ML20211Q194
| ML20211Q194 | |
| Person / Time | |
|---|---|
| Site: | Byron, Braidwood  |
| Issue date: | 10/15/1997 |
| From: | Hosmer J COMMONWEALTH EDISON CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 9710220168 | |
| Download: ML20211Q194 (49) | |
Text
1 October 15,1997 United States Nuclear Regulatory Commission Wash ngton, D.C. 20555 Attention:
NRC Document Control Desk
Subject:
Response to Request for Additional Information Regarding Proposed Technical Specification for the Reduction in Dose Equivalent lodhe Byron Nuclear Power Station Units 1 and 2 Braidwood Nuclear Power Station Units I and 2 NRC Docket Numbers: 50.454. 455. 456 and 457
References:
1)
J. Hosmer letter to the Nuclear Regulatory Commission dated January 31,1997, transmitting Proposed Technical Specif=ation Amendment Request for Byron Station 2)
H. Stanley letter to the Nuclear Regulatory Commission I
dated September 2,1997, transmitting Proposed Technical Specification for Braidwood Station 3)
J. Hosmer letter to the Nuclear F.cgulatory Commission dated August 14,1997 4)
D. Lynch letter to 1. Johnson dated October 7,1997, transmitting Request for AdditionalInformation References 1 and 2 transmitted the Commonwealth Edison Company's (Comed) proposed Technical Specific tion amendments to lown the reactor coolant dose equivalent iodine concentration for Byron and Braidwood Stanon. Refer nce 3 transmitted the voltage dependent growth rate methodology used to predict end of cycle distributions. As a result of the review of the methodology, the Nuclear Regulatory Commission (NRC) transmitted a request for additiona'.
information (RAI) via reference 4. Attached is Comed's response to the RAI.
I Comed has concluded that the results in response to :he RAI support the Braidwood-l 90 Day Ol Report, that the growth rate distributions used in predicting the Braidwood I end-of-cycle 7 steam I
generator tube conditions are sufTiciently conservative. The growth rate distrioutions and resultant predicted leak rates are shown to be conservative by benchmarking the methodology on six different steam generators.
9710220168 971015 PDR ADOCK 05000454 p
((l h\\
h g.
K:niawbmratgen icarain s
A Umcom O,mpan)
(
U.S. Nuclear Regulatory Commission 2-October 15,1997 If you have any questions, plvase contact this office.
Sincerely, 8
John B. Hosmer Engineering Vice President Attachment -
cc:
. Regional Administrator - RIII Senior Project Manager - NRR Byron Project Manager-NRR Senior Resident Inspector - Byron Senior Resident Inspector - Braidwood -
Office of Nuclear Facility Safety - IDNS i
KmlatbyttsdatgenModrai3:2
Response to NRC Staff RAls on Braidwood 90 Day Report BYRON /BRAIDWOOD All predicted SLB leak rate values provided in response to questions in this RAI are equivalent volumetric rates at room temperature (i.e., leak rate at SLB conditions condensed and measured at room temperature).
1.
Requested Information For the predictions of the EOC conditions at Braidwood, Unit 1, Comed used a lower voltage value of 1.1 volts for the highest bin. Since the behavior of voltages between 1.1 and 1.5 appears to be different than the behavior of voltages above 1.5
- volts (ref. Fig. 5-3 of 8/14 submittal), discuss the appropriateness of the lower voltage value of 1.1 volt other than that the 1.1 volt limit provides 200+ indications as mentioned in OL 95-05. Assess the change in the calculated EOC conditions (i.e., voltage distributions and MSLB leakage) for Braidwood-1 as the lower voltage value of the highest voltage bin changes from 1.1 volts to higher voltages, such as 1.45 and 2.0 volts, using NRC approved methodology (i.e., probability of detection (POD) = 0.6).
Response
Annronriateness of Lower Voltage Values for Growth Bins As discussed in the Braidwood-190-day report, the Cycle-6 growth data show an
-increased frequency of larger growth indications with increasing voltage but the magnitude (AV) of the growth values is not strongly voltage dependent. The voltage dependent growth methodology was developed to vary the frequency (probability) of the larger growth rates as a function of beginning of cycle (BOC) voltage. Ideally, the-number of indications used to develop a growth distribution over a given voltage range should be the same as the number of BOC indications over the same voltage range for which the growth distribution is applied. Then, the resulting number oflarge growth occurrences in the projections would be the same as for the population used_ to develop the growth rates.
However, the number of BOC indications in the higher voltage ranges increases due to growth of the indications and the app!ication of a probability of detection (POD) to define the BOC distribution. The net effect of the voltage dependent growth rates is to apply the larger growth rates to larger BOC indication voltages in the BOC distribution than would occur with a voltage independent growth distribution.
The largest influence of the voltage dependent growth rates occurs from the upper one or two voltage growth bins.
Therefore, the methodology has an inherent conservatism in that the number of indications in the upper growth bin ranges is higher in the BOC distribution used for the projection analysis than in the prior t
cycle distribution used to develop the growth rates. This results from growth of the indications over at least one cycle and application of a POD to define the BOC distribution. This difference in the number ofindications between that used for the growth distribution and that for the projected BOC distribution becomes L\\ ROMANO \\0WD-EOC\\RAl\\RaLoire doc 1
10/ 4 5/97. 4 fo PM
Response to NRC F'aff RAls on Braidwood 90 Day Report significantly magnified when a POD of 0.6 is applied as this increases the BOC population by about a 1.7 factor. Thus, this POD has the effect of nearly doubling the large growth indications compared to the popuiation used to develop the growth rates. Increasing the voltage boundary for the upper voltage growth bin increases conservatism in the projections by applying the larger growths to larger BOC indications.
When the conservatism of higher voltage growth bins is combined with a POD of 0.6, the methodology becomes excessively conservative as shown by the analysis results given in later sections of this RAI response.
The selection of the voltage ranges for growth bins must be tied to the number of indications in the growth values within the growth bins. It is also necessary to have fixed guidelines for defining the voltage bins, particularly the highest voltage bin, so that the analyses are repeatable without iterations on the voltage bin size.
A population of about 200 indications for a single growth distribution (bin) provides a high confidence that the distribution is well defined. A lower bound of about 25 indications is necessary to have any reasonable confidence that the growth distribution is acceptably characterized. A minimum of 50 indications in the upper voltage bin may be more appropriate even for a voltage dependent POD so that a variability of a few indications in the growth values do not overly influence the growth distribution. For application with a POD of 0.6, which adds consenatism on the number of indications, the upper growth bin population of about 200 indications was selected and shown to produce conservative projected end of cycle (EOC) leak rates (See also response to RAI Question 3). Further reductions in the number ofindications with an increase in the lower voltage boundary of the upper growth bin results in excessive conservatism in the projected leak rates as shown in the second half of the response to this question. When POD is increased, it is appropriate to increase the upper bin voltage boundary such that a boundary corresponding to about the highest 25 to 50 indication voltages is recommended for use with a voltage dependent POD such as the EPRI POD method (probability of prior cycle detection (POPCD)).
The influence of a POD of 0.6 on the difference in the populations used to develop the growth distribution and the number of BOC indications is shown in Table la for Braidwood-1. For the recommended Braidwood-1, upper voltage e.rowth bin cutoff of 1.1 volt, it is seen that 210 indications are included in the growth distribution but that this growth distribution is applied to almost five times (955 indications) as many indications at BOC-7 for a POD of 0.6. As a consequence, the occurrence of large growth values above 1.1 volt will be almost five times as large as found for Cycle-6 which was used to develop the growth rates.
The benchmarking of the methodology was performed for Cycle-6, which had the number ofindications only 1.5 times the number used to develop the growth rates.
Thus, the benchmarking for Cycle-6 is considerably less conservative than the Cyck-7 application. Similar differences in the number of indications in the growth and BOC-7 distributions are found for the upper voltage bin thresholds of 1 A5 and 1.6 volts. For these higher thresholds, the number of indications in the BOC-7 distribution for a POD of 0.6 is more than a factor of ten higher than in the Cycle-6 distribution used to develop the growth rates and excessive conservatism in the projected EOC conditions can be expected, in the latter cases, the larger growths L \\ROMANG\\BWDEOC\\RAl\\ht01re. doc 2
10/15/97,4.50 PM
Response to NRC Staff RAls on Braidwood 90 Day Report are applied to larger BOC indications with further increases in analysis conservatism. This conservatism will not be apparent in comparisons of projected and actual distributions at EOC-6 since the BOC-6 population is less than twice the number ofindications used to develop the growth distributions.
Chance in Calculated EOC Conditions for Varied Growth Bin Voltages Development of Gmwth Distributions with Varied Bin Voltages During Cycle-6, steam generator C (SG-C) had more indications experiencing large growth than the other three steam generators, and it had the largest growth observed for the cycle. Figure 1 shows the variation of growth with BOC voltage (Vnoc) for SG-C. SG-D had the second largest number of large growths, and its growth vs Vnoc data is shown in Figure 2.
A comparison of Figures 1 and 2 confirms that SG-C had the limiting growth during Cycle-6.
SG-C also had the largest number of indications found among the four SG's at EOC-6. Therefore, sensitivity analysis to examine the effect of varying voltage growth bins on the calculated EOC conditions was carried out using the SG-C data. The methodology presented in the Braidwood-1, Cycle-7 90-day report to select voltage dependent growth requires a hybrid growth distribution that includes the top three growth values from all SG's. SG-C had the top two growth points for Cycle-6 and SG-A had the third largest growth. Therefore, a hybrid growth distribution consisting of all SG-C data and the largest growth in SG-A (to include the three largest growth values a all SGs) was used.
The hybrid growth rate data for Cycle-6 were grouped into several bins using the guidelines established in the Braidwood-1 Cycle-7 90-day report and cumulative probability distributions were calculated separately for each bin. To obtain growth distributions appropriate for use with a constant POD of 0.6 using the recommended guidelines, the hybrirl growth data was split into three bins. The width of the highest voltage bin was selected to include at least 200 indications (per discussion above); it contains the top three growth values for Cycle-6. Figure 3 shows cumulative probability distributions for the three bins selected to represent growth data for a POD of 0.6. It is evident that the distribution representing the highest voltage bin in Figure 3 shows a much higher frequency of large growth occurrence than the other two distributions representing the growth data.
In response to this RAI, two other groupings of growth distributions were also applied with a constant POD of 0.6, and they are illustrated in Figures 4 and 5.
