ML20086A042
| ML20086A042 | |
| Person / Time | |
|---|---|
| Site: | Farley  |
| Issue date: | 10/31/1991 |
| From: | Wootten M WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP. |
| To: | |
| Shared Package | |
| ML19353B292 | List: |
| References | |
| SG-91-10-039, SG-91-10-39, WCAP-13104, NUDOCS 9111180204 | |
| Download: ML20086A042 (51) | |
Text
7.
yg .
ll. '
' ^ '
- g. 2 i j
-a.: .. v. .
i
, _ i ,
y a ;2p.
~
m ,
'g ENi
'm i V k;
x h .
\ p 4 ' ' '; :: -
5-
- . i em
[dl' '[ il (
.,4 +s. ..gr .. h, ' f ' r 3: ; y h p .
3 I
. 1 4
1-
- q/
J- . .
, 7 7 WJsST&$F43E31DyeBSchtsstras -
i WESTINGHOUSECLASS 3 WCAP-13104 SG-91 10-039 ADDITIONAL INFORMATION SUPPORTING SG TUBE SUPPORT PLATE PLUGGING CRITERIA FOR J. M. FARLEY UNITS 1 AND 2 October 1991 Approved by: 7 hA. J. Wootten, Ma' ger Steam Generator Te nology & Engineering C 1991 Westinghouse Electric Corporation All Rights Reserved
l l
TABLE OF CONTENTS SCTO4 IIILE PE E
1.0 INTRODUCTION
1-1 2.0 HESPONSESTOGENERALCONCERNS 2-1 2.1 Leak and Burst Data Base 2-1 2.2 NDE Measurement Uncertainty 2-2 2.3 IGA and Circumferential SCC at TSPs 2-7 2.4 Flaw Growth Allowance 2-8 3.0 RESPONSESTO SECTION2 OUESTIONS e 4.0 RESPONSESTO SECTION 4 QUESTIONS 4-1 5.0 RESPONSESTO SECTIONS QUESTICNS 5-1 6.0 RESPONSESTO SECTION 6 OUESTIONS 6-1 7.0 RESPONSES TOSECTION8 OUESTIONS 7-1 8.0 RESPONSESTOSECTION 9 OUESTIONS B-1 9.0 RESPONSES TOSECTION 12OUESTIONS 9-1
10.0 REFERENCES
10-1 Appendix A REQUEST FOR ADDITIONALINFORMATION A1 i
1.0 INTRODUCTION
This report provides responses to the NRC requeet of Reference 1 for additional information supplementing WCAP 12871 (Reference 2) with regard to plugging criteria for ODSCC at the Tube Support Plates (TSP). In addition to this report, WCAP-12871 has been revised to Rev.1 to incorporate new information and additional details in response to the Reference 1 questions. Where WCAP-12871, Rev.1 incorporates direct responses to the questions, this report references Rev.1 for additional information with only a summary given in this report. The " General Concerns" identified in Reference 1 are addressed in detail in this report.
The major topics updated in Revision 1 of WCAP-12871 include:
o Additional model boiler specimens and pulled tube data which significantly increase the data base for providing revised voltage versus burst pressure and voltage versus SLB leak rate correlations.
o The lower 95% uncertainty limit on the voltage versus burst pressure correlation is used to develop plugging limits rather than the previous lower bound of the test data.
o Voltage limits for tube plugging are independent of bobbin coil depth compared to previous differences below and above 50% depth.
o Voltage growth is apphed as percentage growth which results in added conservatism when applied at the plugging limits compared to absolute voltage growth derived from low voltage indications.
o Based on the increased data base, voltage limits for tube plugging are derived from the structurallimits for tube burst with a!)owances for NDE uncertainties and growth. Additional margins utilized in WCAP-12871 to reduce p'ugging limits pending completion of the additional testing are no longer required.
o increased discussion of NDE uncertainties, similar to Section 2.3 of this report.
1 -1
o Addition of an Appendix providing guidance on eddy current data collection and analysis.
o increased discussion of IGA detected in Farley pulled tubes.
o Revised analyses showing no tubes are susceptible to significant deformation under postulated combined LOCA and SSE events based on recent Model 51 SG seismic analysis and TSP crushing test results.
The responses to the NRC questions of Reference 1 are organized in a manner similar to that of the questions. Section 2 addresses the " General Concems" and Sections 3 to 9 address the " Request for AdditionalInformation" related to specific sections of WCAP-12871. Individual question responses use the page references of the questions.
Due to the length of the questions, the Reference 1 questions are not repeated in this report but are included as an Appendix for completeness of the documentation. Every effort has been made to explicitly respond to each issue addressed in Reference 1.
Section 10 provides a list of references cited in this report.
e 1-2 4
..-i.-a
l 2.0 RESPONSESTOGENERALCONCERNS This section provides responses to the general concerns section of Reference 1.
2.1 Burst and Leak Test Data Base (Ref.1, pages 1-2)
The test data developed for WCAP-12871, Rev.1 has increased the data base for tube H
burst from 17 points (13 model boiler,4 pulled tubes) to 40 points (27 model boiler, "
13 pulled tubes) and for leak rate correlations from 10 points (6 model boiler,4 pulled tubes) to 28 points (24 model boiler,4 pulled tubes). Only laboratory specimens prepared in model boilers, which closely simulate plant operating conditions, are used to develop the tube plugging correlations. The model boiler crack morphologies compare well with the pulled tube morphologies. The voltage / burst trends show good agreement between the pulled tube and model boiler data. Voltage / leak rate trends also show consistency althotgh the lower voltage pulled tubes tend toward lower leak rates than the model boiler specimens. Doped steam methods for sample preparation have been evaluated and found to be less representative of pulled tube behavior than model boiler specimens and therefore have not been used in the plugging limits. The doped steam results are documented in Reference 2 for completeness in describing the tests performed in support of the Farley submittal.
The model boiler specimens necessarily span a much wider voltage range than found in EC inspections as the wider range is necessary to develop meaningful correlations and
. estimate voltages corresponding to structurallimits such as burst pressures at 3 times normal operating and SLB pressure differentials. Sample preparation for most specimens was targeted toward deep and through wall cracks to obtain more limiting conditions for leak rate and burst data. A few specimens were removed from the model boilers pior to leakage and assist the pulled tube data in defining the lower voltage end of the correlations. Plant data are typically limited to partial depth indications as a consequence of the current plugging limits at 40% depth and thus are not adequate for determining tube integrity limits. Thus it was essential to developing plugging limits that the test specimen bobbin coil voltages significantly exceed those encountered in plant eddy current examinations.
Although pulled tube voltages and degradation levels contribute only to the low voltage end of the burst correlation, the pulled tubes provide characteristic crack morphologies 2-1 l
i for simulation in laboratory tests. The pulled tubes typically include ODSCC cracks combined with some level of IGA. Th. rarley pulled tubes showed minor IGA as described in Section 4 of Rev.1 for 3 SAP 12871. The levet of IGA has not changed c.gnificantly oetween tubes pulled in 1986 to 1990 and does not significantly influence structuralintegrity of the tubes For this reason,it was not considered necessary to prepare model boiler specimens with 1GA involvement. Growth of IGA requires significant time to develop even in accelerated laboratoy tests.
The integrity of the tube is limited primarily by the length and depth of the most limiting crack. Other degradation morphology features such as multiple cracks and limited IGA involvement tend to increase voltage levels compared to a single crack. It is recognl:ed that voltage is not a unique function of crack morphology and a given voltage value can be associated with a range of crack morphologies. This spread of morphologies rid associated burst pressures is incorporated into the plugging limits by utilizing the lower 95% confidence bound on the voltage / burst correlation. With this methodology, the lower burst limit envelopes a wide range of morphologies considered typical of many plant conditions including the Farley SGs.
As requested in the Reference 1 question, the plant identities associated with the plant labels used in the WCAP reports are given in Table 2.1.
Overall, the extended data base in Rev.1 of WCAP-12871 provides conservative voltage to burst and leakage correlsr.ons for the Farley SGs.
2.2 NDE Measurement Uncertainties (Ref.1, pages 2-3)
There is no intent in WCAP-12871 to imply that large dent signals such as Table 8.4 would not contribute significantly to voltage uncertainties. The conclusion relative to dents in Sect; >n 8.9 states that large amplitude cracks are detectable with minor denting while small E riplitude cracks can be masked by the dont signal. These considerations are further discussed in the second and third paragraphs of Section 8.3. It is recognized that the dent magnitude can limit applicability of the alternate plugging criteria such as when the dent voltage significantly exceeds the tube plugging voltage limit and the EC signal occurs near the peak of the dent signal. Even when the indication and dent signals overlap but the dent size (voltage amplitude) is less than that of the vc!! age limit for tube plugging, degradatiori detection and quantification is not a significant issue. When l l
2-2
the dent size is large compared to the voltage limit, detection will occur only as a distorted signal. Farley SG inspection experience has shown that indication amplituJes can be quantified at essentially all TSP intersections due to the I,mited prese: s of dents and the spatial separation of the indication and dent signals. Only about 550 tube to TSP intersections in the most significantly dented Farley-1 SG have dents greater than 2 4
volts and about 50 dents greater than the plugging limits. The Farley-2 SGs have insignificant denting. Guidelines on interpreting degradation signals h the presence of dents are discussed in Appendix A of WCAP-12871, Rev.1. Degradation detected by bobbin coil in the presence of overlapping large dents would be evaluated based on RPC inspections and depth limits for tube plugging.
