ML20077C577
| ML20077C577 | |
| Person / Time | |
|---|---|
| Site: | Catawba  |
| Issue date: | 11/03/1994 |
| From: | WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP. |
| To: | |
| Shared Package | |
| ML20077C565 | List: |
| References | |
| NSD-TAP-3093, SG-94-11-001, SG-94-11-1, NUDOCS 9412050235 | |
| Download: ML20077C577 (87) | |
Text
NSD-TAP-3093 SG-94-11-001 l
1 i
i CATAWBA - 1 ]
CYCLE 7 IPC ASSESSMENT AND PROJECTED i EOC-8 SLB LEAKAGE )
l i
i November 1994 l l
I l
i I
WESTINGHOUSE ELECTRIC CORPORATION NUCLEAR SERVICE DIVISION P.O. BOX 158 ;
MADISON, PENNSYLVANIA 15663 9412050235 941129 PDR ADOCK 05000413 P PDR
CATAWDA-1 CYCLE 7 IPC ASSESSMENT AND PROJECTED EOC-8 SLB LEAKAGE I
l TABLE OF CONTENTS TOPIC - PAGE
, 1.0 Introduction 1-1 2.0 Summary and Conclusions 2-1 2.1 Overall Conclusions 2-1 l 2.2 1993 EOC-7 Inspection Results 2-1 j 2.3 Voltage Growth 2-3 2.4 Comparison of Projected and Actual EOC-7 Bobbin Voltage 2-3 j Distributions 2.5 SLB Leak Rate Analyses 2-4 2.6 Tube Burst Probability Assessments 2-5 3.0 EOC-6 and EOC-7 SG Inspection Results 3-1 ;
3.1 EOC-6 Inspection Results 3-1
3.2 EOC-7
Summary of Indications at TSPs 3-1 3.3 EOC-7; Assessment of RPC Inspection Results 3-2 !
3.4 EOC-7
Assessment of New Indications 3-4 ,
3.5 EOC-7
Distributions for Comparisons with Projections 3-6 4.0 Bobbin Voltage Indications Left in Service 4-1 4.1 BOC-7 Indications Left in Service 4-1 4.2 BOC-8 Indications Left in Service 4-1 5.0 Voltage Growth Rates 5-1 i
5.1 Cycle 6 Voltage Growth Rates 5-1 5.2 Cycle 7 Voltage Growth Rates 5-1 6.0 NDE Uncertainties 6-1 7.0 Projected EOC Voltage Distributions 7-1 7.1 Projected EOC-7 Voltage Distributions 7-1 7.2 Comparison of Projected and Actual EOC-7 Distributions 7-1 7.3 Projected EOC-8 Voltage Distributions 7-2 8.0 SLB Leak Rate Analyses 8-1 8.1 Database Applied for SLB Analyses 8-1 8.2 Altemate Leakage Analysis Methods 8 8.3 Projected EOC-7 SLB Leak Rates 8-3 l 8.4 Comparison of Leak Rates for Projected and Actual EOC-7 Distributions 8-4 l 8.5 Projected EOC-8 SLB Leak Rates 8-4
. I I 1 i
l l TABLE OF CONTENTS (centinu:d) l l TOPIC PAGE l
9.0 SLB Tube Burst Probability Analyses 9-1 9.1 Tube Burst Probability for Limited TSP Displacement 9-1 9.2 Free Span Tube Burst Probabilities at EOC-7 for Projected and 9-3 Actual Distributions 9.3 Free Span Tube Burst Probability Projections for EOC-8 9-4 9.4 Contributors to Free Span Tube Burst Probability 9-5 10.0 Alternate Method for Defining BOC Indications Left in Service .10-1 !
10.1 Allowance for Undetected or New Indications 10 2 10.2 Considerations for RPC NDF Indications 10-3 {
j 10.3 Comparison of Alternate Method and' Actual EOC-7 Distributions 10-4 ;
10.4 Projected EOC-8 Voltage Distributions and SLB Leak Rates 10-5
)
11.0 References Il-1 -
l i
l l
i I
i I
l l l ii 1
1.0 INTRODUCTION
This report provides a Cycle 7 interim plugging criteria (IPC) assessment and projected End-of-Cycle-8 (EOC-8) Steam Line Break (SLB) leakage for the Catawba Unit-1 steam generators. Included in this report is information requested by the NRC Safety Evaluation Report for appikation of the IPC for Cycle 8. This information includes comparisons of projected EOC-7 bobbin voltage distributions with actual values found in the EOC-7 inspection, projections of EOC-8 voltage distributions based on indications left in service at Beginning-of-Cycle-8 (BOC-8), projected potential SLB leak rates at EOC-8 and tube burst probability at EOC-8.
I
\
l The EOC-7 steam generator (SG) eddy current inspection represents the first Catawba-1 full
! cycle inspection following implementation of IPC repair limits. Thus, the EOC-7 inspection results provide the first Catawba-1 opportunity to compare actual voltage distributions with projected values. Altemate methods for defining BOC indications left in service are used to project EOC voltage distributions for comparisons with the actual EOC-7 inspection results.
The alternate methods for defming BOC indications evaluated include the NRC recommended method of draft NUREG 1477, which includes RPC NDF (No Degradation Found) indications r.nd adjustments for a probability of detection of 0.6. In addition, an alternate or ,
recommended method for defining the BOC distribution is evaluated. This method assumes that a fraction of the RPC NDF indications may be undetected flaws and accounts for undetected BOC indications based on the prior cycle voltages for new indications found in the
- latest inspection.
i The Monte Carlo methods of the Catawba-1 IPC WCAP-13494, Revision I are applied to the BOC voltage distributions to project the EOC distributions. This is consistent with the NRC guidance given in the Catawba-1 SER. SLB leak rates are calculated using the NRC methodology of draft NUREG-1477, as well as the IPC/APC methods described in WCAP-13494 which apply the APC correlations for probability ofleakage and SLB leak rate.
i The use of the correlation of SLB leak rate with voltage is consistent with the NRC resolution I of draft NUREG-1477 comments, as presented by the NRC in the February 8,1994 NRC/ utility meeting, and the NRC draft generic letter on an APC for ODSCC at TSPs. It is shown that the SLB leak rate correlation with voltage satisfies the NRC guidance for application when it is shown with > 95% probability that a correlation with voltage exists. SG l
C is the most limiting SG for SLB leakage and burst considerations for both Cycles 7 and 8, as this SG has the largest number of indications left in sersice. The latest data base of EPRI l Report NP-7480-L, Volume 2 is applied in this report, with treatment of data outliers consistent with the Catawba-1 SER. Tube bt tst probability analyses apply the WCAP-13494 methodology and the burst pressure versus bobbin voltage correlations.
Consistent with the guidance of the Catawba-1 SER, analysis data such as voltage distributions l and growth rates are given in both graphical and tabulated form. Table 1-1 relates the data j requested in the Catawba-1 SER to the report section, table and figure that provide the data. l l
November 3.1094 DCP1RO 2WP5 1-}
l l 1 l
l l
l l
- j TABLE 1-1 l SG C BOBBIN VOLTAGE DISTRIBUTIONS AND GROWTil RATES l NRC SER Subject Section Table Figure .
I Section 3.5 a EOC 6 voltage distribution - all bobbin 3.1 4-1 3-1 ;
indications !
b Cycle 6 growth rate 5.1 5-1 5-1 l c EOC 6 repaired indications 4.1 4-1 -
d BOC 7 voltage distribution - all bobbin 4.1 4-1 4-1 indications !
BOC 7 voltage distribution - draft 4.1 4-1 4-1 ,
l NUREG-1477 BOC 7 voltage dist. - alternate method 10.3 10-1 -
e BOC 7 voltage dist. - RPC confirmed 4.1 4-1 -
plus not RPC inspected f Cycle 7 NDE uncertainty 6.0 - -
g Proj. EOC 7 voltage dist.-WCAP-13494 7.1 7-1 7-1 Proj. EOC 7 voltage dist.-NUREG-1477 7.1 7-1 7-2 l Proj. EOC 7 voltage dist.-alternate 10.3 7-1 - 1 method h Actual EOC 7 voltage dist.- all bobbin 3.2 3-1, 4-2 3-2 ind.
Actual EOC 7 voltage dist.- RPC 3.5 4-2 3-6 confirmed plus not RPC inspected l
, i Cycle 7 growth rate 52 5-2 5-2 j EOC 7 repaired ind. 3.2 3-1, 4-2 -
k BOC 8 voltage dist.-all bobbin ind. 4.2 4-2 4-2 BOC 8 voltage dist.-draft NUREG-1477 4.2 4-2 4-2 BOC 8 voltage dist.-altemate method 10.4 10-2 -
l BOC 8 voltage dist.-RPC confirmed plus 4.2 4-2 -
not RPC inspected
- m Cycle 8 NDE uncertainty 6.0 - -
l n Proj. EOC 8 voltage dist.-WCAP-13494 7.3 7-2 7-3 Proj. EOC 8 voltage dist.-NUREG-1477 7.3 7-2 7-3 Proj. EOC 8 voltage dist.-alternate 10.4 10-2 -
method 1-2
2.0
SUMMARY
AND CONCLUSIONS 2.1 Overall Conclusions EOC-8 projections were made, consistent with the Catawba-1 SER for Cycle 8l by applying the NRC model of Draft NUREG-1477 to define the BOC distribution with a probability of detection (POD) adjustment factor of 0.6, the NRC SLB leak rate model and the correlation of burst pressure with voltage to obtain the SLB tube burst probability. The resulting EOC-8 SLB leak rate is e:timated at 27.6 gpm which is below the allowable limit of 30 gpm for Catawba-1 Cycle 8. When the leak rate versus voltage correlation is applied using the NRC recommended database, the projected SLB leak rate is only 1.61 gpm by Monte Carlo analyses. The projected EOC-8 SLB tube burst probability by Monte Carlo analysis is 9.6 x 10 4using the EPRI database, which is essentially the same as the NRC recommended database of the SER. The projected EOC-8 burst probability is below the acceptance guideline of 2.5 x 10 4based on NUREG-0844 analyses. Overall, it is concluded that SLB leakage and burst probability acceptance limits were satisfied in Cycle 7 and will be satisfied with large margins for Cycle 8. The Cycle 7 voltage growth rates at Catawba-1 are smaller than those of Cycle 6, although the maximum voltage growth increased from 2.77 volts for Cycle 6 to 4.18 volts for Cycle 7.
Comparisons between projections and actuals are provided in this report for the bobbin voltage distributions at EOC-7 and the SLB leak rate and burst probability calculated from the actual EOC distribution. The best agreement, while retaining conservatism, between projections and actuals is obtained for the altemate or new indication method of defining BOC distributions desenbed in Section 10 of this report. The draft NUREG-1477 methodology with a POD = 0.6 adjustment to the BOC distribution results in similar agreement with the actual voltage distributions and associated SLB leak rates and burst probabilities.
The projections for Cycle 7 were based on SG C, which had the most limiting BOC-7 voltage distribution. The largest EOC-7 RPC confirmed indication was 5.03 volts. The largest voltage projected for SG C at EOC-7 ranged from 3.2 to 4.2 volts depending on the method of defining the BOC distribution. Even though the projections underestimated the maximum voltage found in any SG, the projected SLB leak rates and tube burst probabilities are essentially equal to or higher than the values calculated from the actual EOC-7 voltage distributions for SG C. This conclusion applies for all three methods (POD = 0 6, POD = 1.0 and new indication methods) evaluated for defining the BOC voltage distributions and l demonstrates the conservatisms in the projection methods.
j 2.2 1993 EOC-7 Inspection Results The number of potential flaw indications found in the 1993 inspection is 8308 which includes 1270 in SG A,1543 in SG B,2765 in SG C and 2730 in SG D. The number of RPC confirmed indications was 329 of the 1898 potential indications RPC inspected. For DCPIRO 2 WP5 2-1 Novm ber 16.1994 i
indications above 1.0 volt, the RPC confirmation was 245 out of 1080 potentia' . oin indications RPC inspected. Thus the RPC confirmation rate for potentially repairable bobbin indications above 1.0 volt was 23% For SG C, which had the highest confirmation rate, the RPC confirmation rate above 1.0 volt was 39%. These low confirmation rates are indicative of conservative bobbin indication calling criteria. The largest bobbin indications confirmed by RPC were 5.03,4.00 and 2.91 volts. A potential bobbin indication at 2.67 volts was not confirmed as a flaw indication by RPC No tubes were found for which the bobbin voltage exceeded the maximum bobbin voltage of 2.7 volts permitted to be left in service for unconfirmed or RPC NDF indications.
To assess the need for including RPC NDF indications left in service in the SLB leakage analyses, an assessment was made of the RPC NDF indications in service at BOC-7 that became RPC confirmed indications at EOC-7. Out of 375 RPC NDF indications left in service and RPC inspected at EOC-7,52 or 14% were confirmed as flaws at EOC-7. For indications above 1.0 volt,16% were confirmed The highest rate of confirmation for BOC-7 NDF indications was 23% in SG C. Draft NUREG-1477 methodology requires that all RPC NDF indications, in addition to a POD adjustment of 0.6, be included in the SLB analyses. The low Catawba-1 confirmation rate for RPC NDF indications left in service indicates that the NRC methodology is excessively conservative and only about 25% of the NDF indications should be included in the SLB leak rate and tube burst analyses.
New indications were found in the 1993 inspection that were not reported at the 1992 EOC-6 inspection. Also a number of potential bobbin indications reported at the 1992 inspection were not found in the 1993 EOC-7 inspection. Out of a total over 4 SGs of 6624 indications left in service at BOC-7,3353 or 51% of the BOC potential indications were not reported at EOC-7.
A total of 5030 new indications (not reported in EOC-6 inspection) were reported at EOC-7 which leads to a net number of new indications of 5030 - 3353 = 1677. The RPC confirmation rate for the new indications was only 13% over all SGs and 21% in SG C. Thus the EOC-7 inspection results indicate significant numbers that are new indications and prior indications that do not remain flawlike at EOC. In addition, the new indications show very low RPC confirmation rates of 13 to 21%. These trends of new indications with low RPC confirmation and " disappearing" indications show the variability in calling small indications (dominantly < 1 volt) when conservative bobbin calling criteria are applied. s The probability of prior cycle detection (POPCD) at EOC-6 for indications found at EOC-7 was evaluated. Only RPC confirmed indications are significant for considerations of SLB leakage / burst and were used for the POPCD evaluation. SG C was used for the evaluation as this was the limiting SG for SLB considerations and had the highest RPC confirmation rate.
The resulting 1992 POPCD was 37.5% for indications < 0.5 volt, increased to 81% to 88% for indications between 0.5 and 2.0 volts, and was 100% (with only 3 indications) above 2.0 volts.
The overall average POPCD was 82.4%. These results show that the draft NUREG-1477 POD
= 0.6 is too low for IPC applications and that a constant POPCD should not be assumed for all voltage levels (i.e., the POPCD should be higher at high voltage levels).
1 DCPIR0 2 WP5 2-2 Nov =ber 3.1994 l
l l
i 1
The RPC confirmation rate at EOC-7 was separately assessed for three categories of l indications in SG C. For indications above 1.0 volt, prior bobbin indications that were not RPC NDF indications had a confirmation rate of 67%, prior RPC NDF indications had a l confirmation rate of 23% and new indications had a confirmation rate of 23%. The prior cycle
- RPC confirmed plus not RPC inspected indications contributed about 2/3 of the EOC-7 RPC I confirmed indications even though this group was only about 38% of the EOC-7 RPC inspected indications.
! I 2.3 Voltage Growth The 5.03 volt RPC confirmed indication had a 1992 voltage of 0.85 volts and had the largest voltage growth of 4.18 volts found for Cycle 7. For Cycle 6, the largest growth was 2.77 volts. The Cycle 7 growth rates are lower in average value compared to prior Catawba-1 l cycles while the two largest growth values (4.18,3.18 volts) in Cycle 7 exceeded the Cycle 6 i maximum value. The average voltage growth for Cycle 7 was essentially zero (calculated to j be negative) compared to 0.01 volt average growth for Cycle 6. i i
2.4 Comparison of Projected and Actual EOC-7 Bobbin Voltage Distributions I The 1993 Catawba-1 inspection represents the first full cycle of operation following IPC l implementation. Thus the resulting data permits comparisons of projected and actual EOC l voltage distributions. These comparisons were made applying the NRC draft NUREG-1477 methodology including a POD = 0.6 adjustment to detected indications, a method including RPC NDF indications without a POD adjustment (POD = 1.0 or all bobbin indications left in i I
service) and an altemate method that includes provisions for undetected indications based on new indications found in the inspection as well as a fraction (conservatively 25%) of the RPC NDF indications assumed to develop to confirmed flaws. His altemate method (called the new indication method) is proposed for further evaluation against actual EOC distributions as this method is shown to provide adequate conservatism while applying more realistic, plant specific cdjustments for undetected indications and that only a fraction of RPC NDF indications are likely to develop into significant flaws in the subsequent cycle. The projection methods are compared with the actual EOC-7 distribution for RPC confirmed indications plus ;
i indications not RPC inspected. Since the primary purpose of the voltage projections for IPC/APC applications is to estimate tube leakage and burst probabilities, only indications found to be RPC confirmed (not inspected are conservatively assumed to be confirmed) are used for comparisons with the projected distributions as RPC NDF indications would not have any significant probability ofleakage over the prior cycle.
Both the new indication method and the draft NUREG-1477, POD = 0.6 method applied to EOC-7 projections provide good agreement with the actual EOC-7 inspection results below 1.0 volt and, like the POD = 1.0 method, conservatively bound the distributions above 1.0 volt.
November 3, l994 DCP1R0w2 WP5 2-3
Due to the large number of new, low voltage indications found in the 1993 inspection, the
! POD = 0.6 method is not as conservative as found for other IPC assessments. The new indication method also accounts for the large number of new indications. The low RPC i
confirmation rate for indications < l.0 volt (19% for SG C), and the trend found that many l
BOC-7 potential bobbin indications left in service were not found at EOC-7, tend to indicate that many of the low voltage indications may not be real flaws. The conservative bobbin l i
calling criteria, together with the low actual POD for indications < 0.5 volt, are the likely contributors to the agreement of the POD = 0.6 projection with the < l.0 volt actual EOC-7 l
l voltage distribution. The new indication method, which utilizes the large number of new l indications found at the EOC-6 inspection, also yields good agreement with the actual
< l.0 volt distribution. The POD = 1.0 projection, which does not account for new or undetected indications, underestimates the actual < l.0 volt distribution.
All projection methods conservatively bound the actual EOC-7 voltage distribution between 1.0 and 4.0 volts. However, the projection methods project maximum voltages of 3.2 volts l (POD = 1.0),4.0 volts (POD = 0.6) and 4.2 volts (new indication method) compared to the j 5.1 volt maximum found in the EOC-7 inspection. The underprediction results from the i largest growth of 4.18 volts found in Cycle 7, compared to the maximum of 2.77 volts (based l on Cycle 6 data) used in the projections. As noted below, the SLB leak rates and burst i
probabilities are adequately predicted for EOC-7 compared to values calculated from the actual EOC-7 voltages. :
l l
2.5 SLB Leak Rate Analyses SLB leak rate anrlyses were performed to compare values based on projected EOC-7 voltage distributions with that obtained from the actual distribution as well as to project EOC-8 leak rates for comparison with allowable limits. Applying NRC draft NUREG-1477 methods for calculating leakage, the ctual EOC-7 distribution resulted in a 13.0 gpm SLB leak rate. For the NRC method of d- ng BOC voltages, which includes a POD = 0.6 adjustment and RPC NDF indications, the ;, ,dicted EOC leakage was a factor of 1.5 high at 20.6 gpm. Methods excluding the POD adju,tment (POD = 1.0) predicted 13.0 gpm in good agreement with that obtained from the actual distribution. The new indication method yields a conservative 26.0 gpm, higher than the POD = 0.6 method. Applying Monte Carlo with the APC leak rate versus voltage correlation to the actual EOC-7 voltage distribution yields a 0.57 gpm leak rate, well below the 13 gpm obtained with draft NUREG-1477 methodology. The projected distributions yielded SLB leak rates in good agreement with that obtained from the actual distribution.
