ML20217Q421
| ML20217Q421 | |
| Person / Time | |
|---|---|
| Site: | Farley  |
| Issue date: | 08/31/1997 |
| From: | WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP. |
| To: | |
| Shared Package | |
| ML20217Q420 | List: |
| References | |
| SG-97-08-004, SG-97-8-4, NUDOCS 9709030075 | |
| Download: ML20217Q421 (109) | |
Text
-
Eg7# Tr gass um tu) L@CahKE528KO-U.M. 412 rd2 WI@9 TO B2059925002 P 02/34
~
1 WESTINGHOUSE PROPRIETARY CLASS 3 SG 97 08 004 FARLEY UNIT 1 i
1997 ALTERNATE REPAIR CRITERIA 90 DAY REPORT a
August 1997 1
i i
e i
4 Westinghouse Electric Corporation Energy Systems Business Unit Nuclear Services Division P.O. Box 158 Madison, Pennsylvania 15663-0158 j
i 9709030075 970029 PDR ADOCK 05000348 P
f30 51 oErr/ it] h5 M3 CW) ENG!tEER1NG-W.M. 415 WB Mf5G TO B2059925002 P.0934 SG 97 08 004 Y
FARLEY UNIT 1 1997 ALTERNATE REPAIR CRITERIA 90 DAY REPORT August 1997 J
l 4
d I
AUG 27 '97 12i35 FR (14 ENGINEE6 HrrO.M. 412 722.5883'TO 82059925002 I'.04/34
^
^
~
FARLEY UNIT 1
+
1997 ALTERNATE REPAIR CRITERIA 90 DAY REPORT TABLE OF CONTENTS Pare No.
Glossary iv 1.0 Introduction 1-1 1
2.0 Summary and Conclusions 2-1 3.0 Farley Unit 11997 Pulled tube Examination Results and Evaluation 3-1 3.1 Summary of Farley Unit 1 Tube Ermmination Resulta 3-1 3.2 Evaluation of Pulled Tube Data for ARC Applications 3-7 3.3 Compariran of Farley Unit 1 Data with Existing ARC Correlations 3-9 4
4.0 EOC-14 Inspec' ion Results and Voltage Growth Rates 4-1 4.1 EOC-14 Inspection Results 4-1 4.2 Voltage Growth Rates 43 4
4.3 Probe Wear Criteria 4-7 4.4 Probability of Prior Cycle Detection (POPCD) 4-8 4.5 Assessment of RPC Confirmation Rates 4-10 4.6 NDE Uncertainties 4-11 5.0 Data Base Applied for ARC Correlations 5-1 6.0 -
SLB Analysis Methods 6-1 7.0 Bobbin Voltage Distributions 7-1 7.1 Probability of Detection 7-1 7.2 Cycle Operating Time 7-2 7.3 Calculation of Voltage Distributions 7-2 7.4 Predicted EOC-15 Voltage Distributions 7-2 7.5 Comparison of Predicted and Actual EOC-14 Voltage Distributions 7-3 q:\\ ape \\ala97\\ala90new.wp5 ji
AUG 27 '97 12:35 FR CW) EtEltEERitG-W.M. 412 722 5899 TO 82059925002 P.05/34 8.0 Tube Leak Rate and Tube Burst Probabilities 81 8.1 Calculation of Leak Rate and Tube Burst Probabilities 81 8.2 Predicted ar.d Actual Leak Rate and Tube Burst Probability for EOC-14 8-1 8.3 Projected Leak Rate and Tube Burst Probability for EOC 15 8-2 9.0 Comparison of Probability of Prior Cycle Detection for 19 Inspections, 8 plants with EPRI POD 9-1 10.0 References 10 1 1
w q:\\apc\\ala97\\ala90new.wp5 iii
m 87 '9712i36 FR (WI DCINEERItG-W.M."412 723 5889 TO'82059925002~
~'P.0654
~
~
1 GLOSSARY 1
-ARC Alternate Repair Criteria ASME American Society of Mechanical Engineers BOC Beginning of Cycle CPDF Cumulative Probability Distribution Func%n DI Distoned Indication EFPD Effective Full Power Day EFPY Effective Full Power Year EC Eddy Current EOC End of Cyde EPRI Electric Power research Institute FS Free Span-GPM-Gallons per Minute ICC
. Intergranular Cellular Corrosion IGA Intergranular Attack IGSCC Intergranular Stress Corrosion Cracking -
INR Indication Not Reportable MAI Multiple AxialIndication NDD No Detectable Defect
- NDE' Non Destmetive Examination NOC Normal Operating Conditions NRC Nuclear Regulatory Commission OD Outside Diameter ODSCC Outside Diameter Stress Corrosion Cracking PI PotentialIndication POD Probability of Detection
-POPCD Probability of Prior Cycle Detection pol Probability of Leakage
- R_C; Tube Location by Row, Column RCS Reactor Coolant System RPC Rotating Pancake Coil
- SEM Semnning Electron Microscope SER.
Safety Evaluation Report SG Steam Generator.
SLB Steam Line Break TSP-Tube Support Plate TI'S -
Top of Tube Sheet i
UOA Unusual OD Phase Angle i
g:\\apc\\ala97\\ala90new.wp5 iv v
-^
~
im M iiT206 FUUiliGINEERIN5dM. 4i5322 5859 36 850599h5dO2~~50ME
~
~
s i
r FARLEY UNIT I-1997 ALTERNATE REPAIR CRITERIA 90 DAY REPORT I
1.0 INTRODUCTION
This report provides the Farley Unit-1 steam generaior tube support plate (TSP)
+-
. bobbin voltage data summary, together with postulated Steam Line Break (SLB) leak
. rate and tube burst probability analysis results, in support of the implementation of b
a 2.0 volt Alternate Repair Criteria (ARC) for Cycle 15 as outlined in the NRC Generic Letter 95-05, Reference 10.1. Information required by the Generic Letter is
- provided in this report including projections ofbobbin voltage distributions, leak rates i
_ and burst probabilities for Cycle 15 operation. The methodology used in these evaluations is consistent with the NRC SER, Reference 10.2, Westinghouse generic i
L methodology _ described in Reference 10.3, as well as the methodology reported in the I
prior ARC reports for Farley Unit-1 (References 10.4 and 10,5). Additionally, results
{.
- from an improved methodology previously submitted to the NRC is also provided.
L The results 'of the non-destructive aramination, leak rate and burst testing, and
[
! destructive avamination of the tube section removed during the current (End of Cycle l
14, EOC-14) inspection (R2C85 SG-C) are summarized in Section 3. Eddy current
~
and repair data for TSP indications are provided in Section 4. The actual EOC-14 voltage distributions a well as leak rates and tube burst probabilities calculated for those distributions are compared with the projections-for EOC-14 conditions performed using the EOC-18 data. A large number of tubes in SGs A_ and C 4
previously plugged in accordance with prior repair criteria were deplugged, inspected and repaired in accordance with ARC by installing sleeves or replugged. Only 10 indications, all below the ARC repair limit, were left unrepaired in tubes returned to i
service in SGs A and C No deplugged tubes were returned to service in SG-B.
l Indication population in deplugged tubes returned to service are included in the Cycle -
15 analyses.-. Leak rates and burst probabilities for the projected EOC-15 voltage
[
distributions are reported in Section 8 and compared with allowable limits.
t i
e q:\\ ape \\ala97\\ala90new.wp6 11 l-
p 2.0
SUMMARY
AND CONCLUSIONS i
j SLB leak rate and tube burst probability analysos were performed for the actus1
~
EOC 14 bobbin voltage distributions and for projected EOC-15 dirtributions. SG-C l
which has the highest number ofindications (1140) as well as the largest indication detected (13.74 volts) was found to be the limiting SG at EOC-14. This finding is consistent with the tube integrity prqjections performed during the last (EOC-13) outage. SLB leak rate and tube burst probability calculated for SG-C using the
[
actual EOC-14 bobbin' voltages are 7.9 gpm and 3.8x10d, respectively, and these values are within their allowable limits for Cycle 14, which are 8.2 gpm at room temperature and 10 8, respectively. These results include the pulled tube test results i
- (no burst at SLB conditions, 0.72 gpm leak rate) for the 13.74 volt indication in
- R2C85. - With R2C85 treated statistically - in the Monte ' Carlo analysis, the corresponding leak and burst values are 7.6 gpm and 3.8x104, respectively. These 1
leak rates are calculated assuming that leakage is not dependent on bobbin voltage j-since the current NRC approved SLB leak rate database does not result in a statistically acceptable leak rate correlation. With the inclusion of pulled tube results for R2C85, the projected EOC-14 leak rates and burst probabilities for all SGs are below those calculated using the actual measured voltages. The single indication in SG-C with an unusually large voltage (13.74 volts at R2C85 TSP 1H) accounts for 93%
of the calculated burst probability.-
Since SG-C has both the highest number of indications as well as the largest
-indication returned to service for Cycle 15, it was projected to be the limiting SG.
The predicted EOC-15 SLB leak rate for SG-C based on the present licensing-basis database and method (constant POD of 0.6 and a leak rate independent of voltage) is 15.4 gpm (room temperature) which exceeds the current licensed limit of 13.7 gpm at room temperature. However, the current limnaing methods are very conservative.
The following shows the reduction in leak rate as a result of methodology updates that are currently under_ review by the NRC:
Current basis - Biased leak rate parameters, POD =0.6
-and no leak rate dependency on voltage 15.4 gym Unbiased leak rate parameters with no voltage dependency --
11.4 gpm Use of voltage dependent POD with biased leak rate parameters without correlation 9.9 gy m i-t 1
l q:\\ ape \\ala97\\ala9onew.wp5 2-1
_ _.. _ _..... _.g g 37897 12:37 FR M ENGIEERING-W.M. 412 722 5889 TO 82059935002 P.09/34 3
T
. eExclusion of French data resulting in a voltage dependant
- leak rate correlatics and POD =0.6 2.7 gp m i
- Use of voltage dependent POD and leak rate correlation 1.6 gpm e
j _
It is seen that any one of the methodology updates results in a leak rate below the j
allowable limit of 13.7. gpm.
Using the current licensing basis methods, it is estimated that a projected leak rate equal to the allowable limit would be obtained in about 305 EFPD. This provides ample time for the NRC to complete reviews of the above noted methodology changes. It is expected that one or more of the method F
changes will be approved prior to reaching 305 EFPD. Alternately, the allowable
- Sia leak rate limit can be increased to accommodate the projected leak rate by a reduction in the coolant activity limit.
i A total of 3074 signals classi6ed as potenilal indications (PIs) were found in the EOC-I 14 inspection. In accordance with a recommendation in Reference 10.7, indications -
with unusual: phase angle (UOAs) were called as PIs during the EOC-14 EC inspection, which may partially account for the increase in PI calls from 2571 in
. EOC-13 inspection to 3074 in EOC-14 inspection. All indications above 2 volts in SGs ;
A and B, all indications above 1.5 volts in SG-C, and in all three SGs all indications E
in tubes with indications above the repair limit, 563 indications in all, were inspected
{
with a rotating pancake coil (RPC) probe and 401 were confirmed as flaws. The RPC t
confirmed indications included 270 above 1 volt. The largest number of PIs,1140 indications, was found in SG-C, ofwhich 662 were above 1.0 volt. Three hundred and '
i nineteen (319) of those were inspected by RPC, and 208 were confirmed as flaws.
One hundred and six (106) indications were found above 2 volts in all SGs combined,
}
of 'which 46 occurred in SG-C, and 85 of them were confirmed by RPC.
No volumetric-type RPC signals were found. Three hundred and-eighty-one (381) i indications were repaired in tubes active during Cycle 14, which includes 85 RPC confirmed indications over 2 volts. Accordingly,2693 of the 3074 PIs wen returned j
to service for Cycle 15.-
t' A large number of tubes in SGs A and C previously plugged in accordance with prior repair criteria were deplugged, inspected and repaired in accordance with ARC by i
installing sleeves or replugged. Only 10 indications, all below the ARC repair limit, were left unrepaired in tubes returned to service in SGs A and C..No tubas were j'
. deplugged in SG-B. So, a total of 2703 indications were returned'to service for Cycle 15.
i_-
The results of the non-destructive - er== Nation, leak and burst testing, and destructive cramination of the TSP intersections of tubes pulled during EOC-14 i
L e.pewmwoon.wps -
2-2 1
1 3
AUG 27 '97 12137 FR (W) ENGINEER!NG-U.f1. 412 722.5889 TO B2059925002 P.10/34 e
inspection are summarized in this report. The addition of EOC-14 Farley-1 pulled tube data has a signi6 cant impact on the ARC correlations used for leak and tube burst analyses, An evaluation is being performed to assess the extent ofimpact of the updated database on EOC-15 leak rate and burst probability projections and it will be reported later, I
s q:\\ ape \\ala97\\ala90new.wp5 23
\\
4 3.0 FARLEY UNIT 1 PtnLEn TUBE EXAMINATION RESULT! AND EVALUATION.
3.1 Summary of: Farley Unit 1 Tube Rramination Results t
[
~3.1.1 Introduction Sections from the hot leg of Steam Generator (SG) C tube R2C85, SG A R29C47 and SG B R6C28 from Farley Unit I were sent to Westinghouse Science and Technology
' Center for evaluation of corrosion indications at the first TSP intersection detected i
by field eddy current inspections during Apdl 1997. The as-received sections included tubing from the top of the tubesheet (TTS) region and the Erst tube support plate
- (TSP 1) region, as well as free span (FS) sections of tubing between the TTS and TSP I and between TSP 1 and TSP 2. Ofprimary interest was the presence of a large OD origin indication at the TSP 1 region of Tube R2C85 which had an unusually large bobbin probe voltage increase over the last cycle of operation, suggesting rapid crack growth. Its bobbin signal increased from less than 2 volts to almost.14 volts during 4
the last cycle of operation. The TSP 1 region of R6028 could not be burst or leak j
tested as one of the tube pull field cuts occurred at the edge of the TSP and is not further discussed in this section. This section describes the tube eramination results for TSP 1 with emphasis on R2C85. The following presents a summary of the more significant haings from the aramination.
3.1.2 Nondestructive Examinations j
Table 3-1 presents a summary of field and laboratory eddy current data obtained at the first TSP, as well as a summa 17 oflaboratory UT and radiographic data. The field information presented is that from the odginal field calls,.as well as that.
L obtained from a review of the data when a difference was noted. For Tube R2C85,
. the TSP 1 region indication had a large bobbin eddy current signal voltage increase in the lab data in comparison to the field data (13.7 volts versus 28.5 volts). This -
usually suggests some tearing of ductile ligameats from the tube pull. Plus Point (+
Point) eddy current data suggested the presence of OD; origin multiple axial 4
4
- indications (MAI), with one indication being significantly larger than the rest,0.85 inch long,96 to 98% deep. There was no significant difference between field and laboratory + Point maximum voltages (11.6 and 11.5 volts). Laboratory UT data also suggested the presence of a MAI similar to that observed by the + Point data.
~
For Tube R29047, the TSP 1 region was originally called NDD, but reevaluation of the field data suggested a 20.25 volt DL No field + Point data was taken. In the laboratosy, the TSP 1 region had MAI by + Point.
i espe \\ats7talason.wps 3-1 g
<w e
n
,.,w w
-cm e,, - -+ -
n,-
We tN *@f 884@ 04 GL% Ws8KLWK@-U.N. da8 T88 9Ws VO 82059925002 P.82/34 4
3.1.3 - Leak, Burst and Tensile Data Elevated temperature leak testing was performed on the TSP 1 region of Tube R?C85 -
at differential pressures ranging form normal operating conditions (NOC) to steam line break (SLB) conditions. Table 3-2 presents a summary of the leak test conditions and results. The leak rates for TSP 1 of Tube R2C85 (13.7V Sold bobbin indication) ranged from 0.130 liters per hour (NOC) to 249 liters per hour (SLB). The leak data suggest that the specimen experienced significant crack opening to cause the large range of leak rates. It was shown later by destructive examinations that the indication had a 0.42 inch long throughwall length.
The TSP 1 region of R29C47 was not leak tasted since the NDE indication was small and shallow with a negligible likelihood ofleaking (confirmed by destructive exam).
Following leak t<2 ting, the TSP 1 regions of Tubes R2C85 and R29C47 were burst tested at roo.n temperature. The TSP 1 region of Tube R2C85 was burst tested using foils and bladders, along with an offset TSP lower region simulaut (placed 2 inches l
below the location ofinterest) and also with an upper TSP simulant placed 48 inches above the lower simulant. The lower simulant was clamped tightly around the tube while the upper simulant was placed with a nominal tube.to tube support plate gap.
This semi-restraint condition simulates the field support conditions and prevents excessive bending moments from being placed on the specimen during the burst test.
i-Table 3 3 presents a summary of the data. The burst pressure was 3,990 psi or 36%
ofits free span (FS) control. The TSP 1 region of R29C47 burst at 9,600 psi. The FS sections of each tube, selected as control tensile specimens, had tensile properties typical of Westinghouse mill annealed Alloy 600 tubing from this vintage of tubing (See Table 3-3).
L The burst tests opened and made visible secondary corrosion within the crevice region
-in addition to the main axial macrocrack which initiated the burst opening. Figures 3-1 and 3-2 present sketches of the R2C85 and R29C47 TSP 1 corrosion visually observed on the burst tested regions which were selected for destructive emnia=tions.
Following burst testing, the TSP 1 region of Tube R2C85 was pulled apart using a tensile machine in order to obtain the axial tensile rupture load for a circumferential fracture face through the secondary corrosion-in the crevice region of the burst specimen. The tensile rupture load measurement supports an ARC with " locked" TSPs. The separation load was 9,200 pounds (versus 12,970 pounds for its free span control) with the circumferential fkacture occurring near the center of the burst q:\\ ape \\ala97\\ala90new.wp5 S.2
~..
ng 27 d 9fi2i3fFRlWIENGINEIRfsG-ihi.'h27225889"TiBTOU995005 i3/3h 1
opening and through the crevice region secondary corrosion. - Much of the secondary
- corrosion had an ICC appearance, i
3.1.4 Destructive h=in=tions i
a For R2C85 TSP 1, SEM fractography was performed on the circumferential fracture y
face created by the tensile pulling of the buret specimen and on the axial burst r
opening which was in the form of a top and bottom fracture face since the tenaile l
fracture face bisected the axial burst opening. SEM fractography was also performed for the axial burst opening of R29047 TSP 1. The tensile fracture face fractography
[
was done to characterin the cross sectional degraded area in the TSP region of -
secodary corrosion, and the axial burst fractography was performed to characterize i
the axial macrocrack which initiated the burst opening. Table 3-4 presents a detailed summary of SEM fractographic observations, including crack profiling, microcrack ligament data and positional information for all burst specimens destructively
[
examined.
Note that all axial burst cracks were corrected for burst ductility increases using a linear correction factor based on visible features that were displaced. In the case of the axial crack for the TSP 1 region of Tube R2C85, the use of a linear correcCon factor may have resulted in an over prediction of the throughwall length. The central region of this crack experienced some non-uniform tensile elongation Oocal necking) caused by the post-burst test tenaile separation.
f Instron charts suggest that approximately 0.07 inch ofplastic elongation occurred for I
the TSP 1 specimen. Presumably, most of the non uniform elongation occurred where the axial macrocrack was throughwall, making the measured throughwall length
-somewhat longer than it was prior to tensile separation. The use of a linear L
~
correction factor resulted in the 0.07 inch of tensile related elongation being spread uniformly over the entire length of the crack.
{-
The axial burst opening for R2C85 occurred at a macrocrack that averaged 76% deep l
over 0.81 inch with a maximum depth of 100% that occurred over 0.42 inch. While j
there were uncertainties involving burst elongation, subsequent tensile necking at the tensile fracture location and th exact location of the crack tcp tip, it is judged likely L
[
that the macrocrack extended slightly above (0.03 to 0.096 inch) and below (0.03 inch) the M crevice region where the macrocrack tips were shallow (34% maximum -
depthi The circumferential tensile fracture at the TSP 1 region occurred through secondary crevice region corrosion that covered a 194* area within the crevice region' 4
j with the corrosion occurring from-the 70* position through 0*/360* to the 236'
- position. Five separate corrosion macrocracks within this 194* area totaled 169'long.
The largest macrocrack was 74' long with a maximum depth of 74%. The measured secondary corrosion degraded 20% of the tube cross sectional area. At the TSP 1 l
region of Tube R29047, the burst opening formed on a corrosion macrocrack-that qAapc\\ak97\\ala90new.wp5 3-3 4
i
~.
-r.,
e---,
i averaged 44% deep over a length of 0.39 inch with a maximum depth of 59%. Table 3 5 presents an overall summary of fractographic data for the TSP indications, For the OD origin corrosion, the intergranular corrosion macrocracks (both on the i
axial-burst openings and the _ tensile fracture face) were co:nposed of numerous intergranular microcracks that were interconnected by ligaments. Most of these ligaments had intergranular features, indicating that they had grown together during plant operation. A few of.the ligaments had tensile tearing-(ductile) features, indicating that they tore duririg tube pulling, during subsequent leak, burst and tensilv testing, during pressure pulse. cleaning (PPC), or during destructive examination specimen preparation. In addition, a number of the fracture faces had j
more than one corrosion macrocrack where the individual macrocracks were aligned a
j head-to-toe with non-corroded metal separating them. These non-corroded metal gaps are not defined as ductile ligaments.
An unusual SEM observation on the tensile tacture face was occasional striation like features in and near the intergranular macrocrack tip at a location where the crack l
was 60 to 65% deep. These striation-like features occurred on the surfaces of tho
-intergranular grains and stopped at the intergranular crack tip. SEM examination of replicas pulled from this surface showed fatigue striations that had a 3.5 to 5.0 micro-inch spacing. See Figure 3-3. The direction of pmpagation was approximately 30' from the tube circumferential dimetion. (30' from a tangential direction to the tube in the circumferential crack plane going from the OD to the ID). No striation-l like features were observed by SEM'and TEM araminatir,ns on the ' deeper axial
[
. (burst) macrocrack. The Farley-1 SGs were pressure pulo cleaned (PPC) during the -
l_
outage prior to the inspection and tube pull. It i.= posibiu that the striations are the
- ~
result of fatigue propagation of the oblique cellular corrosion cracks during the PPC operations. Consequently, the bobbin voltage of the indication'may have been p
influenced by PPC.
' The FS burst control from Tube R2C85 was found to have intergranular corrosion.
SEM fractography showed that the corrosion macrocrack was 0.425 inch long with an average depth of 23% and a maximum depth of 37%.
- Transverse and radial metallographic sections were made to characterize the crevice region secondazy corrosion for the burst specimens evamined by SEM fractography.
Table 3-6 presents a. summary of the metallographic data.
Metallographic examinations showed that tha Tube -R2C85 TSP 1 corrosion morphology' was composed primarily of axial IGSCC with significant ICC components. The overall j
contribution of the ICC to the corrosion morphology tended to decrease as a function of depth. Where the corrosion was deepest, only axial IGSCO was present. The q:\\ ape \\ala97\\ala90new.wp5 3-4 w-1p-W
'M
g g.g corrosion cracks observed at the TSP 1 region were considered unusual cracks because of two observations. First the ICC structure never formed camp ete cells, i.e.,
l the oblique angled cracks extending from the axial cracks did not com act adjacent e
axial cracks, even at shallow depths. Furthermore, these branching cracks tended
-to be straighter in alignment than typical,-possibly because of the following obse:vation. The more important of the two met.dlographic observations, bewever, j_
~
was the presence of transgranular crack components along with the dominant
.intergranular corrosion morphology, See Figure 3-4 for an arample of radial metallography showing transgranular cracking along with the dominant intergranular cracking morphology. These transgranular crack components were i_
found at all depths s,t many locations in the TSP 1 region secondary corrosion, not i
just near the crack tips. In Alloy 600, transgranular corrosien is usually associated with either Pb induced corrosion or with fatigue.
Since ' the transgranular.
!~
components were observed at many separatelocations, depths and orientations within the TSP 1 region, Pb induced or Pb assisted corrosion is auspected as the corrosion mechanism that best fits the observations. (Note that Pb induced corrosion can ca entirely intergranular corrosion, a mixture of transgranular and intergranular corrosion, or entirely transgranular corrosion. But Pb induced cracking is the only
[
widely recognized corrosion mechani=m that can cause transgranular cracking or i..
intergranular corrosion with transgranular components.
i.
l The corrosion present at the TSP 1 region afR29C47 was predominantly axial IGSCC with only minor to moderate ICC components.