Figure 4 shows cumulative probability distributions for the case with the lower boundary for the highest voltage bin moved up to just behind the third largest growth value in the hybrid data (1.45 volts) and total number of bins increased to four. This choice of highest voltage bin reduces the indication population for that bin to less than one quarter of that in the recommended distribution for POD of 0.6, which significantly increases the frequency of occurrence of large growth points in that bin. Figure 5 shows growth distributions for a third case where the population in the highest voltage bin boundary is further reduced by about 50% by moving the lower boundary of the highest voltage bin to 1.6 volts. A comparison of L\\ROMANG\\BWWEOC\\RAl\\RaLOlre doc 3
m m so m
Response to NRC Staff RAls on Braidwood 90 Day Report cumulative probability distributions for the highest voltage bins in Figures 4 and 5 with that in Figure 3 shows that the frequency of occurrences of large growth points increased substantially by moving the lower boundary of the highest voltage bin from 1.1 to 1.45 and 1.6 volts. As noted above, growth of the indications and application of a POD substantially increased the population of tubes for which the upper bin growth distribution is applied. The resulting population is much higher than used to develop the upper bin growth distribution and results in projections of more high voltage indications.
Increasing the upper growth bin boundary results in less reliable growth distributions due to the small population of indications while resulting in more conservative projections of voltage distributions.
EOC-6 voltage distributions and SLB leak rate estimates based on the growth distributions shown in Figures 3 to 5 (with POD of 0.6) are discussed in the following paragraphs.
Change in EOC SLB Leak Rates with Van'ed Gmwth Rate Bin Voltages The steam line break (SLB) leak rate a: alysis results using the alternate higher voltage growth bin thresholds are given in Table 2.
Cumulative probability distributions for the different growth distributions used in this evaluation are shown in Figures 3 to 5. The Table 2 results for SG C include the SLB leak rates for the actual EOC-6 distribution, projections of EOC-6 leak rates using each of the voltage growth distributions evaluated and projections of the EOC-7 leak rate for each growth distribution. The best estimate for the leak rate from the actual EOC-6 voltage distribution is obtained for a 5% NDE uncertainty which is more realistic than the 10.3% NDE uncertainty used for predictions since the actual leak rate is dominated by high voltage indications for which sizing capability is improved. The improved sizing capability for large indications is demonstrated by comparison of the field call voltage to the sizing voltage. For SG C, the limiting SO for Cycles 6 and 7, all EOC-6 projections of SLB leak rates applying the voltage dependent growth projection (Cases 2 to 4 in Table 2) exceed the actual distribution leak rate at EOC-6.
The EOC-6 leak rate for the actual distribution of 9.8 gpm is underestimated by 2 gpm when growth independent of voltage is applied.
Application of the larger voltage threshold for the highest bin, as shown for Cases 3 and 4, result in increased conservatism in predicting the leak rate for the actual distribution. The recommended growth bins of Case 2 already overestimate the actual leak rate by over 3 gpm (9.8 actual vs. 13.1 projected).
Since the recommended upper voltage growth bin threshold of 1.1 volt for a POD of 0.6 results in the predicted EOC-6 leakage significantly exceeding that calculated from the actual distribution, further increases in the voltage threshold result in unnecessary conservatism. As shown for Case 4a, the higher voltage growth bin threshold of 1.6 volts is appropriate for application with a voltage dependent POD since the number of BOC-6 indications in the upper voltage range is not increased by the 1.7 factor obtained for a POD of 0.6. Case 4a shows that applying the best estimate of POD with the higher upper bin voltage threshold for growth results in excellent agreement with the leak rate for the actual EOC-6 distribution (9.9 vs. 9.8 gpm).
l nm,umwmccomow a-4 wsms so m
Response to NRC Staff RAls on Braidwood 90 Day Report Although the larger voltage growth thresholds requested for evaluation in this RAI are not necessary for adequate methodology conservatism, EOC-7 projections of SLB leak rates were performed as summarized in Table 2.
The EOC-7 SLB leak rate with indications restricted from burst (IRBs) included is 27.2 gpm with voltage independent growth and 57.1 gpm with the recommended voltage growth bins (Cases 5 and 6 of Table 2). Application of the voltage dependent, EPRI POD with the upper voltage bin threshold of 1.6 volts (Case 7a) results in a leak rate of 54.9 gpm.
When the alternate growth bins (increased upper voltage thresholds) are applied with a POD of 0.6, the SLB leak rates increase to 100.0 gpm for alternate 1 with a 1.45 volt threshold and 98.0 gpm for alternate 2 with a 1.6 volt threshold (Cases 7 and 8 of Table 3). This is the expected result based on the discussion given above relative to the large number ofindications in the upper voltage bins when a POD of 0.6 is applied.
As shown in Table la, the upper bin voltage distribution is developed from about 46 indications for a 1.45 volt threshold and about 25 indications for a 1.6 volt threshold. The BOC-7 distribution for a POD of 0.6 has 504 and 329 indications, respectively, in these upper voltage bins. Application of these voltage dependent growth distributions, which lead to greater than ten times the numbei J BOC indications than used to develop the growth distribution, results in the excessive conserve.hm in EOC-7 leak rates found for Cases 7 and 8.
In summary, it is shown from the results of Case 2 in Table 2 that the recommended voltage dependent growth bins provide conservative predictions in comparison with that obtained from the actual Braidwood-1 EOC-6 voltage distribution. Further increases in the voltage growth thresholds are not warranted for application with a POD of 0.6.
This is further supported by the Byron-1 analyses discussed in the response to RAI question 2 given below.
2.
Requested Information For the predictions of the EOC conditions at Byron, Unit 1, Comed used a lower voltage value of 0.9 volts for the highest bin.
Since the behavior of voltages between 0.9 and 1.5 appears to be different than the behavior of voltages above 1.5 volts (ref. Fig. A-4 of 8/14 submittal), discuss the appropriateness of the lower voltage value of 0.9 volt other than that the 0.9 limit pre. ides 200+ indications as mentioned in OL 95-05. Assess the change in the calculated EOC conditions (i.e.,
voltage distributions ar.d MSLB leakage) for Byron 1 as the lower voltage value of the highest voltage bin changes from 0.9 volt to higher voltages, such as 1.45 and 2.0 volt, using NRC approved methodology (i.e., POD =0.6).
Response
Annropriateness of Lower Voltane Values for Growth Bins As noted in the response to RAI Question 1 for Braidwood-1, the effect of the voltage dependent growth rates is to apply the larger growth rates to larger BOC w>mwwvexwnwo n 5
wmm. m
Response to NRC Staff RAls on Braidwood 90 Day Report voltages than obtained with voltage independent growth. The discussion of this topic given in RAI Question 1 is also applicable to Byron-1. Table Ib provides a comparison of the number of indications in the growth bins used to develop the growth rates with the number of BOC indications. For the largest voltage growth bin, the number of indications in the BOC distributions is about a factor of 2.5 higher (654 vs. 264 indications) for the largest recommended bin with a POD of 0.6.
For the alternate growth distributions, the ratio of BOC to growth bin indications increases to about a factor of three.
These ratios still significantly increase conservatism for the higher voltage cutoffs. As shown below and in the response to RAI Question 3, the increase in the upper voltage bin cutoffs above the recommended procedure add increased and unnecessary conservatism to the analysis method;iogy.
Chance in Calculated EOC Conditions for Varied Growth Rate Bin Voltanes Development of Growth Distributions with Varied Bin Voltages Byron steam generators B and C had a comparable growth during Cycle-7B, and their growth rates were significantly higher than those of the cther two SGs.
Figures 6 and 7 show the variation of growth with he for SGs B and C. The two plots look very similar; however, above a bc of about 1.45 volts, SG-B shows several relatively high growth points and SG C has three indications with high growth between 0.6 and 1.2 volts for which there are no comparable values in the other SG. The methodology presented in the Braidwood-1, Cycle-7 90-day report to select voltage dependent growth requires a hybrid growth distribution that includes the top three growth values from all SGs. With this requirement, the differences between the hybrid growth distributions based on the Byron SG B and SG C data would be small. The hybrid growth distribution based on the SG B data includes a few more growths on the high voltage end and, therefore, appears to be more limiting than that obtained with SG C data. Hence, Byron SG B growth data was applied to Byron SG C in assessing the effect of voltage dependent growth.
However, the results would not be significantly different if SG C's own growth rates were used as shown in the analysis results given below.
To account for growth dependency on BOC voltage, growth rate data were split into several bins using the guidelines established in the Braidwood-1 Cycle-7 90-day report and cumulative probability distributions were established separately for each bin. A Byron hybrid growth distribution that included all SG B data and the largest growth in SG C (to include the three largest growth values in all SGs) was used. The distributions obtained for use with a constant POD of 0.6 are shown in Figure 8. The width of the highest voltage bin was selected to include at least 200 indications (per discussion in response to RAI Question 1) and it contains two of the three largest growth values. Figure 8 also includes the cumulative probability distribution for the case wherein growth is assumed independent of be.
It is evident that the distribution representing the highest voltage bin in Figure 8 shows a much higher frequency oflarge growth occurrence than the distribution based on growth independent of BOC volts.
L.\\ROMANG\\BWD EOC\\RM\\kai_0tre dne 6
10/15/97,4 30 l'M
Response to NRC Staff RAls on Braidwood 90 Day Report n
In response to this RAI, two other groupings of growth distributions were also applied to Byron with a constant POD of 0.6, and they are illustrated in Figures 9 and 10. Figure 9 shows cumulative probability distributions for the case with the lower boundary for the highest voltage bin raised to 1.1 volts, which reduces the indication population for that bin by about fifty percent (in comparison to the distribution recommended for POD =0.6). Figure 10 shows growth distributions for a third case where tht oopulation in the highest voltage bin boundary is further reduced by about fifty percent by moving the highest voltage bin boundary to 1.4 volts. A comparison of cumulative probability distributions for the highest voltage bins in Figures 8 and 10 show that the frequency of occurrences of large growths increased substantially by moving the highest voltage bin boundary from 0.9 to 1.4 volts. A growth rate distribution with the highest bin at 2.0 volts, as suggested in the RAI question was not evaluated because it would include only eight growth points which does not provide an acceptable growth distribution. The effects of applying the growth distributions shown in Figures 8 to 10 (with POD =0.6) on the estimated SLB leak rate are discussed in the following paragraphs.