The draft NUREG-1477 methods were applied to projected EOC-8 voltage distributions (including POD = 0.6 and RPC NDF indications) to obtain a projected SLB leak rate of 27.6 gpm. This is below the Catawba-1 allowable leakage limit of 30 gpm. Thus, potential SLB leakage is within acceptable limits for Cycle 8 operation. When the full APC leak rate 2-4 November 16.1994 DCPIR0-2 WPS
l versus voltage correlation is applied to the EOC-8 voltage distribution, the projected SLB leakage is only 1.6 gpm or a factor of 15 lower than the draft NUREG methodology. This demonstrates the conservatism in the draft NUREG methods which apply the leak rate database l independent of bobbin voltage. Overall, the projection methods yield SLB leak rates in good agreement with that obtained from the actual EOC-7 distribution. This good agreement ,
indicates the conservatism of the projection methods, even when the maximum EOC voltage is ,
underpredicted. ;
l 2.6 Tube Burst Probability Assessments The tube burst probability obtained by Monte Carlo methods from the actual EOC-7 distribution is 2.9x104, Projections for EOC-7 yield good agreement with the actual value.
l The tube burst probability at EOC-8 was estimated by applying the draft NUREG method for l
l defining BOC voltage distributions and the APC correlation methods for burst probability i analyses. The draft NUREG, POD = 0.6 method recommended in the NRC Catawba-1 IPC ,
l i
! SER yields a projected EOC-8 SLB burst probability of 9.6x10, which is below the guideline of 2.5 x 10 based 2 on NUREG-0844 analyses. Thus, the tube burst probability is also within acceptable limits for Cycle 10 operation.
I l .
t l
i l
l l
November 16,1994 DCP1RO-2 WP5 2-$
1 i
,_e , - -
i l
l EOC-6 AND EOC-7 SG INSPECTION RESULTS !
3.0 l
3.1 EOC-6 Inspection Results i
EOC-6 inspection results are required in this report to define the BOC-7 indications left in service that are used as the starting point for projecting EOC-7 voltage distributions for comparison with the actual results from the EOC-7 inspection. The 1992 inspection at EOC-6 was the first Catawba-1 inspection implementing an IPC. This included use of ASME j calibration standards normalized to the reference laboratory standard as described in Appendix A.
l l Figure 3-1 shows the SG C and SG D bobbin indications found at EOC-6 and the indications confirmed as flaws by stPC inspection. The number ofindications found in SGs A and B are j much smaller than SGs C and D. For Figure 3-1 and all figures given in this report, the voltage values gives on the x-axis of the plots represent the upper or right-side value for the voltage bin. Only indications above the IPC repair limit of 1.0 volt were nearly 100% RPC '
inspected. The few indications above 1.0 volt that were not RPC inspected were repaired.
i Only a sample of indications below 1.0 volts were inspected. SG C has been applied for projected EOC-7 voltages and leak rates based on comparative leak rate calculations that showed SG C to be the most limiting. A total of 1971 bobbin indications were found in SG C. Above 1.0 volt,100 indications were found and 171 of these indications were removed from service principally due to confirmation as flaws by RPC. The indications left in service at BOC-7 are discussed in Section 4.1.
The largest bobbin voltage found in the inspection was 3.55 volts on tube R12Clll in SG C.
RPC inspection confirmed the flaw indication on this tube. The next largest bobbin voltage was 2.8 volts in SG D which was also confirmed as a flaw by RPC inspection.
No indications at TSPs were identified that had cracks extending outside the TSPs by either the l bobbin or RPC inspections. In addition, no abnormal indications such as circumferentially oriented indications were found in the 1992 inspection.
3.2 EOC-7
Summary of Indications at TSPs The 1995 inspection at EOC-7 was completed in November,1993. This inspection is the first full cyu inspectio . tit Catawba-1 following implementation of an IPC which permitted leaving indications of 1.0 volt or less in service for indications at TSPs. For this reason, the results of the inspection are evaluated for IPC methodology considerations including comparisons between projected and actual indications in this report. This section summarizes the inspection results.
Nov==ber 2.19a4 DCPIR.1 WP5 3-1
A summary of the inspection results are given in Table 3-1. Figure 3-2 shows the total for all four SGs of all bobbin indications, RPC confirmed indications and repaired indications.
Figure 3-3 shows the corresponding data for SGs C and D. Only a small fraction of the bobbin indications below 1.0 volt were RPC inspected. The number of repaired indications is greater than the number of RPC onfirmed indications and includes indications less than the 1.0 volt repair limit. This results from the presence of more than one indication on a tube and repir due to other causes than TSP indications. Based on Table 3-1, the total number of potertial Oaw indications summed over all 4 SGs is 8308 which includes 1270 in SG A,1543 in SG B,2765 in SG C and 2730 in SG D. The total number of RPC confirmed potential flaw i
indicatons above 1.0 volt is 245 out of about i121 indications RPC inspected for an RPC
! confirmation rate of about 22% Table 3-1 includes the distribution of RPC confirmed indications and repaired indications as well as the total bobbin voltage distribution. A more detailed assessment of the high fraction of unconfirmed RPC indications is gien in Section 3 3 below.
Table 3-2 provides a summary of the largest bobbin voltage indications found in the current l
inspection. RPC inspection result:;, voltages at the prior inspection and whether or not it is a new indication not reported at the last inspection are also included in Table 3-2. The largest bobbin voltage indications found in 1993 were 5.03 and 4.00 volts in SG C and 2.91 volts in SG D For SGs A and B, the largest bobbin indications were small at 1.66 and 1.55 volts and these indications were found to be RPC NDF. The largest RPC confirmed indications in l SGs A and B were 1.25 and 1.44 volts, respectively. The largest new indication (not reported in 1992 inspection) confirmed by RPC was the 2.91 volt indication on R8C4 in SG D. The largest RPC NDF indication was 2.67 volt in SG D. This indication voltage is less than the l
2.7 volt limit (Catawba-1 SER) for leaving RPC NDF indications in service. Of the 8 new indications in Table 3-2,2 were RPC confirmed, I was not inspected by RPC (plugged for other causes) and 5 were RPC NDF. A more detailed assessment of the new indications found j in the 1993 inspection is given in Section 3.4 below.
3.3 EOC-7
Assessment of RPC Inspection Results The largest RPC confirmed indication was on tube R7C41 in SG C with a bobbin voltage of 5.03 volts. This indication was left in service in 1992 with a bobbin voltage of 0 85 volts and has the largest voltage growth of 4.l'8 volts found in the last cycle. As shown in later sections of this report, this indication is the only RPC confirmed indication not closely predicted or bounded by the projections made N EOC-7 based on SG C indications left in service at BOC-7. Fourteen of the 26 bobbm .ndications in Table 3-2 were RPC NDF. A large fraction of the bobbin indications were found to be RPC NDF and most of the RPC indications that l
were found to be NDF in 1992 are also NDF in 1993. Figure 3-4 shows the totals for all 4 l SGs the were RPC NDF at BOC-7, that remained RPC NDF at EOC-7 and the small fraction that were confirmed as RPC flaws at EOC-7. It is readily seen from Figure 3-4 that most RPC NDF indications at BOC remain NDF at EOC.
November 1. t o94 DCPIR.3 WP5 3-2 I
Duke Power has implemented very conservative bobbin calling criteria for identifying potential indications for RPC inspection. Thus, it is to be expected that Catawba-1 would have a very low RPC confirmation rate. Conservative bobbin calling criteria are encouraged by the i Appendix A eddy current analysis guidelines of Reference 2 as applied for the 1993 mspection.
The NRC guidelines of draft NUREG-1477, which are not modified by the NRC presentation of February 8,1994 on resolution of industry comments, require that all RPC NDF indications l left in service be included in the SLB leakage and burst probability analyses. This leads :o l excessive conservatism in the SLB analyses when, as for Catawba-1, conservative bobbin calling criteria are implemented. Excessive conservatism should not be required in the SLB analyses to the pomt that it discourages implementation of conservative bobbin calling criteria.
Tables 3-3 A to 3-3D summarize the RPC inspection results .or SGs A to D from the 1993 l inspection and includes the number of tubes RPC inspected, the number confirmed as flaws by RPC and the percentage of RPC confirmed indications. It is seen that below about 2 bobbin volts, the fraction of RPC confirmed indications is very low. For SG C which had the highest RPC confirmation rate, < 39% of the indications are confirmed between 1 and 2 volts and only 21% of the indications below 1 volt are confirmed by RPC. Even if the bobbin indications not confirmed by RPC are assumed to be real flaws, the indications would be too small to contribute significantly to leakage and burst probability. Thus, only RPC confirmed plus not RPC inspected (last column of Table 3-3) bobbin indications should be included in SLB analyses for EOC-7 conditions or used for comparisons with projected EOC voltage distributions.
For considerations of including the RPC NDF indications left in service for projecting SLB leak rates at the end of the next operating cycle, the relevant question is whether or not the ;
RPC NDF indications tend to remain as NDF at the end of the next cycle Table 3-4 summarizes the largest bobbin indications in each SG that were RPC NDF at EOC-6 (left in l service at BOC-7) and provides the results of the EOC-7 RPC inspection for thm indications.
Also included in Table 3-4 are the largest new RPC NDF indications left in service at E OC 8.
Of the largest 50 RPC NDF indications left in service at BOC-7 as given in Table 3-4, anly 6 of the indications or 12% were confirmed as RPC indications at EOC-7. Table 3-5 provides an ,
overall summary of the EOC-7 inspection results for indications that were RPC NDF at EOC-6 l and left in service at BOC-7. Some of the EOC-6 RPC NDF indications were not inspected at EOC-7. The overall results support the small fraction of RPC NDF indications that become flawlike at the end of the subsequent operating cycle. For all indications RPC NDF in 1992 and RPC inspected in 1993,14% of the 1992 RPC NDF indications became confirmed flaws in 1993. SG C had the highest fraction at 21% of 1992 RPC NDF indications confirmed as flaws in 1993. Thus the large majority of RPC NDF indications continue to be RPC NDF and are most likely indicative of false bobbin calls. A similar trend of RPC NDF indications remaining RPC NDF, although a smaller population of indications, was found in the IPC assessment for the Farley-21993 inspection.
N aember t 1994 DCPIRd WP5 3-3
i l
l l
Based on the large fraction of RPC NDF indications and the fact that the large majority of l RPC NDF indications remain RPC NDF at the end of the next operating cycle,it is concluded l that including all RPC NDF indications left in service in the population of tubes for projecting SLB leakage is excessively conservative This conservatism would further increase in successive IPC cycle applications as the number of RPC NDF indications grows from cycle to cycle For example, bobbin indications that are RPC NDF for two successive inspections should be excluded from the SLB analyses and only a fraction of the new RPC NDF l indications should be included in the SLB analyses This topic is further addressed in the f discussion of Section 10 which provides an alternate model for defining BOC indications left :
in service. It can further be noted that the conservatism for RPC NDF indications becomes further magnified when a POD correction, such as required by draft NUREG-1477, is applied to all bobbin indications found in the inspection. In this case, all RPC NDF indications left in service become 1.67 indications left in service for a POD of 0.6 as required by the NRC.
l
3.4 EOC-7
Assessment of New Indications j 1
Given the occurrence of new indications in the largest bobbin voltages of Table 3-2, the 1993 inspection results were evaluated to categorize the 1993 indications as prior or new (not i
reported in 1992) indications. Tables 3-6A to 3-6E summarize this evaluation for each SG and the sum for all SGs. A number of bobbin indications reported in the EOC-6 inspection and l left in service were not found (NDF in '93) in the EOC-7 inspection. Figure 3-5 shows the l distributions for all SGs and for SG C that were '92 bobbin indications found NDF in the '93 l inspection and the distnbution of new indications in '93. The figure shows that many low l l
l voltage, bobbin potential indications are not found in the subsequent inspection and can be interpreted as false bobbin calls. A similar result was found in the Farley-2 IPC evaluation where these indications were classi6ed as indications not reportable (INRs). This type of indication is given in Table 3-6 as the column called Prior Ind. Not Found. The net number of effective new indications is also given in Table 3-6 and is obtained as the total number of new indications minus the number of prior indications not found. This net number of new indications is more relevant to POD considerations than the total number of new indications when conservative bobbin calling criteria are applied. As would generally be expected from application of conservative bobbin calling criteria, a significant number of the EOC-7 l
indications are new indications. At EOC-6 with the first time application of the conservative critena, nearly all of the indications were new indications.
Table 3 6 also mcludes the number of new indications that were RPC inspected and the number that were RPC confirmed. Again the fraction of indications that are con 6rmed as l Gaws by RPC inspection is small. For SG C, the net number of new indications represents about 35% of the total bobbin indications. For new bobbin indications above 1 volt, only 23%
of the indications were conGrmed by RPC and an even smaller fraction of the indications less f
j than 1 volt were con 6rmed by RPC. Table 3-7 provides a summary of the RPC inspection results for SG C which had the highest RPC confirmation rate of all SGs. In Table 3-7, the l
Nove=ber 2.1994 l DCPIR 3 WP5 3-4 l
l
RPC confirmation rate is separately identined for Prior RPC NDF indications left in service, for new indications and for prior bobbin indications left in service that were RPC con 6rmed and not RPC inspected. It is seen that the RPC confirmation rates are low (20 to 23%) for prior RPC NDF and new indications For prior bobbin indications left in service that were below the repair limit and RPC confirmed or were not RPC inspected, The RPC con 6rmation rate is much higher at 50% for all indications and 66 9% for indications > 1.0 volt. Thus the prior RPC NDF and new indications, while a large fraction (60%) of the total bobbin indications, contnbute only about a third of the total RPC confirmed indications.
The EOC-7 inspection results can be used to estimate the probability of prior cycle detection (POPCD) at the EOC-6 inspection. For considerations of signi6cance to IPC applications, only indications conGrmed by RPC inspection at EOC-7 are important for SLB leakage and burst probability analyses. POPCD is defined as the number of RPC confirmed indications found in the latest inspection, plus confirmed and plugged at the prior inspection, divided by the total number of RPC confirmed indications including new indications at the latest inspection.
POPCD has been developed for the SG of interest, SG C, based on RPC confirmed indications as given in Table 3-8. An overall average POPCD is obtained as 82.4% based on 45 RPC confirmed new indications out of a total of 256 RPC confirmed indications. Thus 211 of the 256 EOC-7 RPC con 6rmed indications were detected at EOC-6. This process defines the POPCD obtained at EOC-6 which can be evaluated as a function of voltage by binning the voltage ranges, as shown in Table 3-8 for SG C. From Table 3-8, it is seen that the POPCD is low at 37.5% below 0.5 volt, and increases to 81% to 88% between 0.5 and 2.0 volts. Above 2.0 volts, the Catawba-1 POPCD is 100% although there are only 3 indications with EOC-6 volts above 2.0 volts The largest 1993 new indication (not reported in the '92 inspection) had a voltage, upon reevaluation, of 1.8 volts It should be noted that this is the EOC-6 POPCD of detecting an RPC confirmed indication at either EOC-6 or EOC-7. The POPCD for detecting an indication that might be large enough to leak at EOC-7 would be much higher smce, even at 1.8 volts, the probability of leakage is only about 4%. For IPC applications, it i; important that the POPCD be related to the signi6cance of the indication for leak or burst and thus the important POPCD is that at higher voltages for RPC confirmed indications which is about the 87.5% found in SG C between 1.5 and 2.0 volts and 100% above 2.0 volts. These results demonstrate that the draft NUREG POD = 0.6, independent of voltage, is excessively conservative.
At EOC-7, there were a net 959 new indications in SG C as given in Table 3-6. Although the fraction of new indications that are confirmed by RPC is small for Catawba 1, the number of new indications implies that allowance must be included in projected SLB leakage analyses for new indications. The allowance for new indications should include considerations of the expected RPC confirmation rate. A proposed model for including new indications in the SLB analyses is described in Section 10. The use of a constant POD independent of voltage is not justified and the proposed method is based on the bobbin voltage at the prior inspection (from reanalyses of prior eddy current data for voltage growth analyses) for new indications found in the current inspection. The Section 10 model for Cycle 8 applies the net number of effective DCPIR4 WP5 3-5 hea i a
I new indications as given in Table 3-6 at the BOC-7 volts to represent undetected indications This assumes that the distribution of undetected indications at the EOC-7 inspection is the same as found at the EOC-6 inspection l'
3.2 EOC-7 Distributions for Comparisons with Projections i
i Voltage projections from BOC to EOC conditions for IPC/APC applications are performed to estimate SLB leakage and burst probability. Thus the desired EOC voltages are those having ,
l significant potential for throughwall cracks or leakage. Indications that are RPC NDF at EOC can be confidently expected to have no potential for leakage over the prior cycle as RPC detectability (as well as bobbin detectability) approaches 100% for ODSCC at TSPS with near throughwall or throughwall indications. It can be noted that the confidence for no leakage over the prior cycle for RPC NDF is much higher than the judgement for leakage potential over the next cycle from RPC NDF indications left in service (i.e. the NRC requirement to include RPC NDF in projected leakage analyses) Therefore, for comparisons of projected EOP 7 witage distributions to be used in leakage analyses with actual distributions, the appropriate actual EOC-7 voltage distribution is the sum of RPC confirmed indications and bobbin indications not RPC inspected (below 1.0 volt for Catawba-1). These distributions are shown in Figure 3-6 for SG C, which is the limiting SG for SLB leakage analyses for both Cycles 7 and 8. The SG C actual distnbutions of Figure 3-5 for EOC-7 are compared with projections from BOC-7 to EOC-7 in Section 7.
I
%=su t wa DCPIR 1 WP5 3-6
Table 3-l: Catawba I, EOC 7, Comparison of Indications Found at TSP's, Confirmed by RPC, and Indications Repaired Steam Generator "C" Steam Generator "D" All Steam Generators Steam Generator "A" Steam Generstor "ll" No. of No. of No. of No of No of No of No of No of No of No of No. of No. of No of No of No. of llobbin Indicatens Indmations llobtnn Indicatens Indkatons 13obtwn Indicatons indw.stums Volts llobtwn Indications Indicatons llobbin indsstens indwatens Repaired Indicatens Confirmed Repaired Indicatmns Con 6rmed Repaired Indicainns Con 6rmed Repaired Indicatens ConGrmed Repaired Indicatens Con 6smed 24 0 2 15 0 0 48 0 2 0.2 7 0 0 2 0 0 0 80 I 6 50 2 5 194 6 11 0.3 42 2 0 22 1 174 7 128 0 11 492 6 23 04 124 1 2 66 0 3 5 6 19 242 23 894 13 50 206 5 6 160 1 2 286 1 0.5 9 77 9 416 4 38 320 4 30 1263 0.6 260 0 0 267 1 492 9 50 332 3 35 1416 19 101 260 4 8 332 3 8 0.7 5 149 0 4 442 2 30 387 1 58 1225 08 173 2 7 223 307 2 38 357 5 47 994 7 92 0.9 110 0 2 220 0 5 4 212 10 49 292 6 35 661 19 90 10 44 1 2 113 2 7 132 40 65 172 23 37 397 68 lil 1.1 21 0 2 72 5 93 39 54 102 14 28 240 56 90 1.2 9 0 3 36 3^^ 5 30 99 23 34 175 50 68 8 2 2 14 2 2 54 23 1.3 44 0 0 18 8 10 102 21 33 129 29 14 4 0 1 5 13 6 10 58 8 12 81 16 25 0 0 9 2 3 1.5 1 0 2 4 26 6 10 33 8 14 1.6 0 0 0 2 0 5 4 I i 17 4 9 22 5 10 0 0 0 0 0 1.7 1 4 9 0 0 0 3 1 3 12 3 6 15
____ _ I 8 _. _ _0 0_ __ _ 0 -
i I 0 3 1 2 0 0 0 0 0 0 l 2 1 20 4 2 5 2 3 0 0 0 0 0 1 1 I 1 2.1 0 0 0 0 0 0 0 1 0 0 2.2 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 2.3 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 2.4 0 0 0 0 0 1 ~ '~
0 0 0 0 0 0 0 0 0 ~2 0 i ~2 ~ 0 l 2.5 -
0 0 0 2 I I 2 I I 2.7 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 1 2.8 0 0 0 1 0 0 0 0 0 1 1 1 I I I 29 0 0 0 0 0 0 0 0 0 0 0 1 I I I i 1 3.0 0 0 0 0 t ! I O O O I I 1 40 0 0 0 0 O O O O O O O O O O 0 48 0' O O O 0
0 0 0 0 1 1 1 0 0- 1 1 1 51 0 0 52 2765 164 473 2730 128 420 8308 329 980 Total > 0 V 1270 17 35 1543 20 47 2201 152 439 2295 125 381 6680 304 894 Total > .5V -891 9 27 1293 18 332 125 184 607 106 176 1121 245 385 44 2 8 138 12 17 Total > l V era siesi,4 9 3s au a m"=== n e t-* n
--- .--- r - _ ._ . _ . _ _ __ _ _
9 Table 3-2.
SUMMARY
OF LARGEST EOC-7 BOBBIN
- VOLTAGES EOC-7 BOC-7 b#O Bobbin Tube TSP Bobbin RPC fn'd Volts Volts Volts
, C R7C41 2 5.03 3.76 0.85 No j C R25C99 3 4.00 5.00 0.82 No l D R8C4 2 2.91 0.82 1.93 Yes D R8C65 2 2.89 1.80 0.93 No i l
C R12C111 3 2.73 N.I.m 1.43 Yes D R9C62 3 2.69 2.36 0.94 No D R29C47 2 2.67 NDF 2.32 Yes D R2C17 2 2.48 NDF 2.14 Yes j D R9C90 3 2.47 N.I.O 1.20 No D R18C18 3 2.31 NDF 1.94 No :
I C R22C60 2 2.28 NDF 1.10 Yes D R20C44 3 2.17 NDF 1.15 No C R32C54 3 2.09 0.39 1.78 Yes 1 D R21C25 2 2.07 0.41 1.83 No D R13C30 3 2.06 NDF 0.44 No l
D R7C26 2 2.01 NDF 1.33 No D R14C43 3 2.01 NDF 1.31 No D R43C33 3 2.00 NI.m 0.35 No C R13C14 3 1.97 1.18 1.47 No D R3C19 3 1.96 NDF 1.88 No D R14C51 4 1.88 NDF 1.00 Yes C R15C65 2 1.86 1.07 1.48 No C R26C18 3 1.86 NDF 1.57 No A R48C62 4 1.66 NDF 0.95 No B R35C40 2 1.55 NDF 1.46 Yes Notes: 1. N.I. = Not RPC Inspected. Tube plugged for other causes.