L Corrosion morphology also can be characterized by crack density and by the degree j
that IGA components are associated with individual IGSCC, frequently measund by D/W ratios.
The two TSP 1 specimens aramined had'.similar overall crack L
morphology;in the low to moderate crack density range, but the distribution of 1
cracking varied around the circumference at the TSP 1. All regions examined had L
ia moderate association ofIGA with individual IGSCC, with typical D/W ratios of 8 8
to 10.
l.
1 4
l
' A high crack density is de6ned as greater than 100 cracks around the carcumference, a moderate i
crack density as 25 to 100 cracks, and a low crack density as less than 25 cracks around the circumference.
[_
8 A DM ratio is measured as crack depth divided by crack width at its mid-depth. DM ratios in the range of 3 to 20 suggest a moderate association ofIGA with IGSCC, DM ration less than 3 j
suggest a high==~i= Hon ofIGA with IGSCO, and DM ratios greate Aan 20 suggest a low-t association ofIGA with IGSCO.
4 q:\\ ape \\ alas 7\\ala9onew.we 3 r 4.
s
.e n.
y e
.c
- -, -.,. - =
=--
.n.o 27 '97.22 40 FR (W) ENGINEERING-U.M.- 412 ?22 5889 TO 820599F.032 P.16/34 i
= 3.1.5 - Conclusions
- The TSP 1 region of Tube R2C85 bunt at 3_990 psi. When adjusted to operating -
temperatures, the burst pressure would be 3636 pal which satis 6es Generic Letter 95-05 burst margin for 1.4xaPm, of 3584 psi for indications at TSP intersections.
- Leak tests were performed on the TSP 1 region of Tube R2085. Only a small leak rate (0.130 litsrs per hour) was measured at normal operating conditions (NOC),
but a relatively large leak rate was measured at steam line break (SIE) conditions (249 liters per hour). Intermediate test coaditions suggest that ductile ligament 2
tearing or a signi6 cant crack opening occurred between the 1560 and 2255 psi differential pressure conditions.
- The R2C85 TSP 1 region had corrosion indications in the Held (13.7 volts) and laboratory eddy current data. The FS corrosion destructively examined had +
Point and bobbin probe indications when the field data were reevaluated.
- The TSP 3 region of Tube R2C85 had a large axial IGSCC burst macrocrack that was 0.81 inch long, throughwall over 0.42 inch and had an average depth of 76%.
L Secondary corrosion in the crevice agien was distributed around approximately half of the crevice and had a moderate depth (percentage degraded area of 20% -
over 360'. This secondary corrosion was detected by edtr current with a + Point MAI call. Shallow crack tips (34% maximum depth) extended for 0.03 inch both above and below the crevice region. The axial burst opening crack was composed of a-large. number of intergranular microcracks joined together primarily by--
intergranular ligaments with only two remaining uncorroded ligaments.
- All areas destructively raminad had OD origin intergranular corrosion with axial IGSCC being the predominnt corrosion morphology. The destructively arawnned areas included the TSP 1 regions of Tubes R2C85 and R29047. The TSP 1 region of Tube R2085 had deep axial IGSCC with significant ICC components.
The R29C47 TSP 1 region had a similar morphology, but with moderately deep corrosion.
- The coransion morphology is typical of most tubes characterized and used as part of the alternate repair criteria (ARC) database.. However, the TSP 1 region of Tube
- R2C85 included moderately frequent transgranular corrosion component observations, in addition to the dominant IGSCC morphology.
- No microcrack ligament tearing occurred in the throushwall portion of the R2C85 axial burst corrosion macrocrack as no ductile ligaments were present there.
eapesal.s7 sal. son.wes 3-6
- - -. - -. - -.. - - - - - - -. ~. -. ~. _ -. ~. - _. - - -. ~ -.... - ~
AUG 27 '97 18 4& FR GO IDGIEER1NG-W.M. 412 722 5889 TO 82059925002 P.87/34
=
4
- Metsllography identi6ed unusual transstanular corrosion crack components that:
were occasionally present along with the predominant IGSCC which suggests that b
Pb could have played a role in the corrosion development. _While Pb can cause
'intergranular corrosion, as wellLas transgranular corrosion and' mixed. mode F
corrosion,.when transgranular corrosion components em observed, Pb can be suspected as having played a role in the corrosion development. Most of the -
transgranular crack component observations were made at the TSP 1 region of Tube R2085. A chemistry analysis of crack surfaces had not yet been completed.
- The Tube R2C85 TSP 1 region fatigue striations, found at the crack tip of the tensile fracture face corrosion cracks, probably were caused by pressure pulse cleani 2g (PPC) performed prior to the tube pull and before the field eddy current i
i inspection. Tube R2C85 was located near the pulser zone for the pressure pulse cleaning. Approxims,tely 32 hours3.703704e-4 days <br />0.00889 hours <br />5.291005e-5 weeks <br />1.2176e-5 months <br /> ofpressure pulse cleaning occurred. All fatigue striations were observed at the crack tip on the tensile fracture face. The deeper axial (burst) corrosior macrocrack had no suggestions of fatigue.
' 3.2 Evaluatica of Fulled Tube Data for ARC Applications This section evaluates the pulled tube av==ination results described above for application to the EPRI database for ARC applications. The eddy current data is reviewed, including reevaluation of the field data, to finalize the voltages assigned to the indications and to assess the 6 eld NDD calls for det-+ ability under laboratory analysis conditions. The data for incorporation into the EPRI database is then defined and reviewed against the EPRI outlier criteria to assure acceptability for the database.
3.2.1 Eddy Ourrent Data Review.
t Table 3-7 provides a summary of the oddy current data ' evaluations for the Farley peilled tubes. - For the large voltage indication, there is little difference (13.74 vs.
13.55 volts)in the bobbin voltage calls between the field and the laboratory reevaluation. For inclusion of the data in the EPRI database, it R desirable' to -
minhnize analyst variability in the voltage calls since this variable is separately accounted for in ARC applications as an NDE uncertainty. Most of the pulled tube -
EPRI database-has been analyzed-by the same analyst that performed the~ field reevaluation of Table 3-7. Thus the reevaluated field bobbin voltages are applied for
. appheation to the ARC correlations.
enape\\ alas 7talano ps 37
nrT2 P 97 12i41 TRl W EA5TAEERING-W.N X 732 5809 TO 82059925002 P.18/34 e
The field bobbin data for the 6 eld NDD calls were reevaluated to derive the most
- appropriate amplitude measurements, where possible, for these very small signals.
This review indicated that the Sold NDD indications for TSP 1 of R29047 and R6C28 -
could be assigned a bobbin fitw voltage.1The reevaluated bobbin data for these indications are >0.25 and 0.38 volt, respectively. The bobbin response for R29C47-1H -
is a highly distortod signal and has been evaluated as the sum of a 0.18 and 0.07 volt response (Figure 3 5). Evaluation es a single peak to peak flaw results in a-somewhat higher voltage and the voltage value is not reliable for this distorted signal..
The increase in the post pull bobbin voltage for R2085 (13.55 to 28.5 volts) is higher than found for most of the pulled tubes in the EPRI database. The post pull + Point resultr show a ~significant increase in response arouv.' the tube circumference compared to the Sold results. This is attributable to cpening of cellular corrosion as
-a result of the tube pull. From Table 3 4, it is een that most of the ligaments.
between microcracks were corroded prior to the tube pull and only one uncorroded ligament remained in the macrocrack. Thus, the tube pull would not have resulted in tearing _ofligaments to increase the voltage although the pull forces could have decreased contact across the faces of the crack.~ Signi8 cant tearing ofligaments or j
crack opening to strongly affect accident condition leakage or burst capability is not expected for this indication and the voltage increase appears to be principally due to I
opening of cellular corrosion cracks. Thus, the pre-pull voltage measurements are acceptable for ARC applications for the indications given in _ Table 3-7.
3.2.2. Farley-1 Data for ARC Application i
The Farley 1 pulled tube results, as developed above, are sn=marized in Table 3-8.
The indication at R6C28-1H was not destructively====Med since the Held cut was too close to the TSP location and is not included in Table 3-8. The measured leak rata data of Table 3-2 are a4ustad in the table to the reference normal operating and SLB conditions by applying the leak rate a4ustment procedure of the EPRI database report (Reference 11.5). The reference SLB conditions are a pressure differential of 2560 paid at a primary pressure of 2575 ' psi and a secondary pressure of15 psi at a l
temperature of 616*F. The measured burst pressures are adjusted to the namin=1 flow stress of 68.76'ksi for 7/8 inch tubing at operating temperature. A tensile rupture force measurement was also made for the TSP 1 indication on R2C85 as a basis for updsting the correlation supporting a higher voltage ARC when limit ~t TSP dispikeement is demonstrated. The data ofTable 3-8 are used in Section 3.3 to assess
=
their influence on the EPRI ARC burst pressure, SLB probability ofleakage, SLB leak rate and tensile rupture force versus voltage correlations.
q:\\apeiala97\\alatonew.wys 3-8
suu FP@fl2ii3]R (W) @G80GERING-W.M. 4&a 722 5889. TO 82059925002 P.19/34 The Farley 1 pulled tube results 'were evaluated for potential exclusions from the database against the EPRI data exclusion criteria. Criteria la to Ic and le apply primarily to unacceptable voltage, burst or leak rate measumments and indications without leak test measurements. Criterion la excludes data for a corrupted bobbin
- signal. The signal for R29C47 is a highly distorted low voltage signal which results in an unreliable voltage but the cause is likely crevice conditions and not damage to the indication or an extraneous response, as intended for exclusion per Criterion la.
The voltage unceri.ainty for this indication was noted in the blind (before destructive exam) eddy cum:nt analyses of Rsference 10.10. However, the R29047 1H indication does not strictly satisfy the guidance of Criterion la for exclusion from the correlations and is further evaluated below against exclusion Criterion 2a.
The remaining criteria do not apply to the indications of Table 3-8 and would not lead to any exclusions from the database. Criterion Id applies to potential tube pull damage I
but is not applicabic based on the discussion given above relative to tube pull damage and the associated bobbin voltage increases. Criterion 2b applies only to indications
> 20 volts which is not applicable to these indications. EPRI Criterion 3 has net been approved by the NRC and is not applicable although application would not lead to exclusion of the indications.-
Criterion 2a epplies to atypicalligament morphology and states that cracks having 5 2 uncorroded ligaments in shallow cracks < 60% deep shall be excluded fmm the database. The indication for R2C85-1H is throughwall and this criterion is not applicable. The bobbin indication at R29047 1H has one remaining ligament and has a maximum depth of 59% Thus, Criterion 2a requires exclusion of this indication
- from the database.
Based on the above assessment, the indication at R2C85-TSP 1 is included in the EPRI database for the burst. probability ofleak, leak rate and tensile rupture force -
carelations. The indication at R29047-TSP 1 is excluded from the database and a correlations based on data exclusion Criterion Ea.
3.3 Comparison of Farley Unit 1 Data with W=g APC Correlations This section reports on the evaluations performed which utilised the results ofles.k -
rate and burst testing of the tube section which was removed firem Farley Unit 1 in 1997. The results of the destructive aramination of the tube is recorded in Section
?
3.1 of this report. The Farley 1 pulled tube, Al-R02C85-1 data germane to the APC
~
correlations, and the bobbin amplitudes for APC applications, are given in Table 3-8.
The results of the destructive mminations, e.g., leak and burst tests, are compared to the database of similar test results for 7/8" outside diameter steam generator -
qnape\\ala97\\ala9onew.wr5 3-9
f
-tubes. The reference _ database consists of the EPRI recommended database as described in Reference 10.8, It is noted that the database was modified to correct the data associated with J1-R05C28 2 relative to the bobbin amplitude (4.58 V), the burst -
pressure (>5.918 kai) and the flow stress (-140.5 kai). Two options for the comparison exist-because of the additional pulled tube data associated with A2-R34C531, evaluated in February,1997. The options are:
i 1)
Treat the reference database as the one that existed prior to the availability of the A2-R34053-1 data and base the comparison on the addition of two data points, or, 2)
Treat the reference database as that of1) plus the R34C53 data and base the comparison on the addition of only the R02085 data.
Both options can be argued to be appropriate. Since the purpose of the evaluation is to determine if the formal database should be updated, the net effect of the data from both tubes, i.e., option 1), should be considered. ' Alternatively, the NRC staff position regarding new data is that it should be added to the evaluation database when it is verified that the data do not meet any of the exclusion criteria. Thus, the l
R34C53 data should be considered to be in the reference database as of February,
-1997, i.e., option 2), regardless of the status ofpublishing a formal update. However, t
option 2) implion that informal updating of the database should occur every time new data become available, which is contrary to efforts aimed at establishing formal-criteria to govern when the database should be updated and published. For this evaluation, the option 1) comparison *vas made for the formal analysis. The option 2)_ comparison was also made, with the results being included in the discussions of '
t the analyses instead of being tabulated and depicted on the figures.
Comparisons with and without French data excluded from the database were per-formed because the NRC staff has not yet responded to a recommendation to omit the French data. In addition, the effect ofincluding the new test data in the reference database was evaluated. _ In summary, the test data from both specimens are consistent with the database relative to the burst pressures, albeit low, and-the probability of leak as a function of the bobbin amplitude. However, the R02C85 specimen leak rate is significantly higher, more than twice the previous maximum value, than the prior database values for the leak rate as a function of bobbin ampli-tude. The compansons and evaluations are discussed in what follows.
4 espes.hsm1. son
..ps 8-10
AUG 27797 32:43 FR (WT ENGINEERING-U.li.'412 723 5889 TO 82059925002.
P.21/34^
3.3.1 Suitability for loclusion in the Databate
- The aported information from the destructive ernmination of the degraded R02C85 tube ~section was reviewed earlier in this report Section relative to the EPRI guidelines for inclusion / exclusion of tube specimen data (R02085 is included in this report, and R34053 was the subject of a previous report) in tho alternate repair
. criteria (ARC) database. This review revealed no information that would lead to conclusion that the data should not be included in the database. Therefore, the resulting correlations should be considered applicable to the use of ARC for indications in' 7/8" diameter tubes in Westinghouse SGs.
3.3.2 Burst Pressure vs. Bobbin Amplitude The results' from burst testing of tube sections which exhibited non sero bobbin amplitudes at a TSP elevation location were considered for evaluation. A plot of the burst pressure of the Farley specimens is depicted on Figures 3-6 and 3-7 relative 4
to the burst pressure correlations developed using the reference database with and without exclusion of the French data.
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 reference
- data regardless of whether or not the French data is included.
2._
The data points fall above the lower 95% one-sided prediction bound of the regression line, hence no statistical anomaly is indicated. This also means that both data are above a simultaneous 95% one-sided prediction bound.
In summary, the visual aramination does not indicate any significant departures &om the reference database, although the burst pressure of specimen R02C85 is less than 1
would have been expected from such an indication since it is well below the mean of the data as given by the regression line. This is true regardless of the consideration.
of the R34C53 tube data.
Since the Farley burst pressure data were not indicated to be kom a separate popula-tion from the reference data, the regression analysis of the burst pressure'on the common logarithm of the bobbin amplitude was repeated with the additional dataL included. A comparison of the regression results obtained byincluding these data in the regression analysis is provided in Table 3-9 and Table 3-10. Regression predic.
The use of multiple tables and figures for the same evaluations corresponds to excluding and not er:luding the French data from the reference database.
-q:\\apc\\ala97\\ala9onew.wp6 3-11
AUG 27 '97 tal44 FR (WJ ENGINEERING-U.M. 412 722 9889 TO 82059925002 P.23/34 tions obtained by including these data in the regression analysis are also shown on -
Figures 3 6 and 3 7.- A summary of the changes is as follows:
The intercept of the burst pressure, PR, as a linear function of the common 1.
logarithm of the bobbin smplitude regression line is decreased by 0.2%, or about 16 pal (regardless of the French data) This has the rifect ofdecreasing the predicted burst pressure as a function of the bobbin amplitude.
2.
The absolute slope cf the regression line is increased by 1.5% to 1.3%, i.e., the slope is more steep, with and without excluding the French data respectively.
This has the effect of decreasing the predicted burst pressure as a function of bobbin amplitude for all indications when coupled with the decrease in the-intercept value.
There is an increase in the standard error of the residuals of1.5% regardless 3.
of consideration of the French data. The effect of this change is reflected in a larger deviation of the 95% prediction line from the regression line.
The net effect of the changes on the SLB structural limit, using 959W95% lower toler-ance limit material properties, is to decrease it by 0.r. V, i.e., from 8.7 to 8.2 V, a decrease of 6%, and from 9.6 to 9.0V with and withet exclusion of the French data-respectively. The decrease in the intercept and Ge increase in the standard error coupled with the fact that the structural limit is also decreased indicates that the probability of burst would also increase for bobbin indications over the structural range ofinterest. Based on tha relatively small, but not minor, change in the struc-tural limit, the increase in the probability of burst would also be expected to be small, but not insignificant.
When the reference database is considered to include the R34C53 results, the effect of the addition of the R02C85 data is to reduce the intercept of the burst pressure correlation by about 8 psi or 0.1%, and increase the absolute value of the slope by 0.1%. The net result is a decrease in the structural limit from 8.6 to 8.2 V. Thus, the change in the structural limit due to the inclusion of the R34C53 data alone is about 0.1V, and that due to the-R02C85 data about 0.4V. h=ination' of the burst pressure plots confirms that the R02C85 data point would exert more leverage on the regression line because it is farther from both the regression line and the centroid of the volts data than the R34C53 data point. Similar results differences are obtained from consideration of tl= database with the EdF data included.
qnape\\ala97\\ala9onew.wp5 -
3-12
'AUG 27 '97 12844 FR dW)'ENGIE ERING-W.M.74s2 723 5889 TO 82059925002 P.23/34 c.
3.3.3 Probability of Leak The data of Table 3-8 were a-==Med nlative to the reference correlation nor the pol -
as a function of the common logarithm of the bobbin amplitude ~ Figures 3-8 and 3-9 illustrate the Farley 1 data relative to the reference correlation.. The specimen exhibited expected pol behavior without including the French data, and greater than the expected including the French data. The prad_isd pol for a 13.6 Vindication is 0.95, thus, roughly nineteen of twenty 13.6 V indications would be' expecte6 to leak.
With the French data, the pol for a 13.6V indication is about 0.60, or roughly two of three 13.6V indications would be expected to leakc In conclusion, there is no sign 16 cant 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 Farley 1 data point 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 3-11 and Table 312. These results indicate:
i 1.
A 0.0% and 3.0% increase (smaller negative value) in the respective' logistic intercept parameters with and without exclusion of the French data in the analyses.
2.
A 1.8% increase and 1.1% decrease in the respect.ive lqgistic slope parameters without and with the French data included in the analysis.
3.
The absolute values of the individual elements of the covariance matrix of the parameters changed by -8.3% to 1.6%. Examination of Figures 3-8 and 3-9 indicate that it is likely that there would be a minor impact on the 95% confi- -
dance bound on the total estimated leak rate from a single SG.
4.
The deviance and the Pearson standard error were not aignificantly affected regardless of choice of reference datebase.
In order to confirm thejudgement that the changes are not signi5 cant, the reference '
correlation and the new correlation were also plotted on Figures 3-8 and 3-9. An av==Mation of Figures 3-8 and 3-9 illustrate a minor change, ~2%, in the pol-correlation if the reference database regardless ofincluding of all of the French data.
It is noted that when the total leak rate i latermined using the leak rate to bobbin volta 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 or insignificant relative to the calculation of the 95% confi-
. qnape\\ala97\\ala90new.wp5 3..
Aus M M X T K 4 9Rl WI ENG'hEERING M M."482 702 5889 TO'82059925002
~ '. M 4
^
P L
,L a
1 dence bound of the total leak rate from a SG. hn the leak rate is considere o
independent of the voltage, the increase in pol directly affects the estimated total i
leak rate,;but, since the change in pol is small, so would be the change in the i
- estimate of the totalleak rate.-
l J
The effect_ of adding the R02C85 data to a reference database expanded by the addition of the R34053 data is much less pronounced than the effect of adding the -
data frian both tubes. This is because the padicted probability for the indication to F
leak was high, about 95%. It is also noted that the bobbin amplitude is higher than all of the indications that did not leak. Both correlation coefficients increased absolute magnitude by about 1%. This is because the effect of the leaking R34C53
{
tube with an amplitude on the order of 7 V is more significant than ths effect of the leaking R02C85 tube with an amplitude of almost 14 V. In summary, the results b
from the R02C85 are in line with expectations regardless of the addition of the L
R34C53 data to the reference database. Furthermore, most of the changes in the i_
predicted pol curves as illustrated on Figures 3-8 and 3 9 are due to including the R34C53 data.
3.3.4 Leak Rate vs. Bobbin Amplitude The R02C85 specimen uhibited a leak rate corresponding to 164 lph for SLB temper- -
ature and pressure difference conditions when condensed and measured at ambient conditions. Without the French data, the correlation ofleak rate to bobbin voltage exhibits a p-value ofl.6% for the slope parameter using the reference database, and 2.0% with the addition of the two Farley data. A summary of the regmssion analyses L
results is provided in Table 3-13 and Table 3-14. With the French data included, the correlation ofleak rate to bobbin voltage exhibits p-values of 6.4% and 7.6% for the L
.same comparison. Based on the requirements stipulated in NRC Generic Letter 95-05, the use of the correlation in performing Monte Carlo simulations to estimate -
the total leak rate would not be considered to be jusCfied if the French data are.
retained. Figure 3-10 and 3-11 illustrate the new data points relative to the distribu-tion predicted ' median and mean using the reference database.
Also illustrated on both Egures is the relation of the data points to the regression fit-and to the expected leak rate (mean of the log-normal distribution) based on the regression analysis of the huk rate on the bobbin amplitude. The R02C85 data value is significantly higher than the median and mean prediction lines with and without i
excluding the French data. The effect of including the two data points in the database on the estimated parameters of the leak rate distribution are tabulated in
_ Table 314. The estimated mean and standard deviation of the population oflog leak i
- rates are increased by 8.3% and 3.6% respectively. The mean and standard deviation 4
4 q:\\ ape \\ala97\\ala9onew.wp6 8 14 4-i
Y of the sample leak rates increased by 36.1%, from 13.3 to 18.11ph, and 66.9%, from
- 20.8 to 34.61ph, respectively. This means that the predicted leak rates from Monte Carlo simulations and the 95% confidence bound on the total leak rate from a single SG will also be increased. Since the Monte Carlo simulations effect a transfor from the logarithm of the leak rate to the actual leak rate u.ing assumed population parameters, the expected mean leak rate from the simulations will be greater than
- 22.0 and 27.91ph for the two databases respectively, see Table 3-14. Furthermore,
- the simulated standard deviations of ths leak rates will be greater than 128.0 and 15.81ph respectively. Unbiased estimates of the population mean leak rates are about 15% lower than listed here, and the unbiased estimate of the standard
& untion is about 50% lower than listed. Ia st amary, the simulations will result in predictions signi5cantly in excess of the observed leak rates.
It is apparent that the significant effect on the leak rate from the addition of the data from the two tubes to the database is due almost entirely to the R02085 data. Tids
-is easily recognized by aramining the relative positions of the results from the two tubes on Figures 3-30 and 3-11. The R34C53 point is almost saaetly on the median leak line on both plots,'while the R02C85 leak rate is larger than any previously observed leak rate from 7/8" tubes.
When the French data are omitted the comparisons may be made on the basis oflinear correlation parameters because the p values are less than 5%. The result is that the shifts in the correlation lines as -
illustrated on Figures 3-10 and 3-11 are due almost entirely to the inclusion of the R02085 leak rate data.
3.3.5 General Conclusions The review of the effect of the Farley 1 & 2 data indicates that the burst pressure correlation is slightly shifted toward lower burst pressures at constant voltage. The -
probability ofleak correlations to the common 1 garithm of the bobbin amplitude were not significantly changed by the inclusion of the data from the R02C85 tube section.-
The mean and standard deviation of the leak rate data were significantly increased for indications that leak by the inclusion of the R02C85 tube section data. Therefore, it is likely that the conclusions relative to EOC probability of burst would not be significantly affected by the addition of the Farley 1, R02C85, and 2, R34C53, data.
The~ total leak rate predictions based on the use of the reference database would be significantly increased by the addition of the Farley 1 data from R02C85-1. If the-NRC staff concurs that the French data should not be in the database, the net effect of the changes would be arnae+ad to be not significant because the use of a correlation of the leak rate to bobbin amplitude would be permitted.
I easesalas7\\.laso
.wes 3-15
(M k
i I!
. Table 3-1. Farley Unit 1 Stease Generator Tube Emman RIDE filusmemory
.g
,-+----n Field Eddy C w;'
I " _y Eddy Carreset I ~"
UT E^'
{.:
1
- D
-1 R2C85 Bobben 13.7V OD Ind.
Bobben: 25.5V. 92% deep OD sed.
OD MAI. largest 0.5"long EAI TSP 1'
+. Point-OD MA1 (priweary led.: 0.85 long. 985 deep.
+ Pw: OD MAI (prietary led. II.5Y, 0.9'* long &
i1.6V) j 96% deep & -0.85" long) 100% seen.
y j;
R29C47 Boldenn: NDD (>0.25V OD DI)
"2i 9.69V G hy UD Ind.
No Data MDD-
' g-TSP 1
+ Point: no data taken
+ Point: 0.18V MAI Y'
- Field data with laboratory reevalestion in parenthesis if significasetly difIesent feene field eveleasson.
-v; Table 3-2. Farley Unit I task Test Data F
!il Specinera Test Type: ILwe leak Rate ("A ~.r)
Test C " ~ _
h Petsaute (psi)
Pp (psig) Pt (psis) Tp (*F) Ts (*F)
< ib R2C55 ssr 1 NOC: 1560 0.130 1566.
.%~
(13.7V field 11 01: 2255 89.4 2290 35 y
~ bobbin Ind.).
TIE 2: 2453 166 2550 97-l SLB: 2563 249 2729 10 p
i NOC = norenal operateeg s',,
Ett. = A esate test conditions: SLB = seeane line break; RT = roorn seenperseure.
h Roent Temeyerature Berst and Temaile Data for Farley Unit 1 S/G Tuldeng Table 3-3.
.O Speciesen
. Borst Pressure.
Borst Burst Opemeng Borst Openseg 0.2% offset Teasde Teasete psig Deceility. %
Lengde incites Width, inches Yield Strenge6 Uhesesse Elongseien. %
psi Stre=g88=. psi 1
R2C55. F5 Control (sosse conesson, 11.000 15A 1.153 0.265.
59.600 99.100 23.3 h
intended to have no corrosion)
B[
ii M ).Isrl' 3.990 C.0 0.592 0.145 R29C47. F5 Control (segligible 13.200 2E.R.
2.14 2 -
9.403..
66.500 I16,400 35.8 i
f cosrosion intended to have ao corrosion).'
., j RZ9C47. sar i 9.600 10.0 1.157 0.250
-
- T-:-^ ' t1h fails and M=Mm -.;J into the speciesen Je sianulant was placed 2 inclees below the TSP or M region and a loose support place siseulent was plac N[
5
- experienced in a SG.
q:\\ ape \\ ale 97\\ ale 90new.wp6 3-16 I'!
a
- 2-
= - - -
r 3
+
Table 3-4, SEM Fractographic Data for Farley-1 Macrocracka 4 ' e-. IAcation Radial Pwitsen (*)C@. (% h G.- ::)
I'
= cent Dame 2M i '_
^ Data J
R2C85' Crack l Crad 3 (cont.)
Crack 4 (coat.)
Crack 5 (cons.)
Macrocre:k i F=a-dr T
4B-3 236#10 29907 L 34W50 32n4 Frase 236'so 261*
- d Center-TSP I 237/43 302M3 -
346f22 3?!71 Ciscomferensial 240f59 304 ADO 34W71 3 &68 Macrocrack 2 Easends M
Tensile Practure 24352.
Crack 4 350R10 3858 Frasa 286* so 293*
(Post-Barst) ~
245M8 '
305J00
. Crack 5 41/M
-]
24848 307S 1-356A10 4442 Macrocrack 3 Extends.
J
- S.
25065 310#43 357G7 46f56 Frem 295* to 304*
25345 313/47 35942
' 49f50 ip
, ci) 256/59 315/37 2M 51/56 Macrocsack 4 Extends 258/50 318/40 5/68
'54/31 From 305* fo 350*
261/00 320f43 7M 57M0 s
Crack 2.
32345 10f9 59f47 Macrocrack 5 Extends -
b
~
284/00 326f50 12/22 62/43 Fross 356' su 360* (0*) so 70* -
L.b 287/31 328/74 15/40 6 &47 i
289f3 (Axial Burst Lecasion) 18/12 67/50 No Ductile Ligassenes oi
?-
292143 331nt 20f56 70m 293AIO 33344 2348
,M Crack 3 336/74 25M2 g
295/D0 339f47 2842 su 297/31 341D4 31M5 Linear Average Dep:h For Crack I = $4% Over 25*
g Linear Average Depth Fo(Cracir 2 = 16% Over 9' Di
- Linear Average Depth For Crack 3 w 28% Over 9' Linear Average Depth For C.
k 4 = 48% Over 45'(Maaimura Depeh = 74%)
Linear Average Depth for Crack 5 =.46% Over 74* (Maxienom Depth = 74%)
-t l> era!! Extent of Cracking = 169*
PDA (percentge degraded ares over 3600) = 20%
$i 4
s!$
~
q:\\npe\\nla97h;a90new.wp5 3-17
~
_-... ~...
I Tame 3 4 (cess.). SID4 FraceegrayMe Dean for Farley-I Macrocracsss L.
i'
~,_
~.
s
[
T6 L - J l (Asee = inches': Os;asestem of IJammans Miesr Amis Reemise o h0maenesst Maior Amis c
in Depees: Onserssium of tigamed Major Amis Reesesse to liste Rashes he Dogmes) l4 RM 000f34;;'
OA0fl0D Top d the Tren=== Imsv s 0.03 d re: Abese the Tay at aar 1 1~
V ^ __.48-2ASA 0.02f28 0.42f909 g
g Buist Opening 0.0er12 0.44/300
-g
, T3P 1 :
0.0601 0.46/108 4
- i y
(AAer Testsile Test *)
GAOf55 0.48fl00 tet: Asse = L3 m It'ince8 3
j 0.1842 0.3C/300 f
Mayor Amis 9 0*; Minor Amis @ W 6
a m
- t 0.82f81 -
0.521100.
0.1447 0.54100 E 2: Asee = 2.2 m 1G* inci' I-2 u - =s e.36fs00 asger Amis e 0*; Minor Amis @ W s
~~
0.1441 0.5E/100 jgi I'
O.IWO4 0.60P800 g
0-O_
g (0.28 4 9 61/I 4 O.22/000 t I Ligament L
0.2er00 0.60s7 2
0.26ftes :-
0.6004 s
O.2ar:0o 0.asma E
R30fl00 OJGI83 g
1 0.32fl0B 03204-m 0.34fl00 034G4 g
0356/I00 0.Mf25 3
Tesinde Itacouse 035/96 036(100 -
0.30 0 -
0 Crack Boeoen (T acased 03E Inch Below flue Tay of TSP I) g 038/100 0.8141 Ut 036fl00 0.SGr3 3
038/900 0.884 U=-_ Average r,'_ = 76% (========*e ashem e-y OA2Incit)
"Deroughwell For 0.42 Inch j
Macsocraci Length = 0.8I Inch (crack extends beyoed te top of te f.-
,_ f_1 specission) 1 j
%near everase coneesien made far seesde ses esonssam.
.Y j
- omente lisseuses eahibst disiple supense resswes oser ease see 50s er seassh.
i
'"A saw cut was made a the tap of tis specumee ese wees duough he top pasion of he cseck. No cracking was c6 served y
j:
c' ^_,.." : does it is beiseved clus sine macrocsack h ey so 0.GM ised abooe te T3P 1 sessen.
e =,c\\eherr m.wys 3 i
Tame 3-4 (Cent.) SEM Frw's. ;"- Desa for Farley-1 Macrocracks w _.m m.=. -. u.a, ma a m u e--
u OncheeP5) ThroughwaR (Area e inches'; OnesserSon of Ligarness Maar Amis Reimmige es Macsecract Major Amis la Degvees; @---
_ af ry_.4 Major Amis Belasswe te lWe Radius im N-)
R29C47 0.002 Crack Top W ; 0.24 Inch Bedow Top of T3P I)
Specimen 4B-2A 0.02/37 y
Bwst Opening 0.04M3 4
TSP 1 0.0659 3
0.00/43 LI: Area = 0.6 a 10
- inch
- 0.lW37 Major Amis @ 25*
5 m
0.12/53 Minor Amis @ 90*
O.14/53 c6 0.16f56
~,
0 18/59 f$l 0.2W53 g
Li Ligenneet b
O_22N7
.b 0.24MO -
3 0.26/50 0.2853 a
"so 030N7 032M7 9U 034/43 036MG 038/25 039M)
Osck Bostoni (Lecated 0.63 Incis Below Tcy of TSP 1)
U l
O8 Linear Average Depth = 44% (snesswenseats taken every 0.02 Inch)
Macrocrack Lengde = 039 loch U
Manicium Depth = $9%
h
' Length corrected for borst deceAiry.
- Dtictile ligaments exhibit di aple rupeare fearates ever siere seen 50% of lengst.
Us N
egaspe\\ala97\\ala9Cnew.wys 3-19
Table 3-5. Sommeery of SEM Fractographic Data Obtained on Fadey Unit 1 *ISP I=dEc=tions
^
h Specanien Type of Fractsee Face Maumment R R.(%
Average Rf. (%
Macrocrack Leegih ifwoughwaff)
C.x,* wall)
Onciees)
RMS, adr i amial burst opemog 100 (for OAZ")
76 0 51 R2C55,13P1 Cm ' - J tensde opeemg 74 20 (over 367)
I67 U
R29C47. TSP 1 antal 6m. gung 59 44 0.39 A
)
n Table 3-6. Meta"q. fl; Data froen Farley Unit 1 *ISP Endications 5
Man. Depen of KE Bseimseed Mas 1Arg.
Transverse and Arial Ag DfW Q
Neect Section blanirmum &mber Depth C -,... (%
Ratio from cf tengde Cracks of Cracks at Mid-(%
Throug6waB le Raed Transverse Di Specames Sectrom Type Cracks (tech) per hoch ma Location h - J.-Q Section)
? :--
b W5 4B Tr-i--
t7 2.44 1.0
-25
+ (T3P8)
Rastal I 6
0.29 21 10 F
depch = 4 %
KC < 50% deep Radid 2 9
0.30 30 depth = 12%
AxialIGSCC2 50% deep b
Radid 3 7
0.30 23 deysh = 22%
Radal 4 7
9.30 23 depsh = 33%
Radial 5 2
0.30 7
depth = 30 %
y RZ9C47 Trarrsverse 23 2.56 10
-30 5507 48 (TSP 1) 5
't 4
o nM 8
N
.w$
bf I
gnape\\ala97\\ala90new.wp5 3-20 L
Table 3-7. Sosneser, of Forfey-1 Fissed Tube Eddy Current Resutes Tube "I5F Fleed Ca5 Lab. Reevelusales of Meid Past Ftsit Deam 3
Bobbien
+ Pelat Bobbin Depth
+ Peine Bobbin
+ Point Volis'8 Vehr Volts
Yeles Yelts Vales d
R2C85 SG C 1
1334 11.5 13.55 97%
I t.5 28.5 31.5 e
R29C47.SG A 1
NDD No Data
>0.25'"
DI No Data 0.69 0.15 m
R6C28.SG B I
NDD 0.26 038 50% :
0.22 3.9 x
i a
Notes:
5
- 1. Schbia voltage data inciede cross calitration of ASME standard to the ih laboratory standsed.
- 2. Bobbin voltage not sefiable doe to distortsca of the signal. Value given as som of 0.18 and 0.D7 voit calls.
Table 3-8 Feeley-I Ftsted Tube Deen for ARC Applicaeless 5
.f TSP Bobbin Data Des'.reeseve Emmen Besidts Leak Race 4hr Burst Pressere Data -last
?
'Debe RFC Max.
Avg.
Crack TW No.
N. O.
SLB Mees.
Mees.
cr/o, Adj.
ABC 1
Veks M
DeptIn Depth Length Lengte Lig "
1300 2544 Burst Tensee Burst Use a
psid'*
paid'*
Press.
Feste Pr#
o=0 an) g R2C85 1
13.55 91 %
11.5 100 %
76 %
0 81-0.42-2 0 06 164 3.990 9200 3.599 B, L.
- e FOL.
g L.
g PS 37 %
23%
0.425*
1 I t.000 r2979 59.dW99.1 9.535 R29C47 I
>0 25 Dt N.L 59%
44 %
039-I'*
9.600 7.220 Nome"'
FS
-05 i
j 13.200 66.5/1I6.4 9.925 N
t %.-
- 1. Measwed leak rates adjuseed so sference coeditions by applying methods of EPRI data report. Reference 10.9.
- 2. Number of swi.JM Iigaments wish > 50% of Egament length resneneing is busw crack face.
- 3. Measured bwst presswe advssted so noniinel, has flow stress of 68.78 ksi for 7/8"dianceer embing.
y
- 4. Data excluded from EPRI a i-u w by Data Exclusion Cneerien 2a (Depth < 60% and < 2 ligamerus). Voltere also not a ecliable sneasurement.
g espesat.rn.t.so e-sys 3-21
_.2
....-..,--...---.._s_
rm W '97 12 48 FffTWT lidifLERifG-U.M. 412 W22 5889 TO 8205992!002 P.32/34 fj F
Table 3 9: Effect of Farley 1 & 2 Data on the Burst Pressure vs. Bobbin An:plitude Correlation (Additional French Data Excluded)
P, = a1 + a, log (Volts)
Parameter Reference Database with New / Old Database Value A2 & Al Data Ratio ai 7.5979 7.5814 0.998 a,
2.3662 2.4016 1.015 r'
81.97%
81.85 %
0.999 o,,,,
0.8236 0.8361 1.015 r
N (data pairs) 80 82 p Value for a, ~
9.6 1042 2.2 1048 0.23 Structural Limit 8.7 V 8.2 V 0.94
__ Reference o, 68.78 ksi Table 310: Effect of Farley 1 & 2 Data on the Burst Pressure vs. Bobbin Amplitude Correlation (All French Data Included)
Reference -
Database with New / Old Database Value 42 & Al Data Ratio a
7.6700 7.6581 0.998 i
a,
-2.3336 2.3635 1.013 r'
79.1 %
78.9 %
0.997 o,,
0.8337 0.8464 1.015 r
N (data pairs) 102 104 p Value for a, 9.2 104' 3.1 10'"
0.34 Structural Limit 9.6 V 9.0 V 0.94 Reference o 68.78 kai r
q:\\apc\\ala97\\ala90new.wps 3-22
5n 3
1 i.
q--
k r-um_
Table 3-11: Effect of Farley 1 & 2 Data on the Probability of Leak Correlation (Additional Freneb Data Excluded)
Pr(Leak) = 1 + e 'l'i'8.l*89[
t I,"'""I'#
Reference Database with New / Old Database A2 & Al Data Ratio Q
-6.2307 6.2286 1.000 j, _
7.7713 7.9142 1.018 V{
2.3006 2.2717 0.987 u
V
-2.6092 2.6185 1.004 u
Vu 3.3013 3.3540 1,016 DoF8I 109 111
(
Deviance 28.96 30.21
- 043 Pearson SD 59.5%
59.6%
1.002 Notes:
(1)
Parameters V are elements of the covariance y
matrix of the coefficients, p,, of the regression equation.
(2)
Degrees of freedom.
Table 3-12: Effect of Farley 1 & 2 Data on the
=
Probability of Leak Correlation (All French Data Included)
Paramter Reference Database with New / Old Database A2 & Al Data Ratio Di 4.7288 '
-4,5870 0.970 S,
4.4613 4.4109 0.989 Y
0.7883 0.7230 0.917 u
V
-0.6944
-0.6455 0.930 u
V 0.6828 0.6463 0.947 is No. of Data 133 135 Deviance 77.3 80.87 1.047
__ Pearson 9D 77.9 %
77.9 %
1.000 q:\\ ape \\ala97\\ala90new.wp5 3-23 i
7
Et & W 1a:49 0@ (U) @@!NEER!!G-W.M. 482 783 G099 TO B2059925002 P 34/34 Table 313: Effect of Farley 1 & 2 Data on the l
Leak Rate vs. Bobbin Arnplitude Correlation i
(All French Data Excluded) log (Q%) -, +
log (Volts)
Heterence Database with New / Old Parameter Database Value A2 & Al Data Ratio p,
0.8392
-0.7557606 0.9016 p.
1.2636 1.2461 0.986 t'
22.9%
19.9%
0.868 o,,,, ( p, )
0.7132 0.7546 1.058 r
N (data pelrs) 25 27 p Value for S, 1.6%
2.0%
1.27 TabIe 314: Effect of Inclusion of Farley 1 & 2 Data on the Reference Leak Rate Database for 7/8" Tube APC Applications (All French Data Included)
Leak Rate Oph)
Log ( Leak Rate )'
Parameter Reference w/ A2 & Al Reference w/ A2 & Al Database Databaqe Database Database Sample Size 27 29 27 28 Sample p 13.31 18.12 0.5696 0.6172 Sample o 20.84 34.57 0.8188 0.8482 Population Estimates Based on Log (Leak Rate) Sample Estimates Biased Unbiased
~
Population p 21.96 lph 27.891ph 18.761ph 23.661ph Std. Dev., o 128.00 lph 185.751ph 63.80 lph 88.981ph p Value 6.4%
7.6%
Notes:
1.
The distribution of sample leak rates bas been previously shown to not contradict a null hypothesis that the parent population is a log normal distrib ation.
s gdapc\\ala97\\ala90now.wp5 82,4
- TOTAL PAGE.34 **
i i Q s
i 8 I
r I
i i
i
-1 i
W 8
\\ !i N
~
3,3 _
fh
- Tor TJP/
E N
i
.s
- s d
? "-l k
o i,
i g
E h.S u ~ - t < J g
an o.oo -.
g a
t ffs a,
w
,N
,i-t 5
y I
2
- j
-o.n
-1bctrtc*t 75f'l 9
lb
- e
- P l
i I
~
l 5
0*
90*
180*
2 7 0.*
360*
a
[
Circumferential Position (degrees)
{
t e
re i
i 4
S
.i ~
. o l N 3
Figure 3-1 Sketch of the OD crack distribution found at the first tube stqport (TSPI) region ofTube R2C85. Also shown is the f
j location of the burst fracture opening. The bar:t opening extended beyond the TSP crevice regic,a rnore than the j
intergranular corrosion. It is estimated that the corrosion cracking on the borst fracture extended 0.03 inch above and below the crevice region. Finally, the post-burst test, tensile fracture face location is shown.
l 3-25 i
R S
I I
I
-a o.m -
(
i
- ror rse i g
i II Y
g' f
1 1
/
b'l N h*l a
g i
i I
I o.oo -
l g
c
'i 0
f
~
' 'l i
5
\\
i w
eE ou-e
. _ me ws 7
2 h
~
g O'
90*
180*
270*
360*
E Circumferential Position (degrees)
B 5
B N
h Figure 3-2 Sketch of the OD crack distribution found at the first tube support (HPI) region of Tube R29C47. Also shown is the location of the burst fracture opening. He burst opening extended beyond the TSP crevice region, but the intergranular corrosion was confined to the crevice region.
3-26
'*)
i
- d,n....,[,,9 T ~,~ ~ [
\\*
' *?S.
~.
'l" 1
S.
% k,, r.';.%". ',U
."7'*
- I
\\
. ' /q..,.
%.. _ e e, ;.. u. *,
>w 1
. ' wk,d,yf,2,. '.
)
t sJ Z.y ~ h e i r g- { j k -
s a. m...
.[L -
, d S t. L g} f L ' & ' ~ ~ k ' ~ '
.A v -"ph.n.
- w..,
& _ ]e
.k m
... 't-t g'r..y - >.---
~ ' Q R{-% W.,_i mb L' f.; i.
~,
g*: ' 'I* '
Tg hr r -a
~
?
. x )...x
. c.'." Q '*.. : aM{@C.":. t u %
-l
'n'
' 2. w.\\
n
~x vt.
g 1..
A
.e d.,~.C MO*..
L SEloSoox '"
.1
.:y 2
t l
l l
l i
A-
- j*w s
mam ' W h
,' f
-p +
~
gy_b
,, y
..g +
~ f'..h
~
y
~
s
^:
A M.
e
[
%e n
4
.F
-.. Q' i
s y
~
-- p-
. ).
t dm,
' i'J,4
[
.SE50000x g
s Figure 3 3 TEM photographs obtained from replicas pulled from the tensile fracture face rnade at the rnid crevice region of TSPI of Tube R2C85. Fatigue striations are clearly observed.
o 3-27 OE/PO*d 200S266S028 01 6885 CEL ETP *W*n-DN!M33NIDN3 (r0 Bd LS:ET L6. 42 Dnd
- .#;s
w,.
u.
g,.f,',
! 1[?
~.
z.
g.;.
r$+4.A..
l
- ..y
- 1) % :...--
g.);L.f,k j;.f
.b~..,' '
i D.'
Teemile Frm tw h
\\
~,
-e:
e.
- 1.
,t x
['%x.
. - v :.
m
- y
.I.\\
.. j.
- N f*. g,
,,,' ;r y.
- t',
., + '
t
'V,
q.o.
.,. lg,,
~
k A. b...:
5 '
'.Q. ' ' t:..' :.e.
x9 9m.5
.i 4 b..'%
a:c<ga
, L. { r*0) i.1 g p '
e.t
..u t
Azlal Freentre F
'. N
.(k.
-A s
c
',' '..O.
4
[:s l'
i, k
I, u
E.
s;
?,':
- 9. ~ *;
g
'M ef
,, d.. }
i
,e
.i e
,t l
s.-
j
,,, r
<) *t a*l
- *~
=
N, 4,...,,,,-
s i 'a s.
>j 6,
c.
?
2 Figure 3 4 Example of radial metallography (14% deep; obtained from the raid.
crevice region of TSP 1 of Tube it2Ct3 following both burst tasting and tensile frsentring of the burst specimen (16X). At highermagnificauons
($00X) transgranular components to the intergranular corrosion att obarved.
3-28 Or/Sa*d c.'00sE66s028 016885 224 ET& *W'n-DN!tf33N!DN3 (m ti3 es:M 46. 43 000
i i I,'.'.'.
'/.'."**
Lr m,_ n L
~
- =
- n...
~
C 1
7_
_b t-d__
z q
g q-6.
y
. f, ;,
. f,,.,,,
r r
i 1
E I I r
~2~~~
~
n
-\\
. ____q
".".'.1 55 a, a,.
..f w
l j;;
a f,
. u
_ _...a A4 Mvm.ht.
- i.,
he.s.1 u. I w. Ippy
~
w...
~
l D
(
)
~ ~
t m.
{
r a
e
~:~
(
7...C 1-
. are r
I
~
~
}
l r--
.: L__
~
~
- iu
_,. g C
)
)
1 c=
,. m
./
J
=
c
... r3 L
I 7
I F
l Figure 3 5 Bobbin Reanalysis for R29C47. TSP 1 3-29 OE M *d 200S2665028 01688S E24 2iP *Wn-DNIM33N10N3 (m 8:1 BS:EI 46. 42 EN
l f
i7 g
a 7
g e
I I
.Y
,)
g g3 A 'o n 7
~7 i
a
.Q g e$
y M,
a a
43-M
^/
o i
e g-y hO M
i; h
0 b'O Ebl*
Io N
8 ihass g,O
=
t 8
o..-
L l
l$
X y
Ae "t
A@
,f ah o
, o f
I
_7 m
,,7n g-..
e
~
+
c 4
11
/
J I
r a
N D
]
i ag I
r 3
d a
J m
o e a n
y[
o o
G=
Y l
k
"?
a c
pM
")
o l]
I O
N l
A or n
a o n J
l M
i
/6 n
T a
0 uU j
D j
i kQ g
/n P
1 b
@[ Un N
I A
"4 a j
i
$h
/
.I i
e
/
/
q
- y) aj n
2 g
1 a
m eG
/
J M
1 p
a r
5 g
'M I
b I,
/
f
)
o
,a e
g g
o o
e o
o o
o M
R (ISK) OESEO.If '}938
-OE/40*d 200S266S028 01688S E24 EIP *W'n-DN!833NIDN3 (N 83 BS 2146i LE EN
Figure 3-7: Burst Pressure vs Volts for 7/8" OD Alloy 600 SG Tubes l
Reference Database, Reference o = 68.8 ksi @ 650*F r
a:
12.0 o EPRIw/EdP Database O
m Plants A2 & Al Data
+ LS EPRIw/EdF 10.0 3
k a
x 95% Prediction @ LTL o
N D
LS EPRIw/EdF & R02C85 N
C
--- 95% P @ LTL w/R02C85 8
y up:o n K t
8.0 n, s
a o
^
"' -s g
t,.,
o a
- )
o
?
i N
4 o
o, 3
6.0 n a k
n w,
k -x'0 7
o D
?
y a :A-.- g M
-]
t 8
3.657 kai o'%
q q
3 W,.x N.
m u
_7 _N s
E a
g 2.0
-ir u
n x
m 4:
- g
_ 9.0 V n
1 0.0 0.1 1.
19.
- 100, Bobbin Amplitude (Volts) 3-31 e.m_--msv_este wwi,
nne m er.m m
g Figure 3-8: Probability of Leak for 7/8" SG Tubes 8
Effect of Plants A2 & Al Data on the Reference Data Correlation a.
1.0
- =-z-
-'= ::::
/
W/
0.9 -
o Reference Database A2-R34C53-1 b
/,['
a Plants A2 & Al Data A1-R02C85-i(l I/
l-8 0.8 -
m
____ POL Reference Data i
d POL w/R02C85 Data f
h M 0.7
/i l5 0
/y 3,
$ 06
[i i
i 5
.j j 0.5 E
E D
I i
b o.4 e
3 1
i t
q 0.3
[
3 g
t O.2
}
x' S
5
[
e 0.1 f
n a
S 0.0
- r=--
r:
~
0.01 0.1 1.0 10.0 100.0 Bobbin Amplitude (Volts)
P8PV 997Ms]M M 3-32
- N'7 W N
g Figure 3-9: Probability of Leak for 7/8" SG Tubes
~
l
}
Effect ofInclusion of Plants A2 & Al Data on 7/8" w/EdF Correlation 1.0
- g:
y
- = c
- = :->
y
/
/
p A2mcsa-1
/
.p 0.9 -
EPRI w/EdF Data j
/
o m
Plants A2 & Al Data
/
/
' A1-R02C85-1 '
0.8 -
POL Reference Data
/
8 o-so a
g 4
POL w/ R02C85 Data g
g W 0.7 u
4,
-l 0.6 jf m,
Q i
5 1
3 05 ll E
t
'g' l
g0.4 g
B 4
/s
.o 0.3 Y
fE A
0.2 Ev
[
~
g 0.1
[
/
~
0.0
~~
~
~
0.01 0.1 1.0 10.0 100.0 Bobbin Amplitude (Volts) p8PV.897.*] PoLEdF 3-33 m sesv7.us m
g -
Figure 3-10: Leak Rate vs Bobbin Amplitude 7/8" Tube Data w/o EdF, Effect of Plants A2 & Al Data 1000.0 Z
^
~
Model Boiler Database a
q a
US Pulled Tubes
~
~
Plants A2 & Al Data 8
a
=
l x
100.0 - E x
Retained PerNRC (MB)
/ :e AJ R.
Z Regression Mean(Q)
- _ _ ft
- l 1_
a x
^
f
+
Regression Median (Q)
'ghf' Ii-
~-
g,-+
g g
Median w/ Al-R02C85-1 4x
~~+
o "f
[
10 0 - E Mean w/ Al-R02C85-1 X r (*
p' q
cy
^
i j
/
x
~
= #y '
^
4n
%4 Y
y 4
-/ x
,1 g
,g 4
x
.- +
a g
7x=
_,g 1.0
/
Cl uf
~
-T 1r s
x s5-r f
x
.'+
3 x
_ e"A v
3:
.- '+
g x
+
O.1 g
_. /
C c
3 3
^, g e
",n
+
+
fB
-$'+
j 0.01 g
0.1 LO 10.0 100.0 Bobbin Arnplitude,V (Volts) dWA_MW serAs{0vsV.ws.EsF1 3-34 not:amar4asav
g Figure 3-11: Leak Rate vs Bobbin Amplitude
'S 7/8" Tube Data w/ EdF, Effect of Plants A2 & Al I'a.a 1000.0 E
o Model Boiler Database The p values for the corthw was yeater than 5%. Ita:s n,
o US Pulled Tubes 2
8 Iq m
Plants A2 & Al Data h
Retained Per NRC (MB)
X%
x 100.0 -
d 5
~'~
N
"="
a EdF Data
" x O
U
- 'G x
-S Reference Mean(Q) e E
+
Reference Median (Q)
,Ix C
a 3
-E 10.g t _ ---- Median w/ Al-R02C85-1
/I
" m, n
~'.,
r-C S
E Mean w/ Al-R02C85-1
-x--
4
^
^
T tx
_ f, p?
^
v e
f i
e x
1
^
gy 1I j
C x
g
/*x-(
-l$
X x
j o
in Oi E-'
e
,3 E
4 9
,.~,x s.
+'
=:
.2
'g
/'y M
-+
f, u 3
u,n
,, 3 +
.+
a e'
O.1 g
m^
=
C
'+
., +
B lN I
l 0.01 Ds 0.1 1.
10.
- too, Bobbin Amplitude, V (Volts) w_maav es7.Wst'3VE*]
3-35 n am or.4m m
4 4.0 EOC 14 INSPECTION RESULTS AND VOLTAGE GROWTH RATES 4.1 EOC 14 Inspection Results In accordance with the ARC guidance provided by the NRC generie letier (Reference j
10.1), the EOC 14 inspection of the Parley Unit 1 SGs consisted of a complete 100%
bobbin probe full length examination of all TSP intersections in the tube bundles of all three SGs. A 0.720 inch diameter probe was used for all hot and cold leg TSPs when ARC was applied. Subsequently, RPC examination was performed for all bobbin indications with amplitudes gnator than 2 volts in SGs A and B, and all indications above 1.5 volts in SG-C. Also, all (TSP) indications in tubes with indications exceeding the apair limit (2 volts) wen also RPC tested in all three SGe.
One hundred and six indicatior.s (106) had a bobbin voltage above 2 volts, and 85 of those were conarmed as flaws by RPC probe. The ARC crito11a in the recent SER for Far!ay Unit 1 (Refonnee 10.2) require that RPC conarmed indications with a bobbin amplitude abcve 2.0 volts and all bobbin indications above 5.6 volts shall be repaired.
Eighty seven (87) indications above two volts including one indication that exceeded the us.per repair limit (5.6 volta) and two RPC NDD indications were repaired. A large number of tubes in SGs A and C previously plugged in accordance with prior npair criteria were deplugged, inspected and repaired in accordance with ARC by _
installing sleeves or nplugged. No tubes were deplugged in SG-B during the EOC 14 outage. Only 10 indications, all below the ARC repair limit, were left unrepaired in tubes nturned to service in SGs A and C.
An augmented RPC inspection was puformed consistent with the NRC aquirements.
All dented intersections with a bobbin voltage greater than 5 volta and de-+ed intersections in one steam generator with a bobbin voltage between 2.5 and 5 volts were inspected with RPC. Large bobbin residual artifact signals weis also RFC inspected. There were no RPC circumferential indications at the TSPs, and no RPC indications with potential ID phase angles. NDE asamination in the field 6d not 3
[
reveal any indications extending outside ths TSPs; however, destructive examination L
of pulled specimen R2C85 TSP 1H showed that the tips of a single indication
[
extended 0.03" on either side of TSP (total indication length was 0.81").
The L
maximum depth for this indication at the edge of the TSP was 34E t
. In accordance with a recommendation made in Referenes 10.7, the SG EC guidelines
}
i-j' for EOC-14 were modiAed to call indications with large phase anglen and 0% depth (formerly UOAs) as P!s and eHmhate the UOA call from the criteria. This change l
in the guidelines is expected to have incroased the nymber cf PI calls.
I qdape\\ala97\\alasomew,wp6 4-],
1 n
OE/EI
- cl 200SE66SOE8 01608$ EE4 EiP *W'l'l-DN!833NIEN3 (l'O Nd OO:ET 46, 48 EN
- a
..a
A summary of eddy current (EC) signal voltage distributions for all steam generators is shown on Table 41. For the tubes that we;w in service for Cycle 14, Table 41
_ provides the number of neld bobbin indications, the number of these Held bobbin indicktions that were RPC inspected, the number of RPC con 6rmed indications, the number ofindications in plugged tubes, the number of Cycle 14 in service indications that truain active for Cycle 15, the numt ar ofindications recovered kom deplugged tubes which were returned to service for Cycle 16, and the subsequent total indication population 11ing retumed to sarvice for the beginning of Cycle 15 (BOC-15). Overall, the combined data for the Farley Unit 1 steam generators show the following:
e totd of 8074 bobbin signals were identi6ed as TSP indications during the 1
Nspection, and they were called PIs.
Of the 3074 PIs, a total of 563 were KPC inspected,385 of which were above
-1 volt. The largest indication was found in SG-C and it had a bobbin amplitude of 13.74 volts. - This indication was found next to the pulser nosale used for pressure pulse cleaning. - The tube section containing this indication was pulled and detailed examination was perfonned during this ou'cne. Details am prwided in Section 8.0.
Of the 563 indications RPC inspected, a total of 401 were RPC con 6nned.
381 indications were amoved kom service due to tube apairs in the present outage. Only 85 ODSCC indications required rep-!r based on exceeding 2 l-volts, and the remaining 296 indications were in tubes npaired for non-i OSDCC causes. Consistent with the 2 volt ARC, indications with bobbin amplitude less than or equal 2.0 volta are not considered for removal from service, regardless of RPC data.
Additionally,10 indications were found in deplugged tubes recovered and returned to service SGs A and C (none in SG B), for a total of 2708 indications retumed to service for Cycle 15 operation in accordance with ARC criteria. All indications in deplugged tubes returned to service had a bobbin voltage below 2 volts.
A review of Table 41 indicates that more indications (a quantity of 1006, with 575 indicationa above 1.0 volt) were returned to service in EG C than the other SGs, thereby it potentially will be the limiting SG at EOC-15. It is noted that SG-C had
. the largest indication (13.74 volts at R2C85 TSP 1H) found in the EOC-14 inspection.
The R2C85 tube' segment containing the 13.74 volt indication was pulled for q:\\ ape \\alnVI\\ala9onew.wp6 42 OE/P1 *d -
E00SE66SOE8 01698S EE4 21P 'W'n-DN1833NIDN3 Tm 8 10:E166. LE EfU
I destructive examination in the recent inspection (see Section 3 for details). Before pulling the tube, a in situ leakage test at nonnal operating pressure differential was performed on the 13.74 volt hdication. The indication did not leak in the in situ test although it leaked in the laboratory tut performed following the tube pull. The 1
I pulled tube withstood a pressure differential equal to 1.4 times SLB differential pressure durir.g b laboratory burst test. Thus, the largest indication detected
[
during the EOC 14 inspection met structural integrity requirement.
i i
Figure 41 shows b actual bobbin voltage distribution for tubes that were in service during Cycle 14, as determined from the EOC 14 EC inspection. Figure 4 2 shows i
the distribution of the EOC 14 bobbin indications which were mpaind and taken out of service, and Figure 4-3 shows b bobbin voltsgo distribution of indications returned to service for BOC-15. It is clear from Figure 4 3 that SG-C has more indications over 1 volt as well as the largest indication returned to service for Cycle 15 (2.9 volts RPC NDD); therefore, it is expected to be the limiting SG. Ten Indications were detected in deplugged tubes returned to service for Cycle 15, all in SGs A and C, and they an included in h distribution shown in Figure 4 3.
Table 62 summarizes EC results for indications in tubes deplugged, repaired by i
installing sleeves and returned to service during the EOC 14 outage.
The j.
distribution of EOC 14 indications as a function of support plate elevation, summarized in Table 4-3 and illustrated on Figure 4-4, shows the predisposition of-ODSCC to occur in the fint few hot leg TSPs (2495 of the 3074 PIs, r.,r about 81%,
4 occurred in the first four hot leg TSPs), although the mechanism does extend to '
i higher TSPs. Only 60 bobbin indications (or about 2%) were reported on the cold leg j
side. This distribution has remained essentially unchanged from the last inspection, and it shows the predominant temperatum dependence of ODSCC at Farley Unit-1, similar to that obsened at ollwr plants.
4.2 Voltage Growth Rates 3
The bobbin voltage growth rates for the Farley Unit-1 steam generators during Cycle i
14 are shown in Table 4-4 in the form ofcumulative probability distribution functions
. (CPDF), and b same data is presented in a graphical form on Figure 4-5. These grorith rates are developed from the 1997 EOC-14 inspection data and a reevaluation of the same indications from the previous (1995) inspection EC. signals. -Table 4-5 shows average growth rates for each SG during Cycle 14. The difference in growth rates among SGe is small, although SG-C has a slightlylarger average growth as well as the indication with the highest gnwth (9.4 volts /EFPY).
r
' =
q:\\ ape \\ala97\\ala90new.wp6 4-3 1
i OC/SI'd 200GE66SOE8 01688S EE4 EIP *W'n-DN1833t410N3 Cr0 Ed 10 ET 46. LE DN
~~ _,_ _._.
I 4
Average growth rates observed for all voltages vary between 11.8% and 13.6%,
between SGS, with an overall average of 12.8%, on an effective full power year (EFPY) basis. The average growth for indications with a BOC bobbin voltage above 0.75 volt is 14.8% per EFPY and for indications below 0.75 volt it is 12.1% per EFPY, Smaller percentage growth observed for BOC volts above 0.75 (relative BOC volts below 0.75) is consistent with data from the last inspection.
The NRC guidelines require that the more conservative growth distribution for the last two operating periods be applied for projecting the next cycle distributions.
Composite growth distributions for the last two cycles are shown on Figure 4 6. Also, average growth rates for the Farley 1 SGs during recent operating cycles are shown in Table 4-6. It is evident that growth rates for Cycle 14 are higher than those for the last cycle. Accordingly, it would be conservative to use Cycle 14 bobbin voltage growth rates for predicting the EOC 15 conditions.
The data in Table 4-6 show that average growth rates decreased steadily between l
Cycles 8 through 12, but that trend appears to have reversed for Cycles 13 and 14.
l Ilowever, a closer examination of the growth data for Cycles 12 to 14 showed that l
Cycle 14 had a substantially smaller number ofindications with negative growths than in the prior two cycles, as shown below.
A m age Percentage Average Growth Number of Number of of for Positive Negative Positive Negative Growths Only Cycle nd Growths Growths Growths (volts /EFPY) 12 841 842 50%
-0.005 0.097 13 740 1830 20%
0.060 0.121 14 28 2785 1%
0.134 0.116 The large decrease in the number of negative growths for Cycle 14 is likely attributable to using long life bobbin probes during both the EOC 13 and EOC-14 inspections. Although the overall average growth, which include negative values, show an increasing trend for the last three cycles, the average growth based on positive growths show only a modest change. Thus, the large number of negative growths reported during Cycle 12 and 13 inspections may have to have caused an underestimation of overall average growth for those cycles. It is noted that all negative values are cone.crvntively treated as zeros in the growth data applied for Monte Carlo analysis for leak rate and burst probability projections. Consequently, the leakage and burst analyses are minimally affected by negative growth values q:\\spe\\ala97\\ala90new.wp5 4-4 OE/91*d 20052665028 01 688S C24 ETP 'W'n-DNIM33NIDN3 (m Ud 39tET 46. LE Dnd
j 4
~
whleh occur at indications having small gmwth values, whereas the large growth i
valum dominste the impact on leak rates and burst probabilities. Other l
conservatisms such as the use oflimiting growth for the'past two cycles provide i
assurance that a conservative gmwth distribution is applied for Monte Carlo l
simulations.
)
Table 4 7 lists the top 30 indicaticas from the standpoint of growth during Cycle 14.
About half of those indications won present at the beginning of the cycle. All 30 indications were conarmed by RPO Inspection. Only four of those indications were i
found in SG-B, and the twt were either in SG A or C.
Figure 4 7 shows a plot ofvoltage growth, AV, versus the BOC voltage, Vsoc, for SG C i
which had a slightly larger growth rate than the other two SGs as well as had the i
largat growth found during Cycle 14.
As noted in the previous section, the indication with the largest growth also had the largut EOC 14 voltage (13.74 volts i
at R2C85 TSP IH) and it was located near the pulser nozzle used for pressure pulse cleaning. The data in Figure 4 7 also includes the second largwt growth found for Cycle 14 and it occurred at R2C77 TSP IH in'SG A. This indication is a new indication and it is also located in the pressure pulser zone. Potential causes for the large growth of R2C85 are discumed in Section 3.1.5. No evidence of fatigue (from
}
pressure pulse cleaning) was found on the axial crack face but was found in a cellular corrosion patch away from the axial burst opening. It is conceptually possible that j
the pressure pulsing led to some separation of the crack faces with a corresponding increase in the voltage and measured leak rata (also potentially afected by the tube pull operation). The fact that top two indledons firom smwth standpoint are in the pressure pulser zone suggests that the pressure pulsing operation may have contributed to the crack voltage increase.
i To ew==ine if the pressure pulsing operation affected growth rates for a mWority of l
-indications in the pulser mone, average gmwth rate was ' calculated separately for L
those indications and compared with that for indications outside the pulser sone.
L Only indications at TSP IH were considered. Tubes located in the pulser zone are identined below..
E PPC Pul-N-la Tube Affected Zone Eaw columns L
1 1 to 30 and 65 to 94 2
1 to 27 and 68 to 94 L
3 1 to 23 and 72 to 94 4
1 to 19 and 76 to 94 q:\\apc\\ alas?\\alasonow,wp6 45 l
OC/Li'd
-200S2669028 01 6889 EE4 EIP 'W'n-DN!83MIDN3 (n) 83 C0:C1 46. 42 Dnd
-i
9 5
1 to 15 and 80 to 94 6
1 to 11 and 84 to 94 7
1 to 8 and 87 to 94 The total number ofindications (at TSP III only) and their average gmwth rates for tubes inside and outside the pulser zone an as follows:
Number of Averare Growth Indications (volts /EFI3')
Indications within the pulser zone 38 0.49 All indications except top two growths inside the pulser zone 36 0.18 Indications outside the pulser zone 482 0.15 It is evident from the above data that although the average growth for all indications in the pulser zone is significantly higher, but the value is inflated by the two large growths observed during Cycle 14 (SG-C R2C85,9.4 volts /EFPY and SG-A R2C77,2.4 volta /EFPY). Without the top two growths, the average growth rate inside the pulser zone is only modestly higher than outalde the pulser zone. Cumulative probability distributions for the two cases are shown in Figure 4-8, and they also a show modest difference. Because of the relatively small population of indications in the pulser zone and modest difference in the two growth rates, no firm conclusions can be drawn i
at present regarding the effect of pressure pulsing on gmwth rate.
According to the Westinghouse ARC analysis methodology presented in Reference 10.5, the larger of the composite growth rate for all SGs and the SG-specific growth rate should be used in projecting SLB leak rate and tube burst probability for individual SGs. Since the growth rates for SGs A and B are below the composite growth rate (see Table 4 5), the composite growth rate is applied to those two SGs to provide a conservative basis for prodleting EOC 15 conditions. However, predictions for SG-C are obtained using its own growth rate since it is higher than the composite rate.
Data from the last operating cycle of Farley Unit 2 showed that deplugged tubes returned to service are likely to han a significantly higher growth rate than the active tubes. Hence it was recommended in Reference 10.7 that when more than a few deplugged tuben are returned to service, the gmwth rates used for EOC voltage projections should take into account potentially larger growth rates for deplugged tubes. If plant specific historical growth data for deplugged tubes are not available, qnspe\\sla97\\mla9onew.wp5 4-6 02/81*d 200SE66SOE8 01688S EE4 EIP 'W*n-DN!M33NIDN3 (N bid E0:El 46, LE 000
i
[
l it was recommended in Reference 10.7 that Farley Unit-2 Cycle 11 deplugged tube
}
growth ratas be used as a consonative growth distribution. Only 10 indications were added to Farley 1 Cycle 15 because of deplugged tubes returned to service; neverthelms, larger grawth rate for the deplugged tubes were included.
As j
suggested in Reference 10.7, a composite growth distribution obtained by weighing i
i the growth distributions for active and deplugged tubes by the number of each type i
ofindication returned to service was used to p.'-
EOC.15 voltage projections.
4J Probe Wear Criteria i
An alternate probe wear criteria discussed in Reference 10.11 was applied during the I
EOC-14 inspection. When a probe does not pass the 15% wear limit, this alternate l
j' criteria requires that all tabes with indications above 75% of the repair limit since l
the last successful probe wear check be minspected with a good probe. Accordingly, l
j all tubes containing indications for which the worn probe voltage was abon 1.5 volts
)
were inspected with a new probe. An evaluation of worn probe and new probe data 4
l-is presented in the following paragraphs.
i j
In accordance with the guidance provided in Reference 10.11, 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 large indications wem detected with new probes; thus, worn probes did not miss significant indications. Figure 4 9 shows plots of the worn probe voltages plotted i
against the new probe voltages for all three SGs. There are only three datapoints for
}
SG-B, so it is combined with the data for SG-A. Data in Figum 4-9 show a consistent i
relationship between the two voltages. Composite data from all three SGe are plotted in Figure 410. Also shown in Figum 410 as a solid line is a linear regression for the j
data, dashed lines apresenting tolerance limits that bound 90% of the population at 95% confidence, and chained lines rppresenting A25% band for the new' probe voltages. The mean regression line has 45' slope indicating that on the average new
}
probe voltages equal to the worn probe voltages. The dotted horizontal line at 1.5 worn probe volta demarcates indications requiring retest from those that do not. The shaded ama at the bottom shows the region where a tube mquiring repair may be left in service because of probe wear. In the Parley 1 EOC 14 inspection, there are no i
occurrences for which a worn pmbe was less than 1.5 volts and the new probe voltage l
exceeded the plugging limit, i.e., no pluggable tubes won missed due to pmbe wear l
}
' considerations.
4 Among the indications requiring ratesting (worn probe _ volts > 1.5 volts), all except' 14 indications fall within the band formed by the ehnined lines representing *25% of l
the new probe voltage. For all those four indications, worn probe voltages are higher i
i papetens.com.wes 4 i I
OC/6T 200S2669028 01680S E24 ETP 'W'M-DN!W33NIDN3 0'D W3 E0ICT 46. LE EN
-,-.- -- - - 'd-4
.-.. ---- --.-..w..-.
4 than the corresponding new probe voltages, i.e., the worn probe voltages are conservative. Four other indleations lie below the lower 25% band and have new probe voltages more than 25% higher than the worn probe values. The maximum new probe voltage for those indications is 1.6 volts and the differences between worn and new probe voltages would not signi8cantly affect leak and burst calculations.
Therefore, data for the eight indications outside the *25% band are acceptable.
Overall, it is concluded that tbs criteria to retest tubes with worn probe voltages above 75% of the repair 11mit is adequate. The alternate probe wear criteria used in the EOC-14 inspection is consistent with the NRC guidance provided in Reference 10.11.
4.4 Probability of Prior Cycle Detection (POPCD)-
The inspection results at EOC-14 permit an evaluation of the probability of detection at the prior EOC 13 inspection. For ARC applications, the important indications are j
those that could signl8cantly contribute to EOC leakage or burst probability. These l
signi6 cant indications can be expected to be detected by bobbin and conSrmed by l
RPC inspection. Thus, the population ofinterest for ARC 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-13 inspection can then be denned as follows.
EOC 13 cycle reported
+ Indications confirmed indications confirmed-by and repaired in EOC 13 RPC in EOC 14 inspection inspection POPCD =
(EOC 13)
{ Numerator)
+
New indications RPC confirmed in EOC-14 inspection POPCD is evaluated at the 1995 EOC-13 voltage values (from 1997 reevaluation for growth rate) since it is an EOC-18 POPCD assessment. The indications at EOC-13 that were RPC confirmed and plugged are included as it can be expected that these indications would also have been detected and con 6rmed at EOC 14. It is also appropriate to include the plugged tubes for ARC applications since POD adjustments to define the BOC distdbution are applied pdor to reduction of the EOC indication distribution for plugged tubes.
qdape\\ala97\\ala9onew.wp5 4-8 BC/02'd 200$E66SOE8 01608S EE4 ETP 'W'n-DN!M33NIDN3 (N &l P0tCT 46, LE 000.
1 It should be noted that the above POPCD de6nition includes all new EOC 14 i
indications not reported in the EOC 13 inspection. The new indications include i
EOC 13 indications present at detectable levels but not nported, indications present j
at EOC 13 below detectable levels and indications that initiated dudng Cycle 14.
Thus, this de6nition, by including newly initiated indications, differs from the l
traditional POD de6nition. Since the newly initiated indications an appropriate for ARC applications, POPCD is an acceptable de6nition and eliminates the need to l
a$ust the traditional POD for new indications.
l
- The above deSnition for POPCD would be entinly appropriate if all EOC 14-L indications wen RPC inspected. Since only a fraction of bobbin Indications an generally RPC inspected, POPCD could be distorted by using only the RPC inspected indications. Thus, a more appropriate POPCD estimate can be made by assuming I
that all bobbin indications not RPC inspected would have been RPC con 6rmed. This de6nition is applied only for the 1997 EOC 14 indications not RPC inspected since inclusion for the EOC-18 inspection could increase POPCD by including indications l
on a tube plugged for non ODSCC causes which could be RPC NDD indications. In addition, the objective of using RPC con 6rmation for POPCD is to distinguish l
detection ofindication at EOC that could contribute to burst at EOC, so that the i-emphasis is on EOC, RPC con 6rmation. This POPCD can be obtained by aplacing L
the EOC-14 RPC conRrmed by RPC con 6rmed plus not RPC inspected in the above de6nition of POPCD. For this report, both POPCD de6mitions an evaluated for j
Farley Unit 1.
The POPCD evaluation for the 1995 EOC-13 inspection data is summarized in Table 4-8 and illustrated on Figure 411. Data based on both RPC con 6rmed. only
[
indications as well as RPC con 6rmed plus not RPC inspected indi:stions are shown in Figure 4-11. Also shown in the Sgure is a generic POPCD distdbution developed j
by analyses of15 inspections in 8 plants and presented in Table 7 4 ofReference 10.6.
It is seen from Figure 411 that the predicted POPCD values for Farley 1 are equal to or better than the generic POPCD in the voltage range 0.2 to 1.2 volts. Hewever, L
between 1.6 to 3.2 volts Farley-1 POPCD remains more or less constant at 0.89 l
whereas the generic POPCD increases from about 0.88 to 0.99. The nasen for this
(
is that 8 new indications detected in EOC-14 inspection were assigned a reevaluated EOC-13 voltage _ exceeding 2 volts, i.e., 8 (out of 48) indications over 2 volts are considered mimaad dudng the EOC 13 inspection, and it has a relatively sign 16 cant impact on the POPCD value for 2 to 3.2 volts bins. POPCD is unity above 3.2 volts i
for both the generic POPCD and the Parley 1 EOC-13 POPCD distributions.
The 8 RPC con 5rmed, new indications found to have reevaluated EOC-13 voltages above 2 volts were further reviewed for possible causes for not being identi6ed in the I
j
_ gnape\\ala97\\alasomew.wps 4-9 ecn e a eeesesssees 01 se s cet ete w n-astwaaitosa cm aa seict ts, te one
_ _.. _. _ _ _ _ _ _ _ _ _ _. -. _. _ _ _. _. _ - _ -. _.. _, ~ _ _ _ _ _ _ _ _,... - _..
prior inspection. Six of the 8 indications have EOC-14 RPC voltages s 0.9 volt which would nominelly be expected to have less than 2 bobbin volta at the EOC-14 inspection and smaller voltages at the EOC-13 inspection. Five of thoso 6 indications do not have a clear flaw indication (as would be expected for 2 volt flaws) for voltage assignment with the flaw signals not clearly sepamted from the TSP lobes or distorted signals. Per the analysin guidelines, the vohages for these indications are conservatively assigned. The sixth indication has a clearer flaw signal bat in the ID plane which implies distortion of the OD flaw signal. While these six indications could have been called at EOC-13, undistorted flaw voltages would be expected to be less than 2 volts at both inspections. The remaining two of the eight indications also have distorted indications for both inspections, but the EOC-14 RFC voltages (21.9 volts) imply true flaw voltages > 2 volts. Both of these indications had significant growth (0.76 and 8.22 volts) and it is possible that the true flaw voltage at EOC 13 were under 2 volts. In summary, the 8 indications, while difficult to identify at EOC.
13, could have been called as distorted indications at EOC-13 and identified for RPC inspection. However, the true flaw voltages would be expected to be < 2 volta at EOC 13 and POPCD ebwe 2 volta is conservatively penalized for these new indications.
In summary, the Farley Unit-1 EOC 13 POPCD supports a voltage dependent POD l
higher than the NRC mandated POD value of 0.G above about 0.6 volts and approaching unity at about 3.2 volts. It is concluded that the POD applied for ARC l
leak and burst projections needs to be upgraded fmm the constant POD value of 0.6 l
to a voltago dependent POD.
1 l
4.5 Assessment of RPC Confirmation Rates l
This section tracks the 1995 EOC 13 indications left in service at BOC-14 relativo to RPC inspection results in 1997 at EOC-14. The composite results for all SGs are given in Table 4-9. For 1995 bobbin indications left in service, the indications are tracked relative to 1995 RPC confirmed,1995 RPC NDD,1995 bobbin indications not RPC inspected and 1995 bobbin indications with no indication found in 1997. Also included are new 1997 indications. The table shows, for each category ofindications, the number ofindications RPC inspected and RPC confirmed in 1997 as well as the percentage of RPC confirmed indications.
Of the 59 RPC NDD indications left in service at BOC-14,57 were RPC tested during the EOC-14 inspection and 26 were confirmed. Thus the overall confirmation rate for 1995 RPC NDD indications is about 46E This is consistent with the value (about 44%) estimated based on Farley 1 EOC-12 and EOC 13 inspection data, gnape\\ala97\\ml:9onew.wp5 4-10 k
evee a aesesssees 01 sees ret en vn-mmmma a s scia ts, te one j
O a
NRC -Generic Lette: 95-05 (Reference 10.1), upon NRC approval, allows for consideration of only a fraction of RPC NDD indications from a current inspection in establishing BOC voltage distribution for the next cycle.
A fractional value appropriate for ARC applications is the largest RPC confirmation rate for prior cycle RPC NDD indications found during the last two outages.
Thus, it would be appropriate to consider only 50% of RPC NDD indications for projecting EOC voltage distributions for Farley Unit-1. However, NRC approval has not been obtained and leak and burst nr.alyses presented in this report are based on 100% of RPC NDD 4
indications.
I 4.6 NDE Uncertainties The NDE uncertainties applied for the EOC-14 voltage projections in this report are those given in the prior Farley Unit-1 ARC reports (Refennees 10.4 and 10,5). 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 used to project the EOC-14 voltage distributions.
4 i
a L
i H
1 4
i q:\\apc\\ala97\\ala90new.wps 4-11 4
c.
ec cz a -
zeesassseze of. sms az4 ate *w*n-mim41m3 cm W 9etti 4s. La EN 4 Table 4 (Sheet 1 of 2)-
Farley Unit 1 April 1997 Outcge g
Sunimary ofInspection and Repair for Tubes in Service During Cycle 14
%vN g
Sksan Generater A Seeams Generseer 5 InJervlee During Chle tocas tocts o ch. t5 Isoerske Derfag Cyde secas socae Q.e,. as e pa es.4 Asto m.,a.s,.4 aswn y
l y,g s,,,
arc arc
= sired e ANTahes Tees Seearned arc RFC hidiradees As Tabes Tess Roomsed c-"*
W c="'-*
- 'ad si.
ee I
to Seretse8 Serstem es serviews servise 03 1
0 0
0 l
0 I
3 I
I I
2 0
2 0.4 15 2
2 2
13 0
13 25 4
3 3
22 0
22 0.5 36 6
4
_4 32 0
32 59 4
2 5
,,_54 0_
54 0
0.6 63 13 9
8 55 2
57 104 9
7 7
97 0
97 F
0.7 88 10 9
9 79 O
79 97 3
3 4
93 0
93 0.8 85 10 10 13 72 0
72 125 13 9
9 116 _
0 116 ' '
0.9 102 12 11 9
93 0
93 113 10 8
13 100 0
100 1
88 6
6 7
St I
82 116 9
8 10 106 0
806 N
~
I.I 84 b
2 2
3 81 0
81 92 9
5 9
83 0
13 1.2 60 8
S 8
52 0
52 71 4
4 4
67 0
67 3
13 62 2
1 4
58 0
58 49 4
3 9
40 0
40 1.4 52
,,._,, L_
2
_2 36 35 4
3 5
30 0
30 f
1.5 41 5
5 3 _
50, _
0_,_
50
.._ 46
__ 5 4
_,, 8 _ _38 _
,_0,_
__38_.,
36 0
f-1.6 25 3
3 3
22 0
22 23 3
1 2
21 0
21 4.7 30 2
1 2
28 0
28 10 0
0 1
9 0
9
't 1.8 21 4
3 3
18 0
18 8
2 2
2 6
0 6
~
1.9 15 2
1 2
13 0
13 9
4 2
3 4
0 4
2 16 5
3 5
11 0
11 5
1 0
2 3
0 3
2.1 8
8 7
7 I
I 3
3 2
2 1
0 i
f" 2.2 9
9 8
8 I
w I
2 2
1 1
I O
I m
2.3 4
4 3
3 1
0 I
6 6
6 6
0 0
0 2.4 4
4 4
4 0
0 0
2 2
2 2
0 0
0 m
3 2.6 2
2
._ _ 2
__2_
0 0
_0 4
4.., __2_,
_ 3_.
I
, _.0 _ _
In 2.5 _
2 2_
v 2
2 0
0 0
2 2
2 2
0 0
0 g
2.7 3
3 I
I 2
0 2
2 2
2 2
0 0
0 2.8 0
0 0
0 0
0 0
2 2
2 2
0 0
0 g
2.9 I
I
. I 1
0 0
0 0
0 0
0 0
0 e
i d
3 2
2 2
2 0
0 0
0 0
0 0
0 0
0 F
3.1 0
0 0
0 0
0 0
I I
I I
O O
O as 3.2 0
0 0
0 0
0 0
0 0
0 0
0 0
0 l -
3.8 0
0 0
0 0
0 0
0 0
0 0
0 0
0 g
6.4 1
1 I
I O
O O
O O
O O
O O
O
~l3.74 0
0 0
0 0
0 0
0 0
0 0
0 0
0 Total 920 131 108 120 800 3
803 1014 113 85 120 894 0
894
>IV 442 72 57 68 374 0
374 372 60 44 68 304 0
304
>2V 36 36 31 31 5
0 5
24 24 20 21 3
0 h
3 4_12
Table 4-1 (Fheet 2 of 2) '
~
Farley Unit 1 April 1997 Outage Selmsmary ofInspectiosa and Repair for Tubes in Service During Cycle 14
~
.N,-
sm c
,e c + e..eassi c
,mm.
In 8.rdee Dustes Cyde soc-ts secs 4 Orne.es Es.ServireD e.gCWee secos
- secas osse. es I
Deptumme A3Tutes Septages Amtshus
.y RFC BPC h
AE hes.
Takes teameesd arc RFC Seebast==a AS Takes Tden asensand Vall.ge m
'"d c*'""'#
I
""'d C"#"'
"d'"#
- e.'eme.o'res s 'r.ame
"'8 h
esserdens saevise 0.3 1
0 5
I i
1 4
9 4-0.4 6
0 _
0 0
1 __
9 I
46 6
3 5
41 9
41 0-0 6
0 6
0.5
- 16 3
2 4
12 0
12 lit 13 a
13 98 e
ss 06 68 7
5 8
60 e
60 2I2 2
' 234 O
~ ~
.7 77 le 3 ~ ~~
6 71
~-
71
~235 29 21 23~ ~ E3" 5'~~
243
~-
0 S
262 23 15 ' ~
19 03 95 8
3 4
98 2
93 305 31 22 26 279 2
281 0.9 IO!
19
- Il 13 88 0
85 316 '
41 30 35 2 88 0
238 I
114 19 15 15 99 I
too 318 34 29 32 286 2
2se N
1.1
' 94 12 8
le 84
.O S4 2'JW 23 15 22 248 0 -
248 0-1.2 106 3
3 2
108 2
106 237 20 12 14 223 2
225 N
1.3 103 13 6-9 94 0
94 284 19 10 22 192 5~
~ "if v
1.4 36 3
7 8
78 1
79 104 16 13 18 34 6 1
167 1.5 72 11 7
4 68 1
69 les 20 15 14 134 I
135 g
1.6
~ 44 44 23 8
36 0
36 92 38 32 13 79 0
79 y
1.8 30 30 18 2
28
_0 k~,_~ __ 72 34
. _ _2 3 9
__63_
O_
73-
{
l.7 32 32 22 6
26 _
9 28 59 36 23 7
52 0
52 1.9 20 '
20 13 2
IS S
IB 44 26 16
'9 35 0
35 2
29 29 23 5
24 0
24 50 35 26 12 38 0
38 2.1 7
7 6
6 1 __ _0 1_
is Is _ ___15 15
_,,3 _
o
_ 3 _.,
g 2.2 13 13 _
9 9
4 9
4
_, 24 24 15 18 6
0 6
G 2.3 7
7 2
2 5
8 5
17 17 11 6
0 6
n 5
5 5-5 0
0 0
Il -. -.
Il 11
.11-0...
0-2.4 0
2.6 3
3 2_
3 0
0
_0 7
7 6
7 0
0 0
g k
17 I
I
-1 1
0-0 0
6 6
4-4 2
0 2
g 2.8 0
0 0 -
0 0
0-0 2
2 2
2 0
0 0
2.9 2
2 I
I I
O I
3 3
2 2
1 0
I g
a 3
2 2
2 2
0 0
0 4
4 4
4 0
0 0
3.1 0
0 0
8 0
0 0
1 I
I I
O O
O g
3.2 I
I I
I O
O O
I I
-l 1
0 0
0 g
. _3.8 1
1 1
1 0'
O O
I I
I,
I.,
0
- __0,
,0 6.4 _
0 0
0 0
0 0
0
,I,_
_,t_,_
,,I I,
_e
.e 0
13.74 I
I I
I O
9 0
1 I
i
_, 1 0
0 0
Total l148 319 205 141 999 7
1906 3074 563 401 381 2693 le 2703
>2V 46 46
__ 169 9L,,.
571 4
, J75,_
_ 147_6 __
,385_,
,._270 227 1249
,4 1253-
> IV
. _ 662 253 34 35 11 8
11 106 106 85 87 19 e
39 n.
4_y3
'Itble 4 2 Farley Unit 1 April 97 Outage
- Sasnanary of laspection a.ed Repair For Tebes Depleased Durina EOC.14 hispection Staase Generator A 8seest Genormeer B 4-Deplussed pertes BOC. le laspessee RTS Cycle 15 Dep6 egged During EOC.14 lasposelen RTS Cycle 15
~
RPC RPC ladcasees
.AB RpC RFC ledicaeans All laspessed Cananned kapeired ledomens lesposwd Conormed Repaired ladicadow
- 04..
0
,,,,,g _
0 0
0 0
0,,
., _, p.
0 0
., OJ
.0 0_
.p,.
0 0
0 0
0,, _
0 0
4 6,.,
3, 2
,1,.,,,
,,,, 1 2
0 0
0,, _
,0 0,
0.?
2 2
2 1
0, 0
0 0
0, 0..
M 2
2 2
1 0
0 0
3_
0 0
.i 0.9 I
I I
I o
0 0
.a 0
0 0
I 2
2 I
I I
O O
O O
0 i
I I
I
.. 0 0
0 0
0 0
__.a t6 1.1 2
2 2_
2 0
0 0
Q,,,,,,,,
0 0,
3.
_. l.3 2
..3..,
2 2
0 0
.0 0
0
.0 t
1.4 :
1 1,,,
I I
O O
J 0
0 0
3 15 2
2 2
2
.p_....
, 0 0
0 0
0 1
,_1.6 2
2 2
2
,_,,,.0 O
O O
0 0
)
.. tJ.
I I
I
.J 0
0 0
0
- 0
....0~
ls8 I.
2 2,,
2 0
0 0
0,,,
.,,0_,,,,,,,,,,,,,,,,,,0, 1.,
.,,.,3 0
0, 0
0 0
0 0
0 0
14 I
.,__I I
O O
0 0
0 0
2J
...I I
I I
0 9..
0 0
0 0
2.7 0
0 0
0 0
0,,,,,,
0 0
0 0
,,.3 1 I.
I I
1 0
0 0
0 0
0 5
15 0
0 0
0 0
0 0
0 0
0 Tal 24
_._25,,
u 2s 3
- 0..,
0 0
0 0
wie 16 16 16 le 0
0 0
0 0
0 seasse Geerster C F
of AN SG Data Deptemed Derlas EOC.14 *,
RTS Cycle 15 Deplesped Durbs 50C.14 lasposehe RTS Cycle 15 vou e tr id riend s
io
{
g,,
g ApC kPC laducadees AD RPC RPC ladecadena An lesposted Coenneed Repened ledaemene tempened Ceanneed Repaired lededoen w,,w, 44
,I,
I I
I 0
I I
I I
O,__.
as i
I I
1 0
I 1.
0..._
l
.. 01, 3,
3 L_,,
,,3 0
6 5
4 4
2 0.7 I
I i
1 0
..L.
,,, 3 3
3 0
- 0. 5._
4 3
3
_.2-7 4
5 5
2
=at 0
0 0
0 0
1 1
. l.
O
.l 4
3.-
3 2
0
,,),
3 3
3
,,,,,0 3
1 6
5 4
4 2
ll
.2 2
2 l
1.2, _,,4,,
4 2
2 2.
,,....e_
6 4
4
,,,2 13. _.. 1.
2 2
.1.
0 d
4 4
0
.. I A 1
1..
..... 0
..0 1
2 2
1
=
l 1
4
..i>
2..
2 _......l
...L i
4 a
3 3
I i
16 I
.. I I
..l. _
0 5
-)_.
.?.
.. 2.
0.__...
l.7
,0,,,,,
0 0
0 0
,,,,,1 I
I I
,,,,,O l
,Js
__0_
0 0
0 0,,,,
2 2
2 2
0 l.9 1
1
.I 1,
0 I
I I
.I; 0
2.4.
0,,
,0
,,,.0 0
0 I
I
,, ).
i 0
2J 0
0 0
0 0
1 1
1 1
0
. 2-L..__
1_.
I I
i 0
I I
I O.
11 0
0
.0...
0 0
1 i
0 33 I
I e
i e
i 1
1
. 1 0
.. To.t.a. l 30 28 23 23 7
56 53 46 46 10
.>1=
15 15 11 11 4
31 31 27 27 4
4
?
%%:een.
4 14 OE/92*d 200S2665028 01688S 224 2IP *W*n-DN!M33NIDN3 CI'D Bfd 40:EI 46. 42 0n0
i Table 4 3 FarleyUnit1 April 1997 TSP ODSCC Indication Distributions for Tabes la Servka During Cycle 14 Steam Generator A Steam Generator B l
Number of Maanmum Averags Largest Average Number -f Mulmom Avenge Largest Averagt Tube Support Indicanons voluge volage Growe Csowe Indications voltage Volugc Growth Growth Plate IH 206.
. 6. 3. 5.. 1.13 3.22 0.23 174 3.10 0.97 1.76
- 0. 1,8...
2H 131 2.32 0.99 0 62 0.10 225 2.50 0.28 0.73 0.10 3H 174 2.67 1.01 0.73 0.11 231 2 66 0 95_,,
0.71,,,,
0.1 1,.,,,
. 4H,
209 _,,, 2.64
, 1.11 0.70 0.15 209 2.67 0.99 0 92, _ _,0 14 5H 99 2.21 1.10 0.37 0.14 93 1.90 0.97,,
0.47 0.12,,,_
6M 60 2.58 1.23 0.57 0.19 50 2.56 1.07 0.32 0.14 7H 7
1.63 1.14 0.28 0.14 16 1.34 0.7s 0.23 0.13
. -. 7.C,,,,,, 3
.., _118,
,,,,,1,._00 0.26 0.12 3
1.68 1.20 0.50,,
0.26
,,,, 6C_
4 0.77 0.70 0.14 0.12
_0,,
SC 12 1.26 0.87 0.37 0.14 7
0.97 0.61 0.24 0.12
._. 4C 8
1.28 0.97,,,,,,,0, 28,,
, 0.15,,
I, 0.49 0.49,
0.10 0.10 3C 1
0.37 0.37 0.01 0.03 3
1-0.74 0.35 0.19 2C 1
0.55 0.55 0.19 0.19 0
i IC 3
1.02 0.68 0.26 0.15 2
0.56 0.81 0.12 0.12 l
~
Teial 920 1014 Steam Generator C Composite of All SGs Number of Maximum Average largest Average Number of Muimum Average Largest Average Tube Suppen Indications Volmgc Volage Growth Otown indications Voltage Voltage Growth Growth Plau IH _
,,,, 1,3 8
_ 13.74 1.15 12.38 0.27 520 13.74 1.08 12.38
,, 0,.22 2H 265 3.76 1.17 1.59 0.15 621 3.76 1.03 1.59 0.12 3H,
, 271
,,_2.83 1.15 0.83 0.!5
,676 2.83 1.04 0.83_,
0.12,,_
,,,,_4 H 260 2.53 1.1s O ss 0.17 67 8,,,,,
, 2 67,_
1.10 0.92 0.15
, 5 H,,,,
127 3.00,,
1.17 0.56 0.19 319 3.00 1.09
, 0.56 0.15,,
,,,,,6,H
$7 2.01 1.16 0.56 0.18 167 2.58 1.16 0.57 0.17 7H 10 2.54..
1.41.._. 1.12 0.34 33 2.54 1.05 1.12 0.20
,,,, 7C 3
1.00 0 83 0.44 0.22,,,,,, J,
,,, 1.68 1.01 0.50 0.20
_, _ 6C 4
1.65 1.19 6.: 5 0.29 8
1.65,__ 0.94 _, 0.55, 0.21 SC 0
19 1.26 0.77 0.37 0 13
,,,,, 4C 1
0.73 0.73 0.21 0.21 10_
l.28 0.90 0.28 0.15
,,, 3C
.,,1 1.09 1.09 0.55 0.55 5
1.09 0.74 0.55...
0.23._
2C 3
0.67
,,, 0.54_
0.26 0.14 4
0.67 0.55 0.26
,,0.15 1C 0
5 1.02 0.73 0.26 0.14 Total 1140 3074 4-15 a-===.mu OE/42*d 200S266S028 OJ. 688S 22A Eit' 'W*n-DN!bf33NION3 (n) tid 80:EI L6. 42 0n0
.h d d;o oo d
6 o e o o6 6o6
- djd, jojo I
i it i i l
!l!
!j!!'!
g i
j j j
-jaiE E Ry el= Hal;- -i- -!= *:=: - - i
6
' I' m
i i
i b b 5
oo o ooo o
ld o 6 o. 6 6 dld l
I l
t i
a j'
o$
O Q
3 h
oj dd n o o ojo o d d d
jo f
i w
I l
l
'l l
l l
h g
j 5
i W
I f
I
!$k h h$ RN N'N -o of-o oe o cle -
k o o eR n $
i i
i t
I sE f n
l I
IIi
~~ 3
{
3 o
E
@ @ E E E E f f'! { { I ( f Es ! RES!
3 D
d, 6 o o 6 6 6 6 6o d o d o 6 o o 6 T40 i
i I
I i
eQ lI g
jab a g E E 5 5 R R E f f f f 5 T
a
=
b
. e e e e g u.
o e de s e I
i a
g a
i
(
% p om n o oo-e o e o oo o ooo t
i i
j o
m k
h 3
3 3
. c o... e e d o!
e o....
gas $sas ssereln's.me'amasI
}
8 a
eoeee ee e,e e e e
. e e e e i
i f
jj oo-.ggig==
I ojo g co No-~~oo oo-i i
f m
l i
E I
e i n n a s s s a e s s e r e s
K 6 6 d d d d 6 O d d e d d i
jj E S S E D D % I f j
2 E E 3 E E E E E 0 $ $
~
a 1
6 OTAE 'd 20052665028 01688S 224 ZIP *W'n-DN!M33NIDN3 (N Md 80:ET 46. Lt
8
.NR Table 4-5 o
Fadey Unit I-Apr81997 Outage Average Vallene Geweeds Diaring Cycle 14 Voltage Number of Amonge Voltage Range ladeselone BOC Esisse Cycle Per EFPY '
ErW8 e Cycle Per EFPY '
Compoelle of AN Steam Generator Date O
Entire Voltage Range 3074 0.91 0.154 0.125 168%
13.7 %
~V noe <.75 Volts 1173 0.57 0.111 0.090 19.6 %
15.9 %
2.75 Volts 1901 1.12 ' ~
~ ~~d.1110 0.146 16.0 %
13.0%
Steam Generater A
.Ny Entire Voltage Range 920 0.92 0.153 0.116
_ _ _1 6.6 %
12.6 %
V,oc <.75 Volts 345 0.56 0.110 0.083 19.4 %
14.7%
j 2.75 Volen 575 1.13 0.179 0.136 15.8 %
12.0%
Stesm Generator B -
Entire Voltage Range 1014 0.82.,_. _ 0.128 0.097
' 15.6%_
_ _ 11_.8%
Z V moc <.75 Volts 489 0.56 0.099 0.075 17.7% _ _
13.4%
2.75 Volts 525 1.06 0.155 0.117 14.5 %
i1.0 %_
]
Steams Generator C a
3 Entire Voltage Range 1140 0.99-0.177 0.134 17.9 %
13.6%
g Y.oc<.75 Volts 339 0.58 0.131 0099 22.4 %
17.0 %
h 2.75 Volts 801 1.16 0.1 %
0.149 17.0 %
12.9 %
a
, a.,ea cyci. is4.
.r4 2.i es,olia2 mervi n
E c err. u sermsw ereM 4-17
Table 4-6 Farley Unit 1 April 1997 Average Voltage Crowth for Cycle 14 Composite of All Steam Generator Data Bobbin Voltage Neanber of AvwageVokege Average Vehage Growth Avwage Persentage Growth konge ladkatlees 50C Entire Cytle For EFFY Entire Cycle Per EFPY Cyde 14 (1995 1997) - 450 EFPD Entire Voltage Range 3074 0.91 0.154 12 %
16.8%
13.7 %
V see <.75 Volts 1173 0.57 0.111 9%
19.6 %
I 5.9%
3.75 Volts 1901 1.12 0.180 15 %
16.0%
13.0 %
Cycle 13 (1994 1995) 489.4 EFPD Entire Voltage Range 2571 0.89 0.085 6%
10%
7%
V poc <.75 1024 0.56 0.101 8%
18 %
13 %
2.75 1547 1.10 0.074 60 7%
_ 5 %,,,
Cycle 12(1992 - 1994) 442 EFPD Entire Voltage Range 1681 0.98
-0.01
-l %
-0%
-0%
V see <.75 466 0.60 0.04 0%
7%
6%
2.75 1215 1.13
-0.03
-2%
-0%
-0%
Cyde 11 (1991 - 1992) - 471 EFPD Entire Voltage Range 1267 0.85 0.22 17 %
26%
20%
V soc <.75 546 0.57 0.21 16 %
37 %
29 %
2.75 721 1.08 0.23 18 %
21 %
17 %
Cyde 10 (1989-1991)
Entire Voltage Range 499 0.70 0.23 N/A 33 %
N/A V soc <.75 -
306 0.51 0.24 N/A 47 %
N/A 2.75 193 1.01 0.08 N/A 8%_,
N/A Cyde 9 (1988-1989)
Entire Voltage Range 431 l
0.62 l
022 l
N/A l
35 %
l N/A Cyde 8 (1986-1988)
Entire Voltage Range 274 l
0.48 l
0.28 l
N/A l
58 %
l N/A Cyde 7(1985 - 1986)
Entire Voltage Range 123 l
0.45 l
0.20 l
N/A l
44 %
l N/A i
4-18 o e
=wm l
OE/DE*d 200S266S028 01688S 224 2TP *W*n-DNIM33NIDN3 (m Ed 60:EI 2.6. 22 000
casw vnpurwnerumRa3#ResMm P,02/02 l
t i
Table 4 7 Far.4y Unit 1 April 1997 Sununary of Largest Voltage Growth Rates for BOC.14 to EOC-14 Staase Generator Bobbia Voltage RPC New SG Row Col Elevation
'ZOC -
BOC Growth ConGrmed 7 Indication 7 C.
2 85 OlH 13.74 1.36 12.38 Y
N
.A 2
77 OlH 6.35 3.13 3.22 Y
Y C
12 48 02H 3.48 0.65 2.83 Y
Y C
21 44 OlH 2.69 0
2.69 Y
Y
=
B 45-42 O1H 3.1 1.34 1.76 Y
N A
12 47 OlH 1.68 0
1.68 Y
Y i
,C 13 63 02H 3.2 1.61 1J9 Y
,_,,,, N A
20 64 OlH 2.31 0.81 1.5
.Y Y
A 16 29 03H 3.1 1.66 1.44 Y
Y A
21 50 OlH 2.99 1,61 '
l.38 Y
N i
A 26 75 OlH 2.45 1.14 1.31 Y
N
..C..
21 44 02H 1.23 0
1.23 Y
Y A
2 85 O1H 3
1.81 1.19 Y
N A
19 64 02H
. 2.5.
1.33 1.17 Y
Y B
29 73 OlH 2.73 1.58..
1.15 Y
.N.
l C,,,,,,.,,, 7 3
07H 234 1,42 1, {2,,,,.
Y _____Y,__
A 12 46 03H 1.1 0
1.1
.Y Y
C 19 28
.... 2H 3.76 2.71 1.05 Y
Y 0
C-25 48 02H 1.42.
0.37 1.05 Y
Y C
41 44 02H
. 2.91 1.95 0.96 Y
N A
12 43 06H 0.95.
0.01 0.94 Y
Y C
9 14 03H 1.19 0.26 0.93 Y
Y A
31 67 O1H 2J IJ8 0.92 Y
N B
45 49 04H 2.67 1.75 0.92 Y
N
~
~
-A 9
34 O1H 1.43 0.53 0.9 Y
Y A
9.
34 03H 1.56 0 67 0.89 Y
Y A
19 63 OlH 2.65 1,77 0.88 Y
N C
23 11 04H 2.44 1.56 0.88 Y
N A
9 34 02H 1.13 0.27 0.86 Y
Y 4
~
'~
B 43 61 OlH 2.74 1.88 0.86 N
f-on-eT.w4.7ame; e m 4 19
- TOTAL PAGE.02 **
Table 4-8 Farley Unit 11997 EOC-14 Evaluation for Probability of Prior Cycle Detection 1
Counposite of All Stessa Geners*9r Data j
- Indcations Detected EOC-13 New indcat!ons Both in EOC-14 and POPCO g;
p EOC-13 inspectons
.g EOC-14 EOC-14 gi inspection inspection RPC j
EOC 14 RPC EOC-14 RPC RFC FIPC Confirmed Voltage inspechon Confirmed inspection Conarmed Confirmed Confimed Plus Not Bin RPC plus not RPC plus not and Plugged inspected Confirmed inspected Confirmed Inspected Frac.
Count Frac.
Count
> 0 - 0.2 0-2 0
0 0
0/0 0.000 0/2 E.@
0.2 - 0.4 9
68' 6
53 0
0A00 6/15 0.438 53/121 h
0.4 - 0.6 10 179 26 355 1
0.730-27/37 0.685 356/535 0.8 4.e 17 168 44 537 1
0.726 45/62 0.762 538/706 0.8 - 1.0 10 98 33 482 0
0.767 33/43 0.831 482/580 1.0 - 1.2 6
49 37 424 1
0.864 38/44 0.897 425/474 1.2 - 1.8 12 44 93 336 4
0.890
.97/103 0.885 340/384 1.6 - 3.2 16
-17 82 100 32 0.877 114/130 0.886-132/149 3.2.-13.8 0
0 0
0 4
1.000 4/4 1.000 4/4 E
TOTAL 80 625 321 2287 43
> 1V 34 110 212-860 41
.~D 8
S 4-20 Poped Tablel 8/26/97 6:40 PM
=
Table 4-9 Farley Unit 1 Analysis of RPC Data from EOC-13 and EOC-14 Inspections Conibined Data from All Seesa Generators Total Total Total Total Percent EOC-13 EOC-14 EOC-14 EOC-14 EOC-14 Group of Indcations Inspection inspectbn inspecilon Jnspection inspection I
Bobbin BotMn RPC RPC RPC Indcation Irdcation inspected Confirmed Conarmed Less then or Equal to 1 A Volt in EOC-14 Inspection EOC 13 Inspection Bobbin Left in Seewtoo
_ tt43
,, _ tt40
,__,,,129 _ _ _, 93
_ _72.1_ _,,
j
- EOC 13 hspecton RPC Confiruned 4
4 3
3 100.0
- EOC 13 Inspedian RPC NDO
._ 0 0
, _.,O_,,,,
,0
- EDC-13 faspecuan RPC Not tns2ected.
1136 1136
., _,,126_,_ _._90, _
71.4 _
- Ne EOC-t4 inspection Bobbin
- 3 New EOC-t4 Inspedion Irw2 cation 458 49 38 77.8 a
Susn ef As EOC 14 inspecqkrilndication 1143 1598 178 131 73 6 Groeter then 1.0 Volt in EOC-14 Inspection Q"
EOC-13 Inspection Bobbin Left in Sen4ce 1273
.,,,j272 317 228,,,,,,
_. 71.9__,_
EOC-13 inspecton RPC ConIhmed_
57 57 52 47 90.4
- EOC-13 faspecilon RPC NDD 59 59 57 26 45.6 h
- EOC-13 Inspection RPC Not inspected I158 1150 208 155 74.5 No EOC 14 Inspection Bobbin
- 1 New EOC 14 hspecebnIndiceGon 204 66 42 61.8 Sumof AN EOC-54 tupeceonindicason 1273 5476 385 270 70.1
$r AllVoltages in EOC-14 inepection EOC-13 Inspecilon Bobbin Left in Service 2416 24I2 446 321 72D
- EOC-13 Inspecten RPC Co'uftrmed 61 61 55 50 90.9 T,
EOC-13 Inspecton,R_PC NDD 69 59 57 26 45.6
_- EOC-131nspection RPC Not inspected 2292,,
,_,2292,,_,
,, _334___
,_ _ 245_ _
,,, 73.4
- No EOC-14 Inspection Bobbin
- __,,4_,
New EOC-14 Inspecson endianton 662 117 80 68.4
-g Sum of AB EOC 14 Inspection inrEcasion 2416 3074 563 401 71.2
- wuncaners spes is neseo en teos ans, tunnenvenage Fogod Tatde 2 8/26&l6:40 Pbt 4-21
n.m Figure 4-1 Farley Unit 1 April 97 Outage Bobbin Voltage Distributions at EOC-14 for Tubes in Service During Cycle 14 f
I40 G
i3 i20 -
b 100 -
23 SG-A E
80 -
O SG-B a
q g
o 3
60 -
E SG-C i
s E
l 40 -
N 8
20 -
N
- N;d;h " l ";""; ~ ;
O A; i
2I55$$$~2I- 0 3 S S S S S ~i U 2 7; Z Z U 9 74
?USE3 b
~
Bobbin Voltage Bebarpe.ntaFigipt/MS7142 PM 4-22
- -.. _. -. -. ~
E Figure 4-2 Farley Unit I April 97 Outage Bobbin Voltage Distribution forTubes Plugged After Cycle 14 Service D
e 16 g
- u w.
-J 14
- o a
E v
12 -
E s
GSG-A f
3 10 -
[
o 3
O SG-B
.b
.s
=
=
z i
~~
h 8-
{
^
s o
N t.
h ESG-C id 2
i a
l s
j l
/
l l
l e
z 6-m m
l l
l l
l l
l
-4 l
O l
l l
l l
m
_____7
- l
. l l
. j
. l
_3
- g
. j 4
rg l
l l
l l
w l
l l
8.-
l l
i 4
l l
l l
l 1
l l
?
l l
?
l s
8:
i i
s i
l l
l l
l 2
l l
- l l
l 7
gj I..
l l
l l
l l
l l
l l
l l
l l
l i
l l
l l
l l
i 1
l
.l i
l i.3.
i
' C.d ^
i i.!i' i f
l
> 5 -
D 0
','m' at w
c
<a t-m en - - es n
-r-c a s-e c~
e,
<~
w n-m c.
a ci o 6 a e 6
= - _.: a = -.:
.. ; -.i vi vi. -~,
.e
\\
,,; -, -i
.._ s
+ =
1; Bobbin Voltage l
l Babasyc.midig21/2WW N 4-23
t 1
j Figure 4-3 g
j Farley Unit 1 April 97 Ontage g
Bobbin Voltage Distributions forTubes Returned to Service for Cycle 15 140 k
E E
120 -
E h
E SG-A 100 -
h I
D SG-B
.9 g 80 -
- B
,r.
b e
3 60 -
E SG-C U
8 m
~
l b
I a
40 -
l
-l M~
20 - -l
-l m:
ki f
1 l
.j
.__=._ s. _..
0 03 0.4 0.5 0.6 0.7 0.8 0.9 l
1.1 1.2 13 1.4 1.5 1.6 1.7 f.8 I.9 2
2.I 2.2 23 2.5 2.7 2.9 i
Bobbin Voltage Beborpr.shfi318eberpsAs rest i
5 4-24 s
. _. _ _ _ _ _ _ ~
_. _ _ _. _ _ _ _. - ~ _ _ -. _ _.. _.. _ _ _ _ _... _ _ _ _ _ _. _.. _ _.. _.__ ___. _
l Figure 4-4
{
Faricy Unit 1 - April 1997 q
ODSCC Axici Distributions for Tubes in Service During Cycle 14 300 E
l 5
w 250-l 200
{
8 ESG-A 7
h!
3 l
-?
150 O SG-B E
,2!
l U!
100-ESGC wl N
50 si E
0
- -- - " = -
U
= -
I III 2if 3H 411 Sif 611 711 7C 6C SC 4C 3C 2C IC j
Tube Support Place mI w
i 4-25 n
Cgure 4-5 Farley Unit 1 Cycle 14 ( Dec.19M to Nov.1995 )
Cumulative Probability Distributions for Voltage Growth on an EFPY Basis g
1.0 m___-
=
=
=
=
=
/ F', f~
W"
~E 0.9 -
/
E i.j Q
I.
0.8-n
/-
5 1.*
SG-A-0 1:
b 8 0.7 E
D I :
I G
E f *:
- c-SG-B
.a iJ:,
h-0.6 w
,8
'yl.:
.. m.. SG-C 0.5 -
t i
m
,! :E E
n
--x-Cumulative 10 w 0.4 1
i E
g u 03 -
g 0.2 =
f e
N-1 0.1-0.0 ::' ' '.
l l
l l
l l
l l
l l
o e
~
o n
m
,r m
e c~
oo e
o
- ~
n n
m w
d 4
6 6
6 6
6 6
6 6
6 6
a a
u d
6
(
Voltage Growth Y
4-25
g
.L
,&mweI. $'
mi'j
\\,l l1l\\. ~ bM N "8e o g88,
? 8da ll11 1
i1i n
. Me
- 9M
- 9a
- Ma s
i 3
4 s
1 a
B k
e l
y yc Y
C C
PF
- H-E o
n
- T-a ne s
- N-no it u
- 1-b ir s t r s o i t Da 7
r 9 ye 9 t n
- 9o iie 1
l h
bG l
e i
-.r a w
i 6 pb s
%o m 7
4 Aos G
2 r e 1
e - Pt eg 4
S e 1 e
- No a ut vl k
l F'q i i A o
nt Ulsf V
eo
- 9o yi e
emt g
lr ui s aCo 9o F
p i
ym r o
- To oC t
s i
.H
- Mo h
t w
- E o
No r
G.
- l;0 l
9o a
l[;
n g
- ;i:
.iS e
- o i
b b
- ? Mo o
B
~
7 No M
P M9 42 3
7N h To 6
0 9
8 7
6 6
A 3
2 1
o.
2
/8 1
0 0
0 0
0 O
0 0
0 0
ig 3
w g*
t F
h wo r
G ll
4 Figure 4 7 g
Forley Unk-1 April 1997 Outage Voltage Growth During Cycle 14 vs BOC-14 voltage for SG-C Plus Largest Growth in SGs A & B E
10 I
u 9-l 0 SG-C (A!Iirdcat'w.s) m i
elargestin Growth SG-A E
8--
E A largest Get.wth in SG-B 7-3 h
- o 16 u
I35 h
6 i
E 4-I E
G
~
o E
- y3 u
E 8
2 m
b a
e j
I s
I-O(
0-
-l g
0 0.5 1
1.5 2
2.5 3
1.5 h
BOC-14 Vokage i
SgtgrvsvC14-78/26#97 6:46 PM
4 Figure 4-8 Farley Unit 1 - April 1997 -
Comparison of Grewsta Rates at TSP IH in Tubes within and Outside the PPC Pulser Zone.
g Caumlative Probability Distributions on an EFPY Basis - Composite of AB SG Data 0
1.0
..+---c o.
...o - ---D,,....u.............o.
____--4,.-.
_on g *...,
- ~~,s 0.9-e j
yp-n m
s c
0.8 -
l
' ~
g s'
[
s'
=
0.7 -
Q s
f x
s
~
0.6 5
E E
- s
/
y
-- n -.oussue Pulser Zone 0.5 -
w r
n 9i n
E
- (
n j o.4 -
W
, Wkhin Pulser Zone 4
I J
u s
0.3 t
o
+
e-e n
0.2 -
a f*
0 s:
i:
N 0.1 -
j.E 8
M s.-
f 0.0 : " - - - - '
m; c1
-1 o
9 9
9 9
9 e.
9 at t
9 t
L.
9 9
o o
o o
o o
o o
o ce es Voltage Growth hj Orewse Fig 3(2) 8/26197 &O9 Phi 4-29
AJG th "tk/ i3818 FR CUL ENGINEERlHG4).N. 412 722 5809' TO'82059925dd2 P.12/27 F1pm &9 Farley Ucit.1 EOG 14 Lespection Comparison of Worn Probe Voltag: ". gainst New Probe Voltage e SG-A Data 4
go
. SG4 Data 8.0 40 f
f.0 g
Z to 14 0.0 0.0 1.0 10 10 4.0 5.0 e0 7.0 New Prete Venege Steam Generator C ss 42
/
4.0
/
/
g,0
/
2.5 54 g
=..
e di-
~
E3 /.
O.0 CA 0.8 1A 1J 2.0
.5 10 3.5 4.0 4.5 6.0 New Probe Voaspe 4-30
Figure 4-10 Farley Unit-1 April 1997 Worn Probe Volts vs New Probe Volts
/
o Field Data s
..o-Linear Regression
.s
~..-
5
- -- -90W95% Tol. Band
/
+/- 25% of New
....- j.-
>o 4
/
e
.o o
M
,. g
f
$3 ti l
Retest Required l
Retest Required l
'f
. U.,' o v
n a
Jk d
h n?x 0*:.-
o
,o.
9 o
s..,,
~
..o. - -
t0. -
p} n.;*
[
r 1
v
,.. a o
3.
a POTENTIAL FOR REPAIRABLE TUBE LEFT IN SERVICE P
/
...o 0
0 1
2 3
4 5
6 7
New Probe Voltage r
in..ns.num 4_31
Figure 4-11
}
Farley Unit 1 1997 EOC-14 Evakestion for POPCD at EOC-13 1.0
-i.....-
- - -~~
"^*_,_.y....u-----',,,,,,.........---
0.9
=-n*__,_________________________%
~
_g
, 'Y 0.8
,P'- - - A u,
Y~~t+ x '- A 0.7 a'
g u
M V
^
50.6 d
h
- 0 F
.5 2;
x RPC Confirmed E 0
/a E
.4
(
- i 0.3
/
6
/
--+--RPC Confirmed Plus Not inspected x
0.2
)y J
0.1
-- * --EPRI POPCD
'd 0.0 l
l l
~
0 0.5 1
1.5 2
2.5 3
3.5 I
M Bobbin Amplitude Pepodrigin/299716:40 PM 4-32
7NViWNTTW3R&W66.T4. m Pa3 @ 3@ To 82S @iB@S3@
P.8,5/27 5,0 DATA BASE APPIRtn FOR ARC CORRELATIONS Correlations have been developed for the evaluation of ODSCC indications at TSP-
-locations in steam generators of nuclear power plants which relate bobbin voltage amplitudes, fires span burst pressure, probability ofleakage and associated leak rates.
The Westinghouse. methodology - used in the calculation of these parameters, documented in References 10.3,10.4 and 10.5, is consistent with NRC eriteria and guidelines of References 10.1 and 10.3.
The database used for the ARC correlations that are applied in the analyses of this report are consistent with the NRC SER applicable to the Farley Unit-1 EOC-14 inspection; it is documented in Reference 10.8. As noted in Section 3.3, the addition of-EOC-14 Farley 1 pulled tube data has a signi6 cant impact on the ARC correlations, especially the leak rate. An evaluation is being performed to assess the extent ofimpact of the updated database on EOC-15 leak rate and burst probability -
projections and it will be reported later. The results presented in this report are based on the ARC database documented-in Reference 10.8, and that database includes pulled tube data from the last Farley pulled tube===mination (EOC-13 L
Farley Unit-1 an1EOC-10 Farley Unit 2 avaminations). The leak rate data in the database represent room measurement ofleakage at prototypic SLB conditions (i.e.,
l
-leakage at SLB conditions was condensed and measured at room-temperature).
L Therefore, SLB leak rate calculated using the ARC correlations provides a volumetric rate at room temperature.
For the SLB leak rate correlation, the NRC recommends that Model Boiler specimen 542-4 and Plant J-1 pulled tube R8074, TSP 1 be included in the database. This database is referred to as the NRC database and is applied for the leak rate analyses of this report. The probability ofleakage correlation of Reference 10.8 for 7/8" tubes is applied in this report. The SLB leak rate data do not satisfy the NRC guidelines for a voltage dependent correlation, as discussed in Section 6.0.
i qAape\\ala97\\ala9onew.wp5 5-1
g
,g, g g g y,,
p
- 6.0 SLB ANALYSIS METHODS
- Monte Carlo analyses are used to predict the EOC-15 voltage distributions and to calculate the SLB leak rates and tube burst probabilities for both the actual EOC-14 voltage distribution and the predicted EOC 15 voltage distribution. These methods are consistent with the requirementa of the Farley Unit-1 NRC SER (Refennce 10.2) and are described in the generic methods report ofWCAP-14277 (Raference 10.3) and the prior reports for Farley Unit-1 (References 10.4 and 10,5), and are in accord with NRC Generic Letter 95-05 (Reference 10.1). Leak rates calculated with the WCAP-14277 methodology provide a volumetric leak rate at room temperature and they are compared with allowable volumetric leak rate at room temperature.
The NRC SER recommended leak rate database does not satisfy the requirement for i
applying the SLB leak rate versus bobbin voltage correlation. As discussed earlier in Section 3-4, the NRC requirement is that the p value obtained from the regression for the slope parameter be less than or equal to 5%. For the NRC recommended database that includes data from pulled tube tests performed in the current -
inspection (as well as 1996 Farley-2 pulled tube data), the p value is about 7.6%;
therefore, a leak rate versus voltage correlation cannot be applied yet. The licensing-basis analyses were carried out using a SLB leak rate correlation based on an L
average of all leak rate data independent of voltage. The analysis methods for applying this leak rate model are given in Section 4.6 of WCAP-14277 (Reference 10.8). A Monte Carlo analysis is applied to account for parametar uncertainties even though the leak rate is independent of voltage. This method ofleak rate analysis is similar to that ofdraft NUREG-1477 except for the uncertainty treatment. Leak and burst database documented in Reference 10.8 was used in the present Monte Carlo analyses and that database does not include the EOC-14 Farley Unit-1 data.
M ara =1. m.i. %..ps 6-1
M & '97 13:20 FR CW) ENGINEERING-W I1 4 C722 5 TO 82059935003 P.17/27
'4 5
k 7.0 BOBBIN VOLTAGE DISTRIBUTIONS i
This section describes prediction of EOC voltage distribution used for evaluating tube leak and burst probabilities at the and of the operating period. The calculation consists of establishing the initial conditions (i.e., the bobbin indication population
-diatribution) based on eddy current inspection data and projecting the indication growth over the operating period. Since indication growth is considered proportional to operating time, the limiting tube conditions occur at the end of any given time period or cycle.
The bobbin voltage distribution established for the BOC conditions is adjusted for m.easurement uncertainty using a quantity termed probability of detection, as described in the following paragraphs. Other input used for predicting the EOC voltage distribution and the results are presented below.
7.1 Probability of Detection
% number of bobbin indications used to predict tube leak rate and burst probability l
is obtained by aqust;ng the number of reported indications to account for l
measurement uncertainty and confidence level in voltage correlations.
This is accomplished by using a POD factor. A4ustments are also made for indications either removed from or returned to service. The calculation of projected bobbin voltage frequency distribution is based on a net total number ofindications returned to service, defined as:
N, N ars " POD-Nw+Nw, r
where:
Nu m = Number of bobbin indications being returned to service for the next cycle.
N, = Number of bobbin indications (in tubes in service during the previous cycle) reported in the current inspection.
POD = Probability of Detection.
Nw = Number ofN which are repaired (plugged) after the last cycle.
(
i Nw = Numberofpreviously-pluggedindicationswhicharedepluggedafter the last cycle and are returned to service.
The NRC generic letter (Reference 10.1) requires the application of a constant POD
= 0.6 to define the BOC distribution for the EOC voltage projections, unless an alternate POD is approved by the NRC.
e spes.1 97i.1.so
.. ps 71
- A.G 59 TTisI20 FR (W) OG!KiL@8K!b-W.M. 412 722 5889 TO 82059925002 p 18/27 7.2 Cycle Operating Time The following operating period values are used in the voltage projection calculations:
1 Cycle 18 = 489.4 RFPD Cycle 14 = 482.1 EFPD Cycle 15 = 485.8 EFPD (estimated) 7.3 Calculation of Voltage Distributions Bobbin voltage projections start with a cycle initial voltage distribution which is projected to the corresponding cycle final voltage distribution, based on the growth rate a$usted for the anticipated cycle operating time period. The overall growth rates for each of the Farley Unit 1 steam generators during the last two operating periods', as represented by their CPDFs, are shown on Figure 4-6. A Generic Letter 95-05 requirement is that limiting growth rate for the past two cycles of operation should be used in the projections. The 1995 - 1997 operation (Cycle 14) growth rates slightly exceed those of the 1994 - 1995 (Cycle 13) operation and are used to predict the EOC-15 bobbin voltage distributions. Further conservatism for the EOC-15 bobbin voltage prediction is provided by the use of the larger of the composite growth rate for all Sgs and the SG-speci6c growth rate in projecting EOC voltages for each SG. - The methodology used in the calculations of EOC bobbin voltage distributions is described in Reference 10.3.
For each SG, tb initial bobbin voltage distribution ofindications being returned to
-i service for the next cycle (BOC-15) is derived from the actul EOC-14 inspection results a$usted for tubes that are either (a) taken out ef:,ervice by plugging, or (b) have been recovered for Cycle 15 eervice by deplugging of tubes plugged in previous outages based on prior repair criteria. The Cycle 15 bobbin voltage popule. tion data is 6ammarised on Table 7-1.
It shows EOC-14 bobbin voltage indicati,ons; the -
subsequent plugged indications (which were in service for Cycle 14 and then taken out of service, albeit not all for reasons of ODSCC at TSP); those indications recovered for service from previously plugged tubes, which were deplugged during this outage, inspected rnd returned to service in accordance with ARC criteria (otherwise they were repaired); and the BOC-15 indications corresponding to a constant POD value of 0.6 as well as the voltage dependent generic POPCD.
7.4 ' Predicted EOC-15 Voltage Distributions i
The licensing-basis calculation for the predicted EOC-15 bobbin voltage distributions is performed for all SGs with a constant POD value of 0.6 in accordance with a NRC requirement. In ad.dition, calculations were also performed using a voltage dependent q:\\ ape \\ alas 7\\ala9onew.wp5 72
P.19/87 generic POPCD distribution developed based on bobbin and RPC data from 15 EC inspections at 8 different plants. Development of generic POPCD distribution is
- described in Reference 10.6. An updated POPCD distribution that includes the latest Farley Unit-1 inspection (EOC-13) data is shown in Section 10, but was not used in the present analysis.
A slightly modined tres.tment of growth data that yields more conservative results than the method applied for the last 90-day report was used to predict the EOC-15 performance. In the previous methodology, growth data supplied in the form of CPDF-was treated piece wise continuous ard the data was interpolated to _ pick growth rates. Such a procedure may not utilize the maximum growth value provided for the simulations at the frequency associated with the growth value. In the modified procedure, growth data is represented by a histogram and all simulations selecting the last growth bin utilize the highest growth rate. Thus the modified procedure yields slightly larger EOC voltages and, thus, more conservative leak and burst results. This level of analysis detail is below that described in the methodology _
report (hference 10.8).
The Farley Unit.1 steam generators BOC-15 voltage distributions used to predict the -
EOC-15 voltages are shown in Table 7-1' As mentioned earlier, the _ EOC-14 j
composite growth rate data shown in Table 4-4 were applied to SGs A and B (since their own growth rates arc vnaller than the composite growth rata) and its own growth rate distribution was usal for SG-C (aince it is higher than the composite growth rate). This approach is recommended in Reference 10.3.
Table 7-2 provides the EOC-15 voltage distributions predicted using the BOC-15 voltage distribution shown in Table 71. - As :.nticipated, the largest number of indications is predicted for SG-C,1766 indications for r. constant POD of 0.6. The assumed BOC-15 and the EOC-15 predicted bobbin voltage frsquency distributions are also graphically illustrated on Figure 7-1 for all three SGs. _ The largest bobbin voltage prah=A for EOC-15 is in SG C, and its magnitude is 14.7 volts for a constant POD of 0.6 7.5 Cornparison of Predicted and Actual EOC 14 Voltage Distributions The actual EOC-14 bobbin voltage distributions and the corresponding predictions presented in the last 90-day report (for EOC-13 inspection, Reference 10.4), are compared in Table 7-3 and on Figure 7-2. SG-C was predicted to be limiting for EOC-14 which is consistent with the actual measurement since this SG has the highest number ofindications as well as the largest indication found in the EOC-14 q:\\ap.\\ alas 7\\ala90new.wp5 7-S
uYE6F 's? i3:SnM (UT ENGINEERING-W.M. 412 722 5809 TO 82059925002 P.20/27 l
^
4
.8 inspection.- The total number ofindications for all SGs are overpredicted by 30% to 40% in the licensing-basis analysis with a POD of 0.6. Also, the licensing-basis analysis signi8cantly overpredicted the actual EOC-14 bobbin voltage population over 1 volt as well as the population above 2 volts in all three EGs. The overprediction for indications in virtually every voltage sise range demonstrates conservatism in the l
projection methodology. However, the single large voltage (13.74 volts) indication in Su-C was not predicted (largest predicted was 7.6 volts). As noted in Section 3.1.5 and 4.2, the indication with the largest growth in SG-C (R2085 TSP IH) as well as the largest indication in SG-A (R2077 TSP 1H) are located close to the pulser nozzle used for presure pulse cleaning. Although there was no direct evidence of fatigue -
1 on the axial crack face ar==hed for pulled tube R2C85, it is conceptually possible i
that pressure pulsing could have contributed to sepat ation of the ODSCC crack faces with a corresponding increase in voltage.
e i
i 1
i i
i
- e. pet.1.en.1.so w.wps 7. -
RJG & '9713:22 FR (W) ENGitEERING-W.f't. 412 722 5899 TO 82059925002 P.88/27 4
,l55:1!ilijlflllill!!!!!!:
5}i:35klit)tilltil l 585 y
I i.
1,...
l 3I!!i!lli I I i lji 11 ! !!! !!! !!!!!!!!!
3l5lSj 5il;!!!i!!i i i !
o 3
., 1
..;,..I i
l
..!.i....
i j
i i
l, i I I
i i
i l,- -
. I i :
.i.
i
! : 1 y
.. c,.=
-1. -._,.:-.,,
iI
-...-..i n n e, n.. 4
=
- -i,- -,-i,..! +,.:-
,]
i n.
,i r
51i111111 1l1i021! !5555555t!!!!!!!l!!l1! ! 5 ll i
g i
i II till'!illli ni,il!!"i,1:s t 5 5 5l5 55!!!;!!! l l 1 s
5 E
y]
i l
4 i
.j.,.l.. 'i tji 1
..=.'i t
y }
. q.
7 i
i.,
.l
.j
.i i
-n i
s s
33
.l,s il, i.
i s.
i i
0 1l l Elli!!i' l l i siil'il l !!! ! !!! ! Int!! !,1 t i l l i
i l 5!!! ill! !!!!!ill!!!! 5255l!!!5tl,11t i l 4
1 I
.l.
]
i 4
i.e..i l
.j-i
. g..
i I
l i
,j -=,...i.'.......
.l.
s t :
1 i
2 3s 3 0 3 5 00 0 $ D 5 3 2 2: 3 5 ] { $
2 2 2 : 0 2 222 j
l
=
mm
.___._A&-.
- -. - -- p-Table 7 2 (theet 1 of 2)
Farley Unit 1 April 1997 Voltage Dietributkm Pro}eetion for EOC.16 mesm meem 94eem Generster A Generator B Generator C Venape Projected indications Distributions et EOC *15. P00 = 0.4 02 0.02 0.04 0.0 0.3 0.54 1.01 0.30 0.4 3.77 6 65 1.71 0.5 14 45 _
24.41 7.21 0.6 34.59 53.M 20 56 0.7 62.70 96.34 45.98 0.8 91.57 18149 77.43 0.9 114.29.
155.40 104.47 1.0 124.31 106.41 131.45 1.1 133.05 165.44 144.29 1.2 129.19 153.71 154.17 1.3 119.18 184 40 156.14 1.4 106.21 111.97 152.28 1.5 32.34 90.37 142.05 1.6 78.85 70 92 136.31 1.7 45.73 54.42 107.84 1.8 54.0?
40.63 84.01 1.9 43.85 29.62 70.41 2.0 34.92 21.26 55.48 2.1 27.39 15.14 43.39 2.2 2],f1 10.92 33.89 3.3 15.79 7.94 35.43 2.4 11.69 5.97 19.04 2.5 4 to 4.65 14.03 2.6 625 3.51 10.20 2.7 4A5 2.72 7.34 2.8 3.33 2.08 8.38 2.9 2.43 1.59 3.76 3.0 1,78 1.19 2.49 3.1 1.30 0.88 1.94 3.2 0.N 0.64 1.39 3.3 0.68 0.46 1.02 3.4 0.49 0.32 0.75 3.5 0.84 0.23 035 3.6 0.24 0.16 0.41 3.7 0.17 0.12 0.31 3.8 0.13 0.10 0.24 3.9 0.11 0.10 0.18 4.0 0.09 0.04 0.14 Table Continues on Next Page swwme.u 9-6
FW t/ 'Y/ 13:23 FR (W3 D0lNEE.Rlik-U.f1. G1d 722 5009 TO 820'59925002 P.23/27 Table 7 2 (Sheet 2 of 2)
Farley Unit.1 April 1997 Voltage Distribution Projection for EOC 16 Steam Steam Steam Generator A Gerarator B Generator C vonege Projected Indications Distributione et EOC.15, POD = 0.8 Table Continues from Previous Page 4.1 0.08 0
0.11 4.3 0 07 0
0.08 43 0.08 0
0.07 44 0 05 0
0.05 4.5 0 04 0.70 0.04 46 0.04 0
0.03 47 0 03 0
0.02 48
_0 02 0
0 02 49 0 02 0
0 01 6.0 0.01 0
0.01 6,1 0.01 0
0 01 62 0 01 0
0 00 6.3 0.01 0
0 00 63 0 04 0
0 00 64 0 09 0
0.00 6.9 0 70 0
0.00 12 9 0
0 0.01 13.0 0
0 0.08 13.1 0
0 0 08 13.3 0
0 0.11 13.3 0
0,.
0.13 13.4 0
0.89, 0.14 13.6 0.30 0.14 13.6 0.14 13.7 0.18 13.8 0.24 13.9 0.01 14.1 0.7 14 7 03 Total 1416 84 1870.00 1766.03 6
6159P1# hanseessetene to e
AUG 27 '97 13823 FR (W) DG!tEERijHO-W.H. 412 722 5809'TO'8205992!c02'~ ~ V.~i M 2 [
~
Table 7 3 (Sheet 1 of 2)
Farley Unit.1 April 1997 Comparison of Predicted and Actual EOC-14 Voltage Distributions Steam Generator A Steam Generator B Steam Generator C EOC'18 I " I' Roc.14 goc.14 Eoc 14 50c 14 Prodspen Predetion Prodction g,,,,,
POD =0.6 Poo.0.6 Poo=0.6 yg "I"
Number of Indications 0.1 0.00 0
0 0
02 0 05 0-0.14 0
0.05 0
0.3 1.77 1
3 62 3
1.03 1
OA 14.14 16 30.81 25 7.63 6
0.5 41.67 36 54.56
$9 29.76 1e 0.6 74.81 63 96.34 104 62.63 88 0.7 104.87 08 132.39 97 98.73 77 0.8 124.06 86 149.18 12$
124.30 95 09 130.00 102 149.00.
Its 141.04 101 1.0
__129.14 88 139.99 ill 144.36 114 1.1 120.90 64 125.81 92 149.06 94
_ 1.2 107.28 60 107.88 71 142.12 108 1.3 90.76 42 87.51 49 127.93 103
_1.4 74.46 52 87.09 46 100.96 46 1.5 to 00 41 49.12 36 91.46 72 1.6 47.90
_ _ _ 25
_ 24.77 10
$9.16 32 36.06 23 74.27 44 1.7 37.07 30 1.8 29.42 21 17.63 0
46.16 30 1.9
_ 22.36 15 12.44 9
36.14 20 2.0 18.57 16 8.84 6
26.09 29
_ _ 2.1 11.99 8
6.31 3
10.85 7
2.2 8.54 9
4.80 2
13.34 13 2.3 6.04 4
3.20 8
9.30 7
2.4 4.33 4
2.33 2
6.43 8
, _ 2.5 3.14 2
1.72 4
4.44 3
2.8 2.32 2
1 2
3.19
_, 3__
l 2.7 1.74 3
1.0 2
2.33 1
I 2.8 1
,,,0 1
2 1.76 0
l 2.9 1.0 1
1 0
1.34 2
30 1
2 0.s 0
1.08 2
3.1 1
0 0.4 1
0.86 0
3.2 0.5 0
0 0
0.71 1
3.3 0
0 0
0 0.69 0
3.4 0
0 0
0 0.49 0
3.6 _
0 0
0 0
0.4 0
3.6, 0
0 0
0 0
0 3.7 0
0 0
0 0.3 0
l 38 0
0 0
0 0
1 39 0
0 0
0 0
0 4.0 0
0 0
0 0
0 Table Continues on Next Page l
e w.=a,>
m==, m s e ti m m 7-8
-,.,n
,n
AUG 27 '97 13123 FR (Wf DG!IEER!fG-li.M. ~413'722 5805 TO B20599250f.'
P.AS/87 Table 7 3 (Sheet 2 of 2)
Perley Unh 1 April 1997 Comparison of Predicted and Actual EOC 14 Voltage Distributions SteamGeneraler A
$ team Generator B SteamGenerster C 300>14 8004 4 EOC 14 acc. 4 e
a p m menen g,,,,,,
m Poe 68 Pop.s.s yg" Number ofIndications Table continues from Previous Page 41 0 11 0
0.18 0
0.19 0
42 0 10 0
0.14 0
0 17 0
44 0.00 0
0,14 0
0.16 0
4.4 0.00 0
013 0
413 0
4.S 0 00 0
O tt 0
0.12 0
de 0 00 0
0.11 0
0.11 0
4 <7 0 07 0
0.10 0
0.10 0
48 0 07 0
0 10 0
0.00 0
49 0.07 0
0.00 0
0.00 0
6.0 0.07 0
0 08 0
0 08 0
5.1 0.00 0
0.08 9
0 OS 0
SJ 0 05 0
0.00 0
6 05 0
83 0 00 0
0 07 0
0.08 0
54 OM 0
OM 0
OW 0
S8 0 06 0
0 07 0
0 00 0
56 0.06 0
0 07 0
0E7 0
5.7 0.00 0
0 07 0
0.18 0
5.8 0 00 0
0 00 0
0.18 0
S.9 0.11 0
0.13 0
0.18 0
80
_ 0.15 0
0.18 0_
0 10 0
0.1 0.18 0
021 0
022 0
63 0.19 0
0.13 0
033 0
6.3 0.03 0
0.00 0
0.23 0
64 0 00 1
0 00 0
OJ2 0
85 J 70 0
0 70 0
0.21 0
$a 0.00 0
0 0
0.19 0
6.7 0 00 0
0.3 0
0.16 0
68 0
0 0
0 13 0
6.9 03 0
0 0 10 0
7J
_. 0 0
0 0 70 0
7.8 0
0 0.30 0
13.74 0
0 1
TOTAL 1278.87 320 130040 1014 1544 07 1140
>10V 654,36 est S$3 96 373 eaaJs
$42
=2Y as 38 38 00 27.48 34.00 71 90 44.00 7-8 p
,e
,ewiin m 4
vJLrbv'Nt/%fRADAJWFLfidR1tWdMB.W. 412 723 5899 TO 82059935003 P.26/27 s
Pigure 71 Farley Unit.1 Aprt!Ipe7 Comparison BOC 18 Bobbin Voltage Distributions with Predicted EOC 18 Distaibutions Dee.a. Ge.w. tar A - POD. 0.0 100l
,w
$4e
[
O Aneumme soc.is i.
- l. '"
'1
~
me soc.is i.,o, q]
t e
a I
i 4
ll j.
- Ib i....
- : : : s a s a s : : s a a.v.:a :. ::::::::::::
e.w si.
o...r.t., s - roo. e.e too 500
...O_
i.o o Aa=====d soc is i.
iro aw soc is
, i o.
g,o j ee E-n o
Il r rI1]La......__._.
t
- : : : : : : : : s: : : : a: :
- 3
- n. w v.n s a oe
- r. = c - Pon.e.s see iso 140 C Amnesnes DOC-il jiao t
,oo
.3 f
l 1
3 Pre 64sied 50018 3
i i
so p
.o 0-l f
s l _I A N
)D)ID$$5 II
$I5..
Nh.
e e
~ ~ ~
7-10
UJ$bttt/ DJ 24 M] (U) DallEER!fG-W.M. 412 722 5009 TO 0205992"002 P.27/27 Mgwe73 Farley UaA61 April 1907 Comeparison of hedleted EOC 14 Bobbia D6stributions with Actual Measwed Distnbutions steam Gomerater A 140 l'
i io.
l 3
o e, o.e t1 4
1 l
,i
,I so s Actuni m--
i 20 Ilbs. _
_ Il l lI i
o D $ I U 5 D 22 U SN E E O I D 3 2 E E U 3 sehWn vennes Steam Generator B
,g 140 120 U
too to j.,
.o no 11 I S h n.a _ m..
E $ $ 0 5 0 22 22N2 3 7 3 D 2 2 3 3 3 senednvenese Steam Otaerster C 160 140 120 O ProdNeed 100 90 1
. Aeauel
'O a
40 -
O E
ao 1
1,h.....
i ndean
- sa: : ense noemen vommee 7 11
- TOTAL PAGE.27 **
tbh%7A'M P.03/21 8.0 TUBE LEAR RATE AND TUBE BURST PROBABILITIES i
8.1 Calculation of Leak Rate and Tube Burst Probabilities This section discusses tube leak and burst probability analyses using voltag distributions projected for the end of the operating period. The calculation utiliz correlations relating bobbin voltage amplitudes (either measured or calculated) to span burst pressure, probability ofleakage and associated leak rata for ODSCC indications at TSP locations. The methodology used is documented in Refe and is consistent with NRC criteria and guidelines of References 10.1. The current NRC approved SLB leak rate database does not result in a statistically accepta leak rate correlation and the analyses are performed with leak rate independent o voltage. The calculated leak rates are volumetric rates at room temperature a should be with compared with allowable leak rates at room temperature.
8.2 Predicted and Actual Leak Rate and Tube Burst Probability for EOC 14 Analyses were performed to calculate SLB tube leak rate and probability of burst for the actual bobbin voltage distribution at EOC 14 '(with no growth projection applie previously presented in this report. The results ofMonte Carlo calculations performed based on the actual voltage distributions including NDE uncertainties are shown on Table 81. Projections for EOC-14 conditions for all three SGs presented in the last 90-day report are also included for comparison in Table 8-1. The allowable SLB rat 4 for the last operating cycle (Cycle 14) was 11.4 gym (at normal RCS coolant avera temperature) or 8.2 gpm (at room temperature).
A comparison of the EOC-14 actuals with the corresponding projections indicates the following:
1.
Tube R2C85 with the 13.74 volt indication was pulled and burst / leak tested in the laboratory. The burst pressure of 3990 psi exceeded the 2560 psi SLB condition. Since this indication was demonstrated to satisfy burst requirements, it is excluded from the condition monitoring (EOC-14) burst probability analysis.
Similarly, the SLB leak rate was measured to be 0.72 gym (room temperature),
and this leak rate is added to the SG C leak rate predicted without including R2C85 in the analyses. The resulting EOC 14 SLB burst probability and leak rate for the measured distribution is 3.8x104 and 7.9 gpm. The pulled tube demonstration of adequate burst capability for R2085 results in a reduction in the burst probability by more than a factor of 10 (from 5.2x10 to 3.8x104 )
compared to statistical analyses including this indication. Inclusion of the pulled q:\\ ape \\ala97\\ala90new.wp5 8-1
w eareiXe P ea e tube leak rata for R2086 results in an increase from 7.6 to 7.9 gpm (room temperature) at EOC-14.
- 2.. SG C was pr(ected to be the limiting steam generator for EOC-14 based ou EOC-13 EC data and 80-0 was also determined to have b highest SLB leak rate as well as burst probability based on actual EC bobbin measurements at EOC 14. The SLB leak rate of 7.9 gym (room temperature) based on the actual voltages for SG-C is less than the priections with a POD of 0.6, and it is below the allowable limit at room temperature (8.2 gpm).
3.
The tube burst probability of 8.8x104 for SG C based on actual voltages and l
pulled tube R2085 results is less than the prqected value of1.4x10'8 l
and much lower than the allowable limiting value of 104 When b R2C88 indication is treated statistically rahr than using the measured burst capability, the l
difference between the actual and pr(ected burst probability in-SG C is attributable to only the R2C85 indication (13.74 volts) found above the predicted peak voltage (7.6 volts).
4.
For SGs A and B, SLB leak rates and tube burst probabilities based on the actual voltages are less than the prdoctions with a POD of 0.6. The leak rates are also below the allowable limit at room temperatum, and tube burst probabilities are also below the NRC mporting threshold of1 x 104 In summary, the limiting SLB leak rate (7.9 gpm) and tube burst probability (3.8x10d) calculated using b actual measund EOC-14 bobbin voltage distributions and pulled tube resul's are below their corresponding limita (8.2 spm and 108, respectively)..
8,3 Njected Leak Rate and Tube Burst bbability for EOC 15 Using the methodology previously descrbd, calculations have been conducted to predict b EOC 15 performance of all thme steam generatora in Farley Unit-1, and b results are s==7 sed in Table 8 2. EOC 15 bobbin voltage distributions as well i
as the leak rates and tube bunt probabilities bassa on those distributions are predicted. The priected leak rates are compared with allowable leak rate at room temperature. The allowable leakage limit for Farley Unit 1 has been inemased to 19 gym (at normal RCS coolant tamperatum) or 13.7 gym (at room temperature)
- As discussed in Section 4.3, an allowance was made to account for potenti ll hi h growth for indications in deplugged tubes nturned to service for Cycle 15.
ay g er A
g:\\apc\\ala97\\ala90new.wp5 8-2
}.__.