The evaluation of dented tube conditions in WCAP-12871 represents a preliminary assessment of NDE and leakage for dented tube conditions. The specimens tested are described in Tables 7.5 to 7.7. Leak test results are given in Table 9.1. The results show that incipient denting and small dents [
]b,c. The dented conditions would also prevent tube burst. The analysis results of Section 11.1.4 show that even the limited corroded or dented TSP conditions in Farley-1 prevent TSP displacement under a postulated SLB event. Thus the assumption of open creviccs used to define tube plugging limits for Farley.1 is extremely conservative. Even if small errors in evaluating voltages at Farley-1 TSP intersections were assumed to occur, the errors would be expected to be inconsequential relative to leakage and burst considerations.
Westinghouse does not ccacur with the Reference 1, page 2 statement that [
]9 circumferential SCC and recent circumferential IGA at ][9 nay be good predictors of future tube degradation at Farley because they have had histories of degradation similar to Farley. [ ]9 degradr..on is substantially different from the Farley SGs. [ 19 denting magnitudes are considerably larger than those in the Farley SGs. The circumferentia' ODSCC is associated with the dented intersections. Axial cracking in the [ ]9 SGs !s dominantly PWSCC occurring also at the dented TSPs (extending beyond the thickness of the TSPs), while Farley degradation is axial ODSCC within the TSP envelope and independent of denting. The
[ ]9 inspection and pulled tubes have not identified any service induced circumferential cracks. The circumferential tearing in the [ ]9 pulled tube R12C8 resulted from the high forces required to pull the tube. IGA is frequently three 2-3
I dimensional or volumetric and thus circumferential involvement can be expected. The examination of 1991 [ ]9 tube R12C8, which was plugged for two years prior to tube pulling, showed axial cracks in addition to IGA and degradation at all 3 Intersections was detected by the bobbin coil inspection prior to the tube pull. [ ]9 tubes ,
R29C70 and R30C64 show negligible IGA involvement. Burst test results for six
[ ]9 TSP intersections are included in the voltage / burst correlation of Reference .
- 3. The data include the third TSP intersection of R12C8, which lies above the best fit to i
all the voltage / burst data (see Figure 9-2 and Table 6.2 of Reference 3), in either case, there is no basis to consider the level of IGA found in [ ]9 tube R12C8 as a predictor of the Farley degradation. Farley Units 1 and 2 have about 10 and 8 effective full power years of operation compared to about 7.8 years for [ J9 so that
[ ]9 should not be considered a predecessor for Farley, in addition, the Farley tube destructive analyses do not indcate that the types of tube wall degradation occurring at [ ]9 are beginning to occur in the Farley steam generators as implied in Reference 1, page 2. The only circumferentialindications found at the Farley TSP intersections are associated with minor branching of the axial cracks. The [ ]G circumferential ODSCC are distinct cracks in dented tubes and not relateo to branching of axial cracks. Circumferential cracks such as those at
[ ]9 are not and should not be considered in the NDE uncertainties for Farley SGs.
The presence of IGA is found at essentially all pulled tube TSP intersections. The differences are in magnitude of IGA involvement. The time sequence of pulled tubes from the Farley SGs between 1986 and 1990 does not show any identifiable growth in IGA involvement. Pending further evaluation of tube integrity with more extensive IGA (large fraction of surface area with depths greater than about 20%), the proposed plugging limits may not be applicable to plants with extensive IGA. Currently available data for indicaticns with IGA involvement and burst tests of laboratory prepared, .
i uniform IGA specimens (Section 9.8 of Reference 3) show consistent voltage response (see Section 8.1 of Reference 3) and voltage / burst trends between IGA, IGA / SCC and SCC specimens. The proposed plugging criteria are applicable for the Farley SGs based on the minor IGA involvement found from examinations of 14 tube to TSP intersections pulled o' 'r a four year span.
24
Specimens with EDM notches or drilled holes have been used as models to evaluate vo!! age trends, perform comparisons or assess sensitivity. For example, holes were i
used to assess eddy current uncertainties associated with reduced probe centering
. resulting from probe wear. The resuhs are implemented by establishing limits requiring probe replacement in the field rather than as a test developed uncertainty All data implemented to establish tube plugging limits for Farley are based on field or '
laboratory cracked tubes.
In most prior evaluations, the SG NDE uncertainty is determined as the difference between bobbin coilindicated depth versus actual depth from destructive tube examinations. This is not the case for voltage measurements such that the NDE uncertainties for voltage do not have such s unique interpretation. For volinge plugging criteria based upon voltage versus burst pressure correlations, the NDE voltage uncertainties affect both 'he voltage measurement and the spread or uncertainty in the burst pressure correlation. The goal for the voltage measurements is to miniml:0 the uncer'alnty on repeatine a measurement so that the uncertainty on the bust correlation f is reduced to the extent practical. The remaining voltage measurement variab!cs end up as part of the burst correlation uncertainty. For example, assume that a number of perfectly identical samples were prepared such that burst pressures would be identical.
If voltage measurements were then made with different probe diameters, calibration standards, open crevices, packed crevices, copper deposits in crevices, etc., the voltage measurement variability would then result in a apread in the voltage versus barst correlation. Clearly the goalis 13 miniml:e the burst ec* relation uncedainty (lower 95% confidence tiroit used for plugging criteria) by controlling the voltage variability.
The voltage measurement procedures must be the same for both laboratory and field implementation to apply the laborator, specimen NDE/ burst data for developing plugging limits. The NDE voltage uncertainty is then defined as the uncertainty in voltage repeatability between the laboratory and the field measurements.
As applied for the plugging limit development, the variables affecting the burst correlation are split into NDE uncertainties in determining voltage and burst correlation uncertainties. This atsessment and the resulting NDE uncertainties are described in Section 8.8 of Reference 3.
Probe design ditierences are eliminated by requiring that only bobbin coil probes with 0.06 inch coils and 0.06 inch spacing between coils be used for voltage measurements.
2-E
These values are commonly used by nearly all probe vendors. The voltage amplitude is a function of coil to coil spacing. For differential responses and a conter to center coil spacing of 0.12 inch, the influence of small changes in coli spacing such as associated with manufacturing tolerances is small. This sensitivity is also given in Section 8.8 of ,
Rcterence 3. Based on these data, a study of different probes of the same probe type, as suggested in Reference 1_. ls not considered to be necessary.
Data analysis guidelir'es for voltage amplitudes will be implemented (see Reference 3 Appendix A for input to Farley guidelines) to minimize the human factu variability in interpreting signal ampMudes. Somo uncertainty will remain and becomes refiocted in the spread of the voltage / burst correlation. The voltage limits move the amplitude of concern for tube plugging to generally higher signal to noise ratios so that human factors and details of interpretation guidelines become less significant. Westinghouse strongly disagrees with the M. J. Gallagher report on voltage variability cited in Reference 1 and expressed this opinion at the workshop. The variability in voltage growth rates cited for Figures 5.3 and 5.6 of Reference 2 reflect the nature of corrosion as well as the variability of measurements. Given that the indications are on the order of the detection 4
limit as noted above and that no criteria were applied for assuring probo centering or calibration standards, the trends in these figures are to be expected. The fluctuations in growth are expected to decrease as the voltage amplituos increases and as the voltage calibration standards are implemented. Data provided in Sections 5.3 and 6.8 of Reference 3 demonstrate the reduction in voltage growth fluctuations as the amplitude increases.
Uncertainties associated with field crevice conditions, I!ke the human factors, are more significant at the low emplitudes near detection threshnids than at the voltage plugging limits. This has been the experience in Farley SGs where distorted indications have been primarily low amplitude indications. Again, the larger amplitudes near voltage plugging limits provide more reliable quantification of the Indications than associated with current experience with depth limits for tube plugging.
Overall, the NDE uncertainty reflects measurement repeatability and is dominated by probe wear allowances which are limited by field implementation of a p.obe wear standard. Burst correlation uncertainties are dominated by crack morphology variations which are accounted for by application at the lower 95% uncertainty on the burst correlation for tube plugging limits. l 2-6
t 2.3 IGA and Circumforential ODSCC at TSPs (Rof.1 Page 3) '
As noted above, circumferential ODSCC at TSPs has only been identified in [
]9 whoh has significantly larger dont sizes than the Farley SGs (see Section 4.0).
