ML20198P659
| ML20198P659 | |
| Person / Time | |
|---|---|
| Site: | Braidwood  |
| Issue date: | 12/31/1997 |
| From: | COMMONWEALTH EDISON CO. |
| To: | |
| Shared Package | |
| ML20198N902 | List: |
| References | |
| NUDOCS 9801220207 | |
| Download: ML20198P659 (33) | |
Text
{{#Wiki_filter:F. . 1 Hraidwood Unit 13.0 Volt IPC Full Cycle Operation Technical Basis Supplement te Hraidwood Unit 1 Cycle 7 Interim Plugging Criteria Report August,1997 Commonwealth Edison December,1997 9901220207 990114 PDR. ADOCK 05000456 P PDR L_.
. t Eucmive Summary - Comed has performed extensive reviews ofindustry growth rate data associated with steam generator tubing outside diameter stress corrosion cracking (ODSCC) at tube support plates (TSP). Based upon these reviews and conservative application of voltage - : F dependent growth rate methodologies, Comed has concluded that all tube mtegrity requirements will be met for Braidwood 1 Cycle 7 and therefore full cycle operation is
- justified. This report provides the supporting information for the conclusion which was discussed in a meeting with the NRC in a meeting on December 11,1997. The conclusions of this report are summarized below:
- 1. An upper bound on growth due to ODSCC at the TSP is 11.5 Volts /EFPY based on .
evaluation of Comed, Domestic and European indications within the range of t Braidwood I repair criteria (3.0 Volts).
- 2. The distribution of growth rates is predictable based upon review of growth rate behavior in almost 23,000 indications.
- 3. A Braidwood 1 EOC-7 maximum leak rate, in the event of an MSLB, of122 gpm (Room Temperature) is predicted based upon leak rate sensitivity studies of voltage dependent growth rate bin cutoff. This increases the EOC leak rate projection, during an MSLB, from 57.1 gpm (Reference 2) to 122 gpm.
- 4. Braidwood I is safe for full Cycle 7 operation with a Dose Equivalent Iodine limit of 0.05 pci/g. Braidwood has administratively limited reactor coolant DE l-131 to 0.05 pei/g and a Technical Specification amendment request for operation at this limit has been submitted for NRC approval.
2 , 1 l
__ - - . - - - ~ -. t .
.i i .
1.0 Introduction The Braidwood Unit 1 sixth refuel outage (AIR 06) eddy current testing (ECT) inspection resulted in a calculated as-found main steam line break (MSLB) leak rate, from tubes w;th - . ODSCC indications at the tube support plate (TSP), which exceeded that predicted at the - ! beginning-of cycle six (BOC-6). The predicted end-of-cycle six (EOC-6) leak rate was !
- based upon the EOC-5 ECT inspection results. The predicted EOC-6 MSLB leak rate as reported in the Cycle 6 90 day report (Reference 1) for steam generator C was 6.99 -
< gallons per minute (gpm). The as-found MSLB leak rate, including NDE uncertainty !
. applied to the distribution, is i 1.5 gpm. This discrepancy in the EOC-6 predicted and as- ,
found leak rate was reported to NRC in the Braidwood 1 Cycle 7 90-day report (Reference 2). A root cause analysis was performed to determine the reason for the under-prediction in the EOC-6 leak rate. The root cause as documented in Reference 3 was the result of voltage dependent growth rates (e g. more large growth rates with increasing BOC i voltage) which first became evident during Cycle 6 due to the implementation of a change
- in the repair criteria from 1.0 to 3.0 volts for a full cycle of operation. :
A change in the NRC GL 95 05 projected EOC distribution methodology was developed to accommodate the voltage dependent growth rates. The revised methodology bins the i growth rates into three or four bins corresponding to the indication BOC voltage. T he I growth rates within each bin cre then applied to the corresponding BOC distribution for the next cycle resulting in the EOC distribution. This methodology provides more conservative results than the previously used voltage independent methods. This is due to , using voltage dependent growth rates where the large growth rate indications are combined with large BOC indications (0.6 of which are artificially created due to the affect ofusing a POD of 0.6). The resulting EOC distribution for each BOC voltage bin is combined and a total distribution is obtained from which the EOC leak rate is calculated. This methodology is discussed in more detail in the Braidwood 1 Cycle 7 90-day report
- (Refertnce 2).