- 2. N.M. = Not measured.1992 data was not reevaluated for this indication.
3-8 sert == ber 16.1994 DCPIR.3 WP5
i l
l l
i i
Table 3-3A: Catawba 1, SG "A", EOC 7 Summary of 1993 RPC Inspection Results RPC Not RPC RPC Fraction RPC Bobbin Volts Performed Confirmed Confirmed Performed N 4 3 0 0.0% 4 0.2 0.3 30 12 2 16.7 % 32 0.4 108 16 1 6.3% 109 0.5 168 38 5 13.2 % 173 0.6 226 34 0 0.0% 226 0.7 229 31 4 12.9 % 233 0.8 141 32 2 6.3% 143 0.9 88 22 0 0.0% 88 1.0 33 11 1 9.1% 34 1.1 1 20 0 0.0% 1 1.2 0 9 0 0.0% 0 ,,,,_
1.3 0 8 2 25.0 % 2 1.4 0 4 0 0.0% _0 1.5 0 1 0 0.0% 0 1.7 0 1 0 0.0% 0 Total 1028 242 17 1045 43 2 3 Total > 1 V 1 RFK 11/494. 9 45 AM
[DCPJOINA XLS] Joined Data
I i
I Table 3-3B: Catawba 1, SG "B", EOC 7 Summary of 1993 RPC Inspection Results RPC RPC Not RPC RPC Fraction RPC
" IS Confirmed Performed Performed Confirmed
~
0.'2 1 1 0 0.0% 1 f- 16.7 % 17 l 0.3 16 6 1 0.4 58 8 i 0.0% ^
58 0.5 143 17 1 5.9% 144 0.6 231 36 1 2.8% 232 0.7 280 52 3 5.8% 283 0.8 192 31 0 0.0% 192 0 0.0% 194 0.9 194 26 _ 85 1.0 83 30 2 6.7%
1.I 1 71 5 7.0% 6 1.2 2 34 3 8.8% 5 1.3 0 14 2 14.3 % 2 1.4 0 5 0 0.0% 0 1.5 0 9 2 22.2 % 2 1.6 0 2 0 0.0% 0 Total 1201 342 20 1221 135 12 15 Total > 1 V 3 RFK 11/494.10 01 AM
[DCPJOINB.XLS) JOINED _B XLS
I l
1 l
l l
l Table 3-3C: Catawba 1, SG "C", EOC 7 j Summary of 1993 RPC Inspection Results RPC Not RPC RPC Fraction RPC Bobbin Volts "'"""
Performed Performed Confirmed Confirmed 0.20 22 2 0 0.0% 22
. ~ . . . -
0.30 69 11 1 9.1% 70 l 0.40 150 24 5 20.8 % 155 !
~ 0.50 265 21 6 28.6 % 271 0.60 385 31 4 12.9 % 389 l
~~
0.70 455 37 9 24 3 % 464 0.80 417 25 2 8.0% 419 '
O.90 288 19 2 10.5 % 290 1.00 177 35 10 28.6 % 187 1.10 4 128 40 31.3 % 44 l 1.20 2 91 39 42.9 % 41 j 1.30 1 53 23 43,4 % 24 1.40 1 17 8 47.1% 9 1.50 3 10 6 60.0 % 9 1.60 1 4 2 50.0 % 3 1.70 0 4 1 25.0 % 1 1.80 1 2 1 50.0 % 2 1.90 0 4 1 25.0 % ,.
I 2.00 0 1 1 100.0 % 1 2.10 0 1 1 100.0 % 1 2.30 0 1 0 0.0% 0
~
2.80 1 0 ~
0 #N/A 1
)
4.00 0 1 l 100.0 % 1 i 5.10 0 1 1 100.0 % 1 Totals 2242 523 164 2406 Totals > 1 V 14 318 125 '139 l
RFK.11/4/94.10.16 AM
[DCPJOINC XLS]DCP Joined C l
Table 3-3D: Catawba 1, SG "D", EOC 7 :
l Summary of 1993 RPC Inspection Results RPC Fraction RPC RPC Not RPC RPC Bobbin Volts Confirmed &
Performed Performed Confirmed Confirmed No RPC 0.2 15 0 0 #N/A I5 l
0.3 42 8 2 25.0% 44 l 0.0% 113 0.4 113 15 0 l
0.5 220 22 1 4.5% 221 0.6 297 23 4 17.4 % 301
~
0.7 300 32 3 9.4% 303 0.8 371 16 1 6.3% 372 ;
l 0.9 328 29 5 17.2% 333 1.0 230 62 6 9.7% 236 1.1 4 168 23 13.7 % 27 1.2 4 98 14 14.3 % 18 11 4 95 23 24.2 % 27 1.4 3 99 21 21.2 % 24
! 1.5 1 57 8 14.0 % 9 l 1.6 1 25 6 24.0 % 7
]'
l.7 2 15 4 26.7 % 6 l 1.8 2 10 3 30.0% '5 j i
1.9 0 5 0 0.0% 0 2.0 1 1 0 0.0% 1 2.1 0 4 1 25.0 % 1 2.2 0 1 0 0.0% 0 l
2.4 0 1 0 0.0% 0
~
2.5 1 l 0 0.0% 1 l 2.7 0 2 1 50.0% 1 2.9 0 1 1 100.0 % 1 3.0 0 1 1 100.0 % 1 4.8 0 1 0 0.0% 0 l Totals 1939 792 128 2067 l Total > 1 V 23 585 106 129 l
I l
RFK: 11/4/94.10 35 AM
, iDCPJOIND XLS)SG "D" Data i
l '
l
1 Table 3-4. RPC NDF INDICATIONS LEFT IN SERVICE AT BOC-7 AND BOC-8 BOC-7 RPC NDF Indications BOC-8 New RPC NDF Indications BOC? EOC-7 BOC-8 Tube TSP Bobbin Tube TSP Bobbin Bobbin RPC IU Volts Volts Steam Generator C R26C18 3 1.57 1.86 NDF R22C60 2 2.28 R22C22 2 1.21 1.R4 NDF R9C7 7 1.63 R46C71(0 3 1.94 1.82 NDF R23C7 3 1.45 RllC84(0 3 1.37 1,77 NDF R28C77 2 1.33 R43C73 3 1.34 1.69 NDF R12C85 3 1.32 R25C63 2 1.78 1.63 NDF R12C65 2 1.33 R10C7(4 2 1.53 1 60 NDF R4C93 3 1.30 R13C66 2 1.25 1.53 NDF R13C103 3 130 R7C84 2 1 39 1.50 NDF R6C18 2 1.29 R6C106m 4 1,43 1.48 0.95 R12C69 3 1.27 R39C68 2 1.66 1.45 NDF R46C68 2 1.25 R7C85 2 1.66 1.40 NDF R7C85 4 1.24 R7C84 3 1.37 1.38 NDF R12C74 2 1.24 R17C22 2 1.31 1.38 NDF R10C28 2 1.23 R7C85 5 0.% 1.34 NDF R12C69 2- 1.22 Steam Generator D R18C18 3 1.94 2.31 NDF R29C47 2 2.67 R21C25m 2 1.33 2.07 0.41 R2C17 2 2.48 R7C26 2 1.33 2.01 NDF R20C44 3 2.17 R3C19 3 1.88 1.96 NDF R13C30 3 2.06 R8C27 2 1.49 1.81 NDF R14C51 4 1.88 R8C15m 3 1.38 1.78 NDF R15C9 2 1.87 RllC61m 2 1.56 1,74 0.36 R14C56 2 1.86 R13C31 3 1.01 1.68 NDF R14C36 3 1.85 R10C57m 4 0.98 1.68 NDF R17C40- 2 1.79 R7Cl4 2 1.11 1.66 NDF R16C62 4 1.75 R7C60m 2 1.42 1.65 0.24 R15C9 3 1.75 DCPIR4 WPS 3-13 No =b= 16.1994
i Table 3-4. RPC NDF INDICATIONS LEFT IN SERVICE AT BOC-7 AND BOC-8 l (continued)
BOC-7 RPC NDF Indications BOC-8 New RPC NDF Indications BOC-7 EOC 7 BOC-8 Bobbin Ti.be TSP Bobbin Tube TSP '
Volts Volt:
1 Steam Generator D (continued) 1.23 1.65 NDF R18C6 2 1.74 l R9C5") 2 I
R14C48 3 1.72 R2Cl10) 2 1.28 1 61 2.01 1.61 NDF R17CB 2 1.71 R36C78 2 1.20 1.60 NDF R9Cl4 3 1.65 R7C26 3 1.18 l 1.60 NDF R13C4 2 1.61 R9C16 3 1.05 i
steam Generator A 1.37 1.22 NDF R48C62 4 1.66 R10C51 2 i
1.12 1.16 NDF R45C53 4 1.46 R49C75") 2 1.15 NDF R45C59 2 1.39 l R13C45"' 2 1.49 l 1.09 1.09 NDF R10C32 2 1.32 !
R36C95 3
~
1.30 1.07 NDF R45C47 7 1.32 R38C68 2 1.22 1.06 NDF R45C48 5 1.29 R27C64 2 R32C58") 2 1.15 1.03 NDF R17C47 4 1.27 R34C39 2 1.05 1.00 NDF R29C102 3 1.26 Steam Generator B R12C51 2 1.26 1,45 NDF R35C40 2 1.55 R10C37 2 0.82 1.42 1.89 R23C28 5 1.52 R21C18 2 0.99 1 26 NDF R36C92 2 1.47 1.23 1.24 NDF R35C35 2 1.44 R27C42 3 R15C74 3 1.12 1.18 NDF R26C52 2 1.4,3, R3tC56 5 1.25 1.17 NDF R8Cl2 2 1.43 4 1.04 1.17 NDF R31C55 2 1.41 R33C51 R12C51 3 1.25 1.07 NDF R31C56 3 1.39 4 1.25 1.07 NDF R26C55 2 1.39 R26C52 R31C46 5 0.90 1.06 NDF R30CS2 4 1.36 R10C98 4 1.17 1.04 NDF R26C56 5 1.34 Notes 1. Tube has been repaired for RPC confirmed TSP indication or other cause.
November 16.1994 DCP!R4 WP5 3-14
Table 3-5: EOC 7 RPC Inspections Results for RPC "NDF" Indications Left in Senice at BOC 7 Steam Generstor "11" Steam Generator "C" Steam Generator "D" All Steam Generators Steam Generstor "A" NDF m NDF m 1992 NDFm NDF m 1992 NDF m NDF m 1992 NDF m 1992 NIF in NDF m 1992 ,
Vohs NDF m m' m " " #* "'" ' " #* '" ' " # * **" 1992
- 1992 1992 1992 1993 Fme=1 m 1991 1992 1992 er 1993 1991 Fourwlir 1993 1993 Found in 1993 1993 Famd in 1993 Found in 1993 3 0 I I O 6 4 0 0.30 0 0 0 2 0 0 3 2 I O 8 4 1 0.40 0 0 0 4 1 0 3 1 1 0 I I O 5 5 0 20 12 0 0 50 0 0 0 14 6 0 2 0 2 2 0 $ $ 0 M 9 0 Ow 0 0 13 4 4 4 0 8 8 0 36 16 1 0 70 0 0 0 24 1 0 6 6 0 3 3 0 27 11 0 0 80 1 1 0 17 2 0 ?! 5 0 2 1 1 8 8 0 31 14 3 OW 0 0 17 2 8 6 2 20 18 2 46 27 5 1 00 I I O 1 to 0 29 22 7 35 31 2 78 67 9 1 10 4 4 0 10 0 25 18 7 32 28 4 62 St Il 1.20 2 2 0 3 3 0 18 13 $ 27 21 5 48 37 10 1 30 1 I O 2 2 0 32 27 3 36 31 3 1.40 0 0 0 0 0 0 4 4 I I 3 2 16 12 4 21 15 6 1.50 0 0 0 2 1 0 0 0 2 2 0 9 7 1 11 9 I IM 0 0 0 0 0 0 0 2 2 0 7 4 2 9 6 2 l.70 0 0 0 0 0 0 1 0 2 I I J 2 I 1 80 0 0 1 0 3 3 0 0 4 4 0 IM 0 0 0 0 0 . 1 1 0 0 0 0 0 0 0 l i O I I O 2.00 0 0 2.10 0 0 0 0 0 0 0 0 0 2 I I 2 I I 0 0 0 0 0 0 0 0 0 1 1 0 I i 0 2 40 38 3 116 92 24 216 184 25 470 323 52 Total 9 9 0 129 I 87 67 20 165 135 23 276 225 44 Total > 1 V 7 7 0 17 16
Table 3-6A: Summary of Prior and New Indications at EOC 7 End of Cycle 7 Volts Beginning of Cycle 7 Volts SG "A"
- '"* New Prior Ind. Not Net New Prior New New Ind.'s RPC Inspected In cations Found In& cations Indications Indications Co ed 0 0 0 2 74 -72 0.10 0 3 2 0 8 97 -89 0.20 4 30 8 2 15 85 -70 0.30 12 38 85 9 1 29 52 -23 0.40 175 32 5 76 36 40 0.50 31 202 26 0 109 25 84 0.60 57 200 22 1 147 1 146 0.70 58 140 24 2 148 5 143 0.80 32 26 84 18 0 165 4 161 0.90 9 35 7 1 117 3 114 1.00 6 15 14 0 69 I 68 1.10 3 6 6 0 46 0 46 1.20 3 5 5 1 28 1 27 1.30 3 3 0 10 1 9 1.40 1 0 0 0 6 0 6 1.50 1 0 0 0 0 2 0 2 1.60 0 0 0 4 0 4 1.70 1 0 0 0 0 0 0 0 1.80 1.90 0 0 0 0 1 0 1 0 0 0 0 0 0 0 2.00 2.10 0 0 0 0 1 0 1 282 983 176 13 983 385 598 Total 29 28 1 167 3 164 Total > 1 V 15
- '~' * ' - RFK 11/4.94.10 03 AM
Table 3-6B: Summary of Prior and New Indications at EOC 7 SG "B" End of Cycle 7 Volts Beginning of Cycle 7 Volts ew In s New Prior Ind. Not Net New Prior New New Ind.'s Its Indications Found Indications Indications Indications RPC Inspected 0.10 0 0 0 0 0 32 -32 0.20 0 2 1 0 3 65 -62 0.30 2 20 6 1 16 114 -98 0.40 14 52 7 0 42 80 -38 0.50 45 115 9 0 71 53 18 0.60 63 204 25 0 143 42 101 0.70 66 266 40 1 152 19 133 0.80 46 177 22 0 169 10 159 0.90 59 160 18 0 165 2 163 1.00 26 87 25 1 149 4 145
~
1.10 18 54 53 3 88 1 87 1.20 7 29 28 1 65 1 64 1.30 6 8 8 0 45 0 45 1.40 0 5 5 0 28 1 27 1.50 4 5 5 0 20 0 20 1.60 1 1 1 0 8 0 8 1.70 0 0 0 0 5 0 5 1.80 0 0 0 0 6 0 6 1.90 0 0 0 0 3 0 3 2.00 0 0 0 0 3 0 3 2.10 0 0 0 0 1 0 1 0 0 2.20 0 0_ 0 1 1 2.30 0 0 0 0 2 0 2 Total 357 1185 253 7 1185 424 761 Total > 1 V 36 102 100 4 275 3 272
[DCPJOINB XLSPOINED_0 XLS RFK.11/494. 9 57 AM
- _ - _ _ _ = _ _ -
Table 3-6C: Summary of Prior and New Indications at EOC 7 SG "C" End of Cycle 7 Volts Beginning of Cycle 7 Volts Prior New New Ind.'s New Pri'ar Ind. Not Net New its Indications Indications RPC Inspected '" " " '""""
Conf 0.1 0 0 0 0 2 57 -55 0.2 15 9 1 0 10 134 -124 0.3 26 54 7 1 29 133 -104 0.4 56 118 15 3 66 121 -55 0.5 99 187 8 3 118 70 48 0.6 146 270 22 3 181 23 158 0.7 187 305 20 5 231 40 191 0.8 178 264 10 1 237 11 226 0.9 143 164 12 1 215 13 202 1.0 120 92 10 2 168 3 165 1.1 82 50 49 9 112 6 106 1.2 52 41 41 12 85 3 82 1.3 38 16 15 2 55 5 50 1.4 14 4 3 1 27 1 26 1.5 12 1 1 0 22 2 20 1.6 5 0 0 0 9 0 9 1.7 3 1 1 0 3 0 3 1.8 2 1 0 0 5 0 5 1.9 4 0 0 0 2 0 2, 2.0 0 1 1 1 0 0 0 2.1 0 1 1 I 2 0 2 2.2 0 0 0 0 1 0 1 2.3 0 1 1 0 1 0 1 2.8 0 1 0 0 0 0 0 4.0 1 0 0 0 0 0 0 5.I 1 0 0 0 0 0 0 Total 1184 1581 218 45 1581 622 959
~
Total > 1 V 214 l18 1I3 26 324 17 307
. ~ a .~. .- w . e .cv c. ....r- ppg- 1114/94.1017 AM
Table 3-6D: Summary of Prior and New Indications at EOC 7 SG "D" End of Cycle 7 Volts Beginning of Cycle 7 Volts New ind 's Prior New New Ind.'s New Prior ind. Not Net New RK '" " "' " ""
Indications Indications RPC Inspected "
Confirmed 0.10 0 0 0 0 0 241 -241
"~
0.20 11 4 0 0 9 579 -
-570 17 .-
0.30 33 4 2 34 481 -447 0.40 73 55 7 0 75 326 -251 0.50 104 138 11 0 _ _ 146 136 10 0 60 154 165 14 4 173 79 94
~
0.70 160 172 14 2 189 40 149 0 80 198 I89- 9 1 229 21 208 0.90 187 170 7 2 157 9 I48 1.00 161 131 23 2 112 5 107
~
1.10 81 91 89 14 71 0 '
71 1.20 73 29 28 5 45 3 42 ~
1.30 52 47 45 12 17 0 17 1.40 68 34 33 8 9 1 8 1.50 38 20 20 4 4 1 3 1.60 21 5 5 0 4 0 4 i 76 14 3 3 1 1 0 I i
~
1.80 7 5 4 I I O I ;
1.90 2 3 3 0- 2 0 2 2.00 2 0 0 0 1 0 1 2.10 ~ 4 0 0 0 0 0 0 2.20 1 0 _.