Change in Bymn EOC SLB Leak Rates with Varied Growth Rate Bin Voltages The SLB leak rate analysis :sults using the alternate higher voltage growth bin thresholds are given in Tabl
'igures 8 to 11 show the cumulative probability
^
distributions for the growt.
... ributions used in these evaluations.
Data in Figures 8 to 10 are based w the SG-B growth data, and Figure 11 shows growth based on the SG-C data. The Table 3 results for SG C include the SLB leak rates for the actual EOC-7B distribution, projections of EOC-7B leak rates using each of the voltage growth distributions evaluated and projections of the EOC-8 leak rate for each growth distribution. The actual Byron EOC-7B leak rates are based on an NDE uncertainty of 10.3% with no value calculated at a 5% NDE uncertainty similar to what was done for Braidwood. This is because the Byron leak rate is not dominated by a small number of large indications like the Braidwood leak rates and therefore the 10.3% NDE uncertainty is used.
Based on the similarity of growth rates for SGs B and C, the growth distributions for SG B were also used for SG C analyses. This is supported by the results of Cases 2 and 2a in Table 3, which show the same leak rates for separate analyses for the SG B and C growth distributions. For SG C, the limiting SG for Cycle-8, all EOC-7B projections of SLB leak rates including the voltage independent growth projection (Cases 1 to 4 in Table 3) exceed the actual distribution leak rate at EOC-7B. This indicates that growth rates in this SG are not strongly BOC voltage dependent and all analysis methods are conservative. It is shown in the response to RAI Question 3 that voltage dependent growth is desirable to improve agreement with the actual EOC-7B leak rate for SG B.
The recommended upper voltage growth bin threshold of 0.9 volt results in the predicted EOC-7B leakage exceeding that calculated from the actual distribution and further increases in the voltage threshold result in unnecessary conservatism.
L\\ROMANG\\BWD EOC\\RAl\\Rai_Olre dw 7
10MW,4 50 PM
Response to NRC Staff RAls on Braidwood 90 Day Report O
Although the larger voltage growth thresholds requested for evaluation in this RAI are not necessary for adequate methodology conservatism, EOC-8 projections of SLB leak rates'were performed as summarized in Table 3. The EOC-8 SLB leak rate with IRBs included is 19.0 gpm with voltage independent growth and 22.8 gpm with the recommended growth bins (Cases 5 and 6 of Table 3), When the alternate
- growth bins with increasing upper voltage threshold are applied, the SLB leak rates increase to 26.2 gpm for alternate 1 with a 1.1 volt threshold and 27.0 gpm for alternate 2 with a 1.4 volt threshold (Cases' 7 and 8 of Table 3).
This is the expected result based on the discussion noted above relative to the large number of indications in the upper voltage bins when a POD of 0.6 is applied. The increases in leak rate for alternate growth bins 1 and 2 for Byron-1 are significantly less than found for Braidwood-1 in Table 2. This results as the number of indications in the BOC-8 distribution for the large voltage bins (Table Ib) are a factor of about three higher than used to develop the growth distributions compared to the factor of ten increase found for Braidwood-1.
In summary, it is shown from the results af Cases 1 and 2 in Table 3 that the recommended voltage dependent growth bins provide conservative predictions in comparison with that obtained from the actual Byron-1 EOC-7B voltage distribution. Further increases in the voltage growth thresholds are not warranted for application with a POD of 0.6.
This result is consistent with that found for Braidwood-1 in the response to RAI question 1.
\\
N 1 1 4
Response to NRC Staff RAls on Braidwood 90 Day Report 3.
Requested Information Discuss why an independent assessment of the voltage dependent methodology is not needed. Provide the basis for Comed's conclusion that
- benchmarking" the methodology using the same data used to develop the methodology provides an adequate demonstration of the ability of the voltage dependent methodology to conservatively predict the EOC conditions at Braidwood, Unit 1 and Byron, Unit 1.
Response
The intent of the benchmarking analyses in the Braidwood-190 day report was to demonstrate the adequacy of the methodology for the most limiting SG C. SG C is the limiting SG with regard to leakage for the actual EOC-6 distribution (4 gpm larger leakage than SG D) and for the projected EOC-7 distribution.
Demonstration of the adequacy of the voltage dependent methodology for SG C 'vas expected to be conservative for the other Braidwood-1 and Byron-1 SGs.
To confirm this expectation and to respond to the RAI concern for additional benchmarking analyses, comparisons of SLB leak rates for projected and actual EOC distributions were made for all four Braidwood-l SGs and for the Byron-1 SGs B and C. The actual EOC-7B leak rates for Byron-1 SGs A and D were too small (< 0.1 gpm) to warrant consideration for methodology comparisons.
The results of the additional benchmarking analyses are given in Table 4.
Comparisons of SLB leak rates from predicted and actual distributions are made for EOC-6 at Braidwood-1 and EOC-7B at Byron-1. The projections for cach SG in Table 4 use the SG specific growth distribution rather than any use of composite of multiple SG growth data or bounding growth distributions. The SG specific values are used to better assess the methodology without additional conservatism from other bounding growth distributions.
The SLB leak rates for each Braidwood-1 SG are given in Table 4 as calculated from the actual and projected EOC-6 voltage distributions. The results for SG C were also discussed in the response to RAI Question 1 based on Table 2. Jor SGs A and D, the projections based upon the recommended procedure for developing voltage dependent growth distributions are 3 to 4 gpm higher than that obtained from the actual distributions even with a 10.3% NDE uncertainty applied for the actual distributions. This conservatism in addition to the 1.6 gpm margin found for SG C projections is partly due to the hybrid methodology of including the largest three growth values from all SGs in the growth distribution. For SGs C, A, D and B, respectively, the hybrid methodology adds 1,2,3 and 3 large growth indications to the SG specific distribution. The impact of the hybrid methodology is particularly conservative for SG B for which the leak rate is overestimated by greater than seven gpm for the projections.
These results tend to indicate that the hybrid methodology is conservative. This topic is further addressed in the response to RAI Question 8.
Overall, it is concluded that the recommended voltage dependent growth methodology is conservative with large margins for all four Braidwood-1 SGs.
tvunum ewo m u w on,am 9
m sw sm
e Response to NRC Staff RAls on Braidwood 90 Day Report 0
For Byron-1, the SLB leak rates based on the actual EOC-7B voltage distributions are 0.27 and 0.13 gpm for SGs B and C, respectively.
For the recommended voltage dependent growth rate bins, the corresponding predicted leak rates are 0.30 and 0.26 gpm which exceed that found from the actual distributions. This confirms the conservatism of the recommended methodology for the Byron SGs.
The results of Table 4 increase the number of SGs for benchmarking the voltage dependent growth methe<fology from one to six SGs.
The recommended methodology provides conservatively projected SLB leak rates for all six SGs evaluated and the methods are adequate for analyses of the Braidwood-1 and Byron-1 SGs.
BRAIDWOOD 1' 4.
Requested Information in the August 14, 1997 submittal, Table 6-2 and Figure 6-5 presented a compacison of predicted and actual bobbin voltage distributions for S/G C, EOC 6, using the voltage dependent Cycle-6 growth and a POD of 0.6.
Discuss how-well the witage dependent methodology works for each S/G and at all voltage levels (i.e., both above and below 3 volt).
A) Provide similar tables and figures for all four S/Gs for all voltage levels, B) Discuss how well the voltage-dependent growth rates conservatively predict the actual voltages for all S/Gs and for all voltage ranges, Discuss how well the voltage dependent growth rates conservatively predict EOC 6 MSLB leakage for each S/G.
Response
A) Predicted and actual EOC-6 voltage distributions for all four SGs are shown in Figures 12 through 15. These distributions correspond to the voltage dependent, EOC-6 projection results given in Table 4 for each SG.
The predicted EOC-6 voltage distributions shown in these figures were obtained applying voltage dependent growth distributions established using the guidelines presented in the Braidwood-1, Cycle-7 90-day report and a constant POD of 0.6.
For clarity, high voltage end of the distributions (over three volts) are presented separately. -The predicted peak voltages are compared with the actual peak voltages in Table 4 (which corresponds to Table 6-2 of 8/14/97 submittal). Further comparisons of the predicted and actual EOC-6 voltage distributions are included in the response to item (B) below.
B) Discussion of Actual EOC Voltage and Associated SLB Leak Rate Predictions Comparisons of tne predictions of the EOC-6 SLB leak rates with those calculated for the actual distributions are discussed in the response to RAI Question 3 given n= = m w m oc m on. =
10 msm no m
)
Response to NRC Staff RAls on Braidwood 90 Day Report above. The comparisons of the predicted and actual voltage distributions based on the figures noted in response item (A) above are discussed herein.
Fa. SG-A, the total number of EOC-6 indications is underpredicted because the indication population below 0.7 volt is underestimated. This is partly because the POD is less than 0.6 below about 0.5 volt. However, above 0.7 volt the predictions substantially exceed the actuals as evident in Figure 12.
For the indication populations exceeding 1 volt,2 volts and 3 volts, the predictions exceed the actuals by about 51%, 66% and 55%, respectively.
Therefore, the predicted voltage distribution is significantly more conservative than the actual distribution. This is reflected in the predicted EOC-6 leak rate for SG-A which exceeds that based on the actual distribution by about 50% (see Table 4).
An examination of Figure 13 clearly shows that the predicted EOC-6 distribution for SG-B is substantially more conservative that the actual distribution. The total EOC-6 predicted population for SG-B exceeds the actual total, and for the indication populations above 1 volt, above 2 volts and above 3 volts, the predictions exceed the actuals by about 94%,149% and 63%, respectively. The predicted EOC-6 leak rate is also very conservative as it exceeds that based on the actual distribution by about a factor of 20 (see Table 4). The primary reason for such large overestimation of EOC-6 voltages and leak ro+e is inclusion of three largest growth values from all SGs which are substantiah, uigher than the largest growth observed in SG-B. SG-B did not have any of the largest 25 growth values for Cycle-
- 6. This is further elaborated in the response to RAI Question 5.
Figure 14 provides a comparison of the predicted versus actual EOC-6 voltage distributions for SG-C, and again the predicted distribution is substantially more conservative that the actual distribution except for the largest indication found in the inspection. Total EOC-6 predicted population exceeds the actual total, and for the indication populations above 1 volt,2 volts and 3 volts, the predictions exceed the actuals by about 63%,68% and 67%, respectively. The predicted EOC-6 leak rate for SG-C exceeds that based on the actual voltages by about 14% to 34%
depending on the NDE uncertainty assumed (see Table 4).
