ML20116H146

Cycle Length Assessment Rept Addendum

ML20116H146
Person / Time
Site:	Braidwood
Issue date:	08/01/1996
From:	COMMONWEALTH EDISON CO.
To:
Shared Package
ML20116H139	List: ML20116H133 ML20116H146
References
NUDOCS 9608090189
Download: ML20116H146 (59)

Text

. . . . - . . . . - - - - - - _ - . - . . - -

l, v l i

i I

i

'l t

l Braidwood Unit I  !

l  !

! Cycle Length  !

l l

l Assessment Report Addendum f i i

i 1

L 6 i  :

i I

l l I

l August 1,1996 l 1

i l

t I

i l

,e

! 9608090189 960802 i PDR ADOCK 05000456 i

G PDR

F o

v' TABLE OF CONTENTS i

I ListOfTables.................................................................................................................i l Li s t O f Fig u re s . . . . .. .... . . . . .. . . . . . . . . . . .. . .. . .. .. ... .. . . . . .... . .. . . . . .. . . . . . .. . . . .. .... . . . .. . .. . .. . . . . .. . . ... . ... . . . . . ... .. 11 BRAIDWOOD UNIT 1 CYCLE LENGTH ASSESSMENT REPORT ADDENDUM Exe c utive S u m m a ry ............................................................................................ 1 r

! 1.0 I n t r o d u c t l o n . .. . . ..... .. . .. . . . . .. . . . .. . ...... . . ... . .. . .. . .... . . .. . . . . . ..... . .. ... .. . . . .. .. . . . . ... .. . .... . ... .. . . . ... . 3 i 2.0 Review of B yron B ou nds B raidwood ................... ............................................. 4 I

l 3.0 B raidwo od Inspection Eval uation ...................................................................... 5 I

l l 4.0 A v e rag e V o l ta ge . . ... . ... ... . ... ..... .. . . .... . .. . .. ... ...... .... . . . . . .. . . . . . . ...... ... .. .. .. .. .. .. . .. .. . . . . . .. ... . 7 1

5.0 Structural Integrity Evaluation for Braidwood ...............................................11 r

6.0 industry Tube Pull /insitu Pressure Test Voltage Analysis .............................15 l 7.0 Le a k R at e A naly si s ........................................................................................... 17 l

l 8.0 B raidwood inspe etion im provement s .............................................................. 18 t

9.0 C o n c l u si o n s . . . . . . .. . . .. ... . . . . . .. . .. . . ... . .. . . .. . . . . .. . . . ... .. .... .. ... . . . .... .... ... . . . . .. . . . .. . ... . . . .... . . . . . . 1 8 l

l 1 0. 0 R e f e r e n ce s . . . . . . ... . .. . . . . . . . . .. .. . . . . . . .. . . ..... . . . . . . . . . . ... ... . . . .. . .. . . . .. .. . ....... . . .. . . ....... . .. . . . . . .. . . . 1 9 j Appendix A End of Cycle Distribution Calculation inputs ..................................... 20 A.1 Braidwood Unit 1 Site Specific POD Calculation ............................................. 20 l

A.2 N D E U n ce rt ai nt y ............................................... ........................................... ...... 21 l 4

i

/

List Of Tables ,

Table Number Table Description 3.1 Byron and Braidwood Units 1 Largest Indication Voltages for Average and i Maximum Voltage '

5.1 Inspection Results for Braidwood Unit 1 February and October 1995 - 0.080" RPC l All Steam Generators l 6.1 Industry Tube Pull and insitu Pressure Testing A.1 Plus Point Probe Wear Assessment Affect on 0.080" RPC Voltage 1996 Data 100% i TW EDM Axial Notch A.2 Plus Point Probe Wear Assessment Affect on 0.080 Coil Voltage 1996 Data 100% TW Hole 5.1 Results of Structural Margin Assessment 1 i L

I i

b l

l I

1

/

• List Of Figures Floure Number Flaure Description Figure 3.1 Byron 1996 SG C Circumferential Indications Detected by + Point only, + Point and 0.080" RPC Figure 3.2a R3C11 SmallIndications ANSER Voltage Contour Figure 3.2b R3011 SmallIndications EddyNet95 Voltage Figure 3.3a R15C99 Medium Indications ANSER Voltage Figure 3.3b R15C99 Medium Indications EddyNet95 Voltage Figure 3.4a R21C54 Large Indications ANSER Voltage Figure 3.4b R21C54 Large Indications EddyNet95 Voltage Figure 4.1a - 4.1d EddyNet95 Voltage Integral Software Displays

Figure 4.2a Braidwood Unit 12/95 and 10/95 Voltage Re-i analysis,0.080" RPC Average Voltage Figure 4.2b Braidwood Unit 12/95 and 10/95 Voltage Re-j analysis,0.080" RPC Maximum Voltage Figure 4.3 Byron Voltage Change /EFPY vs. Relative Frequency, Average Voltage

Figure 4.4 Byron Voltage Change /EFPY vs. Relative Frequency, Vertical Maximum Voltage Figure 4.5 Byron Unit 11996 S/G C Indications

Average vs. i Maximum Voltage Figure 4.6 Braidwood Unit 11995 Indications: Average vs.

Maximum Voltage Figure 5.1 Number of Degraded Tubes vs. Vert. Maximum Voltage Figure 5.2 Number of Degraded Tubes vs. Average Voltage Figure 5.3 Braidwood: Number of Tubes vs. Average Voltage Distribution Figure 5.4 Braidwood: Number of Tubes vs. Vertical Maximum Voltage Distribution Figure 5.5 Average Voltage Structural Limit Figure 5.6 Maximum Voltage Structural Limit Figure 5.7 Braidwood Unit 1 EOC Maximum Voltage Distribution 461 Days Vs. 296 Days Figure 5.8 Braidwood Unit 1 EOC Average Voltage Distribution 461 Days Vs. 296 Days Figure 7.1 Industry Tube Pull and Insitu Pressure Test Leak Rate vs. Maximum Voltage Figure A.1 Distribution of 100 Tube Blind Test Mean Voltage Figure A.2 Distribution of 200 Tube Blind Test Mean Voltage Figure A.3 Distribution of 100 Tube Blind Test Analyst Deviation From Mean Figure A.4 Distribution of 200 Tube Blind Test Analyst Deviation From Mean Figure 6.1 Industry Tube Pull Insitu Pressure Test Average Vs. Max Volts

I I

i Executive Summary Comed has perfonned funher assessment of the Braidwood Unit I steam generator tube degradation to detennine the appnpriate operating cycle length. The objective is to document a technical basis for Braidw(xx11 full cycle operation for a nominal 461 days above 500'F. This represents 165 days beyond Comed's current commitment to the NRC of an October 15,1996 mid-cycle steam genemtor inspection outage. Comed has detennined the requimments to demonstrate steam generator tube integrity, following the methodology of the upcoming steam generator mie, after full cycle operation to be:

1. The structit.d requirements of draft Regulatory Guide 1.121 must be satisfied,

2. ReactF molant leakage must limit site boundary dose to a small fraction of 10CFR100 limi' should a main steam line break (MSLB) event occur.

The assessment documented in this repon uses new information resulting from work being l performed for the Comed /EPRI Circumferential Indication Repair Criteria Program. The new l infonnation includes,

1. Re-analysis using recently developed voltage integral software which detemiines an average indication voltage,

2. End of cycle (EOC) voltage distribution predictions following the methodology of NRC Generic Letter, GL 95-05 (Voltage Based Repair Criteria for Westinghouse Steam Generator Tubes Affected by Outside Diameter Stress Corrosion Cracking),

l 3. Review of industry tube pull /insitu pn:ssure test indication voltages for development of an upper voltage bound structural limit for average and maximum, and

4. Development of a Braidw(xx1 site specific probability of detection (POD) to address Byron Unit I inspection improvements.

Based on this new infonnation Comed believes a tecimical basis is provided for Braidwootl Unit I to operate full cycle (nominal 461 days above 5(XYF). The conclusions of the report are pmvided below.

The structural margin assessment demonstrates adequate margin will be maintained for the degradation distribution for 461 days of operation above 500"F at Braidwood Unit 1.

The leak rate analysis submitted in the previous Braidwood Unit i Cycle Length Assessment remains bounding, and demonstrates margin to site allowable leak ge limits for the combined degradation mechanisms of top of the tube sheet circumferential cracks, F* (no leakage) and tube suppon plate ODSCC.

The additional 165 days of operation greater than 50(fF does not reduce margin to structural or leakage integrity requirements.

Operation of Braidwocal Unit I for a full cycle of 461 days of operation above 50(fF does not challenge steam generator tube structural or leakage requirements.

l l

1

/  :

A i

l

, l Resolution of fouritems required for extension of the Braidw(xxl Unit I mid-cycle steam generator i tube inspection to October 15,1996 are addressed throughout this report. These items are  ;

summarized below: i l 1. Insitu pressure / leak testing during a mideycle inspection. Comed has analyzed Byron and Industry tube pull and insitu pressure test data to establish a relationship to NDE parameters in order to assess the integrity of Braidw(xxl Unit I tubes after operation of 461 days above 5(XTF. This report demonstrates that operation of an additional 165 days above 500"F does not significantly reduce steam generator tube integrity.

2. Pull tubes concurrent with 3.0 volt IPC during upcoming refuel outage. Comed will pull tubes with circumferential indications during the upcoming refuel outage in onler to assess circumferential indication morphology. Results presented in this report regarding the relationship between eddy current parameters indicates that the circumferential indication morphologies at Byron Unit I and Braidwo(x! Unit I are similar.

3. Plug circumferential indications upon detection. Comed will continue to plug all circumferential indications detected.