When these ind', cations do occur, they are frequently not detectable by bobbin coil inspections. The issue of circumferential ODSCC other than branch extensions of axial cracks is not considered to be relevant to application of bobbin coil voltago plugging l limits or the 1.5 volt threshold fer RPC inspection in the Farley SGs.
As also noted above, significant IGA or progression of IGA growth has not boon found in the 14 tube to TSP Intersections removed from the Farley SGs, and the Farley units have greator effective full power years of operation thar [ ]9 SGs. Thus, there is no basis to believe (Reference 1, Page 3) that this damage mechanism is the next step in tube degradation at the Farley TSPs. However, implications of greater 13A involvomont than found in the Farley SGs and tho l ]Q pulled tubo results are being evaluated by Westinghouse. This offort has just recently been initiated and very preliminary data are summarized in Reference 3. Sections 0.7 and 8.1 address IGA detectability and Section 4.4 compares corrosion morphology found in pulled tubos.
it should be emphasized that tube wall degradation depth alone is not an adequate measure of tube integrity. Crack length and volume are needed as well as depth for tube burst ,
considerations. Voltage is strongly related to degradatinn volume. Pulled tube results to !
date indicate that IGA is accompanied by eracks and detectable by bobbin coil Inspection.
The 1.5 volt threshold for RPC inspection is adequate for the degradation found or expected at tho Farley TSPs. Results of further evaluations of the [ 19 indications at TSPs will be assessed by APCo and Westinghouse. If those assessments indicate that a
, modification of the inspection program is needed, the inspection plan will be ;
appropriately modified at that time.
Overall, the available pulled tubo results show comparablo voltige responses relative to maximum depth with no definable dependence on IGA involvement within the scatter of the data. Laboratory uniform IGA samples show significant voltage responses at 30%
depth although pulled tubes with uniform IGA are not available for comparison with the laboratory results. The available pulled tubos with significant IGA levels show IGA with cracks and have been found to be dettctable indications.
2-7
. - . ,a
2.4 Crack Growth Allowance The crack growth rate of 0.5 volts utilized for the plugging limits, as stated in Table 12.2 of WCAP 12781,is an enveloping average growth rate based on Farley Units 1 and 2 operating experience. As stated in Section 5.1, the Farley 1 average growth rate for Cycle 9 (4/88 to 10/89) was 0.4 volts / cycle based on the change in amplitude reported in both years. As stated in Section 5.2, the average growth rate for Farley 2 for the composito of,the three SGs was 0.06 volt in Cycle 7 (ending 10/90) with 0.12 volt as the highest in any SG. Thus 0.5 volts envelopes both units and provides a 0.1 voit margin for potent!al growth rate changes in future cycles, it is further noted in Section 5.1 and 5.2 that there has been little change in growth rates for the last two cycles in Farley 1 and the last three cycles in Farley.2 as shown in Figure 5.7. Since the same plugging limit is desired for both units, the 0.5 volt value was applied for both units.
Average growth rates are considered to provide an adequate (NRC concurrence per Reference 4) allowance for growth in satisfying RG 1.121 structural guidelines for burst pressure capability of three times normal operating pressure differential. Per RG 1.121, ar ellowance for eddy current uncertalnties is included in developing the tube plugging limit ! om the structural requiremont for tube burst as given in Table 12.2 of I
of References 2 and 3. Growth allowance uncertainties are addressed by demog trating large margins against burst at SLB conditions. This margin was discussed in Section 12.4 of Reference 2 and is further developed in Section 12.4.1 of Reference 3. It should be noted that WCAP-12871, Rev.1 revises the growth rato basis by using average percent voltage growth rather than absolute changes in voltage. This change provides more conservative growth allowances at the higher voltage plugging limit compared to the current low voltage data base for obtaining growth rates. However, the approach for growth rate development from prior inspection results and treatment of uncertaintles remains unchangod.
Factors contributing to ODSCC and crack growth at TSP locations are discussed in Section 5.2. Therein, it is noted that the ielatively constant small growth since 1986 is due to the inhibiting effects of boric acid treatment and improved contaminant controls implemented since 1986 at both Farley Units 1 and 2, Tracking of the 1990 indications back in time has shown that most of the indications initiated during periods of cliemical
! irabalance prior to 1986 and have shown modest growth with time. Fluctuations in tube '
i' plugging between recent outages have been loss dependent on crack initiation and growth 28 w.-*w c--, - - - , ,- . - - - .e , . . - - , ----s.
than " inspection transients" resulting from utilizing probes such as RPC, greater knowledge on interpretation of bobbin signals and changes in inspection guidelinos such
. as eliminating the 1.75 volt amplitude critorion.
Axial ODSCC at TSPs is influenced primarily by the crovice environmental conditions !
and local stress conditions are of secondary importance. This is demonstrated by the worldwide occurrence of ODSCC in non dented tubes wi'h dri!!ed TSPs and in sludge pile regions. Packed but undented crevices permit concentration of chemical contaminants which are the primary contributor to ODSCC . Sinco local stresses (inside and outside the thickness of the TSP) such as denting are not a strong influenco on ODSCC, these :
stressos do not drivo axial ODSCC crack growth. In addition, a review of industry operating experience on maximum observed crack lengths would add little useful information for the tube plugging limits based on voltage of this report. Crack length is !
not a plugging limit. ODSCC axla! cracks occur as networks of small cracks separated by ligaments. The voltage limits constrain the crack morphology left in service based on depth and remaining ligaments as wc!! as longth.
Westinghouse strongly disagrees with the Referenco 1 statement: *Given the significant uncertainties it appears the crack growth data is not conservatively bounded".
Westinghouse believes that developing growth rates on 3 to 5 years of past operating eFporience is the most reliable method for determining and bounding plant specific corrosion growth rates. This is consistent with industry wido practices.
The use of growth rates at low voltage levels typical of domestic operating plants to project growth at the higher voltage levels of the proposed plugging limits was evaluated.
The associated data are developed in Section 6.8 of Reference 3 and are summarized here.
The French uso voltage based tubo plugging criteria set at higher equivalent limits than proposed in References 2 or 3. Previously, no plugging limit was applied so that French growth data at high voltage levels can be used to assess growth dependence on initial voltage Icvel. The French data indicate that average percent voltage growth is approximately independent of initial amplitude and show greater scatter at low voltage levels than at higher levels which is consistent with the trend of the Farley data and is expected from detectability/noiso level considerations. The French data tend to show relatively constant percent growth with increasing amplitude while the limited Farley data show a tendency to decrease with amolitude. Overall, the data indicate that use of 29
constar t porcent voltage growth with increasing amplitude is more appropriate and ,
conservative than assurning constant absolute voltage growth. Therefore, percent voltage !
growth has been used in Revision 1 to WCAP.12871,
(
1 1
I l
t e
e i-l l
t 2 -10 f-
-___.,_-......I
t Tablo 2.1 Alphabetic Symbols for Plant Names Plant Letter Plant Name
~ ~
A 9 B !
c D
E F
G H
I J
K L
M i
N .
O s P
e l
2 11
_ . _ . . _ _ . _ . . . _ . - . . _ . - _ . - . _ . _ _ _ _ . - - . . _ - _ _ . . _ . _ _ . . _ - . . _ _ . . = _ - .. ..
_ _._ _. ~ _ ._ . - - _ - _ _ . . _ _ _ _ _ _ _ - . _ _ _ _ _ _ _ _ _ _ . _ . _ _ _ . _ . _ _ _ _ _
3.0 RESPONSESTO SECTION 2OUESTIONS Paco 2.2 Ooottions Relativo to 1.5 Volt Threshold (Ref.1. Paces 44)
The bobbin coil measurements used in support of the voltage plugging limits are
, different!al. Guldelines on interpretation of RPC data for ODSCC are included in Append!x A of Reference 3. The RPC data is evaluated for the appearance of axial crack-like indications. Crack lengths can be estimated from RPC data but are not used as part of the tube plugging criteria. The RPC data will be evaluated to confirm that cracks are within the confines o' 'c, TSP. However, the primary inspection for crack length extensions outside the TSP is based on the bobbin coil Inspection for which 100% of the affected intersections are evaluated. There is no assumed crack length associated with bobbin coilindications of 1.5 volts or less. Shallow cracks with ligaments are longer .i .
1.5 volts than deep cracks without ligaments.
5 D
a 31
4.0 RESPONSESTO SECTION 4 OUESTIOtL9 Pace 4.1 Dent Charaderization (Ref.1. Pace 4) 4 Due to the limited number of dents and small dont sizes present in Farley-1, no profilometry measurements have been performed in ths Farley SGs. The detectability and unceriainty in profilometry measurements are comparable to the Farley dent sizes.
As noted in Section 2, there is no corrolation of axial ODSCC indications with the presence of donts or dent sizes in the Farley SGs or other plants experiencing axial COSCC.