As with SG-A, the total number ofindications at EOC-6 is underpredicted for SG-D because of underestimating the indication population below 0.7 volt, which provides a voltage distribution that is more conservative overall (see Figure 15).
For the indication populations exceeding 1 volt,2 volts and 3 volts, the predictions exceed the actuals by about 51%,73% and 54%, respectively. Also, the predicted SLB leak rate exceeds that based on the actual measured distribution by about 60 %
Overall, EOC-6 voltage distributions for all SGs predicted using a constant POD =0.6 and a voltage dependent growth distribution following the guidelines presented in the Braidwood-1, Cycle-7 90-day report are conservative in comparison to the actual measured voltage distribution.
Consequently, the predicted EOC-6 leak rates are also conservative.
L\\ROMANG\\HWD-lMX'\\RAl\\RaLO 1re dot 11 10/15/97,4 50 PM
Response to NRC Staff RAls on Braidwood 90 Day Report 5.
Requested Information Figure 5-3 of the August 14 submittal depicts the voltage growth during Cycle-6 as a function of beginning of cycle (BOC)- 6 voltage. Provide similar figures for the remaining S/Gs as well as a composite figure. Discuss the voltage-dependency of growth for each S/G based on the data in the figures.
Response
Figures 16 and 17 show the variation of voltage growth during Cycle-6 versus the BOC voltage for PGs A and, B, respectively, and similar data for SG-D are presented 1
in Figure 2. (Figure 5-3 of the 8/97 submittalis reproduced here as Figure 1 for SG C). Each of these plots also includes the top three growth values from all SGs during Cycle-6.
A composite plot containing similar data from all four SGs is presented as Figure 18. Cumulative probability distributions representing growth data for SGs A, B, C and D obtained by applying the growth binning guidelines presented in the Braidwood-1, Cycle-7 90-day report are shown in Figures 19,20,3 and 21, respectively.
SG-A had the third largest growth indication during Cycle-6 among all SGs. An examination of Figure 16 indicates that Cycle-6 growth for SG-A shows a dependency on Vnoc. For Vuoc over about 1 volt, the relative frequency of larger growths (above 1 volt) increases with Vuoc. Figure 17 shows Cycle-6 growth versus Vnoe data for SG-B.
It is evident that the largest growth observed in SG-B is substantially smaller than the three largest growths in all SGs. Without the largest indication with about 4 volts growth, SG-B growth data does not exhibit any dependency on Vnoe for growth above approximately 2.0 volts.
Therefore, it is concluded that the SG-B growth dependence on Vnoc is small and will not significantly affect EOC conditions. SG-D had the fourth largest growth during Cycle-6.
It is evident from the growth versus Vuoc data in Figure 2 that SG-D growth shows dependency on Vuoc.
In summary, Cycle-6 growth data for SGs A, C and D show a dependency on Vnoc, and SG-B shows a much smaller trend.
6.
Requested Information Discuss other characteristics that could possibly explain the apparent dependency of growth rates on initial bobbin coil voltage.
What other factors has Comed evaluated in an effort to understand the high growth behavior of these indications?
Include as a minimum in the response: 1.) location of the indications; and, 2) the age of the indications.
L\\ ROMANO \\ewMoc\\ pal \\pwot re. doc 12 io/is/97,4 so pu 1
e Response to NRC Staff RAls on Braidwood 90 Day Report
Response
As discussed in EPRI reports and GL 95-05, the Alternate Repair Criteria (ARC) was developed wiih the need for a continuing review of whether or not growth rates are dependent upon the BOC voltage. The need for this review was based on the independence of voltage found for domestic data while European data at higher voltages indicated growth dependence on BOC voltage. Prior to data obtained in late 1996 to 1997, these reviews supported the continued use of growth rates independent of BOC voltage for low voltage (less than about 1.5 volt) indications.
The general trend for a higher frequency of large voltage growth for larger volt BOC indications can be expected based on the dependence of voltage on crack depth.
Bobbin voltage increases exponentially with depth, particularly as near through wall indications are obtained. Thus, a small change in depth for a deep indication will have a much larger increase in voltage than from the same change in depth for a shallower indication. On the average, it is expected that growth in depth over an operating cycle is approximately independent of the initial depth. As the BOC voltages increase, the depths of the BOC indications can be expected to be larger and it should be expected that the frequency of larger voltage growth values would increase for the same increase in depth as for lower BOC voltages. The lower BOC voltages require a larger depth increase to obtain the same voltage growth as the higher voltage indications. Based on leakage data and the probability of leakage correlation for %" tubing, throughwall indications begin to occur at around 2 volts.
Thus, it would be expected that the change to voltage dependent growth would also occur around this voltage level as found in the Braidwood-1 data. It is important that voltage growth trends not be assumed to indicate a corresponding per entage change in depta Voltage analyses become increasingly conservativ. as the indication depth increases.
To respond to the RAI request, assessments of the Braidwood-1 voltage data were performed to assess location and age of indication effects. Figure 22 provides a plot of the locations oflarger growth rates on each of the lower support plates. In this figure, the different symbols correspond to the 'AV growth values for Cycle-6 normalized using time at RCS temperature > 500oF (413 days)'(DVCY6) over 1 volt intervals. The results show that the larger growth values are not dependent upon TSP tube location but occur principally at the lower TSP which has the higher temperatures.
The lower TSPs have consistently shown the largest number of indications and the associated frequency of larger growth values should also be expected.
Figure 23 shows the EOC-6 signal amplitude versus the age (time since first detected by bobbin inspection) of the indication in terms of operating cycles (Cycle-1 is EOC-5, etc.). More of the larger voltage indications were initially found two cycles ago at EOC-4. The ARC repair limit at Braidwood-l was increased from 1 to 3 volts at EOC-5. Thus, more of the EOC-4 indications with higher growth rates could be left in service at EOC-5 than for indications found at EOC-3 which were repaired at 1.0 volt at EOC-4. Therefore, the change in repair limits at EOC-5 is more influential on the trends of Figure 23 than the age of the indications.
1 L\\rnmAtu)\\HwMoc\\RAl\\htoir.*
13 io/is/o7.4 so eu l
Re:ponse to NRC Staff RAls on Braidwood 90 Day Report e
7.
Requested Information Provide a comparison between the projected EOC leakage obtained following the approved methodology using the Cycle-6 hybrid growth distribution and the proposed voltage dependent growth rate methodology discussed in the August 14, 1997 submittal. Discuss how the difference between the two values justifies the application of a voltage dependent growth rate methodology.
Response
it is assumed for this response that the requested approved methodology using the Cycle-6 " hybrid" growth distribution is a voltage 9 dependent growth that includes the three largest indications found in all SGs as ned for the voltage dependent growth, in this manner, identical data is used to compare voltage dependent and voltage independent analyses.
Analysis results using the approved methods of WCAP-14277 that do not include the three largest indications from all SGs are given in Table 4. The analyses for Braidwood-1 SG C and Byron-1 SG B result in underestimates of the leak rates obtained for the actual EOC voltage distributions.
The projection for Byron-1 SG C overestimates the leak rate from the actual distribution.
Analysis results for all four Braidwood-1 SGs using the Cycle-6 " hybrid" voltage independent and voltage dependent growth rates are given in Table 5 for the projected EOC-6 distributions. The results for the actual and voltage dependent growth distributions are repeated from the Table 4 data to facilitate comparions with the voltage independent results.
The voltage dependent growth rate projections for all four SGs result in the proj eted leak rates exceeding the leak rates obtained from the actual distributions.
With the voltage independent
" hybrid" growth, the projected leak rates exceed that from the actual distributions for SGs A and B but are less than that for the actual distributions for SGs C and D.
SG C is the limiting SG for leakage and thus the most important SG for assessing the adequacy of the projections.
It is clear from the results of Tables 4 and 5 that the combination of voltage dependent growth rates with the hybrid of the three largest indications from all SGs is conservative. The use of the hybrid includir.n the three largest indications with voltage independent growth rates provides an improvement over the NRC approved WCAP-14277 methodolcgy by adding the conservatism of the hybrid to compensate for the lack of voltage dependence. However, the voltage independent hybrid growth methodology and WCAP-14277 methodology underestimate the limiting SG C leak rate for Braidwood-1 at EOC-6.
wvwwwnwcwwowaa 14 l
non sm. um
Re ponse to NRC Staff RAls on Braidwood 90 Day Report G
8.
Requested h formation t
Discuss the basis for limiting the hybrid growth distribution to include the largest three growth values found in any of the S/Gs. Assess the effect of including the top 5 or 10 growth values on your projected EOC leakage.
Response
The SLB leak and burst analyses are strongly influenced by the largest indications in the distribution. There is no assurance that the largest indications will occur in the same SG between successive cycles. Therefore, the hybrid concept ofincluding the largest growth values in the growth distribution for individual SGs was applied to recognize the potential shift of the largest growth between SGs.
Assuming equivalent corrosion potential between SGs, the probability that the largest growth will occur in a specific SG is about % for a 4-loop plant. The probability that the largest two growth values will occur in the same SG is about 1/16. Inclusion of the two largest growth values in one SG thus establishes about a 95% probability that the largest growth values are included in the projection analyses. However, the methodology was intended to also cover 3-loop plants which would require the largest three indications to obtain about 95% probability of including the largest growth values. Therefort, the methods were defined to include the three largest growth indications in eacn SG growth distribution. For the four loop Braidwood-1 and Byron-1 units, this is added conservatism since there is only about a 2%
probability of the three largest growths occurring in the same SG.
As noted in the response to RAI Question 7, the inclusion of the three largest indications in the hybrid distribution appears to be conservative when combined with voltage dependent growth rates. The fact that the projections overestimate the leak rates obtained from the actual voltage distributions implies the methodology is conservative. Thus, it would be more appropriate to reduce the number of largest growth values in the hybrid distribution to one or two indications than to consider increasing the number in the hybrid distribution. Nevertheless, to be responsive to the RAI request, leak rate calculations were performed with the five and the ten largest growth indications included in the distribution. The probability of the five largest indications occurring in the same SG is less than about 0.1 percent.
Figures 24 and 25 show cumulative probability distributions for growth bins, which include the five and the ten largest growth values.