i
= The voltage dependent growth methodology discussed above was benchmarked against ;
SG C Cycle 6 data using Cycle 6 growth rates and five additional Byron and Braidwood , steam generators. The benchmarking resulted in a conservative EOC distribution and leak rate. Results oithe benchmarking are provided in References 2 and 5. Subsequent to issuance of the Braidwood 1 Cycle 7.90 day report, the stafrissued a Request for Additional Information (RAI) (Reference 4) regarding the sensitivity of the EOC leak rate to varying the growth rate distribution. In particular, the RAI requested Comed vary the cutoff orthe highest bin to the region which demonstrates a significant change in the grow 1h rates to voltage dependency and actificially including the three, five and seven largest growth rates from other SG's into the most limiting SG. A response to 3 l a
- e.i:
9 the RAI was submitted (Reference 5) which showed among other things that changing the
- . cutofTof the highest bin defming the voltage dependent growth rate distribution could -
significantly increase the EOC leak rate. As a result of thee findings the NRC issued a letter on November 13,1997 (Reference 6). The letter states that there was insuflicient information for the NRC to accept the voltage dependent growth rate methodology presented by Comed in References 2 and 5, and that the leak rate used to set the Reactor Coolant Dose Equivalent lodine (DE 1 131) limit was not conservative. 1 This report provides the technical basis to justify full cycle operation for Braidv cod Unit I during Cycle 7. Included in this report is a revised EOC leak rate which requires further reducing the Technical Specification DE l-131 litnit to 0.05 pci/g. 2.0 Growth is Bound and Predictable ~ Comed performed exhaustive reviews ofindustry growth rate information to assess how ODSCC at the TSP growth behaves. The conclusions from the reviews are that 1) growth rates are predictable,2) there is an upper bound to growth rates, and 3) there is a voltage dependency to growth rates. Growth rates over four (4) cycles of operation at Braidwood I are bounded by the industry experience. 2.1 Comed ODSCC at the TSP Growth Experience Growth rates for Byron and Braidwood indications due to ODSCC at the TSP (over 16,000 indications) from previous inspections were evaluated. The evaluation was performed to assess if there had been an increase in the growth rates during Braidwood 1 Cycle 6. Indications reviewed included indications prior to implementation of alternate repair criteria and indict.tions detected during inspec: ions where alternate repair criteria were implemented Cycle 4 at Braidwood is the only other cycle in addition to Cycle 6 where an alternate repair criteria was implemented and the Unit operated full cycle between inspections Figure I shows a chronology of the Alternate Repair Criteria for ODSCC at the TSP for Braidwood and Byron. For consistent comparison growth rates have been normalized to I EFPY and were adjusted to the Braidwood 1 Cycle 7 operating temperature of 612 degrees Fahrenheit using the methodology of Reference 7. Byron and Braidwood Indications with growth rates greater than four (4) volts'EFPY are shown on Figure 2. The data shows that the bounding growth rate is from Braidwood 1 Cycle 4 at 9.62 Volts /EFPY. Cycle 6 browth rates are bound by the previous Braidwood 1 Cycle 4 full cycle experience at 9.62 Volts /EFPY. 4
. 2.2- Domestic ODSCC at the TSP Growth Experience ,
Growth experience for eleven domestic plants implementing an ARC for steam generator tube ODSCC at the TSP was evaluated to provide additional data to supplement the
- Comed database. For consistent comparison, growth rates have been normalized to 1 EFPY and were adjusted to the Braidwood 1 Cycle 7 operating temperature of 612 degrees Fahrenheit using the methodology of Reference 7.
The data was added to the Comed data discussed in Section 2.1. Indications with growth
- rates greater than four volts /EFPY are shown on Figure 3. The data shows that the bounding growth rate is 11,44 Volts /EFPY. Braidwood 1 Cycle 6 growth rates are bound by domestic industry experience at 11.44 Volts /EFPY.
2.3 European ODSCC at the TSP Growth Experience The ODSCC at the TSP indication database was further supplemented with European data from 10 Units with ODSCC at the TSP. The data is from plants with similar designs and materials with Mill Annealed inconel 600 tubing and drilled carbon steel support plates. Required adjustments to the data are further discussed in later sections of this report. Data from European plants is valuable for three reasons. First, voltage based repair limits much higher than domestic plants have been implemented for ODSCC at the TSP indications. Leaving larger indications in senice provides an opportunity to evaluate how the indication distribution grows from cycle to cycle without removing the large BOC indications from senice. Second, there is experience with thousands ofindications of the same magnitude and larger than the Braidwood I repair limit of 3.0 Volts. This prosides the opportunity to assess the impact of a large number ofindications in the distribution growing toward the repair limit. From this an understanding, of how the number oflarge growth rate indications changes with the shift of the distribution towards higher BOC voltages, was developed. Finally, by leaving larger indications in senice a larger number oflarge growth rates (i.e. voltage dependent growth) were observed, similar to that for Braidwood I during Cycle 6. This provided an opportunity to benchmark the voltage dependent growth methodology developed in Reference 2. 2.3.1 Voltage Adjustments In order to use the European data for comparison to Braidwood, indication voltages were normalized to the calibration procedures used at Braidwood. The correlation's used to adjust the European data to Braidwood I follow the methodology outlined in References 8 and 9. Evaluations were performed using the adjusted data. Evaluations of Growth
- rates, which use the change in voltage from cycle to cycle, are not impacted by the voltage adjustments.
5
No adjustment is required for comparison of voltages for tubes of 7/8" diameter x 0,050" wall thickness to 3/4" diameter x 0.043" wall thickness tubes. The basis for not making an l adjustment for the diameter difference is discussed below. Voltage Normalization Comparison For 7/8" and 3/4" Tubing: An evaluation was performed to determine if voltages for Outside Diameter Stress Corrosion Cracking (ODSCC) at Tube Support Plates (TSP) will produce equivalent voltage responses for 7/8" and 3/4" t" Sing. The evaluation shows that voltage responses and normalization routines for 7/8" and 3/4" tubing can be used for growth analysis studies. The technical basis for voltage normalization can be found in EPRI NP-7480-L Volume i for 7/8" Outside Diameter Tubing and Volume 2 for 3/4" Outside Diameter Tubing and EPRI JWR S/G Guidelines Revision 4 Volume 2. The EPRI " Mother" calibration standard for Alternate Repair Criteria (ARC) for ODSCC was used for normalizations and to compare different tube sizes. To compare different tube sizes the similar frequercy rule was applied. The similar frequency rule is that for different wall thickness there exists inspection frequencies that will react identically. For example,400 kI;Iz inspection of 0.050" wall thickness ofInconel 600 mill annealed tubing will give identical responses to 550 kilz inspection of 0.043" wall thickness inconel 600 mill annealed tubing. The ARC normalization routine was used to evaluate the voltage responses. For the 7/8" wall tubing the 400 kilz was normalized to 4 volts on the 4 x 20% flat bottom holes and the prime quarter mix of 400/100 kHz was normalized to 2.75 volts on the 4 x 20% flat bottom holes. For the 3/4" wall tubing the 550 kHz was normalized to 4 volts on the 4 x 20% flat bottom holes and the prime quarter mix of 550/130 kHz was normalized to 2.75 volts on the 4 x 20% flat bottom holes. This voltage normalization is based on using the empirically derived calculation for the tubing's prime frequency and transferring those values to the mixed voltage output This is n.ccomplished through the reiationship with the quarter frequency of 100 kHz for the 7/8" by 0.050" wall tubing and the 130 kHz for the 3/4" by 0.043" wall tubing. Using the quarter frequencies in the mix algorithm the voltage values closely replicate each other at 2.75 volts. This is the same process used as discussed above for the alternate repair criteria for ODSCC at non-dented tube support plates as referenced in NRC generic letter 95-05. The fact that both 0.043" and 0.050" wall tubing will give similar voltage responses compared to the voltage from the mother standard provides evidence that the voltage results are equivalent The above referenced EPRI studies demonstrate that the prime quarter mixes of 400/100 kHz and 550/130 kilz will provide the same ODSCC bobbin coil voltage response for 3/4" x 0.043" wall tubing as the 7/8" x 0.050" wall tubing. The prime quarter mixes are those used to voltage size ODSCC at the TSP. 6
2.3.2 Growth Rate Adjustments For consistent comparison, growth rates have been normalind to 1 EFPY and were adjusted to the Braidwood 1 Cycle 7 operating temperature of 612 degrees Fahrenheit using the methodology of Reference 7, 2.3.3 Growth Rate Bound For each indication the BOC voltage is plotted against its corresponding growth rate in Figure 4. The figure includes over '2,600 indications from ten European units. Three observations are made from the data.