Figure 41 shows a historam of dont voltages for Farley 1 SG C. SG C has the largest number of dented intersections and has dents > 2 volts at only about 250 of the ~42,000 totalTSP intersections. Dent voltages are even less significant in Farley 2 and have not been evaluated. The total dent voltages from Lissajous plots are shown in Figura 41, although half of these values are more rolovant for assessing the potential for dont voltage influence on degradation detectability, it is soon from Figuro 41 that very few dents (-50) have voltages comparable to the voltage plugging limits. Also in Farley-1, the degradation signals tend to be spatially separated from the maximum dont voltago and have boon detectable and quantifiable. This is further discussed in Appendix A of Reference 3. For symmetrical dents, a radial dent size of 1 mil would yield a bobbin coil voltage of about 20 volts. Thus, the Farley SG dents are expected to be less than 1 mil.
Figure 4 2 shows the distribution of dents versus TSP elovation.
Pace 4.2 Circumferenth! IGA (Ref.1. Pace 4)
, Sketches of the crack distributions for Tubes R4C61 (Plant B-1) and R31046 (Farley 2, pulled in 1986) are included in Section 4 of Reference 3. Also included in the revised Section 4 are micrographs showing the crack morphology. Both tubes have a morphology that is dominated by ODSCC with some local areas of IGA. The bobbin emplitude for these indications was 1.9 volts for the 0.75 inch diameter tubing of R4C61 and is shown in Figure 6-1 of Reference 2 together with the RPC trace showing a dominant axial crack indication. The circumferential branching offocts are too small to be detected by RPC. The measured burst pressure for R4C61 was > 6750 psi. The bobbin amplitude for R31C46 was 7.2 volts (Table 6.2) Utilizing crack depths from l the destructive examination, the burst pressure of R31046 was calculated to exceed i
4-1
6000 psi (see response to Question 6.2 in Section 6 of this report). Both tubes are considered to have acceptable crack morphologies (axial ODSCC with minor IGA) for application of voltage based tube plugging limits.
It should again be noted that no circumferential cracking at TSP elevations was found in
[ ]Q. The[ ]Q burst test results are included in the ,oltage/ burst correlation of Reference 3.
Pace 4.7 .ODSCC Crackino Bevond TSP Boundaries (Ref.1. cace 4)
P Farley 1 Tube R20C26 exhibited very shallow,10% deep cracking extending about 0.27 inch above the top edge of the TSP. None of the other Farley pulled tubes or other tubes with axial ODSCC destructively examined by Westinghouse have exhibited cracks
- outside the TSP. Similarly, no plant inspections are known to have identified axial ODSCC outside the TSPs. The crack morphology for the R20C26 cracks outside the TSP was short ODSCC with multiple initiation sites as a continuation of the morphology intemal to the plate. The cracks above the plate edge do not reflect growth of specific microcracks from inside the TSP. Farley bobbin inspection results will continue to be evaluated for cracks outside the TSP as described in Appendix A of Reference 3 and indications outside the TSPs will he evaluated for plugging against current Tech. Spec.
criteria. Tech. Spec. plugging limits based on depth would be applied to any degradation outside the TSPs. Early model boiler specimens, such as 568 4, had frequent occurrences of cracks both extending into and within the Teflon supports used to hold the simulated TSPs in position on the tube. In some cases, cracks were found only in the Teflon supports. For this reason, the Teflon supports have been uud by themselves to produce ODSCC. No cases were found in model boilers that had cracks outside the TSP and Teflon collar combination. Cracks found outside the TSPs in operating 5Gs have been ,
determined by destructive examinations to be PWSCC resulting from the dented tube conditions. These conditions are not applicable to the Farley SGs. The technical justification that significant ODSCC cracking is not expected outside the TSPs is the '
- absence of the packed crevice conditions permitting concentration of chemical l contaminants. ODSCC cracks are initiated at multiple initial sites with loss of ligaments by corrosion to form macrocracks. Crack growth is driven by corrosion rates with l
stress as a secondary factor. Consequently, fracture mechanics analyses for crack growth have not been performed and would not provide meaningful information on growth of ODSCC.
l l 42
_ . _ . - _ . _ _ - - .. -- . - - . . - - . _ - - . ~ .. _ .=. . .-. .-
Pace 4.11 Effects of Indications Near Edces of TSPs (Ref.1 Pace 4)
Positioning of indications near the edge of the TSPs was not reported in Reference 2 as the effects on voltage measurements are small and ODSCC cracks would have to extend from the edge to beyond the center of the TSP to be of significant concem for tube integrity consideration. The positioning of degradation near the edge of the TSP can influence the responses; however, the principal contribution will be due to the mix residual that occurs at these locations. This response is typically small compared to the voltage plugging limits so that the amplitude of significant degradation will not be influenced by this residual.
To demonstrate the TSP edge effects, bobbin coil measurements were made on 1/4. inch EDM slots of 50% and 100% depth. Measurements were made for the slot at the center of the TSP and at the inside and outside edges of the TSP. Results of these measurements are given in Table 4.1. It is seen that the voltage values for the crack within the TSP is essentla!!y the same at the center and at the inside edge of the TSP. Variations in voltage with the slot inside the TSP are <2% for both 50% and 100% deep notches while moving the slot outside me TSP increased voltages by 510%. The bobbin coilindicated depth ,
changed by 29% as the 50% slot was moved from the conter to outside edge of f'le TSP.
These results support the conclusion that amplitude responses to degradation on the order of the 3 4 volt plugging limits will not be significantly impacted by the location of a crack within the TSP.
4-3
_ - . _ _ _ _ _ _ _ _ _ - . - _ _ . _ . . . . _ . _ . ___ .~. , _ , . . . _ -
.. _. _ - ___ .- . . _ _ ~ _ - - __ - . - - . . . . _ _
Table 4.1 Effect of Crack Location On Bobbin Coll Measurements 50% Deep Slot 100% Doop Slot Indication Location Vohace Depin Voltaae Qggb
' a,b,c
- 1. Slot centered in TSP l
- 3. Slot extending from !
TSP edge outside j ofTSP
[
\
44
Figure 41 FARLEY SUPPORT PLATE DENT ANALYSIS lEDA CALLS (2 VOLT THRESHOLD) 120 .
- 100 2
'm . . i - l jQQ-. .......4 . . . . . . . . . . . . . . . . ... . . . . . . . . . . . . . .
o - :
-80 o u.
S/G C 1990 DATA :: ,
p$
m 80- -
--? *-l-t i
-l-i :.
-l-
-70 5 p
- i. .
w cn : !i' : .
- -60 @
CD _-.-.-; . - . ~ . - - . . - - . : - - .-. : . . . ; - . . L - ; . . .- - - - . . . - . . - -
5 o 60 -
I g
- u. i
- i! i i j i :
i -50 O
...........q : : : .
. O c
W 40 i .............:...;........
....+....+......................._.
i i i :
-40 CD :
. i -
2 -30 i : : :- i-3 20-H}
-+ 7 2 ,
i 3 i ! : : i
- : i
- -20 s i ! ! ! ! i : !
l 1U 0 T-- i a i i i i i i i i i i i i i1
~2 3 4 5 6 7 8 9 10 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5 >10 VOLTAGE RANGE (UPPER END) 4-5
. . . ~ . - _ - - - . . - - - - - - . . _ _ _ _ _ . - . . - - . . . - _ . - - - - - - - - - . - - - . - - - . .
i
)
Figure 4 2 l i
i FARLEY SUPPORT PLATE DENT ANALYSIS lEDA CALLS (2 VOLT THRESHOLD) 35- S/G C 1990 DATA 30-
, en
$ 25 #
g . . . . . . .
- u. '*"~'"
O 20- .
! @ m, m
2 15Mw e .:m 5 fidh!;
10;?pjw n n5.,. ,
l N HOT LEG I 5 "if. ~ 2 1
^
3 4 5 6 7 COLD LEG ELEVATION I
I J
1 46 l
i.
5.0 RESPONSESTO SECTm 5OUESTK)NS Pace 5.2 Crack Growth A!!owance (Ref.1. Pace 4)
The questions given here are essentially the same as given in the
- General Concerns" section of Reference 1. Please see Section 2.4 of this report for the related response.
Pace 5.3 RPC Characterization of ODSCC (Ref.1. Pace 5)
The proposed voltage limits for tube plugging apply to axial ODSCC with dominantly axial cracks. When assessments of the RPC measurements indicate that an indication is most probably not ODSCC, the voltage limits do not apply. Some judgment remains in defining whether or not an indication is ODSCC. Discussion of RPC interpretation for ODSCC is also included in the data analysis guidelines of Reference 3, Appendix A. The history of the 14 tube / TSP intersections removed from Farley has shown ODSCC as the degradation mechanism. The RPC data review requirements assure a continuing review for :he potential onset of new degradation mechanisms and would exclude other probable mechanisms such as cold leg thinning from application at the plugging limits. If non ODSCC indications are found, the indications are evaluated for tube plugging at the current Tech. Spec. limits of 40% depth. Further procedures to reconcile or characterize non-ODSCC would be dependent on the state of the-art of NDE technologies such as UT. The need for characterization would depend on the frequency of unexpected non ODSCC indications and would be assessed as the need arises.
