ML20086A849
| ML20086A849 | |
| Person / Time | |
|---|---|
| Site: | Catawba  |
| Issue date: | 06/26/1995 |
| From: | Rehn D DUKE POWER CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 9507050060 | |
| Download: ML20086A849 (4) | |
Text
U
'l
' Q-DukeIbwerCompny
. D. L Rem.
Catawbo Nuclear Generanon Department VicePresident i
4800ConcordRoad (803)8314205 OMce York, SC29NS (803)831-3426 Fax OUKEPOWER June 26,1995 4
U. S. Nuclear Regulatory Commission Attention: Document Control Desk Washington, D. C. 20555
Subject:
Catawba Nuclear Station, Unit 1 Docket No. 50413 Technical Specification Amendment No. 130/124 for FOL NPF-35/NPF-52 Renewal Of Steam Generator Tube Interim Plugging Criteria for Unit 1 Cycle 9 Report on "630" vs. "610" Bobbin Probe Comparison and Analyst Error Assessment i
Reference:
NRC Issuance of subject Amendment Dated March 17,1995 l
4 De following is a Summary Report on the detailed assessment of the 630 vs. 610 Bobbin Probe comparison and analyst error, as committed to in the subject Technical Specification Amendment.
f 630 vs. 610 BOBBIN PROBE COMPARISON ASSESSMENT he application of the Interim Plugging Criteria (IPC) at Catawba Unit 1 involves 100% Eddy Current Test (ECI) of the tube bundle and plugging of >1.0 volt indications which are confirmed by Rotating Pancake Coil (RPC). Plugging of >2.7 volt bobbin indications is performed regardless of RPC inspection results. The IPC requirements for 3/4 inch tubes e
necessitate 0.610" diameter bobbin probes. We inspection was performed principally with 0.630" diameter bobbin probes. Based on comparison of bobbin voltages for a sample of indications inspected with both 0.610" and 0.630" diameter probes, a consensive factor was developed to adjust the 0.630" probe volts to equivalent 0.610" probe volts. Along with this feld ac tivity, an analysis is performed to predict the tube leakage rate and burst probability for a postulated SLB accident at the end of the next cycle of operation.
i l
t 9507050060 950626
\\
l 00r PDR ADDCK 05000413 P
PDR i '
Q new w mn
i s
Nuclear Regulatory Commission June 26,1995 Page 2 j
The EOC-8 tube inspection was performed with a 0.630" diameter bobbin probe (630 probe) in all hot and cold leg TSPs where IPC was applied. To allow evaluation compatible with the IPC database, the inspection results were converted to equivalent 0.610" probe (610 probe) data, on the basis of test and analysis of the two probe sizes; a previous analysis of the suitability of the conversion of 630 probe data to 610 probe equivalence has been previously submitted to the NRC. Additional testing and analysis was performed using the Catawba 1 inspection results. After review of the results of these analyses it can not be stated with certainty that the use of the 630 probes is a direct equivalent to the 610 probes. Due to the limited amount of data available for the comparison, it would not be appropriate to make this conclusion. This review confirms that it was entirely appropriate to use the 630 probes in conjunction with criteria developed for the conversion to the 610 probes. The use of the 630 probe versus the 610 probe would not be expected to contribute to a significar.t difference in the performance of the SGs during normal operation or a postulated SLB event relative to the total expected leak rate fmm TSP ODSCC indications or the probability of burst / overpressure of such indications.
ANALYST ERROR ASSESSMFRT The NRC initially expressed concerns that the data transmitted previously via Reference 1, implied a relatively large analyst team to analyst team uncertainty.
An effort was made to perform a more meaningful study by selecting 49 more indications at random to analyze. For this study the indications were divided in two groups and analyzed by two NDE teams. The first team, Team 1, evaluated the first 25 indications using the 610 probe and the other 24 indications using the 630 probe. Team 2 analyzed the first 25 indications using the 630 probe and the other 24 indications using the GIO probe. The rationale, at the time, for not letting both teams evaluate all of the indications with bcth probes was to ensure that all of the data was from a " blind" examination. Comparisons were made between the probes and analyst teams, and between results from the experts and the analyst teams. The initial implications from the evaluation were sinular to those in the data of Reference 1. A reexamination of the data revealed that only 5 of the first 25 and 6 of the last 24 indications were RPC confirmed. In addition, most of the amplitudes were significantly less than 1 volt. In order to try to get significant information from the test program, further analysis was restricted to only those indications that were RPC confirmed. A plot of the 610 probe volts versus the 630 probe volts was performed and a comparison was made of the Team estimated voltages to those from the expert analyst.
l
~.