~'
~
CG5 tN %N DJ8F0 FR M EPO!KLWPU.M. 413 722 5089 TO B2059925002 P.04/81 composite growth distribution obtained by combining the Cycle 14 growth distribution with the Farley Unit 2 Cycle 11 growth rates for deplugged tubes presented in Refenace 10.7 was used for Cycle 15 projections. Growth rates for active and deplugged tubet were weighted by the number of each type ofindication returned to service. Also, as recommended in Reference 10.7, indications in both active and deplugged tubes were considered together in the projection analyses.
Since SG-C has both the highest number of indications as well as the largest indication returned to servios for Cycle 15, it was prcjected to be the limiting SG.
The predicted EOC 15 SLB leak rate for SG-C based on the present licensing basis database and method (constant POD of 0.6 and a leak rate independent of voltage) is 15.4 gpm (room temperature) which exceeds the current licensed limit of 13.7 gy at room temperature. However, the current licensing methods are very conservative.
The following shows the reduction in leak rate as a result of methodology updates that are currently under review by the NRC:
j Current basis Blased leak rate parameters, POD =0.6 and no leak rate dependency on voltage 15.4gpm Unbiased leak rate parameters with no voltage dependency 11.4 gpm i
Use of voltage dependent POD with biased leak rate parameters without correlation 9.9 gp m Exclusion of French data resulting in a voltage dependent leak rate correlation and POD =0.6 2.7 spm Use of voltage dependent POD and leak rate correlation 1.6 spm It is seen that any one of the methodology updates results in a leak rate below the
-allowable limit of 13.7 gpm.- Using the current licensing basis methods, it is estimated that a projected leak rate equal to the allowable limit would be obtained in about 305 EFPD. This provides ample time for the NRC to complete reviews of the above noted methodology changes. It is expected that one or more of the method changes will be approved prior to reaching 305 EFPD. Alternately, the allowable SLB leak rate limit can be increased to accommodate the projected leak rate by a reduction in the coolant activity limit.
[The EOC-15 SLB leak rates predicted for SGs A and B are well below the allowab limit. Also, EOC-15 SLB tube burst probabilities for all three SGs are below the NRC -
reporting guideline for tube burst probability of 1.0 x 8 i
q:\\ ape \\ala97\\ alas 0 mew..,s -
8-3
03.!fbt'vGJIC9 W2 (W) LWRK_sLMK6U.R GPJ V23 5809 TO B2059925002 P.05/21 The use of a constant POD value of 0.6 is conservative beyond i volt and it is unrealistic beyond about 3 volts where POD is likely to be unity. So, the EOC-15 SLB leak rate and tube burst prebability for all SGs were estimated using a voltage dependent generic POPCD distribution presented in Reference 10.6. For SG C, the EOC-15 leak rate was also estimated using a step distribution for POD with POD =0.6 up to 10 volts and POD =1 above 10 volts. These resulta are also shown in Table 8 2.
The step POD distribution affects only the contribution of the 13.7 volt, repaired indication in SG-C which is large enough to reduce the projected SLB burst probability for SC-C by 50% (from 9.9x10'8 to 6.6x10 ), further increasing margin to 4
the allowsble limit. With the voltage dependent POPCD, the SLB burst probabil for SG-C is further reduced to 4.6x10, and the leak rate is reduced to 9.9 gpm (room 4
temperature) which is within the allowable limit of 13.7 gpm at room temperature.
It was noted in Section 3.3 that the addition ofEOC-14 Farley-1 pulled tube data has a significant impact on the ARC correlations used for leak and tube burst analyses.
An evaluation is being performed to assess the extent ofimpact of the updated database on EOC 15 leak rate and burst probability projections and it will be reported later.
In summary, with the exception of SG-C EOC 15 leak rate, SLB leak rates and tube burst probabilities projected for EOC 15 for all three SGs using the present NRC-approved database and method meet the SER limits for Farley Unit-1. The use of more realistic estimates for non log leak rata distribution parameters yields SG-C EOC 15 leak rate within the allowable limit. Results based on voltage dependent POPCD show that the margin between the EOC 15 predictions and acceptance limits are substantially higher than that indicated by the current licensing method.
qAapc\\ala97\\ala90new.wp5 8-4
p
-'WW '<n'isiseWikrEiei AEERUN6~M.'~55f*f52 5009 iO 820555bIOb2 N.ON2.1 l
i i
1 Table 81 4
Farley Unit 11997 EOC.14 Outage 5mmunary of Calculations of Tube Leak Rate and Burst Probability i
Comepartoon of Prtdootions Based on Actual Bobbia Voltage
}
i Leak Rate Independent of Bobbla Voltage 4
1 i
i 4
Nuamber I
SLB j
Steam POD of Max.
Burst Probability uk
}
Generator Indications Volts?
Rate i
(gym) 1 Tube 2 Tubes i
l I
EOC.14 PROJECTIONS i
}
A 0.6 1276 6.9 7.5 x10'*
2.5 x10*
7.1 B
0.6 1310-6.7 7.1 x10
1.9 x10
6.0 t
i j
0 0.6 1545 7.6 1.4 x10 8 1.9 x10*
- 10.2 4
?
EOC 14 ACTUAIA i
L.
3 A
1 921 6.9 6.2 x10*
- 1.9 x10
5.3 B_
1 1014 3.7 1.9 x10
< 4 x10
4.2 t
C 1
1140
'14.9 5.2 x 1 0~8
< 4 x10
7.6 1
C(s>
1 1139 4.0 8.8 x104
< 4 x10*
- 7.9(*)
}
,htm; i-(1)
Wltages include NDE uneortelaties h Monte Carle analyses and eseeed measured v i
(2)
A4usted for POD.
(3)
RSC85 TSP 1H luie=da= sueluded h Monte Carlo analysis. Fellowing the tube pull, this
.i i'
i indaeation was buret and leak tested. b burst pressure useeded SLB preneurs differentials and can be excluded as a potential rupture at EOC 14.
j
-(4)
N leak rate measured for pulled tube RSCs5 was a4usted to SLB oenditions (0.72 sym) a
(
added to the leek rate predicted by Monte Carlo analysis escluding R2045.
l i
i k
l' e
t k
i L
espe \\atseneiaso
.wps-8-5 i
w-.
_____.____a,_.u_..
UGLifbt' %/ DJ841 OU M (@GitEMH43 ul.W. GW 722 $889 TO 82059925002 P.07/21 4
4 Table 8 3 i
Farley Unit.1 Summary of Prqjected Tube Leak Rate and Burst Pmbability for EOC 15 Leak Rate Independent of Bobb Volts. Projected Cycle 15 (485.8 EFPD) i 2
Steam POD No.of Max.
Generator Indic.
Volts One or Leak f
ations*
1 Tube W re Rate Tubes
(,p )m
)
L ROC 15 Projections Using Constant POD =0.6 _
A*
1416.3 13.5 2.49n10*
2.50=10i 10.6 j-B*
1570.0 13.4 2.21 10i 2.22x104 9.2 WCAP 14277 C"
1766.0 14.7 9.86=104 3
9.92=104 18.7 Method N
14163 13.5 2.49m10*
2.50x10' 7.9 Unbiased leak B*
1570.0 13.4 2.21x104 2.22n104 6.9 Co a4 i-CO 1760.0 14.7 9.85=104 9.92=104 11.4 CO 1766.0 14.2 4.53=104
]
4.53x104 13.4 Dwation. 278 trro t
WCAP 14SM Edwd 00 1765.3 14.1 6.60w104 6.64=10i 15.4 POD =1 Above 10 V 20015 Using Voltage Dependent EPRI POPCD N
1019.1 13.2 1.47 10*
1.47m104
- 6.8 Unbiased leak _
i B*
PopCD 1175.1 13.2 1.56x104 6.1
["
1.57x104 g
C" 1231.5 14.0 4.62x104 4.66=104 9.9 EOC.15 Using Voltage Dependent Leak Rate Correlation
- i' C"
0.6 1766.0 14.7 9,56=104 9.63x10*
2.7 leak rate i
Correlation applied C"
POPCD 1231.6 14.0 4.26=104 4.27x104 1.6 Notan 4 -
(1)
Number ofindications adjusted for POD.
4 (2)
Volumetric leak rate adjusted to room temperature, I
(3)
All SG composite growth rate distribution applied.
(4)
SG C specise growth iste distribution applied.
(5)
SLB leak rate correlation obtained by excluding the French data applied.
1 i
s empa niamalas0new.wps 8-6 i
..,,e,w.y
,.,,,~.-e_
ne.,--,.,--.....,,,,,,,.,..yw,,,-w__,,~,.
,...,.--+.-m
,....-,-,,,,-,,-v-,.--
.-..y.---,
(;MbfTF13i4& FR M ENQINEEQlNO-U.M. 41"t 722 5009 TO B2e59935002 P.08/21 9.0 COMPARISON OF PROBABILITY OF PRIOR CYCLE DETECT FOR 19 INSPECTIONS,8 PLANTS WITH EPRI POD The evaluation of the POPCD for Farley Unit 1 is described in Section 4.4 At this time, POPCD evaluations are available for 19 inspections at 8 plants, includin evaluations for Farley Unit 1. The available data include 12 inspections of plant with 7/8" diameter tubing and 7 inspections of plants with 3/4" diameter tubi section su===Azes these POPCD evaluations for comparison with results of a E study that examined detection probability for a dual analyst team. The POPCD evaluations performed since 1992 show sigal6 cant improvement over the earlier l
assessments which represent the Srst ARC inspections.
Bobbin data analysis guidelines (Appendix A guidelines) havt been revised since the Srst inspections to reflect the initial ARC experience.
Thus, it is appropriate to assess POPCD for i
j inspections performed since 1993, Fifteen of the 19 inspections for which POPCD h been evaluated were p. feed since 1992.
j Table 91 shows the combined POPCD evaluation for plants with 7/8" diameter 1
i tubing and it includes results for 9 inspections performed sinos 1992. These data are
{
also plotted in Figure 9-1, and they include data from the present Parley Unit l'
}
assessment (EOC-13 results representing 1995 inspection) as well as the data for the j
EOG11 and EOC-12 inspections (data for EOC 12 inspection is reported in Reference i
10-2). The POPCD value approaches unity at about 3 volts. The addition of Parley-j
-EOC-13 POPCD distribution to the generic database for 7/8" tubes increases the average value independent of voltage to about 0.71 from 0.69. The increase in the i
average value is noticeable because the population in voltage bins abo.e 1 volt t
increased by 50% to 100% and POPCD values for these bins asceed 0.8. Howev relatively low POPCD values in the 2 to 3.2 volta range calculated for Farley 1 EOC-18 inspection (see Table 4-8) has lowered the composite POPCD for 7/8" tubes in that voltase range. This is evident in Figure 91 where POPCD is compared with the results of the EPRI study on detection probability for a dual analyst team. In the.
i.
volt range 1.2 to 3.2 volts the generic POPCD increases only slightly from 0.86 to 0.89 where as the EPRI POD study results increase from about 0.9 to 0.98.
Table 9 2 and Figure 9 2 show the combined POPCD evaluation performed for plant with 3/4" tubing considering 6 inspections since 1992. These results tend to suppo a POD approaching unity above about 8 volts. The POPCD assessment is in very good agreement with the results from the EPRI sady on detection probability for a e
dual analyst team. The average POPCD independent of voltage is about 0.64 which is in general agreement with-the NRC Generic Letter 95-05 proposed voltage i
independent POD of 0.60.
i qdapc\\ alas 7\\ala90new..,s _
. 9. I r
, - ~,., -, -, -,. -. - -, _. -, - -
-...-.-__n.+nn_.m..n.- -, - - _, - - -,, _,
.-.a.
.n~~~~_scmn,-,.-.
.m,.
-w+-
002 t/tTb/ DJ842 FR (u) ENGllEE. RING-W.M. 412 721 5899 TO G2059925682 P.09/21 N dannition of POPCD includes indications which were not present at the prior I
inspection and thus could be expected to be somewhat lower than the EPRI POD which is based on "eaport" evaluations ofinspection rwults and does not include indications clearly below detectable levels.
i
- N combined data for the 15 inspections since 1992 are given in Table 9-3 and the POPCD evt.luation is shown in Figure 9-3 for RPC confirmed plus not inspected indications.
It is seen that the inspections since 1992 yield a POPCD in good agreement with the EPRI POD which was a-1994 evaluation. POPCD supports a i
POD approaching unity at about 3.5 volts while the EPRI dual analyst detection probability is about 0.98 at 2 volts and unity at 3 volta. Figure 9 3 also includes i
POPCD evaluated at the lower 95% con 6dence limit on the data for indiv j
voltage bins.
The POPCD evaluations shown in Figures'9-1 to 9-3 :.re based on the de6nition of I
" truth" as RPC conErmed plus not RPC inspected indications. Since many of the indications not RPC inspected would be espected to be found NDD ifinspected, this represents a lower bound POPCD evaluation.
Figure 9 4 shows the POPCD i
evaluation for all 15 inspections since 1992 based only on RPC con 6rmed indiostions.
This results in a signi6 cant increase in POPCD below 1.0 volt and a modest increase l
above 1.0 volt. The data of Table 9 3 show about 600 to 11000 indicatims voltage bins below 2 volts, about 300 between 2.0 and 3.2 volts and about 10 indications above about 3.2 volta. Thus, the collective data provide a substantial database for de6ning a POD.
i The results of Figure 9-3 clearly suppott an increase in the POD for ARC applications above the POD = 0.6, independent ofvoltage, required by NRC Generic Letter 95 06.-
For indications above 1.0 volt, the POD exceeds 0.9 and is 0.95 to near unity at 2.0 volta. A POD of 0.6 is only applicable to indications below about 0.6 volts.
t A voltage dependent POD distdbution has been developed for ARC application by -
evaluating POPCD at the lower 95% con 8dence level and the mid voltage of each -
voltage bin. The result is then smoothed to obtain the POPCD distribution as shown in Figure 9 5. This POPCD distdbution is tabulated in Table 9-4 and compared with i
t the EPRI dual analyst detection probability in Figure 9-6. Table 9.4 shows both the POPCD distribution presented originally in Reference 10.6 and the updated POPCD distribution presented in this section which includes results for 5 more inspections including the Farley.1 EOC-13 and Farley.2 EOC 11 inoputions. - Although the updated distribution indudes data from another 10000 indications, the diflistence-between the two seta of POPCD does not exceed 0.01 which provides conadence that the POPCD distribution is based on a sufBeiently.large database.
q:\\ ape \\ala97\\ala90 mew.wp6 92 1
l~~.-,,--
,n.n--w,e.,.-,ca.w---
a.;- ~ n
,.n
..,,,i n d,A,.d_-
e,n.,.,-n---m~,.n,,-
-n--w,-w,,.--,_,
-~
I E l.i.l 5 i E E 3 3 3
- r g
1 4 S R E I E E E E E E 1
o a o a o o a d e a a D
E
- R ! ! B E E 9 3 8 2
- g]
J E.E.EE.l.i E l E E E.
g.
n
}> 3) 1 52 j
. n a s u n n a = a. s s
11 1
E I
B 4 @i l
?
j 1
} Eflf B E E ! ! I I B : = = = l ! j ri 1
i s
a 1
I
}
O
- 8 I S [* h 3 *
- O h h
m
- l_l, ai; e n..... -. _ _ i 3
l N b N R a N E C - m a a E
E a
! ! Y "i ! E E E E E E I d 82 l
-lg t
6 : : : s u n a n n a E i
i
-L.
I
3.gg,,7g g
g 1
I l l i l l l R = =..
li}f i t ! I I ! I F 5 "
- 4 1 1 j n.s.a a l n.e i t.i.g i t
3 e ! ! ! ! ! ! ! s e e e s i n g s p = =
sj 1
I l l i l ! ! ! ! i !
i V
c kl 1
gj1)
L
. n s s e i s s a n g I 11 3
f 2c,E fldj f t} E j 5 1 1 1 i 1 8 $ = - - - l ! , 1 3 g If n ! R 5 E t n e... I g L i ,1 4 t Ig ! k :! $ $ $ $ 9 h h h I
- : s e :
n : e J l l r ,,--...-,m_e., ....e.... .--._y __..~-.,,.m,.-y. ,,re,, ,,,...,..,..,,7,-. ,,mm.- .i4,.y .. _ _ = _,-.,, .,.,,,I.-,
M 27 '97 13143 FR (W) DGitEERitG-W.M. 412 722 5083 TO B2059925002 P.12/21 1 1 !E i l la s. r I la} i l li l i li i k i " "3 1 ! !! ! ! ! ! ! ! ! ! : l ,il I} 1 ! ! i l ! ! ! ! ! ! ! ! : 1 a N [ 1l} l t r t i s s e x = i l 2<l 3 i gei t e.t j ll f* Il}} Il l i i ! ! a a e r - l l 1 R lg} 3 g 1 a l t i E 5 E n e s * ! I 3* X]l} ! i ! ! a i n a - - - - l a 3j il ...., a... -..., >l4 ! ? ? ? ? 2 a? ? ? ? ? n : : ( 3 2 ~ i
m & "Ut/ 13:43 FR (W) DGltEERito-W.M. 482 722 5889 TO B2059925002 P.13/21 Table 9-4 Comparison of EPRI POPCD with EPRI POD Study VnNage EPRl' D NP 7M Bln Study Addendum-1 Updated 0.1 0.30 0.24 0.24 0.2 0.38 0.34 0.34 0.3_ 0.49 0.44 0.43 0.4 0.57 0.53 0.52 0.5 0.62 0.62 0.61 0.6 0.66 0.67 0.67 0.7 0.71 0.73 0.73 0.8 ,, 0.76 0.77 0.77 0.9 0.80 0.81 0.81 1 0.83 0.83 0.83 1.2 0.90 0.88 0.87 -- 1.4 0.93-0.01 0.90 1.6 0.96 0.92 _ 0.91 ~ 1.8 ,0.98 0.93 0.92 2 0.984 0.94 0.93-3 1.00 0.98 0.98 3.5 1.00 1.0 1.0
- Dual analyst detection probability study 9
PocomatTowe H anewassma 9-6
EygWDJiGRUT8TIRRRigypu,M, did Vd2 5099 TO B2059925002 P.84/88 k
i M j l E,g I i k i l 8 = l g -p I u i b j f]. i l u 4 o i _E I Ai t l j i. \\ i s -8 a I.g k I l A ,s i 3 8 y s I .l e v i 1 3 1 3 N I % i - s, a = 1
- s g
i K. ' t ~. s x,,w .a l O h 3 E =n=m pom w a l
Figese 9-2 4 Comebleed POPCD F.valsation for 6 Post-91 Inspections forX4" Die Pl==8* POPCD Besed on RPC CenGrased Plus Not Inspecsed *- " "' ; 1.0 y , g. y. O.9 4 .n 0.8 E i v s < W-pl i 0.7 - .I F D.6 9 .F g T 0.5 - h. k s' Q .i 3< 0.4 h Y I E3 ~I h --+--Conibined POPCD from 6 Assessments e -0 0.2 - g 0.1 --m-- EPRI POD Study 8 y g = 0.0 1 0 0.5 1 1.5 2 2.5 3 3.5 ? Bobbin Anspliende necesurg,e zwiserw:2r eu 9-8
hth 5089 TO 82059925002 P.16/21 a Q IHy .I 1 I I l i i I .. y I 5 [ h I '8 m W e i i jS l sj li W h a. g I i d =E y mu q<4 i .n g \\ jj i d1 s 1 s s{ \\ W k<4p m \\ m s 9 i s 3f 4 al .\\ l \\ My .Li\\ abdp i's b _(,; o
- :s
- % l b' =s
.;g ~4 o a. e-w v. n n o o o o o o o o o o o o golpgeq)e 4!ggeqo4[ n
(;T51W~"vt/T3TL4 FR (W) DGINEERING-U.M. 412 723 5809 TO B2059925002 ' P.N/35 7 q l .7 i 1 j m I e i l n ( I l ~~ M i g e f 5 i 7, j l + 1 1 I i 1 i 4 j}}3 -N i! M i ! o "E 1 1 4 5 4 i I 3- ]l "!\\'t l N u b \\ N 's L \\ ~ s i \\ S =,;,., 4n = ? I _e g S a a a a a a a a a W70 AflM*qoM a
-_____- ggg44mWUUTidninra-u.M. 482 v2a $889 TO B2059925002 P.8B/31 l I i ] i i k l J l i .]f ] ) \\ l b -3 al i i s 8 I jI [i l .g ll 4 o I i 235 't -a h., ?.4 e c h \\ ( 1 a 4 d 4 1 $b .n$ e a f !\\ .iu e ~ h A !N = s; m
- s N --
1 3 i s a a s l 1
P.19/21 ~~ s i l l 1 ng y I J 'l ! l j lg h f !}i ~~ s 11 li1 1 0e I -~ gE i i fj i'1 0 {U hi i A L \\1 -nA 'is un l \\ \\ km )( \\ kA% -9 s"Q s %b %s i o a a.9=mp amqw
tssbrwsurEnirmiiR6Ew. an e CD To s2e5992sma P. w as n? REFERENCES 10.1 NRC Genede Letter 95-05, " Voltage Based Repair Critaria for Westinghouse Steam Generator Tubes Affected by Outilde Diameter Stress Corrosion Cracking", USNRC Omce of Nuclear Reactor Regulation, August 3,1995, 10.2 Safety Evaluation Report, " Safety Ev 4ation by the Omce ofNuclear Reactor baulation Related to Amendment No.128 to Facility Operating License NPF 2, Southern Nuclear Operating Company, Inc., Joseph M. Farley Nuclear Station, Unit 1, Docket No. 50 848", United States Nuclear Regulatory Commission, May 19,1997. l 10.3 WCAP 14277, Revision 1, *SLB Leak Rate and Tube Burst Proubility Analysis Methods for ODSCC at TSP Intersections," Westinghouse Nuclear l Serview Division, December 't#96. l 10.4 SG-96-01-003, "Farley Unit 1,1995 Interim Plugging Criteria 90 Day Report," Westinghoum Electric Corporation, January 1996. 10.5 WCnP 12871 Revision 2, "J.M. Farley units 1 and 2 SG Tube Plugging Criteria for ODSCC at Tuba Support Plates", Westinghouse Electric Corporation, Propdatary Class 2, February 1992, 10.6 Addendum 1 to EPRI Report NP 7480 L, " Steam Generator Outside Diameter Stress Corrosion Cracking at Tube Support Plates Database for Alternate Repair Criteria," October 1996, 10.7 SG-97-03-001, "Farley Unit 2 1996 Alternate Repair Criteria 90 Day Report," Wwtinghouse Nuclear Services Division, March 1997. 10.8 NSD-SGD 1212, "EPRI ARC Databases for 3/4" and 7/8" Dia. Tubes and Updated ARC Co Telation for 7/8" Dia. Tubes," Westinghouse memorandum dated February 26, 1996 Tennessee Valley Authority. transmitted to Duquesne Light Company and 10.9 EPRI Report NP-7480-L, " Steam Generator Outside Diameter Stress Corrosion Cracking at Tube Support Plates - Database for Alternate Repair Criteria, Volume 1: 7/8 Inch Diameter Tubing," December 1993. 4:\\apc\\ah97\\ala90new.wps 10-1
M 27 '9713f45 FR (W) EN3ttEERlHG-W.M. 412 722 5809 TO B2059925002 P.21/21 10.10 Letter from D. Morey of Southem Nuclear Operating Company to U.S. Nuclear Regulatory Ca==l== ton, " Joseph M. Farley Nuclear Plant Unit 1, Tube Pull Eddy Cunent Analysis," Docket No. 50 M8, dated June 20,1997.. .10.11 Letter from B. W. Sharon, Nuclear hgulatory Commission, to A. Marion, i Nuclev Energy Institute, dated February 9,1996. } } g:\\ ape \\ala97\\ala90new.wes .10-2
- TOTAL PAGE.21 **
.-_-.-,_,._,.c---- .-}}