Paae 5.4 Bobbin Coil Crack Deoths (Ref.1. Pace 51 Bobbin coil depths are no longer required for the revised voltage plugging limits of WCAP-12871, Rev.1. The revised criteria specify voltage limits independent of depth.
Thus the Reference 1 questions are no longer applicable. However, with the prior voltage and depth limits, it was intended that when depth could not be called for distorted signals, the lower 3.0 volt limit would be applied independent of depth.
5-1
6.0 RESPG JSESTO SECTION 6OUESDlS Pace 6.2 Leak Burst Pressure Estimated for Pulled Tubes Incfodino Farlev-2.
R31046 (Ref 1. Pace 5)
The burst pressures of steam generator tubes with axlal cracks can be calculated with good accuracy from measured crack profiles. The strengthening effects of ligaments beneath partial depth cracks and of ligaments between adjacent coplanar through wall cracks is included. For through wall cracks, normalized burst pressure, p, is plotted versus normalized crack length, A,in Figure 61, where
' ~
a,c I
Measurement of through wallcrack length and use of Figure 6-1 permits calculation of the burst pressure of a cracked tube.
For par 1ial through wall cracks, the burst pressure is given by
[ Ja c 61
l where [
la.c. Figure 6 2 shows the good agreement of this equation with burst test results on tubes with partial depth EDM slots.
For ligaments between adjacent coplanar through wall cracks, [
Ja c. Figure 6 3 illustrates the success of this procedure, The preceding paragraphs !!!ustrate how burst pressures can be calculated from measurements of crack lengths and depths. Table 6.1 lists the pulled tubes for which burst pressures were estimated. All estimated burst pressures were higher than 3AP at operating temperature. In the absence of crack length measurements, the crack length was conservatively taken as equal to the full tube support plate thickness of 0.75 inches.
No leakage is to be expected from these tubes up to a steam line break pressure differential since in all but one case the maximum crack depths were not through wall.
In the one case of Farley-2 Tube R31046 with through wall penetration, the through wall crack length was very short (0.02 inch) and experience with leak rate measurements has shown that very short through wall cracks exhibit negligible leakage.
Pace 6.8 Conversion of Polted Tube Vohaces (Ref.1. Pace 5)
Voltage conversions for Plant B 1 (3/4 inch diameter tubing) were obtained by using a normalization to 4.0 volts for the ASME 20% hole in the 550 kHz channel and evaluating the data using a 550/100 kHz mix. Westinghouse, under EPRI sponsorship,is further evaluating alternate voltage normalizations for 3/4 inch tubing to compare with the normalization adopted in Reference 3 for 7/8 inch tubing. Comparisons of responses to holes and EDM slots as well as burst correlations will be compared to adopt a 3/4 inch tubing voltage normalization for use in alternate plugging criteria. Until this study is completed,3/4 inch tubing tests are not used in the voltage! burst correlation for 7/8 inch tubing (Farley application). The plant B-1 data of Table 6-2 is considered in SLB leakage considerations as this tube represents the lowest voltage tube found to have a '
6-2
through wall crack. Uncertainties in the voltage normalization for the tube have a negligible influence on the SLB leakage correlation.
The field voltages for plant E in Table 6.3 (Page 6-8) were obtained by normalizing to I
Ja c as applied in Table 6.5. The adjustment f actor is smaller for partial depth indications. Further details of the conversion factors is included in Section 6.6 of Reference 3.
6 gg
)
6-3
Table 6.1 Pulled Tubes for Which Burst Pressures Estimated to Exceed 3AP Maximum PJand Iubg Location Deoth (%)*
PlateC-2 R15C36 1 26 5 76 8 20 Plant C 2 R26C56 2 60 Plant B-1 R4C61 8 35 11 69 12 30 Plant B 2 R6C67 2 57 5 36 8 60 11 35 Plant B-2 RBC57 2 65 5 20 Plant B-2 R4C68 2 50 5 57 8 40 12 47 Plant D-2 RibC77 1 56 2 53 3 35 4 41 5 47 Plant D-2 R11C25 1 32 Plant D 2 R6C40 1 25 Plant D-2 R12C42 3 21 Farley 2 R31C46 1 100 Farley 1 R20026 1 62 Plant M-2 R29C46 1 26
- Depth from destructive examination.
64
Figure G 1
_ a.b 1
h h
6-5
S b
Figure 6-2
. a,I 40 e
b
'h g l
_ /
/
s 6-6
1 Figure 6 3 a,b e
N N
6-7 l
7.0 RESPONSESTOSECTION 8 OUESTIONS Pace B.4 Laboratory Tests of Dented Tubes (Ref.1. Pace Si The purposes of the dented tube tests were to assess the influence of denting on leakage from cracked tubes and to determine the forces required to displace the TSP frc 3 dented intersection. The test results represent background information for considerations of undetected cracking or eddy current uncertainties at dented TSPs. The results are not used in the burst pressure correlation applied to establish tube plugging limits.
The denting tube tests show that even 0.7 inch through wall fatigue cracks do not show significant leakage under SLB pressure differentials. Fatigue cracks were included in the tests as the free span leak rates are predictable and large compared to carrosioq cracks. Voltages are higher for fatigue cracks than corrosion cracks but the test objective was leakage rather than voltage related. The test results and supporting analyses show that even incipient denting can prevent TSP displacement relative to cracks within the TSPs and thus prevent tube burst even under accident conditions. Thus the consequences of EC uncertainties at danted intersections are sn.all relative to tube integrity considerations for leakage and burst.
4 Pace 8.5 Bobbin Coll Detectability of Cradi.e-a > eu w .;r...
,. , , , , , n. * * '% + e' ' w- 8 4 0 3
.s
'84-t,p ;e,; -i - i v -. s __ . , _ _ , . , . ,_
- t. ..e i- e s
..1 2
~
y-
.~
'M - 1 . ,
f
- on ,,.
- -~~~4 >
.e n_
et
- \
P.tist.JLIGLhl s, .. n H r:
m . as N;. 'y e. . .a t. , . . . .: ..%.
.it :. 7. No.y% o 6 ,te . e is ownei W '&L &i
. . . , w , u i . 4 i . n i . . . . . . .. ..,,,,,. i . .
ut ne ut tre r<e t t 'n 'n n se et r i *ut. ni
= < ,+ ., . a.i .t .t..v..e. ~naeen i ., n . . ..
i . o. u n., y i ., , , ..
_ . , _ . . -m-. -.. - , - .s . -.. _
I i
i en...
~ -
s, i.,
I- '
i 4 ea i~a. .
a at. :w, ..a t ro t . .
il i f f' i s
,t.., o. . > i u, - . .
i e : .~ .
A su ou con r i
T * => *a - e -
e a o., - s, w ,,,
c &c
- a. r i u P
.x m r : n .c ier ve e ee iae 0 -
3, s.n. 5. * -
t... c. . t . , ii u ,n um, ,r m m.
test sw .we+
1 r ..ee . , .
- 2.u. , ,
< g s.n n t.se. -
!A H3 W4 35 M6 i.n . ( Vi i . es .
E4 H3 3 f,5
.u. 189 2/2 n_ - . -
um SS M9 5 ;t it Pe itSr'It..'.E? og t't tot 2.',e 4., - .irr . 23e3. +He w.- _ - . , _
Figure 71 Comparison of labor story (upper) and field (lower) ultrasonic inspection results with the rnetalographic (right) examination for S/G D R4 C73 at the first hot leg support intersection.
'4
l nq a p - -' s I -
v [q; {. lu{ i 1 j t. <'
,; y
- _; i i
- iM
,g ,}... . ss o u, - e; -
. t } e ,4 4.
n '.e ;.
' s<< , rt ?. -
' . '"' j
' }1,,,.....-
- . p : t ~s SUPPCf.T
.,,,, > : , , .- .e3 -+ m.te<io,n.,s i : ,., m o<:~ie..
p
, t . 1, f
.a s. sm .* f u en cet
( ' e' ,,'8*- ;*t*
'(*'((C .
- - *4 " ~ _._.7,
/ ;l , , '
i ^-
- . . .. 1
- A. ,l Hll,pl l'f , g,j ) -
L fl.
- a. gg 4 '.O tU
- '. ,'; aste af n ,n crut
- W If? '.$
. , , 82 i N' M ..<. w ii.,,.. . . e w im .
. is )i 1E
' P ,' ,, forma. i t t .," 6 it incse,
, 1 P -
d-H'8 Ht 309 fit ?'t if? fit 15t'?!9 59 ff M f ause * ** a e**:eeets
- e (t n o is .+ tn intem . ie.
