the reference laboratory standard was set to 2.75 volts, the corresponding reading for the 20% holes in the tre.nsfer standard was 3.22 volts. _ The transfer standard was then shipped to the Catawba sitt and eat of the ASME standards used during the 1992 inspection was calibrated against the transfor standard. These tosts were also repeated ten timos. i The results of the cross calibration for the various ASME standards are listed in Tablo 9-5. Jsing this data, correction factors were developed for each ASME standard. The listing of ASME standards used for the various eddy curront data tapes in each steam generator during the 1992 laspection are summarized in Table 9-6. The appropriato correction factor was applied to each TSP indication based on which ASME standard was used during the data acquisition for that tube. The bobbin coil eddy current voltage amplitudes developed in this manner formed the basis for the assessment of TSP indications. 9,4 1992 Inspection Results at TbP Elevations . The oddy current data ana'; sis guidelines applied for the evatuation of indications at TSP elevations are described in Appendix A. Further, the bobbin amphtudes were multiplied by the . appropriato correction factors developed from the cross calibration of the ASME standards as discussed in Section 9.3 above. Thus: 9- 2
. N
- 1) The TSP indication distributions were developed based on oddy current results at the 550/130 kHz differential mix, since this data is availablo from the current (1992) inspection. This frequency mix is consistent with the APC database.
- 2) The oddy current data analysis was performed by normalizing the 550/130 kHz mix signal for the 20% holes in the ASME standards to 2.75 volts.
There were a total of 6941 Indications in the population: 713 in steam generator A,781 in B, 1970 in C, and 3477 in D. With the exception of ono 3.55 voit indication in steam generator C, allindications were loss than 3 volts in amplitude. A frequency distribution of the TSP Indications in all generators is shown in Figuro 9 2. A cumulativo probability distribution is also shown lo the figure.11 shows that over 97% of the indications are less than 1 voit in amplitudo and 76% below 0.5 volt. The averago signal amplitudo of allindications is 0.38 volt. Tho frequency distribution for each of the four steam generators is shown in Figures 9 3 through 9 6. 9.5 Voltage Growth Rates For the 1992 outage, data analysis had first been performed using analysis guidelines typica!Iy used for depth based plugging critoria. In order to apply voltago based plugging criteria, the data from indications at tubo support plates (TSP) were reanalyzed using guidelinos consistent with the application of voltago based plugging criteria (APC). Tho specific guidelinos used for this reanalysis is described in Appindix A. The ovaluation described below was performed using he reanalysis results. Tho voltage growth rato evaluation was perforraed using the results of 541 TSP Indications. These indications woro selected by Duke Power based on tholt (largest) amplitudes from tFe original analysis. Inclusion of all TSP Indications in the growth rate evaluation would have added over a week to the initiation and completion of this critical task. Since the largest amptitude indications from the curront outage were used for this analysis, the calculated growth rates (averages and 90 and 99 percontiles) are likely to be very conservativo. Tho 541 Indications - were made up of 90,117,197 and 137 from steam generators A, B, C and D, respectively, in order to determine growth rates, voltage amplitudes of an indication during successive inspections should be comparod. Therefore, the eddy current data for these 541 indications from the 1991 inspection woro reanalyzed using the same guidelines (At%chment A) so as to get a consistent set. The existing APC database for 3/4 inch chameter tubos uses eddy current results from 550/130 kPz differential mix. In order to accommodate this, the current (1992) Catawba 1 eddy current inspection v,as performed at frequencies including 130,400 and 550 kHz. However, the 1991 Catawba 1 inspection did not utilize the 550 or 130 kHz frequency. The closest frequency mix availablo was 400/100 kHz. This frequency mix was therefore used for the 1991 data. For the growth estimation therefore,it was decided to use the 400/130 kHz rnix from the current inspection.
. To convert the signal amplitudes at 400/100 kHz to an equivalent result at 550/130 kHz, a conversion factor was used. This correlation was developed from the eddy current test results of Catawba 1 pulled tubes performed in the past. The data and the excellent regression fit are shown in Figure 5-3. The correlation is:
V (550/130) - 1.094
- V (400/100) + 0.143 (91) 93
i where V (550/130) and V (400/100) are signal amplitudos at the two corrosponding i frequoncles. This correction is applied to both the 1992 data (400/130 kHz) and the 1991 data (400/100
- kHz). The voltago growth was dolormined for each of the 541 indications.
The above correction factor (Equation 91) is an overcompensation for the 1992 results sinco this data is at 440/130 rather than 400/100 kHz. Comparison of current results from a ; samplo of 96 indications from steam generator C of Catawba 1 (soo Figuro 9 7) suggests a factor of: V (550/130) - 1.038
- V (400/130) 0.047 (92)
In ordor to account for this, the calculated growth ratos were reduced by a factor of 0.95, this being the ratio of the coefficients in the abovo equations (1.038/1.094). This was substantiated Indopondently by data obtained on an ASME standard using the throo mixes. This data is listed in Tablo 5-2. In each case, the mix channel was normalized to 2.75 volts for - the 20% holos and data taken for ihn other machined flaws in the standard. The results show increasing voltage amplitudos for tho 400/100,400/130, and 550/130 kHz,in that order. Further, the magnitude of the difference is comparable to the 5% value used. The resulting growth rates are plotted against beginning of cycle (BOC) amplitudos in Figuro 9-8. A frequency distribution of the growth ratas is shown in Figuro 9 9. This figure also shows a cumulativo distribution function. With tho exception of ono indication with a growth
- rate of 2.7 volts, allindications had voltage growths loss than 2 volts. Indood, only 10 of the 541 Indications (loss than 2%) had growth rates exceeding i volt. ,
Tho avr. ago growth rate foi allindications in the growth rato study was 0.18 voit during the cyclo it may bo noted from Figuro 9-8 that, in general, tho indications with larger BOC amplitudos had loser growth rates. To assess this further, the growth rate averages were dolormined separately for indications with BOC amplitudos below and abovo 0.75 volt. The averago growth for indications with BOC amplitudos below 0.75 volt was 0.21 voWcycle. The average growth for BOC amplitudos at or abovo 0.75 volt was 0.14 volt / cyclo, considorably smaller. Those results as well as some other statics are summarlzod in Tablo 9 7. [ The 199192 operating cycle (Cyclo 6) consisted of 0,80 offectivo full power years (EFPY) of - operation. Per Duke Power; the next operating cycle is planned to be 0.96 EFPY. Thorofore, growth rato projection for the next cycle is obtained using prorating the Cycle 6 growth rates by the ratio of the EFPYs (0.96/0.80). A frequency distribution of the EFPY adjusted growth projecCons (for Cycle 7)is shown in Figuro 9-10. A cumulativo distribution function is also - i
- shown in the figure, i Growth rato during Cycle 5 was determined using reanalyzed signal amplitudos of TSP indications l In tubos plugged during 1991. The reanalysis was performed using the oddy current analysis guidelines shown in Appendix A.158 tubes had been plugged for TSP Indications during tho -
i 1991 outage. Eddy current results from two consecutive inspections are nooded to estimato growth ratos. NDE results from both 1990 and 91 were available for only 126 of these indications. . 9-4 i
. . . _ _ _ . _ _ . . . - . . . - . - ~ m_.~,.~. _.-..-_ _ . , - ~ . , . . . . - - . . _ - .- . . - . - . . - - - - , . . - , , - . ~ . . . , --
i 1 These data ,both 1990 and 1991) Wore taken at 400/100 kHz mix. The adjustment factor obtained from the laboratory NDE of the Catawba 1 pulled tubes (Equation 91) was app!!ed to convert the above signal amplitudes to equivalent 550/130 kHz voltages. Growth rates were a then calculated as the difference between tne 1991 and 1990 values. The average growth rate for the cycle was 0.10 volt, lower than the 0.18 volt t alculated for the last cycle. Those results , are summarized in Table 9-7. Although the average was smaller, the standard deviation _ of 0.35 l volt was comparable to the 0.36 volt obtained for the 199192 cycle. Th0 resulting average percent growth rate is 13% for cycle 5. The growth ratos were higher for indications with BOC amplitudes smaller than 0.75 volt compared to those with BOC amp!!tudes greater than or equal to 0.75 volt. This is consistent with what has boon observed at other domestic plants. A irequency distribution of the voltage growth rates during Cycle 5 is shown in Figure 911. The upper ends of the voitage ranges are shown in the abscissa. Also a cumulative frequency distribution in percentage is shown in the figuro as a curve with the scale shown on the right hand sido. As discussed above, only one intersoction (tube R12 C111 at 2H in steam generator C) had growth rate excooding 2 volts during Cycle 6. Further evaluation of this indication was performed as described below. This indication had amplitudes of 3.39 volts in the 400/130 kHz and 3.44 volts in the 550/130 kHz mix channels. Converting the 400/130 kHz value using-Equation 9 2 results in a value of 3.47 volts at 550/130 kHz. This agrees with the actual , measured value of 3.44 volts. The ASME standard with serial number 50415 was used during - the bobbin coil testing of this tube in 1992 (see Table 9 6). Using the correction factor of 1.033 applicablo for this standard results in a corrected 1992 amplitude of 3.55 volts. Resizing of the 1991 data for this indication performed during 1992 resulted in an amplitudo of 0.78 volts at the 400/100 kHz test frequency. Since bobbin testing at the 550 kHz frequency was not performed during the 1991 inspection, the 400/100 kHz data was ussed along with Equation 91 to obtain the voltage for 1991. This resulted in an amplitude of 1.00 volts at the 550/130 kHz mix. During 1991, the ASME standard 50290 was used for this tube. This ASME standard has a correction factor of 1.236. Thus the corrected amplitude for this Indication from 1991 is 1.24 volts. This results in a growth rate of 2.31 volts (3.551.24) during Cycle 6. This is the largest growth observed at any TSP intersection during Cycle 6. This also reveals that the growth rates for most intersections (particularly in steam generator C) are samewhat lower than calculated above without the detailed accounting for the correction factors in ASME standards. The projected maximum growth rate for Cycle 7, using the adjustment based on EFPY ratios between Cycles 7 and 6, is then 2.77 volts. 9-5 l l l -_ _ _ _ _ _ . _ _ - __ - - - ~ . . _- - . _, _ ,
Tablo 91 Catawba Unit 1 EOC-6 Hot Lc0 Support Plate Bobbin Indications Indications Steam Generator Tota! TSP Indications >1.0 vo!t - 1A 713 6 1B 781 7 1C 1970 100 1D 3477 57 Total 6941 170 b o O 9- 6 l
_..___._.~__>.___.___m__..__..______ _ _ ._. Table 9-2 f Catawba Unit 1 Distribution of Indications Over Support Plates Elevations 1 Hot Leg steam Generator TSP Etovatica JA 1B 1G 10 .Talal - 1H 1 0 1 0 2 2H- 196 155 548 469 1368 3H 139 218 445 1002 1804 ! 4H 109 185 474 839 1607 i SH 60 102 197 347. 7.06 ; 6H 52 40 154 493 736
.. 7H 110 43 134 266 fin 8H 46 38 17 61 162 Total 713 781 1970 3477 6941
- i i
k r k l e 97 l
--e rr,-- ,-, - --- .-,c,. ~ , , . . . , - . , - . - -. -,-,-ey-wm,.r- -*w., ..+-...,v,..-v- w vn .~m -r em- r + w t . -s r - ,es--r m ' J <m~---=,m -
r--e.se .- *ev --
Table 9-3 Catawba Unit 1 EOCG RPC Test Results at TSP Locations RPC RPC Sicam Generator Ey.amined Crnfitmed 1A 50 5 1B 223 7 1C 896 263 10 1358 130 Total 2527 405 O I o 98
t. Tablo 9-4 Number of Tubes Examined During EOC-6 Augmented RPC Program t
~
32A M SIC M Dents , 000 Tubos 14 16 43 5 Ind 16 67 47 6 MRPC ; Tubes 4 58 38 4 Ind 4 64 42 5 i Artifacts (R36) BOB i Tubes 0 0 2 0 Ins 0 0 2 0 l MRPC 1 Tubes 0 0 2 0 l . lnd 0 0 2 0 l l I I l e 1 - l 9-0
- l. .
I.