- 1. From 0 to approximately 5.0 BOC Volts, growth is bound by 11 Volts /EFPY. The Braidwood I repair criteria and therefore the largest indications left ir senice, less than 3.0 Volts, are within this range. This bounding growth rate is comparable to the bounding growth rate for domestic plants when all growth rates are adjusted for temperature. Braidwood I growth rates are bound by the European data.
- 2. There is a BOC voltage dependency to the growth rates, which is bound within a given BOC voltage range.
- 3. The number ofindications with large growth rates does not increase significantly with BOC voltage wv.iin the range of the Braidwood I repair limit.
2.4 Change in Frequency of Large Growth Rates with increase in BOC Voltage As discussed earlier, the growth of a distribution ofindications can be observed at European plants where high voltage repair limits minimize the number ofindications removed from service. The progression ofindications from a single steam generator is shown in Figures 5 through 7 for Cycles 9,11 and 12 (Cycle 10 is omitted because a full SG inspection was not performed). The figures show how the distribution grows toward larger voltages. Similar data is available for Braidwood Cycles 4 and 6, which are the only full cycles of operation at Braidwood where an ARC was implemented. The progression ofindications from Braidwood steam generator C for Cycle 6 and all steam generators for Cycle 4 are shown in Figures 8 and 9, respectively Cycle 5 is not shown due to a mid-cycle inspection where the larger indications were removed from service and all SG's are shown for Cycle 4 because there was a mid-cycle inspection on SG C which also removed the largest indications from senice. A similar trend of the distribution growth is evident. Knowing how the BOC voltage distribution grows in combination with an understanding . of how the number ofindications with large growth rates increases with the increase in BOC voltage provides a means of predicting the voltage growth rates. In order to understand how the number oflarge growth rates increase with an increase in POC voltage the European data was further analyzed. An analysis was performed which determined the growth rate cumulative probability distribution (CPD) for difTerent ranges 7
4 of BOC indications from 0 to 4.0 Volts. The 22,600 indications were sorted, binned and i the cumulative probability of each bin determined for each range of BOC indications of interest. The data is shown in Figure 10 and Figure 10a. Each curve in Figure 10 and Figure 10a represents the growth cumulative probability distribution for a specified range of BOC indications (e.g. all indications with a BOC voltage from 0 to 1.0 Volts have a growth CPD shown by the curve formed by closed diamonds). Figure 10 and Figure 10a shows that for a significant and mature distribution as the BOC increases there is an increase in the growth rate distribution, however the increase is not significant. This data is relevant for Braidwood because it shows how the growth rate distribution will change as the distribution matures and more indications fillin the higher BOC bins. BOC data from Braidwood Cycle 6 and Cycle 7 will be used to demonstrate how this data can be used to assess Cycle 7 growth rates. For Braidwood 1 SG C for BOC-6 there were 191 indications greater than 1.1 Volts, the growth rates at the 95%,99% and 99.5% probability (CPD) for BOC voltages from 1.0 to 1.5 volts (solid diamonds curve) are 1.1,2 2 and 3.0 Volts /EFPY, respectively. For Braidwood i SG C for the BOC-7 there are 181 indications greater than 1.6 Volts, the growth rates at the 95%,99% and 99.5% probability for BOC voltages from 1.5 to 2.0 (solid triangle) are 1.5,2.9 and 3.8 Volts /EFPY, respectively. This indicates that as the distribution shifts from 1.5 to 2.0 Volts BOC from Cycle 6 to Cycle 7 the growth rates will increase on the order of 0.4 Volts (e.g.1.5 - 1.1 Volts /EFPY) at the 95% probability to 0.8 Volts /EFPY at the 99.5% probability. This car, be related to the Braidwood EOC-7 analysis for steam generator C to identify the conservatism in the EOC predictions used to select the conservative reactor coolant dose equivalent iodine (DE l-131) limit. The Braidwood EOC-7 predictions use a voltage dependent growth rate methodology discussed in Reference 2. Steam generator C growth rates are segregated into bins defined by BOC voltages in the 0 - 0.5,0.5 - 1.0,1.0 - 2.1 and > 2.1 ranges. The average growth rates in the highest two bins are 0.59 Volts /EFPY (applied to 1002 BOC indications from 1.0 - 2.1 Volts) and 3.2 Volts /EFPY (applied to 134 largest BOC indications), respectively. The upper two bins result in the majority of the predicted leak rate. The average growth rate of the entire distribution is 0.29 Volts /EFPY. The difference between the average growth rate for the entire distribution compared to that used on the larger indications shows the significant increase in growth rate applied to the larger indications using the voltage dependent growth rate methodology compared to GL 95-05 voltage independent growth methods. Braidwood 1 SG C Cycle 6 growth rate cumulative probabilities were calculated and plotted for similar ranges of BOC voltages as for European plants shown in Figures 10 and 10a. The Braidwood data is plotted in Figure 11. Review of the plots shows that at lower BOC voltage ranges, where there is a significant number ofindications, the Braidwood 1 CPD is comparable with the European data with the Braidwood CPD curve shifled slightly toward the right or higher growth rates. At higher BOC voltage ranges the Braidwood I curve is shifled significantly toward the right or higher growth rates. This is due to the small number ofindications, which have grown to this BOC size, being led by 8
the more susceptible and higher growth rate tubing. The curves for the higher DOC
. voltage ranges are expected to shill to the left as more indications grow into these voltage ranges and the more susceptible high growth rate indications are removed from service.
This conclusion is supported by the data from other units in Figures 4 - 7. 3.0 Voltage Dependent Growth Rate Methodology Benchmark on European I)ata An evaluation of available industry data was performed on a plant by plant basis to identify plants with a higher frequency oflarge growth rates at higher BOC voltages. Plant Z7.-l was identified as such a plant. Figure 7 shows the growth rates plotted versus BOC voltage. This figure shows a highet frequency oflarge growth rates at higher BOC voltages. This condition is similar to that observed at Braidwood I during Cycle 6 which led to the under-prediction of the tail of the EOC distribution. Plant ZZ-1, with its voltage dependent growth rates, is therefore a good candidate to benchmark the voltage dependent growth methodology developed in Reference 2, to predict EOC leak rates for Braidwood 1 EOC-7. The voltage dependent growth rate methodology was benchmarked in the following steps using the EPRI STEIN code to predict the EOC distribution using the methodology outlined in NRC Generic Letter 95 05. Step 1: Predict the EOCn distribution using voltage independent growth rates and n-1 growth rates (case 1), Step 2: Compare the predicted EOCn distribution te 'he as-found distribution. Step 3: If the EOCn distribution is under-predicted determine the EOCn distribution using the Cycle n growth rates (case 2) Step 4: If the distribution continues to be under-predicted use cycle n voltage dependent growth rates to predict the EOCn distribution (case 3). Step 5: The voltage dependent growth rate methodology is described in Reference ?.. The only difference is that instead of the highest growth bin being determined based on a set number ofindications (.>200) the highest growth bin is determined by evaluating the data to determine at what BOC voltage there is an apparent increase in the frequency oflarge groveth rates. For this case the growth rates were binned into 3 separate growth bins sorted by BOC voltage. The bins were from 0 - 5.0 Volts, 5.0 - 8.0 Volts, and > 8.0 Volts. This iesulted in 81 growth rates in the highest growth bin,170 in the middle growth bin and 1218 in the lowest growth bin. The EOC distribution is then calculated for each range of BOC voltages using the corresponding growth rates. The three distributions are combined to obtain a single EOC distribution. The final distribution is then compared to the as-found distribution: NDE uncertainty and the POD is applied consistently for all cases. 9
t
. The results of each case compared to the as-found distribution are shown in Figures 12,13 and 14 for cases I,2 and 3 respectively. The figures show that using voltage independent growth rates, including data from cycle n, under-predicts the as found distribution in the tail (Figures 12 and 13). Ilowever when voltage dependent growth rates are used, the tail of the distribution is conservatively predicted (Figures 14 and 14a). This benchmarks the voltage dependent growth rate methodology developed and described in Reference 2.
This benchmark is in addition to the benchmarking performed on 6 Comed steam generators discussed in Reference 5. In all cases the benchmarking resulted in conservative EOC distributions compared to as found distributions. 4.0 .MSLB Leak Rate Sensitivity to Growth Rate Bin BOC Voltage. An evaluation was performed on the sensitivity of EOC leak rate to voltage dependent growth rate, BOC voltage bin cut-oft. In particular the analysis was performed on Braidwood i EOC-7 projections for SG C. Leak Rate analyses were performed by Westinghouse using the licensing basis methods with the exception of the voltage dependent growth rate methodology described in detail in Refeience 2 and summarized below. The application of voltage dependent growth rates segregetes the growth rates by the indication BOC voltage. The growth rate is then coupled with the BOC voltage, which is binned into 3 or 4 bins (depending on the available range of the bin). When performing the EOC calculations an EOC-7 distribution is determined for each BOC growth bin independently using the binned BOC-7 indications, BOC-7 repaired indications and growth rates, which corresponds to the BOC-6 voltage for the same bin. The EOC distributions of the bins are then combined and a 95%/95% leak rate is calculated for the overall distribution including the consideration of indications restricted from burst (IRB). , The sensitivity analyses that were performed move the BOC voltage cut-ofrof the last bin from 1.1 Volts to 2.3 Volts with analyses performed at 1.1,1.6,1.7,1.9, 2.1 and 2.3 Volts. Table i summarizes the results of the leak rate sensitivity, which is also plotted on Figure 15. 4 l to
. Table 1: Leak Rate Sensitivity Results Top Bin Cut-OfP# # of Growth Points Leak Rate (w/lRB's) Largest Flaw (0.3) (X) (n) (en0 (Volts) 1.1 (90 day case) 210 57.1 14.5 1.25 101 77.1 15.7 1.45 (RAI case) 46 100 17 1.6 (RAI case) 25 98 17.1 1.7 18 104 17.2 1.9 11 109 17.8 2.I 7 122 18,2 23 3 96 17.4
- 90 Day Report Case three bins are used at: 0 - 0.7,0.7 - 1.1, and > 1.1 Volts
# Remaining Cases use four bins at: 0 - 0.5, 0.5 - 1.0,1.0 - X, and > X Analyses were performed using a POD of 0.6 for 475 days. EPRI NDE uncertainty for probe wear and analyst uncertainty was applied.
The sensitivity of EOC-7 leak rate to voltage dependent growth rate BOC voltage bin cut-otThas been performed which resulted in the highest calculated leak rate of 122 gpm at MSLB conditions specified in room temperature gallons. 5.0 Direct Approximation of Leak Rate A method was developed to directly approximate the leak rate expected for Braidwood 1 EOC-7. The method was developed to assess the conservatism in the leak rate calculated using Generic Letter 95-05 methods modified to account for voltage dependent growth rates as discussed in the previous section. A flow chart showing the method is included as Figure 16. The method is based on the fact that there are only two significant differences between Braidwood Cycle 6 and Cycle 7 and the affect on leak rate can be accounted for. The two differences are cycle length and the shift of the BOC indication distribution toward larger indications. Both beginning of Cycle 6 and beginning of Cycle 7 distributions are based on a 3.0-Volt repair criteria. An evaluation of the contribution ofleakage from indications determined that approximately 90% of the as-found leakage came from indications, which had a BOC voltage greater than 1.0 Volt. Braidwood began Cycle 6 with 298 indications in SG C which were greater than 1.0 Volt. Therefore,90% of the as-found leakage of i 1.5 gpm came from indications greater than 1.0 Volt. Braidwood Cycle 7 began with 661 indications greatcr than 1.0 Volt. Therefore, an estimate of the EOC leakage can be made by multiplying the EOC-6 MSLB leak rate by the ratio of the number ofindications greater than 1.0 Volt at the beginning of Cycle 7 to the number at the beginning of Cycle
- 6. This result is then increased to account for the additional days of operation greater than 500"F during Cycle 7 compared to Cycle 6 by multiplying by the ratio of the number of 11
days in Cycle 7 to the number of days in Cycle 6. The resulting leak rate of 30 gpm
. indicates that there is sufficient margin in the MSLB leak rate (122 gpm) used to establish conservative operating limits on the reactor coolant dose equivalent iodine (DE l-131).
6.0 Conclusions
- 1. An upper bound on growth due to ODSCC at the TSP is 11.5 Volts /EFPY based on evaluation of Comed, Domestic and European indications within the range of Braidwood I repair criteria (3.0 Volts).
- 2. The distribution of growth rates is predictable cased upon review of growth rate behavior in almost 23,000 indications.
- 3. A Braidwood 1 EOC-7 maximum leak rate, in the event of an MSLB, of 122 gpm (Room Temperature)is predicted based upon leak rate sensitivity studies of voltage dependent growth rate bin cutoff. This increases the EOC leak rate projection, during an MSLB, from 57.1 gpm (Reference 2) to 122 gpm.
- 4. Braidwood I is safe for full Cycle 7 operation with a Dosc Equivalent lodine limit of 0.05 pci/g Braidwood has administratively limited reactor coolant DE l-131 to 0.05 pei/g and a Technical Specification amendment request for operation at this limit has been submitted for NRC approval.
7.0 References
- 1. Braidwood Unit 1 Cycl: 6 Interin Plugging Criteria Report (90 Day Report),2/96
- 2. Braidwood Unit 1 Cycle 7 Interim Plugging Criteria Report (90 Day Report),
August 1997
- 3. Braidwood Unit 13.0 Volt IPC Leak Rate Under-Prediction Root Cause Report, SGRVP-97-0012, June 16,1997
- 4. U.S. NRC Request for Additional Information Regarding Reductions in the Reactor Coolant Dose Equivalent lodine Levels - Byron and Braidwood Stations, October 7,1997
- 5. Comed Response to Request for Additional Information Regarding Proposed Technical Specification for the Reduction in Dose Equivalent lodine, October 15, 1997
- 6. U.S. NRC Letter to Commonwealth Edison, Projected Braidwood Station Unit 1, End of Cycle Steam Generator Tube Leakage Due to Outer Diameter Stress Corrosion Crack Indications Restricted from Burst, November 13,1997 7, Correlation ofilot Leg Temperature with Rate of Steam Generator Tube Corrosion, Dominion Engineering Inc. Draf) Report, February 1993
- 8. EPRI Report " Steam Generator Tubing Outside Diameter Stress Corrosion Cracking at Tube Support Plates Database for Alternate Repair Criteria, Volume 2: 3/4" Diameter Tubing," NP-7480-L, October 1993
- 9. FPRI Report " Steam Generator Tubing Outside Diameter Stress Corrosion Cracking at Tube Support Plates Database for Alternate Repair Criteria, Volume 1: 7/8" Diameter Tubing," NP-7480-L, Rev.1, December 1993
- 10. EPRI PWR S/G Examination Guidelines, '.'olume 2, Rev. 4. July 11,1996 l 12
Figure 1 ~ Braidwood m and Byron Timeline '-
? G m
E '
& st # 2 e m -
d# pe 28P u .e s,-! -?c
- c. co Q uaU c 15 a o 5 '.. *o
< g o. m
<m ec a u
<cu-m ao u' u'-
m -m Q" m cc v a g-u$" M EUU kiU d b" 3 sv~b m g
.= #
<E3o, $ ."o m E n C._. " - uox > e -oa c. O o x$ o ma 3mm o u
>jt2 >m mY3 "as 20'e c.
-YE .'s .
2m
- o. 5 m
2 2 y <.a m 3/94 2/95 10/95 10/96 4/97 10/98-5A 5B 6A 6B 7 Bra.d. i l
- 6 7A 7B Byron 8 4/96 10/94 10/95 11/97 m%
20 o 522 G x5 ?ux .a x [dgu , gg9 a ve c pd gqx 4-0
.* g$
pn =S c~ 3u <mU gm v g,
>e" Em{ W, e oC Eq
- o. T $ ei 3 Sv
-auh gou %
$g > j .j %
- o. c. e
"' .E Significant Indication of ODSCC at TSP First Identified at Braidwood Cycle 3 and Byron Cycle 4
Figure 2 . Comed Data Growth > = 4.0 Volts /EFPY All Data Temperature Corrected to 612eF 12.00 i
- Brwd Cycle 4 (1 Volt IPC) - .l.i E Brwd Cycle SA (1 Volt IPC) J l1.00 . A Brwd Cycle 5B (1 volt IPC) -
x Brwd Cycle 6 (3 Volt IPC) x Byron Cycle 6 (I Volt iPC) 10.00
- Byron Cycle 7B (3 Volt IPC) !~
i 9.00 d x
- n
'5 8.00 , -X X
> e
'6 s
2 x e i O 7.00 X e , e s
- X 6.00 -x- x- x
- x
- x .
5.00 ,- x a e
- i e x .e
. .N x x 4g A X X XX X # X 0 0.5 I 1.5 2 2.5 3 3.5 l BOC Voltage, Volts l
1/6/93 th50 PNI ComeJGmathv5,x]s Ch Comed data
Figure 3 , Domestic Data Growth > = '4.0 Volts / EFPY
- Bmd Cycle 4 (1 Volt IN "
. All Data Temperature Corrected to 612*F N Brwd Cycle SA (I Volt IPC) a Brwd Cycle SB (1 Volt IPC) i
+ x Brwd Cycle 6 (3 Volt IPC) 11.00 x Byron Cycle 6 (i Volt IPC) . .
I
- Byron Cycle 7B (3 Volt IPC)
+ 7/8" Plant (2 Volt IPC) I 10.00 's 3/4" Plant (I Volt IPC)
- l 3 9.00 N.
W-s x
- 8.00
, X- x 5 .h x
- 7.00
- c
- g. ,
x 1 6.00 x x x x e 5.00 o'
- C 3
e e *
* * * . .N x x
[. 4.00 - O 0.5 i 1.5 2 2.5 '3 3.5 BOC Voltage, Volts k:SGRIMIFS:Comedgrowthv5 > xis
Figure 4 , i . . Combined European Data, BOC Vs. Growth (Volts /EFPY) - (22,647 Indications) I 25 3.0 Volt IPC Maximum Growth I 20 - i 1 . l l 1 k 15 - *
- b i
. o i 3 I . . . .
8 I . e . , . . . , . , , f 10 ,- - g
+
+
- a *,. . .
I: i
. is . . ,0 . 2, . .
+ ,*
1 .i
-5 l 0 2 4 6 8 10 12 BOC Voltage, Volts
- t40te
- Voltages Corrected for Different Ca"bration Procedures and Growth Rates Corrected for Temperature 1/6/98 2.36 PM atidatai xis Ch BOC vs Growth
Figure 5 Cycle 9 BOC Voltage vs. Growth Volts / EFPY
~
18
~
[Cy% 9 Growth Volts / EFP]Y; - 16 14 12 g 10 UJ e z $8 - i } $* a a 4 aa6 oA Aan aj a 2 , A a1
^
O -- a 8" , ,
-2 0 2 4 6 8 10 12 14 BOC Voltage, VoMs 1/7SB 10 37 AM W sis Ch ZZ-1SG1cy9bocvvp
Figure 6 Cycle 11 BOC Voltage Vs. Growth Volts I EFPY 18 io Cycle 11 Growth Volts / EFPY ! 16 14 12 @ 10 & o 2 'o 8 0 6 0 o oo O O o o o e o o o so o o go oo o 0 0 0 o o o o
-2 0 2. 4 6 8 10 12 14 BOC Voltage, Volts
Conclusion:
Indication Distribution Grows to Higher Voltages 117/9810 35 AM euroevoi uts C5 ZZ-1 SG1 cy11bocwp
; Figure 7 ,
Cycle 12 DOC Vs. Growth Volts / EFPY
~
18 i
! . Cyde 12 Growth Volts / EFPYi
+
16 - - = - - l i
- i i 14 !
I 12 i l l [ 10 ! E' ! u U l 88 - 5
- s . *
, O 6 +- ! O +e , l , . . . . . l
. + +
4 ++
,, +--
.- -+ .
O _ p, #_ + , __..
- -2 e i 0 2 4 6 8 10 12 14 BOC Voltage, Volts
Conclusion:
Distribution Continues to Grow to Higher Voltages i 1/7G810 36 AM W wts Ch ZZ-1SG1cy12
l 8 m i s a 5 3 a 1 3
* ?
~
R U o! O h S. h ET i s# 8! Wo i
- s8
<=
eoo y o Em* 0 0 o o
- W ,o 0
k o e ce O O h 00
-{ 0 e-3 i
o o a g
.l
)m O o o
O d j c i, O o o , 85 $ g O 0 O 0 o o o I o oj g
]i
- - 8 !
?
?
?
3
?
?
?
o 8 S 8 o 8 v' 8 N 8 o 8 G l e Add 3/stioA 'tismojo
,I
> 1itliIil i:! !l, ii i [ :tf 'lil!' , ,,t t , ;lI ! ! ii5 f.
' 2 - 2 :
- q :i' 3
a _ a e g _ 5 a t a 2 l o V - 6 C- ^ O , B a Y s A _3 a A a G a 2 S n i a, 6 ht -g l e w a c y o G r a# g, C g . t a^_a n _ s e e g h, g i r r n ^ff t a u , l 9 a e L a - a8a &g 5 Vo D
' 1 6 Y u+
r gC aa A a^# C-O F P i a F G ;- - B E/ S a s t
- , l o
4 a^a$
) -
Y ) s V P G' a a -g 0 F S-E ( l l a a 8 A - ht i n A a4
#g ^ 1 <
w 4; 6 ht o a a r 5 # ' w G L u 3 o 6 ! r e n 3 ' G l c y 3( - j 1 A C G-
)
s # d n S o 1 t 6 n o o i n a i c ht w i 5 d U d I n wu 0 i a O d r r o l l A G ^ B 5 o ( 4 k w : C-d p - n
^
w i a G i. S L o r 3 i s % _ B aE u k - _ l c S' n o C . 1E. 0 - 8 1 6 1 4 1 2 1 0 1 8 6 4 2 o 2-
'e
>ta23 r.e 4 .EO jru I ,
. ! -l ,
i !.-
Figure 10 . European Growth CPD for Selected BOC Voltage Ranges (22,600 Indications) 1.1 T 0.9 -- +++
+ - _
,# j -+-CPDF 0-1.0V l g
# i
-e-CPDF 1.0-1.5V ,
0.7 - -*-CPDF 1.5-2.0V ;_ l
; -*-CPDF 2.0-2.5 1
1'
-o-CPDF 2.5-3.0 l
-+-CPDF 3 0-3 5 !
- u. it ! a
@ 0.5 -
j CPDF 3.5-4 0 o h i l il f 0.3 r 0.1 -- neoaooeoeooueoeoooeoooooo,o,o
-0.1 Growth Rate, Volts /EFPY
Conclusion:
Modest increase in High Growth Frequency with increasing BOC Voltage up to 4.0 BOC Volts 1!7/981154 AM alidatat ris Ch aR data CPDF <4 OV
Figure 10a European Growth CPD for BOC Voltage Ranges 1.01 . 1 2:22 2- ???? ;- ;22:2^22:222^^^^^^^2:22:2);^^;;;-tercree-* _.._s [ ~ g ^:22:22 2: 4 },_ 74???II?b " ' ----- ^' O.99 - - - l--+-CPDF 0-1.0V j
-e-CPDF 1.0-1.5V j 0.98 - - - - - -*-CPDF 1.5-2.0V i
-*-CPDF 2.0-2.5 i
,-e-CPDF 2.5-3.0 !
i o j-+-CPDF 3.0-3.5 g 0.97 - "- ( CPDF 3.5-4 0 0.96 0.95 - 0.94 - - - - - - - - - - 0.93
%p %9 b % b % % % 4 S % N % % % % % % N b A
% %- >N -
%- %- b- %- %- %- A- % >- A- % %~ % %} %b % Q- Q Q-
' Growth, Volts /EFPY f 1,7/9812 06 PM andata1 wfs Ch 0 4V BOC CPD >3 OV-EFPY
-ee -
Figure 11 i i Braidwood 1 SG C Cycle 6 Growth CPD for BOC Voltage Ranges , : 1.1
.................... f ~~~~~~~~~
.. ................................./ __ __ _____ J 0.9
__ I---------------- , 0.7 - l-+-CPD 0 - 1.0 (1778) { l-e-CPD 1.0 - 1.5 (288) l c l-*-CPD 1.5 - 2.0 (26)
- n. 0.5 -
. -2 0 ____ ________ _____
!-*-CPD 2.0 - 3.0 (11) '
_ ..____~_ ..________ b 0.3 - l l a f I 0.1 - l umm 6 - , - , - . .. _,_ __, _m,____. [ 9 @OUd0@h0N@> O @ @ @ @ h Y Y Y @ d @ @ @ ,97,pi
-0.1 ,
Growth, Volts /EFPY l 1/7S8120 PM cce6cpdf W. Ch growth cpd
Figure 12 . Plant ZZ-1 SG-1 Cycle 12 Actual Vs. Predicted EOC Voltage Distribution . Cycle 11 Voltage independent Growth Rates 1000 E EOC-12 Acu2al [O EOC 12 Predicted 800 - 700 - E
$ 600 -
g 5
;E 500 -
4 o' 400 - ' 5 300 - ! i 200 - 100 - - - -- o "'" ! l I E EES#D"*" --- -- - - 4O & 44 & @ @ O &@ $$$$$$$$++9 4 EOC-12 Voltage, Volts 1,7/96 3 31 PM ZZ-1bnchmk ris Ch4. SG-1 Act VS. Pred Cy12 . 1
Figure 120 . Plant ZZ-1 SG-1 Cycle 12 Actual Vs. Predicted EOC Voltage Distribution , Cycle 11 Voltage Independent Growth, BOC Voltages >8 Volts 45 _ {EEOC-12 Actual l
~
'lO EOC 12 Predicted j 40 -
35 - - - 30 - - E is
.2 25 - - - - -
_E_ o E' 20 E - i<
- s _
Z 15 - -- - - -- - ! 10 _ _ _ ._ _. ._. _ ; 5 - - - - - - - - - 0 . . , i i> itoLLu m.
. i . . .
, ,,, , , , .s s.s.s s.o.ocs.. . ..
EOC-12 Voltage, Vohs Im98 3 32 PM ZZ-1bnchmk v1s Ch4a. sgiactnpred cy12>8V
a - Figure 13 . Plant ZZ-1 SG-1 EOC-12 Voltage Distribution Using Cycle 12 Voltage Independent Growth Rates .