'estures. e.ie vita e c t ue everie.4 vest.reo meesheim . toten.
tse 1 ( .itt cha artecientes si,aif
(.isth o' luusat e enotuwhos)
, 7 ,. ,, . , , e n . ~6. . v t e,
! c o 1, - > - ; ">
a s.t. , . _ , , , , o.t . . ,_. . \
t u.....,, , < . , . ~
g n e. i , w.. ** ***$ -
-S'**""
%,r v, u; -
t c n .:
o ise re e+ io* ise
,,im P .__ - . _ _ . . - - . _
?w m ': u 'o 3,,gg,ug emon,,
0 , ,,. .,,,tt..... ,,.a.w
- b. 4.84 1, , t t i .e - nt't' L
t .rs.
! 3.61.
r Q 3.50 9 c. . '- au 94 'W i.. i NE 181 bl2 i.rs.
I M ??.3 4H
. 6e. 17/ >?F i ut si 1?f e ?e it 9e t!t se ite tit 24? t?e ?tt 33e 26e
( am n, m. cv. u, n. .
Figure 7-2 Comparison of laboratory (upper) and field (lower) ultrasonic inspection results with the metallographie (right) examination for S/G B R21 C22 at the first hot leg support intersection, 9
1 1
Initial Scan i 91 43 i 1 JM the CN ) g (L 3 t bt $ 10f j ., ...,
.. i .. . i si.
' \ ( B ?t his 'fC { f(N 4 5
/ flP([D E X 'I ende.g k / 'U
/
1
- .e . It m 134 12e wee 4M m in its I i i i I
'y. y.g-w La L; ci a-. s %_
l ! f ( )
1 e. IM sna CM i *. . 19 3$ als a 4
1 P
)W 0 j&k> 4 f.
I l I
w eesses etens e es's : eso
. > . . m in .
. .. .o en .,. '.
- a a , ;* a --
i i i ! ,w .,,.;..:,v
% % W h %ini s;
% % xc- .
5 j i ; ,
i c- ;,
Repeat Scan au .= o.. Cn i u .. ,. o. g ,, g . g .
,/ ; a r ggs u, agn ,.
- g i,6 i4 N A
- /x i. i .i i tuttat! rec l tta i
'a g'5
/ y -
s.n . . . . ,
-3* m n. ... . . m ,, so i i i ! I i 3
- s y vm--g
- -~ ,s..
-*='=y=
i i i i i !
... ....o., .. ... ,,...i n k
? G.
> t
, G,:, o L
Q /i p ki i I I v .
. . . mu . i ,s m ., --a---
I i ! ; i .,..;, ,,...,.
_ _2 ,
N - -.
r x
.c Y'.-m,. ,;
j r 4i *!
i i l .
Figure 7-3. Repeatability of Field Voltage Measurements
8.0 RESPONSESTOSECTION 9OUESTO3 Pace 9.3 Influence of Temnerature and Procerties on Burst Correlation IRef.1.
Pace 61 M the relationship between burst pressure and bobbin coil voltage the correction for material property strength reduction at operating temperature is made by scaling the
" lower bound" curve by the ratio of the lower tolerance limit (LTL) strength properties o' Alloy 600,7/8 x 0.050 inch, mill annealed tubing at 650*F divided by the strength at room teilperature of the tubing material tested. The strength properties utilized are the sum of the yield and ultimate strengths. The LTL strength (yield + ultimate) of the tubing material at 650*F is 126.0 ksi and the room temperature strength of the i material tested is 147.0 ksi giving a ratio of 0.857.
The tube dimensional variations are considered to be one of the reasons for the variability in measured burst pressures and are therefore consicered in the " lower bound" of the data. The dimensional value that affects burst pressure is R/t and the variation in R/t is expected to be very small.
g.
8-1
l:
9.0 RESPONSESTO SECT!ON 12OUESTIONS i
Pace 12 2 Aeolic.nhility of Criteria to Units 1 and 2 (Ref 1. Pace 6)
For the plugging criteria of References 2 and 3, it is correct that, conservatively x credit is taken for tube reinforcement by the TSP for both F,nty Unit-1 and Ur w
. This assumption of free span ignores the known reinfor.; ament under operating conditions and the demonstrated reinforcement for Unit-1 under accident conditions.
Thus establishing the structurallimit for tube burst at three times normal operating conditions is very conservative for both units due to reinforcement by the TSPs. Under 4:
accident conditions, the analyses to demonstrate no TSP displacement was completed only for Unit-1 and TSP reinforcement is thus applicable for this unit. While it appears to be feasible to demonstrate no TSP displacement for Unit 2 under accident cond8tions, the supporting evaluation has not been perforrud.
The conservative choice applied for the plugging criteria was to assume free span tube conditions and to establish the same tube plugging criteria for both Unit 1 and Unit 2.
Pace 12 4 Throuahwall Crack Assumotion for SLB Leak Rates (Ref.1. Pace 6)
All cracks not exceeding the voltage plugging limit are conservatively assumed to be through wall for purposes of estimating the SLB teak rate. A lower bound voltage level for leakage can be estimated but through wall cracks are currently assumed.
Pace 12.4 Probability for Tube Burst Under SLB Loadino (Ref.1 Oace 6)
The estimated probability for tube burst under SLB loading given uncertainties on NDE, growth and the voltage / burst correlation is given in Section 12.41 of Reference 3. This estimate utilizes the cumulative probability from historical growth rate data and the lower confidence levels associated with the voltage to burst pressure correlation.
Pace 12.6 Bobbin Coil insoection Aeouirements (Ref.1. Pace 6 7)
Farley S/G inspections, upon implementation of the attemate plugging criteria will perform 100% bobbin coil inspections of all hot leg TSP intersections to at least the lowest cold leg TSP with ODSCC indications found in any prior inspection. If a sample 91
t
. Inspection is performed below the lowest TSP with indications and an ODSCC indication is '
found at the lower TSPs, either the inspection will be expanded to meet 100% inspection -
to the new lowest TSP or the alternate plugging limits will not be appiled below the lowest TSP with 100% inspection. Decisions on sampling plans or expansion to 100%
are expected to be outage dependent. In either case, the voltage plugging limits will not be applied at any TSP that does not have a 100% inspection of active tubes.
- Paan 12.7 ' Circumferential Branchino of ODSCC Cracks (Ref.1.' Pane 8)
The descriptions of circuo.5::ential branching of ODSCC cracks found in model boiler specimens is expanded in Section 12.6 of Reference 3. Schematics of OD crack maps and photographs of the burst cracks are included in Reference 3. Only tubes with voltage .
levels well above the tube plugging limits show circumferentialinvolvement in the burst crack. Thus circumferential branching is not expected to be a concem for the tube ^
plugging criteria. On this basis, RPC resolution is considered to be adequate for detection of potential circumferential cracks or branching.
t 4
, 9-2
^
l
't
10.0 REFERENCES
4
- 1. USNRC Docket Nos 50 L,' .0-364, letter ' rom S. T. Hoffman to W. C. Hairston, August 8,1991," Request for Ado?tionalinformation Concerning Steam Generator Tube Support Plate Alternate Plugping Criteria for Joseph M. Farley Nuclear Plants, Units 1 and 2", (TAC Nos. 79818 and 79819).
- 2. WCAP-12871,"J. M. Farley Units 1 and 2 SG Tube Plugging Criteria or ODSCC at Tube Support Plates", February 1991.
- 3. WCAP-12871, Revision 1,"J. M. Farley Units 1 and 2 S/G Tube Plugging Criteria for OSDCC at Tube Support Plates", September 1991.
- 4. NUREG-0491, " Safety Evaluation Report By the Office of Nuclear Reactor Regulation, U.S. Nuclear Regulatory Commission, in the Matter of Westinghouse Electric Corporation Reference Safety Analysis Report", RESAR-414, Docket No.
STN 50-522, November,1978.
- 5. ASTM-STP-722. "Multifrequency Eddy Current Method and the Separation of Test Specimen Variables", Eddy Current Characterization of Materials and Structures.
Amrit Sagar,1981, pp. 269-297.
10 - 1 i
,u-,e, s. ,n 4 + a-.as.
5 APPENDIX A REQUEST FOR ADDITIONAL INFORMATION A-1
O ese
.M % UNITED staff 8 l ,e i (y, h NUCLEAR REGULATORY COMMIS$10N cAsmatoN.o.c sm L #
y ...** August Be 1991 Docket Nos. 50-348 and 50-364 l
l Mr. W. G. Hairston, !!!