0 0 1 0 __, l_
2.40 1 0 0 0 I_ 0 1 2.50 1 1 1 0 _
0 _
- 0 .__ ___
O
'2.70 1 I I 0 0 0 __ ___
0 2.80 0 0 0 0 0 0 0 2 90 1 0 0 0 0 0 0 3.00 0 1 1 1 0 0 0 Total 1448 1281 322 .59 1281 1922 -641 Total > 1 V 367 240 233 46 157 5 152
!DCPJOtND XLS10G TT Data RFK 11'4S4.10 42 AM
t ' ' '
M C
3 2
1 ,
1 4
9 .
/
4 1
1 1
K F
R .
j w"
e 6 2 9 8 2 5 N*N* 6 7 9 4 1 4 0 9 3 2 0 0 9 s 0 0 0 0 1 3 1 3 7 3 3 3 3 3 2 8 4 4 3 3 1 O 0 0 0 0 0 0 8 9 t
t 1 4 6 7 6 5 3 2 1 7 4 1 1 4 3 8 l
o e V N 7
7 e t o
N l
c C y i 9 0 3
5 7 8 O fC 4 5 3 9 5 8 5 O 0 0 0 0 0 0 8 n 0 7 1 7 9 6 0 7 4 2 1 8 7 6 4 3 0 0 0 0 0 0 0 0 0 3 3 2 .
I 4 8 8 5 2 1 1 3 E o r o
t g i r
a n i P
s m -
n r o g i i
t a B e w*
e
- 4 0 3
4 9
2 1
1 1
6 0
9 1
7 3
8 2
0 7
6 4
5 0
4 1
4 2
5 4 4 7
2 5
3 2
3 1
2 1 8 4 4 3 3 1 O 0 0 0 0 0 0
0 30 5
9 7
2 4
3 2
9 i
c N* 2 4 6 7 3 1 d
I n ,
- d w e e m 4 6 7 N fu 0 0 6 4 8 7 9 4 3 6 6 2 8
1 5
1 9 4 0 1 1 0 1 I 0 0 0 O 0 0 0 1 0 0 2 1
0 1 7 d *
- o n
n C ,
a r
i o t s '
r l 9 2 P o 5 3 4 5 8 0 7 5 5 5 3 4 6 6 0 0 0 0 1 0 0 4 7 f V 0 4 2 3 6 8 % 6 5 6 0 2
0 1 7 4 2 4 4 3 1 I 1 I 1
% 8 4 o
7 I
n -
y l e ,
r c a y m fC s
n o 0 6 m o 9
wi e a t
0 8 1
2 0
1 5
1 1
4 3
4 0
7 7 8 5 4
0 1
5 0 6 6 6 6 4 6 3 1 I 0 1 0 I 1 1 0 1 0 0 3 0
6 9 8 u d Ni dc 1
1 3 6 8 9 7 5 3 2 1 7 4 2 5 3 4 -
S n E n
- I E
6- s 3 n r i o 1 8 2 e ot 1 9 0 1 4 5 6 7 5 9 3 5 7 8 7 0 l i a 0 0 3 8 7 2 7 5 1 1 8 3 9 8 5 2 1 9 6 2 4 1 0 1 I 1 0 1 0 1 1 2 7 3 r 3 7 6 Pi dc 2 4 4 4 3 1 1 b 1 4 3 2 a n T I s V V V 0 5 1 G' '
" 1 2 3 4 5 6 7 8 9 0 1 2 3 4 $. 61 7 8 9 0 1 2 3 4 5 7 8 9 0 0 1 S 0 0 O 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 3 4 5 i l a ta a l
l t t l
o o o A T T T
~
6 3
h t
a T
i S
t K
Y R
U S
7 c
o r
t
- = _. . . . .-
i i I
l Table 3-7. Summary of RPC Confirmation for S/G C Group of Bobbin RC Indications In8Pected Confirmed l Indications Confirmed l
All Voltages Prior bobbin ind. 1068 189 95 50.3 %
not RPC NDF Pnor RPC NDF 116 116 24 20.7 %
ind.
New indications 1581 218 45 20.6 %
Sum of all ind. 2765 523 164 31.4 %
j Voltages > 1.0 volt l Prior bobbin ind. 127 118 79 66.9 % ;
not RPC NDF ,
Prior RPC NDF 87 87 20 23.0 %
ind. ,
New indications 118 113 26 23.0 % :
- Sum of all ind. 332 318 125 39.3% '
l .__
i l
DCPIR4 WP5 September 16,1994 l
I
I l
l i
! Table 3-8. Catawba-1, SG "C", EOC-6 Probabiltiy of Detection
- l c j
'93 RPC '92 RPC '93 New
'92 Voltage Confirmed Confirmed Bobbin Binned Bins '92 Bobbin '92 Bobbin Indications POPCD POPCD Indications & '92 '93 RPC i Left in Plugged Confirmed -
Senice l 0.50 6 0 10 37.5 % 37.5%
6/16 Ind.
0.60 8 0 2 80.0 % :
0.70 9 0 6 60.0 % !
88.6 %
0.80 7 4 2 84.6 % 109/123 Ind 0.90 13 21 2 94.4 %
l.00 25 22 2 95.9 % I i
1.10 16 15 13 70.5 %
l 1.20 21 12 5 86.8 %
81.1 %
1.30 11 3 0 100.0 % 86/106 Ind. l l
1.40 0 4 1 80.0 % l l.50 3 1 1 80.0 %
1.60 1 0 100.0 %
l l 1.70 4 0 100.0 %
87.5 %
1.80 1 1 50.0 % 7/8 Ind l 1,90 0 1
1
! 2.00 1 100.0 %
l 2.60 1 100.0 % 100 %
2/2 Ind.
2.70 1 100.0 %
3.60 1 100.0 % 100 %
1/1 Ind.
l l
Totals 119 92 45 - 82.4%
211/256 Ind.
I l
DCPIR 3 WP5 September 16,1994 I
l l
r l
l Figure 3-1: Catawba 1, EOC 6 Voltage Distribution i SG "C", All Bobbin Indications and RPC Confirmed Indications 400
" O AllIndications 350 E RPC Confirmed 300 -
9 250 "
i .E 1
i 200 - - - -
l C
- w .
E 33o _ _ _ _ _
5 z
100 ,
50 - - - - - - - -
0 : "; '; : : : ; ": ; E: b : ' : ^- : - - ' : : : .
- N 9 M n c h W e o - N q v n e o a e o e o e o o o o o o o o o a a a - a a a a a a H naa Bobbin Amplitude (Volts)
SG "D", All Bobbin Indications and RPC Confirmed Inlications 900 0 AllIndications 800 - 5 RPC Confirmed 700 -
E -
.e 600 - -
.s g 500 - -
C 400 - -
2 -
! 300 - - -
z -
200 -- - - - - -
100 - - - - - -
0 : : "; "+ : : ; ; :' ':~;-: : : : : :
o o o o o o o o o - - - - -*.- n- e-h-= no nN n=
9 N 9 *. n
- n a e o m N 9 Bobbin Amplitude (Volts) oce_ PLOT X13. Fg 31 AFK 3/2194
Figure 3-2: Catawba 1, EOC 7 ODSCC Analysis, All SG's 1600 0 Indications Found B RPC Confirmed E Repaired 1200 - -
m 8 1000 - -
- ii
.g ._ _ _ _
E CO 800 - - - -
y _.. __ _ _
.o
@ 600 - - - - -
z 400 -- - - - - - -
200 -- - - - - - - - - - -
0
+ - - - " + S+ i bht"l -+ l -- l- l- l- - --- l l-l l l l n n
- n 9 m a e o o o o o o o o - - - - - - - - - -
- o. -
n n *n
. w n a e N
- o. -
m n n *m. n m m n= m m m m a m v v n
- o. o = -
Bobbin Amplitude (Volts)
EOC7SMRY XLS prK 11/494
Figurt 3-3: Crtawb21, SG's "C" & "D" EOC 7 Vcitage Distdbutica l SG "C", All Bobbin Indications and RPC Confirmed Indications
! 500 l
450 0 AllIndications -
400 E RPC Confirmed J
l E Repaired l l g 350 - -
l l .$ l l
j 300 ,
l 1
l =o 250 - - - -
l l 6e l 2 200 - - - - -
z 150 _ _ _ _ _ _
100 - - - - - - -1 50 - - - - -
0 0';k ", -: , , ; ;"'; :- '- : ; .
N 9 9 9 9 D 9 E 9 "U 9 T. 9 9 6 9 E 9 ""
9 9 9, o o o o o o o o - - - - - - - - - - n n a n m l Bobbin Amplitude (Volts) l SG "D", All Bobbin Indications and RPC Confirmed Indications
! 400 0 AllIndications 350 -
E RPC Confinned ePaked l 300 - - -
E C
gm ,
1
= 200 - - - - - --
l C '
s.
f150
=
l z -
100 - - - - - - - ----
50 1
0 0 ,
s ,, I, f, ,,
, ,, k, , ,, a ,,b ,b ,, n. ,. . , _ , _ _ , _ , _ , _ _ , _ _ , __ , _ _
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.5 2.9 4.8 Bobbin Amplitude (Volts) ocP,,PLOTES. Fag 3 3 RFK 3/2MM l
i
Figure 3-4: Catawba 1, EOC 7 ODSCC Analysis, All SG's RPC "NDF" Indications Left in Service at BOC 7 80 -- -
O RPC "NDF" at BOC 7
~~
70
- - - -- 1 ORPC "NDF" @ BOC 7 & EOC 7 E Confirmed at EOC 7
_ _ _ _ _ _ _. _ . - - ~ ~ - - -
~
~
g___._ _.__
~ -
.! 50
.2 -
+
.o c -
l 40 1
[o --
= -
g
= _
=
y
_ = _ _ . _ _ . _ . .
? . _ _ _ _ .
.o
~
4 30
Z ; -
_ _ 3
~
~
__ _ _ _ . _ __ / _ __
1 __
._ g _ . . _ _ . . _ . _ . _ _ _ _ _ . . . _ _ ~ _ _ . .___.._ _ .
I
-[
~
? -
? - ---
20 -- -- :
{
+
_ g_
~
i _ _
- = =
- = _
~
= - - - - - - - -
$ 1 - - --
10
_ ~
s _
7_
- =
0 -
- --t -
"4 -
-t - -
-0 "1 -
--i " t -
- i -
-t -
t-+ - +- 4 -
"+ Et + -t m-tO~" i m - !
R ? R 8 R S 8 8 9 R R ? 8 8 R S 8 8 9 ?
6 6 6 6 6 6 4 A a a a A J u a a d d d 6
Bobbin Amplitude (Volts)
RFK 1114/94 errv 7enanv vi m rn 14
l Figure 34 (A): Cat wbs 1, All SG's, New '9312dicati:ns i and '92 Indications "NDF" in '93
- 900
. E New '93 Indications i g "
O'92 Indications NDF in '93 i
700 -
- "III i ! 600 -
4
- s m 500 - -
1IIIi l
i
?_
.Illil S
I
} - - -
- 300 --
. 2ee ._ _ _ _
lillili
! '" ~ - - ~ Elllili
! . .I I I,LLt.lltt....... ._
n e n:::r: var:oroann:
4 Bobbin Amplitude (Volts) j Figure 3-5 (B): Catawba 1, SG "C", New '93 Indications 4 and '92 Indications "NDF" in '93 l 250
! aI 5 New '93 Indications l
200 D'92 Indications NDF in '93 l _
150 E 9c:
t l _ _ _ __
I E
y 100 - -
e _. -
_l,l __ -
3
~- =
50 - - -
- ' ' - ' = ' - - ' -
0 - '
-n 'n','e'e'e.'e'e'o' 'n'n','e'e'e.'e'.'
6 d 6 6 6 6 6 6 6 4 4 4 4 5 4 4 4 4 4 d d d
'n'n Bobbin Amplitude (Volts)
DCP_ PLOT.XLS, Fg 3 S RFK. 3/28,94
Figure 3-6: Catawba 1, EOC 7 Voltage Distribution SG "C" Bobbin Indications Confirmed by RPC or Not RPC Inspected 500
.464 - - . . - -
I 450 419 . . . . . - -
389 400 350 c
2w .
$8 300 u 271 m '
c Co 250 y _. .
.o E 200 187
- $ 155 ~
- ~~ - - ~ ~ ~
150 -
100 70 44 50 -41 r
. . . . _ _ _ _l gN 9 __ 9 __ 3 2 1 1 1 1 1 1 1 0 . .
0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.40 1.50 1.60 1.70 1.80 1.90 2.00 2.10 2.80 4.00 5.10 Bobbin Amplitude (Volts) ioce.oncusi c=== .3 nne m.m 1
4.0 BOBBIN VOLTAGE INDICATIONS LEFT IN SERVICE 4.1 BOC-7 Indications Left in Service The indications left in service at BOC-7 are used to project the EOC-7 voltage distribution for comparisons with the actual distributions found at the EOC-7 inspection. Two representations of the BOC-7 distributions are utilized in this report. The Catawba-1 NRC SER requires that the indications found in the inspection be divided by a POD of 0 6 and then reduced by the indications repaired to define the BOC distribution. Bobbin indications found to be NDF by RPC inspection are conservatively included as indications left in service. The second distribution used in this report is the same as the NRC required distribution except that there is l no adjustment for POD. That is, the second distribution includes all detected bobbin indications not repaired independent of RPC confirmation. The BOC-7 distributions are given for SG C which was the most limiting SG with regard to the number and size of indications left in service for potential leakage considerations in Cycle 7.
The BOC-7 indications left in service for SG C are given in Table 4-1 and shown in Figure 4-1. The figure shows the NRC required distribution with a POD of 0.6 applied and all bobbin indications without a POD adjustment. Both distributions include RPC NDF indications left in service. Table 4-1 includes the BOC-7 distribution of RPC confirmed plus not RPC inspected indications. Since the POD adjustment is applied before the distribution is ,
reduced for repaired tubes, the POD adjusted distribution shows more indications than obtained by dividing all bobbin indications by 0.6. For example, the adjusted distribution includes 0.7 indication at 3.55 volt which is obtained by dividing the one indication (RPC NDF and l
repaired) by 0 6 to obtain 1.7 indications which is then reduced by 1.0 since the indication was j repaired. The effect of this conservative methodology is equivalent to assuming that a 3.55 volt indication was missed in the inspection even though the largest indications were found and repaired. The largest bobbin indication retumed to service was 1.9 volts. l 4.2 BOC-8 Indications Left in Service Bobbin voltage distributions for BOC-8 were developed in the same manner and for the comparable two distributions as described above for BOC-7. The reference distributions include all bobbin signals independent of RPC confirmation that were left in service. SG C is the most limiting SG for Cycle 8 leakage considerations although SG D is very similar in number and size of indications left in service. The determination that SG C was most limiting was made by performing leakage calculations for the actual EOC-7 voltage distributions.
SG C had a higher EOC-7 leak rate than SG D and it would be expected that the same result would be obtained at EOC-8. Therefore, SG C is judged to be the most limiting SG for Cycle 8 although differences from SG D are insignificant.
November 3. I M4 DCPIR 4 6 WP5 4-l
The BOC-8 bobbin voltage distributions are given in Table 4-2 and shown in Figure 4-2. It is {
seen that the POD = 0.6 adjusted distobution includes 0.7 indications at 5.1,4.0 and 2.8 volts {
that are not present in the unadjusted data. This results because the three largest indications l I
were repaired. The largest bobbin indication left in service at BOC-8 was 2.3 volts compared to 51 volts for the NRC POD = 0 6 model. The distributions of Figure 4-2 are applied in l Section 7 to obtain the projected EOC-8 voltage distributions for leakage and burst probability analyses given in Sections 8 and 9. ]
i l
l l
1
\
l l
l so e der 3.1994 DCPI R4.6 WP5 4-2 l
l l
Table 4-1: Catawba 1, SG "C", EOC 6 & BOC 7 Bobbin Voltage Distributions EOC 6 Inspection Results BOC 7 Indications Left in Service
' 'I Draft RPC Confirm No. of Bobbin All Bobbin Volts n ca ns Indcadons NUREG- & Not RPC Indications Indications Confirmed Repaired 1477 Inspected 0.1 88 0 3 85 143.7 66 0.2 240 6 7 233 393.0 186 0.3 324 14 14 310 526.0 249 0.4 358 33 11 347 585.7 280 0.5 247 28 8 239 403.7 201 214 24 14 200 342.7 153 0.6 _
0.7_ 173 42 10 163 278.3 50 0.8 93 32 9 84 146.0 26 0.9 91 42 23 68 128.7 22 1.0 47 22 25 22 53.3 0 1.1 37 15 17 20 44.7 0 1.2 22 12 13 9 23.7 0 l l
1.3 14 3 3 11 20.3 0 1.4 10 4 5 5 11.7 0 l 1.5 7 1 1 6 10.7 0 !
1.6 4 1 1 3 5.7 0 !
1.7 4 4 4 0 2.7 0 1.8 1 1 1 0 0.7 0 1.9 1 0 0 1 1.7 0 2.0 1 1 1 0 0.7 0 2.6 1 1 1 0 0.7 0 2.7 1 1 1 0 0.7 0 3.6 1 1 1 0 0.7 0 Totals 1979 288 173 1806 3125.3 1233 RFK: 10/29/94,4:26 PM
[DCPEOC6C.XLS] Tables
l 1
I Table 4-2: Catawba 1, SG "C", EOC 7 & BOC 8 Bobbin Voltage Distributions EOC 7 Inspection Results BOC 8 Indications Ixft In Service All Bobbin Draft No.of Bobbin " " * " I" "'
Indications Indications NUREG-1477 Inspected i
Volts Confirmed Repatred 0 0 0 0.0 0 i 0.1 0
~~
22 2 22 38.0 20 0.2 24 70 6 74 127.3 65 0.3 80 7 167 283.0 148 0.4 174 155 271 19 267 457.7 252 0.5 286 38 378 655.3 355 0.6 416 389 464 50 442 770.0 416 l 0.7 492 419 80 362 656.7 345 0.8 442 290 38 269 473.7 253 0.9 307 187 49 163 304.3 142 l 1.0 212 44 65 67 155.0 0 l 1.1 132 41 54 ~ 39 101.0 0 I 1.2 93 24 30 24 60.0 0 1.3 54 f 9 10 8 20.0 0 1.4 18 l 9 10 3 11.7 0 1.5 13 3 4 1 4.3 0 1.6 5 3 5.7 0 1.7 4 1 1 2 3 0 2.0 0 1.8 3 ~
2 2 4.7 0 l 1.9 4 1
~
1 0 0.7 0 )
2.0 1 1 1 0 0.7 0 2.1 1 1, 0 0 1 1.7_ 0 2.3 1 1 0 0.7 0 2.8 1 1 1 0 0.7 0 4.0 1 1 I
1 1 0 0.7 0 5.1 1 2406 473 2292 4135.3 1996 i Totals 2765 l
l l
i RFK 9/14'94. 610 PM l OCPJOfNC.XLS!DCP Joined C
k l Figure 4-1 (a): Catawba 1, BOC 7 Voltage Distdbution j SG "C", All Indications, pod = 0.6 4 _
i E 4
4 E
.2 400 j .s - - -
- i 300 3l 2
4
- m. .. ,
p _ _ _
$ 200 th .
- EEE
_ i
- O H..iIf..il..li..r.........._........
S.
O 9 6 9 I
9 O C N O O O T C 9 C9 CD C9 CE -9 == 9 N* O* M 9- b - N N N m
' Bobbin Amplitude (Volts) 4 Figure 4-1 (b): Catawba 1, BOC 7 Voltage Distdbution J
SG "C", All Indications Independent of MRPC Confirmation 500
.g 400 1 "
Co 300
.s.
= 200 ...
l-H+h _ ._ .
0-- l l l l : : : : : : ";";- -:-: : : -: : : :
- N - N nM 9 9 m 9 e 9 9 h 9 C C OC MC 9C 9C m9 C eo C C - - - - - - - - - - N N N M Bobbin Amplitude (Volts)
RFK 3/21/94 (DCP PLOT.XLS] Fig 4.1
t j
Figure 4-2 (a): Catawba 1, BOC 8 Voltags Distribution
- All Indications, pod = 0.6 4 -
t 700 I
i c
rh
- 3* E E.E1 g
85 E_ _ _a a E E
""' 400
- s E~E E E~
l EEEE _
}e " 8 i 1 200 I t + 1- hE
- o M
.. . i I1..I..H.1.1.h......._._.
- 9 9 9 9 o M N 9 ?
- 9 ". 9 9 o
a aM a 9n 9 o M o No 9 o? o o o o o o - - . . - - - . y e
! Bobbin Amplitude (Volts) i" Figure 4-2 (b): Catawba 1, BOC 8 Voltage Distribution All Indications Independent of MRPC Confirmation
- 800 i
e
! 700 i
- e 600 a c j .2 a
l 8 500
- Ea g i
400 i
1 %
o BE-
).
e j g._g_g_l_
3 Z 200 j _." - E E E._
j 100 4 _ _ . _
J
- O ;"
- : ; ; ; ; ; ;
o No 9o oT o*9o o to 9o 9-o-M -N * - - - - - N a a n , o M
- ";.
- *: 8 9 9 9 O
M 9 9 o
M Bobbin Amplitude (Volts) i RFK: 3,71/94 l [DCP_ PLOT.XLS] Fag 4.2 5
]
i 1
l
! Figure 4-3 (a): Catawba 1, BOC 8, Voltago Distribution i SG "D", AllIndications, pod = 0.6
! 600 E
! 500 1
400 a
J g 300 i
i I= -
l l 100 EE 4
1 E N i
E; ;;; ';;; ;E;";" - '-- ' - ' - :-- ' . .
l 0 ; ;
923239a0$20032505SS?%%222%0 Bobbin Aruplitude (Vohs) i Figure 4-3 (b): Catawba 1, BOC 8, Voltage Distribution l SG "D", All Indications Independent of MRPC Confirmation
! 600 l
l 500 l 400 L &
.I l g
! I=
i 1
E 100 us - m 0 "; ' ; ;
l l E252$$$$320333UUU$$2Ud20$MU Bobbin Amplitude (Vohs) i i
l l
! RFK 3/21/94 4 IDCP, PLOT.XLS] Fig 4 3 i
4
5.0 VOLTAGE GROWTIl RATES 5.1 Cycle 6 Voltage Growth Rates Voltage growth rates for Cycle 6 were developed at EOC-6 by the field analysts as described in WCAP-13494, Revision 1. Voltage growth rates were developed for all SGs and for each SG individually. The individual SGs could be independently analyzed due to the significant number of indications found in each SG. It was found that SG C had the highest growth rate and this growth rate was applied for the EOC-7 projections. The measured growth rates from Cycle 6 were adjusted to the projected EFPY for Cycle 7 by the ratio of 0.96/0.80, which is the ratio of the planned Cycle 7 length based on maximum fuel loading to the actual cycle length for Cycle 6. Figure 5-1 shows the SG C adjusted Cycle 6 growth to obtain the projected Cycle 7 voltage growth distribution. The largest growth value was 2.77 volts and found for only one indication. Eleven indications had voltage growth values > 1.0 volt. The Cycle 7 growth distribution of Figure 5-1 is applied to project BOC-7 voltage distnbutions to EOC-7 conditions for comparisons with actual distributions as described in Section 7.
Table 5-1 summarizes the average growth rates from prior cycles and the 1993 results for l
l Catawba-l. The Cycle 6 (1991 to 1992) average growth was highest in SG C at 0.13 volt or 27% of the BOC-6 average of 0.49 volt.
5.2 Cycle 7 Voltage Growth Rates l
Cycle 7 voltage growth rates were developed by the field analysts for all 1993 indications and growth distributions determined for each SG. For each 1993 indication, the bobbin data from the 1992 inspection were reevaluated to obtain a growth rate. As for Cycle 6, the largest growth rates were found in SG C, although the average growth rate is higher in SG D than the other SGs.
The Cycle 7 grcwth distributions for SG C and SG D are shown in Figure 5-2. The largest growth value in SG C is 4.18 volts, which is greater than the largest value of 2.77 volts projected for Cycle 7 based on Cycle 6 growth rates. The next largest growth of 318 volts is similar to the largest value projected for Cycle 7. The largest growth rate in SG D is 1.96 volts, which is the third-largest found in the 1993 inspection. Based on the 1992 (EOC-6) inspection, SG C was expected to have 11 indications with growth rates above 1.0 volt during Cycle 7. In the 1993 (EOC-7) inspection, SG C had 5 growth values above 1.0 volt and SG D had 6 values above 1.0 volt. With the exception of the two largest growth values in SG C, the observed growth rates for Cycle 7 are smaller than projected from Cycle 6, as shown by the cumulanve growth distributions for SG C shown in Table 5-2.
The Cycle 7 distribution for SG C shows a larger peak at the s 0 volt bin than observed in Cycle 6, while the SG D peak of the growth distribution occurs between 0.0 and 01 volts l DCPIR4 6 WP5 $-[ hen 1 W
The SG C growth distribution of Figure 5-2 is applied to project BOC-8 indications left in service to EOC-8 conditions in Section 7. For Cycle 8 applications, the Cycle 7 growth distributions are increased by the ratio of 390/347 = 1.12, which is the ratio of the EFPD for the fuel loading in Cycle 8 to the actual operating days of Cycle 7. For Cycle 8, the largest growth in the cumulative probability distnbution is 4.18 x 1.12 = 4.68 volts for SG C.
Table 5-3 shows the bobbin and RPC results for all indications with a growth rate 2 0.9 volts Also shown is whether the indication was a new indication reported in the 1993 inspection.
Seven (including the 4 largest growth values) of the 17 indications with the largest growth rates were confirmed by RPC,7 were NDF by RPC and 3 were not RPC inspected (tube was plugged for other causes). Six of the indications were new indications in the 1993 inspection and only one of the new indications was confirmed as a flaw by RPC inspection. The largest growth rate in SG B was 1.07 volts and the largest in SG A was only 0.71 volts.
Table 5-1 summarizes the average voltage growth rates for Cy:le 7 (1992 to 1993) The ;
average growth rate is negative for SG C (also found for SGs A and B) and is only 0.08 volts or 11% in SG D for Cycle 7, whic'a is smaller than the 27% found for Cycle 6. The average growths show the overall trend fo reduced growth in Cycle 7 con. pared to Cycle 6.
I l
I DCPI R 4-6 WP5 $.2 N=e* A m4
t f
i l
l Table 5-1. Average Voltage Growth Per Cycle for Catawba-1 l
Number Average BOC Average AV Cycle l Growth / Cycle EFPY Cycle Indications Voltage l
l Cycle 5: '90 '91 All Indications 126 0.74 0.10 l
l Cycle 6: '91 '92 0.80 All Indications 6941 0.55 :0.01 SG C 1970 0.49 0.13 Cycle 7: '92 '93 0.95 All Indications 8299 0.78 -0.06 SG A 1265 0.78 -0.17 l SG B 1542 0.82 -0.12 SG C 2765 0.81 -0.10 SG D 2727 0.74 0.08 i
l j
l l
l l
DCPI R4-6 WP5 $.3 NmMn i 1994 l
t l
i I
Table 5-2. SG C: Cyde 6 and Cycle 7 Voltage Growth
( Cumulative Probability Distributions Cycle 6 Cycle 7 Voltage Growth
- Voltage Growth
- AV, volts No.of Cum. Prob. No. of Cum. Prob. l Indications (%) Indications (%) .
O.0 545 27.66 1896 68.57 ,
0.I 399 47.92 450 84.85 0.2 301 63.20 219 92.77 0.3 228 74.77 108 96 67 0.4 195 84.67 47 98.37 0.5 119 90.71 19 99.06 0.6 76 94.57 8 99.35 0.7 52 97.21 6 99.57 ,
0.8 24 98.43 3 99.67 0.9 12 99.04 3 99.78
( i 1.0 8 99.44 1 99.82 1.1 1 99.86 1.2 4 99.64 1 99.89 1.3 1 99.93 1.4 2 99.75 0 99.93 1.6 2 99.85 0 99.93 1.8 2 99.95 0 99.93 l 2.77 1 100.00 0 99.93 3.2 0 100.00 1 99.96 4.2 0 100.00 1 100.00 ,
Notes:
(1) Includes scaling by 1.2 EFPY ratio for Cycle 7.
(2) Does not include scaling by 1.12 EFPY ratio for Cycle 8.
I DCPIR 4 6 WPS $.4 . b e ba A 1994
l l
Table 5-3. Summary of Largest Bobbin Voltage Growth Rates for Cycle 7 EOC 7 BOC 7 AV SG Tube TSP Bobbin Growth Bobbin RPC d*
Volts Volts 5 its Volts
~ l
- l. R7C41 2 5.03 3.76 0.85 4.18 No -
C 4.00 5.00 0.82 3.18 No C R25C99 3 No l D R8C65 2 2.89 1.80 0.93 1.96 j
2.69 2.36 0.94 1.75 No D R9C62 3 l R43C33 2.00 N . I .") 0.35 1.65 No ]
D 3 D R13C30 3 2.06 NDF 0.44 1.62 No R12Cl11 2.73 N.I. 1.43 1.30 Yes C 3 ;
2.47 N.I. 1.20 1.27 No 1 D R9C90 3 C R22C60 2 2.28 NDF 1.10 1.18 Yes C R6C5 3 'l.73 1.94 0.64 1,09 No 1
B R33C40 4 1.42 NDF 0.35 1.07 Yes D R20C44 3 2.17 NDF 1.15 1.02 No D R8C4 2 2.91 0.82 1.93 0.98 Yes B R30C52 4 1.36 NDF 0.42 0.94- Yes C R10C84 3 1.64 2.58 0.69 0.85 No A R48C62 4 1.66 NDF 0.95 0.71 No 1.23 NDF 0.60 0.63 Yes l A R27C58 3 Note (s) 1. N.I. = Not RPC Inspected. Tube was repaired for other causes and was not RPC inspected.
1 I
- "*6"*
oce.6 wn 5-5 l
1 l
i i .
1 I
l Figure 5-1: Catawba 1, Voltage Growth Projection for Cycle 7 j SG "C", All Indications t
t 600 l 545 l
i 500 i
e g 399
- .g 400 i I
.E . . . _
3 301
! C 300 i o
- y _ _ 228 g 195
= 200 Z
. _ 3 39 -. l
! M 76 l 100 --
.- 24 g - -- j 0 l l l l l l l l "; : - l --- l l -- i a o n n m e o e r- e e o n v e e o i 6 6 6 6 6 6 6 6 6 6 a a a a A W.
Bobbin Amplitude Growth (Volts) i i
i t l j Note: Cycle 6 growth rates adjusted to projected EFPY for Cycle 7 by ratio of 0.96/0.80 (planned Cycle 7 length / actual Cycle 6 length) l
\
J l
1
- 1 i i J l
'e j
(
4 i
i -
j l t
i i
i a 1 t'
i KP PLOT XLS] Fig 51 RFK 3.'21194
Figure 5-2 (a): Catawba 1, SG "C", ODSCC Data Analysis Growth Rate, BOC to EOC 7, TSP Indications 600 100 %
500 -
-- 80%
l So 400 .!
-- 60%
! .) 33:
- i C 300 a 3 50 % y 4
=
i o j zos .
21' -- 40%
" =
$ 200 -- 30% U z
120 .
- -- 20%
g ,, n .
47
-- 10%
43 3 3 i : :
i 1 1 5 6 4 l"8' s 6 m- -- : 0%
0 : . .
49E997$$$$ISISSIE...222525EEE302220 ODSCC Bobbin Amplitude lncrease (Volts)
Figum 5-2 (b): Catawba 1, SG "D", EOC 7, ODSCC Data Analysis Growth Rate, BOC to EOC 7, TSP Indications 694 700 - 100 %
~
-- 90%
~~
506 509 n
$500 -- 70% $
=
-- 60%
400 50 % o
'E 271 g'300 -- 40% m g 226 (. . _.
200 E ~~
155 108 -- 20%
100 64 14 8 ~ 3 1 l~~
2 1 ~'
5 '4 ~ ~1 5 ~ 4 ~ 7 ~ 12 - 1
@$$$$$4$553222333EEE322O53 ODSCC Bobbin Amplitude (Volts)
AFK; 3/21/94 fDC A PLOT XL.S1 Fq 5 2
i l
6.0 NDE UNCERTAINTIES NDE uncertainties for voltage measurements were developed for Catawba-1 h WCAP-13494, l
Revision 1. Section 5.7 for Cycle 7 and in WCAP-13854, Section 4.4 for Cycle 8. The NDE uncertainties between Cycles 7 and 8 differ due to the inclusion of the probe wear standard in the EOC-7 inspection and the probe wear standard was not used in the EOC-6 inspection The net NDE uncertainty developed in WCAP-13854 for Cycle 8 is essentially the same as developed in the EPRI Report TR-100407, Revision 1 (Draft of August 1993).. While the WCAP and EPRI reports differ slightly in the components of the net NDE uncertainty and in the development, the net NDE uncertainty is not significantly different (12% standard deviation in the WCAP versus 12.5% in the EPRI report) and the WCAP NDE uncertainty is applied for both Cycles 8 and 9 for Catawba-l.
For Cycle 7, the NDE uncertainty, without application of a probe wear standard, was a standard deviation of 16% with no cutoff for probe wear. The analyst variability uncertainty '(
has not changed between Cycles 7 and 8. For the reanalyses of Cycle 7 in this report, the l
analyst variability is applied as a 10% standard deviation with a cutoff at 20% as described i below for Cycle 8.
I
! For Cycle 8, the NDE uncertainty is principally due to probe wear with a standard deviation of l approximately 7% about a mean of zero and analyst variability with a standard deviation of 10%. These distributions are applied as normal distributions and combined to obtain a net NDE uncenainty of 12% for one standard deviation. The upper bound on the probe wear ;
uncertainty is limited to 15% by the Catawba-1 IPC requirement to replace the bobbin probes when measurements on the probe wear standard for a worn probe differ from that found for the new probe by 15% The upper bound on the analyst variability uncertainty is limited to 20%
by the eddy current analysis guidelines which require lead analyst resolution of bobbin voltages (with one or more reported above 1.0 volt) differing between analysts by more than 20%. An l upper limit on the net NDE uncertainty of 25% is obtained by combining the 15% and 20%
upper limits by the square root of the sum of squares.
The net NDE uncenainty applied to obtain projected EOC-8 voltages is then a 12% standard deviation about a mean of zero with 2 maximum cutoff at 25%. Normal distributions are l
applied for the NDE uncertainty. 1 I
i l
l l
DCPIR4 6 WP5 6-l W ebuEim l - -.-
7.0 PROJECTED EOC VOLTAGE DISTRIBUTIONS 7.1 Projected EOC-7 Voltage Distributions Consistent with the methodology of Reference 3 (as suggested in the Catawba-1 SER), Monte Carlo analyses are applied to develop projected EOC distributions from the BOC distributions.
The BOC voltages are increased by allowances for NDE uncertainties (from Section 6) and voltage growth (from Section 5) to obtain the EOC values. In the Monte Carlo analyses, each voltage bin of the BOC voltage distributions (Figure 4-1 for example) is increased by a random ,
l sample of the NDE uncertainty and growth distributions to obtain a EOC voltage sample.
Each sample is weighted by the number of indications in the bin. The sampling process is repeated for each BOC voltage bin and then repeated for a large number of samples across the BOC distribution. In the present analyses,10,000 samples of the BOC distribution were applied to obtain the EOC voltage distributions. Larger samples are used to obtain the Monte Carlo burst probabilities from the EOC distributions. The EOC projections were performed for SG C which is the limiting SG for both Cycles 7 and 8.
The projected EOC-7 bobbin voltage distributions are given in Table 7-1 and shown in Figure 7-1 for the three categories of BOC distributions discussed in Sections 4.1 and 10.0.
For the POD = 0.6, adjusted distribution, the maximum projected EOC-7 voltage is 4.0 volts.
Since the Monte Carlo analyses yield a cumulative probability distribution of EOC voltages, a method must be defined to obtain a discrete maximum EOC voltage value. The method adopted in this report is to integrate the tail of the Monte Carlo distribution over the largest 1/3 of an indication to define a discrete value with an occurrence of 0.33 indications. For N indications in the distribution, this is equivalent to evaluating the cumulative probability distribution of voltages at a probability of (N-0.33)/N. The largest voltages for all distributions developed by Monte Carlo in this report have been obtained with this definition for the maximum EOC discrete voltage. The largest EOC-7 voltage for the distribution of all indications without POD adjustment (POD = 1.0) is 3.2 volts. The new indication method described in Section 10 results in a maximum EOC-7 voltage of 4.2 volts. In the 1992 SLB ,
leak rate evaluation, a maximum EOC-7 voltage of 2.65 volts was projected for all indications l without POD adjustment. The difference between 3.0 volts and the 2.65 volts of this report is due to the method (1/3 of an indication in this report) used to define the reported voltage.
7.2 Comparison of Projected and Actual EOC-7 Distributions l For purposes of comparing projected and actual distributions, it is necessary to consider the l applied purpose of the projections in order to define the actual distributions appropriate for the comparison. This results as the Catawba-1 EOC inspection results include many RPC NDF indications. The projected EOC voltage distributions for IPC applications are applied to project SLB leakage and tube burst probability. As discussed in Section 3.5 in developing the EOC-7 distribution for comparisons with projections, RPC NDF indications have a negligible kmte M, Im DCPIR7 8 WP5 7-I
likelihood of potential SLB leakage over the prior cycle and should be ignored in comparing IPC projections with the actual distributions. Thus the comparisons of IPC projections with
, actual distributions are made for EOC-7 RPC confirmed indications summed with indications not RPC inspected since it cannot be stated with confidence that the latter indications would not have leaked over the prior cycle. The appropriate SG C EOC-7 actual distributions for this comparison were given in Figure 3-6.
Figure 7-1 and Table 7-1 show the comparison of projected (POD = 0.6, POD = 1.0, new indication method) and actual EOC-7 voltage distributions (excluding RPC NDFs as discussed above). The NRC POD = 0.6 and new indication models lead to projected voltages of 4.0 and 4.2 volts, respectively, and generally lead to similar EOC-7 voltage distributions even though the method of developing the BOC distributions is substantially different between these two methods. The POD = 1.0 method leads to a maximum EOC voltage of 3.2 volts. All three projection methods conservatively bound the actual EOC-7 voltage distributions above 1.0 volt except for underestimating the maximum actual voltage indication of 5.1 volts. The POD = 0.6 and new indication methods are in good agreement with the actual distribution below 1.0 volt while the POD = 1.0 method tends to underestimate the low voltage distribution.
Underprediction of the low voltage indications would be expected for POD = 1.0, since this method does not include an allowance for new or undetected indications which tend to be low voltage indications. Overall, the agreement between projected and actual distributions is very good, particularly for the POD = 0.6 and new indication methods. This supports the projection methods acceptability even for the Catawba-1 case with a very large number of bobbin indications. Similar agreement between projections and actuals has been obtained for other plants (three other plants with IPC assessments comparable to this report) although the POD = 0.6 method has been found to be excessively conservative for the other plants evaluated.
Figure 7-2 shows the comparisons between projections and the actual EOC-7 distribution for all bobbin indications including RPC NDP indications left in service. It is seen that the projection methods remain adequately conservative even when compared to this EOC distribution, which is unnecessarily conservative for IPC applications.
l 7.3 Projected EOC-8 Voltage Distributions i
The BOC-8 voltage distributions are described in Section 4.2 and Figure 4-2 with and without the POD = 0.6 adjustment included in the distributions. Section 10 describes the BOC-8 ;
distribution for the new indication method. Monte Carlo methods are applied using the Cycle 7 voltage growth distribution of Figure 5-2 to obtain the projected EOC-8 distributions.
As above for Cycle 7, projections to EOC-8 have been made for three distributions including the NRC model with a POD = 0.6, all bobbin indications left in service including RPC NDF (POD = 1.0) and the new indication method of Section 10.
DCP!R7-8 WP5 7-2 smen a. w
Figure 7-3 and Table 7-2 show the projected distributions for all three definitions of the BOC distribution. De NRC model projects to an EOC maximum voltage of about 5.7 volts compared to 5.1 volts for POD = 1.0 and the new indication method. ' The larger projected maximum EOC voltage results from the RPC NDF indication of 5.1 volts found in the inspection even though this indication was removed from service. The largest bobbin indication left in service was 2.3 volts. The NRC model yields a maximum projected voltage larger than that found in the inspection independent of whether or not the tube was repaired. This results as the POD adjustment is unrealistically assumed to be independent of voltage.
The EOC-8 distributions given in Figure 7-3 are applied to SLB leakage analyses in the following section.
P I
i I
DCPIR7-5 WPs 7-3 wa=t,- 4. m4 l
Table 71. SG C: Actual and Project 3d EOC 7 Voltage Distnbutions Actual EOC-7 Proiected EOC-7 Volts All Bobbin RPC Conf. All Bobbin All Bobbin New Ind.
Indications + Not Insp. POD = 0.6 FOD = 1.0 Method 0.1 19 11 10 02 24 22 106 61 59 03 80 70 236 138 153 04 174 155 335 198 284 0.5 286 271 392 231 387 06 416 389 374 221 390 07 492 4 64 361 212 361 08 442 419 310 181 301 09 307 290 257 149 245 10 212 187 201 115 186 1.1 112 44 151 84 133 12 93 41 112 61 95 13 54 24 78 41 62 14 18 9 54 29 41 1.5 13 9 39 19 28 l 16 5 3 26 14 17 1.7 4 1 18 8 11 1.8 3 2 12 6 7 1.9 4 1 9 5 6 l 20 1 1 6 3 4 j 21 1 1 5 2 3 2.2 3 2 3 l l
2.3 1 2 1 1 l 24 2 2 25 1 1 1 26 1 27 1 28 1 1 1 1 29 07 I 30 1 3.1 1 32 03 3.3 1 34 1 36 07 _
3.8 07 40 1 1 0.3 42 0.3 5.1 1 1 Totals 2765 2406 3114 1794 2795 DCP!R7-8 WP5 7-4 November 11,1994 l
Table 7 2. C:tawba-1 SG C: Projected EOC-8 Voltage Distributions Volts All Bobbin All Bobbin New Ind.
POD = 0.6 POD = 1.0 Method 02 14 8 8 0.3 67 39 37 O4 179 105 104 05 338 198 252 06 507 293 4?9 i 0.7 620 356 544 O8 644 365 555 09 543 304 451 10 412 -226 316 11 294 156 195 1.2 191 96 106
]
13 122 58 55 i
14 76 34 28 l
15 44 19 15 q 16 28 12 9
'17 18 7 6 l 18 10 5 4 1.9 8 3 2 1 2.0 - 4 2 2 21 4 1 2 22 3 1 1 ;
i 23 1 1 i 24 1 1 25 1 1 ;
2.7 8 29 1 i 30 1 3.4 1 35 j 3.6 1 j 42 1 1 4.5 1 4.6 1 48 0.7 l J
50 0.7 03 !
5.1 1 52 03 53 0.7 57 0.3 Totals 4136 2292 3076 l
i DCPIR7 B WP5 7-$ Wudu 4. W4 I l
i l
, 1 l Figuro 7-1 l Cetawba 1: S/G C Comparison of Projected and Actual EOC-7 Voltage Distributions l
' Actual RPC Confirmed Plus Not RPC Inspected i
- 500 i !
l l POD = 0.6 5
i E ! j j S 3oo .;
l 1
1 3 r '
! O f l j
" E
-l l 1
i E 200 - -
J 1
l
! o' l z r i
)
l , i
) joo ,\
3 , _ _ .
0
'I - - -
I ' ' " - - --
! 01 03 05 07 09 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 di 43 45 47 49 51 4 Volts 1
i 500 i l i POD = 1.0 i
? 4M -
I e m j C O
- w 300 _ _
4 8 i 5
_c o 200 , -- -
o ,
l,i l
- z [ l i i t
I g l -
1M i l
' l j , ! r E '*---
0
- 01 03 05 07 09 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 4 Volts i
t 4
500 New indication Method i
1 400 -
! T i
l
! E ,
300 -
l 0 '
]
1 -
l "c ! Projected EOC-7 l ~
o 200 , I r
{
, o Actual EOC-7 RPC Cont + N I
- z i l !
l :
j 100 j 1
- < f ; ,
O E ' ' " - - - - - - --
)
01 03 05 07 09 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51
' Volts i ^
7-6 i
Figuro 7-2 Catawba 1: S/G C Comparison of Projected and Actual EOC-7 Voltage Distributions for All Bobbin Indications 500 ROD jP =0.6
-- - -4 - - - - - -- - - - - - - - -
400 to l C l
.9 l p- -- -- - ---- -
} 300 lp --
. m , i i S i I
3 200 i
- 4 i o ; I ! l i z !
l 1 j 100 4 j 4 i l 0' E ^"^----
! 01 03 05 07 09 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 J
Volts j 500 m i
3 POD = 1.0 - - - - - - - - - -
400 l 1
I .9 '
g 300 l
i s ;
i s j B 200 -- -- ' - - - - - -
4q l 6 1 i
i z . ! ; i l
! 100 l --r !
4 i
) - - - - -
- - - - -- + l j ,
,l i 1
"- E' " -
0' - - - -
- 01 03 05 07 09 11 13 15 17 19 21 23 25 2.7 29 31 33 35 37 39 41 43 45 47 49 51 j Volts i
!. 500 i
i New Indication Method l 400 f--
) :
I! E I !
O _ _ _ _ .-- _ _
e 300 _ _ _ _ _
' 7 i .0 . ) i y , , '
! 5 E 200 h- - -
. 1 i 6 !
r ! l i
- z .
E Projected EOC-7
! l ; i :
l 4
[
4 100 i l
- - I
? r j , ! ,
i ,
- l l l Actual EOC-7 (RPC Conf + N I ) l 0' " "- LL* --- - -
01 03 05 07 09 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 40 51 Volts 3
1 7-7 i
i I
f
) Figura 7-3
- Cetawba 1
- S/G C Projected EOC-8 Voltage Distributions j 1000
- POD = 0.6 j
l 800 1 m C
O
-[
i
% 600 -
! .9 l E 1
o 400 4 i
o' !
, I Z ;
'I j 200 -
4 i i -
0
' 'I " " - ~
! 01 03 05 07 09 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 Volts l
l 500
} POD = 1.0
} 400 3 C o
3
- 300 -
.h -
- u
_C o 200 i d
z 100 4 4 0
01 03 05 07 09 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 Volts 800 New Indication Method 600 -
.9 lii
.9 y 400 o
6 Projected EOC-8 z '
l 200 4 4
ir l l BOC-8 0
--4 s^
lJ..__
- 01 03 05 07 09 11 13 15 17 19 21 23 25 17 29 31 33 35 37 39 41 43 45 47 49 51 53 Volts
i 8.0 SLB LEAK RATE ANALYSES This section provides the results of SLB leak rate analyses for EOC-7 and projected EOC-8.
Leak rates are presented using the draft NUREG-1477 methods and using the EPRI leak rate j versus voltage correlation. The database used for the analyses of this report are described in l l
Section 8.1.
i 8.1 Database Applied for SLB Analyses l This section identifies the database supporting the alternate repair criteria burst and leak rate j l
correlations as applied for the Catawba-1 SLB anakses. The database for 3/4 inch diameter 1 tubing is described in EPRI Report NP-7480-L, Volume 2 (Reference 1). However, at the February 8,1994 NRC/ Industry meeting, the NRC presented resolution of industry comments on draft NUREG-1477. The NRC identified guidelines for application ofleak rate versus i voltage correlations and for removal of data outliers in the burst and leak rate correlations. The EPRI outlier evaluation for 3/4 inch tubing has been provided to the NRC in WCAP-14046 (Braidwood-l IPC). The EPRI database and APC correlations of this WCAP are applied in this report for analyses using the EPRI SLB burst and leak rate correlations as a function of voltage.
However, the Catawba-1 SER specifies the database to be used for SLB leakage analyses and this "NRC database" differs for three specimens from the EPRI database of WCAP-14046. The SER specifies that model boiler specimens 598-1 s.nd 598-3 are to be included in the leakage database. In addition, the SER requires that Plant S pulled tube R28C41 should be included at l
the leak rate calculated with the Westinghouse CRACKFLO code. At the time of the SER, a
' preliminary estimate of 2496 liter /hr was identified for R28C41 at the SLB pressure differential e of 2560 psi. For the SER required leakage analyses using draft NUREG-1477 methodology, these three data points are included in the analyses and described as the NRC database.
A more detailed analysis for the R28C41 SLB leak rate is givcn in WCAP-14046 with a resulting leak rate of 1251/hr. This lower leak rate is included in the EPRI database as applied for the leak rate correlations in this report. Thus the differences between the NRC database and ,
l 4
the EPRI database, as applied in this report, are that the NRC database includes specimens 598-1 and 598-3 which are not in the EPRI databt.se and the NRC database uses 24961/hr for R28C41 while the EPRI database uses 1251/hr. SLB leak rate analysis results are given for both the EPRI and NRC databases applying the EPRI leak rate correlation and draft ,
NUREG-1477 methodology as required by the Catawba-1 SER.
For the burst pressure correlation, there is very little difference between the NRC and EPRI databases. The difference is that the NRC database includes model boiler specimen 598-1.
Burst probability calculations given in Section 9 use the EPRI database. Comparative burst probability analyses with both databases result in differences of less than 25% with the higher value between databases dependent on the voltage distribution being evaluated.
DCPIR7-5 WP5 8-) Vovember 4.1994 o
The draft NUREG-1477 leak rate methods (Section 8.2) apply the leak rate data averaged l
independently of voltage. The NRC database leads to an average leak rate of 149.21/hr with a standard deviation of 503.11/hr for use with the NUREG methods.
8.2 Alternate Leakage Analysis Methods Two methods for leak rate analyses are applied in this report. These include the draft NUREG-1977 methodology as required by the Cttawba-1 NRC SER and the appbeation of the EPRI leak rate versus voltage correlation as described in the previous section.
Draft NUREG-1477 Methodolory The NRC methodology of draft NUREG-1477 obtains the number of indications at the beginning of an operating cycle as:
N = N, + N" ' - N' = N' + ( 1 - oD ) y _ y' , N, _ y' pod (8-1) pod l
l where, N, = number of detected bobbin indications at the end of the preceding cycle, N, = number of repaired indications at the end of the preceding cycle,
! N., = number of indications not detected by the bobbin inspection, and pod = probability of detection (0.6 for NUREG-1477 methodology).
4 The above adjustments for pod have been incorporated in the BOC and EOC voltage distributions developed in Sections 4 and 7 so that no further adjustmwes are required for the leakage calculation. Section 3.3 of draft NUREG-1477 states that the to:rJ leak rate, LR, should be determined as:
LR = p P + Z cpp2p _ { y,p,2 2
, (8-2)
- % < l where, l-( P = { N, P, , (8-3) l is the expected number of indications that leak summed over all voltage bins, and l l
p = mean of the leak rate data independent of voltage o = standard deviation of the leak rate data independent of voltage P, = probability that a tube leaks for the i* voltage bin _
N, = number of indications (after pod adjustment) in the i* voltage bin l 2 = standard normal distribution deviate corresponding to the one-sided level of. l confidence on the total leakage rate.
DCPIR7-8 WP5 8-2 Nove=ber 4.1994
l l
For the total leakage, the first term of the above equation represents a mean expected leak rate while the square root term is an effective standard deviation for the total leakage. Draft j NUREG-1477 recommends that Z be applied as 2 which corresponds to a level of confidence of j 98%. Leakage data for the above equation are described in Section 8.1 above.
Leak Rate Correlation Methodolorv l
The leak rate versus voltage correlation can be applied by Monte Carlo methods in conjunction l l
with the EOC voltage distributions obtained by Monte Carlo methods or by applying the POL l correlation and leak rate correlation to the EOC voltage distribution obtained by Monte Carlo l methods as applied for the draft NUREG methodology. It is shown in WCAP-14046 that both the full Monte Carlo leak rates and the application of the correlations to the EOC voltage distribution yield essentially the same leak rates. A similar equivalence of the two methods was found for a Catawba-1 analysis. Thus it is adequate to apply the correlations to the EOC voltage distributions. ,
8.3 Projected EOC-7 SLB Leak Rates l SLB leak rates were calculated for the projected EOC-7 voltage distributions of Figure 7-1.
The results are given in Table 8-1. The full NRC methodology with a POD = 0.6 adjustment and draft NUREG-1477 leak rate independent of voltage yields a SLB leak rate of 20.6 gpm for i the NRC database and 13.9 gpm for the EPRI database. Applying the POD adjusted EOC l voltage distribution with the APC leak rate correlation and deterministic analysis yields 0.81 and 0.081 gpm for the NRC and EPRI databases, respectively Without the POD adjustment, the SLB leak rates are 13.0 and 0.38 gpm for the draft NUREG and APC leak rate methods
! with the NRC database. This difference (factors of about 20 to 30) is typical of the differences between the NRC method which ignores the leak rate dependence on voltage and methods applying the leak rate versus voltage correlation which shows a large reduction due to the strong dependence of leakage on voltag3 Both methods utilize the same probability of leakage correlation. Application of POD = 0.6 is seen to result in about a 50% increase in the leak l
rates compared to POD = 1.0. Following the 1992 outage at EOC-6, the projected SLB leak rate at EOC-7 was < 0.1 gpm based on applying the APC leak rate versus voltage correlation to l
l all RPC confirmed plus not RPC inspected indications left in service without a POD l adjustment. This result is consistent with that found for the EPRI database in Table 8-1 although lower than found by applying the NRC database.
Also shown in Table 8-1 are leak rates obtained with the new indicatie method developed in Section 10 for defining the BOC distribution. This methodology accounts for new and
' undetected indications based on the prior cycle distribution of new indications and estimates the RPC confirmed EOC distribution. Due to the large number of new indications found at EOC-6, which was the first IPC application, the new indication method results in higher leak rates than obtained for a POD = 1.0 and the results are similar to that obtained for the POD = 0.6 analysis.
8-3 m ember 16.1994 DCPIR7-5 WP5 i
4 8.4 Comparison of Leak Rates for Projected and Actual EOC-7 Distributions SLB leak rates were also calculated for the actual EOC-7 voltage distribution of Figure 3-7 which includes bobbin indications confirmed by RPC plus indications not RPC inspected. This distribution and leak rate represents the target that the projections are attempting to predict.
The resulting leak rates are given in Table 8-1. The leak rate for the NRC methodology is 13.0 gpm compared to 20.6 gpm for the projected methods with the POD adjustment. Without the POD adjustment, the projected leak rate is 13.0 gpm which is in good agreement with the leak rate for the actual distribution.
When the APC leak rate correlation is applied, the leak rates calculated for the actual distribution are higher than or about equal to the projected leak rates. This results as the projections underestimate the maximum EOC-7 bobbin indication of 5.1 volts as noted in Section 7.2. For the small leak rates resulting with the application of the leak rate correlation, the single largest indication has a significant influence on the total leak rate due to the strong dependence of the leak rate on voltage. For this type of difference between projected and actual voltages being largely due to a single indication, it is more appropriate to look at the absolute difference between leak rates rather than percentage differences. Based on the Monte Carlo analysis results of Table 8-1, the projection methods underestimate the leak rate obtained from the actual voltage distribution by less than 0.1 gpm, which is a negligible difference.
4
~
Overall, good agreement is obtained between projections and leak rates calculated from the actual distributions. For the draft NUREG-1477 methods, which are strongly dependent on the number of indications, the projection methods are conservative. For the APC leak rate correlation methods, the projections agree with the leak rates from actual distributions to with a few tenths of a gpm even though the maximum EOC-7 voltage indication was underestimated by the projection methods.
- 8.5 Projected EOC-8 SLB Leak Rates 4
Leak rates at EOC-8 were calculated in the same manner as described above for EOC-7. The results are given in Table 8-1. Per the Catawba-1 SER, the reference SLB leak rate at EOC-8 using the draft NUREG-1477 methodology and the NRC database is 27.6 gpm. This result t includes +he POD adjustment and RPC NDF indications left in service. During the outage, the projected leak rate was calculated to be 29.4 gpm. Differences between the 29.4 and 27.6 gpm values are due to finalization of the voltage distributions and growth rates. The projected leak i rate is below the allowable leak rate limit of 30.0 gpm agreed upon by the NRC in the Catawba-1 SER.
P is seen from Table 8-1 that the results based on the leak rate versus bobbin voltage correlation yield about a factor of about 20 lower leak rates than the NRC methodology.
Applying the NRC database with POD = 0.6 for the voltage distribution and Monte Carlo DCPIR7 8 WP5 84 November 16,1994 i
analyses for the APC leak rate correlation yields a SLB leak rate of 1.61 gpm. This analysis method and database was accepted by the NRC in the Braidwood-l IPC SER and would be acceptable under the requirements of the NRC draft generic letter for ODSCC at TSPs.
Application of the APC leak rate correlation for Cycle 8 at Catawba-1 would eliminate the need for the coolant activity reduction from 1.0 to 0.58 microcurie per gram dose equivalent I-131 currently applied for Cycle 8.
i i
I l
1 l
l l
Novmba 16,1994 DCPIR7-8 WP5 85
i I
i 1
1 Table 8-1: Summary of EOC-7 and EOC-8 SLB Leak Rate Analyses: SG C SLB Leak Rates (gpm)
Draft NUREG- APC Leak Rate EOC Voltage Distribution 1477 Method Correlation EPRI NRC EPRI NRC Database Database Database Database Determ. M. C. Determ. M. C.
EOC-7 Results
- Actual Voltage Distribution
- RPC confirmed plus not RPC insp. 8.84 13.0 0.18 0.14 1.34 0.57
- All bobbin indications 10.7 15.8 0.18 0.17 1.36 0.55
! - Projected Voltage Distributions
- All bobbin indications, POD = 0.6 13.9 20.6 0.081 0.10 0.81 0.57
- All bobbin indications, POD = 1.0 8.83 13.0 0.030 0.38 i
- New indication method 18.4 26.0 0.074 0.08 0.58 Projected EOC-8 Results All bobbin indications, POD = 0.6 18.6 27.6 0.34 0.36 2.42 1.61 All bobbin indications, POD = 1.0 11.1 16.4 0.15 0.11 1.16 0.51 New indication method 12.3 18.2 0.20 0.17 1.52 0.81 Note: Determ. = Deterministic; M. C. = Monte Carlo DCPIR7.s WP5 8-6 Nonmbn 16.1994 I
i i
- 9.0 SLB TUBE BURST PROBABILITY ANALYSES 1
j 9.1 Tube Burst Probability for Limited TSP Displacement !
l Analyses for TSP. Displacement under SLB conditions have been performed specifically for j Catawba-1 as described in References 2 to 4. These analyses based on the conservative *
- assumptions of zero friction at the tube to TSP crevices and at the TSP wedge to SG wrapper contact points show that the maximum TSP displacement at any location on any plate is 0.525 inch. Additional conservatisms in these analyses include the a
- sumptions of as-designed tube to TSP gaps (i.e., no TSP corrosion or packed crevices), tubes centered in the gap (i.e., no random tube misalignment) and bounding SLB hydraulic loads assuming a double ended pipe ;
break and hot standby conditions to maximize the loads. In Reference 2, an upper bound estimate of the tube burst probabilitj was developed under the assumption that the TSP displacement uncovered a throughwriil crack at every TSP intersection equal to the local displacement of the TSP in a SLB event. This is equivalent to assuming that the tube at every TSP intersection has a throughwall crack approximately equal to the TSP thickness. The tube burst capability for a 3/4 inch long throughwall crack is approximately equal to the free span i burst probability for the length of the crack exposed outside the TSP. For the FDB with larger tube to TSP gaps, supplemental burst tests were performed to establish the reinforcing effect of
- the FDB and these results were included in the tube burst probability estimate of Reference 2.
For the above bounding assumptions, the tube burst probability was developed in Reference 2 and found to be 3.6 x 10-5 As noted in Reference 2, this estimate bounds the burst probability for the Catawba-1 SGs and free span burst probability estimates should not be required for Catawba-1 IPC applications. However, in the Catawba-1 SER, the NRC notes that they have l l
not completed the review of the Catawba-1 SLB displacement analyses and requires that a free span burst probability estimate be made for EOC-8. ~ A free span estimate is provided in Section 9.3 although Westinghouse does not believe free span assumptions are appropriate for the Catawba-1 SGs.
If realistic assumptions are used in the SLB TSP displacement analyses, it is expected that the displacements would be much smaller than the Reference 2 results and tube burst would be prevented by the constraint provided by the TSP. Tubes in the SG are randomly misaligned sh associated tube to TSP contact forces. Axial loads on the TSPs, such as in a SLB event, ut then resisted by friction forces and bending of the misaligned tubes. Under operating l conditions, the temperatures of the TSPs are hotter than the wrapper such that the differential thermal expansion leads to a contact force and associated friction force between the wedges 4
.nd L wrapper. In an SLB event, this friction force would resist TSP displacement at the periphery of the plates where displacements are the largest. For the hydraulic loads in the existing analyses, the loads have been very conservatively maximized by assuming hot standby conditions and water levels below the expected water level. As shown in WCAP-14046, Rev.1 (Braidwood-l IPC), the hydraulic loads on the TSPs increase significantly as water level is assumed to drop below expected levels as controlled by the water level set points.
DCPIR911 WPs 91 serie.d., 27.1994
l Since the time at hot standby conditions is very short compared to operating conditions, the associated probability of an SLB event at hot standby would be very small and the contribution of hot standby to the total probability of an SLB event would be negligible. When the more realistic conditions are included in the TSP displacement analyses, the resulting displacements are likely to be negligible (< 0.35" estimated) and the associated tube burst probability would also be negligible. Since these effects are ignored in the Catawba-1 displacement analyses, the existing analyses for an upper bound burst probability should obviate the need for free span tube burst probability analyses.
The importance of limited TSP displacement in a SLB event can be demonstrated by comparing the free span burst probability (applying the burst pressure versus voltage l
l correlation) to the burst probability associated with the crack length exposed by the TSP l displacement. Table 9-1 provides this comparison for the largest voltage indications found in SG C. The first two columns of Table 9-1 show the field call and bobbin voltage for the largest SG C indications. Columns 3 and 4 provide the nominal burst pressure from the l
burst / voltage correlation and the probability of SLB burst at the indication voltage value l
l including effects of temperature and material properties variations. The sum of the free span 4
burst probability for the 28 indications evaluated is 4.48 x 10 The next three columns of i
Table 9-1 show the burst probability estimate assuming the cracks are located at the maximum SLB displacement position on the plate as obtained frorr. the SLB TSP displacement analyses.
Column 5 shows the maximum displacement anywhere on the plate, column 6 shows the expected burst pressure for a throughwall crack length equal to the maximum TSP displacement and column 7 shows the burst probability for the assumed throughwall crack l length. For this assessment, it is assumed that the exposed through-wr I crack ler,gth is equal to the maximum TSP displacement of column 5 even though this length may be larger than associated with the bobbin voltage of column 2. For example, a less conservative assumption is to use the expected burst pressure of column 3 based on voltage to estimate the throughwall crack length for this burst pressure. With this assumption combined with limited TSP displacement, the calculated burst probability for the 28 indications would be < 10 " For the assumption of throughwall crack length equal to the maximum TSP displacement of column 5, the tube burst probability for the 28 indications is 3.98x10' (bottom of column 7), which is a decade smaller than obtained from the voltage burst correlation. The last 3 columns are similar to columns 5 to 7 except that the expected 'lSP displacement (column 8) at the location of the indication is used for the exposed crack length rather than the maximum displacement anywhere on the plate as given in Column 5. In Column 8, the maximum exposed crack tength of any tube is 0.353" compared to a maximum displacement of 0.525" in Column 5.
The local TSP displacement of Column 8 still exceeds, in many cases, the equivalent through-wall length associated with the expected free span burst pressure of Column 3. While remaining conservative, the burst probabilities of Column 10 are the most realistic Table 9-1 for the limited TSP displacements expected at Catawba-l. From Column 10, the t . , of the burst probabilities is 1.02 x 10'" which is much smaller than the 4.48 x 10" obtained for the free sprn burst probability. This example analysis demonstrates the benefits on reduced burst probability resulting from the limited TSP displacements in an SLB event for Catawba-l.
DCPIR911 WP5 9-2 m e a.i994
f 9.2 Free Span Tube Burst Probabilities at EOC-7 for Projected and Actual Distributions Monte Carlo analyses have been performed to project EOC voltage distributions, as described in Section 7.3, and to estimate the free span tube burst probability based on the projected EOC voltage distributions. The NRC model of Reference I has been applied to define the BOC indications left in service. This model very conservatively applies a probability of detection (POD) of 0.6 to the detected indications at EOC to define the total potential indications at EOC. The POD adjusted distribution is then reduced by the tubes repaired to define the BOC voltage distribution. The NRC model includes RPC NDF indications as bobbin voltage flaw indications left in service. The assumptions that a POD = 0.6 is applied to all detected indications and that all RPC NDF indications are flaws lead to very conservative projections of bobbin voltage distributions. Analyses were also performed using POD = 1.0 and the new indication method of Section 10.
For the Monte Carlo tube burst probability analyses, the EPRI data of Reference 5 were used to define the burst pressure versus bobbin voltage correlation. Per the NRC SER for Catawba-1, model boiler specimen 598-1 would be included in the data to develop the NRC database burst correlation. The difference between the NRC and Reference 5 correlations are, however, negligible and the EPRI database was used for the burst probability analyses. Burst probability analyses using beh databases were performed to support this conclusion although, for simplicity, only the EPRI da,sbase results are given in this report. Differences in the burst probabilities calculated for the two databases are bounded by 25% with neither database resulting in uniformly higher burst probabilities.
The Monte Carlo analyses use the the EOC-7 voltage distributions of Section 7.1 (Table 7-1) which include the NDE uncertainties of Section 6.0 and the voltage growth distribution obtained for Cycle 6, SG C (Figure 5-1). SG C was found by comparative analyses with SG D to be the most limiting SG for SLB leakage and was also used for the burst probability analyses. For the Monte Carlo analyses, the SG EOC voltage distribution is sampled 100,000 times to obtain a statistically significant SG burst probability. Bobbin voltages for each indication in the SG are rounded to a bin voltage with a bin width of 0.1 volt and sampling of each indication in the bin assumes a uniform distribution of voltages within the bin. For each EOC voltage indication, the distributions for material flow stress and burst pressure are sampled to obtain a burst pressure. The sampling process is repeated for each indication in each bin and then for all bins in the SG. The sample burst pressures are used to develop a sample SG probability of burst pressures less than AP3t3 = 2560 psi. The SG sample burst probabilities are ordered into a cumulative probability distribution and this distribution is evaluated to obtam the upper 95% confidence on the SG tube burst probability. For 100,000 SG samples, the 95% confidence value is close to the 95% value of the cumulative probability distribution. This Monte Carlo process can be applied to both actual and projected EOC voltage distributions.
?
DCPIR911 WP5 9-3 seri 6.,27.iv94 ;
Deterministic burst probability analyses are also performed using a statistical combination of the material property and burst pressure distributions which is then applied to obtain a burst probability for each voltage bin. This deterministic method, while less accurate than the Monte Carlo analyses, provides reasonable estimates as shown later and can be applied for survey analyses as an alternative to the more elaborate Monte Carlo analyses.
Table 9-2 provides the results of the burst probabilities calculated from the actual EOC-7 and projected EOC-7 voltage distributions. The results indicate that the deterministic methods result in conservative estimates compared to the Monte Carlo analyses. The projected EOC-7 burst probability for POD = 0.6 is about 50% higher than the 2.9x10'8 (Monte Carlo result) obtained for the RPC confirmed actual EOC-7 voltage distribution. The burst probability for the new indication method is in very good agreement with that obtained from the RPC-confirmed actual distribution. The burst probability calculated for the POD = 1.0 voltage distribution is about 30% lower than obtained for the actual distribution. The burst probability calcult.ted for the actual all bobbin indication (including RPC NDF) distribution is only slightly higher than obtained for the RPC confirmed indications. This close agreement results as the l
larger indications contributing to the burst probability were RPC confirmed indications such that the differences would be expected to be minor. Even though the largest 5.1 volt bobbin indication was underpredicted by the projection metliods, conservatisms in the new indication and POD = 0.6 methods result in an overestimate of the burst probability compared to that for the actual distribution.
9.3 Free Span Tube Burst Probability Projections for EOC-8 The projected EOC-8 tube burst probability from the Monte Carlo analyses for the NRC model is 9.6 x 10 as shown in Table 9-2. The new indication method yields about a 30% lower burst probability and the POD = 1.0 distribution yields almost a factor of three lower burst probability. As shown in the following section, the major contributors to the burst probability are the five largest EOC-8 voltage indications (which largely result from the NRC POD correction leaving 0.67 indications left in service for indications that were detected and plugged) and the smaller per tube burst probabilities weighted by the large number of indications left in service. This can be seen from Figure 9-1 which shows the free span probability of burst as a function of bobbin voltage. This burst probability trend was developed by statistically combining the material property distribution and the burst pressure versus voltage correlation for a SLB pressure differential of 2560 psid. The NRC model applies the POD to the EOC voltage distribution, including plugged indications, and the resulting BOC voltages are increased by growth and NDE uncertainties. For the NRC model, this results in a maximum EOC-8 voltage of 5.7 volts. From Figure 9-1, the burst probability for a 5.7 volt indication is about 3 x 10-5 Thus the very conservative assumption that a POD = 0.6 applies to a 5 volt indication dictates, by itself, a burst probability greater than 10
such tl at the cr.lculated burst probabilities from the NRC model are excessively conservative.
November 16,1994 DCPIR911 WP5 9-4
l 1
Any adjustment for undetected indications must reflect the high probability of detecting high voltage indications to provide a meaningful tube burst probability.
Even with the excessive conservatism of the NRC model, the projected EOC-8 tube burst probability is acceptable without additional tube repair as it is less than the guideline of 2.5 x 10 2 given in Reference 2 based on the NUREG-0844 analyses. The calculated burst probability of 9.6 x 10-' is also adjusted below the NRC draft generic letter value of 1.0 x 10 2 for reporting results to the NRC.
a 9.4 Contributors to Free Span Tube Burst Probability The Monte Carlo analyses provide the most accurate estimate of the tube burst probability as the process develops millions of burst pressure samples include the extreme low burst pressure ;
tails of the burst correlation. A good approximation to the burst probability is obtained from the deterministic analysis given in Table 9-2. This application uses the Figure 9-1 probability of tube burst .s a function of voltage. The contribution from each voltage bin is obtained by l weighting the burst probability for each voltage bin by the number of indications in the bin. l The total burst probability is then the sum over the contributions from each voltage bin. Tube burst probabilities were obtained for the NRC model (POD = 0.6), for the alternate model of Section 10.0 for POD adjustments, for all bobbin indications left in service and for the actual voltage distribution at EOC-7. The associated burst probabilities are given in Table 9-2.
Based on the results for the NRC model, the deterministic methods yield about 25% higher burst probabilities than the full Monte Carlo method of analysis.
The deterministic voltage method for burst probabilities permits an estimate of the contribution l to the burst probability for each voltage bin. The analysis results for the NRC model at EOC-8 are tabulated in Table 9-3. The contribution to the burst probability from EOC-8 indications (4 indications) above 4 volts, a typical maximum EOC value, is 6.9 x 10-' The remainder of the total probability of 1.64 x 10 2 results from indications below 4 volts with significant contributions from the large number of indications between 0.8 and 2.0 volts. Table 9-2 also indicates a burst probability of 2.9 x 10'$ for the measured voltage distribution at EOC-7. This is more likely to be typical of the EOC-8 value than that obtained from the NRC model.
5 DCPIR911 WP5 95 so...deri6.i,94
Table 9-1: Catawba 1, SG "C", Reduction in Burst Prwhability Due to Limited SLB TSP Displacement 1993 Bobbin Inspection Based on Burst Pmssure vs Based on Burst Pressure vs'Ihroughwall length Based onTSP Displacement Analysis Results Volts Analysis Analysis Maxinum Pmbebility of Pmbebility of Maximum TSP Expected Pb @ Probability of Expected Pb p Pb SLB Burst Displaament 650*F SLB Burst Exposed length @ 650*F St.B Burst Indication Vg (650*F)
ODI 5.0.5 5.443 1.53E-03 0.525 3.932 3.02E-05 0.005 10.269 7.72E-28 NOI 4.78 5.507 1.25E-03 0.499 4.104 4.92E-06 0.2 % 5.852 238E-14 ODI 4.00 5.728 5.98E.04 0.222 6.721 5.90E-18 0.222 6.721 5.90618 ODI 2.91 6.124 1.48E-04 0.525 3.932 3.02E-05 0.290 5.921 1.18E-14 ODI 2.89 6.132 1.43E-04 0.525 3.932 3.02E-05 0.004 10.298 6.79E-28 ODI 2.73 6.203 1.10E-04 0.222 6.721 5.90E-18 0.222 6.721 5.90E-18 ODI 2.69 6.221 1.03E-04 0 777 6.721 5.90E-18 0.222 6.721 5.90E-18 ODI 2.67 6.231 9.%E-05 0.525 3.932 3.02E.05 0.010 10.171 1.20E-27 NOI 2.48 6322 7.06E-05 0.525 3.932 3.02E-05 0.010 10.158 1.27E-27 ODI 2.47 6327 6.92E-05 0.222 6.721 5.90E-18 0.222 6.721 5.90E-18 NOI 231 6.411 5.04 E-05 0.222 6.721 5.90E-18 0.222 6.721 5.90518 ODI 2.28 6.427 4.74E-05 0.525 3.932 3.02E-05 0.041 9.557 233E-26 ODI 2.17 6.488 3.74E-05 0.222 6.721 5.90E-18 0.222 6.721 5.90E-18 ODI 2.09 6.535 3.12E-05 0.222 6.721 5.90E-18 0.222 6.721 5.90E-18 ODI 2.07 6.547 2.97E-05 0.525 3.932 3.02E-05 0.087 8.719 2.43E-24 NOI 2.06 6.553 2.91E-05 0.222 6.721 5.90E-18 0.222 6.721 5.90E-18 NOI 2.01 6.584 2.58E-05 0.525 3.932 3.02E-05 0.031 9.748 8.93E-27 NOI 2.01 6.584 2.58E-05 0 ??? 6.721 5.90E-18 0.222 6.721 5.90E-10 NOI 2.00 6.590 2.52E-05 0.222 6.721 5.90E-18 0.222 6.721 5.90E-18 ODI 1.66 6.821 9.98E-06 0.288 5.943 9.46E-15 0.288 5.943 9.46E-15 ODI 1.55 6.907 7.06E-06 0.525 3.932 3.02E-05 0.004 10.298 6.79E-28 NOI 1.52 6.931 639E-06 0353 5.281 1.01 E-11 0353 5.281 1.01E-11 ODI 1.47 6.972 539E-06 0.525 3.932 3.02E.05 0.234 6.573 2.22E-17 NOI 1.46 6.981 5.20E-06 0.288 5.943 9.46E-15 0.288 5.943 9.46E-15 NOI 1.45 6.989 5.02E-06 0.525 3.932 3.02E.05 0.009 10.187 1.12E-27 ODI 1.44 6.998 4.84E-06 0.525 3.932 3.02E.05 0.062 9.171 1.80E-25 NOI 139 7.042 4.04E-06 0.222 6.721 5.90E-18 0.222 6.721 5.90E-18 MOI 139 7.042 4.04 E-06 0.525 3.932 3.02E-05 0.013 10.106 1.62E-27 Total: 4.48E-03 Total: 3.98E-04 Total: 1.02E-11
[DCPBURST.XLS] Nommt RFK 944/94. 8.24 PM
I d
I i
l Table 9-2. Summary Results for Tube Burst Probabilities: SG C
- Tube Burst Probability 1
Method of Analysis !
EPRI Database Deterministic Monte Carlo Actual EOC-7 Voltage Distribution ,
1 I
- RPC confirmed plus not inspected 5.4x 10 2.9x10
i All bobbin indications 6.6x 10 4.1x10
Projected EOC-7
- NRC model with POD = 0.6 8.0x10-' 4.7x10
i
- All bobbin indications, POD = 1.0 3.6x 10
New indication method of Section 10.0 5.0x10 3.0x10
1 i
Projected EOC-8
- NRC model with POD = 0.6 1.6x10 2 9.6x 10
- All bobbin indications, POD = 1.0 6.4x 10-' 3.5x10'$
- New indication method of Section 10.0 8.2x10 ' 6.2x10
Note:
Monte Carlo analyses with 100,000 samples of total voltage distribution evaluated at +95% upper bound.
l sove d.,16.1994 DCPIR911 WP5 97
l l l l l Table 9-3 l
l l i i Deterministic Estimate of Tube Burst Probability at EOC-8 l
\
l Bin Bin Expect StDev Student Probability
, volt Count Burst Burst Deviate of Burst I
0.20 14.00 9.516 1.040 6.689 2.7313D-08 1 0.30 67.00 8.999 1.024 6.287 7.1170D-07 0.40 179.00 8.632 1.014 5.989 6.53e9D-06 l 0.50 338.00 8.348 1.006 5.751 3.2675D-05 l 0.60 507.00 8.115 1.000 5.553 1.0938D-04 l 0..) 620.00 7.916 0.996 5.382 2.6473D-04 l 0.80 G44.00 7.748 0.992 5.232 4.9768D-04 0.90 543.00 7.598 0.988 5.098 7.0875D-04 1.00 412.00 7.464 0.985 4.977 8.5957D-04 ;
1.10 294.00 7.342 0.983 4.866 9.3731D-04 ,
- 1.20 191.00 7.231 0.980 4.764 8.9635D-04
! 1.30 122.00 7.129 0.978 4.670 8.1653D-04 I 1.40 76.00 7.035 0.977 4.582 7.0602D-04 1.50 44.00 6.947 0.975 4.500 5.5419D-04 )
1.60 28.00 6.864 0.973 4.422 4.6842D-04 l 1.70 18.00 6.787 0.972 4.349 3.9279D-04 1.80 10.00 6.714 0.971 4.280 2.8010D-04 1.90 8.00 6.645 0.969 4.214 2.8350D-04 2.00 4.00 6.580 0.968 4.151 1.7705D-04 2.10 4.00 6.517 0.967 4.091 2.1854D-04 2.20 3.00 6.458 0.966 4.034 2.0019D-04 2.30 1.00 6.401 0.965 3.979 8.0721D-05 2.40 1.00 6.347 0.965 3.926 9.6775D-05 2.50 1.00 6.295 0.964 3.876 1.1500D-04 2.70 1.00 6.197 0.962 3.780 1.5919D-04 3.00 1.00 6.063 0.960 3.647 2.4700D-04 3.60 1.00 5.830 0.957 3.416 5.2029D-04 4.20 1.00 5.634 0.955 3.219 9.6024D-04 4.60 1.00 5.517 0.953 3.102 1.3673D-03 i 5.10 1.00 5.386 0.952 2.968 2.0252D-03 5.30 0.70 5.337 0.952 2.918 1.6380D-03 5.70 0.30 5.244 0.951 2.823 9.1942D-04 Cumulative Probability of Burst = 1.641D-02 I
I DCPIR91 I WP5 9-8 ser==b- i s.1994 )l l
4
l Figure 9-1: Probability of Burst vs. Bobbin Amplitude 3/4" OD x 0.043" Thick, Alloy 600 MA, SG Tubes @ 650*F 1.0E-02 ....... . _ . . . . . . . . .
_.-s. 4'-
Database includes MB 598-1 !.__
1._.
1.0E 03 ,
l l
y ,
/ -
t -- s 4_ j_
, /
/
1.0E-04 . . _ . = ._, . , -
/ ._.:
, .--- . .f
+-F-t-
7 ._ t.7 ,t ;_..i_- ,f,i 7
-4'
.4.. _ . _
I I I I tIl i' l l 1.0E-05 . :_ : '
t-H - + .
', _ _a i
5 :
M._ .
}
- j. 4 .
4_yt \. q-
% ~T. T~
g 1.0E-06 . _, ,, . :.
.5 l --+q. _a
_+- - ,
> _}_d - .
b . _
-. L ]
i 1.CE-07
./
I I I-
/, +-+ . -r
,C4 v4 p* ._.:_-.
i
/
1.0E-08 ,-
/ . . . .
l i
+-._
f . 4 - - + , _._
r 1 .
/- i
/ .L L- ._
1.0E-09 ,-
/ I
. . -._.4.
I
~ . _ . . _ -
f p_ _
+ 4 t._.i . . ,
t ._.
,f ,
n.: g _.t__. -
t,
_.3 _3.__.r.. ,.__ . 7_ _.
y l _[... i ._.j_ ! ,
I ! l 1.0E-10 1.0 10.0 0.1 Bobbin Amplitude (Volts)
RFK: 3r&94 rupv NRC.XLS) PoB NRC ;
- i l
I 10.0 ALTERNATE METHOD FOR DEFINING BOC INDICATIONS LEFT IN SERVICE I The application of a POD = 0.6 to obtain BOC voltage distributions can lead to excessive conservatism when EOC-7 projections are compared with actual inspection results for the distributions. For Catawba-1, the degree of conservatism for POD = 0.6 is less than typically i d found due to the large number of small voltage, new indications. Tjhe latter have not been RPC inspected and a low confirmation rate would be expected based on the trend of Table 3-3. l The BOC distribution based on all bobbin indications left in service (POD = 1.0), including (
RPC NDF indications, yielded acceptable agreement with the actual EOC distribution above 1.0 volt but underestimates the low voltage distribution. However, this method compensates for ignoring new indications by the inclusion of all RPC NDF indications and equivalent compensation may not be likely in future outages. Thus it is desirable to develop an improved l method for defining BOC distributions that accounts for potentially new indications without the extreme conservatism of the POD = 0.6 adjustment and that recognizes that some RPC NDF indications might develop to flaw indications over the next cycle. It is unlikely that new or
- RPC NDF indications at TSP intersections will develop to leakers over the next operating cycle such that inclusion of these indications at any magnitude in the BOC distribution is, by itself, conservative. New indications at a given outage or time include indications missed at the prior inspection, indications grown from nondetectable to detectable levels and growth from no I indication to a detectable indication. For IPC/APC applications, there is no need to distinguish I these sources for new indications, as the only consequence of importance to leakage or burst is l j that the new indication is sufficiently large to be detectable by both bobbin and RPC probes.
In addition, some low voltage indications reported at the prior inspection are no longer found at the subsequent inspection, particularly when conservative bobbin calling criteria are applied. ,
This has been found at the EOC-7 Catawba-1 inspection, as well as the last other plant !
inspections. For projection considerations, the prior inspection indications not found in the l current inspection should be subtracted from the new indications (but not excluding RPC confirmed new indications) to obtain a net number of new indications. This subtraction typically applies to indications < l.0 volt that are not RPC inspected.
The objective for a desirable DOC distribution is that projections to EOC conditions yield good overall agreement with the larger voltage (greater than 1-2 volts) RPC confirmed indications which have finite probabilities of leakage or burst. Only RPC confirmed indications would have a significant probability ofleakage over the prior cycle and are consistent with the leak rate and burst correlation database, which is dominantly comprised of RPC confirmed indications. The methodology should recognize that new indication and POD considerations are potentially plant specific due to varying degrees of the influence of residual signals on detectability of small indications, enmcements in eddy current analysis guidelines and corrosion controls implemented at a specific time. In addition, the methodology should recognize that historical data, as quanti.ied for Cycle 7 in this report, indi .ates that only a few RPC NDF indications tend to develop to RPC confirmed indications over the subsequent operating cycle. The following sections define a BOC distribution methodology consistent DCP1R911 WP5 10-1 %=ba is, i994
l l
with these objectives that is then applied to obtain EOC-7 projections for comparison with the actual distributions. This methodology is found to result in good agreement with the 1993 EOC-7 voltages and is proposed as an altemative to applying POD adjustments for IPC or I APC in plementation.
l 10.1 Allowance for Undetected or New Indications i
An undetected or new indication is defined as an indication found in the current outage
! inspection that was not identified at the prior inspection independent of the causative factor for the new indication. The method used to develop growth rates in this report is based upon reevaluating the prior outage eddy current data for all indications found in the current inspection. For most indications, when given the somewhat larger flaw indication in the current inspection, reevaluation of the prior data permits identification of the smaller flaw and l
assignment of a voltage at the prior inspection, even if the flaw was not reported at the prior inspection. This process permits assignment of voltages to indications which were not reported at the prior outage and leads to a distribution of prior cycle voltages for new indications which can be described / applied as a voltage distribution for undetected (NDF) indications as an l
alternative to applying more arbitrary POD adjustments to detected indications. This process l
l can lead to a highly conservative distribution of new or undetected voltages when the current f inspection involves an " inspection transient" resulting from implementation of significantly l more conservative eddy current analysis guidelines than applied at the prior inspection. This implementation of more conservative guidelines occured at the first time IPC implementation in 1992 for Catawba-l. Since the 1993 inspection is the second application of IPC guidelines, the l
currently developed distribution of voltages for new indications not detected at the prior outage provides a meaningful description of an undetected voltage distribution. This method for defining undetected indication voltage distributions has been presented to the NRC in WCAP-13692 (April 1993), prepared in response to the NRC APC Task Team questions on analytical models for SLB leakage analyses.
As bobbin analysis guidelines are made increasingly more conservative, the potential for false bobbin calls increases. Current inspection practices and IPC implementations utilize RPC inspection to assess the significance of the bobbin call. If the bobbin indication is not detected by the RPC inspection, the bobbin call is either a false call or the indication is too small to be detected by the RPC probe. An indication not detected by RPC would not challenge tube integrity (burst or leakage) and need not be considered as a significant undetected indication at the prior outage. Thus, the latest inspection results for indications confirmed by RPC inspection or not RPC inspected provide the population of tubes for which the prior cycle eddy l
current data is reviewed to develop the undetected or NDF voltage distribution. Applying the prior inspection undetected voltage distribution as the current inspection NDF indications for the next BOC distribution is conservative, since the data analysis guidelines tend townd the same or more conservative guidelines in successive outages. The undetected, RPC confirmed voltage distribution is added to the detected distribution for indications left in service to obtain DCPIR911 WP5 10 2 Nov..b.,16.1994
i the net BOC voltage distribution adjusted for POD considerations on a plant specific and time dependent basis.
Based on the above, the following guidelines are used to estimate the voltage distribution of !
- undetected or NDF indications:
- Population for Evaluation: Latest inspection indications confirmed by RPC or not RPC inspected that were not reported in the prior inspection (i.e., new indications).
The only assumption for this population is that NDF indications have grown in one cycle to detectable indications to be of a concern for leakage considerations.
Alternately, an indication not detected in two successive inspections can be assumed to result in negligible leakage over the next cycle.
- Process: For the above population, the prior cycle eddy current data is reevaluated by applying the latest analysis guidelines to assign a voltage to the indication at the prior cycle. This process is the same as that applied to develop voltage growth rates for Farley SG indications.
- Undetected (NDF) or New Indication Voltage Distribution: The reevaluated voltages at the prior inspection in which the indication was not detected are used to define the new indication distribution for inclusion in the next BOC distribution.
The number of new indications in each voltage bin that were not RPC confirmed (not RPC inspected) is reduced (but not allowed to be < 0) by the number of indications reported at the prior inspection but not reported at the latest inspection (i.e. assumed false calls at prior inspection) to obtain the net number of new indications.
Based on these guidelines, voltage distributions wsre developed for undetected indications to ,
be included in the BOC-7 and BOC-8 distributions. These distributions were developed for all l new indications in SG C. SG C has a sufficiently large number of new indications such that averaging of indications over all SGs is not necessary to obtain a representative distribution for new indications. When smaller populations are involved, it would be recommended that the new indication distribution be averaged over all SGs. The resulting undetected indication voltage distributions are given in Tables 10-1 and 10-2 for BOC-7 and BOC-8, respectively. j Note that the BOC-7 distribution represents the undetected indications at EOC-5 (prior to IPC i implementation) and the BOC-8 distribution represents the undetected indications at EOC-6 (first IPC inspection). 'Ihe undetected maximum voltage is < 2 volts for both cycles.
10.2 Considerations for RPC NDF Indications As previously discussed, only a few, if any, RPC NDF indications develop to RPC conbrmed indications at subsequent inspections. In the current Catawba-1 inspection, < 25% (22.7% in DCPIR911 WP5 10-3 November 16. i994
i l
SG C above 1.0 volt, lower in other 56s; of the prior inspection RPC NDF indications i
developed to RPC detectable indications. Although RPC NDF indications less than the maximum voltage repair limit are left in service, it would be excessively conservative to include all of these indications in the BOC distribution as potential leakers at the end of the next operating cycle. Therefore, a conservative assumption that 25% of the RPC NDF indications left in service in SG C will develop to RPC confirmed indications was applied to define the contribution of RPC NDF indications to the BOC distribution. This assumption provides adequate conservatism and avoids the excessively conservative assumption that all RPC NDF indications left in service will develop to confirmed indications and only a fraction i
of the RPC NDF indications should therefore be included in the BOC distribution. The resulting contributions to the BOC-7 and BOC-8 voltage distributions are included in Tables 10-1 and 10-2. The BOC-7 and BOC-8 voltage distributions resulting from applying the attemate methods described above are shown in Figure 10-1. Both cycles include bobbin voltages up to 1.5 volts left in service for RPC NDF indications. :
Indications that are RPC NDF in the last two inspections are excluded from the BOC distributions. It can be expected that none or very few of these indications will become RPC detectable at the next inspection.
I 10.3 Comparison of Projections Applying Alternate Method with Actual ;
EOC-7 Distributions Section 7 describes the methodology used to develop the projected EOC-7 distribution of !
indications for the alternate (new indication) method described above. Table 7-1 and Figure 7-1 show the comparison of the actual EOC-7 voltage distribution (excluding RPC NDFs as discussed in Section 7) with the projected EOC-7 voltage distribution based on the new indication method. As shown in Table 7-1, the maximum projected EOC-7 voltage is i 4.2 volts. The new indication method (as well as the POD = 1.0 and POD = 0.6 methods) conservatively bounds the actual EOC-7 voltage distribution above 1.0 volt, except for l
! underestimating the maximum actual voltage indication of 5.1 volts. Below 1.0 volt, the new indication method is in good agreement with the actual distribution. Overall, the agreement between the projected and actual distributions is very good. This supports the new indication projection method's acceptability even for the Catawba-1 case with a very large number of bobbin indications. Similar agreement between the new indication method projections and actual distributions has been obtained for three other plants with IPC assessments comparable to this report.
l Figure 7-2 compares the alternate method projections and the actual EOC-7 distribution for all bobbin indications including RPC NDF indications left in service. It is seen that the alternate projection method remains adequately conservative, even when compared to this EOC distribution, which is unnecessarily conservative for IPC applications. This agreement with all bobbin indications tends to occur when a large number of new indications not RPC inspected 10-4 %=bn 16. im DCPIR911 WP5
l are included in the new indication distribution as is the case for Catawba-1 and represents a conservatism in the new indication method.
Sections 8 and 9 included comparisons of the SLB leak rates and tube burst probabilities based on the actual EOC-7 voltage distribution with the projected voltage distributions. It is seen in l
l these sections that the leak rates and burst probabilities calculated with the projected distributions using the new indication method are in good agreement with the values calculated from the actual voltage distributions. His further supports the new indication method for APC applications.
l l 10.4 Projected EOC-8 Voltage Distributions and SLB Leak Rates Using the BOC-8 distribution for the new indication method described above, Monte Carlo l
methods are applied using the Cycle 7 voltage growth distribution of Figure 5-2 to obtain the projected EOC-8 distributions (see Section 7). Table 7-2 and the bottom plot of Figure 7-3 show the projected EOC-8 distribution obtained by applying the attemate method for defining BOC distributions. A maximum EOC-8 voltage of 5.2 volts is projected for the new indication method.
l l The EOC-8, new indication method distribution leads to a SLB leak rate of 18.3 gpm for the draft NUREG-1477 methodalogy with the NRC database and 0.78 gpm using the APC leak i rate correlation. These results can bc ;,ompared to the Table 8-1 results at EOC-8 of 27.6 gpm for the NRC methodology and 1.61gpm for the APC leak rate correlation meihodology where both methods include the POD = 0.6 adjustment. Thus the more realistic treatment for l
undetected and RPC NDF indications leads to the order of a factor of two reduction in projected SLB leak rates.
l 10 5 sove der is.1994 DCPIR911 WP5 I
l l
Table 10-1: Catawba 1, SG "C", Alternate BOC Distribution for Cycle 7 (~350 EFPD)
- Indications (23 " * " #* "
Tested"'
0.10 66 0 0 19 8.9 0 74.9 186 6 14 47 22.1 5 223.1 0.20 _
0.30 249 13 157 63 29.6 9 439.6 O.40 280 28 320 70 32.9 2 658.9 0.50 201 24 280 39 18.3 0 523.3 0.60 153 18 180 49 23.0 0 374.0 0.70 50 27 78 117 55.0 0 210.0 0.80 26 25 33 61 28.7 0 112.7 0.90 22 32 11 47 22.1 0 87.1 1.00 0 13 4 24 11.3 0 28.3 1.10 0 l1 2 22 _
10.3 0 23.3 _
l.20 0 8 1 10 4.7 0 13.7 130 0 2 0 11 5.2 0 7.2
~
1 40 0 2 0 6 2.8 0 4.8 1.50 0 1 0 6 2.8 0 3.8 1.60 0 0 0 3 1.4 0 1.4 1.70 0 3 0 0 0.0 0 3.0 1.80 0 1 0 0 0.0 0 1.0 1.90 0 0 ,
0 1 0.5 0 0.5 2.00 0 1 0 0 0.0 0 1.0 2.60 0 1 0 0 0.0 0 1.0 2.70 0 1 0 0 0.0 0 1.0 3.60 0 1 0 0 0.0 0 1.0 Totals 1233 218 1080 595 279.6 16 2794.6 Notes:
- 1. EOC 6 indications that were left in service and were either RPC confirmed or not RPC inspected.
- 2. New indications found at EOC 6 that were confirmed by RPC (4/91, EOC 5, resize voltage levels used).
- 3. New indications found at EOC 6 which were not RPC inspected (4/91, EOC 5, resize voltage levels used).
mena .a . e v. " T** RFK: 11/1/94.11:40 AM
Table 10-2: Catawba 1, SG "C", Alternate BOC for Cycle 8 (390 EFPD)
- *" #* "' "* ^'#"*
Initial BOC RPC NDD Volts Confirmed Not RPC . . NDD Indications NDF Method BOC Distribution'" Indications IMi io n"' TWo, Indications at EOC 7 Indications 0.10 0 0 2 0 0.0 57 0.0 0.20 20 1 15 2 0.5 134 21.5 030 65 2 58 9 2.3 133 69.3 0.40 148 4 124 18 4.5 121 159.5 0.50 252 3 203 15 3.8 70 391.8 0.60 355 2 264 26 6.5 23 604.5 0.70 416 6 272 28 7.0 40 661.0 0.03 345 2 240 23 5.8 11 581.8 0.90 253 2 130 17 4.3 13 376.3 1.00 142 2 52 24 6.0 3 199.0 1.10 0 13 0 87 21.8 6 34.8 1.20 0 5 0 51 12.8 3 17.8 1.30 0 0 0 30 7.5 5 7.5 1.40 0 1 1 9 2.3 1 3.3 1.50 0 1 1 4 1.0 2 2.0 1.60 0 0 0 2 0.5 0 0.5 1.70 0 0 0 3 0.8 0 0.8 1.80 0 1 1 1 0.3 0 13 1.90 0 0 0 3 0.8 0 0.8 2.00 0 0 0 0 0.0 0 0.0 2.10 0 0 0 0 0.0 0 0.0 2.30 0 0 0 1 0.3 0 0.3 2.80 0 0 0 0 0.0 0 0.0 Totals 1996 45 1363 353 88.8 622 3134.8 Notes:
- 1. EOC7 indications left in service including: RPC confirmed, not RPC inspected, and new RPC NDF indications mtritiplied by 0.25. Indications RPC NDF in two successive cycles are not included.
- 2. New indications found at EOC7 (EOC6 voltage levels) confirmed by RPC.
- 3. New indications found at EOC7 (EOC6 voltage levels) which were not RPC inspected.
--....-..e.n-n .u..'* RFK: 10/26/94.11:13 AM
l
\ .
l r Figure 10-1: Catawba 1, BOC 7 & BOC 8 Indication Distributions Alternate Method for Estimating Distributions t 700
-. E BOC 7 Indications OBOC 8 Indications a - i 500 g _- --
O
, g 8 400 =
Ec -
m _ . - .
O
~
$ 300 E
- s __
Z 200 E _ _ _ ._
100 i
, , , , , , , , .E , dl , _ . . ---
c >
0 . - - - - . . . - - -
S 8 R % R S 8 5 8 ? @ 8 8 %. 8 % 8 % 8 ?
~
?
m 8
~
m e e e e e o e 0 0 - - - - - - - - - -
Bobbin Amplitude (Volts)
RFK- 11/1/94,12:17 PM erv'pruott' vi m Fin to 1
I
11.0 REFERENCES
The references used in this report are:
- 1. Draft NUREG-1477, " Voltage Based Interim Plugging Criteria for Steam Generator Tubes
- Task Group Report," United States Nuclear Regulatory Commission (NRC). June 1,1993
- 2. WCAP-13854, " Technical Suppon for Cycle 8 Steam Generator Tube Interim Plugging ,
Criteria for Catawba Unit 1," Westinghouse Electric Corporation, September,1993
- 3. WCAP-13494, Revision 1, " Catawba Unit-1 Technical Support for Steam Generator Interim Plugging Criteria for Indications at Tube Support Plates," Westinghouse Electric i Corporation, March 1993
- 4. WCAP 13684," Analysis to Determine Relative TubeHube Support Plate Displacements l Under Steam Line Break Loads for Catawba Unit-1 Steam Generators," Westinghouse l Electric Corporation, March,1993. J i
- 5. EPRI Report NP-7480-L, Volume: 2, Steam Generator Tubing Outside Diameter Stress Corrosion Cracking at Tube Support Plates - Database for Alternate Repair Criteria, l Volume 2,3/4 Inch Diameter Tubing," October,1993 i
I l
1 l
l l
DCPIR911 WP5 l1-1 October 13.1994
[
l i
i
11.0 REFERENCES
The references used in this report are:
f
- 1. Draft NUREG-1477, " Voltage Based Interim Plugging Criteria for Steam Generator l
- Task Group Report," United Sta:es Nuclear Regulatory Commission (NRC), June !
- 2. WCAP-13854, " Technical Support for Cycle 8 Steam Generator Tube Interim Pluggin Criteria for Catawba Unit 1," Westinghouse Electric Corporation, September,1993
- 3. WCAP-13494, Revision 1, " Catawba Unit-1 Technical Suppon for Steam Generator Interim Plugging Criteria for Indications at Tube Suppon Plates," Westinghouse Elect Corporation, March 1993 l
- 4. WCAP--13684, " Analysis to Determine Relative Tube / rube Suppon Plate Displacements Under Steam Line Break Loads for Catawba Unit-1 Steam Generators," Westinghouse Electric Corporation, March,1993.
1
- 5. EPRI Repon NP-7480-L, Volume 2, " Steam Generator Tubing Outside Diameter Stres l Corrosion Cracking at Tube Suppon Plates - Database for Anemate Repair Criteria, Volume 2,3/4 Inch Diameter Tubing," October,1993 l
1 l
i l
I l
l l
cewsu n.i994 11-I ocm9:t wr5