4. Upcoming Bmidwood Unit I steam generator tube inspection will be perfomied using

)

l inspection and analysis techniques equivalent to those used at Byron Unit I in 1996. 1 l An assessment of procedurrs and guidelines to assure an equivalent inspection is in l pacess. l l

l l

l l

l l

I i

l 2

l l

1

,' j l.0 Introduction ne objective of this report is to document a technical basis for Braidwood I full cycle operation for 461 days above 500'F. This represents 165 days of operation beyond the currently approved ,

October 15,1996 mid-cycle inspection outage. De requirements which must be satisfied to l demonstrate that steam generator tube integrity is maintained after full cycle operation are:

l. De structural requirements of draft Regulatory Guide 1.121 must be satisfied. l

2. Reactor Coolant leakage must limit site boundary dose to a small fraction of l 10CFR 100 limits should a MSLB event occur.

The Braidwood Unit I full cycle operation assessment uses a combination of probabilistic and deterministic approaches. Suf6cient data is available to predict a probabilistic end of cycle distribution; detenninistic lower bounds are used to predict burst pressure and leakage. Additional supporting information for the deterministic approach is the information summarized in Section 2.0 that Byron Unit I level of degradation bounds BrMdwood Unit 1. l Dis report provides new infonnation which is the result of work being perfonned for the Comed /EPRI Circumferential Indication Repair Criteria Program. The new infonnation to be presented in this repon includes:

1. Re-analysis using recently developed voltage integral software which detemlines an average indication voltage,

2. End of cycle (EOC) voltage distribution predictions following the methodology of GL 1 95-05 (Voltage Based Repair Criteria for Westinghouse Steam Generator Tubes Affected by Outside Diameter Stress Corrosion Cracking),

3. Review of industry tube pull /insitu pressure test indication voltages for development of an upper voltage bound stmetural limit for average and maximum voltage, and

4. Development of a Braidwood site speci0c POD to address Byron Unit 1 inspection improvements.

He Byron Unit I look-back infonnation previously presented in the Braidw(xxl Unit i Cycle i Length Assessment (Reference 1) and an addendum to this report (Reference 2) is used to predict a realistic EOC voltage distribution of circumferential indications. The predicted EOC distribution is used to assess the steam generator tubes compliance to Regulatory Guide 1.121 and to site allowable leakage limits.

The structural margin assessment demonstrates adequate margin will be maintained for the degradation distribution for 461 days of operation above 500'F at Braidwood Unit 1.

He leak rate analysis submitted in the previous Braidwo(x! Unit 1 Cycle Length Assessment remains bounding, and demonstrates margin to site allowable leakage limits for the combined degradation mechanisms of top of the tube sheet circumferential cracks, F* (no leakage) and tube support plate ODSCC.

He additional 165 days of operation greater than 50(TF does not reduce margin to structural or leakage integrity requirements.

3

/

Operation of Braidw(xxl Unit I for a full cycle of 461 days of operation above 5(X)"F does not challenge steam generator tube stnictural and leakage requirements.

2.0 Review ofIlyron Bounds liraidwood Infonnation presented in References I and 2 develops the basis that Byron txnmds Braidwood.

Byron has demonstrated that there is substantial margin to safety limits after operation for a full cycle of 448.5 days above 500"F. The technical basis for this period of operation has been accepted by NRC. A summary of these two reports is pmvided below.

• Based upon look-back of 1996 0.080" Rotating Pancake Coil (RPC) data. there was slow growth of Byron Unit i 1996 circumferential indications; after one full cycle of operation the maximum voltage of the distribution grew from 0.55 volts in 1994 to 1.1I volts in 1996.

Based upon the number and size of indications, Byron degradation rates bound Braidwood.

• Byron Unit I tube pulls and insitu pressure testing demonstrated that structural integrity of the Byron steam generator (SG) tubes is not threatened. The size of the Byron Unit 1 indications bounds those detected at Braidwood Unit I and therefore structural integrity at Braidwood is not threatened.

• The increase in the number ofindications at Byron Unit 1 in 1995 and 1996 are due to an inspection transient resulting from changes in inspection and analysis techniques and not from accelerated growth ofindications. The number ofindications at Braidwood Unit I has been adjusted by a POD to account for similar inspection transients.

• The number of Braidwood Unit 1 indications identified during the last Braidw(xxl Unit I refueling outage, A1ROS,(23) are far fewer than identiGed at Byron Unit 1 in their 1994 refueling outage (132), Byron Unit I mid-cycle (2578) inspection and the Byron Unit 1 1996 refueling outage (3478). Therefore Braidwood degradation is bounded by Byron.

• Inspections at Braidwood have detected and repaired circumferential indications well before they become structurally significant; the size of the largest indication detected decreased from 2.53 maximum volts to 1.36 maximum volts fmm the Spring to Fall 1995 outages.

• 1995 Byron tube pull indications had margin to the structural limit after minimally 342 days of operation above 5007. In addition, the 1994 tube pull results support full cycle operation above 500"F. subsequently Comed has demonstrated that structural integrity is maintained with significant margin after 448.5 days or longer with temperatures greater than 500"F.

• There was increased industry and Comed awareness of circumferential indications during the Braidw(xx! A1R05 inspection as a result of the Byron 1994 inspection where circumferential indications were first detected.

• The maximum estimated leak rate for Braidwo(xl is 15.8 gpm during a MSLB (including leakage from TSP ODSCC and F*), using conservative assumptions, which is less than the l

4 i

J

1

= 1

' j l

26.8 gpm site allowable limit (from Reference 1), based upon the new infonnation, tids analysis remains bounding. j l

• The chemical environment widdn the Braidwood Unit 1 SG's has not influenced the top of tube sheet initiation or the propagation of these indications (from Reference 1). l

• Byron and Braidwood have had no measurable primary-to-secondary leakage attributed to circumferential cracks.

• Compared to industry circumferential indication experience, the number of Braidw(xxl I ,

circumferential indications is similar with other plants (from Reference 1). I 3.0 Braidwood Inspection Evaluation in order to form a technical basis for Braidwood Unit I full cycle operation, it must be demonstrated that the inspection perfomled during the fifth refueling outage (A 1R05) was of a high quality to ensure the stmetural and leakage integrity of the steam generator tubes is maintained. l The A1ROS steam generator inspection was the last inspection prior to the current cycle of operation. The A1RO5 inspection used the three coil RPC probe (circumferential sensitive, axial sensitive and 0.080" RPC) with the ANSER analysis software. This was the second inspection using these techniques and was preceded by the 1994 Byron tube pull confinning the presence of circumferential cracks. The tube pull data resulted in a higher level of circumferential crack awareness at Braidwo(x1.

1 An assessment of a predicted begimdng of cycle (BOC) POD for the Braidwood Unit 1 A1RO5 outage, for use in detemiination of an EOC Braidw(xxl Unit I voltage distribution, is presented in l Appendix A.

3.1 Size ofIndications Detected Data obtained from the Byron Unit 1 1996 k>ok-back analysis for SG C was used to assess the sensitivity of the 0.080" RPC to detect degradation levels before they challenge structural integrity (Defined in Section 5.4). Tlds is determined by comparing the number of tubes where degradation was identified by both the + point and 0.080" RPC with the number of tubes where degradation was l identified by the + point coil alone. The results from this comparison an shown in Figure 3.1 where the number of tubes identified by tuth + point and the 0.080" RPC, and the number of tubes j identified by + point alone are shown as a function of the + point vert max voltage for each tube. l The data represents tubes with indications detected and repaired in 1996. The broken line in Figure 3.1 represents indications which would have been detected if the 0.080" RPC was used alone. The solid line represents those indications which would not have been detected if the 0.080" RPC was used alone.

Figure 3.1 demonstrates that: (1) degradation was detected by both the + point and 0.080" RPC for the largest degradation levels where the + point voltages are greater than approximately 0.98 volts, (2) most flaws (>80% +ivyint detected flaws) detected by + point also were detected by the 0.080" RPC for + point voltages from 0.8 :o I volt, and (3) approximately half the degraded tubes that 5

i /

were detected by + point were detected by the 0.080" RPC for voltage in the range from 0.610 0.8 volts. The + point coil identified substantially more tubes as being degraded than did the 0.080" RPC for voltages less than 0.5 volts.

l Industry experience over the past 6 years also supports that indications which may challenge structural integrity are found and removed from service the first time steam generator tubes are inspected with a coil sensitive to circumferential indications. Because growth rates are low (from Byron look-back results), large indications do not occur in one cycle of operation. This experience l is supported by the inspection results at Byron and Braidwood Units 1.

Table 3.1 presents infonnation on the largest average and largest maximum voltages detected at Bymn and Braidwood Units I for the inspections completed to date. This data demonstrates that indications which could challenge the structural / leakage integrity of the tubes have been detected and repaired and the size of indications being detected is smaller in each successive inspection.

3.1.1 Conclusions

1. Growth rates for OD initiated circumferential cracks at the roll transition are low.

2. Circumferential indications in tubes have been detected and repaired using 0.080" RPC well before they challenge structural / leakage integrity.

3. Based upon Byron Unit I look-back results, all + point vertical maximum voltages greater than 1 volt can be detected with 0.080" RPC.

34 Braidw(xxl Unit I look-back results demonstrates for the largest 10/95 voltage circumferential indication a signal is present in the 2/95 inspection data.

3.2 Inspection Comparison to Byron In addition to the probe improvements discussed above, the inspections perfonned at Braidw(xxl Unit I used ANSER analysis software. Analysts using ANSER analysis software have been used tiuoughout the industry and have been successful in detecting circumferential indications before they challenge the structural limit, following full cycle operation. An evaluation of the detectability of indications using ANSER compared to EddyNet95 was perfonned. The comparison to EddyNet95 is being made because it may have contributed to an inspection transient at Bymn Unit i 1996. Figures 3.2,3.3 and 3.4 show the eddy current testing (ECT) analysis graphics for small, medium and large indications as detected by ANSER and EddyNet95. The graphics show that the indications are detectable using ANSER but not as pronounced as with the filter capabilities of EddyNet95. The maximum voltages of the indications are consistent between ANSER and EddyNet95.