Senior Vice President Alabama Power Company 40 Inverness Center Parkway Post Office Box 1295 -
Birmingham, Alabama 35201
Dear Mr. Hairston:
L
SUBJECT:
REQUEST FOR ADDITIONAL INFORMATION CONCERNING STEAM GENERATOR TU
$UPPORT PLATE ALTERNATE TUBE PLUGGING CRITERIA FOR JOSEPH M. FARLEY l NUCLEAR PLANT, UNITS 1 AND 2 (TAC NOS. 79618 AND 79819) l l By your letter dated February 26. 1991, you submitted an amendment request for to allow the use of an Joseph M. Farley Nuclear Plant (Farley), Units 1 and 2,ithin the boundariss of alternate plugging criteria for steam generator tubes w the tube support plate. Your request was supported by Westinghouse topical reports entitled "J.M. Farley Units 1 and 2 SG Tubo Plugging Criteria for ODSCC at Tube Support Plates" (WCAP.12871, Proprietary and NCAP.12872, Non-proprietary).
1 Your submittal is being reviewed for us by our contractors Pacific Northwest Laboratory and Oak Ridge National Laboratory. Their review has identified a need for additional infortaation which is described in the Enclosure. We have reviewed and concur with the enclosed request for additional information.
Please provide a response to the general concerns and specific requests contained in the Enclosure. Your response is requested within 30 days of receipt of this letter so that our review of your submittal can continue.
l t The reporting and/or recordkeeping requirements contained in this letter l' affect fewer than ten respondents; therefore, OMB clearance is net required under P.L.96-511.
Sincerely, AA u-1 Stephen T. Hoffman, Project Manager Project Directorate II 1 l
Division of Reactor Projects - !/II i
Office of Nuclear Reactor Regulation cc: See next page l
A2
~
e
W, B a tt e ll e
~
l Pacific Nrf thwnt Laboratorin Sanete lowles ard P.O Boe 999 aicMand wsAiegioa t9352 Tecomm 375 3812 August 2, 3971 Mr. Emmett L. Murphy Mail Stop 7 D4 U. S. Nuclear Regulatory Com.ission Washington, D.C. 20555
Subject:
FIN L1855, Task Assignment 91-02, Review of $tsam Generator Tube Support Plate Alternate Plugging Criteria Proposal for Farley 1/2.
Daar Er. matt:
In accordance with the Statement of Work for Task Assignment 91-02 of the subject As FIN we have reviewed anc evaluated Alabama Power Company's TSP APC proposal.
a result of Our review we have enclosed a re::uest for additional information.
If you have questions or comments about the enclosed material please give me a call on (509) 375-3812.
Best Regards, Richard J Xur 2 Automation & Measurement Sciences Department re.v a.,e,<-, w .m a ew w A-3 . ..... .. .. ..
REQUEST FOR ADDITIONAL INFORMATION r Prepared by q Pacific Northwest Laboratory and Oak Ridge National Laboratory.
REVIEW 0F STEAM GENERATOR TUBE SUPPORT .
PLATE ALTERNATE PLUGGING CRITERIA FOR FARLEY 1/2 Inteeductien A review and evaluation of Alabama Power Company's proposed alternate plugging region of the Farley 1 and.2 criteria (APC) for the tube support plate (TSPsteam generators was cond Power Company's: letter to the NRC dated February 26, 1991 and a-Westinghouse l .
report (WCAP-12871) which presented detailed technical information.on the-development of the APC. The concerns and questions given below include those '
from Pacific' Northwest Laboratory staff along with additional input from a review performed by Dr. C. V.~ Dodd.cf Oak Ridge. National Laboratery.
.Under the . licensee's.aroposal, the APC would only apply to outa diameter stress corrosion crac(s (0DSCC) which are contained within i,he thickness of l the . TSP. All other portions-of the tube would be covered by the current 4%
L plugging criteria. The APC utilites the eddy current. bobbin coil voltage and l phase angle indicated'deptl. as an. indication of crtak severity within the TSP.
Tubes with an indicated depth greater than er. equal' to 53 will be plugged if the eddy current bobbin-volta e exceeds three volts. Tubes with an indicated depth less than 53 will be p ugged if the eddy current' bobbin voltage exceeds 4
eight' volts. All tubes with efect voltages greater than 1.5 volts will be inspected with-a rotating pancake coil (RPC) to determine if 005CC is the main L cause of-the signal.
-feeBePRI CBScorns From our initial review of the-proposed APC and supporting technical documentation we have a number of general concerns which should 'be highlighted. ,
F lt is of particular' concern that the burst and leak test database is very small . Only 13 data points plotted in Figure 9.1 are from actual burst tests.
Similarly, only 10 data points comprise the data set for the leak rate signal .
amplitude correlation shown in Figure 9.3. In addition, very few of the burst and leak tests are for pulled tubes with service induced degradation from the g
Farley steam generators. The-pulled tube database was expanded by including I tubes from other plants, but- this must be dont cautiously since these tubes may not be representative of the conditions present in the Farley steam generators. Thus,:it is important to know the identity of all the plants included in the pulled tube database. It is recognized that Westinghouse 1
l A-4 j
l
n attempted to sup]1ement the pulled tube database with laboratory degraded specimens, but tiese specimens, in some cases, were not comaletely typical of pulled tubes because tie bobbin coil voltages were higher tian these-typically encountered in plant eddy current examinations and crack morphologies were different when the tubes were highly stressed during flaw fabrication, in summary, we are concerned that'the relationships of signal voltage i with from f data burst and leaktat m pulledtubes.ightnotbeconservativegivenalargercollectono Another concern involves the development of the NDE measurement uncertainty
' allo,<ances presented in Section 8 of the re) ort. Table-8.6 does not appear to
- include all of the significant factors whici might affect NDE measurement-accuracy. For example, the wide variation in voltages with denting for the doped steam samples presented ~1n Table 8.4 conflicts with the conclusion The that denting does not contribute significantly to measurement uncertainty.
signal amplitude of two out of the three samples changed significantly, with one-of the samples decreasing by a factor of four, anc the other increasing by-a factor of two.
It appears from the report that many of the evaluations of NDE measurement uncertainty utilized specimens with idealized defects such as EDM notches or drilled holes. Based on our experience the NDI measurement uncertainty increases substantially when realistic defects are examined under realistic conditions. 'The presence of multiple cracks. deposits, tube support corrosion and denting must all be considered in realistic combinations to obtain In addition, the occurrence of reasonable estimates of the measurement error.
multiple of defects such as intergranular attack (!GA) and circumferential StC in combination with.0DSCC will also have a significant effect on the voltage measured by a differential bobbin coil. It should be noted that and recently, circumferential SCC has been fcand at the TSP in North Anna, Trojan. Both of significant circumferential ]GA was discovered at the TSP in these plants may be good predictors of future tube degradation at Farley
- because they have had histories of degradation similar to Farley.- As mentioned in the report destructive metallographic analyses of recently pulled Farley tubes indicates these other defect types are beginning to occur in the Farley steam generators, but these factors are not included in the assessment of the reasurement uncertainty.
Signal voltage will also be a strong function of the coil to coil spacing in and calibration to the ASME Section XI standard will probably not remove all o the influence of this factor. A study of the differential bobbin probe,f
-different probes of the same probe type from the same manufacturer should be
. conducted.
Perhaps the most significant source of measurement uncertainty that was not addressed in the report is the human factor. Repeatability of the voltage readings on actual field defects using the present data analyst guidelines is an important question. There appears to be a certain amount of judgement
- involved in determination of signal amplitude for a particular flaw. A report by M. J. Gallagher at the Ninth /,nnual EPRI Steam Generator Workshop notes that reported voltages for identical signals varied by a factor of six depending on the measurement methods employed. The variability of the t
A-5 -
4
val %:ges be% ween inspections shown ja Figures 5.3 aad 5.6 inGicate % hat voltage readings on real steam generator defects may not be too accurate.
- 'i
. art aisc concerned that edetien of the APC may allow more deep through-wall IGA or circumferential SCC in the vicinity of the TSP to remain in service than would be otherwise. This is because the APC would accept flaws of g.y depth with signal amplitudes less than three volts and inspection with an RPC probe would not be required unless the signal amplitude exceeds 1.5 volts. Both IGA and circui.farantial SCC type flaws generally give low voltage signals, which the differential bobbin coil probe does not detect reliably unless the flaws are quite deep. Thus, dees 1GA or circumferential SCC zay be misclassified as acceptable ODSCC because the signal amplitude is too low to
- trigger an RPC inspection. We believe the increasing occurrences of TSP IGA and circumferential SCC may indicate that thess damage mechanisms are the next step in tube degradation at the TSP. Recent experience at Trojan and North Anna t mest this may be the case. Thus, it seems prudent in the development of an
- 4ection program and APC to consider these possible modes of cegradation.
The last majer area of general concern involves the flaw growth allowance. It is very difficult to determine from the discussion in Section 5 how the flaw growth allowance is calculated. A number is given in Table 12.2 which appears to be based on average flaw growth rates (the 0.5 veh growth allowance .is aither an average value or a somewhat conservative value depending on which unit is being considered). The voltage change data in Section 5 is presented as representing flaw growth from one inspection to the next. The analysis does not censicer the influence of NDE measurement uncertainty or any of the other f actors noted above which could contribute to the observed voltage changes. The Westinghouse reocrt does not discuss factors that contribute to
- 00 SCC and crack growth at the' TSP locations. Such a discussion is needed te better assess the tilowances for crack growth between inspections. Nothing is said about the levels and distributions cf the denting induced stresses that drive the crack growth. There is no discussion of how stresses decay outside the boundaries of the support plate, and thus limit the amount of potential crack growth. There is no mention of industry operating experience on the maximum observed '2ngths of cracks at TSP locations. Given the significant uncertainties it appears the flaw growth data is not conservatively bounced.