- 1400 lO* Tubes Predicted j
'E Actual !
l 1200 1 1000 8 z j 800 _E o n j soo __ _ E 5 400 -- - 200 - -- -- -- - 0 E b1'~ - - - A S EOC-12 Voltage, Volts II7S6 3 34 PM ZZ-1bnchmk uts Ch S. cy12cy12 grwttrAtrwjeg gr
Figure 13a Plant ZZ-1 SG-1 EOC-12 Voltage Distnbution Using Cycle 12 Voltage Independent Growth Rates, BOC Voltage > 10.0 Volts 60 O Predided ; 5 EOC-12 Adual l 50 40 E 2 - li N I y 30 - - 8 E E z 20 - - 10 - -- - - - - - 0 " " " - * " " " 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 EOC-12 Voltage, Volts 1/7/98 3 34 PM ZZ-1bnchmk uts Ch Sa.cy12 w cy12 G-4w oti, l
Figure 14 - Plant ZZ- ; G-1 EOC-12 Voltage Distribution Using Cycle 12 Voltage Dependent Growth Highest Bin BOC Cutoff at 8 Volts 1400
;OPredicted Total j lEActual !
1200 1000 D .5 E = f 5 800 3 l 0 ' D 600 - -- _ 3 xn . E n o 2 400 -- - 200 - - - - 0 = E % Tb rs rt_ m m _ _ _ _ _ _ ' Top 0 1 2 3 4 5 6 7 8 9 1G 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 of Bin , EOC-12 Voltage, Volts
..a 3 36 PM Ch 6 at cyc12cy12grwm vtdepZZ-1bnctrnk A
Figure 14a . Plant ZZ-1 SG-1 EOC-12 Voltage Distribution Using Cycle 12 Voltage Dependent Growth Rates . Highest BOC Growth Bin Cutoff 8 Volts, BOC Volts > 10.0 Volts 45 , lO EOC-12 Predcted i 40 - - - - i 35 - n 30 - C 1 g 25 - c ' g 20 - -
.c
~
b Z 15 -- - - -- 1. g 10 - - - i 5 -- -- - -- - 0 - - - N n 1 - 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 EOC-12 Voltage, Volts
Conclusion:
Voltage Dependent Growth Methodology is Benchmarked Using Cycle 12 Voltage Dependent Growth Rates on BOC-12 with 81 Indications in the Highest Bin ! 177S8 3 37 PM Ch sa cy12wty12,--^..vup3V ZZ-1bnchmk uts 1 s , -,, -
Figure 15 , Braidwood 1 SG C EOC-7 Leak Rate Sensitivity to Variation in Top Bin Cutoff and Number of
~
Growth Points in Top Growth Bin .
.140 120 --
Approx. 50 Points Required for Statistical Significance 100 - l-+-IRB LP.. (gpm) l 3 Growth Points in Top Bin Cutoff at 2.3 VBOC E
- o. 80 -
2 2 E . I 60 J 210 Growth Po'. in Top Bin , Cutoff at 1.1 VBOC
. GL 95-05 Guidelines 40 4 20 0
0 0.5 1 15 2 2.5 BOC Voltage at Top Sin Cutoff L 1r/G8 3 22 PM tunsensdrvty als Ch IRB LR Sensevty
,s _, ., - - .,, -, , _,y. , - _,.4.
.. - , . __.,,.m .
Figure 16 Cycle 6, 3.0 Vcit Cycle 7, 3 0 Volt - Repair Criteria Repair Criteria Majority of Cycle 6 Leakage from Ind.'s V tx t 1.0 to 3.0 V I T Number of BOC-6 Ind.'s V rw me 298 Indfs Leak Rate is 1.0 to 3.0 \ l1.5 gpm v ! i Number of BOC-7 - Leak Rate is Ind.'s V 3,c "* 26 gpm 1.0 to 3.0 V Cycle 7 60 Days Cycle 7
"" Leakage 30 Longer than Cycle 6 gpm
-_ _. __ . _ _ _ _ .. .. - . _ ~ - _ - - - -- - .-
M 13 '90 10:30 FR COPED SEC HRIJD 015 450 3595 TO 16306637171 P.02/02
'Mcmerandum i
Dats:' January 13,1998 To: Irene Johnson Licensing Operations Director
Subject:
Braidwood Unit 13.0 Volt IPC, Full Cycle Operation Tec}mical Basis Report, Supplement to Braidwood Unit ! Interim Plugging Criteria Report August,1997
Reference:
- 1. November 13,1997 letter from D. Lynch, Office of Nuclear i Reactor Regulation to I. Johnson, transmitting the NRC's Review of Comed's Response to RAI l
- 2. August 14,1997 letter from E Standy to NRC Docur.wnt Control Desk,"Braldwood Station Unit 1. Steam Generator Interim Plugging Criteria 90 Day Report"
- 3. December 11,1997, Meeting between the Commonwealth Edison Company and the Nuclear Regulatory Commission in Reference 1, the NRC stated that there was Insuscient data to support the Braldwood Unit I projected End-of-Cycle 7 IPC leak rate value of 57.1 gpm (Room T/P), as submitted in Reference 2. Based upon this feedback, Comed performed a sensitivity study to determine how the application of voltage dependent growth rates effect the projected End-of-Cycle IPC leak rate value. The study dete:Ynined that projected leak rate value continues to increase to a bounding value of 122 gpm (Room T/P), then become statistically unrollable. The findings of this study was presented to the NRC in a December 11,1997 meeting. Reference 3. At the conclusion of the subject meeting the NRC requested Comed submit a report documenting the rnecting presentation and to omcially revise the Braidwood Unit 1 projected End-of-Cycle 7 IPC leak rate value to 122 gpm.
The attached report documera the presentation from the December 11,1997 meeting and presents the technical basis I
- the revised Braidwood Unit 1 paiected End-of-Cycle IPC leak rate value of 122 gpm. Braidwood Station has reviewed this report and finds it to be acceptable for transmittal to the NRC. If you have any questions regarding this matter, please contact Mike Sears at Braidwood extension 2251.
Sincerely, h p J.Meister T Engineering Manager Braidwood Nuclear Generating Station Attachments
==
cc: J. Meister T. W. Simpkin D. Saccomando - Licensing Operations
.- - .--}}