As stated above, it was concluded in Reference 2 that the large number of indications detected at Byron in 1996 were the result of an inspection transient. The inspection transient was partly due to the use of the + point probe, filter capabilities of EddyNet95 which was used in 1996 (ANSER was used in 1995) and partly due to improved analyst sensitivity to circumferential indications, gained through experience from Byron Unit i 1995 tube pulls. Appendix A provides furthcr evaluation of the differences in inspection techniques I etween Braidw(xxl Unit I and Byron Unit I and the resulting site specific Braidwo(xl Unit i POD.

6

i

• 1

/ ,

l

)

l l

3.2.1 Conclusion l Improvements in analysis tools and analyst sensitivity have improved detection of circumferential indications. Smallindications are detectable with the use of ANSER analysis software. Analysts using ANSER analysis software have been used throughout the industry and have been successful in detecting circumferential indications befom they challenge the structund limit, following full '

cycle operation. Appropriately trained analysts using 0.080" RPC and the specific ECT procedures applied in the h)ok-back have and will succeed in detecting all circumferential indications befom they challenge structural integrity.

l 4.0 Average Voltage 4.1 Hasis for Average Voltage Eddy current maximum voltage is one of the primary non-destructive examination NDE parameters used to assess the condition of steam generator tubes. This parameter has proven to be an effective indicator of degradation in steam generator tubes for a wide range of degradation mechanisms and l morphologies. For circumferential cracking, maximum voltage can be an effective indicator for i structural and leakage integrity assessments, especially when the crack is asymmetric and a j segment of the crack is either through wall or substantially deeper than the remairkler of the crack front. This morphology is prevalent in most OD circumferentially cracked tubes. However, service experience indicates there can be a small percentage of the tubes where the circumferential I crack front is essentially symmetric. In these instances, maximum voltage may underestimate the level of degradation, by measuring the eddy current response over a portion of the tube l

cin umference, and does not characterize the degradation over the entim tube circumference at one time. Average voltage provides a measure of the average degradation over the entire tube circumference, and can be used as a compliment to maximum voltage to ensure symmetrical crack fronts are properly assessed.

Comed evaluated an average voltage relationship and its possibilities for establishing a comparison criteria for voltage versus burst and pressure tests (Section 6.0). An objective of this evaluation included conelating data with test results from consecutive eddy current inspections as well as industry tube pulls and insitu pressure test tubes.

4.1.1 Conclusions Average voltage can be used with maximum voltage in predicting EOC distributions for use in assessing steam generator tube integrity to ensure that all cmck morphologies are assessed.

4.2 Average Voltage Look-Hack and Re-analysis  ;

I 4.2.1 Introduction Comed performed a kuk-back and re-analysis evaluation of the Bymn and Braidwood Unit I circumferential tube sheet indications detected during previous cddy current inspections. An eddy 7

i

% l l

i current evaluation was also perf(mned on industry insitu pressure test and tube pulls data from different utilities. All of the eddy current evaluations were perfonned using a Voltage integral (voltage averaging) software developed by Zetec, Inc. The maximum voltage from the integral voltage software were also calculated as part of the evaluation. The objectives of the look back are to:

i l

1) detennine the maximum and average voltage for the indications at Braidwotxl and Byron Unit 1, l 1

2) assess the average and maximum voltage growth, and j

3) compare the average voltage results to structural data.

4.2.2 Scope The look-back analysis consisted of analyzing the 0.080" RPC and + Point data for indications detected during the 1995 (SG B) and 1996 (SG C) eddy current inspections. One hundred three (103) indications identified during the 1994 eddy current inspection were re-analyzed with the 0.080" RPC probe. Re-analysis of thirty-nine (39) Braidwood Spring and Fall 1995 0.080" RPC j indications was performed for all SG's. Fifty-Six (56) industry tubes that were either pulled,  :

pressure or burst tested were re-evaluated. Further discussion of the industry tubes is presented in Section 6.0 1 4.2.3 Voltage Integral Discussion i

A voltage integral value is the area between the plot line and the voltage threshold line (see Figure 4.la) and is calculated with respect to one complete circumferential scan line that produces an l average voltage. These measurements are made in conjunction with the Eddynet95 RPC program and are discussed in more detail below.

I l

4.2.4 Voltage Integral Features

a. Plotting a single circumferential plot line in the voltage integral window requires an initial active data plot in the RPC C-scan window of Eddynet95 Figure 4.la. This plot line has been nulled or" zeroed" with respect to the lowest data point in the line by the software. This plot line is selected for the largest venical component of the indication.

b. A Volts-at-cursor plot is calculated as the difference between the vertical voltage threshold line and the data value where the cursor intersects the circumferential plot line.

I

c. When a voltage integral plot is perfonned by selecting a Plot Cursor Scan, the program j calculates and displays the voltage integral value. The voltage integral value is the area between the plot line and the voltage threshold line. This value is calculated with j respect to one complete circumferential scan. Examples: A straight line elevated to a l

! constant voltage value of 1.0 volts over the entire circumference of the tube would yield l a voltage integral value of 1.0 volts. A straight line elevated to a constant voltage value of 1.0 volts over one-half of the entire circumference of the tube, then dropping to 0.0 l volts for the remaining half of the tube would yield a voltage integral value of 0.5 8

. i

/

volts. This would effectively have half of the area between the plot line and the direshold line as would the first example plot.

l l

4.2.5 Voltage Integral Measurements Voltage integral measurements are derived from a plot defined as a single circumferential scan plot.

For each adjacent data point pair in the scan, the arithmetic mean (average) of the venical components for the two points is computed. This value is divided by the total number of data points between the stan and end locations of the scan. All of these values are summed. This is the data point average sum. The voltage integral value is the data point average sum multiplied by the voltage scale. This value is the same as the area undemeath the circumferential scan line.

4.2.6 Voltage Integral Data Analysis Guidelines The look-back and re-analysis evaluation for the Byron Unit I circumferential indications reported during the 1995 and 1996 eddy current inspections were perfonned by Rockridge Technologies Inc., at Benecia California fmm June 17 - June 25,1996. The Braidwood Unit I and Industry tube evaluations were perfonned by Comed's Level 111 ECT analysts. Data analysis guidelines were developed and administered prior to the look-back evaluation. The course included eight hours of classroom instmetion, with an additional eight hours of hands-on training with the voltage integral software.

1 The purpose of the guidelines is to provide general instmetions and to define specific requirements for the look-back analysis. The hx)k-back analysis guidelines were developed to provide a i structure to ensure the data is analyzed (a)in accordance with the appropriate techniques and practices that reflect current industry experience, (b) in a consistent and repeatable manner, and (c) '

in compliance with Comed requirements.

To provide voltage consistency, the voltage was nomialized on the 1007c axial notch to 20 volts peak-to-peak for the 0.080" pancake coil in the main Eddynet95 Lissajous window as shown in Figure 4.1b. After setting die 0.080" pancake coil to 20 volts, the set volt units in the circumferential lissajous RPC window was set to the Circumferential to Main Eddy in the Circumferential Voltage Scale Menu popup.

After *he nonnalization was perfonned, the data was analyzed in the RPC display. A C-Scan display was plotted for the expansion region of interest as shown in Figun 4.lc. Using the Axial and Circumferential Strip Chan Cursors, isolate the C-Scan line associated with the maximtim venical channel displacement as shown in Figure 4.lc. Clicking the Plot Cursor Scan button in the RPC Voltage Integral display window replicates the circumferential line scan generated in the main RPC DISPLAY. The cursor is then positioned to the maximum venical displacement as shown in Figure 4.10 The Volts-at-Cmsor will change as the cursor is moved along the trace while the Voltage Integral value remains constant. The circumferential line-scan baseline trace detennines the zero reference point for voltage measurements.

9

I 4

l t

l For consistency, the data analysts were instmeted to suppress a circumferential scan line above the i tube sheet which was free of any possible indications. The chosen line is subtracted from every circumferential scan line in the C-scan plot. It acts as a nulling function for all of the l circumferential scans and removes unwanted signals that occur consistently in each scan line.

Reporting the voltage values was accomplished by using the Report Entry display. Pressing the RPT ENTRY button activates the Repon Entry Display with the Voltage Integral value automatically entered in the ARC field. Also included in the report was the Volts-At-Cursor i " maximum voltage" using a vertical maximum measurement from the circumferential lissajous window. The analyst would type the three letter code " MAX"in the percent field of the report entry display.

4.3 Average Voltage Look-Ilack Results Figures 4.2a and 4.2b show the average and maximum voltage distributions for the Braidw(xxl Spring 1995 mid-cycle and Fall 1995 refueling outage. These distributions are increased by a POD of 0.2 to address inspection arxl analysis improvements made since the inspection was perfonned. This POD is discussed in more detail in Appendix A. The Braidwood Unit i 1 distribution is discussed in more detail in Section 5.0.

Figures 4.3 and 4.4 show the average and maximum voltage growth for the Bymn Unit I indications. Braidwood Unit I growth rates were not calculated due to difficuhy converting the 2N5 data for the 10/95 indications, efforts will continue to convert and analyze the data to assess j growth rates for the 10/95 circumferential indications. The Bymn Unit I growth rates were used in l the Braidwo(x! Unit I cycle length assessment discussed in more detail in Section 5.0. It is important to note that the 1996 Byron Unit I population of SG C 0.080" RPC circumferential indications is much smaller than 1995 (see Figure 3 of Reference 2).