Recuest fer Additieesi infomatien The following list of questions resulted from the review of W;AP-12871. We believe the information requested is needed to complete a thorough evaluation of the proposed APC.
PAGE 2.2 It is stated that all tubes with bobbin coil indications >1.5 volts at TSP intersections shall be reinspected using RPC probes and that the RPC results shall be evaluated to support ODSCC as the dominant degradation mechanism. Are the bobbin coil measurements differential or both differential and absolute? How '
specifically will the RPC data be evaluated to determine if the i
3 A-6
indications are in fact due to CDSCC7 Will the RPC data be used to estimate crack length and to determine if the cracking has progressed beyond the confines of the TSP 7 !s it assumed that all bobbin coil indications of 1.5 volts or less are short cracks? If so, what data supports this conclusion.
PAGE 4.1 Have any profilometry measurements been taken to quantify and characterize the extent of denting in the Farley steam generat. ors?
If so, to what extent does dent severity correlate with the occurrence of indications? Both dent voltages and a corresponding a diameter change should be furnished for the tubes in the affected region.
PAGE 4.2 Areas of circumferential htergranular corrosion were observed on Tube R4C61 which had a mr 91ogy more like IGA than ODSCC. What was the amplitude of tN 91s from these flaws? Please provide sketches of t'2 s ;. ;tribution for Tubes R4C61 and R3)C46 similar to the t.
- vided in Section 4 of the report. Also please e e< c 'cetions of the recent discovery of deep citew 2 . 4' : +,ing at the TSP in Trojan on the application of the / :p .
It was noted on page 4.7 tw e 7. b Steam Generator C, PAGE 4.7 9 iSP. Also, for one of Tube R2CC26 exhibited crackir4 the model boiler samples (Spec 3 men C61. esp 514) considerable crack growth beyond the support plate t>ov.daries was observed.
Operating experience at North Anna-1 has also shown crack growth beyond the support plate. What is the technical justification for assumina that ODSCC cracks will not extend beyond the boundaries cftheISP7 Have fracture mechanics calculations of crack growth at TSP locations been aerfomed? Do predict that the lengtiwise growth of(or ODSCC cracks would) decrease or such calc stop once they extend beyond the TSP 7 PAGE 4.11 From the information given in the report it would appear that the effect of positioning the flaw near the edge of the TSP was not investigated. What would be the effect of signal distortion on detection and sizing of ODSCC when the cracks are located near the top or bottom of the TSP 7 PAGE 5.2 Section 5 presents change in signal amplituda information which is used in Section 12 as a flaw growth allowance for datermining the tube plugging limit. Table 12.2 indicates that the flaw growth allowance is 0.5 volts. This allowance appears to ha based on the average voltage change plus one standard deviation on the data as mentioned at the top of page 5.2. Please present a clearer description of how the flaw growth allowance is deterr.ined and a technical justification for why a more conservative allowance is not used, such as two standard deviations on the data. The description of the flaw growth allowance determination should also l
talude a description of how NDE measurement uncertainty is treated in the flaw growth-signal a p11tude relationship.
4
' A-7
' PACE 5.3 On page 5.3 it is stated that '... characterization of the degradation'at-these locations with the rotating pancake coil L
probe (RPC) often clarifies'the dimensions and distribution of the-major crack features.* What procedure will be used to reconcile bobbin coil indications if they cannot be clarified by RPC inspection?
PAGE 5.4 The data presented in Table 5.1 shows that all, except one of the bobbin signals from the 1989-inspection of the Farley 1 steam ,
generators were distorted. How can flew deaths be reliably.
determined from distorted signalst Since 15e proposed tube plugging criteria depend on measurement of flaw depth, what is the' NDE measurement uncertainty for this phrameter? J PAGE 6.2 Referring to the pulled tube database. 'it was stated that crack morphologies for eleven tubes having no leak or burst test measurements were reviewed to determine their potential for leakage and to estimate their burst pressure. How was the review performed to estimate the burst pressure and reach the conclusion that no leakage would be expected for these tubest Also please identify the specific plants contributing infarmation to the pulled tube database.
PAGE 6.2 How were the leak and' burst pressure information estimated for the Farley 2 tube with the 7.2 volt indication?
PAGE-6.8 In Table 6.3 the data from a number of foreign tube pulls are used to- supplement the data from Farley pulled tu5es. A conversion of the foreign voltages to the Farley voltages is made.- Please provida details of how this conversion was performed.
PAGE 8.4 Six fatigue crack and three doped steam corrosion crack samples were leak tested to evaluate the influence of the dented TSP crevice condition. These specimen types were declared to be ;
unrealistic in that they gave bobbin coil voltages that were high 1 relative to field cracks. What-is the technical justification for !
including data from.these samples in the database?
PAGE 8.5 It is stated that bobbin coil detection of cracks at dented TSP intersections is unreliable. Please present information on the probability of detection as a function of crack size for CDSCC at dented TSPs.
PAGE 8.6 Have any other NDE methods been used to evaluate the ODSCC in Farley 1 and 27 If so what were the results and conclusions of these evaluations?
PAGE 8.7 It was stated in the report that proper implementation of the tube -
plugging criteria depends on data acquisition and data analysis in the field and in the laboratory being the same. In order to evaluate the proposed plugging criteria a copy of the calibration procedure and data analyst guidelines that will be used in the 5
A-8
field is needeG. Hoa repeatable are the voltage readiags on actual field defects using the present data analyst guidelines?
Please provide copies of the eddy current scans used to size the defects.
PABE 8.17 As shown on page 3.17 the signal amplitude does not change with increasing slot le' .h for part through wall slots greater than about 0.25 inch. ace the signal amplitude is not sensitive to increasing crack length for part through wall fisws how will crackgrowthbeyondtheboundariesofthesupiertplatebe
' monitored? In order to thoroughly evaluate tSese data readable copies of Figures B 2 and 8 3, along with the referenced phase shift figures are needed.
PAGE 9.3 In the relationship between burst pressure and bobbin coil voltage, how were corrections for the influence of temperature on tube material properties made? In acidition, were the tsffects of tube dimensional variations considered in dettreining the lower bound relationship?
PABE 12.2 Based on the report it appears that no credit for tube rainforcement by the TSP was taken for deriving the plugging limit. Please confirm if this impression is correct for both Farley-1 and 2. In other words, are OD$CC cracks at TSPs always treated as ' free span
- for both normal operation and accident conditions? Are there any differences in the actual plugging criteria for Unit i vsrsus Unit 2 -(i.e., reflecting the conclusion that TSP displacements can occur for Unit 1 but not for Unit 2)?
PAGE 12.4 As depth measurements of 005CC cracks seem to be unreliable are all crack indications not exceeding the APC considered to be through wall for purposes of estimating the SLB leak rate?
PAGE 12.4 The report uses a conservative bounding approach to demonstrate carains relative to tube burst under $LB loadina. However, it is difficult to judge if these calculations are sufficient to demonstrato no possibility of tube rupture given the existence of the large number of degraded tubes in each steam generator and the fact that large increases in degradation (voltage increases) have been observed between inspections. Furthermore, the data supporting the correlation of burst pressures with voltage are sparse and show a large amount of scatter about a mean curve.
What is the estimated conditional probability for tube burst under
~
SLB loadina given the uncertainties inherent in the derivation of thetubep$uggingcriteria? Please provide the technical basis for this estimate.
PAGE 12.6 The report states that 100% inspection is required for all het leg TSP intersections and for cold leg intersections down to at least the lowest TSP with ODSCC indications. Please describe how the extent of inspection is determined. Are inspection results from the previous ISI used to determine the extent of inspection on the 6
' {
A-9
[
cold leg side or is a sample of tubes inspested full length to make this determination? If the latter is the basis, then that is the initial sample site for this full length inspection?
PAGE 12.7 It is stated in Section 12.8 that circumferential branching of f
' ODSCC cracks is acceptable. How much circumferentini crack growth is allowed before the alternate plugging criteria would no loncer apply? How will reliable detection of an unacceptable level c?
circumferential branching be ensured?
l i ,
l l
l l
l l
l l
l l
l l
[
W 4
7 1 A 10
_ _ _ _ _ _ - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ . _ _ _ _ _ . _ _ - - - - _ _ _ _ _ _ _ . _ _ _ _ . _ _ _ _ _ _ _ _ . _ _ _ _ _ . _ _ _ _ _ _ _ _ - _ . - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _