ML20199G596

Cycle 7 Voltage-Based Repair Criteria Rept for Jan 1999

ML20199G596
Person / Time
Site:	South Texas
Issue date:	01/31/1999
From:	WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
To:
Shared Package
ML20199G594	List: ML20199G596 NOC-AE-000405, Forwards voltage-based Criteria 90-day Rept for SG Tube Exam Performed Under NRC GL 95-05,during Refueling Outage.Rept Contains Info Required by Section 6.b of Attachment 2 to GL 95-05
References
SG-99-01-002, SG-99-1-2, NUDOCS 9901220347
Download: ML20199G596 (100)

Text

- - _ _ __ ..- . - _ - - - - . _ _ _ _ _ ___ . . - _ _ - . - - - - _ - - . - _ - - - -

SG-99-01-002 1

SOUTH TEXAS UNIT - 2 CYCLE 7 VOLTAGE-BASED REPAIR CRITERIA REPORT l

January 1999 0

Westinghouse Electric Corporation Energy Systems Business Unit l

l Nuclear Services Division i F.O. Box 158 Madicon, Pennsylvania 15663-0158 9901220347 990119 PDR ADOCK 05000499 P PDR i

. -- .- . - . ~ . . . . - . _ - - - = - . . - . - - . . - . . . . - - . - . - . - . . . . - - . . - - .

I l

i SG-99-01-002 t

i l

1 SOUTH TEXAS UNIT - 2 CYCLE 7 VOLTAGE-BASED REPAIR CRITERIA REPORT January 1999

1 Table of Contents l

Page No.

.1.0 Introduction 1-1 2.0 Summary and Conclusions 2-1 3.0 South Texas Unit-21998 Pulled Tube Data for TSP Locations 3-1 3.1 South Texas Unit-2 Pulled Tube Examination Results 3-1 3.2 South Texas Unit-2 Pulled Tube Evaluation for Voltage-Based Repair Criteria Applications 3-5 L 3.3 Comparison of South Texas Unit-2 Data with the EPRI l Database 3-7 1

4.0 EOC-6 Inspection Results and Voltage Growth Rates 4-1 4.1 EOC-6 Inspection Results 4-1 4.2 Voltage Growth Rates 4-3 l 4.3 NDE Uncertainties 4-4 4.4 Probability of Prior Cycle Detection (POPCD) 4-4 4.5 Assessment of RPC Confirmation Rates 4-7 4.6 Probe Wear criteria 4-8 5.0 Database Applied for Leak and Burst Correlations 5-1 6.0 SLB Analysis Methods 6-1 7.3 Bobbin Voltage Distributions 7-1 7.1 Calculation of Voltage Distributions 7-1 7.2 Probability of Detection (POD) 7-2 l 7.3 Limiting Growth Rate Distribution 7-2 )

7.4 Cycle Operating Period 7-3  !

7.5 Projected EOC-7 Voltage Distributions 7-3 l 7.6 Comparison of Actual and Projected l EOC-7 Voltage Distributions 7-3 l l

8.0 . SLB Leak Rate and Tube Burst Probability Analyses 8-1 8.1 Leak Rate and Tube Burst Probability for EOC-6 8-1 8.2 Leak Rate and Tube Burst Probability for EOC-7 8-2

! 9.0 References 9-1 i

Q: \ ape \ thz \ thx98 \ taparc \ thsc690d. doc l

i

, - ~ , , - . - - . ,.. , , , . - , , - - - . . - - - .--- , ,

l South Texas Unit - 2 Cycle 7 Voltage-Based Repair Criteria Report 1.0 Introduction This report provides a summary of the South Texas Unit-2 steam generator (SG) bobbin and rotating pancake coil (RPC) probe inspection at tube support plate (TSP) intersections, together with postulated steam line break (SLB) leak rate and  !

tube burst probability analysis results, in support ofimplementation of a voltage- ,

based repair criteria for Cycle 7 as outlined in the NRC Generic Letter 95-05 I (Reference 9-1). A 1.0-volt repair criterion for outside diameter stress corrosion cracking (ODSCC) indications at the TSP intersections is being implemented for i the first time starting with the current cycle (Cycle 7) for Unit-2. Information  !

required by the Generic Letter is provided in this report including SLB leak rates and tube burst probabilities calculated using the end of cycle (EOC) conditions for the last cycle (Cycle 6) and projection of bobbin voltage distributions, leak rates and burst probabilities for the EOC-7 conditions.

Analyses for Cycle 6 were carried out using the actual bobbin volti ge distributions measured during the EOC-6 outage and the results compared with corresponding results from projections based on the EOC-5 bobbin voltage data presented in the technical report submitted to justify the 1.0-volt repair criteria (Reference 9 2).

Westinghouse generic methodology based on Monte Carlo simulations presented in Reference 9-3 was used in these evaluations, and this methodology was also utilized for the analyses performed for Unit-1 after its recent outage (Reference 9-6).

Analyses were also performed to project leak rates and tube burst probabilities for postulated SLB conditions at the end of the ongoing cycle (Cycle 7) based on the 1.0 volt repair criteria. These analyses utilized bobbin voltage distributions measured .

during the recent (EOC-6) inspection and a limiting growth rate distribution from I the last two inspections (EOC-5 and EOC-6 inspections). l Two other supplemental evaluations are also presented in this report. One of them examines probability of detection for Cycle 5 inspection (probability of prior cycle detection - POPCD) and the other assesses the fraction of the indications that showed no degradation during the RPC inspection in 1997 (EOC-5 inspection),

were left in service at beginning of Cycle 5 (BOC 5), and were RPC confirmed in 1998 at EOC-6.

Two tube segments (R18C100 and R19C83) in SG-A each with 4 TSP intersections (TSP 1 - flow distribution baffle (FDB)- and TSPs 2 to 4) were pulled during this Q Aapc\thx\thx98\tsparc\thxc090d doc 11

inspection for detailed laboratory examination. Results from leak and burst tests  ;

~

and metallurgical examination are presented in Section 3. Eddy current and repair data for EOC-G TSP indications are provided in Section 4. The leak and burst databases applied and the Monte Carlo analysis used to wumate leak rate l l and tube burst probability are briefly described in Sections 6 and G. The actual  ;

I EOC-6 voltage distributions as well as leak rates and tube burst probabilities  !

j calculated for these distributions are compared with the preiactions for EOC-G l conditions (performed using the EOC-5 data) in Sections 7 and 8 Leak rates and

! burst probabilities for the projected EOC-7 voltage distributir os are reported in l Section 8 and compared with allowable limits.

l l

l l

l l

i i

Q:\a pe\thx\thx98\tspa rc \thxc690d. doc 1-2

2.0 Summary and Conclusions A total of 1485 indications were found in the EOC-6 inspection of which 40 were over 1 volt and 6 of these 40 indications were above 2 volts. Thirty-nine of the 40 indications over 1 volt were found on the hot leg side and one indication on the cold leg side. Forty-four TSP indications, including all 39 indications on the hot leg side over 1 volt, were inspected with an RPC probe and 34 were confirmed as flaws. The single cold leg indication over 1 volt was not RPC inspected, but was treated as a RPC-confirmed indication and repaired; this indication is further described in the

, paragraph below. SG-B had the largest number ofindications among the four SGs with 500 bobbin indications; however, SG-A had the highest number ofindications above 1.0 volt, and it also had the 4 largest indications found during this inspections, all above 2 volts. All 11 indications above 1.5 volts and 23 out of 28 indications between 1.0 to 1.5 volts inspected by RPC probe were confirmed as I flaws and were repaired. No ID or circumferential indications at the TSP intersections, or indications extending outside the TSP were found in this inspection. A total of 47 TSP intersections in all 4 SGs combined with a mixed residual signal that could potentially mask a 1.0 volt bobbin indication (residual signal voltage 1.5 volts or greater) were inspected with a RPC probe and 4 of them were found to contain single axial indications (SAIs), and they were repaired.

In SG-C, the bobbin signal for the first pre-heater baffle plate intersection on the cold side (220)in tube RIC102 was initially called as a wear indication and was assigned 2.05 volts. A later reexamination of this bobbin signai indicated that it may be a potential crack.like signal and its voltage was revised to 1.23 volts. As a crack and potential ODSCC indication subject to GL 95-05 requirements, the )

indication' would be required to be RPC inspected. By the time this reassessment j was completed, equipment needed for RPC examination of this intersection had I been removed from the steam generator. Based on discussions with the NRC, it was concluded that the RPC inspection could be omitted and this tube (R1C102) was repaired, which is equivalent to assuming RPC confirmation of the indication as a crack-like flaw rather than wear. To ensure proper classification of cold leg indications in future inspections, all pre-heater baffle plate intersection indications on the cold leg side will be inspected with an RPC probe. Both bobbin and RPC l data will be used to classify the indications as ODSCC or wear. Indications extending outside these baffle plate intersections will also be RPC inspected.

SLB leak rate and tube burst probability analyses were performed for the actual EOC-6 bobbin voltage distributions as well as the projected EOC-7 bobbin voltage distributions. T5a e.nal sis took credit for the availability of pressurizer PORVs by l' using a primary-to-secondary pressure differential of 2405 psid for the design-basis SLB event. The actual number ofindications detected dunng the EOC-6 inspection .

are about 12 % to 42% below the corresponding projections performed I i

l Q:\ ape \ thx \ thx98 \ tsparc \ thxc690d. doe 2-1

--r - ,. . . , , , . . - , , , - _ _ --

__m _ ._ . _ . _ _ _ _ . _ _ _ . _ . _ _ . _ - . _ _ . _ . _ . _ _ .

using the EOC 5 data and POD =0.6 for all SGs except SG-C for which the actual l number is about 5% highe: than the projected number ofindications. While the peak EOC-6 voltage measured for SG B was below the value projected, 4 indications in SG A and 1 each in SGs C and D exceeded the peak voltage projected l for them. Because of detection of a 4.1 volt indication in SG-A, the leak rate and l tube burst probability based on the actual voltages are higber than projected for l l that SG. However, the absolute magnitude of the SLB leak rate (4.Gx104 to

! 3.2x10-2) and tube burst probability (1.2x104 to 3.8x10-4) values based on the actual conditions are small, and they are almo:t 2 orders of magnitude below the

! acceptance limits (15.4 gpm at room temperaturc and 10 2) for all 4 SGs.

The leak rate and tube burst probability projections at the EOC conditions for the i current vele (Cycle 7) are also well within their acceptable limits. The limiting SLB leak rate projected for the EOC-7 conditions using the standard analysis methodology (Reference 9 3) and a constant POD of 0.6 is 0.033 gpm. This value is projected for SG A which had the largest indication found in the EOC 6 inspection, and it is more than 2 orders of magnitude below the allowable EOC-7 leakage limit of 15.4 gpm (room temperature). The highest tube burst probability, 4.2x10-4, is also predicted for SG-A, and it is more than a decade below the NRC reporting guideline of 10-2. Two sensitivity analyses were also performed for the limiting SG (SG A). The EOC 7 projection for SG A was repeated using leak and burst correlations updated to include new data from the tube specimens pulled during the present inspection. While the EOC-7 tube burst probability did not change l significantly, the SLB leak rate increased from 0.033 to 0.045 gpm. The Cycle G growth data for SG-A appear to show a dependency on the beginning of cycle (BOC) voltage. Therefore, EOC-7 projections for SG A were also repeated using the methodology recommended in Reference 9-4 to account for growth dependency on BOC voltage. The projected EOC-7 leak rate increased from 0.033 to 0.040 gpm and tube burst probability increased from 4.2x10-4 to 5.5x10-4 The magnitude of increase in SLB leak rate and tube burst probability in the above two sensitivity analyses are small in comparison to the margins available to their respective acceptance limits. Thus the GL 95-05 requirements for continued plant operation for the projected duration of Cycle 7 are met.

As the magnitudes of the projected EOC-7 leak rates and tube burst probabilities are very small, there is some potential for the leak and burst results based on the l actual EOC-7 conditions to exceed their projections. As in Cycle 6, occurrence of just one indication in the modest voltage range of 3 to 4 volts, which is not considered highly improbable, can result in the actuals exceeding their projections.

However, even if the SLB leak rates and tube burst probabilities for the actual EOC-7 conditions exceed their projections by a factor of 5 to 10, they would still be an order of magnitude below their respective limits.

l Q.\apc\thx\thx98\tapa rc\thscG90d. doc

(

! 22 l

Probability of detection (POPCD) for the EOC-5 inspection was assessed using EOC-5 and EOC-6 inspection data. Although a voltage-based repair criterion was not applied at EOC-5, the eddy current data were evaluated using the same

procedures as applied to plants using voltage based repair criteria. Therefore, l EOC-5 inspections can be used for POPCD evaluation. The results support a l detection probability greater than the NRC mandated value of 0.S. Four l indications with no degradation found (NDF) by RPC during the EOC-5 inspection i

were tested again in the EOC 6 inspection and 3 were confirmed yielding a RPC confirmation rate of 75%. Currently, the database for the RPC confirmation rate for prior cycle NDF indications in the South Texas units is too sma'l to recommend l a confirmation rate for use in the projection analyses. All RPC NDF indications are included in the EOC-7 projections presented in this report. ,

i i

i 4

i l

Q.\apc\th x\thx98\t oarc\th xc690d. doc 23

3.0 South Texas Unit-21998 Pulled Tube Data for TSP Locations 3.1 South Texas Unit-2 Pulled Tube Examination Results 3.1.1. Introduction Two tubes removed from SG-A of South Texas Unit 2 (R180100 and R19C83) were examined in a hot cell at the Westinghouse Science and I'echnology Center.

The pulled tube segments. included the following areas of interest for TSP intersections from each tube: first tube support plate (TSP-1 or FDB), and the TSP-2, TSP-3 and TSP-4 locations. Prior to tube removal, field eddy current inspections showed potential indications (PI) by bobbin probe at the TSP-2 and TSP-3 locations of Tube R18C100 and at the TSP-2 location of Tube R19C83.

Field 80 mil pancake coil single axial indications (SAI) were observed at each of

, these locations. In addition, the TSP-3 region of tube R19C83 had a PI call from l the bobbin data, but not the pancake coil data. The FDB and TSP-4 regions of l R19C83 were retained as archive samples and were not destructively examined.

l j 3.1.2. Nondestructive Examinations l

l The tube sections were inspected in the laboratory by eddy current using techniques similar to those used during the field inspection. The tubra had been cut in the field into lengths between 23 and 34 inches long to allow the tube segments to be removed from the steam generator and to fit into a shipping I container. Specifically four sections of each tube were inspected using a 0.610 l

inch diameter differential bobbin coil probe, and a Zetec + Point probe. The inspected tube sections were associated with the TSP-1, TSP-2, TSP-3 and TSP-4 locations. The data were collected using a R/D Tech TC 6700 and recorded on  ;

Optical disks. Analyses of the data were conducted using the Westinghouse Anser system.

l A review of the field eddy current data for the removed tubes showed essentially i no difference between the original field calls and the review of the data. Table 3-1 summarizes the eddy current results for the areas of interest. Note that the

+ Point results are for the 300kHz channel and that the bobbin coil results are for the 550/130 kHz MIX channel. The table shows the presence of a large indication at the TSP-3 location of Tube R18C100 and at the TSP-2 location of Tube R19C83 as noted by both bobbin and + Point data. A smaller indication was noted by both bobbin and tPoint data at the TSP-2 locr. tion of Tube R18C100. At '

the TSP 3 location of Tube R19C83, only a bobbin indication was noted.

The laboratory eddy current data showed the same indications as in the field data with a small increase in eddy current voltage for the large indication at Q \apc\thx tthx98\90 day \thxc690d.dec j 3-1

l TSP 3 of Tube R18C100 and a large increase in eddy current voltage for the TSP-

2 of Tube R19C83 over that of the field results. This increase suggests the possibility that noncorroded ligaments present between individual corrosion microcracks had separated during the tube removal for these two locations. The e

bobbin coil indication identified at the TSP-3 location of Tube R19C83 was a distorted indication that could be a tube deposit response rather than a response i

from corrosion degradation. However, the guidelines for NDE analyses to support voltage-based repair criteria recommend calling this type ofindication as

a potential indication for RPC inspection, which is consistent with the field call.

F 3.1.3. Leak, Burst and Tensile Data

Following NDE testing, elevated temperature leak testing was performed on the two large voltage indicationn (TSP-3 of Tube R18C100 and TSP-2 of Tube R19C83). Both specimen developed leaks at all of the tested differential pressures designed to si tulate conditions ranging from normal operating conditions (NOC) to steam line break-(SLB) conditions. Actual differential pressures ranged from a minimum of 1325 psi to 2581 psi. Table 3-2 provides a summary of test conditions and measured leak rates for the various test conditions. The TSP-3 region of Tube R18C100 had a smaller leak rate at NOC than did the TSP-2 region of Tube R19C83. However, exposure to higher differential pressures apparently caused ductile ligament rupture to the extent that subsequent leak rates were higher for the TSP-3 region of Tube R18C100 than for the TSP-2 region of Tube R19C83. Later SEM fractographic data showed that the crack network for the TSP-3 region of Tube R18C100 was very complex with two or three throughwall cracks' that probably interconnected during leak testing. Measured leak rates ranged from 0.000145 gpm at NOC to 0.033 gpm at SLB conditions for TSP-3 of Tube R180100 and from 0.000792 gp

n at NOC to 0.0185 gpm at SLB conditions for TSP-2 of Tube R19C83.

Following leak testing, sections of both tubes were burst tested at room temperature. The TSP 1, TSP 2, TSP-3 and TSP-4 locations of Tube R18C100 and the TSP 2 and TSP-3 locations of Tube R19C83 were burst tested along with a control free span (FS) section of each tube without NDE indications. Table 3 3 l presents a summary of the burst data. All burst pressures were well above safety

' limitations required by R.G.1.121 with the two large voltage indication locations I having the lowest burst pressures: 5,006 psi for TSP-3 of Tube R18C100 and 5,958 psi for TSP-2 of Tube R19C83. The TSP-2 region of Tube R16C100 was the only other burst specimen with an obviously reduced burst pressure: 7,196 psi burst pressure.

A visual examination performed on the burst tested specimens showed no corrosion on the single FDB region burst tested (TSP-1 of Tube R18C100), and Q.\apc\thz\thx98\90 day \thxc690d. doc 3-2

corrosion present on all true TSP regions (TSP-2, TSP-3 and TSP-4 of Tube R18C100 and TSP-2 and TSP-3 of Tube R19C83). The corrosion observed on the TSP regions was entirely conGned to the TSP crevice region. The TSP-4 region of Tube R18C100 had only field and laboratory NDD calls and a burst pressure and ductility similar to that of its FS control specimen. The TSP-3 region of Tube R19C83 had only a small bobbin PI and a burst pressure and ductility similar to its FS control specimen. It is judged that the corrosion present at these two locations was shallow, probably on the order of 10% deep. It was decided to destructively examine only the three specimens with reduced burst pressures:

the TSP-2 and TSP-3 region of Tube R18C100 and the TSP-2 region of Tube R19C83. Figures 3-1,3-2 and 3-3 provide sketches of the burst openings and of the secondary corrosion observed on the three specimens chosen for destructive examination. Note that the TSP-3 region of Tube R18C100 had a complex burst opening. It will be more completely described in the next section.

Finally, Table 3-3 includes room temperature tensile test data obtained on additional FS sections from both tubes. The tensile properties appear typical of MA Alloy 600 steam generator tubing of this vintage.

3.1.4. Destructive Examinations SEM fractography was performed on each of the burst opening fractures faces chosen for destructive examination: the TSP-2 and TSP-3 region of Tube R18C100 and the TSP-2 region of Tube R19C83. Table 3-4 presents a summary of the results in the form of crack depth profiles and ductile ligament data. The TSP-2 region of Tube R18C100 and the TSP 2 region of Tube R19C83 had simple axial burst openings. The TSP-3 region of Tube R180100 had a complex "H" shaped burst opening that formed from two close-by parallel axial corrosion macrocracks. These two cracks joined by a tearing through a region of intergranular corrosion that separated the two axial macrocracks. Table 3-4 provides the crack profiles for both of these two parallel axial corrosion macrocracks. The horizontal bar of the "H" shaped region was not characterized by fractography. However, the horizontal bar was measured as approximately 0.1 inch long, while the two parallel macrocracks were approximately 0.64 and 0.38 inch long.

Each of the four axial burst fracture faces (one each from TSP 2 of Tubes R18C100 and R19C83 and the two fracture faces from TSP 3 of Tube R18C100) had OD origin intergranular corrosion thet occurred as a macrocrack composed of a number of OD intergranular microcracks joined together by ligaments. Most of these ligaments had only or mostly intergranular features, indicating that these particular ligaments grew together during plant operation. Each of the four burst corrosion macrocracks also had ligaments with predominantly ductile Q.\apc\thx\thx98\90 day \thxcG90d doc 3-3

i features, indicating that these particular ligaments formed (tore) during either tube pulling, leak testing, burst testing, or subsequent laboratory handling. The two macrocracks from TSP-3 of R18C100 also had ductile ID lips that probably acted similar to ductile ligaments.

The largest corrosion macrocrack was the left hand crack of the "H" shaped crack network for the TSP-3 region of Tube R18C100. It was 0.637 inch long, averaged ,

68% deep, was 100% throughwall over 0.045 inch and 98% deep (ID tensile lip on fracture) over another 0.135 inch. The right-hand crack of the network was at l least 0.38 inch long (mechanical damage at the bottom of the crack prevented l exact length determination, but burst photographs suggested a crack length of 0.38 inch), averaging 79% deep over the 0.325 inch with fractographic data, was 100% throughwall over > 0.155 inch and 97% deep (ID lip on fracture) over another 0.032 inch. (Again, these two parallel axial macrocracks were joined l together during burst testing (possibly during leak testing) by a 0.1 inch long horizontal crack.)

The TSP-2 region of Tube R19C83 was 0.436 inch long, averaged 77% deep and was 100% throughwall over 0.2138 inch. The TSP-2 region of Tube R18C100 was 0.621 inch long, had a maximum depth of 93% throughwall and averaged 55%

throughwall.

Based on the appearance of the cracks examined by SEM, it is believed that the corrosion morphology was composed primarily of axial intergranular stress i corrosion cracking (IGSCC) with some intergranular cellular corrosion (ICC) also l present. ICC is a crack structure composed of a mixture of axial, circumferential and oblique angled IGSCC. It is further suggested that the OD intergranular corrosion present is typical of that in the EPRI data base gathered in support of l alternate plugging criteria.

1 1

3.1.5. Summary The true TSP crevice regions of Tubes R18C100 an'd R19C83 had OD intergranular corrosion. The FDB (TSP-1) regions did not have corrosion. The TSP-3 region of Tube R180100 and the TSP-2 region of Tube R19C83 had throughwall corrosion. Leak testing for these two specimens produced leak rates that ranged from a low of 0.000145 gpm at NOC (TSP-3 of R18C100) to a high of 0.033 gpm at SLB conditions (TSP-3 of R18C100). Burst testing showed that the corroded TSP regions all had strength properties exceeding regulatory guidelines.

Of the remaining three true TSP regions, one (TSP-4 of R19C83) was archived after NDE examination and the other two were burst tested without performing i destructive examinations. Visual inspection of these two TSP regions (TSP-4 of I R18C100 and TSP-3 of R19C83) showed that intergranular corrosion was Q.\apc\thx\thx98\90 day \thrc690d. doc 3-4

present. From their apparently unaffected burst properties, it is assumed that their corrosion depth was probably on the order of 10% throughwall. No field eddy current detection would be expected for corrosion of this depth. No degradation was called by eddy current for the TSP-4 region of R18C100, while a small voltage PI was called by bobbin inspection for the TSP-3 region of R19C83 (NDD field pancake coil and laboratory + Point call).

3.2 South Texas Unit-2 Pulled Tube Evaluation for Voltage-Based Repair Criteria Application The pulled tube examination results were evaluated for application to the EPRI database for ARC applications. The eddy current data were reviewed, including reevaluation of the field data, to finalize the voltages assigned to the indications

, and to assess the Seld NDD calls for detectability under laboratory conditions. The I

data for incorporation into the EPRI database were then defined and reviewed j against the EPRI outlier criteria to provide acceptability for the database.

i 3.2.1 Eddy Current Data Review l Table 3-5 provides a summary of the eddy current data evaluations for the South Texas Unit-2 pulled tubes. These NDE data results have been discussed in the above Section 3.1.2. As noted above, the field and laboratory reevaluations of the field bobbin data are in very good agreement for both voltage magnitudes and NDD calls. The reevaluated field bobbin voltages, including the adjustment for cross calibration of the field ASME standard to the laboratory standard, are used I

for the EPRI ARC database. The reevaluation was performed by the same analyst that performed a large part of the EPRI pulled tube database and the use of these voltages minimizes analyst variability in the database, which is separately accounted for in ARC applications as an NDE uncertainty.

The post-pull laboratory inspection results show a 30% increase in bobbin voltage for R18C100, TSP-3 and almost a factor of two increase in bobbin and + Point voltage for R19C83, TSP-2. These increases tend to indicate that some ligaments likely tore during the tube pulling operation. However, increases of these magnitudes in the bobbin voltage are not unusual and do not impact the use of the data for the ARC correlations.

3.2 2 South Texas Unit-2 Data for ARC Applications The pulled tube leak test, burst test and destructive examination results are summarized in Table 3-6. The leak rates in this table have been adjusted to the reference conditions using the EPRI leak rate adjustment procedure commonly applied for data in the ARC database. Both R18C100, TSP-3 and R19C83, TSP-2 i

Q.\apc\thx\thx98\90 day \thacG90d. doc 3-5

l were found to have modest leak rates at normal operating and SLB conditions.

leak rate data are given in the table for SLB pressure differentials of 2405 and l 2560 psi. The 2405 psi leak rates are applicable to South Texas-2 ARC analyses due to applicability of the PORVs for limiting the accident condition pressure l differential. Although the indication at TSP 2 of R19C100 was not leak tested, it can be inferred that this indication would not leak at SLB conditions. The maximum depth of the indication is 93% and more than 90% deep only over about l a 0.03" length. No ODSCC indications that have been leak tested in the ARC )

database at depths less than about 98% have been found to break through to l throughwall at SLB conditions with resulting leakage. The short length of the deeper part of the indication supports an insignificant likelihood of break through and leakage at SLB conditions. Even a throughwall indication of 0.03" length would not leak at SLB conditions. Therefore, this indication can be included as a non-leaker in the probability ofleakage correlation.

The South Texas Unit-2 pulled tube results were evaluated against the EPRI data exclusion criteria for potential exclusions from the database. Criteria la to le apply primarily to unacceptable voltage, burst or leak rate measurements and indications without leak test measurements. Criteria la to le are not applicable to the South Texas Unit 2 indications. Criterion 3 applies to potential errors in the leakage measurements and is not applicable to the South Texas Unit-2 indications since there are no known errors in the measurements and the leak rates are not low relative to leak rate correlations.

EPRI Criterion 2a applies to atypical ligament morphology for indications having high burst pressures relative to the burst / voltage correlation and states that high burst pressure indications with 5; 2 uncorroded ligaments in shallow cracks < 60%

deep shall be excluded from the database. Table 3-6 identifies the number of remaining ligaments and the maximum depths for the indications. The three indications dectructively examined have maximum depths > 60% and Criterion 2a is not applicable. However, the R19C83, TSP-3 indication does not have destructive exam data to define the crack profile and the presence or absence of ligaments cannot be determined. The burst pressure for this indication is high on the burst correlation and Criterion 2a could be applicable if the crack profile was available for assessment. Since the destructive exam profile is not available for this indication to permit evaluation against Criterion 2a, the indication is excluded j from the ARC database and is not used for either the burst pressure or probability l ofleak correlation.  !

As shown in the last column of Table 3-6, the TSP 2 indication of R18C100 is to be j included in the probability ofleakage and burst correlations. The indications at R18C100, TSP-3 and R19C83, TSP-2 are included in the burst, leakage and probability of leakage correlations. The R19C83, TSP-3 is excluded from the Q:\a pc\th x\t h x98\90 day \t h xc690d. doc 3-6

database per EPRI exclusion Criteria 2a due to lack of a destructive exam profile to assess the indication. The impact of the indications on the ARC correlations is ,

further discussed in Section 3.4. l 3.3 Comparison of South Texas Unit-2 Data with the EPRI Database This section reports on the evaluations performed utilizing the leak rate and burst pressure test data described in the previous section. The data obtained from the tests are compared to the reference EPRI database for nominal 3/4" by 0.043" SG tubes as identified in Reference 9-4. The NRC staff concurred that the recommended database was appropriate for use via Reference 9-5. The results of I the destructive examinations of the tube sections are delineated in the previous two sections (Sections 3.1 and 3.2). Those results revealed no information that would lead to a conclusion that the data should not be included in the database.

This section presents results from the evaluations carried out to examine the effects ofincluding the leak rate and burst pressure test results from the South Texas Unit 2 pulled tube specimens on the reference database probability ofleak, leak rate, and burst pressure correlations to the bobbin amplitude. In summary, the test data are consistent with the database relative to the probability ofleak, the leak rate, and the burst pressures as a function of the bobbin amplitude.

These comparisons and evaluations are discussed below. Furthermore, the resulting correlations based on including the data in the database should be considered to be applicable to the use of voltage-based repair criteria for indications in 3/4" diameter tubes in Westinghouse SGs.

3.3.1 Burst Pressure vs. Bobbin Amplitude The results from the burst tests, performed on tube specimens which exhibited a non-zero bobbin amplitude at a TSP elevation location, were considered for ,

evaluation. A plot of the burst pressures of the South Texas Unit-2 specimens is i depicted on Figuras 3-4 and 3-5 relative to the burst pressure correlation develored using the reference database.

1. A visual examination of the data relative to the EPRI database indicates ,

that the measured burst pressures fall within the scatter band of the i reference data. Figure 3-4 shows that the data fall well within a 95%

confidence, two sided tolerance bound for a 90% portion of the underlying population.

2. The data points fall relatively near the regression line and no statistical anomalies are indicated, i.e., the data are visually remote from the prediction and tolerance bounds. It is noted that three of the four data l

Q:\ ape \thx\thx98\90 day \thrc690d doc 3-7

l l

l points fall below the regression line, however, this is not of statistical significance.

In summary, the visual examination doesn't indicate any significant departures from the reference database.

l l Since the burst pressure data from the South Texas Unit-2 tube specimens were not indicated to be from a separate population from the reference data, the regression analysis of the burst pressure on the common logarithm of the bobbin l amplitude was repeated with the additional data included. A comparison of the I regression results obtained by including these data in the regression analysis is l provided in Table 3-7. Regression predictions obtained by including these data in the regression analysis are also shown on Figure 3-5. A summary of the changes .

is as follows: l

1) The intercept of the burst pressure, Ps, as a linear function of the common logarithm of the bobbin amplitude regression line is decreased by 0.3%, or about 25 psi. Because of the logarithmic scale, the intercept corresponds to a bobbin amplitude of 1 V. The change has the effect of uniformly decreasing the predicted burst pressure as a function of the  :

bobbin amplitude by a minuscule amount.  !

l

2) The absolute slope of the regression line is increased by 0.1%, i.e., the i slope is more steep. This has the effect of decreasing, albeit very slightly, the burst pressure as a function of bobbin amplitude for large indications, i.e., for those indications with a log-amplitude greater than the mean log-amplitude (about 2.5 V).

l

3) There is no meaningful change in the standard error of the residuals.

The net effect of the changes on the SLB structural limit, using 95%/95% lower tolerance limit material properties, is to decrease it by 0.09 V, i.e., from 4.79 to 4.70 V. For a SLB differential pressure of 2405 psi, the structural limit decreases from 5.80 to 5.69 V. This results from the increase in the slope coupled with no meaningful change in the intercept and standard error. When coupled with the fact that the structural limit is decreased indicates that the probability of burst (PoB) would decrease for bobbin indications with amplitudes less than about 1 V (the logarithm is zero and it is near the centroid of the logarithm of the volts),

would increase for indications greater than about 1 V up to the upper bound of l the structural range of interest. Based on the relatively small change in the structurallimit, the change in the PoB would also be expected to be small. The l effects of the changes on the PoB are illustrated on Figure 3-6. As expected the PoB is decreased up to about 1 V and increases slightly for indications with Q \apc\thx\thx98\90 day \thxc690d. doc 3-8

1 larger amplitudes.

3.3.2 Probability of Leak Correlation The data of Table 3-6 were examined relative to the reference correlation for the probability of leak (pol) as a function of the common logarithm of the bobbin amplitude. Figure 3-7 illustrates the South Texas Unit-2 data relative to the reference correlation. One of the specimens exhibited expected pol behavior,i.e.,

the indication had a calculated low probability of leak and did not leak. The other two indications had estimated probabilities of leak of about 0.2 and 0.6.

Both of these indications leaked at the SLB differential pressure. These results are not statistically different from the expectation. Had the test results been significantly different from the expectation, statistically anomalous behavior might have been suspected. Thus, based on the data examination, there is no significant evidence ofirregular results, i.e., outlying behavior is not indicated.

In order to assess the quantitative effect of the new data on the correlation curve, the database was expanded to include the South Texas Unit-2 data points and a Generalized Linear Model regression of the pol on the common logarithm of the bobbin amplitude was repeated. A comparison of the correlation parameters with those for the reference database is shown in Table 3-8. These results indicate:

1) A 8.0% increase (smaller absolute of a negative value) in the logistic intercept parameter.

2) A 4.3% decrease in the logistic slope parameter.

3) The absolute values of the variance and covariance of the parameters changed by 15% to 22%. Examination of Figure 3-7 indicates that it is likely that the probability of leak of all indications in the range of interest is increased by the inclusion of the South Texas Unit-2 data.

However, the pol equation generally has a small effect on the total estimated leak rate and it would be expected that there would be no significant impact on the 95% confidence bound on the total estimated leak rate from a single SG.

4) The deviance of the regression increased by 9.8%. An increase is expected when additional data is added. The Pearson standard error decreased by almost 14%, i.e., from 1.12 to 0.97, indicating an improvement in the models predictions.

In order to examine the changes to the pol, the reference correlation and the new correlation were also plotted on Figure 3-7. An examination of the figure Q. \ a pc \ thx \ thx98 \ 90d ay \ thxc690d. doc 3-9

1

! indicates an increase in the pol on the order of a few percent for all indications I over about i volt. It is noted that when the total leak rate is determined using l the leak rate to bobbin volts correlation, the resulting value can be quite '

insensitive to the form of the pol function. So, the effect of the changes in the parameter values and variances would be expected to be small relative to the calculation of the 95% confidence bound of the total leak rate from a SG. j 3.3.3 SLB Leak Rate Versus Bobbin Amplitude Correlation As previously noted, two of the removed tube specimens exhibited leakage under l SLB conditions. The leak rates, described earlier in Section 3.1, are depicted on l Figure 3-8 relative to the correlation obtained using the reference database. The

! two data points from South Texas Unit-2 exceed the reference correlation curve for the median leak rate. It is implied from the visual examination, using the relative distance from the 95% confidence bound on the arithmetic average to the arithmetic average, that the data would fall well within a 90% non-simultaneous, two-sided prediction band. Thus, the visual appearance of the data indicates strong support for the trend of the prior correlation. A summary of the l parameters of the correlations is provided in Table 3-9. The following changes resulted from the addition of the new data:

1) The intercept of the correlation curve increased (smaller absolute of a i negative value) by 14%.

2) The slope of the correlation curve decreased by 7.6%.

The standard deviation of the common logarithm of the leak rate l 3) residual errors increased by 2.5%.

4) The p value for the slope coefficient increased by a factor of 2.7 to 210-11 This change is of no statistical significance.

The net effect ofincluding the additional data is to slightly increase the expected, i.e., arithmetic average, leak rate for bobbin amplitudes over most of the range of the data. In practice, the change to the estimated leak rates would be expected to be not significant rdative to allowable values.

It is important to note that the pressurizer power operated relief valves (PORVs) at South Texas Unit-2 may be considered to be available and operable during a postulated SLB event. Therefore, the predicted leak rates at a differential pressure of 2405 psi (based on a PORV set-point of 2335 psi plus 3% for accumulation) are applicable for the calculation of total leak rates. The parameters of the regression fit of the leak rate data for a differential pressure of Q: \ ape \ thx \ thx98 \ 90 day \ thme690d. doc 3-10

- .- .- . - . . - . _ - _ _ - - - .= - - _ - _- ._.__ .. - -

2405 psi are listed in Table 3-10, and the correlation of the leak rate to the l

logarithm of the bobbin amplitude is illustrated on Figure 3-9.

1 3.3.4 Conclusions Relative to ODSCC Correlations The review of the effect of the South Texas Unit-2 data indicates that the l correlations of the burst pressure, the probability ofleak, and the leak rate to the common logarithm of the bobbin amplitude would not be substantially changed by the inclusion of the data. Although, both the pol and expected leak rates would increase, the effect on total predicted leak rates can be expected to be relatively small. It is judged to be likely that the conclusions relative to EOC probability of burst and EOC total leak rate based on the use of the reference database would not be significantly changed relative to results obtained from corrclations developed after adding the South Texas Unit-2 data to the database. j i

l i

l Q: \ ape \ ths\ thx98 \ 90 day \ thxc690d. doc 3-11

1 l

Table 3-1. Summary of NDE Data on South Texas Unit 2 S/G Tubes Field Eddy Current Field Data Review Lab Eddy Current  !

Location Bobbin 0.080" Bobbin + Point (300 Bobbin + Point (300 Coil Pancake Coil kHz) Coil kHz) l (600 kHz) volts /%TW volts /%TW volts /%TW volts /%TW/ volts /%T volts /%TW/

length (") W length (")

R18C100, NDD NDD NDD NDD NDD NDD I TSP-1 R18C100, 1.25/PI 0.56/SAI 1.21/PI 0.66/<20/0.5 1.1/PI .28/72/0.69 TSP-2 i R18C100, 4.03/PI 3.72/SAI 4.08/83 2.2/81/0.50 5.2/75 2.5/93/0.7 l

TSP-3 0.13/<20/.2 R18C100, NDD NDD NDD NDD NDD NDD i TSP-4 R19C83, NDD NDD NDD NDD NDD NDD  !

TSP-1 j R19C83, 2.76/PI 2.78/SAI 2.77/92 2.02/83/0.45 5.2/82 3.99/95/0.61 TSP-2 R19C83, .24/PI NDD .23/70 N/A NDD NDD TSP-3 R19C83, NDD NDD NDD NDD NDD NDD TSP-4 NDD - No Detectable Degradation SAI- Single AxialIndication DI - Distorted Indication N/A - Not Appropriate PI- Possible Indication l

I i

Q Aa pc\thx\thx98\90 day \thxc690d. doc

! 3-12 t

I. .

i l

I I 1 Table 3-2. South Texas Unit 2 Leak Test Data  !

Specimen Test Type

Leak Rate Test Conditions  ;

l Differential (liters /hr & gpm) P,(psig) P,(psig) T,('F) T,('F) I Pressure (psi)

R18C100 NOC: 1344 0.033/0.000145 2330 986 618 618 TSP-3 ITC: 1917 0.585/0.00258 2438 521 598 612 SLBl: 2512 7.50/0.0330 2720 208 587 520 l R19C83 NOC: 1325 0.180/0.000792 2277 952 604 612 TSP-2 ITC: 1974 0.248/0.00109 2502 528 596 607 j SLBl*: 2527 4.20/0.0185 2734 207 580 479 i SLB2*: 2581 3.70/0.0163 2789 208 576 544 NOC = normal operating conditions; ITC = intermediate test conditions; SLB = steam line break. )

I

• The leak rate for the SLB1 test of Tube R19C83 TSP-2 was observed to decrease during the second half of the test. This test was terminated and test SLB2 was then run at similar conditions. The leak rate was 13% lower than for the SLB1 test, even though the differential pressure was slightly higher. Differing leak rates for the same test conditions occur relatively infrequently and usually for low to moderate leak rates (tighter cracks); but when differing rates happen, they occur such that lower leak rates are experienced with increasing time. This observation is probably related to ID crud particles becoming trapped in the crack and partially sealing the leak. This hypothesis is supported further by previous test observations where instances ofjarring the specimen caused a restoration of higher leak rates. Note that all other leak tests in this series of tests appeared to have a constant leak rate.

l l

Q Aa pc\th x\thx98\90 day \th xc690d. doc 3 13

i Table 3-3. Room Temperature Burst and Tensile Test Data for South Texas Unit 2 S/G Tubes t 2

Location Burst Burst Burst Burst 0.2% Offset Tensile Tensile t Pressure, Ductility, Length, Width, Tensile Yield Ultimate Elongation, ,

psig  % inches inches Strength, psi Strength, psi  %

R18C100,FS 11,000 33.7 1.414 0.321 53,700 101,600 44.2 R18C100, TSP-1 11,100 35.0 1.502 0.319 RI 8C100, TSP-2 ,

Run #1** 6,373 2.5 Mult. Cracks 0.004 Max

Run #2" 7,196 11.1 0.826 0.225 R18C100, TSP-3* 5,006 15.0 0.552 0.272 R18C100, TSP-4 11,000 35.5 1.305 0.379 R19C83,FS 11,800 34.5 1.630 0.391 57,400 106,300 37.5 .

l R19C83, TSP-2* 5,958 9.7 0.718 0.230 R19C83, TSP-3 11,700 39.9 1.5772 0.434 TSP = tube support plate; FS = free span; S/G = steam generator ,

• Tested with foils, bladders and extensions.  ;

• • Initially tested without a foil and bladder. No significant burst opening or tearing at the burst opening tips was observed. As a consequence, the specimen was re-tested using a foil and bladder to obtain a wider burst opening with I ductile tears at the crack tips.

• • • Burst opening occurred away from TTS location near a Swagelok fitting which reduced the ductility and burst opening i dimensions. No corrosion was observed on burst fracture face. The burst near the fitting indicates that no corrosion >

was present at or near TTS. ,

i Q-Napc\thx\thx98\tsparc\thxcG90d. doc 3-14

Table 3-4. SEM Fractographic Data for OD Intergranular Macrocracks on S. Texas Unit 2 Tubes Length vs. Depth * & Ligament Positional and Ductile Ligament Data Specimen, d Location Location (inches /% throughwall) (Area == inches: x 10 ; Orientation of Ligament Minor Axis relative to Macrocrack Major Axis in degrees; Orientation of Ligament Major Axis relative to Tube Radius in degrees **)

i R18C100, TSP.2, 0.000/00 Macrocrack bottom 0.049' above TSP bottom @ 180=

barst fracture at 0.0485/49 O.097/41& Ugament I Ligament 1: Area = 3.4; Minor Axis @ 90'; Major Axis @ O' 180' l 0.1455/70 0.194/78 0.2425/81+ Ugament 2 , g g, 0.291/936 Ligament 3 Ligament 3: Area = 2.1: Minor Axis @ 90s; Major Axis @ O' Ligament 4: Area = 1.3; Minor Axis @ 908; Major Axis @ O' 0.3395/81* L'gament 4 O.388/70 0.4365/55 0.485/49 ,

0.5335/14 0.582/29 Macrocrack top located 0.670" above TSP bottom @l80*

Ma rack LAD * = 55%, Maximum l

Depth = 93%, Macrocrack length = &

0.621")

• Average depths may be calculated by a number of different methods depending upon the need. Methods used are LAD = linear average depth; ATD = average throughwall depth (length weighted average depth); PDA = percent degraded area (relative to cross sectional area of an nondegraded tube).

" Note that the ductile ligaments in the table are described by both a major and a minor axis orientation. The ligaments are usually considerably deeper / longer (major axis) than wide (minor axis). He ligament major axis is that going from the OD to the ID of the tube wall (or from the to the OD in the case ofID origin cracks) and is usually close in orientation to the radius of the tube. He orientation of the major axis is relative to the tube radius. He minor axis of the ligament is the observed orientation where the ligamentjumps from one microcrack to another microcrack as viewed from the OD. He orientation of the minor axis is relative to the tubing major axis. Usually the minor axis is close to perpendicular to the tube major axis.  !

[

I i

f Q:Na pe\th x\thx98\ tsp arc \thrc690d. doc 3-15

Table 3-4 (Continued). SEM Fractographic Data for OD int-rgranular Macrocracks on S. Texas Unit 2 Tubes Specimen, Length vs. Depth * & Ligament Positional and Ductile Ligament Data Location Location (inches /% (Area = inches x 10"; Orientation of Ligament Minor Axis relative to throughwall) Macrocrack Major Axis in degrees; Orientation of Ligament Major Axis relative to Tube Radius in degrees **)

R18C100, TSP-3, 0.00/00 Macrocrack top located 0.735' above TSP bottom @ 295*

left-hand 0.025/41 macrocrack of"H" 0.05/39 shaped burst 0.075/55 opening, overall- 0.10/64 burst fracture 0.125/68 centered at 310' O.15/76 0.175/73 0.20/88 (0.21/98) + Ugament 1 Ligament 1: Area = 0.8; Minor Axis @ 90*: Major Axis @ 0*

0.225/984- ID Lip #1: 98% TW for 0.04" Ligament 2: Area = 1.3; Minor Axis @ 90*; Major Axis @ O' 0.25/100 + usament 2 O.275/100(- Max Depth =100% for 0.045" (0.295/100) 0.30/98(- ID Lip #2: 98% TW for 0.095*

0.325/98 0.35/98 + Ugament 3 Ligament 3: Area = 0.6; Minor Axis @ 90*: Major Axis @ 10*

(0.390/98) 0.40/82 0.425/77 Ligament 4: Area = 4.8; Minor Axis @ 90*; Major Axis @ Da 0.45/73 + Ugament 4 Ligament 5: Area = 0.4; Minor Axis @ 90*; Major Axis @ 0*

0.475/64 + Ugament 5 0.50/60 0.525/49 0.55/49 5 8 + Wama Ligament 6: Area = 0.9; Minor Axis @ 90*; Major Axis @ O' O.625/32* Usament 7 (0.637/00)

(Macrocrack LAD

• 68%, Maximum Macrocrack bottom located 0.098" above TSP bottom @ 295*

Depth = 100% over 0.045",98% deep over another 0.135", Macrocrack length 0.637")

Q:\ a pe\ th x\th x98\tspa rc\ thxcG90d. doc 3-16

Table 3-4 (Continued). SEM Fractographic Data for OD Intergranular Macrocracks on S. Texas Unit 2 Tubes Specimen, Length vs. Depth * & Ligament Positional and Ductile Ligament Data Location Location (inches /*/* (Area = inches: x 10d ; Orientation of Ligament Minor Axis relative to throughwall) Macrocrack Major Axis in degrees; Orientation of Ligament Major Axis relative to Tube Radius in degrees **)

R18C100. TSP-3, 0.00/00 Macrocrack top located 0.601" above TSP bottom @ 325' right-hand 0.025/31 macrocrack of"H" 0.05/60 shaped burst 0.075/62 opening, overall. 0.10/78 burst fracture 1.125/81 centered near 310* (0.138/97) + liganent I Ligament 1: Area = 1.6; Minor Axis @ 90*; Major Axis @ 10' O.150/97 (0.170/100) 0.175/100 0.20/100 0.225/100 0.250/100 0.275/100 0.300/100 0.325/100 ND = no data, mechanical damage prevented a depth determination 0.350/ND 0.375/ND Macrocrack bottom located 2 0.221" above TSP bottom @ 325', based on burst (20.38/ND) macrophotographs (LAD

• over top 0.325* = 79%,

Maximum Depth = 100% over 2 0.155",97% deep over another 0.032", Macrocrack length > 0.38")

Q.\spc\thx\thx98\tsparc\thxc690d. doc 3-17

Table 3-4 (Continued). SEM Fractographic Data for OD Intergranular Macrocracks on S. Texas Unit 2 Tubes Specimen. Length vs. Depth * & Ligament Positional and Ductile Ligament Data Location Location (inches /% (Area = inches x 10d ; Orientation of Ligament Minor Axis relative to 2

throughwall) Macrocrack Major Axis in degrees; Orientation of Ligament Major Axis relative to Tube Radius in degrees **)

R19C83, TSP-2, 0.00/00 Macrocrack bottom located 0.151" above TSP bottom @ 250' burst fracture at 0.0455/49 250* (0.090/61)

• Ligament 1 Ligament 1: Area = 2.8; Minor Axis @ 90*; Major Axis @ 20-0.091/100 0.1365/100 0.182/100 0.2275/100 UgameM 1 ha = Q hoW4 m % % @ &

0.273/100+ Ligament 2 (0.3048/100) 0.3185/92 0.364/72 '

0.4095/56 Macrocrack top located 0.587" above TSP bottom @ 250' (0.436/00) '

(Macrocrack LAD * = 77%,

Maximum Depth = 100% over 0.2138", Macrocrack length =

0.436")

I l

Q.\a pckth x\th x98\tspa rc\thxc690d. doc 3-18

Table 3-5. Summary ofSouth Texas-11998 Pulled Tube Eddy Current Results Field Call Lab. Reevaluation of Post Pull Data Tube TSP Field Data Bobbin 80 mil Bobbin Depth + Point Bobbin + Point Volts"' Pancake Volts"8 Volts Volts Volts Volts 1 NDD NDD NDD NDD NDD R18C100 FDB 2 1.25 0.56 1.21 PI 0.66 1.10 0.28 SAI SAI SAI 3 4.03 3.722 4.08 83 % 2.20 5.20 2.50 SAI SAI 0.13 MAI 4 NDD NDD NDD NDD NDD NDD R19C83 1 NDD NDD NDD NDD NDD NDD FDB 2 2.76 2.78 2.77 92% 2.02 5.20 3.99 SAI SAI SAI 3 0.24 NDD 0.23 30% N/A NDD NDD Notes:

1. Field and laboratory data include cross calibration of ASME standard to the reference

_ jaboratory standard.

Q:\a pc\thx\th x98\tspa rc\th xc690d. doc 3-19

i Table 3-6. South Texas-21998 Pulled Tube Data for ARC Applications T Bobbin Data Destructive Examination Results Leak Rate-/hr Burst Pressure Data - ksi Use in Tube S + Poi Corr.

P nt Max. Avg. Crack No. N. O. SLB Mess. s, e, Adj.* nom m Volts Dept Depth Depth Length Lig.* 1300 2405 Burst Burst Volt h , inch Psid psid Press. Press.

FDB NDD NDD 0% 11.100 10.231 None R18C100 2 1.21 P1 0.66 93 % 55 % 0.621 4 0.0* 0.0* 7.196 6.632 B, 290% deep POL for = 0.03" 3 4.08 83 % 2.20 100 % Notes 6, =0.38* 4 1 0.0265 3.26 5.006 4.614 B, L, 7 >0.155 TW 7.24m POL 100 % 68 % 0.637 7 0.045 TW 0.135 @ 98 %

4 NDD NDD = 10% Not destructively examined i1.000 10.139 None FS I1.000 53.7 101.6 10.139 FDB NDD NDD None R19C83 2 2.77 92 % 2.02 100 % 77 % 0.436 2 0.147 1.59 5.958 5.209 B, L, 0.214 TW 3.68* POL 3 0.23 30% N/A = 10% Not destructively examined 11.700 10.231 None FS l l 11.800 57.4 106.3 10.318 Notes:

1. FS is freespan section of tubing with no tube degradation to obtain tensile properties and undegraded tubing burst pressure.

2. Number of uncorroded ligaments with > 50% ofligament length remaining in burst crack face.

3. Burst pressures adjusted to 71.57 ksi, average flow stress at 650a F for 3/4" diameter tubes.

4. B = data to be used in burst correlation, POL = data to be used in probability ofleakage correlation, L = data to be used in leak rate correlation.

5. SLB leak rate at 2560 psid.

6. Mechanical damage to specimen prevented fractography and depth measurements over fulllength of the crack.

7. Burst opening had complex H shape formed from two closely spaced cracks that joined by tearing of the corroded region (= 0.17 between the two cracks.

8. Judged to be a non-leaker based on short length between 90% and 93% depth with uncorroded ligaments remaining in the deep section. This indication would not be expected to tear throughwall under SLB conditions.

Q Aspc\ thx\thx98\tsparc\thxcG90d. doc 3-20 1

Table 3-7: Effect of South Texas Unit-2 Data on the  ;

Burst Pressure vs. Bobbin Amplitude Correlation Pg = a o +a, log (Volts)

Parameter Addendum 2 l Database with l New / Old l Database l South Texas 2 l Ratio ll ao 7.42817 7.40278 l 0.997 l l

at -2.91207 l -2.91382 l 1.001 l r2 82.27 %

l 81.88% l 0.995 l o ctror 0.86118 0.86077 1.000 l l l l Mean log (V) 0.408302 l 0.407375 l 0.998 ll SS log (V) 36.91777 37.06576 1.004 N (data pairs) 93 96 Str. Limit (2560 psi) 4.79 V 4.70 V 0.981 Str. Immit (2405 psi) 5.80 V l 5.69 V l 0.981 l 0.210 I l p Value for a2 3.010-3G 6.2 10-37 l

Q.\ ape \thx\thx98\taparc\thxc690d. doc 3-21

i l

Table 3-8: Effect of South Texas 2 Data on the Probability of Leak Correlation 1

Pr(Leak) = -[beb log (Volts))

1 y 2 Parameter Addendum 2 Database with New / Old Database , South Texas 2 Ratio bi -5.2246 -4.8082 l 0.920 l b2 8.8034 8.4215 l 0.957 l Vii(1) 1.4990 1.1712 l 0.781 l Viz -2.1391 -1.7218 l 0.805 l V22 3.4198 2.8917 0.846 Number of Data 120 123 Deviance 41.75 45.90 MSE 0.354 0.379 1.071 Pearson SD 1.124 0.970 0.863 l Notes: (1) Parameters Ve are elements of the covariance matrix of the coefficients, bi, of the regression equation.

i Q Aapc\thx\thx08\tsparc\thxcG90d. doc 3-22 l

l l

1 l

I 1 \

l l

Table 3-9: Effect of South Texas 2 Data on the Leak Rate vs. Bobbin Amplitude Correlation (2560 psi)

_ g[h +h3 log 4 (Volts)]

Parameter Addendum 2 Database with New / Old Database South Texas 2 Ratio ba -1.90061 -1.63838 l 0.862 b4 3.18325 2.94093 l 0.924 r2 64.8 % 61.6% l 0.951 ozrror(br,) 0.59132 0.60638 1.025 Mean log (V) 0.938156 0.921007 j '

SS log (V) 2.795994 3.134826 i N(data pairs) 46 48 p Value for 62 7.71012 2.11011 2.65 l

l 4

Q.\spc\thx\thx98\tsparc\thxc690d. doc 3 23

l l

I Table 3-10: Effect of South Texas 2 Data on the Leak Rate vs. Bobbin Amplitude Correlation (2405 psi)

= 1 0 I63+6 8 0' "")l Parameter Addendum 2 Database with New / Old Database South Texas 2 Ratio b3 -2.087379 -1.870836 l 0.896  ;

b4 3.176887 2.976689 l 0.937 l r2 64.7% 62.8% l 0.971 l orrmr(bs) 0.59169 0.597912 1.011 Mean log (V) 0.938156 0.921007 SS log (V) 2.795994 3.134826 N (data pairs) 46 48 p Value for b2 8.31012 9 o.1012 1.15 Q:\ ape \thx\thx98\ttpsrc\thxc690d. doc 3-24

I l

l r' '

I I

i 0.75 TSP Top l l l (S

l

-i g

Dictance i From TSP <

Bottom /

- (Inches) l l N i l

T i 0.0 TSP Bottom i

i i I 00 9 00 1800 2700 3600 Circumferential Orientation (Degrees) i Figure 3-1 Sketch of OD crack distribution observed on the TSP 2 region of Tube

R18C100. The burst opening also had OD crack features on its fracture face i

that were confined to the TSP crevice region.

i Q.\apc\thx\thx98\tsparc\thxc690d. doc 3 25

0.75 TSP Top II h

l 1

- Distance l From lI ,

I TSP 3 Bottom

}

. (Inches) jl

\

l I ll U

0.0 TSP Bottom i

i l l Oa- 90a 180a 270a 360a Circumferential Orientation (Degrees)

Figure 3-2 Sketch of OD crack distribution observed on the TSP 3 region of Tube R18C100. The burst opening also had OD crack features on its fracture face that were con 5ned to the TSP crevice region.

Q:\apc\thx\thx98\tsparc\thsc690d. doc 3-26

! l l

l 0.75 TSP Top Di:tance From h

TSP Q r Bottom (Inches) (l 0.0 TSP Bottom i

l i I 0o 9 00 1800 2700 3600 Circumferential Orientation (Degrees) l Figure 3-3 Sketch of OD crack distribution observed on the TSP 2 region of Tube R19C83.

The burst opening also had OD crack features on its fracture face that were confined to the TSP crevice region.

l l

Q:\aje\thx\thx98\taparc\thxc690d. doc 3-27 l

> ' 1  ;! ;e i . ,

0 0 .

,.  ! l i '

_. 0 .

1

,!q  ! lr! I a I t i  ! ,i ,!

t a I_ -

,'ij D i,;

s , pI? -

s - .

,*'p d de  !

T NI i n d s _ u d

. s'l g' 2 o BA k _. .

,p 5

6 t a m u c C e 2 l

.bl N ,lb_ .

5 _. Dadn nA

_ e ar t N _.

1 t s d e na  % o e7 ,,4 TATP ed ol l

Iso s c=

l

,Ia _

nf .

~ o o s aS o _ s - .

r

, _ oa ~

leF ~

~ _

2 h ;N 's"0 o _

_ 0 To *0 N p 0 _

1 5 4' In d@

e

,6

.[_.

l~

,l

{

!o

) ,-i f

IE N 3

,. )

t s

s l ue t s

' l

.I'o ' D,

,c ?9 V

o -

i l a . y Ii .l l

>b o ' ,i B s

(

e pb 1.b{l h@ _

a , p  ! , ;I d mt

! l u

o 4-AaD t

_ o _

D i

i , Il (iI lI a

,, ;f l

p 3 _

n s i

e y

o-u "N .

_. A m e r 8 u 2-b b -

f Bh5 b n g 3 b u ,

l, R o U li4 li -

_ i i BoTG b F

_ ' u s _. b o

B sS v _ g'

- % os eA _

u -

" s 0

rM u .yl l:N , 'ii ,-li il!

1 s0 ' !i lyP I r lQ ,I s0  ; l .i  ! i t4  ; I e6 _.

,No " D r y ,

,  ;.! l Po l sl,l ,

iLl tl _ -

sA ,!;1 N  ? dl l .; !I r"

u3

.\

B40 l i D

0 -

o s- -

x N q_ l ,

l I lt-

}

c o

d.

" d -

4 s s 0 9

/ s " 6

_. c -

3 ,

x

_ s -

s h N;

t

- \

_ c r

_ ' I.

1 a 0 p s

0 0 0 0 0 0 0 t

\

2 0 8 6 4 2 0 8 1 1 9 x

h gu" 2 eum E5 c t

\

x h

t

\

e p

n

\:

Q

Burst Pressure vs Bobbin Amplitude 3/4" x 0.043" Alloy 600 MA SG Tubes Database @ 650*F, Sf = 71.565 ksi 12.0 '

Il i

1 i 'ii  !  ;  ; .

l

~

t

}' j  !

I a Test Data -! '

3 l I i x Addendum 1 I l t

I i

10.0 1 r

! i j i + 95% Pred @ Th --

l D !i  ! I a Plant AC2 Added Data I o i rj .,._

I ,  !

New Data Regression g ,

6 ol]0 g l

-- New 95% Pred @ LTL l j , ,  ! a l 3l ~

g 8.0 p ,

i al o i

• I q

.a

- 0 m Bo - n l

~ ~

h s~+]* s g

"% I o u f , oli o i

' a . .u  !

E g

60 i N.

i s 4,1 ou j i 4

• l

(

l I ' +,a a b h a m -+- u a i

'* a .

sg

$i 54 s $i goll i i % o I i

' i

-L L-.-.- .3 +$

_ . _ . 3.657 ksi - . L . 1 - . - . - - ------

ug }

1 g ao i

?+ v I I il a_o

_ . . 2.560 ksi 1. - . .-..- .. ----

I iL.! !.. _'f-%

I 2.0 I i .! .'*J

! l x'A m s

i i 4.71 V 12.0 V . '+ j i 'I I

!I 0.0 1.0 10.0 100.0 O.1 Bobbin Amplitude (Volts)

Figure 3-5 Q: \ a pc \ thx \ th x98 \ tspa rc \ thxc690d. doc 3-29

Probability of" Free Span" Burst vs, Bobbin Amplitude 3/4" OD x 0.043" Thick, Alloy 600 MA, SG Tubes @ 650 F 1.E+00 1.E-01

--- Prior Database at 2560 psi 1.E-02 -- New Database at 2560 psi g' 6

- -- New Database at 2405 psi g

1.E-03 ,f, a '

g 1.E-04 [*

$ //

% 1.E-05 [

3a 1.E-06 i -

/ ,

g 1.E-07 ,[ ,

f./

1.E-08 1.E-09 ['

,ip-1.E-10

/'

/

/

1.E-11 O.1 1.0 10.G 100.0 Bobbin Amplitude (VcIt.s)

Figum 3-6 Q:\ ape \thx\thx98\;tpatre\;':,xce90d doc 3-30

w-Probability of Leak for 3/4" SG Tubes @ 650*F, AP = 25G0 psi Comparison of New Data with Addendum 2 Reference Database l ! !!!

• l ! Il! Wt  ! ; ij ;I l

g,g ,_

o Addendum 2 Database i l l l j ll l l llPfl l ll lll m Plant AC-2 Data l l l l lll l l l l/'/ ll l i

llil l l // ll l --- L E Logistic Fit I I I l lll l l l l lltl l 0.8 -

Log Logistic w/ New Data  ! l !!l}  ! l // ll l }  ! ll m ll lI I I I /[ i l l N0.7-- ---- 90% Conf w/ New Data I ! Ill

@ l I i l I I I!l} l l l /// !ll l il lll

'S

! l ll l  ! l i !Il l l l !l/l ll! l  !! lll

$ o,e I I I I ll! l  !

i! I!! '

' !// !  !

! I l ll Eo.5  ! I I I !II I  ! II 'II I  !! // I I I  ! I II d I i '

l l i i ! !!!!  ! . '! ! l! I I l  ! ! !!!

ll ' il i l  ! iI!l l i l l  ! i !l b0.4  : ,1

=

r; lI l

_ - _- l l t

! I i

_ _.L_.t}i n _ ___. 1 _ a i

.8 o.3

. Il i l l Iil\ \ l

/!l,y\

t 8 \ \\  !

l lj j -

i I 0.2 II ll  ! !I I ef // l I l lll l .

I II i II I I/!l l I llI I o.1

! I ! ll!  ! l!  !!i - 'gI  !! l l ll'.' ,

0.0 I 'I Ill}

m l

-xxm 5r.

ii 1.Lf 9l % l lI!

j l-1If 0.01 0.1 1.0 10.0 100.0 Bobbin Amplitude (Volts)

Figure 3-7

4. \ a pe \ thx \ th x98 \ tsparc \ thxc690d. doc 3-31

0 0

L! -

-  ! l ,  !  ! i t-e

=E, l'l 8i 1

e 1I l i! l i!  ! , i t )

a ) k I  ! . l l ,t !l R A a ri

-1 k .(

AA AL e e n (r

! i!l .  !  ! .i z - a

' .' e et Lt aR a oP

/

l .  ! l , .i l  ;

t a

a geR e*

c I

s k n et ak

.e N .l l ;i I l ,

D r e eLiR a aede a l i

s [i -

a2 vL taAC A n ge onka f

}[

ep

• d0 i

ii  !  ; i il i D t 1 a aC . i r

e L iI u6 t 5 tseland d ed ev5% A i

2 - '

TPAMA9 A l

p= n o a+

/

1[

mP AA i I l

. f. c

.i t l i

i'

)

s nF

^

,7 n

t l

o i .

b*0 _# V

[s-

~

(

u -

b5 2 e

o6 a

/ a d a.,yk Jo n ' u B@ . _

0 i t

s s *;;'

E u~

1 l

p 8-veb  ! ' ii  !

@~e>n[ DG ..g'!l

\ .i ,  ;' l m 3

)

i u i '  ! i  !  !  ;

. + 1  !  ;, l' l A e r 2 -

sT -

,d[. . n u 3-p ,  !! . i  !

t '

g i '  ! l .

i g 3 0 G

= b  !

, eill b F

'M 6 S -  ! .  !  ! ,

g, $ .'I l ,' l o

50 '

z

' f B 20 -

', u/ #

h. l l

?  ! i i  ! ,i i l

( 6 ey talo l , r l  ;

n

i  !

,/

k

1. ,E.'

a . ,

l' l RlA .

4*f- : .

ka "3 ,l l . ii l l i i  !

~

a l ,l'l e4 L 00

_W I/ WiI '

Bx L

/ '

.#i iT

,T; S "4

, iI l . ,  ! l ,:i i ,J

!i

/.

/ '

3

~

- / j

1. /

c o

- d.

/:

d

- , 0 9

6

' c

. O x

-' ht e '

1 . 1 \

c r

1 1 a 0 0 0 0 1 0 p 0 0 0 1 0 t s

0 0 1 0 \

0 1 8 9

1 x ht m' g"@e$ 3oM yc)amam s c \

x h

t

\

c p

a

\.

Q

0 0 _

I: ~

i 1 -

!i i i!I  ! I I ,!i  ;

)

il

_' )

A k

- . 1i I II I l ,i i a

AA e

(

_-' I i I  !  !  !

AL ,!  !

e ge n r (

. i

~

, /

1xI aI a ra oP l  ! . i i t

- R e e* c

. I ,ii v n te ,!!

/ iI kAa e cd R e a

!v i

t aLi t f a n e nk t

,  !< I ,i  !!l  : ,!i 1

o a ep i

s .

/ .

D t

.aih mC L e s d t e

ei% A d5 fc.I ,

s>  !

e ,

TMA9r 5A

. i } .

I i II u0 '

t 4 ,

i l 2

/ '

a -

p= F /

/ -

AA mP 2 i  ! . i e

t.

x L

1  !

1

! l

, 1i

) -

' r n t s

i nF .

7- c w' _

l V

o b

b5 0 -

f u '/ ( .

._ e _

o6

.- o d 1 n u u B@

m '

/ -

gT n 0 i t

s s ,?-

- 1 l

p 9-veb  ; i

!o-.

W!l"lo 1 ;j r

i ti m 3

_ )

i u .i iI i  ! s l if4 t

,~ ' l!l 4; A e r 3 sT p

'- n u g

3-

, i i I 8 , l .4 II@ ,l o l ,i , t t i 3

5 G

.- b i

- f / b F 0 S ,I.'c a

, t iI , l -

. , " o o 40 r , B 20 ,

. / f9 f -7 it i

. i! I i f i . . t  ! I

( 6 ey f .

r '.,,

O yj t o 1 r -,:

al l  ; I ,j i I .

,i RA

=

s ka "3 4

.  !  ! i , i . ii I  ! +

f-i u

s eA e :a l e _b '

L0 0 s

/' ; .

Bx

. - /

L /

S "4 f i !i I jl I i l

. i' .

n(

A

/ ~

3 .

/

'- c

. . o

~

~

, s d.

d

- ' 0

' 9 6

' , c x

. , ht e 1 \

c r

1 1 a 0 0. 0 0. 1 0 p s

0 0 0 1 0 t 0 0 1 0 \

8 0 1 9 1 x

. ht mu$ gO . h e3Ma51 c mam \

x ht

\

ep s

\

Q

l l

4.0 EOC-6 Inspection Results nnd Voltage Growth Rates l

4.1 EOC-G Inspection Results l

l According to the guidance provided by the NRC Generic Letter 95-05, the EOC-6 inspection of the South Texas Unit-2 SGs consisted of a complete,100% eddy current (EC) bobbin probe full length examination of the tube bundles in all four SGs. A 0.610 inch diameter probe was used for all hot and cold leg TSPs where voltage-based repair criterion was applied. RPC examination was performed for all indications on the hot side with amplitude above 1 volt. Thirty-nine l indications on the hot leg side had a bobbin voltage above 1 volt. They were all l inspected with a RPC probe and 34 were confirmed as flaws and repaired. Only l one indication on the cold leg side had a bobbin voltage above 1 volt. This l indication was initially called as a wear indication and later reclassified as a I potential crack-like indication. It was not RPC inspected but was treated as a RPC-confirmed indication and repaired. This indication is further discussed in the paragraphs below.

Fifteen tubes in SG-D are excluded from voltage-based repair criteria as they are made of thermally treated tubes. As noted in Reference 9-2, tubes in the wedge regions are not excluded from the repair criteria as they are not expected to deform excessively under design-basis SLB conditions.

No RPC circumferential indications at the TSPs, no indications extending outside the TSPs, and no RPC indications with potential ID phase angles were found in this inspection. Also, no signal interference from copper deposits was found. A total of 47 TSP intersections in all 4 SGs with a mixed residual signal (MRI) that could potentially mask a 1.0 volt bobbin indication (MRI voltage 1.5 volts or greater) were inspected with a RPC probe and 4 of them were found to contain single axial indications (SAIs), and they were repaired.

A summary of EC indications for all four SGs is shown on Table 4-1, which tabulates the number of field bobbin indications, the number of those indications that were RPC inspected, the number of RPC confirmed indications, and the number ofindications removed from service due to tube repairs. The indications that remain active for Cycle 7 operation is the difference between the observed and the ones removed from service.

Overall, the combined data for all four SGs of South Texas Unit-2 show the following.

• A total of 1485 TSP indications identified during the inspection of which 39 indications on the hot leg side and one indication on the cold side were over 1 volt and 6 of these 39 hot leg indications were over 2 volts.

Q: \ a pe \ thx \ thx98 \ ts parc \ thxc690d . doc 4-1 l

l l

l i

• All 39 hot leg indications over 1 volt and 5 additional indications under 1  !

volt were inspected with a RPC probe,38 were confirmed. ,

All 34 RPC-confirmed indications over 1 volt (bobbin) were repaired. In addition 10 more indications were also removed from service because they l were present in tubes repaired for non-ODSCC causes. A potential ,

indication on the cold leg side (220) in tube RIC102 of SG-C with a bobbin voltage above 1 volt was not RPC inspected, but it was treated as a RPC- l confirmed indication and repaired. Another cold leg indication with voltage under 1 volt was not repaired. Consistent with the 1 volt repair criteria, )

indications with bobbin amplitude less than or equal 1.0 volt was not ,

considered for removal from service, regardless of RPC data, i

- A review of Table 4-1 indicates that SG-C had the highest number ofindications i returned to service for Cycle 7 operation (498 indications, none above 1.0 volt).

However, SG-A had 4 of the 5 indications over 1 volt returned to service (i.e., RPC NDD indications), and it also had larger average growth rate in the last cycle (Cycle 6). Therefore, SG-A may well be the limiting SG at EOC-7 from the standpoint of SLB leak rate and tube burst.

Figure 4-1 shows the actual bobbin voltage distribution determined from the EOC-6 EC inspection; Figure 4-2 shows the population distribution of those EOC-6 indications removed from service due to tube repairs; Figure 4-3 shows the distribution for indications returned to service for Cycle 7. Of the 44 indications removed from service, 34 indications are in tubes repaired because of the TSP voltage-based repair criteria. The rest are in tubes plugged for degradation ,

mechanisms other than ODSCC at TSPs. l The distribution of EOC-6 indications as a function of support plate location is summarized in Table 4-2 and plotted in Figure 4-4. The data show a strong predisposition of ODSCC to occur in the first few hot leg TSPs (1366 out of 1484 or ,

about 92% of the indications occurred at hot leg intersections in the first three TSP above the flow distribution baffle plate), although the mechanism extended to higher TSPs. Only one indication was initially called on the cold leg side (in SG-C).

Another bobbin signal at the first pre-heater baffle plate intersection on the cold side (220) in tube R1C102 in SG-C was initially called as a wear indication and was assigned 2.05 volts. A later reexamination of this bobbin signal indicated that it may be a potential crack-like signal and its voltage was revised to 1.23 volts. As i

a crack and potential ODSCC indication subject to GL 95-05 requirements, the j indication would be required to be RPC inspected. By the time this reassessment l was completed, equipment needed for RPC examination of this intersection had

! been removed from the steam generator. Based on discussions with the NRC, it Q: \ a pe \ thx \ thx98 \ tsparc \ thac690d. doc 4-2 i

was concluded that the RPC inspection could be omitted and this tube (R1C102) was repaired, which is equivalent to assuming RPC conSrmation of the indication as a crack-like flaw rather than wear. To ensure proper classification of cold leg indications in future inspections, all pre-heater baffle plate intersections on the cold leg side will be inspected with a RPC probe. Both bobbin and RPC data will be used to classify the indications as ODSCC or wear. Indications extending outside these bafIle plate intersections will also be RPC inspected. In summary, the distribution of indication population at TSPs in South Texas Unit-2s how predominant temperature dependence of ODSCC, similar to that observed at other plants.

A total of 73 dents with a bobbin voltage over 5 volts were found at TSPs in all 4 SGs combined. (Dents called within

• 0.5" from the TSP center line are considered to be within TSP.) All dented TSP intersections above 5 volts were inspected with j a RPC probe in this inspection, and no degradation was found at those locations. '

1 1

4.2 Voltage Growth Rates  !

1 I

For projection of leak rates and tube burst probabilities at the end of Cycle 7 '

operation, voltage growth rates were developed from EOC-6 (October 1998) inspection data and a reevaluation of the EOC-5 (May 1997) inspection EC signals for the same indications. Table 4-3 shows the cumulative probability distribution for growth rate in each South Texas Unit-2 steam generator during Cycle 6 (July l

'97 October '98) on an EFPY basis, along with the corresponding Cycle 5 growth rate distributions. Cycle 6 growth data are also plotted in Figure 4-5. The curve labelled ' cumulative' in Figure 4-5 represents composite growth data from all four l SGs. Cycle 5 growth rates were established using field bobbin data resized per the standard method for bobbin signal evaluation established for plants utilizing voltage-based repair criteria with an exception that history data for new EOC-5 l indications under 1 volt were not reevaluated and those indications were not included in the Cycle 5 growth data.

Average growth rates for each SG during Cycle 6 are summarized in Table 4-4. It is evident that the absolute magnitude of average growth in all SGs is relatively small (less than 0.2 volt). Among the four steam generators, SG-D had a larger average voltage growth during Cycle 6, but SG-A had 4 out of the 5 largest voltage growth during Cycle 6 (see Table 4-3). The average growth rates over the entire voltage range vary between 18% and 48.7% (of BOC voltage) per EFPY, between SGs, with an overall average of 27.1% per EFPY. The small average BOC voltages l (between 0.25 to 0.44 volt) leads to the relatively large percentage growth even l when the average growth (s; 0.122 volt per EFPY) is very small. The average

' growth for indications greater than or equal to 0.75 volt is 10.9% per EFPY and for indications less than 0.75 volt it is 28.8% per EFPY. A smaller growth for l

l Q: \ a pe \ thx \ thx98 \ ts parc \ thxc690d . doe 4-3

was conduded that the RPC inspection could be omitted and this tube (R10102) was repaired, which is equivalent to assuming RPC confirmation of the indication as a crack-like flaw rather than wear. To ensure proper dassification of cold leg indications in future inspections, all pre-heater baffle plate intersection indications on the cold leg side will be inspected with an RPC pmbe. Both bobbin and RPC data will be used to classify the indications as ODSCC or wear. Indications extending outside these baffle plate intersections will also be RPC inspected. In summary, the distribution ofindication population at TSPs in South Texas Unit-2s how predominant temperature dependence of ODSCC, similar to that observed at other planta.

A total of 73 dents with a bobbin voltage over 5 volts were found at TSPs in all 4 SGs combined. (Dents called within

• 0.5" from the TSP center line are considered to be within TSP.) All dented TSPintersections above 5 volta were inspected with a RPC probe in this inspection, and no degradation was found at those locations.

4.2 Voltage Growth Rates For projection of leak rates and tube burst probabilities at the end of Cycle 7 operation, voltage growth ' rates were developed from EOC-6 (October 1998) inspection data and a reevaluation of the EOC-5 (May 1997) inspection EC signals for the same indications. Table 4-3 shows the cumulative probability distribution for growth rate in each South Texas Unite 2 steam generator during Cycle 6 (July

'97 - October '98) on an EFPY basis, along with the corresponding Cyde 5 growth rate distributions. Cyde 6 growth data are also plotted in Figure 4-5. The curve labelled ' cumulative' in Figure 4-5 representa composite growth data from all four SGs. Cyde 5 growth rates were established using field bobbin data resized per the standard method for bobbin signal evaluation established for plants utilizing voltage-based repair criteria with an exception that history data for new EOC-5 indications under 1 volt were not reevaluated and those indications were not induded in the Cycle 5 growth data.

Average growth rates for each SG during Cyde 6 are summarized in Table 4-4. It is evident that the absolute magnitude of average growth in all SGs is relatively small (less than 0.2 volt). Among the four steam generators, SG-D had a larger average voltage growth during Cyde 6, but SG-A had 4 out of the 5 largest voltage growth during Cycle 6 (see Table 4-3). The average growth rates over the entire voltage range vary between 18% and 48.7% (of BOC voltage) per EFPY, between SGs, with an overall average of 27.1% per EFPY. The small average BOC voltages (between 0.25 to 0.44 volt) leads to the relatively large percentage growth even when the average growth (s; 0.122 volt per EFPY) is very small. The average growth for indications greater than or equal to 0.75 volt is 10.9% per EFPY and for indications less than 0.75 volt it is 28.8% per EFPY. A smaller growth for Q:\ npe\thx \thx98\ tsparc\thsc690d. doc 4-3 l

l'

indications 2 0.75 volt is not consistent with the data for other plants; however, as noted in the previous paragraph, Cycle 5 growth distribution does not include EOC-5 indications under 1 volt, which could exclude low growth indications and result in overestimation of average growth for the under 1 volt population. Also, since the number of indications with BOC voltage 2 0.75 volt (47 indications) is small in comparison with those below 0.75 volt (1437 indications) this growth trend may not be reliable.

Figure 4-6 is a plot of voltage growth during Cycle 6 vs. BOC-6 voltage. An examination of Figure 4-6 indicates that the Cycle 6 growth data for SG-A seem to show a dependency on BOC-6 voltage since essentially all large growth values (say, over 0.5 volt) occurred at BOC voltages greater than the mean BOC voltage (0.44 volt). However, SG-A also had many indications towards the high end of the BOC voltage spectrum with growth well below 0.5 volt. To examine the impact of the voltage-dependent growth trend observed for SG-A on tube integrity projections, SLB leak rate and tube burst probability projection for the EOC-7 condition for SG-A was carried using the methodolog.,y recommended in Reference 9-4 for considering growth dependency on BOC voltage, and the results are discussed in Section 8.0.

l Averaged composite voltage growth data from all four steam generators for the last two operating periods are summarized in Table 4-5. The guidelines in Generic Letter 95-05 require the use of more conservative growth rate distributions from the past two inspections for projecting EOC distributions for the next operating cycle. It is evident that the average growth rate /EFPY for Cycles 5 and 6 are comparable, with Cycle 5 having a slightly higher growth. However, as noted before, Cycle 5 growth distribution does not include EOC-5 indications under 1 volt, which could exclude low growth indications and its average growth may be overestimated. Furthermore, Cycle 6 data includes 3 growth values over 1 volt and a value over 2.1 volts while the growth rates for Cycle 5 are all equal to or less '

than 1 volt (see Table 4-3 and Figure 4-7 where cumulative probability distribution for the composite growth rate data from all SGs during Cycle 6 is compared with that for Cycle 5). Hence, SLB leak rate and tube burst probability projections for the EOC-7 condition based on the Cycle 6 data would yield more conservative msults; therefore, Cycle 6 growth distribution was applied to obtain EOC-7 projections.

From Table 4-3 and Figure 4-5 it is evident that the Cycle 6 growth rates at larger growth values (> 0.3 volt) for SG-A are higher than the composite growth distribution. Also, at lower growth values (below 0.6 volt) Cycle 6 growth rates for SG-D are higher than the composite growth rates. Per the methodology described in Reference 9-3, SG-specific growth rates are to be used for SGs A and D, while the composite growth rates should be applied for SGs B and C. The SG-specific Q \spc\thx\thx98\tsparc\thxc690d doc 4-4 l

growth data for SG-D do not include any of the top 4 growths observed for Cycle 6.

Since a few relatively high growth values observed during each cycle can be expected to occur randomly in any SG, it is not considered highly improbable that highest growth for the ongoing cycle would occur in SG-D. To account for such a possibility, the top 3 growth values for Cycle 6 were added to the SG-specific growth distribution applied to the EOC-7 projection for SG-D.

Table 4-6 lists the top 30 indications on the basis of Cycle 6 growth rates in ,

l descending order. Twenty-three of those indications were RPC confirmed and the I remaining 7 were either not inspected or no degradation was found (NDFs). All but one of the 30 indications shown are new indications, and the EOC-5 voltages used to estimate growth rates for them were obtained by reevaluating the prior inspection data. The result that the indications are new is to be expected since only RPC NDD indications were left in service at BOC-6 (repair on detection criterion was used at EOC-5).

4.3 NDE Uncertainties i The NDE uncertainties applied for the Cycle 6 voltage distributions in the Monte l Carlo analyses for leak rate and burst probabilities are the same as those previously used for the South Texas Unit-1 voltage-based repair criteria report of Reference 9-6 and NRC Generic Letter 95-05 (Reference 9-1). They are presented in Table 4-7 as well as graphically illustrated in Figure 4-8. The probe wear uncertainty has a standard deviation of 7.0 % about a mean of zero and has a cutoff at 15 % based on implementation of the probe wear standard. The analyst variability uncertainty has a standard deviation of 10.3% about a mean of zero with no cutoff. These NDE uncertainty distributions are included in the Monte Carlo analyses for SLB leak rates and tube burst probabilities based on the EOC-6 actual voltage distributions as well as for the EOC-7 projections.

4.4- Probability of Prior Cycle Detection (POPCD)

Although a voltage-based repair criteria is being applied for the first time for South Texas Unit-2, bobbin and RPC data evaluated consistent with the Generic Letter 95-03 , uidelines are available for two inspections (EOC-5 and EOC-6 inspections).

As pt rt of preparing the technical justification report for 1 volt repair criteria (Reference 9-2), the EOC-5 field bobbin data were reevaluated using a standard procedure developed for plants using voltage-based repair criteria. Therefore, with availability of EOC-6 inspection results, probability of detection at the prior EOC-5 inspection (POPCD) can be evaluated. For voltage-based repair criteria applications, the important indications are those that could significantly contribute to EOC leakage or burst probability. These significant indications can be expected Q \apc\thx\thx98\tsparc\thxc690d. doc 4 I) 1 1

l to be detected by bobbin and confirmed by RPC inspection. Thus, the population of interest for voltage-based repair criteria POD assessments is the EOC RPC confirmed indications that were detected or not detected at the prior inspection.

The probability of prior cycle detection (POPCD) for the EOC-6 inspection can then be defined as follows.

EOC-5 cycle reported + Indications confirmed indications confirmed by and repaired in EOC-5 RPC in EOC-G inspection inspection POPCD =

(EOC-5) ( Numerator) + New indications RPC l

confirmed in EOC-6 inspection POPCD is evaluated at the 1997 EOC-5 voltage values (from 1998 reevaluation for growth rate) since it is an EOC-5 POPCD assessment. The indications detected at l

EOC-5 that were RPC confirmed and plugged are included as it can be expected

~

that these indications would also have been detected and confirmed at EOC-6. It is also appropriate to include the plugged tubes for voltage-based repair criteria applications since POD adjustments to define the BOC distribution are applied prior to reduction of the EOC indication distribution for plugged tubes.

It should be noted that the above POPCD definition includes all new EOC-6 indications not reported in the EOC-5 inspection. The new indications include EOC-5 indications present at detectable levels but not reported, indications present at EOC-5 below detectable levels and indications that initiated during Cycle 6.

Thus, this definition, by including newly initiated indications, differs from the traditional POD definition. Since the newly initiated indications are appropriate for voltage-based repair criteria applications, POPCD is an acceptable definition l and eliminates the need to adjust the traditional POD for new indications.

l l

The above definition for POPCD would be entirely appropriate if all EOC-5 indications were RPC inspected. Since only a fraction of bobbin indications are generally RPC inspected, POPCD could be distorted by using only the RPC l inspected indications. Thus, a more appropriate POPCD estimate can be made by l

assuming that all bobbin indications not RPC inspected would have been RPC confirmed. This definition is applied only for the 1998 EOC-6 indications not RPC l

inspected since inclusion for the EOC-5 inspection for repaired tube could increase POPCD by including indications on a tube plugged for non-ODSCC causes which could be RPC NDF indications. In addition, the objective of using RPC Q: \ a pc \ thx \ thx98 \ tsparc \ thxc690d. doc 4-6

I l

confirmation for POPCD is to distinguish detection of indication at EOCn.i that  !

could: contribute to burst at EOCn so that the emphasis is on EOCn RPC l confirmation. This POPCD can be obtained by replacing the- EOC-6 RPC  !

confirmed by RPC confirmed plus not RPC inspected in the above de6nition of l POPCD. For this report, both POPCD definitions are evaluated for South Texas Unit-2.  !

I It can be noted that many of the new indications not RPC inspected can be false l calls and are not found at the subsequent inspection. It would be appropriate to l define new indications as the net increase in new indications at EOC-6 minus i indications reported at EOC-5, but not found at EOC-6. This would represent the j net new number of unconfirmed indications. Ignoring this effect leads to  ;

conservative POPCD distribution.

The POPCD evaluation for the 1997 EOC-5 inspection data is summarized in  ;

Table 4-8 and illustrated on Figure 4-9. As seen from Table 4-8, during the EOC-5 inspection a large number ofindications under 1 volt were RPC tested and those  ;

confirmed were repaired since " plug on detection" criteria was applied then.

However, relatively few indications under 1 volt were RPC tested in the EOC-6 since 1.0-volt repair was applicable. Because of this disparity in the RPC inspection of indication under 1 volts, POPCD based on RPC-confirmed only indications is not reliable. Therefore, only the results based on RPC confirmed  ;

plus not RPC inspected indications are shown in Figure 4-9. It is evident that l South Texas Unit-2 POPCD values support a POD signi6cantly higher than the I NRC mandated value of 0.6. A generic POPCD distribution developed by analyses l of 18 inspections in 10 plants and presented in Table 7-4 of Reference 9-4 is also )

shown in Figure 4-9. It is seen fmm Figure 4-9 that the POPCD values for South Texas Unit-2 are comparable to the generic POPCD in the voltage range 0.2 to 0.6 volt, and between 0.6 to 1.5 volts it is below the generic data. The POPCD value reaches unity at about 1.5 volts.

In summary, the South Texas Unit-2 EOC-6 POPCD supports a POD higher than the NRC mandated POD value of 0.6.

l i

4.5 Assessment of RPC Confirmation Rates

- This section tracks the 1997 EOC-5 indications left in service at BOC-6 relative to RPC inspection results in 1998 at EOC-6. If sufficient plant-specific data is  !

available on RPC confirmation rates for prior cycle NDFs, NRC approval may be obtained for considering only a fraction of unconfirmed (RPC NDF) indications m l

Q: \ a pe \ thz \ thx98 \ taparc \ thac690d. doc 4-7 j

l l

l

! I the BOC voltage distributions used for SLB leak rate and tube burst probability l projections. The object of this evaluation is to build such a database for later submittal to NRC.

The composite results from this evaluation for all 4 SGs are given in Table 4-9. For 1997 bobbin indications left in service, the indications are tracked relative to 1998 RPC confirmed,1998 RPC NDF,1998 bobbin indications not RPC inspected, and 1997 bobbin indications with no indication found in 1998. Also included are new 1998 indications. The table shows, for each category ofindications, the number of indications RPC inspected and RPC confirmed in 1998, as well as the percentage of RPC confirmed indications. l Only 4 out of 310 EOC-5 RPC NDF indications in service at BOC-7 were RPC l tested during the EOC-6 inspection and 3 were confirmed. Therefore RPC confirmation rate for prior RPC NDF indications is 75%. This RPC NDF database for South Texas Unit-2 is still too small to recommend a confirmation rate for use in the projection analyses. All RPC NDF indications are included in the EOC-7 projections presented in Section 8.0.

i 4.6 Probe Wear Criteria An alternate probe wear criteria approved by the NRC (Reference 9-7) was applied during the EOC-6 inspection. When a probe does not pass the 15% wear limit, this alternate criteria requires that only tubes with indications above 75% of the repair limit since the last successful probe wear check be reinspected with a good probe.

As the repair limit is 1 volt, all tubes containing indications for which worn probe voltage was above 0.75 volt were inspected with a new probe. An evaluation of worn probe and new probe data is presented in the following paragraphs.

In accordance with the guidance provided in Reference 9-7, voltages measured with a worn probe and a new probe at the same location were analyzed to ensure that the voltages measured with worn probes are within 75% of the new probe voltages. No new indications were detected with new probes; thus, worn probes did not miss any indication. Figures 4-10 and 4-11 show plots of the worn probe voltages plotted against the new probe voltages for all 4 SGs, and the data in these two figures show a consistent relationship between the two voltages. There are 2 indications with a worn probe voltage of about 2 volts for which the new probe voltage is about 17% to 25% higher. Composite data from all 4 SGs are i

plotted in Figure 4-12. Also shown in Figure 4-12 as a solid line is a linear l regression for the data, dashed lines representing tolerance limits that bound l 90% of the population at 95% confidence, and chained lines representing 25%

band for the new probe voltages. The mean regression line has slightly less than

(

1 Q hapc\thx\thx98\tsparc\thxc690d doc 4-8

45a slope indicating that on the average new probe voltages are slightly higher i than the worn probe voltages. The dotted horizontal line at 0.75 worn probe volts l demarcates indications requiring retest from those that do not. The shaded area l at the bottom above 1 volt shows the region where a tube requiring repair may be left in service because of probe wear. In the South Texas Unit-2 EOC-6 inspection, there are no occurrences for which a worn probe was less than 0.75 l

volt and the new probe voltage exceeded the plugging limit, i.e., no pluggable i tubes were missed due to probe wear considerations.

Among the indications requiring retesting (worn probe volts > 0.75 volt), 4 indications < 0.6 volt with the new probe fall outside the 90%/95% tolerance limit bands and *25% of the new probe voltage bands. All these 4 indications lie above the upper 90%/95% tolerance band as well as the upper 25% band; i.e., the worn probe voltages are higher than the corresponding new probe voltages and the worn probe voltages are conservative. Therefore, the data for these 4 indications are acceptable. Also, there are 5 indications below the lower 90%/95% tolerance band for which the new probe voltage exceeds the worn probe voltage. However,  !

all these 5 ind.ications are small (<0.75 volt with the new probe) and the new and l worn probe voltages differ by only few tenths of a volt; a voltage variation of this magnitude can be expected if the measurement is repeated with new or worn probes. Therefore, the data for 5 indications below the lower 90%/95% tolerance band are acceptable.  !

Overall, it is concluded that the criteria to retest tubes with worn probe voltages above 75% of the repair limit is adequate. The alternate probe wear criteria used  ;

in the EOC-6 inspection is consistent with the NRC guidance provided in Reference 9-7. l I

l l

l l

Q.\apc\thx\thx98\tsparc\thre690d doc 4-9

Table 4-1 (Sheet 1 of 2)

South Texas Unit 2 October 98 Outage Summary of Inspection and Repair For Tubes in Service During Cycle 6 Steam Generator A Steam Generator B Steam Generator C RTS for Cycle 7 In-Service During Cycle 6 RTS for Cycle 7 In. Service During Cycle 6 RT S for Cycle 7 In-Service Ih. ring Cycle 6

'O E A A = 1 f_ E A A = "

f_ E A A - A f_

0 2 0 0 0 2 2 _4 0 0 4 4 0.1 Q 0 0 0 Q _

_O_

0 5 59 0 0 0 59 59 38 0 0 0 38 3F 0.2 5 0 0 5 ,

0.3 40 0 0 1 39 3_9 125 0 0 0 125 125 96 ,

p 0 1 i 95 9; 0

39 124 0 124 124 109 .O_ 109 109 0.4 40 0 0 1 _39 Q .O. O.

0 0 0 34 34 70 0 0 0 70 70 82 0 0 1 81  ; 81 0.5 34

• 0 0 18 66 0 0 0 66 66 60 0 0 0 60 60 0.6 18 0 18 _

0 I 11 28 0 0 0 28 28 24 0 0 1 23 23 0.7 12 Q 11 ,

0 0 0 7 7 12 0 0 0 12 12 18 0 0 '8 18 0.8 8 1 0 0 6 6 9 0 0 0 9 9 10 0 0 0 10 10 0.9 6 0 4 3 2 5 4 2 0 0 0 2 2 5 0 0 0 5 5 1 7 0 2 0 1.! 3 3 2 2_ _t 0 2 2 2 2 Q_ 2_ _ _l_ 1 1 0 0 1.2 3 3 2 2 1 0 0 0 0 0 0 0 3 3 3 3 2 0 0 1.3 3 3 3 3 , Q 0 1 1 1 .I O _0 _ 2 1 1 0 0 1.4 0 0 0 0 0 0 0 0 0 _0 0 0 1 ,1 1 1 2 0 0 0 0 0 0 Q 0 0 0 0 0 0 1.5 3 3 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1.6 0 0 0 0 0 0 0 1.7 0 0 0 Q ,, 0 0 0 0 0 0 0 _0_ 1 1 1 1 ,

0 0 0 0 0 0 0 0 0 0 0 0 0 1.8 2 2 2 2 Q 0 0 .l. I 0 0 1.9 0 0 0 0 _0 0 0 0 Q 0 0 0

.Q .

0 1 _.

1 0 : 0 2.3 0 0 0 0 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 j 0 2.5 I i 1 1_ 0 '

0 0 0 0 0 2.6 1 1 j. 1 0 0 0 0 0 _0 0 .Q 0_

0 0 0 0 0 0 0 0 0 Q 0 0 0 0 2.8 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4.2 1 1 1 1 500 3 3 i 3 497 i 497 457 10 9 13 ! 444 443 Total 188 22 17 20 168 163 ' ~~

' 0

~

0 10 9 9

>lv 18 18 : 14 14 4 0 3 3 3 f3 3 0 0 10 0

1 0

>2v 4 4  : 4 4 0 0 0 0 0 0 1 1 1 1

% 1 . n , un- . u ~

4-10

Table 1 (Sheet 2 of 2)

South Texas Unit 2 October 98 Outage Summary of Inspection and Repair For Tubes in Service During Cycle 6 Steam Generator D Composite of All SGs in servke penne cyde 6 RTs ter cyde 7 In.sereke Dunes cyde s RTs ter cyde1

• • '" " wrc nec an nee mec an pampseted cesaraud Reyserad gamessens tasyse.rd casArund Sepdred Isersemos 0 0  : 0 l 2  : 2 8  : 0  : 0 0  : 8  : 8 0.1 __2  !  !

25 127 :

0

0

0

127 :

127 0.2- 25  : 0  : 0 25 0.3 84 l 0  ! 0 l~0d ~~ ! 84 l~~ 84 345 l ~6~l 0 [ 2 l 343 $ 343 0.4 71 } O  ! 0 l 1 70 i 76 344 i 0 } O 2 i 342 i' 342 OI-- 64 I O j 0 i ~ 'i 64 i 64 250 i-~0 i 0 1 I 249 : 249 0.6 29 i 0 i 0  : 0 29 i 29 173 I ~ D- ~~l 0 m 0 i 173 i 173 29 0 0  :. 0 29  : 29 93  ; O  : 0  :. 2  : 91  : 91 0.7  :  :. .

0 0 12  : 12 50  : 0  : 0  : I  : 49  : 49 0.8 12  : 0  :  : m 0 0  :  : 1I 36 1 0  : 0  : 0  : 36  : 36 0.9 11  :

0  : 11 17.... .~.16 0  :, 5 19 5 1. .. 4--.-.2.

..I. . 5.  :

.. .I. .:- 1  :

. 5...

0

6 1 6 2 j 0-

I  : I I j 0 j 8 _8 1.1 1.2 1

2"~~!~

2  !~~~i-l 2' i

' D~ ~ j 0

8_._6 8 I ~i~ 6

~Y' 6

~~ l 0

i. 0_

0 13 I  : 1  : 1  : I  : 0  : 0 7  : ,

,_ 0 .

0 0 0  :  : I  : 1 0  : 0 I.4 0  : 0  :  :  :  : 1 1 2 2  : 2  : 0  : 0 5  : 5 i 3  : 3 -l 2  : 0 1.5 2  :  :

1.6 1  ! I i I  : I  : 0  : 0 I j , __ I

I I

I  :

! 1 I

O 0

i 0]

0 i.7 0  : 0  : 0  : 0  : 0  : 0 1

-~j ,p O  : 0  !~~U~ O  : 0 2  : 2 2  : 2  : 0 O 1.9 ~~6^-;l 0 0 l~  !~ 'U *! ~~D ]! O I ^! I I

~f I '~~U D __

0

,i_ gi_7i__U' i 0 2 L 2 L 2  : 2  : 0  :

i 2.3 7..-O 1 1 1 g g  ; g 7 g ,  ;-- I~~ I 1 i 1 0 F 0 i ~~ U~I l i 0 0 2.6 0 0 I~0 0 O I  :  : 1  : i' 0

0 0 0 0 0 0 I  : 1  : I  : I  :  :

2.8 0 l  :  :  :

i 0 i 0 0  : 0  : 0 I  : 1 i 1 i i  : 0  : 0 4.2 0  :

331 1485 : 44 -: 38  :- 44  : 1440 :-. 1434 Total _340 ,

9 . .9 l

. 9.. :. 331 5 0

>I v 8 8  : 8  : 8  : 0  : 0 40  : 39  : 34 34~~ : 1 6 ~~I 0 i l -~ ~ ! ~ l ~! j ~ ~ ! ~ ' ~U ~ ~ ! -' ~ 0 6  ! 6 f 6

>2 v r . .., im . n.

4-11

Table 4-2  :

South Texas Unit 2 October 1998

.t TSP ODSCC Indication Distributions for Tubes in Service During Cycle 6 a

Steam Generator A Steam Generator B Steam Generator C -

- Tube Average I

. Average Imgest Average Number d Maximum Average Largest Support Number d Maximum Average largest Average Number of Manimum '

voltage voiage Grows crown inacanons vanage vahage - Grows artme Plaie inacanons vahage voltage crowe crows indications t 0.20 179 1.29 039 0.67 0.12 166 1.84 0.47' ' t .40 0.14 02H 77 2.76 0.62 - 2.23 '

0.09 170 1.08 0.42 0.69 0.09 142 2.26 - 0.42 _f .86_ _0.12 _

03H 74 4.12 0.60 3.41 115- 0.94 037 035 0.07 102 134 0.41 0.98 ' 0.11 24 0.85 _ 0.40 0.27 0.03 ,

04H_

._12_,. 0.99 0.40 0.20 _ 0.04_ 33 _ 0.70 033 035- 0.08 28- 0.90 039, 034 - 0.08_ 1

._05H _ 0.41 031 0.09-0.29 0.29 0.00 0.00 0.23 0.23 0.03 0.03 15 0.84 ~

06H 1 1 '

0 0 - - - - 2 0.49 032 0.14 0.08 07H - - - -

-2 0.15 0.13 0.02 0.02 0 - - - .-

. 08H 0 - - - -

0 - - - 1 0.18 0.18 0.05 0.05 11C 0 - - - - - 7 0 - - - 1 1.23 1.23- - - -

~22C 0 - - - - -

188 500 457 l

~ Total Steam Generator D Composite of All SGs.

Tube  !

Average Largest Average Number of Maximum Average 1mgest Average S*PPort Number of Maximum [

l Plate Voltage Voltage Growth Growth In& cations Voltage , Voltage Growth Growe Indications 02H 164 'I.53 0.42 1.20 0.18 586 2.76- 0.45 2.23 0.15-d3i 2.26 ' 1.44 0.21 512 4.12- 0.46 3.41 0.13

> 03H l2I6 039 268 _l34_ 0.98 ,. 0.09 0.65 _ _.0.18 ;

i 04H _27 _ _ 0.85 __039 05H 14 0.68 037 038 0.13 87 0.99 037 038 0.08 06H 8 0.63 038 034 0.16 25 0.84 039 034 0.11 07H I 0.29 0.29 0.15 0.15 3 0.49 031 0.15 0.10 '

0 2 0.15 0.13 0.02 0.02 08H - - - -

0 - 1 0.18 0.18- 0.05 0.05 11C - - -

22C 0 - - - - 1 1.23 1.23 - -

Total 340 1485 t

Grumerrdeel Q1117t tM9811900 4-12

- _ _ - _ _ - _ - - = _ _ -__ _ _ _ _ ___ - _ _-_ _ _ .

Table 4-3 South Texas Unit 2 October 98 Signal Growth Statistics For Cycle 6 on an EFPY Basis Steam Generator A Steam Generator B Steam Generator C Steam Generator D Cumulative Delta Cycle 6 Cycle 5 Cycle 6 Cycle 5 Cycle 6 Cycle 5 Cycle 6 Cycle 5 Cycle 6 Cycle 5 CPDF CPDF CPDF CPDF CPDF CPDF CPDF CPDF CPDF CPDF ,

i 0.005 0.0 0.002 0.0 0 0.0 0.0 0 0.0 2 0.001

-0.2 0.0 _

1 .

0.032 0.015 0 0.002 0.011 2 0.004 0.006 0 0.0 0.009 7 0.006

-0.1 0.0 5 44 0.266 0.215 65 0.132 0.09 49 0.112 0.194 14 0.041 0.137 172 0.122 0 0.021 0.782 0.738 334 0.8 0.534 289 0.746 0.823 170 0.541 0.62 890 0.722 0.1 0.326 ~ 97

]04_

~

~U.847 ~308 ~ ~ U.fi@

d0.3 0}736 0.868

. _ 27_

4 0.926 0.947

_0.974]

0.99 8[~

15

' 0 79 _6~I 0.815_ _ _ 94 [ ~U 935 ~0_.96 _

0.996 0.942 16 0.987 0.989 34 0.947

_ 0.87[

0.952 69 0.976 0.938 0.963 0.99 0 0.996 0.984 1 0.989 _ 0.989 8 0.971 0.977 12 0.984 0.4 3 0.5 _ _0.972_ __2_ _0.973 0.995_ _2__ l.0 _ _ 1.0_ _ __

1 _0.991 _ _0_.989_ _6_ _ 0.988 . __0.99 _ _ _1 1_ _ _0.991 0.979 0.995 0 0.991 1 0.991 0.996_ _2__ 0.993

_ 0.6_ 0.986 _

1 ______ _ _ _ _ __ 0 _ _1.0 0 0.979 1.0 0 0.993 1 0.994 0.999 2 0.994 0.7 0.993 1 0 0.979 0 1 0.996 1 0.997 1.0 2 0.995 0.8 1.0 0 0.979 0 1 0.998 1 1.0 2 0.997 1

2 0.989 0 0 0.998 0 2 0.998 1.1 0 0.989 0 1 1.0 0 1 0.999 1.2 0.995 0 0 1.0 0 1 0.999 1.5 1 2.2 1 1.0 ' 0 0 1.0 0 1 1.0 500 456 340 1484 Total 188 Cmh CTDFdata 12/21/98 II:1) AM

Table 4 4 South Texas Unit 2 - October 1998 Outage Average Voltage Growth During Cycle 6 Voltage Numtw of Average Voltage Range Indications BOC Entire Cycle Per EFPY ' Entke Cycle Por EFPY '

Composite of AII Steam Generator Data Entire Voltage Range 1484 031 0.13 0.08 42% 27 %

1437 0.29 0.13 0.08 44 % 29 %

V soc < .75 Volts 2.75 Volts 47 0.93 0.16 0.10 17 % i1%

Steam Generator A 188 0.44 0.12 0.08 28 % 18 %

Entire Voltage Range

, _ _ ],, _ _ __ _ 0,_

.5 V

. Y 80C 5 . _ _ _ , _ _ _ _ ._ . _ _ _ ., , , _ _ - _ _

Steam Generator B Entire Voltage Range 500 O.29 0.10 _

0.06 33% 21 %

490 0.28 0.10 0.06 35 % 23 %

V soc < .75 Volts 2.75 Volts 10 0.85 0.02 0.01 2% 1%

Steam Generator C Entire Voltage Range 456 0.31 0.12 0.08 39% 25 %

445 030 0.12 0.08 41% 27 %

__V soc < .75 Volts _

7%

2.75 Volts 11 0.89 0.09 0.06 11 %

Steam Generator D 0.19 0.12 75 % 49 %

Entire Voltage Range 340 _ ___ 0.25 338 0.25 0.19 0.12 75 % 48%

V sac < .75 Volts 2.75 Volts 2 0.80 0.83 0.54 104 % 68 %

1. Based on Cycle 6 duratum of 564.9 EFPD (1.547 EtPY) wmunWWi1.H AM 4,14

Table 4-5 South Texas Unit 2 October 1998 Avenge Voltage Growth Satatistics Composite of All Steam Generator Data Average Voltage Growth Average Percentage Growth Bobbin Voltage Number of Average Voltage Range Indications BOC Entire Cycle Per EFPY Entire Cycle Per EFPY Cycle 6(1997- 1998) - 564.9 EFPD 1484 0.31 0.13 0.08 42 % 27 %

Entire Voltage Range 1437 0.29 0.13 0.08 44 % 29 %

V soc < .75 Volts 47 0.93 0.16 0.10 17 % 11 %

2.75 Volts Cycle 5 (1995 - 1997) - 450 EFPD 703 0.31 0.12 0.10 39 % 31 %

Entire Voltage Range 6% 0.31 0.12 0.10 39 % 32 %

V noe < .75 Volts 7 0.91 0.20 0.16 22 % 18 %

2.75 Volts ,

4 o-wrewum2mmn us m 4-15

Table 4-6 South Texas Unit 2 October 1998 l Summary of Lagest Voltage Growth Rates for BOC-6 to EOC-6 Steam i;en<.ata Bobbin Voltage RPC New  !

3--

SG Row l C e l Elevation EOC BOC Grow:h Confirmed ? Indication ?

-n..'.

A 18 -.

i X)

.T H 4.12 0.71 3.41 Y Y  ;

I A 19 _ 83 02h 2.76 0.53 2.2'4 Y N C 27 88 03H 2.26 0.4 1.86 Y Y A 25 88 03H 2.54 0.88 1.66 Y Y A 20 36 02H 2.43 0.78 1.65 Y Y l D 16 100 031; 2.26 0.82 1.44 Y Y C 8 111 0211 1.84 0.44 1.4 Y Y C 25 40 03 1.62 0.42 1.2 Y Y D 20 36 02P. ' e2 0.22 1.2 Y Y l D 16 101 -

03H 1.44 0.45 0.99 Y Y C 17 72 'WH I.34 036 0.98 Y Y D 16 100 m et I/ 0.63 00 Y Y A 25 40 02H _ l.78 0.97 0.81 Y Y A 24 104 02H 1.23 0.48 0.75 Y Y C 15 86 02H 1.14 _ 0 41 0.73 Y Y D 29 43 03H 0.99  : <.27 0.72 Y Y A 25 85 02H I 29 0 39 0.7 Y Y D 18 40 02H 0.E2 0.12 0.7 N Y j B 13 113 03H 1.08 039 0.69 Y Y B 10 102 02H 0.85 _ 0 18 0.67 l N Y i D 19 38 03H 1.1 . (1.45 0.67 Y Y D 24 68 04H 0.85 0.2 0.65 N Y D 29 21 02H 0.79 0.15 ,_ 0.64 N Y D 18 101 03H 13 0.51 0.63 Y Y D 23 74 03H 0.91 03 0.61 N Y D 42 89 03H 0.8 , 0.21 0.59 N Y A 24 106 02H 0.9" 0.4 0.58 N Y C 34 76 03H _ l.01 0.46 0.55 Y Y A 21 104 02H I .0' O.% 0.53 Y Y D 19 104 03H s.03 0.5 0.53 Y Y cno-o T bies imman ii:is Au 4-J G

Tchle 4-7 ,

Probe Wear and Analyst Variability - Tabulated Values I Analyst Variability Probe Wear Variability Std. Dev = 10.3% Mean = 0.0% Std. Dev = 7.0% Mean = 0.0%

No Cutoff Cutoff at +/- 15% ,

Value Cumul. Prob. Value Cumul. Prob. )

-40.0% 0.00005 < - 15.0% 0.00000

-38.0% 0.00011 _ -15.0% 0.01606

-36.0% 0.00024 -14.0% 0.02275  !

-34.0% 0.00048 ~ -13.0% 0.03165 I

-32.0% 0.00095 -12.0% 0.04324

-30.0% 0.00179 -11.0% 0.05804  ;

-28.0% _

0.00328 -10.0% 0.07656

-26.0% 0.00580 -9.0% 0.09927

-24.0% 0.00990 -8.0% 0.12655 i

-22.0% ~

0.01634 ~

-7.0% 0.15866 l

-20.0% 0.02608 -6.0% 0.19568

-18.0% 0.04027 -5.0% 0.23753 i

-16.0% 0.06016 -4.0 % 0.28385

-14.0% 0.08704 -3.0% 0.33412

-12.0% 0.12200 -2.0% 0.38755 l

-10.0% 0.16581 -1.0% 0.44320 l

-8.0% 0.21867 0.0% 0.50000

-6.0% 0.28011 1.0% 0.55680

-4.0% 0.34888 2.0% 0.61245

-2.0% 0.42302 3.0% 0.66588 i 0.0% 0.50000 4.0% 0.71615 2.0% 0.57698 _ 5.0% 0.76247 4.0% 0.65112 6.0% 0.80432  !

6.0% 0.71989 ,

7.0% 0.84134 ,

8.0% 0.78133 8.0% 0.87345 l 10.0 % 0.83419 9.0% 0.90073 12.0 % 0.87800 10.0% 0.92344 14.0% 0.91296 _

11.0 % 0.94196 16.0 % 0.93984 12.0 % ,

0.95676 18.0% 0.95973 13.0% 0.% 835 20.0 % 0.97392 14.0 % 0.97725 22.0 % 0.98366 15.0 % 0.98394 24.0 % 0.99010 > 15.0% l.00000 26.0 % 0.99420 28.0 % 0.9 % 72 30.0% - 0.99821 32.0 % 0.99905 34.0 % 0.99952 _

36.0 % 0.99976 38.0 % 0.99989 _

40.0 % 0.99995 NDEuncert Table 3 712/21/9811:31 AM 47y

'A r

Table 4-8 South Texas Unit 21998 EOC-6 Evaluation for Probability of Prior Cycle Detection Composite of A81 Steam Generator Data

" "'" POPCD New Indications In ion 1 7n o Bobbin 1998 1998

' Inspection Inspection RPC 1998 RPC 1998 RPC 1997 RPC Confirmed Inspection Confirmed Inspection Confirmed Inspection Confirmed Plus Not Voltage plus not RPC plus not Confirmed inspected Bin RPC Confirmed Inspected Confirmed Inspected and Plugged Frac. Count Frac. Count 0 414 0 45 32 1.0 32/32 0.157 77/491

> 0 - 0.2 4 531 1 196 289 0.986 290/294 0.477 485/1016 i 0.2 - 0.4 9 161 2 52 203 0.958 205/ 214 0.613 255/416 0.4-0.6 16 57 0 16 104 0.867 104/120 0.678 120/177 0.6 -1.0 6 6 0 0 16 0.727 16/22 0.727 16/22 1.0 - 1.5 0 0 0 0 1 1.000 1/1 1.000 1/1 1.5 - 2.0 TOTAL 35 1169 3 309 645

> 1V 6 6 0 0 17 PopedTable1(2)12/216811:21 AM 4 yg

Table 4-9 South Texas Unit 2 Analysis of RPC Data from 1997 and 1998 Inspections Combined Data from All Steam Generators Total Total al Peet Total 1998 1997 1998 1998 U" Int,pection inspection Group of Indications inspection Inspection ,

Bobbin Bobbin RPC RPC Indication Indication Ins W ed Confirmed Confirmed Less than or Equal to 1.0 Volt in 1998 Inspection '

1997 Inspection Bobbin Left in Service 345 306 0 O -

- 1997 Inspection hPC Confirmed 0 0 0 0 -

- 1997 Inspection RPC NDD 306 306 9 .g_ -

.1997 Inspection RPC Not in_ speed _ _.. ___0____ ____0 _ _0 , .__ __0 _ _

- No 1998 Inspection Bobbin

• 39 - - -

- 1139 5 4 80.0

_ New 1998 Inspection Indication Sum of All 1998 Inspection Indication 345 1445 5 4 80.0 Greater than 1.0 Volt in 1998 Inspection 1997 inspection Bobbin Left in Service 4 4 4 3 75.0

- 1997 In_spection RPC Confirmed . O_ _ _0 _ _0 , _ . ._

0, _

._ 75.0

- 1997 Inspection RPC NDD 4 4 4 3

- 1997 Inspection RPC Not inspected 0 0 _

0 0 -

- No 1998 Inspection Bobbin

• O - -

35 35 31 88.6

._ New 1998 Insq Indication, , _ _ _ .

Sum of All 1998 Inspection ledication 4 39 39 34 87.2 All Vol'= gas in 1998 InsWlon 349 310 4 3 75.0 1997 Inspection Bobbin Left in Service

- 1997 Inspection RPC Confirmed _0 _ _ _ O_ . , __O _ __0 _ _ -

- 1997 Inspection RPC NDD 310 310 4 3 . 75.0

- 1997 Inspection RPC Not inspected 0 0 0 0 -

- No 1998 InspEtion Bobbin

• 39 - - -

- 1174 40 35 87.5 g ._New 1998 Inspection Indication 38 86.4 l Sum of All 1998 Inspection Indication 349 1484 44

• Indications spht is based on 1997 Inspection botMn voltage Poped Table 212/21/98 If:21 AM 4-19 i

Figure 4-1 South Texas Unit 2 October 1998 Outage Bobbin Voltage Distributions at EOC-6 for Tubes in Service During Cycle 6 140 i

120 --

100 E SG-A E

O SG-B i S

5 e

80 L-5 ,

E SG-C i i

t i

.$ 60 -

E l -l  :

- -l ESG-D I

n -

t Z

40 - -

t i

20 - l

- i 0 l l l l l l l lhl' *

" l - l' "l -l - l" l -l - l- l- l- l-

• m N 9

o N

o lIn o

l 4.

o 9

o

'o o

N o

o 9

o N

O-M 9

'o N

9 9

m 9

N 9

N N -r Bobbin Voltage

% nsu 4-20

t Figure 4-2 South Texas Unit 2 October 1998 Outage Bobbin Voltage Distribution for Tubes Plugged After Cycle 6 Service 3.5 j i

i 3

E SG-A E

O SG-B 32- -

E SG-C l

=

$ l.5 -

j ESG-D

J L ll I I E ll llI,1,I,1,1,1, _

l ll LI

,1L,I,1,1,Illi,,,Ill

= =========,;===

Bobbin Voltage

%n 2 4-21

Figure 4-3 South Texas Unit 2 October 1998 Outage Bobbin Voltage Distributions for Tubes Returned to Service for Cycle 7 140 f

120 -

1 e

j 100- -

ESG-A 3

i O SG-B 5 80 - -

E SG-C 1

!# ESG-D 40 -

1 --- --

20 - -

E i i l

1 5 5 i o "=lE  :

E

=

:  :  : M:__ _

0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.5 0.1 01 Bobbin Voltage i

i

" " * ' 4-22

!l !

i, il! .! i l! ' I .;i ;l' :l!; j-!  !;i! I ii

._ C l l

A- B -

C- D- ,

G G G G _

S S S S .

E O E D H 8

0 6

l e -

c y - H 7

C 0 g -

i n -

r u .

8 D

9 e H 9i c S e 1 v r r EE: 1 1

6 t

a e e 0 P l

b S on t i t

r c o 4Osbe p

p 3 4 - u u 2-e 2,T S rt uir EN I i' e

4 gn o 5EEE 1 1 b u

iUf F s 5

0 T s n a

xi o et Tub hi t r ut s oi SD 5E=E=5 E== 2=5E5= 1 1

l 4 i

a -] 0 x

A C  :

C S

D O 5E2 - - - E 5 =- = ======- 5E" E H 3

F ll 0

$agagShaahaa$ 3Q

- E - - - - E- - - E - - - ======E=522 = H 2

0 0 0 0 0 0 0 0 0 0 o 0

2 8

1 6

1 4

1 2

1 0

1 8 M 4 2 gg 8g$o w8.E$

Figure 4-5 South Texas Unit 2 Cycle 6 ( July 1997 to Oct.1998 )

Cumulative Probability Distributions for Voltage Growth on an EFPY Basis

_ .:. : - u g :w -= - r- r i

. 5 y -

s' O.9 -

• w'

l/

0.8 ------ --- -------- -

t-i a */ .

o i

---

)

t 0.7 - /

c

• 2 5 i,k /

---

------- 1 r*----------------------------------------------------------------------------------------

I-8 0.6 x ,

= /Y

.c  : SG-A

'f-------------------------------------------------------.....

T 0.5 ------ --- ------

4

) I Q --o-- S G-B .......

1. 0.4 ------ --- -----

=n

- i .I

- - x - - SG-C

$0.3 ------ -------

2 I U

- *- - SG-D 0.2 - ------------!------------------------------------------------................ .......

1

--x- Cumulative 0.1 ------- - -

d--- ----------------- --------------------- ------ ------------_... _. ..___..

I a

0L -  : l l 0.2 0.3 0.4 0.5 0.6 0.7 0.8 1 1.1 1.2 1.5 2.2

-0.2 -0.1 0 0.1 Voltage Growtii Growth Fig 212/21/9811:17 AM 4-24

Figure 4-6 South Texas Unit -1 October 1998 Outage Voltage Growth During Cycle 6 vs BOC-6 Voltage 3.5 ,

3 o SG-A 4

t i e SG-B 2.5 ,

e w

0 + SG-C

[O on 2

i i x SG-D

.g l

• l

.g 1.5 x

$ +

g x

+ x gi #

h x x., x

. +x

• o x 05 -  ? +" ^

$ '+ o*

+ ,

+

l A i

l O  ! =

i

-0.5 ,

0 0.2 0.4 0.6 0.8 I l.2 1.4 1.6 BOC-6 Voltage t

l

% % vimunii9:ii.i,au 4-M  ;

Figure 4-7 South Texas Unit 2- October 1998 Bobbin Signal Growth History - Cumulative Probability Distributions on an EFPY Basis Composite of All Steam Generators l.0 -

--- - --- - m - .

. a- ----

i n-0.9 -----

0.8 ----------------------,!-------------------------------------------------------------------------

C 2 --- - -------- ---- Y. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

% 0.7 -

C

• m

• g g 0.6 -----------------l---------------------------------------------------------------------------------

=  :

m .

.m

• 1: n.5 ..................; .................................................................

I - - O - - Cycle 6

? 0.4

---

l------------------------------------------------------------

2  :

m ---------- ----------- ------------ ------- ------

5 0.3 ---- --- ---- -~~------- ------- -- -

Cycle 5 r"

0.2 0.1 - -----

0.0 c - - - - -  :  :

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 1 1.1 1.2 1.5 2.2

-0.2 -0.1 Voltage Growth Growth Fig 312/21/9811:18 AM 4 26

0

-. 4

- li ,.

- . 0 3

0 2

)

%(

A 0 e 1 g t

a s l o

n i

o V l

t u a n

b g i

r %iS t 0 s n 8i i 4

D n o

e y i t

7 2-rt n i a

ui ga r 4 i

F re t

? g %a 0 V 1 t c - n e

n c r

U t y e E

i l

i P

D b a

N i r r a a  %

0 V e 2-t s

y e W

l a b AP n or &

^ . -  %

- 0

_ .- a

.. u

_ ..~  % u

. 0 4 n 1 9 8 7 6 5 4 3 2 1 0 - n i

0 0 0 0 0 0 0 0 0 i

^iiiaj I>=!mEmU a n.

E2

!. c 4

m i

n.

m c

4 n

J

Figure 4-9 South Texas Unit 2 1998 EOC-6 Evaluation for POPCD at EOC-5 1.0 _ _ _ _

- ~~~'___

0.9 ,~

/

s 0.8 ,'

/

0.7 j r

i a /

.2

+* /

e 8 0.6 '

d

/

y 05 .

5

.C /

/

.E O

0.4 /

/ .

l 0.3  :

RPC Confirmed Plus Not inspected

/ .

t 0.2

+-c

--- EPRI POPCD (NP-7480, Addendum-2) 0.1 0.0  :  ;

0 0.5 1 1.5 2 2.5 3 35 Bobbin Amplitude roccerigui2nimH1:D AM 4-28

. . . =_ . _ . . - .

1 1 '

1 Figure 4-10 i South Texas Unit-2 -- EOC-6 Inspection  !

Comparison of Worn Probe Voltage Against New Probe Voltage

! Steam Generator A '

3.0 )

l i

I i

2.5 ,

i i

g 2.0 l 4

> o ,

1.5 )

- o

• ,n - -

,a o, no a

o' O.5 ". *.

8 ,#o " ,

on a 50 0 0.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 New Probe Voltage Steam Generator B 3.0 2.5 2.0 1.5 5

N 1 .0 l .

l

= .

0.5 o

sa 0.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 New Probe Voltage

- . . - - 4-29

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

Figure 4-11 South Texas Unit 2 -- EOC-6 Inspection Comparison of Worn Probe Voltage Against New Probe Voltage l

Steam Generator C 3

I 2.5 2.0

- l i

1.5 E

D 1.0 o o 0.5 ., ,

o 0.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 New Probe Voltage 1

Steam Generator D 3.0 2.5 l

2.0 ,

1.5 ,

t 1.0 = e o o 0.5

1. . o e

0.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 New Probe Voltage

-w-4-30

I t

l Figure 4-12 i South Texas Unit-2 October 1998 j Worn Probe Volts vs New Probe Volts  :

4.0 l a Field Data ,. -

Linear Regression ,-

l

- ---- 90%/95% Tol. Band - -

t 3.0 ,.- , ..

---

+/- 25% of New ,-

.s

..- .. l g 2.5 ,

,..- y ...

3 -,.

. ,... - l, o -

e .s . . ..- o

.o ,.-

g 2.0 ,,..

. . , .o a

% /

.s p

o

.s

.. ~ ...o ... ....- _. '.

y 1.5 - -

Retest Required -

Retest Required

.- D'... -

{ j j

. s[.-

l ..

{ .. -, ' .

k' .'{ { h k e

1.0 .-,..g- , - ,

o a * . . , . .

c . .D 3. .

n .s.t1 -p PC p '-

,,..D'

~~

0.0 1.6 2.0 2.5 3.0 0.0 0.5 1.0 New Probe Voltage .

-n m n . m 4-31

5.0 Database Applied for Leak and Burst Correlations The database used for the leak and burst correlations that are applied in the analyses of this report is the same as the voltage-based repair criteria database for I  %" tubes approved recently by the NRC (Reference 9-5), and it is documented in I

Reference 9-4. Plant S pulled tube indication R28C41 is included in the leak rate correlation at a revised SLB leak rate of 1250 lph consistent with Reference 9-5 Leak rate data fer Model Boiler specimens 598-3 and 604-2 are excluded from the repair criteria database based on application of EPRI data exclusion Criterion 3a which permits exclusion of leak rate data that lie below the one-sided, 99-percent statistical confidence intervals of the mean regression line relating leak rate to both throughwall crack length and bobbin coil voltage.

South Texas pulled tube data from 1993 and 1995 inspections are included in the voltage-based repair criteria database The database meets the NRC requirement l

~ that the p value obtained from the regression analysis ofleak rate be less than or equal to 5%. Therefore, a SLB leak rate versus voltage correlation is applied for the leak rate analyses of this report.

l The following are the correlations for burst pressure, probability of leakage and leak rate used in this report (Reference 9-4).

Burst Pressure (ksi) = 7.4234 - 2.9920 x log (Volts)

I Probability of Leak =

, ( 5.1721 - 8.6705 x log (volts))

f N

- 2.119 + 3.3162 x log (volts)

= >

Leak Rate (l/hr) 10A Additional leakage and burst pressure data are available from 3 TSP indications pulled from South Texas Unit-2 SGs during this outage. Two of the indications are  !

in tube R18C100 in SG A (TSPs 2 and 3) and the other indication in tube R19C83 l of SG-A (TSP 2). An evaluation of the effects of adding the new South Texas Unit- l 2 data to the reference database in Reference 9-4 (described earlier in Section 3.3) indicates that the burst pressure, leak rate and the probability ofleak correlations I to the common logarithm of the bobbin amplitude would not be significantly changed. Therefore, SLB leak rates and burst probability analyses were carried out using the reference database presented in Reference 9-4. As a sensitivity study, EOC-7 projections f- the limiting SG (SG-A) were also calculated using s Q.\npc\thx\thx98\taparc\thsc6W da:.,

51 i

leak and burst correlation based on an updated base that included the data from the above pulled specimens, and those results indicate are presented in Section 8.

The upper voltage repair limit applied at the EOC-6 inspection, documented in Reference 9 2, was developed using the latest NRC-approved database presented in Reference 9-4. The structural limit (V.i> for the TSP indications established using 3 times normal operation AP value (3675 psid) is 5.45 volts, and V.i for the FDB intersections using 1.43 times the SLB AP of 2405 psid is 4.47 volts. The allowance for voltage growth used is 49%/EFPY, which is the highest average growth rate on an individual SG basis for South Texas Unit-2 Cycle 5 operation, which is above the minimum value (30%/EFPY) specified in the Generic Letter 95-05. For the expected 0.9 EFPY (329 EFPD) for Cycle 7, the growth allowance becomes 45%.

The allowance for NDE uncertainty is 20% per Generic Letter 95-05. The upper voltage repair limits then becomes 3.30 volts for TSP indications and 2.71 volts for FDB indications. These values were applied at the EOC-6 inspection to assure that indications exceeding these limits were repaired independent of RPC confirmation.

I l

4 l

s Q.\apc\thx\thx98\tspa.. 2.&Od. doc 52 l

+-

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

6.0 SLB Analysis Methods  ;

s.

Monte Carlo analyses are used to calculate the SLB leak rates and tube burst i probabilities for both actual EOC-6 and projected EOC-7 voltage distributions. The '

Monte Carlo analyses account for parameter uncertainty. The analysis methodology is described in the Westinghouse generic methods report of Reference 9-3, and it is consistent with the methodology applied to obtain the leak rate and' tube burst probability results presented in the last 90-day report for Unit-1 (Reference 9-6)

In general, the methodology involves application of correlations for burst pressure, probability ofleak and leak rate to a measured or calculated EOC distribution to ,

estimate the likelihood of tube burst and primary-to-secondary leakage during a postulated SLB event. NDE uncertainties and unce-tainties associated with burst

.nressure, leak rate probability and leak rate correlations are explicitly included by considering many thousands of voltage distributions through a Monte Carlo sampling process. The voltage distributions used in the projection analyses for the next operating cycle are obtained by applying growth data ta :he BOC distribution. i The BOC voltage distributious include an adjustment for detection uncertainty and occurrence of new indications, in addition to the adjustments for NDE uncertainties. Comparisons of projected EOC voltage distributions with actual distributions after a cycle of operation have shown that the Monte Carlo analysis technique yields conservative estimates for EOC voltage distributions and as well as leak and burst results based on those distributions. Equation 3.5 in Reference 9-3 was used to determine the true BOC voltage.

Q.\apc\thx\thx98\tsparc\thscG90d doc G-1

_ - . _.. _ __ m ____ _ . _ ___ _ _ _ ______. _ _ _-_.

l' 1

l 1

7.0 Bobbin Voltage Distributions  !

l This section describes the salient input data used to calculate EOC bobbin voltage distributions and. presents results of calculations to project EOC-7 voltage

distributions. Also, EOC-6 voltage projections performed during the last outage based on-EOC-6 inspection bobbin voltage data are compared with the actual bobbin distributions from the current inspection.

7.1 Calculation of Voltage Distributions I

The analysis for EOC voltage distribution starts with a cycle initial voltage l distribution which is projected to the end of cycle conditions based on the growth I rate and the anticipated cycle operating period. The number of indications assumed in the analysis to project EOC voltage distributions, and to perform tube leak rate and burst probability analyses, is obtained by adjusting the number of reported indications . to account for detection uncertainty and birth of new indications over the projection period. This is accomplished by using a POD factor, which is defined as the ratio of the actual number ofindications detected to total number ofindications present. A conservative value is assigned to POD based on historic data, and the value used herein is discussed in Section 7-2. The calculation of projected bobbin voltage frequency distribution is based on a net total number of indications returned to service, defined as follows.

NTot nTs = Ni/ POD N,,p.irea + Naepioggea where, NTot RTs = Number of bobbin indications being returned to service for the next cycle, Ni = Number of bobbin indications (in tubes in service) identi6ed after the previous cycle, POD = Probability of detection, N,,p.ir a = Number of Ni which are repaired (plugged) after the last cycle,  ;

Na.pioggea = Number ofindications in tubes deplugged after the last i cycle and returned to service in accordance with voltage-  !

based repair criteria. i f

There are no deplugged tubes returned to service at BOC-7; therefore, Naepioggea = 0.

i Q:\npe\thx\thx98\tsparc\thxc690d. doc

'71

j I

The methodology used in the projection of bobbin voltage frequency predictions is described in Reference 9-3, and it is. same as that used in performing EOC-8 predictiom, during the last (EOC-7) outage for Unit-1 (Reference 9-6). Salient input data used for projecting EOC-7 bobbin voltage frequency are further discussed below.

7.2 Probability of Detection (POD)

The Generic Letter 95-05 (Reference 9-1) requires the application of a constant POD value of 0.6 to define the BOC distribution for EOC voltage projections, unless an alternate POD is approved by the NRC. A POD value of 1.0 represents the ideal situation where all indications are detected. A voltage-dependent POD would a more accurate prediction of voltage distributions consistent with voltage-based repair criteria experience. In this report both NRC mandated constant POD of 0.6 as well as a voltage-dependent POD developed for EPRI (POPCD) are used.

The EPRI POPCD is developed by analyses of 18 inspections in 10 plants and is presented in Table 7-4 of Reference 9-4. The POPCD values applied represent a lower 95% confidence bound, and their distribution is graphically illustrated in ,

Figure 7-1. l 7.3 Limiting Growth Rate Distribution l 1

As discussed in Section 4.2, the NRC guidelines in Generic Letter 95 05 stipulate that the more conservative growth rate distributions from the past two inspections should be utilized for projecting EOC distributions for the next cycle. It is evident from Table 4-5 that the average growth rate /EFPY for Cycles 5 and 6 are  ;

comparable, with Cycle 5 having a slightly higher value. However, Cycle 6 data I includes 3 growth values over 1 volt and a value over 2.1 volts while the growth rates for Cycle 5 are all equal to or less than 1 volt (see Table 4 3 and Figure 4-7).  ;

Hence, SLB leak rate and tube burst probability projection for the EOC-7 condition based on the Cycle 6 data would yield more conservative results; therefore, Cycle 6  !

growth distribution would be applied to obtain EOC-7 projections. )

As noted in Section 4.2, the Cycle 6 growth rates for SGs A and D are higher than the composite growth distribution and, per the methodology recommended in Reference 9 3, SG specific growth rates are to be used for SGs A and D while the i composite growth rates should be applied for SGs B and C. The growth data for l SG.D does not include any of the top 4 growths observed for Cycle 6. Since a few ,

relatively high growth values found in each cycle can be expected to occur l randomly in any SG, it is not considered highly improbable that highest growth for Q Aapc\thx\thx98\taparc\thxcG90d. doc 72

-. .-- - - . - - . . . - - - - - - . ~ - . - . . . . - - . - . - -

the ongoing cycle would occur in SG D. To account for such a possibility and thus provide additional conservatism, the top 3 growth values for Cycle 6 were added to  ;

the SG-specific growth distribution applied to the EOC-7 projection for SG-D.

7.4 Cycle Operating Period The operating periods used in the growth rate /EFPY calculations and voltage projections are as follows.

Cycle 6 - BOC-6 to EOC 564.9 EFPD or 1.55 EFPY (actual)

Cycle 7 - BOC-7 to EOC 374 EFPD or 1.02 EFPY (estimated) .

7.5 Projected EOC-7 Voltage Distribution Calculations for EOC-7 bobbin voltage projections were performed for all four SGs based on the EOC-6 distributions shown in Table 7-2. The BOC distributions were adjusted to account for probability of detection as described above, and the -

adjusted number ofindications at BOC-8 are also shown in Table 7-2. Calculations were performed using a constant POD of 0.6 as well as the EPRI POPCD distribution (presented in Table 7-1). As d.iscussed in Section 7-2, EOC-6 growth rates shown in Table 4-3, were applied. The EOC-7 voltage distributions thus projected for all four SGs are summarized on Table 7-3. These results are also shown graphically on Figures 7-2 to 7-5. In general, results based on a constant POD of 0.6 are more conservative than those using the voltage-dependent EPRI POPCD.

7.6 . Comparison of Actual and Projected EOC-6 Voltage Distributions Table 7-4, and Figures 7-6 and 7-7 provide a comparison of the EOC-6 actual measured bobbin voltage distributions with the corresponding projections performed using the last (EOC-5) inspection bobbin voltage data. The EOC-6 i projections, originally preFented in Reference 9-2, are based on a constant POD of  !

0.6 and the assumpLun that a 1.0 volt repair criteria was applied during the EOC-5 inspection. However, " plug-on-detection" criterion was actually applied in the EOC-5 inspection and, therefore, a large number indications under 1 volt confirmed by RPC were repaired. Therefore, the actual number of indications  !

found at EOC-6 in all SGs are significantly below the projections presented in  :

Reference 9 2. However, as projected SG-B was found to have the largest number  !

. of indications and SG A was confirmed to have the largest number of indications over 1 volt. The actual measured voltages include 3 values above 2.5 volts that  !

l s

Q:\apc\thx\thx98\taparc\thac690d. doc l

73

were not projected. j A comparison of the actual and projected voltage distributions in Figures 7-6 and 7-7 show that in general the indication population above 0.5 volts is substantially overestimated in the projections based on a constant POD of 0.6. This POD value is conservative for voltages above about 0.5 volt but non-conservative below 0.5 volt as seen in Figure 7-1.

Q:\ ape \thx\thx98\tsparc\thxcG90d. doc 7-4

~.-._ _ _

_ . . _ _ _ ._ ._ ._ , _ __ _ . . _ . ~ . . _ . .

Table 7-1 EPRI POPCD Distribution Based on Data from 15 Inspections in 8 Plants Voltage EPRIPOPCD*

Bin 0.1 0.24 0.2 0.34 0.3 0.44 _

0.4 0.53 0.5 0.62 0.6 0.67 l 0.7 0.73 0.8 0.77 0.9 0.81 1 0.83 1.2 0.88 1.4 0.91 1.6 0.92 1.8 0.93 2 0.94 3 0.98 3.5 1.0

1. Data fromTable 7-4 in Reference 8-4.

l-l I

I e

l apnpoped Tem e-i tarmee 12:o2 PM 75

.~ _ .. ._._ .

Table 7-2 (Sheet 1 of 2)

South Texas Unit 2 October 1998 EOC-6 Bobbin and Assumed BOC-7 Bobbin Distributions in SLB Leak Rate and Tube Burst Analyses Steam Generator A Steam Generator B Voltage EOC 6 BOC 7 EOC-6 BOC-7 B'" need sown inac.uo.s indusuuns me,.wed Pop e4 rised name and.c canas inecanens u pan.d POD e4 l mp 0.1 0 0 0.00 0.00 2 0 3.33 8.33 0.2  % 0 8.33 14.71 59 0 98.33 173.53 0.3 40 1 65.67 89.91 125 0 208.33 284.09 0.4 40 1 65.67 74.47 124 0 2 %.67 233.96 0.5 34 0 56.67 54.84 70 0 116.67 112.90 0.6 18 0 30.00 26.87 66 0 110.00 98.51 0.7 12 1 19.00 15.44 28 0 46.67 38.36 0.8 8 1 12.33 9.39 12 0 20.00 15.58 0.9 _ 6 0 10.00 7.41 9 0 15.00 11.11 1 7 2 9.67 6.43 2 0 3.33 2.41 1.1 3 2 3.00 1.51 2 2 1.33 0.34 1.2 3 2 3.00 1.41 0 0 0.00 0.00 1.3 3 3 2.00 0.35 I i 0.67 0.12 1.4 0 0 0.00 0.00 0 0 0.00 0.00 1.5 3 1 4.00 2.28 0 0 0.00 0,00 1.6 0 0 0.00 0.00 0 0 0.00 0.00 1.7 0 0 0.00 0.00 0 0 0.00 0.00 1.8 2 2 1.33 0.15 0 0 0.00 0.00 1.9 0 0 0.00 0.00 0 0 0.00 0.00

,2.3._ ._ _ _ .0 0 0.00_., _ 0 00 _ 0 0 0.00 _ _0.00_ .

l 2.5 I I 0.67 0.04 0 0 0.00 0.00 1 2.6 I I 0.67 0.04 0 0 0.00 0.00

.. 28 _ ._ __

I _9 __ ._ 0J __ 0 0 0 . 00 _.

Total 188 188 188.00 188.00 188 188 188.00 188.00

>IV 18 18 18.00 18.00 18 18 18.00 18.00

>2V 4 4 4.00 4.00 4 4 4.00 4 00 l

l l

I Predcomp Table 1 (3) 1/11/99 9-d6 AM 7-6

i l

i Table 7-2 (Sheet 2 of 2)

South Texas Unit 2 October 1998 l EOC-6 Bobbin and Assumed BOC-7 Bobbin Distributions in SLB Leak Rate and Tube Burst Analyses l

Steam Generator C Steam Generator D Voltage EOC 6 BOC-7 EOC-6 BOC 7 OI" Field Bobbin Indicadons POD Field Bobbin ladicauons POD Indicadons Repaired 0.6 Indicadons Repaired 0.6

_0.1 4 0 _ 6.67 16.67 2 0 3.33 8.33 1 0.2 38 0 63.33 111.76 25 0 41.67 73.53 0.3 96 1 159.00 217.18 84 0 140.00 190.91 0.4 109 0 181.67 205.66 71 1 117.33 132.96 0.5 82 1 135.67 131.26 64 0 106.67 103.23 0.6 60 0 100.00 89.55 29 0 48.33 43.28 0.7 24 1 39.00 31.88 29 0 48.33 39.73

! 0.8 18 0 30.00 23.38 12 0 20.00 15.58 0.9 10 0 16.67 12.35 11 0 18.33 13.58 1 5 0 8.33 6.02 5 0 8.33 6.02 1.1 2 1 2.33 1.34 1 1 0.67 0.17 1.2 3 3 2.00 0.41 2 2 1.33 0.27 l

1.3 _ _ l__ __l 0.67 0.12 _ _ _ _ 1 1 .

_ _0.67 _. _0.12 _

1.4 1 1 0.67 0.10 0 0 0.00 0.00 1.5 0 0 0.00 0.00 2 2 1.33 0.19 1.6 0 0 0.00 0.00 1 1 0.67 0.09 1.7 1 1 0.67 0.08 0 0 0.00 0.00 1.8 0 0 0.00 0.00 0 0 0.00 0.00 1.9 1 1 0.67 0.07 0 0 0.00 0.00 2.3 I I 0.67 0.05 1 1 0.67 0.05 2.5 0 0 0.00 0.00 0 0 0.00 0.00 2.6 0 0 0.00 0.00 0 0 0.00 0.00 2.8 0 0 0.00 0.00 0 0 0.00 0.00 4.2 0 0 0.00 0.00 0 0 0.00 0.00 Total 188 188 188.00 188.00 188 188 188.00 188.00

> lV 18 18 18.00 18.00 18 18 18.00 18.00

> 2V 4 4 4.00 4.00 4 4 4.00 4.00 l

Predcomp Table _1 (2) 1/11/99 9:46 AM 7-7

i Table 7-3 South Texas Unit 2 October 1998 Voltage Distribution Projection for EOC - 7 Steem Generator A l Steam Generator B l Steam Generator C l Steem Generator D Voltage Projected NumberofIndications at EOC-7 0 0 PoPCD PoPCD PoPCO ,

PoPCD 6 ,

O.1 0.11 0.18 1.55 3.20 1.15 2.41 021 0.44 0.2 3.70 5.55 18.90 33.62 13.20 23.72 4.08 7.57

' O.3 22.10 29.83 85.52 131.48 59.61 90.78 28.96 45.17 0.4 48.76 60.25 159.76 210.61 123.79 161.21 80.14 107.40 ,

55.20 61.48 174.92 204.68 151.59 174.61 104.14 125.41 0.5 135.69- ~140.06 99.48 107.72  !

0.6 46.41 46.69 140.98- ~147.69

-'.7~O 32.64 ~56 27 165'.58 98.09 100.50 94.93 7726 76.40 i O.8 21.99 19.04 66.45 58.96 64.17 56.60 55.61 50.89 0.9 15.53 12.66 37.90 32.04 39.00 32.48 38.18 32.73 1.0 11.39 8.67 20.44 16.62 23.56 18.74 25.31 20.69 1.1 8.15 5.72 10.80 8.53 13.83 10.59 15.92 12.57 1.2 5.51 3.52 5.60 4.32 7.79 5.68 9.35 7.15 -

1.3 3.72 2.17 3.00 2.33 4.41 3.03 5.29 3.96

_5 2_,__

_ _ 2 _ , _ , , ,_ . . _ 1 85_ __. _ _. .__ 7 _._ ___.

~

_ 1.6 ' ~ ~ ' 2.68 ' ~ ~~ 1.39 _ 0.8_2 __.0.76~ ~ 1.31 _ _ _0l80 ~ ~'.66 1 1.08 0.75 _

1.7 1.54 0.95 0.53 0.45 0.93 0.52 1.18 ,

1.8 1.19 0.76 0.38 0.36, ,

0.69 0.38 _ . , _ _ 0.98 _ _ _ _ 0.71_ _

1.9 . _ 1.07_ _ 0.77 0.30 _ 0.29 _ _ _

0.55 0.30 _ . _0.87 _ _ . 0.71  :

2.0 0.84 0.57 0.03 0.05 0.44 0.13 0.64 0.50

> 2.1 0.62 0.38 0.00 0.00 0.34 0.00 0.45 0.32 2.2 0.44 0.22 0.00 0.00 0.26 0.00 0.31 0.18 ,

2.3 , . 0.34_. _ . , _ 0.13 0.70 0.00._ _ ,, , _ 020 0.70, _ _ 0.23 _ _ 0.11 2.4 0.30 0.09 0.00 0.70 0.09 0.00 0.19 0.09 -  ;

2.5 0.35 0.15 0.00 0.00 0.00 0.00 0.26 024  ;

2.6 0.59 0.44 0.00 0.00 0.00 0.00 0.46 0.49 i 2.7 0.60 0.22 0.00 0.00 0.70 0.00 0.37 0.11  !

2.8 0.52 0.00 0.30 0.30 0.00 0.30 0.00 0.70 ,

2.9 0.38 0.70 0.00 0.00 0.30 0.00 0.70 0.00 3.0 0.28 0.00 0.00 0.00 0.00 0.00 0.00 0.30 3.1 0.21 0.30 0.00 0.00 0.00 0.00 020 0.00 t 3.2 0.16 0.00 0.00 0.00 0.00 0.00 0.00 0.00  ;

3.3 0.05 0.00 0.00 0.00 0.00 0.00 0.00 0.00  ;

3.8 0.70 0.00 0.00 0.00 0.00 0.00 0.00 0.00 4.5 0.30 0.00 0.00 0.00 0.00 0.00 TOTAL 293.32 296.74 835.36 957.61 748.68 820.96 557.67 608.11  !

>1V 35.49 22.12 25.56 20.62 36.42 25.42 44.30- 33.69

>2V 5.84 2.63 1 00 1.00 1.89 1.00 3.27 2.54 Prodcomp Table _2 (2) 1/11.99 592 N 78 [

S.

I i

l l

l Table 7 4 i South Texas Unit 2 October 1995 l

Comparison of Predicted and Actual EOC-6 Voltage Distributions ,

1 i

l Steam Generator A Steam Generator B Steam Generator C Steam Generator D 1 Number of indicatione l i

EOC4 EOC4 EOC4 EOC4 EOC4 Voltage Prediction Prediction Prediction Prediction POD = 0.6 POD = 0.6 POD = 0.6 POD = 0.6 0.1 0.0 0 1.0 2 0.1 4 0.4 2 0.2 0.2 5 10.1 59 2.7 38 10.1 25 0.3 1.9 40 36.1 125 11.6 96 41.6 84 l

0.4 10.4 40 73.8 124 27.9 100 89.7 71 0.5 23.4 34 105.0 70 53.9 82 106.4 64 0.6 36.0 18 107.8 66 60.0 60 81.2 29 0.7 43.7 12 86.4 28 73.0 24 492 29 0.8 43.5 8 59.3 12 61.6 18 27.3 12 0.9 37.9 6 36.9 9 47.6 10 12.9 11 1.0 30.8 7 20.9 2 34.3 5 5.8 5 1.1 23.9 3 11.6 2 23.0 2 3.5 1 1.2 18.4 3 6.1 0 14.5 3 3.2 2 1.3 14.1 3 3.0 1 8.4 1 2.2 1 1.4 11.0 0 1.4 0 4.6 1 1.4 0 1.5 8.4 3 0.8 0 2.4 0 0.8 2 1.6 6.2 0 0.7 0 1.0 0 0.0 1 1.7 4.3 0 1.0 0 0.0 1 0.7 0 1.8 2.9 2 0.9 0 0.7 _

0 0.3 0 1.9 1.9 0 0.6 0 0.3 1 0.0 0 2.0 1.2 0 0.5 0 0.0 0 0.0 0 2.1 0.7 0 0.2 _

0 0.0 0 0.0 0 2.2 0.1 0 0.0 0 0.0 0 0.0 0 2.3 0.7 0 0.7 0 0.0 1 0.0 1 2.5 0.3 $ 0.3 0 0.0 0 0.0 0 2.6 0.0 1 0.0 0 0.0 0 0.0 0 2.8 0.0 1 0.0 0 0 0.0 0 4.2 1 0.0 0 0 0 TOTAL 321.7 188 565.0 500 436.7 456 436.7 340

>1V 94.1 18 27.6 3 54.9 10 12.1 8

>2V 1.8 4 1.2 0 0.0 1 0.0 1 l

l l

1 l

e,= % wi m m 7-9

Figure 7-1 Generic POPCD Distribution Based on 15 Inspections in 8 Plants

[ Presented in EPRI Report NP-7480, Addendum-1]

1.0 0.9 -

0.8 -

0.7

.! NRC Mandated POD = 0.6 j 0.6 == - -

w 0.5 E

5

.l 0.4 2:

0.3

/

l 0.2 0.1 l  :  :  :

0.0 l l 0.5 1.5 2 2.5 3 3.5 0 1 Bobbin Amplitude w r#mm'""" " 7-10 t

1 Figure 7-2 South Texas Unit 2 SG A Predicted Bobbin Voltage Distribution for Cycle 7 Combined Data for Hot and Cold Leg Indications l

POD = 0.6 ro -

go . .

so - --

e 5

40 - -

O tsCO-7 E Pred EOC-7 l 5 '

m _ _ __

to I' to f-I o.~.I+. I driI[hr.n.. . . _ . _ _ . . _ _ _ _ . .._ _

54S $$$$$UUU0 0IDEUUEN$ $$$$$$5NN$ U Bobbin Voltage EPRI POPCD 80 m

70 -

D BOC-7 gm - -

1:

" so _ .

m Pred EOC-7 3

40 - -

y so .

20 - - -

10 - - - - - -

0

^ "^ # "D'"'---- - - - ^ -

a

a::::::::::::::::;aa::::::;

Bobbin Voltage l

. .o ie.= u. =

7,yy

Figure 7-3 South Texas Unit 2 SG.B Predicted Bobbin Voltage Distribution for Cycle 7 POD = 0.6 250 200 --

lk-

{

r 150 -

D BOC-7 m Prod EOC-7 1

o l

t30 -

Z I I

,w 11.. _

E  : : :  ::  :  :  :  :  :

Bobbin Voltage EPRIPOPCD

! 300 250 D BOC-7 8

E -

150 - - -

E Prod EOC-7 1

M ion _ _ _

So _ _ _ _

'M '

0

^ ' ' - ^ - - ^ --

i  ; a: : : : ::: : : : : : : : : : : : : :

l Bobbin Voltage

"'" 7-12 " " ' * " "

l Figure 7 4 l South Texas Unit 2 SG-C Predicted Bobbin Voltage Distribution for Cycle 7 POD = G.6 goo _

~

180 160 0 80C-7 140 -

E o

i120 o

j u Prod EOC-7 100 _ ._ __

o ao _ .- _

E '

60 - - - - -

do ---- - - - - -

20 - - - - - -

,< I rl am ._ - -

22

: : : : : : : : : : : : : : : : : : : : : :

Bobbin Voltage EPRI POPCD 25o 200 _-

O BOC-7 E

o

$150 -

@ E Pred EOC-7 5

li fioo - _ _

l so - _

E o

^ ^ ^ ' ^ " ^ ~ ^ ^ ^ - '

5 $ = $ 2  : I 2 Bobbin Voltage

?

i

• 7-13 "*"""

1

Figure 7 5 South Texas Unit 2 SG-D Predicted Bobbin Voltage Distribution for Cycle 7 POD = 0.6 140 120 l -

goo _ _ O BOC-7 E

8 so - -

) E Pred EOC-7 I

so - -

i .

.o _ _ _ _ _ _

E w _ _ _ _ _

, n n. .[ [La m . __ _. _. _. __ _ __

e N a *. 9 e N M

• O

• N et *.

• e % e e o e N et

• q e t- e e o o o o o o o o o e e- e e e <- e e e e n n n N u m N N N n Bobbin Voltage EPRI POPCD 200 100 I 160 O BO C-7 140 e

8

%120 o

j B Pred EOC-7 100 - ---

o 80 - - -

E E

n_ _ _ _ ._

l 40 - - - -~ ' --

20 - - - - - -

O n t - I I. t .. _ _ _

r a a *. * * % 9 m o *- N et e * =M ew h ee a<- o e .- e a et

• e o P- e o o o o o o o o o o <- .- e u n w w w n n w w n Bobbin Voltage

• 7-14

" " * " " ~

l

,ta.a ,ah , s_ , _...~i.,A.--4 s # ..ad ., ea .sm,,h-aE m i-h- 1,,s. . . J. 44, . -m-Y -,-4 ..a..audae A- a4 M 4 t

i 1

• en

{

h l N

• e

, b ~

}

4

. O ej . ~

a I I i ~ $

! 2 E

M i _E k- 1

e 1 i ;

< o. O E i

O E h 4

l a l

t'

• i e i

e  ?

l -

M n 4 us I 9 e

l %oiN4 < M r s> T0 g

• c N >

s 1 9 g 'C w c *'

i se5 S munum -3o 2 ]o g

b936 I

I l l *

• 2" i $s g 1 1 "  %

,EI $ M $  ;

($ l *

$ 9 I as s lII  : i g

ml l l l

Ml :

ml IIIlll : l g -

l l l l I I I I I I l'i e ,

II I I _

llIlll l  :

I I I

• II I I ,

l -

l1 O O

R I

- : l5 .

I

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

1 1

J J

• n

" ~

l l 4 I e

~

i J

j en .

O l - e) y

' D e- A i l

' 1 e

< E  : a I

0 R E O 1

3 ('

E 1 0 <

4 u I I

• I  !

' l U i 4D "

. m

, e ti ~s g o>

! e g

c Db -

h O h N E $

j Em3 B  ? E E

E 57 sc g s e e _ .

o S e E E

a y  :

p$ ' E l

i suuruumm:

W

• l l

8 NM  :

4 m- : >

i i i I

"9'====  :

Ii .,

M 1 I

l =o I 3 o

[

! l F

c I  !

8 8 8 8 2 *

= 5 r 3 8 s s e a i l

suopeogput go JeqiunN suoprolput go JequJnN

i. j

)

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

8.0 SLB Leak Rate and Tube Burst Probability Analyses This section presents results of analyses carried out to predict the leak rates and tube burst probabilities for postulated SLB conditions using the actual voltage distributions from EOC-6 inspection as well as for the projected EOC-7 voltage distributions. The methodology used in these analyses is described in Section 6.0.

SG-A with the largest total number of indications over 1 volt is expected to yield the limiting SLB leak rate and burst probability for Cycle 6.

8.1 Leak Rate and Tube Burst Probability for EOC-6 Analyses to calculate EOC-6 SLB leak rates and tube burst probabilities were performed using the actual bobbin voltage distributions presented in Table 7-2.

The results of Monte Carlo calculations are summarized in Table 81. A comparison of the EOC-6 actuals in Table 8-1 with the corresponding predictions presented in Reference 9-2, indicates the following.

a) SG-A was predicted to be the limiting steam generator for EOC-6 based on a voltage distribution projection performed using tho EOC-5 outage. SG-A was conSrmed to have the highest tube leak rate and burst probability based on actual EC bobbin measurements for EOC-6.

b) For the limiting SG-A, leak rate and tube burst probability predictions based on the EOC-5 inspection data are below those obtained with the actual measured EOC-6 voltages. However, the magnitude of the differences are small (~2x10 4 for burst probability and 0.02 gpm for leakage) and they are about 2 orders of magnitude below the acceptance limits for leak rate and tube burst probability.

c) Leak rate and tube burst probability predictions for all four SGs based upon EOC-6 actual bobbin measurements are well within the allowable limits.

In summary, the limiting values for SLB leak rate (0.032 gpm) and tube burst probability (3.8 x 10-4) obtained using the actual measured voltages are nearly two orders of magnitude below the allowable Cycle 7 SLB leakage limit of 15.4 gpm (room temperature) and the NRC repor.. y guideline of 102 for the tube burst probability.

Q.\a pe\th x\thx98\tspa rc\thxc690d. doc 81

8.2 Leak Rate and Tube Burst Probability for EOC-7 Calculations to predict SLB leak rate and tube burst probability for the limiting steam generator in South Texas Unit-2 at the EOC-7 condition were carried out using two values for POD: 1) NRC required constant value of 0.6, 2) voltage dependent EPRI POPCD distribution. Projected results for EOC 7 conditions are summarized in Table 8-2. With the standard calculation methodology presented in Reference 9-3 and a constant POD of 0.6, the largest EOC-7 SLB leak rate projected is 3.3x102 gpm (room temperature), and it is predicted for SG-A which has the largest number of indications over 1 volt returned to service for Cycle 7 operation. This limiting SLB leak rate value is nearly 3 orders of magnitude below the allowable SLB leakage limit for Cycle 7 of 15.4 gpm (room temperature). The highest tube burst probability, also predicted for SG-A, is 4.2x10 4, and it is about 1/25th of the NRC reporting guideline of 10-2, With EPRI POPCD total number ofindications predicted are higher than those for POD =0.6. The reason for this is that below about 0.5 volt, the detection probability calculated from EC inspection data could be significantly below 0.6 as shown by the EPRI POPCD distribution in Table 8-1 and Figure 8-1. Nearly 80% percent of the indications returned to service for Cycle 7 operations are below 0.5 volt.

However, SLB leak rate and burst probability values more strongly influenced by indications over 1 volt and therefore leak and burst values based on EPRI POPCD are all below those corresponding to POD =0.6.

As noted in Section 4.2, the Cycle 6 growth data for SG-A seems to show a dependency on BOC-7 voltage since larger growths (say, over 0.5 volt) occurred at BOC voltages greater than the mean BOC voltage for the SG. To examine the impact of the voltage-dependent growth trend observed for SG A on tube integrity projections, SLB leak rate and tube burst probability projection for the EOC-7 condition for SG-A was carried using the methodology recommended in Reference 9-4 and the results are included in Table 8.2. The Cycle 6 growth data for SG-A was divided into two bins: s 0.4 volt and over 0.4 volt. Since there only 188 indications in the Cycle 6 growth data for SG-A, the Reference 9-4 recommendation that growth bins should include at least 200 indications could not be met, and therefore the voltage-dependent growth distribution utilized herein is conservative.

As shown in Table 8-2, with a voltage-dependent growth assumption, the EOC-7 SLB leak rate prediction for SG-A increased from 3.3x10-2 to 4.0x10 2, and the corresponding tube burst probability increased from 4.2 x10- 4 to 5.5 x10 4 It is evident that the magnitude of increase in both SLB leak rate and tube burst probability are small in comparison to their 1 to 2 orders of magnitude margins relative to their acceptance limits.

Q.\apc\thx\thx98Ntsparc\thxc690d doc 82

_ _ _ __ ~

Additional leak rate and tube burst pressure data are available from the tube l specimens pulled during the recent inspection. An evaluation of the impact of the new data on the leak and burst correlations, described in Section 3.3, indicated that the new data would not significantly affect tube burst probability and the SLB leak rate may increase slightly. In accordance with the NRC-NEI protocol for determining whether the voltage-based repair criteria leak and burst database should be updated to include the latest data, EOC-7 leak rate and tube burst probability calculations for SG-A were repeated using correlations developed in ,

Section 3.3 including new data, and these results are also included in Table 8.2.

While the tube burst probability essentially remains the same, inclusion of the recent South Texas Unit-2 pulled tube data in the leak and burst database increases SLB leak rate from 3.3 x10 2 to 4.5 x10 2 Again, the increase in the SLB l leak rate is negligibly small in comparison to the margin to the allowable leak rate In summary, SLB leak rates and tube burst probabilities predicted for EOC-7 are 1 or 2 orders of magnitude below their respective limits.

l l

l Q \npc\thx\thx98\taparc\thxc690d. doc 8-3 l

Table 8-1 South Texas Unit-2 1998 EOC- 6 Outage Summary of Calculations of Tube Leak Rate and Burst Probability j Based on Actual Bobbin Voltage  !

i SLB l Steam Number Max. Burst Probability Leak Generator POD ofIndi- Volts (2) Rate cations (0 1 Tube 1 or More (gpm)(3)

Tubes ,

l EOC - 6 Projections Reported in Reference 9-2 A 0.6 321.7 2.5 1.7 x10 4 1.7 x10-4 1.4x10-2 B 0.6 565 2.5 9.0 x10- 5 9.0 x 10- 5 5.5x10-3 l

C 0.6 436.7 1.9 6.0 x10 5 6.0 x 10- 5 4.2x10-3 D 0.6 436.7 1.8 2.4 x10- 5 2.4 x10 5 1.3x10 3 1

EOC-6 Actuals A 1 188 4.1 3.8 x 10- 4 3.8 x10- 4 3.2x10-2 B 1 500 1.3 2.5 x10 5 2.5 x 10- 5 1.8x 10-4 C 1 457 2.3 5.8 x10 5 5.8 x10 5 2.0x10-3 D 1 340 2.3 1.2 x10 5 1.2 x 10- 5 4.6x10-4 Notes:

(1) Number ofindications adjusted for POD.

(2) Voltages include NDE uncertainties from Monte Carlo analyses and exceed measured voltages.

l (3) Equivalent volumetric rate at room temperature.

Q-\apc\thx\thx98Ntsparc\thxc690d doc 8-4 l

Table 8-2 South Texas Unit-2 October 1998 Outage Summary of Projected Tube Leak Rate and Burst Probability I for EOC 250k Simulations i l

"* SLB Steam No. of Max. Proba[oility Leak Comments )

Generator POD Indic- Volts (2) Rate ations(l) 1 Tube 1 or (gpm)(8)

More Tubes EOC - 8 PROJECTIONS A 293.3 4.5 4.2x10 4 4.2x10-4 3.3x10 2 Standard leak rate and tube burst B 835.3 2.8 7.8x10-5 7.8x10 5 2.7x10-a probability

'

• EY C 749.3 2.9 8.8x10-5 8.8x10 5 5.2x10 3 Addendum-2 D 557.7 3.1 1.2x10 4 1.2x10-4 8.2x10 a database A 293.3' 4.6 5.5x 10-4 5.5x10-4 4.0x10 2 Voltage-dependent 0.6 E#

• A 293.3 4.5 4.4x10 4 4.4x10 4 4.5x10 2 Updated database .,  !

with present pulled tube data included  ;

i A 296.8 3.1 1.1x10 4 1.1x10-4 6.8x10 3 Standard leak rate I and tube burst B 957.6 2.8 5.8x10-5 5.8x10-5 2.5x10 3 probability meGodolog C 821.0 2.8 5.3x10-5 5.3x10 5 2.8x10 3 .

Addendum-2 D 608.1 3.0 3.1x10 5 3.1x 10-5 6.4x10-8 database 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 Q.\ ape \thx\thx98\tsparc\thxcG90d. doc l

l 85 l

9.0 References 91 NRC Generic Letter 95-05, " Voltage-Based Repair Criteria for the Repair of Westinghouse Steam Generator Tubes Affected by Outside Diameter Stress Corrosion Cracking," USNRC Office of Nuclear Reactor Regulation, August 3,1995.

92 SG-98-01-004, " South Texas Project (STP) Unit-2, Technical Justification for License Amendment to Implement NRC Generic Letter GL 95-05 Voltage-Based Repair Criteria Steam Generator Tube ODSCC,"

Westinghouse Company, January 23,1998.

9-3 WCAP-14277, Revision 1, "SLB Leak Rate and Tube Burst Probability Analysis Methods for ODSCC at TSP Intersections", W,ntinghouse Nuclear Services Division, December.1996.

9-4 EPRI Report NP 7480-L, Addendum 2, " Steam Generator Tubing Outside Diameter Stress Corrosion Cracking at Tube Support Plates Database for Alternate repair Limits," Electric Power Research Institute, April 1998.

95 " Evaluation of Proposed Update to SGDSM Database and Modifications to the Methodology to Assess Steam Generator Tubing Outside Diameter Stress Corrosion Cracking," G. C. Lainas (USNRC) to D. J. Modeen (NEI),

November 20,1998.

9-G SG-97-12-006, " South Texas Unit-1, Cycle 8 Voltage-Based Repair Criteria Report," Westinghouse Electric Company, December,1997.

l 9-7 Letter from B. W. Sheron, Nuclear Regulatory Commission, to A. Marion, Nuclear Energy Institute, dated February 9,1996.

l Q.\ ape \thxsthx98\tsparc\thxc690d doc 9-1 l

t i

9.0 References 9-1 NRC Generic Letter 95-05, " Voltage-Based Repair Criteria for the Repair l of Westinghouse Steam Generator Tubes Affected by Outside Diameter l Stress Corrosion Cracking," USNRC Office of Nuclear Reactor Regulation, August 3,1995.

9-2 SG-98-01-004, " South Texas Project (STP) Unit 2, Technical Justification for License Amendment to Implement NRC Generic Letter GL 95-05 Voltage Based Repair Criteria Steam Generator Tube ODSCC,"

Westinghouse Company, January 23,1998.

9-3 WCAP-14277, Revision 1, "SLB Leak Rate and Tube Burst Probability Analysis Methods for ODSCC at TSP Intersections", Westinghouse Nuclear Services Division, December.1996.

i 9-4 EPRI Report NP 7480-L, Addendum 2, " Steam Generator Tubing Outside Diameter Stress Corrosion Cracking at Tube Support Plates Database for Alternate repair Limits," Electric Power Research Institute, April 1998.

9-5 " Evaluation of Proposed Update to SGDSM Database and Modifications to the Methodology to Assess Steam Generator Tubing Outside Diameter Stress Corrosion Cracking," G. C. Lainas (USNRC) to D. J. Modeen (NEI),

November 20,1998.

9-6 SG 97-12-006, " South Texas Unit-1, Cycle 8 Voltage-Based Repair Criteria l Report," Westinghouse Electric Company, December,1997.

07 Letter from B. W. Sheron, Nuclear Regulatory Commission, to A. Marion, Nuclear Energy Institute, dated February 9,1996.

l l

Q.\apc\thx\thx98\tsparc\thxc090d doc 9-1 l

L __