A comparison of the average voltage to the maximum voltage is presented in Figure 4.5 for Byron I Unit i 1996 SG C 0.080" RPC indications and Figure 4.6 for all Braidwo(x! Unit 12/95 and 10/95 0.080" RPC indications. 'The figures show a good relationship between the average voltage and the maximum voltage detennined during the average voltage kok-back. This mlationship is expected because of the crack morphology of the Bymn tube pulls which show many shon non- I coplanar cracks in a band typically around the circumference of the tube and not a single deep crack. A tube with deep cracks, corresponding to a large maximum voltage, would,in this case, also have a large average voltage. As discussed above, the average voltage hiok-back software selected the null point for calculation of the maximum voltage. This feature provides a more consistent analysis and reduces the analyst variability. The fact that the slopes of the average voltage versus maximum voltage for Byron Unit I and Braidw(xxl Unit I are similar leads to a conclusion that the degradation morphology is similar.

As discussed in Reference 2, the tube with the most limiting circumferential indication detected at Bymn or Braidwood was identified as being mw 23 column 44 from steam generator at Byron.

This was based upon maximum voltage and tube pull metallographic crack sizing msults. This tube was pulled in 1994. During re-evaluation ofindustry tube pull data using the average voltage software and nonnalized to the same procedure, this tube has a maximum voltage of 4.14 volts compared to the 5.5 volts using previous techniques. This does not change the conclusions that this is the limiting tube indication detected at Byron or Braidwood and the other indication voltages detected are significantly smaller than this indication. The next largest 0.080" RPC indication 10

I

< j s 1 maximum voltage is 2.53 volts at Braidwood and 2.23 volts at Byron, txith detected the first time a l 1(X)% top of tube sheet (TTS) RPC inspection was perfortred. i 5.0 Structural Integrity Evaluation For Braidwood Bie following summarizes the evaluation procedure used to demonstrate that Braidw(xxl can l operate for one full cycle (461 days above 500"F). l i

• Demonstrate that the 0.080" RPC used during the previous inspection for Braidw(xx! Unit 1

)

has a POD that ensures structurally significant defects were detected and removed from service in a manner equivalent to the inspection technology used at Byron. j

• Detennine a POD that reflects the inspection technology used at Braidwo(xl relative to that i used at Byron.

1 1

• Determine a BOC distribution for Braidw(xx1 diat is equivalent to that which would have been 1 found if the Braidwood SG's were inspected using technology similar to that used at Byron. l

• Predict the degradation distribution at Braidwood at the end of one complete cycle of l operation (461 days of operation above 5(XTF) using: (1) die Braidw(xxl BOC and POD, I which have been adjusted to simulate an inspection technology equivalent to the Byron inspection, (2) the degradation growth rates from k>ok-back to 1994,1995, and 1996 SG B l and SG C inspection results from Byron Unit 1,(3) analyst uncertainty obtained from the blind I l

test of Byron tubes, and (4) probe wear from a study perfonned of the Byron 1996 Unit i ECT data.

1

• Demonstrate that Industry and Byron experience bounds the predicted EOC degradation l distribution at Braidwotx! and there is an adequate level of safety for the EOC degradation distribution after one complete cycle of operation.

5.1 Detection Capability and POD The results from the 1996 and 1995 inspections at Byron were used to assess the inspection capability of the inspection techniques used at Braidw(xx1 and to define a POD for the 0.080" RPC relative to die + point coil.

An overall POD was defined for the inspection technology used at Braidw(xxl in October of 1995 relative to the inspection technology used at Byron during the 1996 inspection. The POD was defined from consideration of the differences in the coils (+ point vs. 0.080" RPC) and differences in the analyst sensitivity and software used to evaluate the data (EddyNet95 vs. ANSER). The calculation of an overall Braidw(xxl Unit 1 POD, based upon Byron Unit 1 inspection data,is presented in Appendix A. The result is that a conservative POD of 0.20 is applicable to the Braidw(xxl Unit 1 Fall 1995 inspection.

11

I 5.1.1 Conclusion  !

A conservative POD of 0.20 is applicable to the Braidwo(xl Unit 1 Fall 1995 SG tube ECT l inspection and addresses the Bynm Unit 1 inspection improvements not utilized at Braidw(xx! Unit  !

I during the last inspection. L i

5.2 Braidwood Unit 1 Degradation Distribution and Comparis(m With Byron Inspection {

Results Table 5.1 summarizes the February and October 1995 inspection results for Braidwood Unit 1.

As indicated by the table, very few tubes with circumferential indications were identified during the inspections (16 tubes in February 1995 and 23 tubes in October 1995). The voltage data in Table r 5.1 was obtained by re-analysis of the circumferential indications using the new Zetec average voltage software. A 100% hot leg TTS RPC inspection of each steam generator at Braidw(xxl Unit I was performed to detect the indications. i Figures 5.1 and 5.2 present a comparison of the frequency distribution for the Braidwood Unit 1 inspection results adjusted for the POD and the inspection results obtained at Byrun Unit i for steam generator B in October of 1995 for vertical maximum and average voltage, n spectively. j The infonnation in Figures 5.1 and 5.2 demonstrate that the initial inspection at Braidwood Unit 1 in February of 1995 with the 0.080" RPC detected the largest and most structurally significant ,

flaws. These flaws were removed from service. 'Ihe figures also show that the degradation levels  !

at Braidw(xx] Unit 1, adjusted for POD, are within the degradation levels at Byron Unit 1.

i Because the degradation at Braidwood Unit 1 is bounded by the degradation at Byron Unit 1 in size i and number of indications the Byron Unit I growth rates can be used for application in predicting a Braidwood Unit 1 EOC voitage distribution.

5.2.1 Conclusion l The Braidwood Unit 1 Febmary 1995 inspection techniques detected the largest and most structurally significant flaws and removed them from service. Braidwood Unit 1 indication distribution is bounded by Byron Unit 1.

5.3 End of Cycle Distribution For Braidwood Unit 1 The procedure used to predict the EOC voltage distribution for Braidwood after one full cycle of operation (461 days atxwe 500"F) followed the metixxiology in NRC GL 95-05 and implemented in Byron and Braidw(xxl IPC submittals (WCAP-14277), or l

N i = Nm / POD - Ns . (1) where N, = the number of flaws at the beginning of the up coming cycle for the idi degmdation level,

! Nm = the number of flaws detected by NDE at the ith degradation level, and l Nn = the number of flaws repaired at the ith degradation level.

12 l

I Implementing repair on detection,i.e., Nm = Nn , Eq. I becomes l

N, = Na (1/ POD - 1). (2) l A POD = 0.20, as described in Appernlix A, and the Nm vertical maximum and average voltage distributions presented in Table 5.1 for the Braidw(xx! October 1995 inspection were used in Eq. 2 to obtain the BOC distributions.

The BOC voltage distribution represented by Eq. 2 was adjusted for probe wear, analyst uncertainty, and degradation growth rate to detemiine the EOC distribution following the methodology in GL 95-05. The probe wear and analyst uncenainty were developed as nonnal distributions, the standard devhtions for probe wear and analyst uncenainty were 0.06 and 0.20, respectively. No cut off of the distribution tails were used for either of the distributions. The bases for the probe wear and analyst uncertainty are presented in Appendix A.

The degradation growth rates were obtained from look-back to 1994,1995 and 1996 Byron inspection results using the meduxtology in GL 95-05. The combined growth rates from the inspections are presented in Figures 4.3 and 4.4 for average voltage and, venical maximum voltage respectively. Consistent with GL 95-05, the negative gmwth rates seen in the figures wem set equal to zero for development of the EOC distribution.

The EOC voltage distributions for vertical maximum and average voltage were obtained using a Monte Carlo sampling pmcedure. The results from the computations are presented in Figures 5.3 and 5.4 for average voltage and, venical maximum voltage respectively, where the EOC distribution is presented along with the adjusted number of flaws detected by the 0.080" RPC as described in Section 5.2.

l 5.3.1 Conclusion Conservative EOC voltage distributions are predicted using the metinlology of GL 95-05 for use in structural (Section 5.4) and leakage assessments (Section 7.0).

I 5.4 Structural Margins An assessment was made to determine if there was adequate stmetural margin for the predicted EOC voltage distribution. To accomplish this evaluation, relationships between burst pressure and vertical maximum and average voltages were detennined fmm available Byron (including two EDM simulations) and industry burst tests for pulled tubes and insitu tested tutus. Tirse relationships are presented in Figures 5.5 and 5.6 for average and venical maximum voltages, respectively. These relationships weie developed using lowest tolerance level (LTL) material strength propenies (Braidwood Unit 1,95/95 lower bound propenies at 650"F divided by industry room temperature sample mean (Reference 3)), to adjust burst /insitu pressures, and a curve that bounds all of the LTL burst points. Further discussion on tids assessment is presented in Section 6.0.

l l

13

1 1

l l

The stmetural limit for Braidw(xx! was based on the margins specified in Regulatory Guide 1.121, i.e., the higher of 1.43 times postulated faulted load and 3 times nonnal operating pressure differential. For Braidwo(x1, the limiting condition is 3 times the nonnal operating pmssure l differential which is 4,035 psi.

I l Tubes with voltages less than or equal to the structural limit at EOC have adequate margin of l safety against burst. Tubes with voltages greater than the structural limit at EOC must be

! evaluated to detennine if there is adequate margin against burst.

l l 1he basis for this evaluation is that the conditional probability of burst is acceptably low for the

! sum of all tubes with voltages beyond the structurallimit at EOC. Because a stati;tically based l burst correlation cannot be developed with the available data, the sum of the voltage frequencies beyond the structural limit is used to assess the margin. The acceptance criterion is that the sum of l all frequencies for voltages beyond the structural limit is less than 2x10-2 per steam generator.

The value of 2x10-2 has been defined to limit the number of tubes with voltages beyond the stmctural limit, and to be consistent with the criterion that the conditional probability of burst for d

the sum of the tubes with voltages beyond the structural limit is less than approximately lo The sum of the fmquencies equal to 2x10-2 was selected to maintain frequencies no greater than 10-2 to )

4 10 near the structurallimit where the burst probability would be less than 102, and frequencies no 4

greater than 10 to 10" for higher voltages where the burst probability would be greater than 10-2 l In addition, this criterion limits the number of tubes with voltages greater than the stmetural limit 1 to 2% or less of the total number of tubes in the EOC distribution. It also ensures that the large majority of tubes have voltages significantly less than the structural limit, and consequently, have low burst probabilities.

l Ninety Nine (99) percent of the Braidwood Unit i EOC indications have voltages less than the structural limit. The sum of the frequencies for voltages beyond the structural limit for all four 4 d steam generators at EOC for Braidwood is 6.6x10 and 1.1x10 for average and maximum voltages, respectively. Consequently, it is concluded that the EOC distribution for Braidw(xxl has acceptable structural integrity.

14

5.4.1 Conclusions The voltage structural limit for the limiting pressure of 4,035 psi is shown on Figures 5.5 and 5.6.

The results from the margin assessment are sunmarized in Table 5.2.

Table 5.2. Results of Structural Margin Assessment Venical Maximuu Average Voltage Parameter Voltage Structural Limit Voltage 3.64 volts 0.91 volts Limit for Sum of Voltage Frequencies > 2.0x10 2 2.0x 10-2 Stmetural Limit (per Steam Generator)

Sum of Voltage Frequencies > Structural- 4 1.1x10" 6.6x10 Limit (461 Days) 4 Sum of Voltage Frequencies > Structural 1.09x10" 3.7x10 Limit (296 Days)

The results from the structural margin assessment presented in Table 5.2 demonstrate that the stmetural criteria were met for both the venical maximum and average voltages, and adequate margin will be maintained for the degradation distribution through the end of 461 days of operation above 5007 at Braidwood.

Based upon distribution ofindications detected at Braidwood and Byron Unit 1, pesented in Reference 2, this distribution represents significant conservatism in the tail of the EOC distribution.

The change in the EOC voltage distribution from the currently approved October 15,1996 mid-cycle to the refueling outage scheduled 165 days later is presented in Figures 5.7 and 5.8 for maximum and average voltage, respectively. The results in the Figures and Table 5.2 indicate that there is no significant reduction in margin from an October 15,1996 mid-cycle inspection to a March 29,1997 refueling outage. i 6.0 Industry Tube Pull /Insitu Pressure Test Voltage Analysis 6.1 Scope of Review 1 Voltage analysis ofindustry outside diameter top of the tube sheet circumferential indications, which were pulled or insitu pressure tested, was performed. The analysis was perfonned to i detennine an EOC voltage limit for tubes with circumferential indications, to demonstrate compliance to Regulatory Guide 1.121 structurallimit of three times NOP differential, or4035 psi.

A breakdown of available industry tube pulls and insitu pressure testing ECT data is provided in Table 6.1. EddyNet95 was used for the analysis.

15

4.

6.2 Correction Factors in order for the analysis to provide consistent results between plants thme correction factors were anticipated. The first correction factor addresses the voltage nonnalization. The second correction factor was for tube wall thickness. Tubes with wall thickness of 0.(43 and 0.(M8 inch are included in the industry data-base. The third correction factor addmsses the difference between data acquired with the 0.080" RPC Vs. the 0.115" RPC.

6.2.1 Voltage Norinalization Correction Factor Voltage nonnalization was addressed by setting the voltage m 10 volts on the 100% thruugh wall ASME drilled hole which was available on all calibration standards with the exception of Byron Unit I tube pull R23C44. Because the Braidwood Unit I and Byron Unit I look-back and m-analysis were nonnalized by setting the voltage to 20 volts on a 100% EDM axial notch, a correction was required. The correction factor was detennined by the following procedure:

1. Set volts to 20 volts on a 100% axial EDM notch

2. Read voltage of the 100% through wall hole = X volts

3. X volts /10 volts = voltage nonnalization cormction factor Check data:

4. Set volts to 10 volts on a 100% through wall hole

5. Read voltage of the 100% axial EDM notch = Y

6. 20 volts /Y = voltage nonnalization correction factor The voltage nomialization correction factor was calculated using calibmtion standards used for I analysis of Byron, Braidwo(xl. ANO and Calvert Cliffs data. The voltage nonnalization correction factor used to compare the industry tube pull analysis data to Braidw(xxl and Byron Unit I look-back data is 0.58. Byron tube R23C44 was analyzed with a nonnalization of 20 volts on the 100% through wall EDM axial notch with no cormction required.

6.2.2 Tube Wall Thickness Correction Factor in order to assess the voltage affects of tube wall thickness, the voltage of the 100% axial EDM notch and 100'/c drilled through wall hole were measured for tubes of different wall thickness, after consistent voltage normalization. The voltages from the different tube wall thicknesses were similar (Byron (0.043"), Braidwood (0.(M3"). ANO (0.048"), and Calven Cliffs (0.(48").

Therefore no voltage correction for tube wall thickness was required. All tubes analyzed were 3/4 inch and themfore no correction for tube diameter is required.

6.2.3 Rotating Pancake Coil Correction Factor l Based upon Byron Unit i 1996 kx>k-back results, on average, the O. I 15" RPC voltage is 24%

gmater than the 0.080" RPC voltage. Therefore, the 0.115" RPC voltages are corrected by a factor of 0.76.

16

i4 l

l i

6.3 Conclusions l

'Ihe average ami maximum voltage is plotted in Figures 5.5 and 5.6 against pressure for tubes l which have been burst or insitu pressure tested. Die voltages have been adjusted for nonnalization l to 20 volts on the 100% through wall EDM axial notch (same nonnalization of Byron and Bmidwood hx)k-back re-analysis) and for differences in coil (0.115" vs. 0.080" RPC). No corrections were required for wall thickness or tube diameter (all tubes 0.75 inch). Test pressures have been adjusted using LTL material stength properties (Braidw(xxl Unit 195/95 lower bound properties at 650"F divided by industry nxwn temperature sample mean (Reference 4)). From this data, cuives are drawn to establish an upper bound voltage limit for the Regulatory Guide 1.121 stmetural limit of three times nonnal operating differential pressure (4035 psi) and for the main steam line break limit (2560 psi). These limits are identified in Table 5.2. The data points for insitu pressure testing represent the lower limit for the test since the tubes were not taken to burst.

The actual burst pressure of the tubes is bounded by this point.

1 As discussed in Section 4.3 a relationship between average and maximum voltage (Figure 6.1),

similar to Bymn Unit I and Braidw(xxl Unit 1, exists. This implies a similar degradation morphology between industry circumferential cracks.

7.0 Leak Rate Analysis Figure 7.1 shows the leak rate measured from industry tube pull leak testing and insitu pressure testing versus the indications maximum voltage. The leak rates have been nonnalized to operating temperature and the MSLB pressure of 2560 psi. The voltages were obtained as discussed in Section 6.0. Maximum voltage provides a g(xxl measurement of the crack's depth especially when the crack is asymmetric and a segment of the crack is either through wall or substantially deeper than the remainder of the crack front. Average voltage provides a measure of the integrated degradation over the entire tube circumference and there may not be segments where the crack is either through wall or nearly through wall.

The maximum voltages for which leakage has been detected thmugh tube pull leak testing and insitu pressure testing (Figure 7.1) is greater than 1.1 volts with a maximum leakage of all tests of 0.162 gpm. The number of tubes from the Braidw(xxl Unit i EOC distribution above this threshold is 25. The leak rate assessment provided in Reference I remains conservative and demonstrates that the maximum estimated leak rate for TSP ODSCC,'ITS degradation, and F*

mechanisms of 15.8 gpm during a MSLB, using conservative assumptions,is less than the 26.8 gpm site allowable limit (takes credit for Technical Specification reduced primary coolant i(xtine levels).

l 7.1 Conclusions j The leak rate analysis submitted in Reference I remains bounding and demonstrates the margin to

{ site allowable leakage limits for the combined degradation mechanisms of top of the tube sheet circumferential cracks, TSP ODSCC. and F*.

i 4

! 17 1

,, ,n= -- , . - -- , . , . . - --

l I

I l H.0 Braidwood Inspection Improvements j f

l The next TTS inspection to be perfonned at Braidw(xx! Unit I will be comparable to the inspection l perfonned at Byron Unit I in 1996 with regants to probe, analysis software and analyst training.

l This will include using the plus point probe with tne plus point,0.080" and 0.115" RPC. The analysis software to be used is still being evaluated. The Braidwood Unit 1 inspection will be l performed to the same guidelines as Byron's 1996 inspection. Training of analysts will be perfonned to these same guidelines and Byron Unit i 1996 and Braidwood Unit 1 1995 data will be used for the analyst site specific perfonnance demonstration test.

9.0 Conclusions The results from the structural margin assessment presented in Table 5.2 demonstrate that the structur:d enteria wen: met for both the vertical maximum and average voltages, and adequate '

margin will be maintained for the degradation distribution through the end of 461 days of operation above 5007 at Braidw(xxi.  ;

l The change in the EOC distribution fmm the currently approved October 15,1996 mid-cycle to the l refueling outage scheduled 165 days later does not significantly reduce the margin to the structural l requirements.

The leak rate analysis submitted in Reference 2 remains bounding and demonstrates the margin to l site allowable leakage limits (a small percentage of 10CFR100 limits) for the combined  !

degradation mechanisms of top of the tube sheet circumferential cracks. TSP ODSCC, and F*. 1 Operation of Braidwcxxl Unit I for a full cycle of 461 days does not challenge steam generator tube structural requirements.

Steam Generator tubes with TTS degradation will be repaired.

l l

18 l

l

I i,

5

( 10.0 References j 1. Braidwood Unit 1 Cycle Length Assessment Report, Transmitted to U.S. NRC February i 4

23,1996 j 2. Bmidw(xx! Unit 1 Cycle Length Assessment Repon Addendum, Transmitted to U.S. NRC-j May 17,1996

. 3. EPRI Draft Repon NP-6864-L Revision 2, Dated August 1993, PWR Steam Generator l Tube Repair Limits: Technical Support Document for Expansion Zone PWSCC in Roll  ;

Transitions i

b ,

i 1  !

1 i

i i

l l

l i

19 l J

I s '

i l

Appendix A

)

End Of Cycle Distribution Calcu'ation Inputs A.I Braidwood Unit 1 Site Specific POD Calculation I 1

The results from the 1996 and 1995 inspections at Byron Unit I were used to assess the inspection 4

capability of the inspection techniques used at Braidwood and to define a POD for the 0.080" RPC relative to the + point coil.

i An overall POD was defined for the inspection technology used at Bmidwo(xl in October 1995 relatiize to the inspection technology used at Byron Unit I during its 1996 inspection. The POD was defined from consideration of the differences in the coils (+ point vs. 0.080" RPC) and I differences in the analyst sensitivity and software used to evaluate the data (EddyNet95 vs.

ANSER). The calculation of an overall Braidw(xxl Unit 1 POD based upon Byron Unit 1 inspection data is presented here. The result is that a conservative POD of 0.20 is applicable to the Braidwood Unit i Fall 1995 inspection.

Two Methods were Used to Detemline a Braidwood Unit 1 POD:

Method 1:

1. As reported in Reference 2,78 pertent of 1996 plus point indications were present in the SG C i look-back to 1995 using the EddyNet95 software. The original 1995 analysis was perfonned I with ANSER and without the knowledge of the 1995 Byron Unit I tube pulls. Applying the .

78% factor to the 860 repaired SG B 1996 indications results in 670 additional indications l

which could have been detected in 1995 if the conditions of the 1996 inspection were present in 1995. Because the plus point probe was used in both 1995 and 1996, this represents improvements due to the software and analyst sensitivity, which wen: the major changes between inspections.  ;

2. During the k>ok-back of SG B 1995 indications, it was determined that 745 of the 978 SG B 1995 + point indications were present in 1995 with 0.080 RPC data. This represents improvements due to use of the + point pmbe in 1995, the first inspection with + point.

3. Combining the data pmvides a 0.080" RPC POD for the 1995 inspection at Byron. This value also represents the improvements between the Byron i 1996 inspection and the Braidw(xxl 1 Fall 1995 inspection which was performed with the RPC probe (no plus point coil), without experience gained frum Byron 1 1995 tube pulls and with ANSER software. The POD calculation for SG B is shown below:

Number of 1995 SG B Repaired 0.080" RPC Ind's Total Number of 1995 SG B Repaired Ind's + Number of 1996 Ind's Present in 1995 745 978 + 670 = 0.45 20

e l 1

1 l

This results in a POD of 0.45 to address differences fmm the inspection at Byron Unit I and Braidwood Unit 1 in 10/95. This POD multiplied by a POD of 0.6 to address other unknown issues for use at Braidwocxl Unit 1 provides an overall POD of 0.27. ,

I Method 2: l l

1. 1996 Indication look-back showed that 34% of 1996 indications are detectable with the 0.080 l RPC. Bierefore a POD of 0.34 could be used. I

2. Using a POD of 0.33 is conservative. The 1996 Byron Unit 1 inspection is the second inspection using the plus point probe. The size of the indications detected and repaired in 1996 is smaller than those detected and repaired in 1995, so it is expected that a smaller percentage of the plus point indications to be detected with the 0.080" RPC in 1996 than 1995 based on Figure 3.1. Sixty seven percent of plus point indications had an 0.080" RPC indication in 1995.

Use of 0.34 combined with a POD of 0.6 to address other unknown issues for use at Braidwood Unit I results in an overall POD of 0.20 for all voltage ranges.

The value of this overall POD is dominated by the smaller levels of degradation and is malistic for small flaws. The results in Figure 3.1 indicate that a POD of close to 1 is more representative for larger flaws, and that using POD = 0.20 will be very conservative for tubes with relatively large I degradation levels.

l l

A.2 NDE Uncertainty l Two factors were considered for NDE uncertainty input into the cycle length assessment. The two factors being probe wear and analyst variability.

A.2.1 Probe Wear An assessment to quantify the effects of probe wear on indication voltages was perfomied. The i probe wear was determined fmm the trends of calibration group voltages from the first tube to the last tube in the calibration group. During SG tube inspection analysis, voltages are nonnalized for each calibration group and therefore this trend represents the greatest indication probe wear uncenainty which could occur. The results of the assessment are presented in Table A.1 and A.2.

Data from the 0.080"RPC on the plus point probe for the Byron Unit i 1996 Inspection was used in the assessment. Table A.1 identifies the voltage of the 100% EDM notch for the first tube and the last tube for different calibration groups. Table A.2 identifies the voltage of the 100% through wall drilled hole for the first tube and the last tube for different calibration groups. The percent deviation is then calculated for each calibration group from this data. Finally, the standard deviation of the percent pro 5e wear of the different calibration groups is calculated. The msults of i this assessment are 5.62% & 5.9% probe wear for the 100% TW EDM notch and the 100% TW l

drilled hole, mspectively. The input in the cycle length assessment used for the probe wear measurement uncertainty is six percent. Because tubes were not removed from service based on a specific level of probe wear, no cutoff in the distribution was used in the Braidwood evaluation.

21

l This procedure is similar to what is used for the Byron and Braidwood ODSCC interim plugging '

criteria. The last tube voltage is generally higher than the first tube voltage which provides conservatism to the calibration group voltages.

A.2.2 Analyst Variability i

In Reference 1, Comed provided information regarding two blind tests perfonned at Byron Station.

The blind tests were conducted with four different analysts. The scope , development, protocol and results are discussed in detail in Section 3.0 of Reference 2. The blind tests included a 100 tube and 2(X) tube test. Data from three years (1994,1995 and 1996) were included in the test. As a part of the blind tests, the voltage of the indications detected were recorded by the four analysts.

The deviation between the four analysts recorded voltages from the blind tests were used to determine a percent analyst variability for contribution to the NDE uncenainty.

The mean of the voltages arported by the analysts for each tube was calculated and considered to be the reference for that tube. A distribution of the mean voltages is presented for the 100 tube and 200 tubes test in Figure A.1 and A.2, respectively. The deviation from the mean for each analysts reponed Oge for each tube was calculated.1he analyst variability was calculated by dividing the devia ne mean. A distribution of analyst variability from the mean is presented for the 100 tube a,m J ;0 tube tests in Figure A.3 and A.4, respectively. The standard deviation of the i analyst variability for the tests topulation is then calculated. This procedure is similar to what is I used for the Byron and Braidwood ODSCC interim plugging criteria.

The total number of observations was 396 for the 100 tube test and 1127 for the 200 tube test. Of I the observations, thirteen were re-evaluated through a resolution process for the 100 tube and 2(X) tube tests. The criteria used to select observations to be re-evaluated was any voltage which exceeded the mean plus two times the standard deviation of the mean voltages. These tubes were then re-evaluated by a LevelIll ECT analyst and the resolved value substituted into the data. This l process is similar to the resolution process utilized for ECT primary and secondary analysis where discrepancies in tiie analysis between analysts are submitted to a resolution analyst for disposition.

Two additional criteria were used to screen the data. First, any indication which had only one call by the four analysts was not included in the sample. Secondly, observation voltages from different coils for the same indications were excluded.

The results of this assessment are a standard deviation of analyst variability of 0.19 for the 200 tube test and 0.22 for the 100 tube test. A value of 0.20 analyst variability standard deviation was used in the NDE uncertainty input for the cycle length assessment calculation.

22

Table 3.1 Byron and Braidwood Unit 1 Largest Indication Voltages For Average and Maximum Voltage Byron 0.080 Voltages Braidwood 0.080 Voltages Circ Sensitive l Inspection Outage Maximum Integral Outage Maximum Integral t

l First B1 RO6 (1994) 4.14 1.02 A1M05 (2/95) 2.53 1.21 1

Second B1 PO2 (1995) 1.24 0.49 A1 R05 (10/95) 1.36 0.52 Third B1 RO7 (1996) 1.02 0.57 - - -

i

C Inspection Results for Braidwood Unit 1 February and October 1995 - 0.080" Coil All Steam Generators I Table 5.1 All Data Normalized to 20 volts - 100% EDM Notch YEAR PROBE ROW COLUMN MVIR VIR Feb-95 610 4 48 0.37 0.21 Feb-95 610 16 111 1.3 0.6 Feb-95 610 4 107 0.77 0.28 Feb-95 610 12 73 0.97 0.27 Feb-95 610 22 44 1.9 0.69 Feb-95 610 21 44 0.71 0.32 Feb-05 610 21 45 2.53 1.21 Feb-95 610 20 46 0.75 0.26 Feb-95 610 19 50 2.5 0.71 Feb-95 610 20 51 1.3 0.4 Feb-95 610 38 48 0.78 0.2 Feb-95 610 26 65 0.64 0.16 Feb-95 610 45 24 1.99 0.41 Feb-95 610 21 39 0.54 0.31 Feb-95 610 23 40 1.8 0.64 Feb-95 610 19 45 0.73 0.21 _

Oct-95 610 27 38 0.59 0.21 Oct-95 610 22 41 0.56 0.28 Oct-95 610 21 41 0.65 0.23 Oct-95 610 20 42 0.73 0.23 Oct-95 610 25 49 1.15 0.42 Oct-95 610 46 45 0.38 0.2 Oct-95 610 29 61 0.67 0.26 Oct-95 610 23 70 0.26 0.09 Oct-95 610 25 70 0.53 0.21 Oct-95 610 2 109 1.17 0.47 Oct-95 610 3 111 0.33 0.18 Oct-95 610 4 111 0.39 0.16 Oct-95 610 4 109 0.84 0.32 Oct-95 610 9 110 0.61 0.23 Oct 95 610 21 54 1.36 0.52 Oct-95 610 13 102 1.06 0.46 Oct-95 610 15 99 0.69 0.29 Oct-95 610 9 68 0.45 0.1 Oct 95 610 11 49 0.82 0.32 Oct-95 610 16 69 0.88 0.4 Oct-95 610 24 53 0.44 0.15 Oct-95 610 30 53 0.61 0.25 Oct-95 610 8 66 0.6 0.27 l

Industry Tube Pull and insitu Pressure Testing Table 6.1 Total Number Available ECT Data Pulled Tubes 33 24 Number Burst Tested 15 11 (Note 2)

Number Crack MET Size 33 24 Insitu Pressure Tests 33 32 Total Number of Tubes 66 56 Note 1: The largest tube pullindication with no ECT data has a MET average crack depth of 35%, these tubes will not contribute new data to this evaluation Note 2: Includes 2 Byrcn tube pull simulations l

l l

l

i c.

i Table A.1 Plus Point Probe Wear Assessment 1

Affect on 0.080 Coil Voltage i

1996 Data 100% Throughwall EDM Axial Notch l

100% EDM Notch Voltage

) First Tube Last Tube Percent Number of volts volts Deviation

} Cal Group Tubes (x) (y) [(x-y)/x]

i 10 90 20 22.44 12.20 %

14 100 20 21.35 6.75 %

18 46 20 20.95 4.76 %

22 112 20 21.54 7.68 %

26 109 20 20.58 2.92 %

30 96 20 18.32 -8.39%

34 25 20 19.87 -0.65%

4 49 20 21.97 9.85 %

8 31 20 20.61 3.06 %

12 77 20 21.82 9.12 %

16 90 20 20.63 3.14 %

20 44 20 21.38 6.92 %

24 98 20 22.44 12.21 %

Total Tubes = 735 Standard Deviation = 5.62%

i i

i Table A.2 i

Plus Point Probe Wear Assessment i Affect on 0.080 Coil Voltage 1996 Data 100% Throughwall Hole 100% Hole Voltage First Tube Last Tube Percent Number of volts volts Deviation Cal Group Tubes (x) (y) [(x-y)/x]

14 100 10 10.62 6.20 %

18 46 10 10.07 0.70 %

22 112 10 11.33 13.33 %

30 96 10 10.43 4.31 %

34 25 10 10.15 1.52 %

96 29 10 11.32 13.20 %

100 6 10 9.47 -5.25%

102 45 10 10.58 5.84 %

4 104 29 10 10.75 7.45 %

106 41 10 11.13 11.29 %

Total Tubes = 459 Standard Deviation = 5.90 %

)

i i

i i

Byron 1996 SG C Circumferential Indications Detected by + Point only and

+ Point and 0.080" RPC Figure 3.1 300 250 -

200 --

e E

g  : Without 0.080 Indication (+Pt. Only)

[ 150 -- -- * --With 0.080 & +Pt. Indication E

z 100 --

,.N' ...

,. ., 0.98 Volts (+ Point 50 __

p349 Tubes & 0.080 RPC

/ , Same Detection

.- 's. , Level) 1.47 Volts

! l l a.,~'"----1

,; _...i .. .. .... ....___.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 Plus Point Vert Max Volts

~ ~

M __ /

6 ~~.

I Figure 3.2a R3 C111 SmallIndications ANSER 0.29 VOLTS 1

41- N r __ - _ _ -  %

4- R -. ~~

w \

..s -- * /

1 24o 120 o Figure 3.2b R3 C111 SmallIndications Eddy Net 95 0.30 VOLTS

X 1"M_.

Figure 3.3a RIS C99 Medium Indications ANSER 0.45 VOLTS

..w -

L 3, f U- W- _

_f_^

w no .

I Figure 3.3b R15 C99 Medium Indications Eddy Net 95 0.45 VOLTS I

I.

=z -

~

f MP

-M M

-"A~

Figure 3.4a R21 C54 Large Indications ANSER 0.66 VOLTS t.p_

C

~~~D a

.y. W W

_ i

. .i .

Figure 3.4b R21 C54 Large Indications Eddy Net 95 0.67 VOLTS i

.J

4 l

1 l

1 1

l

-l we v*w%e soiancatesses sou n : e .4 nza guejl . tu  ;

e.e 1.e s.e y S.0" I '$

s e PLOT LINE

• 2.e 1.e

..; '..,.. ..,.. . . . . ..,.. . . . . . . . . . . . . .r-

. . . .. m.. . *- VOLTAGE THRESHOLD LINE g

cmsne e(4 ) i na s.* si.=- she.s nn l /4 n l vm. As c 4.w ,

v** he Fd LSE UUW j a f,l -

na c= nw m. va, r@ "

d 8""

• D=**=88 f an v.w 1

l Figure 4.la: Cursor Scan Lme l

i I

l

__ . . . _ _ . _ _ . . . _ _ . . . _ _ _ _ _ _ _ _ m _ _ _ . . . _ _ _ _ _ _ _ _ _ _ . _ _ . . _ . _ _ _ - _ _ _ . _ _ _ _ _ . _ ._

i

• 1 i

9 i.

4 i

i 1

_j - - -

a e,* (C)-ismass i u n. (usi l to i

3 p

_e. _4 e,* _s,.== 1.* t.,- Ass omis.y. _ Heir i.e.c e l me I i Sc1ANCaLe0Ms WED24125 OCi6 3G1A ftDWseeCOLese ids 3 1

< t-6. > ses canc it is came si e aos li c- l ci i rame  ;

l i

l t e 270 t c- ci vore c- ci vore s.es v/s ., tas I 1

? t i

i 4

i L

-o.m f L

L o.u mm -

una M N N lise}

mas n.se ,,,,,

A Y + uess ><

C a M O r ,,

l Figure 4.1b: Normalization for the 0.080" RPC to 20 Volts

r t

l S

l 1

I f

i i

l I IJI  !

,_i nec cana soime= === so a a n c es nn Iwi  !

mter ev etwa nr -a ne===? m.t.es zht.m m .a -+ ,

u,=m al c I  !

-0.00

1. l

• =  ! [El, ,

i e

l @.

't.

N g- ? l

= l

sia mia l

se

/

/

220

@ $W 11 i

0

/ t 0..

"' mw.

1. win a

m I - ,

• 9.11 .

. ._.- a i

Figure 4.1c: RPC display C-scan Showing Strip Chart Cursors Located at Maximum Vertical Channel Displacement and Expansion Zone Circumferential Indications f

i 4

l l

i i

l 1

l s

l l '

-i are v.a w.,e sowcassues sou a ae 44 nn tien . u 7.M s.o-y s.o-

. d.e-3.o-t g,o-

• t.o- ,

o.. ..,.. ..,.. ..,.. ..... . . , . . . . . . . . . .

o.o es.o m.o tm.o sao.o m.o no.o us.o me.o Chemise.renetsiE. test (dog)

Pts Seats St.eums aigleSeanetst i laterwl ** ** c-

• m l va t.,4 too se Is=* / el -

l n.t c s.

rw anaus r=*.

vw.o. M , -

Is== n el n.**is v.zu I i 1

l i

1 1

< l l . I I

I I  ;

Figure 4.id:

I Circumferential Line Scan as Displayed in the RPC Voltage <

l Integral Window Showing Maximum Displacement t ,

I i

l l

l l

i

r Number of Tubes 1

' o N A cn co o N A cn co

! 0.00 - , ,  ;  ;

l l l 0.05 n.....

0.10 - .....,~'- m .-

0.15 < r..'

O.20

> -a y

1 m

(

0.25 < ....-...........***~.....n

~.......

n lE

• ~~ O 0.30 ..m. o l

0.35 C

I t

0.40 < r 2 0.45 - L '

o .

• N 0.50 < e un

> 0.55 < a ... ~:n s 3

e 0.60 >

B

'O 0.65 - <> 9 e

ut o

< 0.70 - <>

i <

in 0.75 - <>

n. .O.

e um m 0 80 ~ CD 8

e 0.85 - 3 0.90 - -

g 3

0.95 - m_

1.00 - 9,,

m 1.05 O 1.10 -

1.15 - D 1.20 - ]

1.25  : >

e m

.<, 2 co cn o

,~+

S3 N 6E C E E e o

8. 0.

-m 9

wub, 01 -

b.

>=>

"C (n ~~n

~O Om a

e 9

_ oQ D

(,m 1,oo C2% , ,o e@ g. 2eW <o29ee he"m**D Tpom Nm

• o 1 E#M.3e3 <oWmOe m oh

• A0 aA o M : # -

O : -

s i

s - -

e ,

b u W: .

~ -

T ,~

f

. ,. '4 e.

sw6 l ~Eu-i o

>= C~Om o . - . '

1 r

e .-:

- ' > $ . .mB6bg -. coU b >=(n~Om b .

mG  : a .

u .

N ..

. n n '

~

A .

a - -

. 'm .

N : -

a.

• o (

I

" }'

~ ~

o " 4 0 1 2 3 4 5 6 7 8 9 0 1 2 - 3 41 5 6 7 8 9 0 i. 2 3 4 5 6 0 0 0 0 0 0 0 0 0 0 1 1 $ 1 1 1 1 1 1 2 2 2 2 2 2 2 o 8 o'1O o3 yex 3 4ozw0e 1

\ i ,\ i .' . l- :l, 1

.o AVolts/EFPY vs. Relative Frequency, Average Voltage Figure 4.3 0.50 Total Data Points = 757 Byron S/G B: 94 to 95 Byron S/G C: 94 to 95 & 94 to 96 0.40 -

D y 0.30 E

u.

I 3 0.20 -------------------

0.10 - -

0.00

-0.5 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 AVolts/EFPY, Average Voltage

I AVolts/EFPY vs. Relative Frequency, Vertical Maximum Voltage  :

Figure 4.4 0.30 t Total Data Points = 757 '

Byron S/G B: 94 to 95 0.25 7.

x 0.20 o

c if .

E 0.15 - -

l I

=

.!!! t e

" 0.10 -

0.05 ,

0.00 ' ' ' ' ' ' '

I

-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 AVolts/EFPY, Vertical Maximum Voltage

Byron Unit 11996 S/G C Indications: Average vs. Maximum Voltage Figure 4.5 3.0 Total Data Points = 357 2.5 2.0 CU 5

E 1.5

3 -

.E -

a -

lE -

1.0 , , $ ,

+ +o +

0.5 0.0 O.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Average Voltage

. i;lll4 i e

4 1

l

• ' 2 1

0 e

g 1

t a '

l o '

V m '

u 8

i m 0 g e

x a ,

a ' t l

M6

• V o

s 4 .

e ver g a

eu gg .

r e

ai v rF e

1

+ 6 0 A v

A d

o .,

o '

w + . 4 d

i l

+ ,+

a 9 0 r 3 B =

s  :

t n * '

i o ,

P . . '

a .

i$.'N.

t a +. 2 D ,'

0 la .

t o ** '

T -

- - . . . . . . - . - - .- - - : - . . 0 0

0 5 0 5 0 5 0 3 2 2 1 1 0 0 g, o> El3E7=2 ll 1!1

.o l

?

Number of Degraded Tubes vs. Vert. Maximum Voltage Figure 5.1 200 '

. Number of Degraded Tubes al Braidwood = Number of 0.b80 Indication's / 0.20 175 .

- + - Braidwood 2-95 ISI, All S/Gs 150 ~

E - 0 Braidwood 10-95 ISI, All S/Gs

.o -

i

-e-Byron 10-95 ISI, S/G B l 125 1  :

3- -

o

? 100 Q -

o

~

$ 75 E  :

z  :

50 .

25 .9 0

?'Y  :* =- -  : :*

': : : :W? - - - - - -

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Vert. Maximum Voltage, volts

- 1l 0

0 2

0 0

h. 8 0  :

1

/

s  :

n o  :

it 0 e

g b i 6

d  :

1 lt a In 0  :

o 8 V 0  :

0 D

e g f 4

o 1 a

r r

e s e b s G  :

v m G S

/ '

A u /

S ' 0 s N

l l

t

. l A B 2

s =

l A I, 1 l

o v2 d I, S G/

v s5 o S I S e e o I 5 I, g

b er w 5 9 S  : a u id 9- -

0 Tg u 0 I t a

r 2 1 5 0 o l

de F i

B d d 9 -

1 V t

a o o 0 o o e d

a s ww 1 n

g a

r e id id o  : r g b a r

ar r y 0 e e u B- B B 8 v h A D d -

0 e

f o d a

+ : :

r r -

e g b e 0 b 6 m f o - 0 u r N e  :

b '

m u 0 b

L.. ' 4 0

E .

0 2

0 0

0 0 5 0 5 0 5 0 5 0 0

_ 0 7 5 2 0 7 5 2 2 1 1 1 1

,8, ee8Q 7) y E :s l'

Braidwood: Number of Tubes vs. Average Voltage Distribution 461 Days, POD = 0.2, Analyst Uncertainty = 0.2, Probe Wear = 0.06 Figure 5.3 20

-+-NDE

- O EOC 15 g O O

[ .

}10

.8 -

E .

O 5 -

0 0 5

O O

O O

' ' ' ' ' ' ' ' ' ' ' ' *

• 0 " ^

0 0 0 ' 0 0 "

O -

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Average Voltage, volts

Braidwood: Number of Tubes vs. Vertical Maximum Voltage Distribution 461 Days, POD = 0.20, Analyst Uncertainty, = 0.2, Probe Wear = 0.06 Figure 5.4 25 ,

-*-- N D E

, o EOC 20 g 15 F

o a- I o o l10 z

l o o o

O

^

5 o O O o

oo oo 0 0 - - -

"^^**^ 0 0 O.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Vertical Maximum Voltage, volts

Average Voltage Structural Limit vs. Adjusted insitu or Burst Pressure Corrected to Braidwood LTL Properties Figure 5.5 12 i o insitu (No Burst) m Burst 10 = m . Limit Curve E

a 3 '

x 8 p e a E. '

[6 V

o

^"

o o Lower Limit 4035 psi 2 0

$ e o ao oo 0

g , ,

4 4 o jo d OG 2

Structural Limit at 4035 psi = 0.91 volta 0

O.0 0.2 0.4 0.6 0.8 1.0 1.2 Average Voltage, volts

,o Maximum Voltage Structural Limit vs. Adjusted insitu or Burst Pressure Corrected for Braidwood LTL Properties Figure 5.6 12 o insitu (No Burst) 10

• m m Burst

% , Limit Curve 8

. a m a

g 6 o fo e 0

g Loyver Limit 4035 psi

" o 90 @ O 4 o oo O o o o o, o 2

Structural Limit at 4035 psi = E.64 volts 0

O.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Maximum Voltage, volts

_ _ _ - _ _ _ - _ . _ _ _ _ _ _ _ _ - _ _ ~ -.- . - . - _ __. _ _ _ _ _ _ _ _ -

~

Braidwood Unit 1 EOC Maximum Voltage Distribution 461 Days Vs. 296 Days Figure 5.7 12

, s O

O 10 O

g 5 m

0

.8 o

O o 461 Days m 296 Days

$ o Z m , g O E O u O 2

u U o "o 0 :" = a e a a a a a 3 ..

0.0 0.5 1.0 1.5 2.0 2.5 3.0 Maximum Voltage

s y y s

a a d d 1 6 6 9 4 2 ou

0 2

s y  : 81 a  :

D 6  :

9 2  :

6 s  : 1 V  :

s y _

a  : _

D e 1 4 6 e 1 4

n .

i o  :

t u  :

b i

t r  : 21 _

i8 s 5 -

e D 5 5 g a

ee gr

• lt o

t aug 0 V l

• 1 e oi F g V

• a r _

e 0 e

g v a A r

e 8 v 0 A O C E O

E o E 1 6 _

o E 0 it n O U _

0 m d

o j o 4 w , _

0 .

i d

a r

B " O m o 2

o 0 m o

,o

=

0 0

6 4 2 0 8 6 4 2 O 1 1 1 1 eE y oE>o.E 2*

f e

7 a

s sn

." o o N u

(5 E

sn

$O o 9

e On E

e En ee

+ 8 6

e .,

E5 .

+

g 39 N u. o x

+

• 2 I E o 3 * . H

~ , o N $

(lL.

g .

o 3 h-g T W

• 6

~

6 .

=

g . .. .

b 45 t.-. ** 8 6

. *f

+ .t..

8 o o o o o o o 6 9 9 9 o 9 O M N N - W o o Silo A 6 A Y

M Industry Tube Pull and insitu Pressure Test Leak Rate (Corrected for Temperature and Pressure) Vs. Maximum Voltage Figure 7.1 0.18 0.16 0.14 0.12 E

$ 0.10

+ l + Leak Rate (gpm)l h

a:

g 0.08

.5 0.06

+

0.04 Leakage Threshold = 1.1 volts, maximum leakage = 0.164 9Pm 0.02

• +

0.00 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 Voltage

- . . _ _ - _ _ - _ _ _ _ _ _ - _ _ - - _ _ _ _ _ _ - - _ - _ _ _ _ _ _ _ - - _ _ - .__ _ - - . , ____-____w _

Distribution of 100 Tube Blind Test Mean Voltage Figure A.1 25 20 -

1 j 15 --

lE# of Tubes l 1

$ 10 - -

5- -

0 ' '

i 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 Volts

e s.

C m

c>

llE

_E E ".-

m: '

1 C

! 2=

1 i

$ 5 I

C 1

. .O i

-e aw -

0 O :s a >

' 13 C

a z=

1 >

a J

4 4

4 a

o a

i 4

i o

i 1

f i t 3

. , n, , , ,

4 c- o o o o o o c.

F- O O 4 (9 N w-1 segnilo JeqmnN 4

5 1

1 i

. l T

$ I d'

'O l w l

_E l

l CL eL L*L  :

C e w

$ $ I N ll ci c 6'O is .9 I 5 gf l 8'O j

= z. ,

.!!! m u) '

> 's a EO I O

e mO 1 O_

O g p 9'O am VJ DD x cE -

9'O c e 5a

• cq mz {

44 . -

po o2 5 _S) C'Oz{

z u. -

.5 E'O 'g b o es .

5 M' L'O Y j i

e 0 F

. L'0-m g o-3:s ..?.

h -

j.:

.- .f.. c o.

8

, . ;; p o.

-z;: g.0-l 9'0-

[0-l l l , , l  : l 8 8 8 R 8 8  % 8 8 S seqn.L 10 JaqmnN

,, l l

! i I a

~

c.

l n

o it ia v

- e _

D l

E

.=

7 2

- n 1 a 1

- e 2 =

M 9 s .

1 n

m o -

o =i -

na t _

- r _

f n o vr '

i o it e as -  : -

t a ivb i

v eO -  : .

Df _

e dr r o

- D _

- t s

ae db

- y nmu a " : na l

a t

- n SN '

e A 4

Mm

- f A o e o nr r i

o ug  : f n

t i uF i o

t _

- b i  : i av r

t e i

s D

- D t

s -

e  ;

T '

d i

n - _

l B

e a _

b u  ;

T 0

0 2

~

. 0 0 0 0 0 0

. 0 0 5 0 5

. 3 2 1 1 S 3EE5

-