The leak rate analysis results for hybrid distributions including the three, five and ten largest indications are given in Table 6 for SG C. As expected, the use of the five or ten largest indications results in increased conservatism in the leak rate projections. The leak rates from the actual EOC-6 distribution are overestimated for the recommended hybrid distribution, and the overestimates become larger as the number of largest indications in the hybrid distribution is increased.
This 1
L\\ROMANG\\HWD40C\\RAl\\RaL0lre doc 15 msm.4 w m
Response to NRC Staff RAls on Braidwood 90 Day Report 3
T) e effect is smaller for SG C than would be obtained for the other SGs since SG C already includes three of the largest five growth values and six of the largest ten values.-
It is concluded that the hybrid - distribution using the three largest growth indications cttrrently applied results in sufficiently conservative leak rates. Larger numbers should not be considered due to the increased conservatism that results in the leak rate analyses.
L:\\WOMANG\\f3WthEOC\\RAl\\Ratoire dw 1h 10D SM. 4 50 W
0 Response o NRC Staff RAls on Braidwood 90 Day Report Table la Braidwood-1, SG C Number ofIndications in Growth Bin and BOC Distributions Growth Rate Growth Bin No. Ind.
No. Indications in BOC-6 No. Indications in BOC-7 Bin Option Voltage in Growth Distribution Distribution Range Bin Actual
> 1.1 210 191 320 551 955 Alternate I s 0.5 734 319 534 374 631 0.5 to 1.0 1059 843 1414 976 1641 1.0 to 1.4 266 240 401 377 635
>1.4 46 58 97 284 504 Alternate 2 s 0.5 734 319 534 374 631 0.5 to 1.0 1059 843 1414 976 1641 1.0 to 1.6 287 273 456 480 810
> 1.6 25 25 42 181 329
- Distribution Adjusted for Repaired Indications Table Ib Byron-1, SG B Number ofIndications in Growth Bins and BOC Distributions Growth Rate Growth Bin No. Ind.
No. Indications in BOC-No. Indications in BOC-8 Bin Option Voltage In Growth 7B Distribution Distribution Range Bin Actual
> 0.9 264 284 523 375 654 Alternate I s 0.5 825 483 824 615 1042 0.5 to 0.8 551 562 965 584 991 0.8 to 1.1 249 269 462 267 456
> 1.1 122 133 261 214 377 Alternate 2 s 0.5 825 483 824 615 1042 0.5 to 0.9 658 680 1166 690 1170 0.9 to 1.4 212 228 413 290 499
> 1.4 52 56 110 85 155
- Distribution Adjusted for Repaired Indications L\\ROMANG\\BWD ROC \\RAl\\Rai_Ol re doc 17 MH 5M. 4 M N l
Re:ponse to NRC Staff RAls on Braidwood 90 Day Report Table 2 Braidwood-l Summary of EOC-6 and EOC-7 Stil Leak Rate Projections SG - C llL Indications Only SLB Leak Rate Case No. of M ax.
(spm)*
Cycle 6, Growth Voltage Bin Widths No.
POD Ind.*
Volts
- Comments g;g 7
IRBs Span Actual EOC-6 "Best Estimate" for Comparison with Projection Methods Ref.
N. A.
I 2098 11.3 9.8 8.0 5% NDE unc.
Ref.
N. A.
I 2098 11.9 11.5 9.3 10.3% NDE unc.
Per GL 95-05 Projected EOC-6 for Comparison with Actual EOC-6 Dependence on Voltage Range for Voltage Dependent Growth I
WCAP-14277 Voltage Indep. Growth 0.6 2382 9.7 7.8 2
Recommended Growth Bins 0.6 2382 10.3 13.I 10.6 Fig.-3 growth Os.7, 0.7-1.1, > l. I 3
Alternate 1 Growth Bins 0.6 2382 10.8 14.2 11.8 Fig.-4 growth Os.5, 0.5-1,1-1.45, > l.45 4
Alternate 2 Growth Bins 0.6 2382 11.1 14.2 11.8 Fig.-5 growth Os.5, 0.5-1,1-1.6, > l.6 4A Same as Case 4 EPRI 1973 10.9 9.9 8.0 Projected EOC-7 Dependence on Voltage Range for Vol: age Dependent Growth 5
WCAP-14277 Voltage Indep. Growth 0.6 3411 12.8 27.2 24.2 6
Recommended Growth Bins 0.6 3411 14.5 57.1 52.2 Fig -3 growth 0s.7, 0,7-1.1, > l.1 7
Alternete 1 Growth Bins 0.6 3411 17.0 100.0 92.3 Fig.-4 growth Os.5, 0.5-1,1-1.45. > l.45 7a Same as Case 7 EPRI-2722 12.8 54.9 50.1 8
Alternate 2 Growth Bins 0.6 3411 17.1 98.0 90.4 Fig.-5 growth 0s.5, 0.5-1,1-1 6, > 1.6 Notes:
(1) Number ofindications adjusted for POD.
(2) Voltages include NDE uncertainties from Monte Carlo analyses and exceed measured voltages.
(3) Equivalent volumetric rate at room temperature.
L\\ROMANG\\0WD400\\RAl\\kal_Olre doe 1b 10/15/97.4 50 PM
o Resporme to NRC Staff RAls on Braidwood 90 Day Report Table 3 Byron-1 Summary of EOC-7B and EOC-8 SLB Leak Rate Projections SG - C llL Indications Only, Cycle 7B Growth Data SLB Leak Rate Case Cycle 7B. Growth Bin Width Voltage No. of Max.
(gpm)*
No.
Range POD Ind.m Volts" Comments With Free IRBs Span Actual EOC-78 "Best Estimate" for Comparison with Projection Methods Ref.
N. A.
I 2035 3.5 0.13 0.13 10.3% NDE unc.
Per GL 95-05 Projected EOC-7B for Comparison with Actual EOC-7B Dependence on Voltage Range for Voltage Dependent Growth I
WCAP-14277 Voltage Indep. Growth 0.6 2606 3.6 0.17 0.17 2
Recommended Growth Bins 0.6 2606 3.8 0.26 0.26 Fig.-8 growth.
O s 0.5,0.5-0.9, > 0.9 SG B growth.
2a Same as Case 2 with SG C growth 0.6 2606 3.9 0.26 0.26 Fig.-l 1 growth.
distribution SG C growth.
3 Alternate 1 Growth Bins 0.6 2606 4.0 0.29 0.29 Fig.-9 growth 0 s 0.5,0.5-0.8,0.8 1.1, > l.1 4
Alternate 2 Growth Bins 0.6 2606 4.2 0.31 0.31 Fig.-10 growth 0 $.5, 0.5-0.9, 0,9 1.4, > L4 Projected EOC-8 Dependence on Voltage Range for Voltage Dependent Growth 5
WCAP14277 Voltage Indep. Growth 0.6 3317 11.5 19.0 16.0 6
Recommended Growth Bins 0.6 3317 11.0 22.8 19.6 Fig.-8 growth 0 s 0.5,0.5-0.9, > 0.9
-7 Alternate l Growth Binc 0.6 3317 11.3 26.2 22.8 Fig.-9 growth 0 s 0.5,0.5-0.8,0.8 1.1, > 1.1 8
Ahernate 2 Growth Bins 0.6 3317 11.7 27.0 23.7 Fig.-10 growth 0 5.5, 0.5 0.9, 0.9-1.4, > 1.4 Notes:
(1) Number ofindications adjusted for POD.
(2) Voltages include NDE uncertainties from Monte Carlo analyses and exceed measured voltages.
(3) Equivalent volumetric rate at room temp rature.
L\\ROMANG\\HWD ILOC\\RAl\\RaL0lre doe 19 ion syn. 4-so eu
Response to NRC Staff RAls on Braidwood 90 Day Report Table 4 Braidwood Unit-1 EOC-6 and Byron-1 EOC-7B Comparison of SLB Leak Rate Results Based on Actual and Projected Voltage Distributions SLB No. of M ax.
Leak Rate Comments SG Growth Voltage Bin Widths POD Indications")
Volts")
(gpm)*
Braidwood-l EOC-6 ACTUALS
.A N. A.
1 1754 10.2 6.39 10.3% NDE unc.
B N. A.
1 801 5.8 0.36 10.3% NDE unc.
C N. A.
I 2098 11.9 11.5 10.3% NDE unc.
C N. A.
I 2098 11.3 9.8 5% NDE unc.
D N. A.
I 2099 9.7 6.99 10.3% NDE unc.
EOC-6 PROJECTIONS A
SG A data: O s 0.5, 0.5-1.0, > 1.0 0.6 1679 10.2 9.4 Fig.-19 growth B
SG B data: 0 s 0.5,0.5-0.8, > 0.8 0.6 916 10.0 7.5
- Fig.-20 growth C
SG C data: 0 s 0.7,0.7-1.1, > l.1 0.6 2382 10.3 13.1 '
Fig.-3 growth C
WCAP-14277 Voltage Indep. Growth 0.6 2387 9.7 7.8 D
SG D data: O s 0.5,0.5-1.0, >l.0 0.6 1769 10.2 11.2 Fig.:21 growth Byron-1 EOC-7B ACTUALS B
N. A.
I 1779 4.5 0.27 10.3% NDE unc.
C N. A.
I 2035 3.5 0.13 10.3% NDE unc.
EOC-7B PROJECTIONS B
SG B data: 0 s 0.5,0.5-0.9, > 0.9 0.6 2513 4.1 0.30 Fig.-8 growth B
WCAP-14277 Voltage Indep. Growth 0.6 2513 3.8 0.21 C
SG C data: 0 x 0.5, 0.5-0.9, > 0.9 0.6 2606 3,9 0.26 Fig -l 1 growth C
WCAP-14277 Voltage Indep. Growth 0.6 2606 3.6 0.17 Notes:
(1) Number ofintlications adjusted for POD, (2) Voltagesinclude NDE ancertainties from Monte Carlo analyses and exceed measured voltages.
(3) Equivalent volumetric rate at room temperature.
t.:\\ROMANOVr#W ANAL \\ht0lre doc 20 wmm.4 so eu
o Response to NRC Staff RAls on Braidwood 90 Day Report Table 5 Braidwood Unit-1 EOC-6 Projections Comparison of SLB Leak Rate Results for Voltage Dependent and Voltage Independent Growth SLB-No. of Max.
Leak Rate Comments SG Growth Voltage Bin Widths POD Indications
- Volts *
(gpm)*
Braidwood-l EOC-6 ACTUALS A
N. A.
I 1754 10.2 6.39 10.3% NDE unc.
B N. A.
I 801 5.8 0.36 10.3% NDE unc.
C N. A.
I 2098 11.9 11.5 10.3% NDE unc.
C N. A.
I 2098 11.3 9,8 5% NDE unc.
D N. A.
I 2099 9.7 6.99 10.3% NDE unc.
EOC - 6 PROJECTIONS A
0 s 0.5, 0.5 1.L, - 1.0 0.6 1679 10.2 9.4 Fig.-19 growth Voltage Independent flybrid Growth 0.6 1679 9.6 6.5 B
0 s 0.5,0.5 0.8, >0.8 0.6 916 10.0 7.5 Fig.-20 growth
,Vnuage Independent Itybrid Growth 0.6 916 9.6 6.1 C
O s. 7, v.7-1,1, >l.1 0.6 2382 10.3 13.1 Fig.-3 growth Voltage Independent Hybrid Growth 0.6 2382 9.7 8.7 D
0 s 05,0.5-1.0, >l.0 0.6 1769 10.2 11.2 Fig.-2I growth Voltage Independent Hybrid Growth 0.6 1769 9.5 6.7 otes:
(1) Number ofindications adjusted for POD (2) Voltages include NDE uncertainties froni Monte Carlo analyses and exceed measured voltages.
(3) Equivalent mlumetric rate at room temperature.
- L\\ROMANG\\DWD4CC\\RAl\\RaL0lre doc 21 IofIS/97,4 m) FM
Response to NRC Staff RAls on Braidwood 90 Day Report Table 6 Braidwood-l Comparison of EOC-6 and EOC-7 Projections for Varying flybrid Growth Distributions SG - C llL Indications Only SLB Leak Rate Case No. of Max.
(Epm)*
Cycle 6, Growth Voltage Bin Widths POD Ind!"
Volts
- Comments g;g IRBs Span Actual EOC-6 "Best Estimate" for Comparison with Projection Methods Ref.
N. A.
I 2098 11.3 9.8 8.0 5% NDE unc.
I Ref.
N. A.
I 2098 11.9
) 1.5 9.3 10.3% NDE unc.
Per GL 95-05 Projected EOC-6 for Comparison with Actual EOC-6 Dependence on Number of flybrid Growth Values for Voltage Dependent Growth
! l Recommended Hybrid Growth 0.6 2382 10.3 13.1 10.6 Fig.-3 growth Hybrid = 3 largest growth values Bins = Os.7,0.7-1.1, >l.1 2
Alternate i Hybrid Growth 0.6 2382 10.3 14.4 11.8 Fig.-2 ' growth Hybrid = 5 largest growth values")
Bins = 05.7,0.7-1.1, >l.1 3
Altc. ate 2 Hybrid Growth 0.6 2382 10.3 15.4 12.9 Fig. 25 growth Hybrid = 10 largest growth values Bins = Os.7,0.7-1.1, >l.1 Projected EOC-7 Dependence on Voltage Range for Voltage Dependent Growth 4
Ruommended Hybrid Growth 0.6 3411 14.5 57.1 52.2 Fig.-3 growth Hybi d = 3 largest growth values Bins = Os.7,0.7-1. I, >1.1 5
Alternate 1 Hybrid Growth 0.6 3411 14.7 62.5 57.1 Fig.-24 growth
. Hybrid = 5 largest growth values")
Bins = 0s.7, 0.7 1.1, > 1.1 6
Alternate 2 Hybrid Growth 0.6 3411 14.9 67.7 61.9 Fig.-25 growth Hybrid = 10 largest growth vahies Bins = 'K7, 0.7-1,1, > l.1 Notes:
(1) Number ofindications adjusted for POD.
(2) Voltages include NDE uncertainties from Monte Carlo analyses and exceed measured voltages.
(3) Equivalent volumetric rate at room temperature.
(4) Including 5* largest growth is the same u including 4* largest growth as the 5* to 7* largest growth values occur in SG C L: \\ ROMANG \\fiW D. EOC \\ RAl \\ RaL0 l re. doc 22 ionsm.4 m m
4 e
O O
O 0
a
___o Om $.
O es
+
"O P
UO o
O5 O
O O
C
-O N
no ^
o O
anE o
a 2 O
b O
sa O
$5g e
o E
o O
E
[g O
O O
co g
a:::
gggg
_ _7__
_ o-3s
__0_
o MSy a
fo o
e
'W.T m
I O
=
- B 5, F e
e&
o 9
8 o
a 9
U DC, O
$ Q.I O
O 3yb o
O O
g 3g3 O
O _.O_
_ O,g Qh O
5 O
,O c
oo
- 0g c
- 8 St OZ O
O Oe 00 E, E*
c6
- f n
+
.E q
'c c
.b
- c
}E o
O
.5 9
O E
3b O l G
83 o
ep O
2 8
3e O
E e-4
\\
m
=
3 m
o c
n v
O N,
O p
qiwoao 9 apta paz!puuoN
v-L Figure 2 Braidwood Unit -1 April 1997 Outage Voltage Growth During Cycle 6 vs BOC-6 Voltage - Hybrid of SG D + 3 Largest Growth Values Growth Normalized Using Time at RCS Temperature > 50(F (413 days) 8 3
i i
O SG-D (All Indications) l E
E 7
,f E 2 Largest in SG-C l
g A Largest in SG-A O
i O
O l
2
[
0 1
0 0
O l
]
I e
,g O
00 o
h O
O O
.]3 o
0O f
j O
t a
h i
O
$.2 0
00 O
O O_
O l
g c
g O
O i
-1 0
0.5 1
1.5 2
2.5 3
HOC-6 Voltage cescatfotiass72 4s eu
t
' Figure 3 Braidwood Unit -1 Cycle 6 Ilybrid Growth - Normalized Using Time at RCS Temperature > 500 F (All SG-C+ Largest SG A)-Cumulative Probability Distributions for Use with POD =0.6 1.0
- - ). x - x- -x -X ' " ' y 5dMC
=
- dM-dF5:
x.x-r-,,,.
i 4SA
,..g. x. x- - x- -x -x.
y 0.9 -
--g
,x..x.
,.8 I x..x..x x'
/
o 0.8 -
-Fi x
1g i
,i a
0.7 -
jo
,/.
f n
E
.'t
=
W 06-I Up to 0.7 volt (1230) e f
/
e k
x 2
II
- c- 0.7 to 1.1 volts (665) 3 0.5 -
t X
m5 e
a l
-- r --Over 1.1 volts (209) 0.4 _
x
=
l E=
- e-- SG-C Only (All volts)
U 0.3 -
x:
0.2 -
I.If 0.I -
- --.k'
.r, u.s. '
0.0
, m c.
-o c.
m.
, e e e-
= w - -
e, m, m o e-
= a c.
c.
m e ao - m, e a e -
e.
m m a e-9 e o g o o o o o o o o o ce es ce ce ce m m m e m m or
, m m e e e-Voltage Growth c,.c e.,s 2,an, s n m 1
t s
Figure 4 Braidwood Unit -1 Cycle 6 Hybrid Growth - Normalized Using Time at RCS Temp.> 500 F (All SG C+ Largest SG A)- Alternate 2 Cumulative Probability Distribution 9,,,g,7;g amsw 1.0
,,g_,g_gg3_;,p, _;,p, g ;,,;
T. ~~
.,_,,..x..x. x - -
,.A x-0.9 -
X' ' #
p--#
- x..r W.
.K-*,
3
,4
//
.r.Y e,.-- F
~~,,,,,,,,_..,-
gt
- e
- '
0.8 -
,f-X' m-ur' 1
G$
/
x s
c: 0.7 -
4 y
4 F
E-W V
u 9
E
- )
j0.6-r y
g Y
a
/
l 5
n 1
W Up to 0.5 volt (734) j0.5-5 i
e h,-y
/
-c- 0.5 to I volt (iO59)
E I
g 0.4 -
y 3
l.ul E
lF
-- * -- I to 1.45 volts (266)
=
.f U 0.3 -
J ll
- *-- Over 1.45 volts (46)
I m
0.2 -
4 y'
- <:- SG C Only (All Volts)
- q:
,/
0.l -
//
.K 0.0.l.t-f.f' ~
e.
3 m ao e-w m e,
- o - es m e m e e-oc e - -- ce m
,y o
r-ao e.
ca --
- c. m o
=-m
, e e
e m m m o e-o q q q o o o o o o o o o ce ce es en es m m m m m m Voltage Growth cwe
>=.7. n =
t e
Figure 5 Braidwood Unit -1 Cycle 6 Hybrid Growth -Normalized Using Time at Temp > 500 F (AllSG-C+ Largest SG A)- Alternate 2 Cumulative Probability Distribution 1.0 -
,ig,g _,qtg;,t, 3,g y ;,,
, g y,g,pf,p g i r-
.x-r_
/
y ".=
0.9 -
Y
'-I' q"
y -x. x - r '
W x
g'
//
f n
,,,a 0.8 -
-p-7
~-
g a"
g
,.x '
y.
= 0.7 --
t t
.E c
I o
i
" 0.6 -
c E
I o
o y
a I 0.5 -
Up to 0.5 volt (714) af
=
5 Y
ej0.4-
/
- c- 0.5 to I volt (1059)
K a
E a
-- x -- I to 1.6 volts (287) v 0.3 -
it 0.2 -
i
- *-- Over 1.6 < rolts (25)
,w y
45 4
- <:- SG C Only (All Volts) 0.1 -
.f IV m-My 0.0 W i" l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l
-s x
m x
-r ca
-r c
w ce ce c
-r ca o
en
-r c
x ci ci 5
5 5
-i d
i 6
6 6
6 6
6 Voltage Growth ew,a twe-,s x,~
7
L Figure 6 Byron Unit-1 October 1996 Outage Voltage Growth During Cycle 7B vs BOC-7B vohage - AII SG-B + Largest in SG-C Normalized Using Time at RCS Temperature > 50(F (102 days) 7 X
x SG4 (AilIndcations) 6 o SGC (Largest Growth) 5 5s 4
X I
X X
o X
3 X
X X
X Y E X '-
X X
bN X
2 x
x x
X M
g
^
i
~'
j X
-2 0
03 I
I5 2
2.5 3
HOC-7B Voltage w w.c44or,a763ru
Figure 7 Byron Unit-1 October 1996 Outage Voltage Growth During Cycle 7B vs BOC-7B voltage - All SG-C+ 2 Largest in SG-B Normalized Using Time at RCS Temperature > SOFF (102 days) 7 e-O l
o SG-C (A!! Indications) 6 e
O O
e 2 Largest in SG-B 5
l O
S O
4 C
O O
O E
0 3
e OO O
o O
O O OO O
b
-o oooog 2
O g
ano-o o
o de g
o O O O
O o
O
?
O O
G I
e
.O Swr O
0 m
Oog g 0800 O
O O
b*
OP g 8
0 O
O g-_ O o
_0o O
O O
-2 0
0.5 I
I.5 2
2.5 3
HOC-7B Voltage s e<ns.gunnme wru e
' ' ' ' ' ~ '
- - - - - - - - - - - - - - = - -
t
- 4 j
Figure 8 Byron Unit -1 Cycle 7B Hybrid Growth - Normalized Using Time at RCS Temperature > 500 F All SG-B + Largest in SG-C -Cumulative Probability Distributions for Use with POD =0.6 1.0
.g a w =.
_ - - m., s. - - o-er- - - --
w p - -o -o.o _. c - - a g4, -
D 0.9 -
-[
," y -
s m
g p -o i
.,g
.D 0.8 -
.c
-g F
g.jl'
.D d
i e 0.7 -
5 o
/:
P i
l' f
=
<a
=
/.
o
'* 0.6 -
,=
g
=
/
.E M
r 0.5 -
y W
5
/
$, OA -
i
[
--o-- Up to 0.5 voit (825)
~
d E
=
/
p u o.3 -
p
- c- 0.5 to 0.9 volt (658)
./
/p-g
- o-- Over 0.9 e shs (2M) 0.2 -
f 1
_t P.
/
-- m --SG-B (Growth Independent of BOC volts)
- 0. n.,
y p a
0.0 !,.
m m n
-o c.
n
,m o n =
a.
---e.
.,e
<n
,,e.
e.
m, e e n, e - m e n -,,
q o y o q o o = = o e o = c c.
~ n n ~ ~ n ~ n - - e n, m - e e Voltage Grouth w~,. m..
L
- 4 i
Figure 9 Byron Unit -1 Cycle 7B Hybrid Growth - Normalized Using Time at RCS Temperature > 500 F AII SG-B + Largest in SG-C - Variation 1 of Recommended Distributions for POD =0.6 1.0 3 5. G ";
- o. c a-o* o.o o -o -o -D O O O 4.qM
,a miE A U-0.9 -
P*
P W o-o d p n.m-a 7
d P, _ m.-
,er or 0.8 -
7 JI o'
f n'
e 0.7 -
- ar e
W A
/
.z a
p c
a
/
s 0.6 -
d P
b V
/
.o s
S 1
/
05 '
5 o
P i
x T
d g
=_,f f o:_
.N
-o-Up to 0.5 volt (825)
/
o
/
E O
d
- c- 0.5 to 0.8 volt (551) s
/
/
f U 0.3 -
D _f
- g
-c- 0.8101.1 volts (249)
/ /
0.2 -
0, l
p
-o-- Over 1.I volts (l22)
/ !
p J, d,0
-- m --SG-B (Growth Independent of BOC volts) i 0.1 -
f,g - /
-g
-G O.0. d:lr.*
m, e n - o
-n m,
-e n a >
--n n
,-e n a
, +.
.u.
n
, e e n,, - e e n -,. m n 9 o q q q o o o a o o o o o n
n n n n ~ n n m - n n, -
-c e
Voltage Grow th w.
.w..
L Figure 10 Byron Unit -1 Cycle 7B llybrid Growth - Normalized Using Time at Temperature > 500 F All SG-B + Largest in SG-C - Variation 2 of Recommnded Distributions for POD =0.6
- - r-e-e-s 1.0
.w, w a. -_ _ _ _ e -o-B-c-e -=
.c u-y aX a-
,,o -o -o -o X
ULO 0.9 -
y D'*
Y
.a
.o.o-y
.cr
.tr 1
p Y
.0
,fr 0.8 -
a P.o-a 1
9 1
c 0.7 -
n-d o
\\ 't p
o' t
5 c
g 6
6 s 0.6 -
/
p'-
a D.6
.o' c'
-o-- Up to 0.5 volt (825) 5 9
-/
3 0.5 -
p a
5 9
f
.D 0.5 to 0.9 volt (658)
$0.4-
-p'
/
f
}
I[N f
,d
-- 0 --0.9 to 1.4 voit (212) f
,D-G
- s
--p -- D U 0.3 -
g - Over 1.4 volts (52)
/
I i
g P
a2-
- p,0 o'
- *- SG-B Only (All Volts)
I
/
/
0.1 -
/
,o -
P
.D-D- 0*
0.0 1 S$$E$E$$--"212iEEI"2%%%%%%%%%C%?!372E 5 $ 5 2 e-o Voltage Grow th s
.... _-. m
L Figure 11 Byron Unit -1 Cycle 7B Hybrid Growth - Normalized Using Time at RCS Temperature > 500 F All SG-C + 2 Largest in SG-B - Cumulative Probability Distributions for Use with POD =0.6 10 g g a+ a -x-a--x--
p y _= 4 4 = - e --- -
y.z-4'
, m - - - -e *~
0.9 -
z r
5
/
Y
,5 m m-E,a 0.8 -
/
/
Ju' l
a a 0.7 -
-p' w
/
('
E JC
'A 0.6 -
N' Y.
8=
i i
i i
P Up to 0.5 volt (955) r 0.5 -
2 Y
Q
/
?
Y
- c- 0.5 to 0.9 volt (840)
= 0.4 -
n
/
k
/
/
8 o
s'
-f
- e-- Over 0.9 volts (230)
U 0.3 -
f
,41
.s r; F
/ '
O.2 -
- -x --SG-C Only ( All volts)
$i oI_
IF T
0.0 '
E E
E E
E E
2 1
E E
2 0
0 I.
O
~
Z E
Voltage Gron th ww,-,...
Figure 12 Braidwood Unit 1
- Steam Generator A Comparison of Predicted and Actual Bobbin Voltage Distributions for SGA, EOC. 6 Voltage Dependent Growth Rates Applied. POD =0.6 Staam Generator A. Up to 3 volts 250 0 Actual EOC-6 (Up to 3 volts) 200 m Predicted EOC-6 (Up to 3 volts)
Cycle 6 Growth
, 175
~
150
}125 5
ioo z
75 50 -
-~
25 a: : ::: ::: :
Bobbin Voltage Steam Generator A. Over 3 volts O Actual EOC 6 (Over 3 volts) 5-a Predicted, POD = 0.6 (Over 3 volts)
Cycle 6 Growth
- 4 e
.I 3-g 2 2-l
\\
.il llii..ilfi i
11 li..
Bobbin Voltage n.,
%i ie.,ii u u
I Figure 13 Braidwood Unit 1 -- Steam Generator B Comparison of Predicted and Actual Bobbin Voltage Distributions for SGB, EOC. 6 Voltage Dependent Growth Rates Applied. POD =0.6 Steam Generator B. Up to 3 volts 110 O Actual EOC4 (Up to 3 volts)
~
~~ ~
~
~
E Predicted EOC4 (Up to 3 volts)
-~~
g
_=_-_
Cycle 6 Growth 70 so h
50 -
40
---~~
z 3o 20
-r to
-~ - -
- i:'::a:::::::::
Bobbin Voltage Steam Generator B. Over G volts 2
O Actual EOC4 (Over 3 volts)
E Predicted, POD = 0.6 (Over 3 volts)
Cycle 6 Growth e
z lii....-__i I i....___.I liil i
- T.:::::::
- I
- I*3**3~.
I*I I;
- O:*2;2
- i Bobbin Voltage w wwsw n u a
g Figure 14 Braidwood Unit.1... Steam Generator C Comparison of Predicted and Actual Bobbin Voltage Distributions for SGC, EOC. 0 Voltage Dependent Growth Rates Applied. POD =0.0 Steam Generator C. Up to 3 volts 250 225 0 Actual EOC4 (Up to 3 vol;s) 200 5 Predicted EOC4 (Up to 3 volts)
Cycle 6 Growth 175
~
- 350 Ji 125 100 Z
75 50 -
25 EESo $EiE$ICUU
$IUIIN550
$$S$$$
Bobbin Voltage Steam Generator C Over 3 volts e
O Actual EOC4 (Over 3 volts) 5 E Predicted, POD = 0.6 (Over 3 volts)
Cycle 6 Growth
- 4 1
s 2 2
! l l o...__.Il li.....lill i.l!
l 1
- : : : : : : : : : : : : : : :::::: a: : : : :::::::::;
Bobbin Voltage
% eweenuw
e I
(
Figure 15 Braidwood Unit 1 - Steam Generator D Comparison of Predicted and Actual Bobbin Voltage Distributions for SGD, EOC. 6 Voltage Dependent Growth Rates Applied. POD =0.0 Steam Generator D. Up to 3 volts 200 200-0 Actual EOC 6 (Up to 3 volts) pso.
220 5 Predicted EOC 6 (Up to 3 volts)
Cycle 6 Growth 200 in 380 14o o
ipo ion eo 60 -
40 po 0
- I-
-I
'I 'I d'
" M ~ " "" *
-i
- a:::::::
Bobbin Voltage Steam Generator D. Over 3 volts O Actual EOC-6 (Over 3 volts) y 5 Predicted, POD = 0.6 (Over 3 volts) 3-Cycle 6 Growth 1
i a g
r IZ 1
i il lii....
,ill iiii I;
Bobbin Voltage
%,.mvn==
o e
b$
3 4 7$
e n
o k
^
0 %
o o
b qE o
h.e g oo g
~$f O**
2 o o
- I'
-.-s
.- 0
- o...
vi ;$
.n.
L* 2 5 o
O
.> a o
m gf*p hy O$
o n o 9
0 o
o l e *e oco 4>E o
o og
{m.D o
o FE 9
o o
O $
.E o
k 2e
.g a
o 3
3 0
I i$j 3
g o
o g
a o
i o
l o
e.
=
3 ymnao 9 al343 par llcuuoN
<l ill1
\\l1 1
i.