4 Nuc1 car Regulatory Commission I
June 26,1995 i h the exception of two Page 3 In both of these cases a reasonable correlation is shown, w t 630 probe.
data pairs, which were overestimated by th based on the expert h
evaluation of the voltages was performed. T e pro that of allof the data.
here to assess the d
Because of the small sample size, no attempt is ma ei ted with th aignificance of the NDE analyst uncertainty assoc at determination, the statistics when the teams are significantly over called, relative to the experthe 630 eadings. Team i significantly b
compared to the expert using all of the 630 pro e rti n one indication u under called, relative to the expert determ na o,the teams are compare i
h probe. This value dominates the statistics w enthe dominant values, th 0.5% with a standard expert using all of the 610 probe data. Without difference between the teams' readings and Overall, there does
)
l not appear to be any significance associated ana ys use of the 630 probe data for the IPC analysis.
l i ns may be made Based on the evaluation of the data, the following conc us od h
relative to the use of the 630 probe for t e eCata DSCC indications:
f The results of the initial analysis of the data that th i
the 630 probes in conjunction with the con 1) ted adjustment factor) instead of the 610 probe wou f mance of the SGs contribute to a significant difference in the i
t the total i
expected leak rate from TSP ODSCC ind fh dj sted b
asurements 630 probe measurements instead of actual 610 pro e m would be expected to result in overestimates o t eEnd of Cycle the PoB during a postulated SLB event at the The nondestructive examination error associat of the 630 probe appears to be significa 2) litudes ofless than 1 volt (by both probes) and which were not
c,..
Nuclear Regulatory Commission f
June 26,1995 Page 3 In both of these cases a reasonable correlation is shown, with the exception of two
.l data pairs, which were overestimated by the team using the 630 probe.
Finally, a comparison of the 610 probe to 630 probe voltages based on the expert evaluation of the voltages was performed. The probe to probe scatter is similar to that of all of the data.
Because of the small sample size, no attempt is made here to assess the significance of the NDE analyst uncertainty associated with this test data. Team 2 significantly over called, relative to the expert determination, two indications with the 630 probe. These two data points dominate the statistics when the teams are compared to the expert using all of the 630 probe readings. Team i significantly '
under called, relative to the expert detennination, one indication using the 610 probe. This value dominates the statistics when the teams are compared to the expert using all of the 610 probe data. Without the dominant values, the average difference between the teams' readings and the expert was 0.5% with a standard i
deviation of 3.7% (using the combined data for both probes). Overall, there does not appear to be any significance associated analyst variability to invalidate the use of the 630 probe data for the IPC analysis.
l Based on the evaluation of the data, the following conclusions may be made 3
relative to the use of the 630 probe for the eddy current examination of the Catawba 1 tubes at the TSP elevations for the disposition of ODSCC indications:
- 1).The results of the initial analysis of the data that the use of the 630 probes in conjunction with the conversion criteria is appropriate.
The use of the 630 prche (in conjunction with the data supported adjustment factor) instead of the 610 probe would not be expected to contribute to a significant difference in the performance of the SGs during normal operation or a postulated SLB event relative to the total expected leak rate from TSP ODSCC indications or the probability of burst / overpressure of such indications. In fact, the use of the adjusted
=
630 probe measurements instead of actual 610 probe measurements would be expected to result in overestimates of the total leak rate and the PoB during a postulated SLB event at the End of Cycle 9.
2)
The nondestructive examination error associated with the use of the 630 probe appears to be significant if the entire database, which includes a large number of indications which had amplitudes of less than 1 volt (by both probes) and which were not RPC confirmed, is i
I
.3
- u
,f Muclear Regulatory Commission June 26,'1995 Page 4:
considered. If the database is restricted to RPC confirmed indications,-
the average error between two teams of analysts was found to be.~7%.
When compared to an expert determination of the true voltages that.
should have been reported, the NDE error is lessened.. Due to the small' sample size, it would not be appropriate to conclude that a systematic error is present which could significantly affect the outcome of analyses perfe aned to estimate the end of cycle totalleak rate in a SG or the PoB -
L.ang a postulated SLB event.
The outcome of the analysis of the Catawba 1 data would not be considered to be generically applicable relative to the analyst error. The conclusion of the' analysis is based on a limited data sample. - Compliance with the NRC's desire that..
absolute blind testing be achieved by not allowing analysts to examine the same indications with the different probes made the evaluation more complex, and may.
l have resulted in the estimate of the difference between the two probes including analyst differences in addition to any real differences.
REFERENCE 1.
Letter to U. S. NRC, Document Control De' k, dated Ma:ch 12,1995, from s
Duke Power Co. (Dave Rehn) on the subject of application of the 0.630" bobbin probe with respect to the IPC submittal for Catawba Unit 1.
A full report with all supporting data will be provided by July 10,1995.
Pursuant to 10 CFR 50.91 (b)(1), the appropriate South Carolina official is being provided a copy of this amendment request supplement.
Very truly you gr D. L Rehn Attachments l
DT/
_