- n.m,- , ,,~.,-vv,nm-An.-m.,,-e.,,,-,,--.e.,_,, e, - - - - - . . . ~ . . ~ , . , ,-,-..n-.- .---,mm- ,--,--- . - - - - , - . . , , , - - - - - - - - - - ---~r-
Table 9 5 Cross Calibration of ASME Standards Against theTransfer Standard , Mt0BE 590 610 Is3LC 610 SFist 630 isA.C Pi P1 1 P1 1 P1 standerd 1 i 7 APE 1 AS 01591 4.560 3.246 4.A 3.236 4.560 3.241 4.560 3.244 2-9617 3.316 2.377 3.382 2.389 , 3.329 2.345 3.?74 2.386 _ TAPE 2 50390 4.287 2.994 3.980 2.783 4.032 2.806 4.112 2.871 50391 3.954 2.760 3.844 2.710 3.853 2.670 3.786 2.636 50392 4.408 3.131 4.125 2.939 4.039 1.829 4.055 2.869 , 50415 4.229 3.141 3.992 2.841 3.916 7.749 3.971 2.815 50416 3.965 2.839 3.852 2.736 3.885 2.750 3.797 2.714 50417 4.469 3.178 4.365 3.134 4.293 3.067 4.199 3.004 50418 4.1% 2.967 3.895 2.746 3.840 2.675 3.831 2.690 50419 3.823 2.661 3.607 2.54 3.619 2.507 3.617 2.562 A5'01591 4.561 3.183 ' 4.560 3.224 4.560 3.207 4.560 3J256 1: 550 Hz - P1: 550/130 Hz Mix e 9-10 1
Table 9 6 Catawba Unit 1,1992 Inspection ASME Standards Used and Corresponding Calibration Corrections Steam Tapes Numbers Calibrated Calibration Generator ASME Standard .Usino the ASME Standard Correction g4 gyv A 50391 1 to 37 0.9855 {
-q; A 50417 39 to 84 1.1396
-fu
. a.i B 50415 50 to 85 1.0331 ij B 50417 1 to 49 1.1396 B 50418 86 0.9985 C 50415 1 to 48 1.0331 C 50417 49 1.1396
+ C 50418 50 to 85 0.9985 D 50391 38,39,41 0.9855 D 50416 39,41 to 50,52 to 60,82,83 0.9949 D 50!17 84,85 1.1396 D 50419 1 to 37,40,51,86,87 0.9258 m 9-11
Table 9-7 Catawba Unit 1 Growth Rate Statistics Number of Avorage Growth Rate Percent Projected 92-93 (1) IDdications BOC Vohs Average Stdj2cy. Growth Growth Rate - Cycle 6 (199192)
. I.ota! Pooulation of Bobbin IndicatioDS 4' p w NA Entire Sample 6941 0.55 0.01 0.23 2 0.02 ?N./l f BOC Rancos: .f;e.
g.e _ 1.. M 5j.! VDOC < 0.75 volt 5846 0.48 0.03 0.19 6 0.05 pl;Q M' VBOC z 0.75 volt 1095 0.92 -0.16 0.33 -17 -0.15 VDOC s 1.00 volt 6699 0.53 0.01 0.21 2 0.03 Pletiminaryltudy of 541 Laroest Bobbin Indications Entire Sample 541 0.71 0.18 0.36 27 0.19 BOC Rancos: VBOC < 0.75 volt 318 0.53 0.21 0.31 40 0.21 VBOC 2 0.75 volt 223 0.96 0.14 0.41 15 0.15 ' VBOC s 1.00 volt 471 0.64 0.18 0.31 28 0.19 Cycle 5 (1990-91) Entiro Sample (2) 126 0.74 0.10 0.35 13 BOC Ranoes: VBOC < 0.75 volt 78 0.55 0.20 0.24 37 VBOC z 0.75 volt 48 1.06 -0.07 0.42 -7 Vgoc s 1.00 volt 101 0.62 0.18 0.27 29 Noter
- 1. Projected from the 199192 growth rate based on 0.80 EFPY for the 1991-92 cycle versus 0.956 EFPY for the next cycle. Negative voltage changes were not factored up.
- 2. Tubos plugged during the 1991 outage. -
9-12
CATAWBA-1: 1992 Indications at TSP's 2000 1804
, 1600- . .$1400 1369 m
h1200-0 1000-o 5 800- 739
.a w 5 600- [i 553 Z
400- f 200- {j 162 2 ffM ' ' ' ' 1H 2H 3H 4H 5H 6H 7H 8H Support Plates 8 S/G A M S/G B E S/G C E S/G D Figure 9-1. Distribution of Indications with TSP Elevation / Location
~
CATAWBA-1: 1992 TSP INDICATIONS-ALL SG ALL BOBBIN COILINDICATIONS
= ;
= = = = = 100 1600 1431 -90 1400-1298i e
-80 .O T '
1240 r 3 - o
;4: voltag..
m 1200- 4: 8 y ly:::i Avg o.38 Std Dev 0.25
-70 5 u,,,
'c ,
- ?- :-
k > C f1000- 3. -60 O a c) -
'w ' D m , c' .
o
? 800
.O O N o. .-
. -50 5 4 .m g .g.
' ; 8'
&v, 602 613 -40 c)
C) -. m E 600-506 4: -
.2 -
5.-.
-30 33 h
- @ :N ;f ::;g Z 400
, .hz:
X
- e
- E -
282 -20 3
. f. -
}-:', f '( ;, . . , ' Y 3 0
A :B 4 :es' .- M 200- -c
-10 M' $ ; @f - w
, f. , 109 4:- E 9- 4:. 5 40 4; w, :& q 11 4 1 1 1 1 1
, , , , , 0 0
i i , , , , , , , . 0.1 0.3 0.5 0.8 125 1.75 2.2 2.7 3.55 0.2 0.4 0.6 1 1.5 2 2.6 2.8 Bobbin Voltage Figure 9-2. Frequency Distribution of TSP Indication Voltage Amplitudes in All S/Gs
~
CATAWBA-1:1992 TSP INDICATIONS-S/G A ALL BOBBIN COIL INDICATIONS 160 - - 100 145 ,.48 1, 7 35': 90 140- ;
/c
,:r : .:
g : : c p, -
,< a m 120- .
3
# T D fpj Voltagee g 'z {
m 104 ',; . f: Avg 0.32 70 Lt. O if#;j Std Dev 0.21 h 100-- E :g: I6 i $n, q y c o q ; .,,:; y': 3;; z w '-
- M; -60 D 97 .,i.
h 8& :$ h ,. h 72 O g
.. m 1: y o ;g .:f;, gj; s :- R -50 5 u a;; M');: >,, : ;;f; o o so- y : :: -'c :, ;
e 56 >
.O o,' : w$ .. t: :w: ~40 '"
E * - i: /@e:41 5 a ' f'; W: g: 29 40- p %g -30 E iW :hh) B p$u[ 4: a y%u y: a ig>:. r 26
- 0 2& [ yg; jfg p gj llg f jj -20 0
hhIhIkhh.h,$Mh.h.I$ 3 0 11 10 - 0.1 0.3 0.5 0.7 0.9 1.1 1.3 0.2 0.4 0.6 0.8 1 12 1.4 Bobbin Voits Figuie 9 3. Frequency Distributic" af TSP in6 cation Voltage Amplitudes in S/G-A 1 1
CATAWBA-1: 1C_e TSP INDICATIONS-S/G B ALL BOBBIN COlLINDICATIONS
- = = 100 200 180- I
-90
- ' e 160- !% 80 .9 m :i 145 F 3 U
$ 7 ""** -70 E b 140- [' J: Avg 0.37 LL, .;s :: '!
O F~ Std Dev 0.20 c 115 t ; -60 '9 h120-112 k'dj x, u) - a e z :,y a h100- h $ h -50 $ 3 m . a m 5 -40
.=
D [h h[j ;-{ fh g . j S E 60-N
- : :'! ,I !$~; :.
d Ii : -30 .9 2 z 42
!? : '1 ! ',: M*: i; 3
E
- q- L,j 'i i 39 3 40- M ;
- - -20 28 o 1 f % fl !' Jj gj !'*
2e $gg fj [] [j g ljj l 3, 2 y0 c i" M . L 1 .' . , Ol f , # F51, 3 1 1 g 0.1 0.3 0.5 0.7 0.9 1.1 1.3 0.2 0.4 0.6 0.8 1 1.2 1.4 Bobbin Volts Figure 9-4. Frequency Distribution of TSP Indication Voltage Amplitudes in S/G B
CATAWBA-1: 1992 TSP INDICATIONS--S/G C ALL BOBBIN COIL INDICATIONS
= = = =
100 400 358
- ' -90 350-324 326 i:dfji c E o
'dy !y I'$ r 3 -80 g u) 300- -
, :' - J; voltages c g yg$ 264 0.46 3
2 :; !L j ;, Avg -70 1L M
* :$ I24 !p,! 24 ~ ;; std Dev 0.30 C h M $ j f q J .9
$ n d'
$ ;# @is r ':j 215 [W! - 60 '. D
.o ;gy?,4
- r .o E
O 200- !':R M: si f?s %@ o ,Ni I':, $ [jff -50 $ f y ,f'f;g u 1 ',
$ l# i g):'i h,$hy!y gja i2 e ~4 m
- E i ! 'l 3 Z 100- #: ,,
g@u w!p,ip: 1 @i L, :s -30 E m a :: 3, ; s. a q; 4 e e m
?q iz; 3,: o 4/ o: E a m 33
.].??g):j :;j $
Se -20 e i!6i I:
?& 24 M:
w
,if,'d
$:g$
i if.$.i s,,i
,i I', :: :
y$, 33 5 2 1 1 1
~;, . e-m. 10 c . . . . . . . . . . . . .
0.2 0.4 0.6 1 1.4 1.8 2.59 3 55 0.3 0.5 0.8 1.2 1.6 2 2.69 Bobbin Volts Figure 9 5 Frequency Distribution of TSP Indication Voltage Amplitudes in S/G-C 1
CATAWBA-1: 1992 TSP INDICATIONS--S/G D - ALL BOBBIN COIL !NDICATIONS
= = = =
100 900 - - 801 -90 800- r 4: 772 7
'x 7/ c
%m %
-80 a0 700- ( y u) : :'g -
0 C ly& i F Voltages (. - - . -
-70 3 cO 600- M Avg 0.34 LL.
Y W( R['? Std Dev 0.23 c f$. [I[?W: ' ( ) -60 .g 8 500- %c; :6-
;w
.a g
, ,' 0f -b0 'C o 400-u jfi
?
N
% g
.I 368 Q
Q fs :;,s: 7
-40 Q ,Q 9 %;'4:
f: .: ' . E 30& 276 2: w 9 a 7 !,s @
.s 243 239 -30 .9 3
Z 'q'
- g s $nb;fd g# x f,, ":y, r
E 200-m: :g ;z:. yy ;: 20 3 -
- .y ;w v. g;ys.e., ::';/ g'7 ,. ..
Q s y 9,: a i&'is : 104 100- ?. g; w. y;3 y ?? K: M v -tc
..- /s ,2 . , . 19 5 33 h, :s? '$ fAl ,$ q 2 3 o , , , , . .
I 0.8 12 1.6 1 2 1 1 2.8 0 ! 0.1 0.3 0.5 l 0.2 0.4 0.6 1 1.4 1.8 2.19 Bobbin Vorts l 1 i l 1 1 1 - i i t Figure 9 6. Frequency Distribution of TSP Indication Voltage Amplitudes in S/G D
CORRELATION OF 550/130 TO 400/130 kHz . RESULTS FROM 1992 CATAWBA-1 INSPECTION 4.5 4-
- 3. 5-
-y 3-x 8
- 2. 5-6 g 2-ui 0 1.5-N d =
> 1-
- 0. 5-0- Y --- - - -- - - - - ----- ---- - - - -- -- - - -- - - - - --
-05 , , , , , , ,
O O.5 1 1.5 2 2.5 3 3.5 4 - VOLTAGE AT 400/130 kHz e DATA y - 1.038 x - 0.047 Figure 9 7. Correlation of Voltage Amplitudes Between 550/130 iMz anJ 400/130 kHz Using a Sample of 96 TSP indications in S/G-C from the .192 Outage
3 . O
- ~ ~ - - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ - - -
2.5- ~ ~ ~ - ~ ~ - - ~ ~ ~ - m 2- - - - - Y . O d ui 1.5- - ~~~~~~~~~~ ~~~~~~~~~ ~ ~~~~~~~~ d O O D D D
,_...... O....... g ...... .. .......................................................................
O u u ng O g z os
- pc .a.....g..........................a..........
O die U OO
-_7g @n D 0, 7
- o. g g 8. D.....D....D..................................
U
,r---
O k O O Ub O
~~~-O~~-~~~~-~--~~ -0. 5 - ~ ~ - ~ ~ ~ ~ -
OO O U O 1- -- i i , i i i i i O.2 0.4 0.6 0,8 1 1.2 1.4 1.6 1.8 2 1991 (BOC) AMPLITUDE, VOLTS Figure 9-8. TSP Indication Voltage Growth Rates During Cycle 6 verras BOC Vohages - for a Sample of 541 Largest Indications from 1992 Inspection
i CATAWBA-1: 1992-1991 TSP INDICATIONS VOLTAGE GROWTH RATES DURING CYCLE 6 2087 200G - 100 f;j 180C-f IN
-90 C
1600- j'j -80 $ '
- iti 8 1400- /;! 1 -70 3
.~ /,,
L.L
@ ,hl i C f V ff -60 33 6 120Cr '
m % b 1000- % % [h -50 h a & ,I .9
' 800-(y;765 -40 0 . $ $)
h : 3) h 3 5 600 583 y,;' g 30 g Z !j$ [h 45,8 y 400- l' 'A -
-20 -5
!, / [j- fj 280 239 O 200- 33 h ki h? -10 o
2
.f , ,
,h, ,,
h ,,,,,,,,, 12 3 ,4 3 0 1 1 0
-1.0 -0.4 -0.1 0.1 0.3 0.6 1.0 1.4 1.8 2.31
-0.6 4.2 0.0 0.2 0.4 0.8 1.2 1.6 1.84 Voltage Growth
. Figure 9 9. Fri.quency Distribution of TSP Indication Voltage Growths During Cycle 6 (All S/Gs)
l l l CATAWBA-1: 1992-1991 TSP INDICATIONS . VOLTAGE GROWTH RATE PROJECTIONS FOR CYCLE 7 2000 - : : = = = = 100 1796 180& r -90 E C 1600- @ -80 .@ g 0: g
-70 3 o 1400- '. LL
.:s C
191 h 1200- p 7 Q. e
$ 10185 l 5 O 1000- T :s -50 3 o $i h hl $
5N 698 $ k i 693 0
' # f!j I T E E 600- 5 '
% , j.
- 'O -30 .!9 3 z%: y- u 6 y z r c w e 440 E 400- d! !$ fh ! 3 339 325 -20 o 189 )r R
- 3i
' is :$! %m f
y 1 -10 0 200- -/g i# i f" E'd' 6 [' i'! if $34 c 63Mf$$i:}i,$,h,h,5,37 8 3 3 4 1 1
, , 0
-1.0 -0.4 -0.1 0.1 0.3 0.6 1.0 1.4 1.8 2.77
-0.6 -02 0.0 02 0.4 0.8 1.2 1.6 2.21 Voltage Growth I Figure 9-10. Frequency Distributbn of TSP Indication Voltage Growth Projections ,
for Cycle 7 Using EFPY Ratios and Cycle 6 Results (All S/Gs) l ---- -- - - - ---- - - -- _-_- - - - - _ - - - - _ _ _ - _ - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
CATAWBA UNIT 1 VOLTAGE GROWTH DURING CYCLE 5 (1990-91) 40 f 100 36- -------------------------3G-----------90 32- - - - - - - ---- -- --- - - - - - ......... ... ..........-.... ~80
@ pg_ ... .............................................., , ...................... . . . . . . . . . . . . . . . . .
-70 Q , 26 O g4- ......... .... ....... ... .....................
/ -
-60
{ 22
@5 2} - - -- ------- - - - -
-50 b.
u.
- a. ........................................ . ..
...1s.............................-40 g h 13 ,
H g a g_ ................................... ,. ...
.. .. ............................. 30 0- - - - --- -- - ----- '- -
-20 6 -
4_ ....... ..................... /k / . . . ... .. L ..
....a...........a.......... _1o C , 1 0
C ., , , , , , ., , , , , 0
-1 -0.6 0.2 0.2 0. 6 1
-0.8 -0.4 0 0.4 0.8 1.2 AMPLITUDE GROVTTH (1990-91), VOLTS
~
Figure 911. Frecuency Distribution of TSP Indication Voltage Growths During Cycle 5 (All S/Gs) 9-23
Sect:on 10 CATAWBA UNIT 1 IPC EVALUATION 10.1 Introduction This section provides the tube integrity evaluation performed for the Catawba-1 IPC to demonstrate margins against R.G.1.121 criteria. The Catawba 1 IPC are given in Section 10.2 including the tube repair basis, inspection requirements and operating leak rate limit. An , equivalent Catawba-1 voltage repair limit for fullimplementation of an APC is developed in Section 10.3. S ; limit is aplied to the IPC as the upper limit for leaving bobbin coil flaws in service even if not mnfirmed by RPC inspection. The Monte Carlo analysis methods used in thia section are general;y described in Section 10.4 and are used to project the EOC voltage distribution given in Section 10.5. This section also develops the maximum expected EOC 7 bobbin voltage. Sections 10.6 and 10.7 provide the tube burst margin assessment and the SLB lohkage evaluation. Development of tne operating leak rate limit is given n Section 10.8. and tha principal conclusions of the Catawba-1 IPC evaluation are summarized in Section 10.9. 10.2 Catawba 1 Interim Plugging Criteria (IPC) The Catawba 1 Interim Plugging Criteria (IPC) follow the precedent approved by the NRC for application to J. M. Farley and D. C. Cook 1 steam generators. The IPC include the tube repair basis, inspection requirements and operating leak rate !!; nit as described below: Tube Repair Basis o Bobbin coil indications having flaw voltages greater than 1.0 vo't and confirmed as flaws by RPC inspection shall be ropaired. o Bobbin coil indications having flaw voltages greater than 2.5 volts shall be repaired independent of RPC confirmation of a flaw. o Projected leakage for a postulated steam line break (SLB) event at end of cycle (EOC) conditions shall be less than 1.0 gpm for the most limiting S/G. Bobbin coil flaw , indications inspected by RPC and found to have no RPC indication do not need to be included in the leakage analyses, o Tubes identified as subject to significant coformation at a TSP elevation under a postulated LOCA + SSE event snail be excluded from application of the IPC at that TSP location. Inspection Requirements . o The inspection shall includo 100% bobbin coil inspection of all hot leg intersections ard cold leg intersections down to the lowest TSP for which the IPC is to be applied.
, o All bobbin coil flaw indications above 1.0 volt and below 2.5 volts shall be inspected by RPC to evaluate for detectable RPC indications and, for indications, to support ODSCC as the degradation mechanism.
10-1
o Eddy current analysis guidelines shall be consistent with guidelines utlitzed in prior NRC submittals supporting APC for ODSCC at TSPs. o An RPC samplinu program of at least 100 TSP intersectio: ., will be performed emphasizing intersections with greater than 5 volt (bobbin coil) dents and including some intersections with artifact bobbin indications or Indications wim cual phase - angles. Operating Leak Rate Limit o The norrnal operating leak rate requiring plant shutdown shall be limited to 0.1 gpm (150 gpd) per S/G. The remainder of Section 10 demonstrates that he IPC provide significant margins against the
- tube integrity criteria of R.G.1,121 The 1992 Catawba 1 inspection satisfied the above inspection requirernents including the RPC sampling plan as described in Section 9.1. The operating leak rate limit is di.veloped in Section 10.8.
10.3 Equivalent Catawba 1 APC Repair Limit The equivalent APC voltage repair limit for full implementation of alternate plugging criteria (APC) is utilized in the IPC to establish the maximum bobbin coil flaw voltage indication to be left in service even if not confirmed by RPC inspection. The tube repair criteria are developed - to preclude freespan tube burst if it is postulated that TSP displacement would occur under accident conditions. No analyses for the Model D3 S/Gs in Catawba 1 have been performed for ~ SLB conditions to determine if significant TSP displacement would occur to uncover the tube degradation occurring within the TSPs under normal operating conditions. If significant TSP displacement does not occur, the constraint provided by the TSPs would prevent tube burst and the principal repair criteria would be based on limiting leakage rather than tree span burst, The equivalent APC repair limit is developed to provide R.G.1.121 tube burst margins. The lower IPC repair limits incorporate additional margins against R.G.1,121. For the equivalent voltage repair limit, the voltage structural requirement for burst capability at three times normal operating pressure differential (3APNO) is reduced by allowances for NDE uncertainties in the voltage . measurements and by voltage growth between inspections as described below. EOC Voltace Limit for Structura! Recuirement The recommended correlation between burst pressure and bobbin voltage, as adjusted for temperature and minimum material properties, is developed in Section 7.3. At the lower 95% prediction interval, a bobbin voltage of 4.1 volts establishes the structural requirement for 3APNo (3750 psi) tube burst capability as shown on Figure 71. , Affowance for NDE Uncertainty The Catawba 1 NDE uncertainties for bobbin voltage sneasurements are developed in Section 5.8.2. NDE uncertainties at +90% cumulative probability are applied to develop the tube repair liniits. l'or the final Catawba 1 uncertainty evaluation given in Table 5-8, the NDE io- 2 v -. - - - - . - - m
l uncertainty of 90% cumulative probability is 22% of the voltage measurement. This uncertainty is applied at the tube repair limit voltage to de'/elop the equivalent APC repair limit.
. AHowance for Crack Growth Voltage growth rates for the Catawba-1 S/Gs are developed in Section 9.5 and summarized in . Table 9 7 The voltage growth rate average over the entire BOC voltage range is 27% for the largest 541 EOO indications. When averaged over only BOC indications greater than 0.75 volts, the growth rate is15% for the 541 largest EOC indications. These results show that the Catawba-1 growth rates, as a percentage of BOC voltage levels, tend to decrease with increasing BOC volts. This trend is consistent with other domestic p' ants as shown in Figure 6-12. The data from European plants indicate that percent growth may be approximately independent of amplitude, it is thus conservativa to assume that percentage growth is independent of amplitude and to use overall average growth from Catawba-1 operating experience for the growth rate allowance in the plugging limits. The Catawba-1 average growth of 27% is conservatively increased to 42% to established the equivalent APC repair limit.
Ecuivalent APC Recair Limit Table 10-1 summarizes the development of the equivalent APC repair limit based on reducing the structural voltage limit of 4.1 volts by allowances for growth and NDE uncertainties. The resulting equivalent APC repair limit is 2.5 volts. For IPC applications, the equivalent APC repair limit is used to define an upper bobbin flaw voltage limit for leaving unconfirmed RPC indications in service rather than as the tube repair limit. 10.4 Monte Carlo Methodology Monte Carlo sampling is the primary methodology applied foi projecting EOC voltage distributions. The Monte Carlo method permits combining any number of distributions to obtain a frequency or cumulative probability distribution of EOC voltages. A maximum EOC voltage can then be defined that properly reflects the number of indications left in sCvice by integrating the tail of the frequency distribution to one complete indication (or evaluating the cumulative probability distribution at (N-1)/N where N is the number of indications left in service). The Monte Carlo methods for EOC voltages applied in this report includes the following steps:
- 1. Establish frequency distribution of indications vs voltage (0.05 volt intervals applied) for BOC indications (s1.0 volt) left in service.
- 2. Define cumulative probability or normal distributions for voltage growth and NDE uncertainties (probe wear, analyst variability).
- 3. For each BOC voltage interval, randomly sample growth and (JDE uncertainty distributions and, for each sample, add to mid-interval voltage value to obtain one EOC
. voltage sample.
- 4. Weight each EOC . ample by the number of indications in the BOC voltage interval.
- 5. Repeat steps 3 and 4 for each BOC voltage interval to obtain one sample of the EOC voltage frequency distribution.
lo - 3 I i
- . . - - ~ - - . . - - . .- .. . _- -. _ . .-.. - - -
- 6. Repeat steps 3 to 5 above 100,000 times, order distributions and divide by 100,000 to
- obtain overall frequency.
- 7. Integrate frequency distribution of step 6 over voltage intervals to obtain a histogram of EOC indications.
P. Determine maximum EOC voltage by lntegrating tail of frequency distribution of step 7 to one complete tube, if SLB leak rate is desired from the Monte Carlo process, the first 3 steps abova are the same and are followed by: 4a. At each EOC voltage sample, randomly sample SLB leak rate distribution (mean 1. uncertainties) to obtain a sample leak rate. Sa. Weight each SLB leak rate sa. r.,e by number of indications in the BOC voltage interval. 6a. Repeat steps 3,4a and Sa for each BOC voltage interval and sum weighted leakage for each interval over all intervals to obtain one sample of total EOC leakage per S/G. 7a. Repeat steps 3 to 6a above 100,000 times and devdp leak rate per S/G into a cumulative probability distribution. 8a. Evaluate result of 7a at 90% cumulative probability to obtain reference SLB teak rate - per S/G. The tube burst probability is obtained sc:n,arly to the SLB leakage with the burst pressure vs voltage distribution sampled at step 4a and no summation is made at step 6a. It can be noted that the maximum EOC voltage, as defined by step 8 above, increases as the number of indications left in service increases when all distributions are fixed. This results as the maximum voltage moves to a higher cumulative probability as the number of indications increase. It is important in app'ying the Monte Carlo process that the cumulative probability distribution for voltage growth be developed from a sample reprecentative (in approximate number and voltage range) of the distribution of indications left in service. For example,if the growth distribution is developed only over a small sample of the largest indications and applied to a large population including a range of indications, excessive conservatism can result. This results as the growth distribution for only the largest ECC indications would overestimate the probability of large growth values. This is shown by sensitivity analyses in Section 10.5. 10.5 Projected EOC Voltage Distribution The most limiting S/G for tube integrity (tube burst, SLB leakage) considerations is between S/Gs C and D. S/G D has beer' found to have an RPC confirmation rate for bobbin indications of about 10% based on RPC testing of about 1350 indications. This indicates that the bobbin calls , rt.sult from very short and shallow degradation or are false bobbin calls. Both bobbin and RPC probes are expected to approach 100% detectability (see Plant L WCAP-13129) for significant t 10- 4
flaw lengths (greater than about 0.2") having average depths greater than 40-50% Thus indications not detected by both bobbin and RPC probes would not be expected to challenge tube integrity over one operating cycle. This EC detection consideration forms the basis for
+
application of RPC inspection for resolving distorted bobbin signals in applying the 40% depth repair limits. S/G C has had an average confirmation rate of about 30%. Since S/G C has about 2000 bobbin Indications compared to about 3500 in S,3 D, S/G C would be expected to have about 70% more RPC confirmed or significant indications than S/G D. In addition, the average bobbin voltage growth rate during Cycle 6 in S/G D is negative while positive in S/G C. This negative average growth in S/G D may be another clue that the S/G D indications are small or marginal. Based on these considerations, S/G C is the most limiting S/G and is used for reference , tube integrity assessments and sensitivity studies. An assessment ;s also provided for S/G D based on all bobbin indications less than or equal to one volt for a general comparison with S/G C. The results of this analysis confirm that S/G C is more limiting. In S/G C, about 45% of the bobbin indications were RPC inspected. RPC confirmed indications having bobbin voltages >1.0 volt have been repaired. Of 1871 bobbin indications less than or equal to 1.0 volt in S/G C,793 were RPC tested and 229 (29%) were confirmed as flaws. The remaining 564 bobbin ;dications found to be NDD can be ignored for the reference tube integrity assessment, as noted above, the indications not found by RPC are either too small or are false bobbin calls and would not be a challenge to tube integrity. Thus the reference S/G C assessment is based on BOC Indications left ;n service as the sum of indications not RPC tested ' plus those coi,iirmed by RPC with bobbin voltages less than or equal to 1.0. The total number of indications left in service for this reference case is 1307. For consistent Monte Carlo analyses, it is necessary that the growth rates be developed from a population comparable to the BOC indications left in service. Therefore, the growth rates for the reference analyses were also based on 1992 S/G C RPC confirmed indications plus indications not RPC tested in 1992. The growth rates include RPC confirmation of bobbin indications >1.0 volt which were repaired and not include 'in the BOC 7 indications left in service. Growth rates for 1991 to 1992 were increased by the ratio of 0.96 EFPY potential Cycle 7 operation to the actual 0.80 EFPY operation for Cycle 6. All growth rates used for EOC 7 projections include this adjustment for the potentia:!v longer Cycle 7 operation. As a sensitivity analysis for comoarison with the above reference analysis, a calculation was carried out based on leaving in service at BOC 7 only S/G C RPC confirmed indications having bobbin voltages less than or equal to one volt (229 indications). Growth rates were derived from the S/G C population of tubes having 1992 RPC confirmed indications. This assessment of the most limiting BOC indications might be expected to yield the maximum EOC voltage. For additional sensitivity analyses, the more common pravice previously used for APC applications (7/8" tubing, WCAP-13464 and WCAP-12871) was also performed for S/Gs C and D. These analyses assume all bobbin Indications in a S/G less than the repair limit of >1.0 volt are left in service at BOC 7. The number of BOC 7 indications are 1871 for S/G C and 3420 for S/G D. The growth rate applied is the Crowth distribution obtained for the total population of TSP indications in all S/Gs during Cycle 6, adjusted to the potentially longer Cycle 7. For - comparisons of burst pressure margin ratios between Catawba-1 and a typical plant with 7/8
. inch diameter tubing (WCAP-13464), the more limiting S/G C results are utilized. S/G C has the higher RPC confirmation rate which is more typical of the prior 7/8 inch experience given in WCAP-13464.
~
Based on the above, four analyses are performed: 1) the S/G C reference model with RPC confirmed (<1.0 volt) and not tested by RPC at BOC 7: 2) the S/G C model with only RPC 10- 5
, , , _ _ , m., - - ,_- -.- - - , -
confirmed (<1.0 volt) indications at BOC 7; 3) the more common S/G C model including all bobbin indications below the repair limit at BOC 7; and 4) S/G D including all bobbin indications below the repair almit. For each of these models, Monte Carlo analyses (as described in Section 10.4) were performed to project EOC 7 bobbin voltages distributions. The BOC 7 distributions, . growth ratos and EOO 7 distributions are shown in Figures 101 to 10-4. The EOC 7 results for models 1 to 4 are, respectively,2.34,2.53,2.24 and 2.33 maximun,200 vdtages and 2,3, 2, and 3 for the the number of indications above the SLB leakago threshold of 2.0 volts. - For deterministic tube burst margin assessments, the estimated EOC 7 voltages based on a 1.0 volt indication left in service and increased by allowances for growth and NDE uncertainties are also required. The dolorministic assessments utilize allowances at +90% cumulative probability for comparisons with 3APNO tube burst capability and at +09% for comparisons with SLB burst requirements. EOC 7 voltages for the +90% and +D9% allowances are given in Table 10-2 for the reference case and for the S/G C, all bobbin indication case. The S/G C result
-for all bobbin indications left in service is used in Section 10.6 for comparisons with typical results for a plant with 7/8" tubing. The Monte _ Carlo analysis results for maximum EOC voltages are includod in Table 10-2 for comparison with the deterministic assessments, it is soon that there is little difference (s0.1 volt) between the two deterministic analyses for projecting EOC 7 voltages. Also the sum of +99% cumulative probabilities is within 0.15 voit of the Monte Carlo results. It can be noted that the sum of +99% probability levels has a net confidence on the order of 99.8% for the present application. The reference caso Monte Carlo maximum voltage corresponds to 99.92% for the Monte Carlo EOC 7 voltago cumulative probability distribution.
The above reference and sensitivity results indicate that the maximum EOC 7 voltago for indications <1.0 voit left in service would be in the range of 2.54 (reference case) to 2.53 (RPC confirmed case) volts and can be bounded by 2.53 volts, For tube burst margin assessments in . Section 10.6,2.53 volts is applied as the maximum EOC voltage. Table 10-2 also includes the preliminary analysis results reported at the Catawba-1 NRC meeting of August 28,1992 (WCAP 13496). This calculation utilized all S/G D BOC 7 indications less than or equal to 1.0 volt as left in service (same as Case 4 above). The growth rate distribution applied was that obtained for the largest 541 Indications identified in the 1992 inspection. The use of the small sample of SA1 large growths overestimates the probability of large growth when applied to the total population of 3420 indications left in services in S/G D. Foi example, the largest 20 growth indications having increases >1.0 volts in Figure 10 3 are the same 20 Indications in the largest 541 population. The probability of the largest 20 in the 541 Indication subset of growths is about 3.7% but would only be 0.3% of the total population. Thus the use of the small population of Isrge growths for the larger population leads to a (0.037-0.003)
- 3420 - 116 indication overestimate of growths greator thaa about 1.0 volt.
The result is to over estimate the large voltage indications at EOC 7. This result is seen by the 3.28 maximum voltage and 33 indications above 1.8 volts for the proliminary assessment compared to the Figure 10 4 result of a maximum EOC 7 voltage of 2.33 volts and 5 indications above 1.8 volts. Thus the preliminary EOC 7 projections were excessively conservative compared to the more consistent analyses of this report. The preliminary analyses wore intended to be conservative to establish the feasibility of a 1.0 volf IPC for Catawba-1 although - the degree of conservatism could not be assessed prior to completing the final 1992 inspection voltages and associated growth ratos. 10- 6
~ - , ,,
10.6 Tube Burst Margin Assessment Application of the equivalent APC repair limit developed in Section 10.3 would result in meeting
- R 3.1.121 criteria at EOC conditions. The objective of the IPC repair limit is to establish additional margins befond that included in the equivalent APC repair limit.
The limiting R.G.1.121 criterion for Catawba 1 is to satisfy the 3aPNO tube burst margin requirement. Thus the additiotialIPC margias can be expressed as burst pressure margin ratios relative to 3aPNO. That is, the ratios of BOC and EOC burst pressses to 3APNO. The burst margin ratios are developed adding +90% cumulative probability on growth and NDE uncertainties to the 1.0 volt repair limit and evaluating the resulting voltages at the lower 95% orediction interval of the burst / voltage correlation (Figures 7-1 to 7-2 for tube burst capability.) These uncertainty levels provide that only a few, if any, indicaticns would exceed 3aPNO at EOC conditions. It is necessary to establish higher confidence levels to prevent tube burst at APSLB. The burst margin ratios relative to APSLB are therefore developed applying
+99% cumulative probability on growth and NDE uncertainties to the 1.0 volt repair limit and the lower 99% pred;ction interval of the burst / voltage correlation (Figure 7-3) for tube burst capabi!!!y. In aculon, the burst margin ratios relative to APSLB are provided for the maximum projected EOC voltage of 2.53 volts (Section 10.5).
Table 10-3 summarizes the tube burst margin assessmet,t relative to 3AP NO for Catawbc-1. At
+90% cumulative probability on allowances for growth and NDE uncertainties, the projected EOC volts are 1.56 for the limiting S/G C based on all bobbin Indications s1.0 volts left in service (Figure 10-3 data). For the reference case (Figure 10-1 data) based on RPC confirmed plus not tested ir.dications <1.0 volts left in service, the projected EOC volts is 1.66. These two cases yield EOC burst pressure capability of 4810 and 4740 psi. The EOC burst pressure margin ratios relative to 3aPNO are 1.28 and 1.2G or substantial margins against this R.G.
1.121 criterion. At BOC, the burst margin ratio is 1.41. A typical case for the 1.0 volt IPC apolied to 7/8 inch diameter tubing is also shown in Table - 10-3. It is seen that the Catawba 1 burst margin rat:os are essentially the same as the 7/8 inch tubing example. For a 1.0 volt IPC, th: 900 burst margin ratios are comparable for the two tubing sizes which shows that the 1.0 volt 4 "' establishes essentially equivalent margins between 3/4 and 7/8 inch tubing. While the 1.4 margin ratios are the same for th!s particular case, it should be recognized that variations about 1.4 will result for plants with different steam pressures and the 1.4 ratio may be somuwhat modified upon opdating the burst / voltage correlation for 3/4 inch tubing. The EOC burst margins should not be compared for assessing equivalency of the 1.0 volt IPC limit between tubing sizes as modest changes in growth distributions can significantly a!!er the tube size comparison. It is necessaif for IPC margin demonstratiot, that the EOC margin ratio exceed unity to show that R.G.1.121 is satisfied with some margin. For fullimplemer;tation of the equivalent APC repair !imits, the EOC burst margin > ratio relative to 3aPNO would be expected to be near unity. The tube burst margin ratio assessment relative to AP SLB is given in Table 10-4. The Catawba-1 results are the growtn rate deve!oped for all bobbin indications (Figure 10-3) and for the maximum projected EOC 7 voltage of 2.53 volts. The EOC margin ratios were found to be 1.35 to 1.41 for Catawba-1. This demonstrates substantial margin against burst at SLB for EOC 7 conditions. As a supplemental demonstration cf burst margins, the probability of tube burst at , 10- 7 l
l APSLB was calculated for S/G C (Figure 10-1 data) using, Monte Carlo methods and found to ba sbout 1.1 X 10 5. Thus large margins against burst at SLB conditions a' 3rovided by the Catawba 1 IPC. It can be noted from Table 10-4 that both 3/4 and 7/8 incu tubing provide , large margins against APSLB although the 7/8 inch rnargin ratios are somewhat higher than that for 3/4 inch tubing. The above assessment shows that the Catawba 1 IPC repair limit of >1.0 bobbin coil voltage provides significantly large margins against R.G.1.121 structural criteria for tube burst. These assessments utilized the recc nmended voltage / burst correlation of Figure 7-1. Section 7, Figures 7-4 to 7 9 provide sensitivity of the correlation to various conservative s assumptions on finakation of the 3/4 inch burst data. For all o' these conservative correlations, the EOC voltages of Tables 10-3 and 10 4 provide burst pressure margin ratios above unity against 3APNO and APSLB, respectively. The correlation of Figure 7 7 based on including the unreliable Catawba-1 burst data provides the lowest burst pressure capability for all sensitivity correlations examined including use of the Belgian data without any voltage adjustment for cross calibration of ASME standards. With this correlation, the burst pressure margin ratios are 1.13 relaHve to 3APNO and 1.15 relative to APSLB using the methodology of Tables 10-3 nd 10 4. Thus the R.G.1.i21 criteria would be satisfieo by the Catawba 1 IPC repair limits even under the most limiting assumption on effects of current burst eta uncertainties on the voltage / burst correlation. 10.7 SLB Leak Rate Assessment , The IPC require that potentialleakage under SLB conditions at EOC 7 be less thsn 1 gpm. This section provides the results of the leak rate analyses to demonstrate that this requirement is . satisfied for the Catawba-1 IPC repair limit of >1.0 bobbin coil volt. The methods of analysis are consistent with the guidelines provided in the SER for the D. C. Cook-1 IPC. These guidelines recommend SLB leak rate analyses based on Monte Carlo analyses using a SLB leak rate vs bobbin voltage correlation and a separate analysis using EOC voltage cWributions that consider tails of the growth distribution together with a more deterministic leak rate calculation based on a bounding, stepwise change in leak rate with voltage. The results of these two analyses are provided in this section. In addition, the results of sensitivity analyses for SLB leak rates are provided. The results of applying the two SLB leak rate analysis methods are given in Table 10-5. This table also provides the source of the input data (Tables, Figures) epplied to cabulate the leakage. The BOC voltage distribution for the reference SLB leak rate analysis is thst of Figure 10-1 and includes all RPC confirmed indications and indications not inspected by RPC that have bobbin voltages <1.0 volts. The EOC voltage distribution is shown also in Figure 10-1. For the deterrninistic assessment, the SLB teak rate relation applied (Table 7-4) leads to no leakage for EOC indications below the leakage threshold of 2.0 volts and a leax rate of 1.0 liter /hr per indication between 2.0 and 3.5 volts. No EOC iridications are expected to approach the 3.5 volt limit for a 1.0 liter /hr leak rate and only about 2 indications are projected above the leakage threshold of 2.0 volts. Thus the oeterministic method leads to an ECC 7 SLB leak estimated at 2 liter /hr or <0.01 gpm. The Monte Carlo analysis utilizes the same input data as the deterministic analysis except that - the continuous EOC voltage distribution is applied to the SLB leak rate correlation of Figure 10- 8
74 3 (including uncertainty distribution). The 2.0 volt threshold for SLB leakage is applied in the Monte Carlo analysis, if the EOC voltage sample being evaluated is less than 2.0 volts, it is assigned zero leakage and the Figure 716 distribution is sampled for EOC volts >2.0. At 90% cumulative probability of the Monte Carlo SLB leak rate distribution, the EOC 7 leakage is projected to be approximately zero. Thus both the deterministic and Monte Carlo methods project negligible SLB leakage at EOC 7. To assess the sensitivity of SLB leakage to the input data, a number of alternate cases were run with both the Monte Carlo and deterministic models applied for the leak rate calculation. The - four BOC voltage distributions of Figures 10-1 to 10-4 were assessed as the ieference case and Cases 3,5 and 6. All of these cases show essentially no SLB leakage (s0.01 gpm) for both the-deterministic and Monte Carlo analysis methods. These cases apply the 2.0 volt SLB leakage threshold. Supplemental cases applying a 1.8 volt leakage threshold also yielded zero gpm leakage by Monte Carlo and <0.022 gpm by the deterministic model. Case 2 which applies the SLB leakage correlation including the Belgian data without a voltage adjustment yields the same Monte Carlo results of zero leakage as obtained for the reference as Cases 3,5 and 6. Thus the SLB leakage is insignificantly dependent on the input data a sng as a leakage threshold is included in the analyses. Leakage thresholds <1.8 voK were not evaluated. As bounding sensitivity cases, the unrealistic assumption of no leakage threshold was applied for cases 1 and 4. Cases 1 and 4 use the S/G C BOC voltage distributions of Figures 10-1 and 10-2. Even under the no threshold assumption, the S/G C leakage is bounded by 0.67 gpm (Case 1) for the recommended SLB leak rate correlation of Figure 716. Given that a leakage threshold of >2.0 volts (Section 7.5) for a measurable or quantifiable leak rate is supported by all model boiler and pulled tube data, it is concluded that the SLB leak rate for Catawba-1 at EOC 7 is expected to be nearly zero gpm and is conservatively within the - allowable limit of 1.0 gpm. 10.8 Operating Leakage Limit R.G.1.121 acceptance criteria for estab%hing operating leakage limits are based on leak before break (LBB) consideration such that plant shutdown is initiated if the leakage associated with the longest permissible crack is exceeded. The longest permissible crack is the length that provides a factor of safety of 3 against bursting at normal operating pressure differential. As noted above, a voltage amplitude of [ 19 voits for typical ODSCC cracks corresponds to meeting this tube burst requirement at the lower 95% confidence level on the burst correlation. Alternate crack morphologies could correspond to [ ]9 volts so that a unique crack length is not defined by the ! burst pressure to voltage correlation. Consec,uently, typical burst pressure versus through wall crack length correlations are used below to define the " longest permissible crack" for evaluating operating leakage limits. The CRACKFLO leakage mudet has been developed for single axial cracks and compared with leak rate test results from pulled tube and laboratory specimens. Fatigue crack and SCC leakage data j- have been used to compare predicted and measured leak rates. Generally good agreement is obtained between calculation and measurement with the spread of the data being somewhat
- greater for SCC cracks than for fatigue cracks. Figure 10-5 shows normal operation leak rates ,
including uncertainties as a function of crack length. l The through wall crack lengths resulting in tube burst at 3 times normal operating pressure 10- 9 l
t differentials (3750 psi) and SLB conditions (2650 psi) are about [ la, respectively, as shown in Figure 10-6. Nominal leakage at rormal operating conditions for - these crack lengths would range from about [ la would cause undue restrictions on plant operation and result in unnecessary plant outages, > radiation exposure and cost of repair. In addition,it is not feasible to satisfy LBB for all tubes , by reducing the leak rate limit. Crevice deposits, presence of small ligaments and irregular - fracture faces can, in some cases, reduce leak rates such that LBB cannot be satisfied for all tubes by lowering leak rate limits. An operating leak rate of 150 gpd (-0.1 gpm) will be implemented in conjunction with application of the tube plugging criteria. As shown in Figure 10 5, this leakage limit provides for detection of [
]a. Thus, the 150 gpd limit provides for plant shutdown prior to reaching critical crack lengths for SLB conditions at leak rates less than a 95% confidence level and for 3 times normal operating pressure differentials at less than nominalloak rates.
The tube plugging limits coupleo v,lth s 00% inspection at affected TSP locations provide the principal protection against tube rupture. The 150 gpd leakage limit provides further protection against tube rupture. In addition, the 150 gpd limit provides the capability for detecting a rogue crack that might grow at much greater than expected rates and thus provides additional protection against exceeding SLB leakage limits, 10.9 Conclusions - Based on the above evaluation of the Catawba-1 IPC repair limit of >1.0 bobbin volt, it is concluded that: o R.G.1.121 criteria for tube integrit) _ conservatively satisfied at EOC 7 for an IPC repair limit of 1.0 bobbin volt. o At EOC 7, burst pressure capability (expressed as margin ratios relative to 3aPNO and APSLB) is expected to have ratios of about 1.25 relative to 3APNO at 90% cumulative - probability levels and about 1.35 relative to APSLB at 99% cumulative probability levels. A burst pressure margin ratio of 1.4 relative to 3aP NO for Catawba 1 at BOC conditions is comparable to typical values for plants with 7/8 inch diameter tubing with an IPC repair limit of 1.0 vc't. Thus the two tubing sizes can be considered to have equivalent margins for IPC repair limit of 1.0 volt. o Potential SLB leakage at EOC 7 is expected to be negligible (~0.01 gpm) as supported by both Monte Carlo and deterministic evaluations including sensitivity analyses, o The maximum EOC 7 bobbin voltage indication resulting from indications left in service below the repair limit is projected to be abnut 2.53 volts. - o The operating leak rate limit of 150 gpd implemented with the IPC satisfies R.G.1.121 guidelines for leak before break. This limit provides for plant shutdown prior to , reaching critical crack lengths br SLB conditions at a -95% confidence level on leak rates and for 3AP conditions at less than nominalleak rates. 10- 1 0
rates and for 3aP corditions at less than nominalleak rates. 9
==
e 4 10- 11
Tabte 10-1 Equivalent APC Repair Limits to Satisfy Structural Requirements - Item Vo!!s Basis FAaximum Voltage Limit to 4.10 Burst Pressure vs. Voltage Satisfy Tube Burst Correlation at -95% Structural Requirement confidence level (Fig. 7-1) Allowance for NDE -0.55(22%)(1) From Table 5-8,22% Uncertainty uncertainty at 90% cumulative probability. Allowance for Crack -1.05(42%)(1) Table 9-7 shows average growth / Growth Between cyc!e of 27%. Allowe.1ce Inspections conservatively increased to 42% of Tube Plugging Limit. 4 Equivalent APC Repair 2.50 Voltage Limit o Acceptable Limit to Meet Structural Requirement d2!d:
- 1. Voltage percentage allowances for NDE and growth rate / cycle applied to Equivalent APC Repair Voltage Limit of 2.5 volts.
4 10- 12
Table 10-2 - Maximum EOC Voltago Sensitivity Assessment EO% Cumulative Prob. 99% Cumulative Prob. Monte Carlo BOC Volts 1.00 1.00 1.00 NDE Uncertainty 0.22 0.42 --- BOC + NDE Unc. Volts 1.22 1.42 ---
~
Reference One S/G C:RPC Confirmed + RPC Untested o Growth 0.44 0.78 - o EOC Maximum Volts 1.66 2.20 2.34(1) Case for Comoarison of Burst Marcin Ratics with Tvoica! 7/8 inch Tubino REsutts S/B C:RPC Confirmed Only o Growth 0.34 0.79 --
, o EOC Maximum Volts 1.56 2.21 2.24 Preliminarv Assessment S/G D:All BOC Bobbin, Largest 541 Ind. Growth, Preliminary NDE Unc.
o Growth 0.62 1.4 - o EOC Maximum Volts 1.78 2.81 3.28 N21e:
- 1. Maximum EOC volts based on integrating EOC distribution to one indication (typically >0.999 cumlative probability) 10- 13
i l 4 l Table 10-3 Catawba 1 Tube Burst Margin Assessment for 3APyo Catawba-1:3/4" Tubina 7/8" Tubina Examo!e(1) i BOC Voits 1.0 1.0 Allowances @ : 3% Cum. Prob. o Voltage Growth 0.34 0.60 o NDE Uncertainty 0.22 0.16 EOC Volts (+90%) 1.56 (1.66)(2) 1.76 Tube Burst Capability (psi) o 3aPNO Requirement 3750 4380 . o Capability at -95% Pred. Int.: At BOC - 1.0 volt 5290 6200 At Projected EOC Volts 4810 (4740)(2) 5660 Burst Capability Ratios to 3APNO o At BOC - 1.0 volt 1.41 1.41 o At Projected EOC volts 1.28 (1.26)(2) 1.29 l l N_gte:
- 1. Example for 7/8" Tubing for Plant A-2, WCAP-13464. .
- 2. Reference analysis results.
10 14
Table 10 4 Catawba 1 Tube Burst Margin Assesement for SLB Conditions O Catawba 1:3/4" Tubina 7/8* Tob!no Examoto(1) 800 Volts i.0 1.0 Allowances @ +99% Cum. Prob. o Voltago Growth 0.79 2.0 o NDE Uncorialnty 0.42 0.25 EOC Volts (+99%) 2.19 3.25. Maximum ProjecteG at EOC Volts o Monte Carlo 2.53 --
, Tubo Burst Capability (psi) o APSLB Roquiremont 2650 2650 o Capability at 99% i' rod. int :
At +99% EOC volts 3730 4420 At Maximum EOC Volts 3580 -- Burst Capability Ratios to APSLB o At +99% EOC volts 1.41 1.67 o At Maximum EOC volts 1.35 -
, Note 1. Example for 7/8" Tubing for F ant A 2, WCAP-13464 10 - e 5
,c,p.w,- -
- - - , - , - , , , , - - - , . . . , _ , . . _ , . - , . - , , , , - . - , .-,_,,,-m -
Tablo 10 5 SLD Leak Rato Assessment for EOC7
' Deterministic
- Attettment Monte Carlo Astettment inputs to Leak Rato Analysis o DOC Voltage Distribution Figuro 101 Figure 101
. RPC Confirmed or not
- Inspected with BOC bobbin volts si.0 volt ,
f o NDE Uncertainties Tablo 5 8 Table 5 8 l Monto Carlo Distributions Final Values Final Values ; t o Voltage Growth Figure 10-1 Figure 101 Distribution for all RPC confirmod or not RPC inspocted o ECC Voltago Distribution Figure 101 Continuous Distribution Consistent with i Figuro ; o SLB Leak Rato Correlation !
- Throshold Voltago for Leakago 2.0 2.0
( - Correlat'on Tablo 7 4 Figure 716 -' Discreto Step Continuous from l Changes Rogression- , Analys!s Projected EOC 7 SLB Leak Rato <0,01 ppm 0.0 ppm l l l l l i i ( t I
- 1. 10- 16
- . . - -~.c . -~.....:. . - - . -. --.-. .-.. ..,- - -...-- .. - - -.- ---.- - - . .. .. ~..-.- ~ - - - .
i Tabl010 6
*- SLD Leak Flate SensitMiy Asscss'nont 2 tGr,L .S.G D IkMwa Cus_1 .Ceg2 Pre 3 Cmed Cad Cmd inputs to Leak flate Analysis o DOC VoNage Distrib Jtion Fg.101 Ref. Ref. Fig.5 t c U. Fg.173 F6g.10 3 RPC+ Uninsp. RPO L ny All<1.0v All <1.0v i o NDE Uncertainties Tabh68 Ref. Ref. Ref. Rof, Ref. Ref.
5 o Vohnge Growth Fig.101 Ref. Ref. Fig.10 2 Caw 3 Fig.10 3 Fig.10-4 ; SG C RPC4 Unt. SG C RPC AllSG Alls.G 1 o EOG Vohages Fg.101 Ref. Ruf. Fg I'>2 Case 3 Fig.10 3 Fig.104 , Continuous Dis for LAC. I o GLD Leakage Correlation
- Threshold Vofts 2.0 0.0 2.0 2.0 0.0 2.0 20 ,
Monte Carlo Corr. Fig,716 Ref. Fig 718 Ref. R ef. _ Ref. Ref. Monte Carlo EOC-7 o Maximum EOC Vohago 2.34 2,34 2.34 Tc.53 2.53- 2.24 -2.33 o SLD Leak Ratospm 0,00 0.67 0.00 0.00 0.41 0.00 ' O.00
' Deterministic
- EOC 7 -
o No.ind. > Leak Threshold 2 N.A. 2 3 N.A. 2 3 o SLD Leah Rate gpm <0.01 N A. <0 01 0.01 N . A. . <0.01 0.01 i l~ l l e 8 10 17 w = w.-r vswt- e v- m.y -,nr,--gmn, w,,- g--,97+-,sr e -
,_y., ogq.y.,w- -q 9 -y., r emcwe,-poi,-94 .- g weder 'fp*vg'=v*"v q - = - - =v'wwe shT ww ehT eg y y ' y= 9 -
WTg'~-- 'W19pSy.gp7
-.- . _ - . . - . - . . . . . . . - . . - - .-. -. - - . .-- - . - .-. - - . . - - . ~ . . - , _ . - i, CATAWBA.1: tan TSP (NDICAT10NS-SG C RPC OONFIRMED PLUS NOT RPC TE STED IND. yo - mn *' 's:
;% T M
. &, . &, c: c w ,
'<: . 'e g;:
x ., n
~
, / +,- a .- ,,
f, # # % Lu ' '
- tw i fi y :sj:
I .m 3: Z i t '* /. . $, ,
, w , E,, . / ,,. -
y , .s <
- i. y, gy ra, ,ys :/, i .. y, ,
'O' 's M 4:
' b ,, -- '
9, y< ,;;: ,,-$:, 4: '? ' ;p); P1
. I:
- .e p 4 2 % n , , ,
L 02 04 08 1 14 in 2.59 3 5$ 03 01 0 4*. 12 te 2 2 80 httei V(es , CATAWpA t 1r 91 MTAGE GROWTW SG C RPC COr4FIRMID PLUS NOT RPC TE STED IND. - JuD y .- _ _ _ 100
+
' 90 2% 241 g n , no
?19 e -;:
- a.w.e.
. . . . v.o.i.. ee gn ,
g i. oc . ,0 - 1 E l,
- q o 1%4 so l d 110 "1 m
v.a o aa so T so 3, i. i *.*~*u
. ;[ i
- 13) '
ao . rs 30 3 gg 0: m g y, M ;, .:. ?
- 5
- f j: ,o 20 y La
? ) -
/
?
',6 5.
'?:
//
71 i34 e i , , , , pm 10 C' 45' ~41' ' o 1' ' o.3' 'os' 'o r' 'o9' '12' ' 16' '2.77 42 00 02 04 06 08 to 14 16 votage Geomei CATAWBA 1:PACUECTED EOC 7 BOBBIN ETAGE OtSTRIBtJDON SG C RPC CONFIRMED PLUS NOT RPC TESTED IND.
?oc g too .c 179 19 ~
'p 160' f ,, Sgt l 137 : "
(g # ~ '
~?, , l, W .k ;, ' :
o 100 y ij.i g, ,
'~'
, j,. A6
" j a -
4: 4: 4 - he 4: ,, 6: / fi, sq
'O' , 4 !I
% so. .V ,, V m f l[ -
7': /: '. , ,\ 4: 41 8 ';['. * d C [ T h 2,1 3,: m ps a is :< ry -
. r. -
4: ,;. ';: 1 / 0 p 4 + 3 - 4 ;p ',- :y: :p :p [/,j r3 [ ,4, q g ,
' O.1 ~ ' O.3 ' '05' '07' '09 ~ ' t .*' '14' '18' '22 02 04 06 De 10 12 to 20 2 34 Pngded Ette1Votage Figure 13-1. Steam Generator C EOC 7 Voltage Distribution Based on RPC Confirmed and Not Tested Indications
CATAWBA 1' 17/2 TSP INDICAT10NS-LG C RPC COrdiRMED INDICATK/dS , i
~ g 9
W ' M '
, 2,: -
, - ~s ,
,. to ...
y3'.
.m y,; ~~.a
** ,. 6:
- 2 ?. 3:
3w . , , :y: . n .. n >< v
~
X; ' 11 t* f i t, s V: 'ln 1/.
,w :l % % $,, .
.. s. ', %.
p +,.. '6 : s s p*] :n > 2: 4: ;< p A .n , , , ,
- c. . .
U2 . 04 - . 06 1 to 18 . 2 5v. 3 55 03 05 08 12 16 2 2M tbsto Vcan f cATAgcNMb#ggggyps c i z no 49 10 4& . r
. - ~ . .
~ UO h q
g* 3 }] .h j :::l :P ' 70 .g l" e ,, x 2 d
- . I:~
E
, I s0 5
tM -
.v
/ -
l' 'A 4; m m f. , 40 m y (
., ,. it I 19 'S { ' . 1 I .x
,, 'y 2 -
4 ' s :; to 2 6 2 ,' T
'V '2U u s s l.
- ') f: 5 h4,,iii 10 ,
(- jg -
-g,- '03~ ' O S' '07' '09' '12' ' 16' '2 77 42 00 02 04 06 08 to 14 18 V<duve Groven CATAW1W1: PROJECTED EOC 7 DOEON VOLTAGE DISTRIBUTON 50 C RPC CONFIRMED INDICATONS y
** U U i p
& N '
's,. N.
19 is
" , ,; ' . ?-
33 15 - 1& :h . 9 2 [ l ,' 4 % f j p ,. q 7' 4 3 ,
'7 y ,; , ';_ y y p i , / ' ,' -
ppfJf 03 05 0F 09 11 1J 15 17 2 24 04 06 06 10 12 14 16 1.8 il 2 53 , Pete.ded Iktte1Votage Figure 10-2. Stcam Generator C EOC 7 Voltage Distribution Based on RPC , Confirmed Indications
. . . . . .. -_ . . ~ . ._ - .-- - . - ,_ . - - . .- . . - .- .. - -- - .- -_ - . - - . . . . _
i CATAWBA 1' 1932 TSP INDICATONS.*A C ALL DOBEUN COIL INDICATONS WK: 3r4
- r4 r,s Ry
, ,w -
f: ; ,
. :y :p % 34 ~.-
8 1 y $';: L45 T :3. !!! jm - 2 J? .5 ;: " z x. + +
. ,, .m n- ./.
r tw
'c : g,,:
~
tw I 2:
%: 9: , W Ew -
g: %
+ 9:
S
- X: -:
- ,
- 3 a
9 5 e :p: ;<
.9
~
+ m24 ,n s 2 1 1 1 02 04 06 1 14 1.0 2M 3 55 03 06 04 12 16 2 2 00 them vots e e sa cATAwsj;tggdN%EINbicNs-A e,m - _
= - = - - Ino 1796 its:Xr " 90 im , , 80 !
,1 *x+
g
- l ;ll 7o .!
5 im
,o,e 2'
' 181
- .a1
.- r, - n
. uxo .
e, ...
. .so I y so .- .':/.; m
~~*a .
4o 5 xs 5 un i
, 440 e se .:/
m 1% r5 20 $ it0 k ' - I I!
. p 75
'; /. 114 to 6 s ' ) Q 22 8
- 3 4 1 1 IO '4 4 ' ' 01' -
01' '0 3 ' '06' '10' '14 " '18' '2.71
-06 42 00 02 04 08 12 16 2 21 votaom cwtmm CAT AWBA 1: PAOJECTED EOC-7 00PPIN VOLTAGE DiSTRBUTION SG C ALL BOB 81N Coll. lNDICATONS at
$ 2Q N 7 P30 @> - 7 10 7 '
l h, 'h . k) W .,,, . f. T n # '
,33 m
$ iso ? L
'[
.' h h 15 tw 2
/
. L?;
+
e i
;t:
a> T sg
- 37 h I i8 ; $
-f ;k' ;h $M 35 h01 h
' 0.3 ' '05' 1
O7'
'09'
'; $! 2
' 1,1 ' '14' h ,$_
~ 18'
_2
'22 1 1
- 02 04 06 08 1.0 12 16 20 2.24 Pregaded Batt*1Votagn Figure 10 3. Steam Generator C ECC 7 Voltage Distribution Based on All Bobbin Coil indications
. - - - _ -. _. , . , . . . _ , m .- _ -
- _ ~ . .
_-..___ .-_._. _ . _ . __ _ . _ _ _ _ _ . - _ . _ . . - _ _ _ , _ _ . _ _ _ _ _ _ _ _ . . - . _ . ..m_. CATAWBA t 19'/2 TSP INDICATONS-LG D AtilOBbN CDiL IND6CATONS w a, av l 7(D ...y ; lh 617 m,s . .. .
,, w .
ew w e ns { 9V 9; f'
~
v4 i t 4ce
- I 4: :': 4! 7
{2 1" ' 4: W 743 ~73 v
,, :.. x:
e, an. - s. ,... f ',, 19 7: $: 4;: Inn ite i, M , X - M
: ' ' S' '/[. _16, p 3 3 g 9 02 04 06 1 14 18 2 19 03 06 08 12 16 2 20 t6 tic hats t
l C ATAWBA 19241 VOLTAGF GnQWiWALL SG All LOBBIN COIL INDICAHONS -
= - 1(o 7xx- _
17 % ilt t? " - 90
, ifd e r
.~
, -80 f .
-70 ,,
g 14?NP llll ll 1Me Y r -
-60 h*
1018 , If11r 9 I' **M -82) 1,.. _, yi '.
.01 *
-X)
% 14 0 .
2 440 yo -
'/ >. N am y,% -3 ,
~
ap yg MN
/
k '
- ' 114 -10 i O28 3 3 4 1 1
. t' O' '4 4' M 1' '01 ' '07 '06' ' 10 ' '14 'IS' '211 46 42 00 02 04 08 12 16 221 vas ap cutwi CATAWBA 1 PROJECTED EOC 7 BOBBN VOLTASE DSTRBt. MON 4G D ALL BOB 84N COfL INOiCATONS 7tr ese tsn 7 g4
?
? fin *f
, , 'm , ,
1 < m l *n $ $ ya 5 , ; ,; e B are ',
, A' -
, ;wa
~
I .i f. h no E i' F 'N
, / tr ilt 2 T ' ', 3 gg
., /, ';
10L* j, J,1 ~ ggs 7, / i i
.,c y 31
' ' 'A
@ g g 13 4 3 i , ,
01' ' 0 3' '0S' 07' ' 0 9' '11' '14' '18' '2 15'
'2.33 02 04 06 08 10 12 1.6 20 225
- r,*o aum we, Figure 10-4. Steam Generator D EOC 7 Voltage Distr:bution Based on All ,
Bobbin Coilindications e.--- ,--iv-y,-gw y--se% +---s,v-w,hy-*.i-v ,---e+- --e*-
- wem o w -e--. 2-w. w-n -ey-.w-giw.nd-+.- ap ,+%,.. y.,-s--uss-qsw w y-- m raw erw.4<* gr-. am-,m.e-.z-wtg w- ----- y y-y yw'--wyay
a s
- Figure 10-5. Leak Rate Under Norms: Operating Conditions versus Crack Length for 3/4 inch Tubing
~. - . - . - - . - _ - . - . _ _ _ - . _
i
. l l
BURST PRESSURE VERSUS CRACK LENGTH l 3/4X0.043 INCH TUBING 10 DELGIAN UNSERLED BURST ORTA BRSE 9 - l e 7 Q y 6 - Pn Fu 5 - x E 4 _ __375_0__PS_I___\_ i m 3 j ____2g50_fgL__j_______ . i 2 l l 0.48'l 10.70' l i l I D
' ' ' ' ' ' ' ' ' ' ' ' ' ' ' d 0.00.10.20.30.40.50.60.70.80.91.01.1 1.21.31.41.5 m oon.m RXIRL CRACK LENGTH, INCH e
Figure 10-6. Burst Pressure versus Crack Length for 3/4 inch Tubing
Appendix A < Catawba Nuclear Station, Unit 1 Hot Leg Tube Support Re sizing Analysis Guidelines and { Hot Leg Tube Support Re sizing Analysis Guidelines for Growth Trending This appendix summarizes the guidelines used for re analysis of the hot leg tube support plate (HL TSP) calls from the Catawba Unit 1, EOC 6 bobbin coilinspection program. The initial oddy current data analysis was performed in accordance with
- Eddy Current Analysis Guidelines, Catawba Nuclear Station Unit 1", Revision 2, dated 7/9/92.
Following the initial inspection program, the guidelines presented in this appendix were used to obtain voltage values consistent with Westinghouse recommendations for measuring bobbin voltages for ODSCC degradation at HL TSPs. This appendix also , includes guidelines used to obtain ODSCC growth trending information,
, The following pages show the HL TSP re sizing cnalysis guidelines and the guidelines for growth trending. Figures A 1 through A 13 are provided to support the application of these guidelines, F
4 a 1
- , , + . - - . - - . . . . . . - - - . - . - . , , . . - . . ~ . . - < . . - , - . , - . .,..,..,,-,,,.,w- .-.- -,c..- ,, , , . - - . - -, . , , - , , , -..x.-m-.-,m-,-=.-.--m-..-,-.. ,-. , , -
liotleg Tube Support Re-sizing Analysis Guidelines Technical Approval em //t./h, - + ~ Dated J L
}{DE Level III, Eddy Current NPD Concurrence: )
Support Engineer, Nublear Sere cos Date kN Revision 0 - 08/12/92 P l* . l
i l ! Catawba Nuclear Station ' t
' Unit 1 Hot Log Tube Support
,. Ro-sizing Analysis Guidelines 1.0
Introduction:
Catawba Unit 1 is implementing a Hot Leg Tube Support voltage critoria for dispositioning Hot IAg TSP indications. This will require Re-sizing of all reported H/L TSP indications ! using a 550/130 Kilz differential mix. All analystn will be trained in Re-aizing TSP Indicaticns and the uno of thn resolution modo prior to analyzing data. 4
hese guidelinen are in addition to the present Catawba Unit 1 Guidelines Rev. 2 dated 7/9/92.
2.0 Mixes
' - only one additional mix will be required for Ro-wizing, a 4
$$0/130 KHz differential mix (Mix-5) shallThio be established as the primary H/L TSP nuppression mix. mix shall be ,
- accomplished by using the drilled TSP in the calibration ,
standard.
3.0 Normalization
The voltage of the 550/130 KHz differential mix channel will be net at 2.75 volts on the four 20% flat bottom holes, and naved/ stored to all other channels. i 4.0 Re-sizing To assure that a proper voltage is obtained only the ODSCC signal.will be sized and not the TSP residual. This shall be 4 accomplished rf loviewing the TSP using the 550/130 KHz mix differential channel and setting the measurement " BALLS" so only the indication signal is being sized using P/P volts-If the ODSCC signal cannot be distinguished from the mix residual- .
; then the 550 or 400 KHz differential channel can be used to set the P/P voltage then flip to the 550/130 KHz mix to report ;
the indication. Insure that the P/P voltage points go from one edge of the indication to the other edge. All indications shall be reported off of Mix 5 550/130 KHz differential mix. (reference pictures attached for correct sizing.)
-re- ,, r mvr -- c e---= - , ---,ve, -..ww-- am, ,,--w-- ---,.c,.r--,,-r, -,-m-, ,,vr ,w,a,,,- s--,-<r -o,--,,ve--,,e-mw -w.- r ,,-e m v r-s , m r- m m ,,re,-wn. -n wm --
5.0 Reporting
- Ro-nizing vill be accomplished using the lot Re-oval ,
mod e .
- only tube supportn will be Re-sized and the call will ~
be ontered into the report using the lot Re-oval.
- A final report by taps will be used to determine what tubes have TSP calls. Oraly the tubes with TSP calls need i
review. 6.0 Graphic Printoute: How graphic printouts will be required. A graphic of the i reporting channol 400/130 XY Liss and D * '*111 be made. The l D-8 shall be comprised of channels 1, 3, d Mix 1 on top-l and 2, 4, G and Mix 5 on the bottom. The -a shall be auto scaled in order to maintain a constant representation of the l indications being report 6d. 7.0 Data Flow:
- A sign off tapo log will be maintained. When a taps in
! complete it will bo signed off by the analyst. ' 1
- Now get shells will be printed by the analyst once the tapo is complote. This data will bo printed and saved. .
l
- Prior to turning in the new get shell and new pictures, the analyst will compare his get shall to the firnt final l
report to onnure he has identified all TSP calln. l
- All got shells and new pictures will be put into a folder and put into a box marked Re-Eval-in the lead analyst cube.
( 0.0 Lead and Resolution Analyst Responsibilities:-
- Compara original get shell to now get shell. Make corrections as required.
- Review all pictures to ensure the indications aro l reported correctly. If not, resolve as required.
t
- Sond new got shells to Duke Power to be loaded into tho l data base.
i I i
o
}!otleg Tube Support Re-cizing Analysis Guidelines For Growth Tronding e
< Technical Approval: w . M1 d,m.to. .. Date 8 3 L
}{DE Lovt 1 III, T.ddy Current NPD Concurrence VN Date O Support Engineer, Nu'oloar Services Revision 0 - 08/12/92 l
l s-4 l-l
Catawba Nuclear Station Unit 1 i Hot Leg Tube Support Re-sizing Analysis Guidelines ' For Growth Trending . 1.0
Introduction:
Catawba Unit 1 is implementing a Hot Leg Tube Support growth trending study for dispositioning Hot Leg TSP indications. This will require Re-sizing of all reported H/L TSP . indications using a 400/130 KHz differential mix. All analysts will be trained in re siz1ng TSP indications prior to analyzing data. These guidelines are in addition to the present Catawba Unit 1 Guidelines, Rev. 2 dated 7/9/92.
2.0 Mixes
Mix 1 will be used for growth trending. This mix is a 400/130 KHz differential mix. This mix shall be accomplished by suppressing the drilled TSP in the calibration standard. l A 550/130 kHz dif ferential mix (Mix 5) shall be established as an additional H/L TSP suppression mix. This mix shall be - accomplished by suppressing the drilled TSP in the calibration standard. 3.0 Normalization; The voltage of the 400/130 KHz differential channel will be set at 5.75 volts P/P on the four 20% flat bottom holes, and saved / stored to all other channels. 4.0 Re-sizing To assure that a proper voltage is obtained only the ODSCC signal vill be sized and not the TSP residual. This shall be accomplished by reviewing the TSP using the 400/130 KHz differential mix channel and setting the measurement " BALLS" so only the indication signal is being sized using P/P ensuring that the P/P voltage points go from one edge of the indication to the other edge and that the maximum voltage is displayed. If the ODSCC signal cannot be distinguished from the mix residual then the 550 or 400 KHz differential channel can be used to set th. P/P voltage. Then, flip to the 400/130 Kitz fferential mix to report the indication. All indic .lons shall be reported of f the 440/130 KHz dif ferential , mix ( aference Figures 1 thru 13 for correct sizing).
5.0 Reporting
- Re-sizing will be acccuplished using the 2ND_RESIZE
, window.
only tube support intersections with existing calls shall be resized.
- A listing of existing calls will be provided to determine which tubes have TSP calls.
- The analyst shall visit only those tubes with existing TSP calls and resize tne indications at those intersections.
6.0 Graphic Printouts: New graphic printouts will be required. A graphic of the reporting channel 400/130 XY Liss and D-8 shall be made. The D-8 shall be comprised of channels 1, 3, 5 and Mix 1 on top and 2, 4, 6 and Mix 5 on the bottom. The D-8 nhall be auto scaled and refreshed in order to maintain a constant representation of the indications being reported. 7.0 Data Flow:
- A sign off tape log vill be maintair.ed. When a tape is complete it will be signed off by the analyst.
The analyst shall print a build report for each cal group.
- Prior to turning in the results and new pictures, the analyst will compare his build report to the first final report to ensure he has identified all TSP calls.
All build reports and new pictures will be sent to Duke Power at ncGuire. > 8.0 Resolution Analyst Responsibilities: Compare original report to the build report. Make-corrections as required.
- P= view all pictures to ensure the indi0ations are reported correctly. If not, resolve as required. .
Print /Save get_ shells under 2nd Re-size. Mark the get shells as "2ND RESIZE" and turn in to Data Management in the "2ND RESIZE" box. Put pictures in the "2ND RESIZE" TO BE FILED BOX.
-,,. , m, . . .. _ . -,,_m. , _ . - . , _ . . ~ _
m ,-
N Q( 0 9 S=a f
< Q 1,4 t t6 ,
U $ l 3 c. k1 ~I e $$i I3 g $ E e y[
-.${1 t [
" a T
2 -
> C f
ij 5{3 21 j- 1 j l ,
, i x .
h% g p
,' A W \';* %W n / h R m W- .
l ,, e
- a/ 1 e e e
[
~
e b s D .c ~ l du
~
s g ,
"o p -
3 e, -- ( D c ~ -. a 2
~
a [, /
/ I '
s E V, 5 e ) E
\
i }
,; E e
T
'a i
is t m x) g s x m- - I
$'d [ < *T
.w
- $= -
M 3 E5 i A y
) '[\,m E5 '
R
- - - }
e 9 k ? n l 0" ( v' ~~ s ,_.._..^- -- { f.
?
,' 1, 1
~= v I, p, E b
, a v it J fu-*c
] .
h
; y ; , 7 j -
a: ,
;' l U 5
n ! ea m g .. g! T'_ I'
\ h
,gj i i h --. p . _.-m _._ , , , ,. _. __
(r _1 _ ( .: j .
" L, ..
e es s s* a s> a s a s* s I r --
- e- *-
4 *. I CE D
@ 1 1, 3 [ d b b i s Ta n . x e m 3 ,
T -
< i
/ .
1 -
$@r f
4_ ( Gc f- .. G M T Y O L % m t 4 e o vT o 7(3 NM l A t O t e o n e .
*' a l s ,e
=_
t i s L a ) w nhC 3 4 s , r ** e r e u 3 t
- - t e e
e t a e g #" 3 t 0 3 nr i t t e s h <
- t e e
.e i
t e S v _
- S r
c H ,N 0 0_ . 0 F 3 9 _. F I e n , D C C 1 2 r _ v
=
= W. t
- e \
3 4 % y. m 1 1: d \ m 3 ; 7 P / c 3 = t L 3. M O g d C 2 O 4 S I e 1 (, N/ t F e n F a W G O R I C 1
=
W 1 g 0 = w 3 p ~ v C 1 s W - 5 : m 2. 0 1 G 3 d m 5 1 3 3 ,4 c /.
/
C f E
- 6 s
a d - 2 9 1 I W - l I ! 2- e e 4 5 0 9 5 2 9 0 A wLU e
- 2 1 1 D
. e r s o c J F e m 4 n 2 l s F G 5 u a I g 0 3 3 D [ i F 4 1 8 5: 0 0 6 o v
=
= W e
1 4 s 7 w ~ p I m 9 S. R 3: d m " S F 2 E E L
/
v 3 F 1 9 it 0 1 9 ,, M
- e. a n c. 3 o aa 0 e A 2 t
h600 , e n 2 1 ts e e i a c 6 p ! se a D a 4t t
. 5 0
3 p A' % v
=
= W t
a 5 o ' v o s 3 J t : R 0 a 1 = C d m +
/
v 9 4 C B 0 M- - i C 2 L C S k t
' O n t
' = C 3
r e - { 1 f ! v D O M)P 6 l 0 ; t ,o 5 e e bM1,lk! ' < F
,?":i q 1 )k+ -
3 N , 1 P r T T-
^-
r
.P P P % p P
o Hs 3 $ 5 S ,s S
- - t 1 ! T T i I t 5 k
r i T Y m 3 H n n - H I 5 R y ,T L a 1 . 3 T 4 S n e 9 E ', E: ,I g b
!l
, .iIj ii ij' <;ib j ; i ; > I e
ijI'f .I{.' ;lI, 'lr - ;;[f,!e! ![v ;(f : ; 1 I Efob C M t CmQ [ v y C H
~
N (Y Y 0 , 4 N t e O wte t s r
" .s . r Le w F
' 2 s s . . ,.
e
- > 1 M Q . . -
L w w t t u 4 C u - e t m-
- r o e a
- t. g r
e a z N 2 x a t n a. -
< t '
s ,- t e a L a e C a ' D ' e P .
% M i
L { 3ff. . o F
?
0 2
% s-i
" f i
D 5 1 e o e c v 9 d
- s 2 1 a- 1 '
S d ' 7 P / 1 . t 2 i L 0 s O j o C e F F I 2 4 1 r o t c a e r d 0 4 ( - A. D 3 8t 1 - 0 1 - , , 3 v 9 o C 1 s % a 1 3 5
/
a_ n S G 5: d - 0 e y
/
e
- t 0 s 6 u u
3- . 2 1 - i 9 - P , A - e o S 5 e o r u g wZU J L T F I 2 9 e
\
s w 1 9 9 T D N 2 tr . i . e F e m:5 ai 2 3 D 0 0 8 6 p o N v
=
i d w 9 .
- s. R mF a
I t 3 d 9
/
v
/ t u
- 5. -
3 J , . oo l 0 g
- t. a o n c. 3 e
[
- .e
- oo f F
9 2 e
\' t a
n G s 3
- =, cw F c n
=
4r t a t D 0 3 \ ( 0 5
- 5 p v v
= 5 o 2 e e 4 z l : a.
a 1 C d I
- t 1
pn 8 C 4 u g B C \ C 2 C *x 5 k D # t e { i 0 e 3 s 1 w D C : 6 D S 3 s t W e e - I$l h'
'd*
- <QL ; * ;; ;jst [ }
t - P . ll[ - -
~ - - -
y P P
*r 4, L NY S
T ' P S T D Y' I S I H t w S T
'H S
T M L a T S n 2
- T d
t S t 6 t 7 e fI
- 4) i5 lj) .j J4,a
! - 1 $ I: 1 l
- ij1!I: j- : f; :; i .
1 1 9 m 914,. t i Y :
' a *#
g e t Et y c. d' o x - kx 1 4 ~
, 6us m h t. s& q* k- . . - . . .
~j i '
i 2 i ?1t g2 I ; , 1l E v4 - . e I
~
a u-z v v ia f
; :g 1
r - r a: ' t* Ik
, n f_ I s
yp 3 g W"Wh a i
% ]2 s $5 d d
, y .
9\- s
.1
-g
%(-
h - u y \ g
}
t u
- "g .
g. t\ 3 'l n I d b -
\
l tt 1
\i s y J l . \:
T
< 3 k, 0 8 b [
g k g 6 > g g $ - a E 3xi a 8, a !
. a
)[
. d.
. . J ,
g og .. tt E Aq $ -J . s E s s.
- - ~
v h .a 3.- g e 2 a 3- ., a , 5 a 7 s / 1 s 3 U d a 4 e em , 8 2
; ; % = u. m .
. ._ L
- E l y th h b b h h h b h G " 3 5 3 1 - r. m 3, 3,, I.
~ . - . _ .- .,_ %. , - , - , - _ , , , - -...y. . - . , , , . . - - , - - - , ,
_a..e4-.h 4 4. .J. 43_pu.A-.t-4 *-4'm- hsh2 EadL.M-M ia. ap.m.gwhwa,,w_4.AA4_A .W hh gJAa p;4..w.%.4.+ha--awa-M W4 mp-,d.Lewem .ea. r em_ .L 4m.m,,3.3%._,.saw.w..Mwa..a.Je.-..Jma.nj._,,,A,wgma, Ne J f o% - o
~h a 2%
5. b5n
?,_
i e - i < j# 1 0 ~ R W 3 je T
$ f. 5 ij j a .
~ g
~
P - y -- , g ,. .- e d( 2 a* 2 O
$ [
y 8
~
f
~
2 . i 6 c g g 3 - l E o 8 g g l u k- 2 8 e . = -' I S 6 g I [. 2, \ l I g o - h:
- 5 s J I . (
l , m
}5 s h' E-h 14 i . , a g - .
I b O f7" 2 8
~
j fJ g .o , [ sy m % a 3 J Aq - g. g g
%- ' . g A
'd O. O e e
]g= { *
- --- ; g -- -,
- \ g
~
e! a 5 g , 4
. 2 1 ., a -
a
~
~-
- 1
- o u ? '
n u e 5 i a N/ . a . R e -
~
a
- li f-
-n 3 5 fl .=::=.- : _, _ _ , wn 3 , L _
U h1 h, k h. h. a - 4 A n . ~ 1 i e e . - . . . - . ~ .- -...,we. -w---w . , - , , - , - . - - - + _ -, --
= _ . _ _ . . . . _ _ _ _ _ _ . _ _ _ _ _ . _ .. . . . ._._m . _ _ _ . . _ _._ _ _ . _ _ _ _ _ , _ . _ . , _ _ . , _ _ . _ _ _ _ _ _ . . _ _ . _ , , _ , . _ ,
,, h 'd l w Y l e
g, '- #v< a hp e p a ^s v. t f% hb d C d Y e t. E d.,. ) d. Q l x d~ q f o
~
{ n i
- t. d 6 e _
3 3 i- 6 @~ 8 JG
- g
.. g I 2 :a } , 31 a: s r 3 J ,
1
%yJ r , u e !
i e, g
~
g ? -* 4 % % l i . . 8, A -__ -_ -_ _ _ _ _ e _
- v. < g I n
A g j '
" [* '
w
)
N
~
g
- 5 v - g
}"
a 3 , i e /\\-_ 5 Dl I e
!l
, s 8
f
?
1 8 E 5
/U -
3 - o . . O '
.. l h N E ='. .
2 3s
/ [ E b l* R J !
3 s
- JL
%r{ - ,
n : n (y a t - i a R . . _ , - - - ,, - - - - U Ji l a I
'. ,u m . ~ ., .
,, la u Aam - .
L--
'-1
.[-
n , , m' 81 - 1 m- .d.h..,,,.m,- b ha - - , - -
)I i il i' 4 ,, , , , , , , , ,
b b b b b b b b l } EBR R ? A i ! A i l 6 _. ..s_ - - - - - - I l l-. - _ . _ _ , - , . - - . _ - _ . - - . _ . . - - , . _ _ - _ . . _ - __
-%a_ #+.. - 4- 4e>e..4-oA 5-- ----
+ ,s.--.=---E----L--44 J-.J -AWl. AmJ Jhe -.4AA*4 M4aW.AAm,.- #ehm u4 J 4-A+ 4 444-- Ar 4*eJ4-- 4-. ----
0, d k,(e
$($q
- 0 v
x h\ 9k - r> u. v= yo 3; .) Q Lu c- 2 L .h O 3 __ , s: 11 - yi a 1t - a s r :. I I ?6 y e e- 1 :1 e s: ~ g w:a 1) y ea a: ' f ;g ! )[ r -
. t b
Y QJ j %)V
%/,
e ; ! i
- 4 5 e
.~
, e" q g m _*[_- ._
.ij d ~ / g I. p .
t
\ / j i !;
f C g N# D >
.,1a. .
u - g g E 6 = I d 8, a l iA 3 g
?? E h
D
< ,7 5
F. , [ 8- - O w b . l C 5 b B t 3 O E 3 &" n 1 { g.. 8, t , s , 8 E s $ x ; - g. g 5 'd N N
. b s
[>, ii m
!! 5 I !=
l i!
= " s i u
- d / _
$} ,
E , i I;ati e l l ,, j ! = g . , . . l 8 s l }
- e 1
y Mgh'- .'^y
a_ 1
,mawi ia ,
r : i .
! i i i i - , , , , . .
( ,(h h - - h h h
- h_ h h
h
> Pa >
a - A A = n . ~ m Li l 1 - - - - - - -
a - 6.s. .s,, os ....m.m_m_..a4 m 4.,4.a-. a 4,_a,..est,ma.4--uam.u. p ra,.m._a..u,a-w.__. ,A,mm..--.,--e4.e ma.-Ama .m A_mm m
- n. m e
e a e og - U S' d0awh' v , 0 -3 $, 9
. el 8 e[k ef t15 &&s ee s&
j 2 5 y r t g 5
. ~= Ig T w
1 { l
~g 9
y - C y 4 i
%
- i 1
- t
~
j w t ; a - El t M m.
~
s '4 3, g s% Cu
~
~
L i
~
e
~.
m j ~0 I
~
f, ' s E $ E b : . 3 v 6 1 , g -
*a e e 1 E u
!e i a f 7 i
, n a e a
$ !f i O 5 5 E a a: s
> a i,- i - e YS k. ie _ _ ._
jj 'j \ B N
! h,2!
1 g g
- M lU l
3 f ! ! A Sr
, e 4 e. g a g >
r l a j f ****"rr' Mit.Ji;/. wW @ { 5I;
*
- M *' "
i ; i
; )
=u- g g i e
sn se , s s. ; s
a e m Ahh,3-4 e sM,_ _J.c_s-.'.. Am y 4 -Mh46h4 4J a-JA M- W A M JE M w,mA.-nud,h,.h +66-0 .M re A .am-+ M'A e -vA 4-e-h.MeMm_A s s..M a1AdA._-aAm-
- vmaM m%em.s h e _ A h' A .44O AJ6A.eJ.
1< e ;s. g eo4 ev: 9 o vg. 095 hh $h esi b .9 s Ag d 3, o:. > ze i
~g e
a u
- a y
o 4 a i-yl :,i ?l
- t a - . . ei r, .
.m__.
A_ b a
, s u
w- . 5 2 I $ .!
- % i
* - %g -
g o 5 - , 4 ,
~
~
j r g
=
5
~
8 a
\ -
JM m , j a "
; \ 1 2
g. m !? J E . _k / i 4 !? / ! b
;n 1,4 .E
=
5
,n. g m .
(- -- E 2 eg A , ; S. a _J w! - E : h! . N ft m t c k
) !n
~. .
)
u ' s ; -
^
}
H d
-a . ,
R j -
~ _
i
-5 h i
[-
% 5 -
, ed*Nw%me %pNM lg +
=
a > e 1 > a , s 4 4 3 3
>*
- M >
- - - i s =u 4 - A 1, i n
i i
~ s. ,
e
~
r
h..A& -,A_i -, 4 <--- A4 -i- 4'M-+-_w,& N J--e4 44 m.A,,a.e, .4wh.A_,,44,,.4,,,J, J,.a__,aahJ.--Aw-,4AM-a4% mmh-*sme.-- t-w r'-4a-JltJs umdad-.. 4s=44A 4 4 4rldim.m 2 ~4.-ua3,mmia.m k J ( ; 3 " .oh9 og - a . t
$ )
2 M a o t O e . Q.m $ t l 4 2e et [ vl y ga
. j a3 i !! 11 13 1x
~ g H
j , u h 9, ,
, ) - - -
A g d u g \
~
( l k a u* h (O , i
=
a O a-8 o u i - la t d o
.~
-e _
- I - /
N, M . e _d l w ' g 5
- k e 5 8 D
" M O y E n e
;e E-
~
a - SE E
-i: -
g a
- )'
5 5 E e I. r ' - . g
; s y / 2 i E C % 5 u
i
. a V Y rA, i ,
.l
.. , - (-
o 6 3 l n > ,
~_ x_. . .
' 11 e f i Y$ $ $ $ $ $
3 ;- - - 5
, s
{ ~
~m - { [ E, E n
3 ~
$ 5 me,, in ---+s-3.- ....w..,_.--.y,r._,_e--, - _ _ . - , . - - . , , ,,_r . - , , , , , , - . , . . . - , , , ,. , . . , , < . % .~. ---, .,_ e
t t f L i i f , 'ii fl,[t f: r{ : - .l:- t ' T ? E a
- fwNNr_
l l [ 1Is d Lt _
' /
/
C 659 ,0 F
\
M 1 6(j a T _ / y , M t s
, n t.
t e nu c
, a a er t
= n a
-,r -
s
- nh C L % c T e e .o 4e 0 - - n 1
- - t t
~-
e t
=
y
~ 0 2 ,
2 x t t
= h a 4 4
s e1 t S v c n . c s
.e, e C a i u ,
- L M c F F t(f:-
- 7 0 0 0 1
F 2 F a t e e D I a 6 'i! G 0 9 1 = a w If 5 o w 2
- s 1 1
= 0 5
1 5 F d
/.
2 "t
=
[z < I L 0 S 23-O f C 2 4 y Q c e t
-kJ'N 1 1 1 -
F r a ( a d W
?
l G 4 O 2 C 3 - R 5 = 0
=
b 3 o v e C 1 s s : e. R. 2 3 a 1: 5 0 s 5 d a
/
.h v 0 e
t e 2 e. M - 9 C s - i 1 P e4 5 0 9 S 1
- n 2 - 9 2
1 T A a L J 1 D y.
== P e n P **
o J F a 5 9 3 ' l Y r (- r e. i G u 7 C e ig m5 s 3
- 0 0
3 o v
=
=
d 3 F a , t 4 s 4
- o 1 : a 2 s 8 3 mF d h m 1 t t
luc t t 5 0
/.
M + .
. a 0
- c. 3 e 0 8
1 w4 t 4 2 e n 0 1 .(
- weca C T r
i [ a c 9 \
, d a D 4 m t 1 A 0 .= 5 ,
fy* 5 a v ht
" s ?
w e a -
- 3 e 1 s m
" C /
0
" B 2 3
' C B 0 3 .f
- C
- 2 L O S b B *
- u. I A t 0 r ,
3 e ,LI e 1 v
- C C
6 C _.- P 5-t r 6 - ,' I fp5 ,[4 ( 6;ri ' r c gdLR' I<c'F ,1 - e - 3 1 i d ]" )< 5]s g' I 3 v 1
- -n _
P r- t P P P P S P P S 7 ' P S P S n- S S T S T T S T T , ' T 1 t , m i T D O H H M " H ., r
- T f , 1 S M P f L
o I 2 3 4 S 6 7 " 9 o
^ '
^
F f i l1
< ! * , 4 ,.* l' ; 1
~....._4._..-_,..a %, ....v ..uu -m.on-u.-+=._%+.,=-.-4-s-....2.u. m um -* - o m --.s . mm,- .-w.s+e+e.. .,..i- a m .__a , - - . . A_..%.q.,.. ..A_m.s_
, ~
V 14 - i Y a N E c h p o ' O a s 6
=
s -= i a ~ i n 11 : 2 I ; y
- 4 IR 1
~a iI vi
- )
! "i i *' : :
a1-t l
. 3 2 :s i
a- fIu t x
$ $ .A
" s
$ h j
1 i e Ay - 4 3 d w b I-t y A Q , E . r- I
\
u a - g
. 4,
/<.
b i j $ g 3 o Tg .
.y _ E , \/'
$ '? "
/
, 4 1x :
3 5 X
.- ~
e a O i~ t =a d o a O
=
g w 1 f. T [
- d gg a e
+ - 1 ".! Vhp N 5n
~"
E g u n 8 j ' se go j 5 E
- f. , 5 1- a ~h >
4s 3 g 5 1 s. o --
- 3 2 o I a
d g g = a a
\
3 a MADh 1 y pv w::.4=n: r 2.2. ;z , . ,., 2 :' 1 :- . y.m c :rw r i e 6 La , , , , .
= c- . . - 1 e i "e i t
f at A A : o i e ~ =
- 2
-w- .
9
- .---.,n . - . , - _ . _ - _ _ . ._.r s -, - , ~ _ .. ~ , -, .y.., 7 3 -- t . c.., -----..-...,._,.-,.,ww--<
, [;il ; i * ' i~ j ,Ii [ i I : ;t$ l ;!,i
.t>?
[{y 7- P,> ,
'J 8sU b
V 0 bl > u-1s J J 4I O D* 4 W _ d $' 3 C I C * ^9 l a ~s 7 . T % c 0 d b 6 C~ J 0
)/
t y M~ ~
> D i
K C*.I C Di% i bNR ,1 f( Ae L kO f g t e n a t s !
/ c r n . e
=
F a I a e i u . I F g,,,M m 2, h D e . - t-e o C L n n - - f n t
" t o e el n n t
=
y ,3 ( Z
' nt a t
s a h t a P a h A r e , L . e C D a C P . H . N F g-7 0 9 0 1 F 2 F I e i r e. A D 1 M p/ G n a d 3 4 1 v 5 p v 2
- o 3 1
R t = 0 5 d M ' 6 P / 1 w 2 L 0 M O
- C 2 0 4 0 1
7 1 3 e u a p F F I D 1 h* t 0 3 1 2 o w
- t
, : R f G = 8 W\p 5 d M o 1
$ I .
/ }
v 0 4 2 g C W - 3 - P 1 2 4 5 8 2 9
/N 0
8 1 S T t _ A t u j e n 6 s H T F e r F A G 1 7 .
, d u 2 4
D 1 a 2 ig : 0 3 m 1 4 0 p V F 4 o v . j R 9 p a: = 1 r d M
/ 1 lI w
2 l 0 M a 3 0
- c. 0 0 A 2 1 E L O e n 11 .
O F \j A 2 e p F I G a D d t 1 _ n 2 0 a 5 1
= 5 p V 5 o _ v e R 1 l :
C a 1 d k M
= 4 a
u m 9 C O B 2 O 5 L
/
2 0 5 M O
)
- k .
y D A t 2 = r
'O e
* :1 v u
z 0 0 6
' *(
C 5 3 V t r e :L._ Pf(tNl< jr g> q' -, i
,4,d 1' i d d
[ i, j-1 P a g P P P P P
. 1 5
1 i S Y" I S T 5 7 S T k t . r t D D H H H a N R I" T T T t 2 3 4 S 6 7 L e g ._ ~ m' ;^}}