^
5 2
s G
Sr e
h t
O m
,ilI' l
!ii
- 1f i 4
i, rl 2
o) r s f y t
a s d O
eg3 r 1 o
a 4 L(
3 (F
+
0 O
BS O
eG>
g O
aS er ul u
,i',i
.*I1{
- +!!l1 i'
t l 5
O A ta 1
f r
o 7 o e o
9 p
9 d e
m g
i 1 rb e a
l T
t 7i y l
r o
pHS s.
1 eA - C O
o V
r e R o
6 C-u1 g n
o t
g -
a a O
i t t o
B Fil e
no UV m aW i
d6 T
)
1 o C-o g
s n
wOi G
s S
o dB U l
o ias l
r v d A
B 6 ze n
i
)
i t
s O
o e
l s
G g
l a
e y m g
S o
c r
la C r a
l o
L o
gN d
as n
r l
o h
3 ir
(
le t
u w A
w D o
)
r s
G sa 3
5 h
G n
S
(
0 n
C o_
t w
io t
i o
a r
c h
G s
t G
d w
S n
o n
I r
i eg G
t a
A se t
t
(
s o
B e
g l
r g
a M
V G
L r
P a
2 S
L 2
4 2
o s
+
79 O-
%0
~
t t
to 0
P_
8 7
5 4
2 G
t 1
o I
c S
5sEc*
0 5~ E k ty h
t6y C
l ll!ll!
.1
s Figure 18 Braidwood Unit-1 April 1997 Outage Voltage Growth During Cycle 6 vs BOC-6 Voltage - Composite of AII SG Data Growth Normalized Using Time at RCS Temperature > 50(fF (413 days) 8 D
D a SG-A 7
SG-B a
6 i
0 D SG-C D
i x
0 0
.c 5
x x
~
{
x SG-D x
x O
g n
D e 4 x
0 g
g f
x 0
x O
D x
a k
00 x
A O
D e
D x
x j 3 x
x 3
x a U
a x;
- 9 ta a
Z.2 v 9 Uh o*---
- o a'
x3 a3 m
0 x
a x.
00 ^t 0 a 3
a a
.x a
o Da#
3
_a_8 O c_A Da
^x b
i D
k, A
40 O
- b
^
O g
n u
. m uuu
.x o
l
-1 0
0.5 I
I.5 2
2.5 3
BOC-6 Voltage ensgscasem_rwaw a 22 eu
t Figure 19 Braidwood Unit -1 Cycle 6 Hybrid Growth - Normalized Using Time at RCS Temperature > 500 F All SG-A + 2 Largest in SG-C - Cumulative Pn>bability Distributions for Use with POD =0.6 1.0
,g_4 p.g.;.;._;_;__;_;-.;--; -
,-g,g,..=
m-E
- x.. x.. x. x - -x -*
pf
,..x. r x -*.
[.A
..x ~,
0.9 -
.x
.x-l 0.8 -
- I x*
.r.
/I K'
n 0.7 -
-I c'
e t'
.x x
!i
.x
=
~
Up to 0.5 volt (840)
La.
06~
~
r' g
f s
P, e
.C
.de
-c-0.5 to I volt (721) g 05-
.j Ei II o
- I f
--* --Over i volt (21I) 11 OA -
- t C
4 e
E j
y'
-* - SG-A Only (All volts) uu 02 -
0.2 -
-1 0.1 -
1
~
0.0 !! -F
,., ~ _ o - -.,,
m
. -. e _ _ n
.-.,n o q q o o o o e o e o o o
~ n ~ ~ ~ ~ ~
n
- e e n Voltage Growth Ceespe#W F*trW9' 2 43 8W w
Figure 20 Braidwood Unit-1 Cycle 6 Hvbrid Growth - Normalized Using Tiene at RCS Temperature > 500 F All SG-B+ 3 Largest in All SGs - Cummiative Probability Distributions for Use with POD =0.6 1.2
) _h 4 44 ~Y
[.[2.[.~..x.~--[r-5-5-~"
s.y T-x -,....x - - -x g./
_ x.
- a...
/
5 r
. x*
DE -
/
x Up to 0.5 volt (338) re.
/
e f
.c 5
.=
x
-C-0.5 to 0.3 volt (284) 3 0.6 -
//
b j!
-- m --Over 0.8 volt (189) i o
- 3.,
Ei E
_._. SG-B Only ( All volts) 0.4 -
3 O.2 -
ll
..x/
..x/
0.0,
o
~
a n
=
e
~
e
~
n E
=
=
M e
b a-E C
d C
O C
D d
C d
h
(
Vohage Growth C N M% m W$44N 7
O O
Figure 21 Braidwood Unit-1 Cycle 6 Ilybrid Growth - Normalized Using Time at RCS Temperature > 500 F All SG-D + 3 Largest in All SGs-Cumulative Probability Distributions for Use with POD =0.6 1.2 A*-ve-* w,~-_,==-_*
P,a,..e + + + 4 4fr m y v., y, 4 "-
x-x-x x-x 2'*****'*'**
~- -
f.0 -
l N:
x x. x x-x x-x- x,
y w x, x. _x. -
7 0.8 -
- -g
.x,x. x-g 1
=
0 w
x-
=
U to 0 5 velt (953)
P
=
.9 5
e t
-c-n 5 to I volt (940) 3 0.6 -
[
x I
.. * Overivoitg2ggy c
y
?
=
SG-D Only (AII vorts) 5 0.4 -
t --.-
~ * - -
u
?
_i O
O e
s 0
0.2 -
e
/
0.0 " -
- n - C --
~n
-e s.
=, -
-n
.-, - e e,
-<n
-n
.- -,, - e,,, n,
Q Q k C
C C
C C
Q C C
C
- - *= -*
e-
- A M
M M
M
- 't M
- 't
- t s-e
- W 9
e g
b Voltage Growth cewn wenwn m
...s_
L Figure 22 Braidwood Unit-1 Location of Large Growths During Cycle 6 at TSP Elevations Cosnbined Data frosa All Stease Generators g Q.,'
a::s
=
_l L
E ii i,
- J.
....i a
a u=a
- t
__.....p........I.......,
4.
._. g.....-si,,,-
Note: Elevation is mapped
=!
ihi.
to support plates as:
=
sn:.
- i!!
1 = 03H
=
=
/-
.; =
._.,_..+....._.~U~
2 - 05H
~
I
+.
.J
= * -
i.b... :i 3 = 07H
!!:T]..
4 = 08H
~
E'
'3 5=09H a
a es a
~
/
ei.
=
sti a' O.
tt;.ir O
siu o
33 g.
o; o
- I
%.T i
/ u....._... U..._. I........ U.. _ 9 i U....,.. ~..
s;
~iii o
o c;::
- o
- !===
ye
!si.
c
- t:. :
o a
W o
3!;.
pa;.....,...,g... K... m........_..,p..._a&....a i...*o._._
siisi
. o :
.n DVCY6>=2 and DVCY6<3
_g 4.._..iN
=
DVCY6>=3 AND DVCY6<4
- 7.._ _.....
DVCY6>=5 AND DVCY6<6
.. _.. L..._. _...;...
.. J
.._.i _ _.._ l...
~d
=
+
DVCY6>=6-
~
=
2 Plate outline
If
\\
l l
o
]6V l
5 3
e.,
3 e
g A
n o
i tac s 5_
idr 2
o nt I a n
f r o
oe it n n c
e e
o t
iG e
D tcm e
1 n
c a
s.
2 n t
L iu e nFt is 3 Ua S s
e 2
l.
l c
edsl r o aA y
os C
u wem g
s go i
n d
r o
F iaf it al 5
t a
a rot c
1 BVa d
i n
6D I
f Cd o
e e
On g
A Ei f b om m
no
+
e 1
oC i
t u
b r
i t
s i
D 5
0
<O 0
2 0
8 6
4 2
o 1
1 Q
r a E " _ g e. : o @
l(l l
..._t L
Figure 24 Braidwood Unit-1 Cycle 6 IIybrid Growth - No malized Using Time at RCS Temperature > 500 F Ilybrid SG-C Data - Largest 5 Growth Values from All SGs Included
,o
,_f.
- y -6.-8~.S:8 2'B=8 *'*- W * * " ~ ~ ~ ~ ~
.. ~"
x-O.o-D Jr-y -----"....x-m*
w 0.9 -
u" g
X I
g D'0
.. - - -x...
/
_K /
F ~..x
,,. r og g
i
!/
~
sa y
c 07 -
.-G x
=
E ff 3
06-
~#
~
~ ~ ' - ~
Up to 0.7 volt (1230)
,i
.x. 2 e
.o 1
-j 16 g
i
-o-0.7 to I.I volts (665)
E 03 -
j1 e3 1
X 1
e t
---x --Over 1.1 volts (211) f f,
o.,3 i
3
- .1
=
l
- *-- SG-C Only (All volts) c U
0.3 -
1 0.2 -
1 0.1 -
r
.g -
u 0.0 :P-
, n n - c - ~.," - e
- - -,- n.,.--e e
,e n
=,--n
.., e a n - e - - n a.
e. - e,
-e c -
y q q q o o o o o o o o =
~ n n ~ ~ n n -
Voltage Growth w s
-n.,~
Y L
o, 4
Figure 25 Cycle 6 H brid Growth - Normalized Using Time at RCS Tempec w e > M.'?
Braidwood Unit -1 3
Hybrid SG-C Data - Largest 10 Growth Values from AII SGs Included
'*F #
- ~
-hem 4MM" 1.0 t
,.-,e-N
5 ------ --
-/'g
^^'
0.9 -
3 --
~~**x.x*
l o
D-I
/
"***"'"".....x-l 0.8 -
-M c' ff
,x *..x....-
t p
(
l no 1
l c
0~
.f
.K*x 5,
f I
j 04-
,g' Up to 0.7 voit (1230)
E
,8 to y
^5 0.5 -
j!
-c- 0.7 to 1.1 volts (665) f m
I T
a
'r
-- x --Over 1.1 volts (213)-
d 5
0.4 -
t
-h 2
5
~r.
- e-- SG{ Only (All volts)
U 03 -
..x 0.2 -
.1
- )
1 0.1 -
-f f
-Q *
^
^ ^
^
^
^ -
u 0.0 :P-
^
, m n - o
-n
,m e. ~ -, - -
e,
,-e n a e n - n m a a
-,e a, - n
-e
. n,
q q q q c e e o e o e a c
~ ~ ~ ~ ~....- - - - n,, - - - e n Voltare Grow th wem
- w = =
_