ML20134P218

Rev 0 to Fracture Mechanics Evaluation & Weld Overlay Design for Recirculation Sys Piping in Ja Fitzpatrick Nuclear Power Plant

ML20134P218
Person / Time
Site:	FitzPatrick
Issue date:	04/26/1985
From:	Kuo A, Riccardella P, Tang S STRUCTURAL INTEGRITY ASSOCIATES, INC.
To:
Shared Package
ML20134P200	List: ML20134P198 ML20134P218
References
SIR-85-015, SIR-85-015-R00, SIR-85-15, SIR-85-15-R, NUDOCS 8509060199
Download: ML20134P218 (50)

Text

_ _ _ _ _ _ _

s- .

Report No. SIR-85-015 Revision 0 51 Project No. PASNY-05 April 25, 1985 -

E l

Fracture Mechanics Evaluation and Weld Overlay Design for Recirculation System Piping in the James A. Fitzpatrick Nuclear Power Plant Prepared by Structural Integrity Asscciates Prepared for New York State Power Authority l

l Prepared by: Date:

A. Y. Kuo "

Date: k Wf85 S .,. 5 Tan Reviewed and ,

Approved by:

,7 bh  !.//!!Af T Date: (/25 I /

Ef

_P '~ Ricc#d&M a EuS Date: d 8f 1annuzzi U

'E%

85 99 850509 6 i PD o OCK 050 3 MMME i

_ _ _ _ _ _ d

g ..

TABLE OF CONTENTS Page ii Lisi 0F TABLES ........................

iii LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . .

1.0 INTROCUCTION . ...................... 1-1 2.0 StM1ARY OF INSPECTION RESULTS .............. 2-1 3.0 FLAW EVALUATION ..................... 3-1 3.1 Methodology . ................... 3-1 3.1.1 Input Stresses . . . . . . . . . . . . . . . 3-1 3.1.2 Stress Intensity Factors . . . . . . . . . . 3-3 E 3.1.3 Crack Growth . . . .............

3.1.4 Allowable Flaw Size ............

3-4 3-5 1

3.2 Evaluation and Results .............. 3-6 3.2.1 Weld 12-4 ................. 3-6 E 3.2.2 Weld 28-112 ................ 3-8 3.2.3 Welds 12-17, 2S-48, 28-53,28-113 ..... 3-9 4.0 WELD O','ERLAY REPAIRS . . . . . . . . . . . . . . . . . . . 4-1 4.1 f'ethCdology . ................... 4-1 4.2 Weld Overlay Designs ............... 4-2 E 4.2.1 Weld 22-2-22 . . . . . . . . . . . . . . . .

4.3 Aplicat ion Recocnendations ............

4-2 4-3 4.4 AS-Buill Measurements . . . . . . . . . . . . . . . 4-4 4.5 Shrinkage Stresses ................ 4-4 5.0 CONCLUS'ONS ....................... 5-1 6.0 REFERE'.CES . . . . . . . . . . . . . . . . . . . . . . . . 6-1 E

E P

w , s-mmm sINTEQUTY l - ASSOCIATEEINC

a i LIST OF TABLES {

Page .

TABLE 2-1 James A. Fitzpatrick - Results of Spring, i 1985 IGSCC Inspections . . . . . . . . . . . . 2-2 TABLE 2-2 James A. Fitzpatrick - Results of Fall, 1984 IGSCC Inspections . . . . . . . . . . . . 2-3 l TABLE 2-3 James A. Fitzpatrick - Results of Fall,  !

g 1983 IGSCC Inspections . . . . . . . . . . . . 2-4 j E i TABLE 3-1 Summary of Piping Stresses .... .... .. . 3-11 Allowable End-of-Evaluation Period Flaw Depthl to E TABLE 3-2 Thickness Ratio for Circumferential Flaws - Normal i

Operating (Including Upset and Test) Conditions . 3-12 l TABLE 4-1 Weld Overlay Sizing for Weld 22-2-22 ... .. . 4-5 TABLE ;-2 Weld Cverlay As-Built Measurements - Weld 22-22 . 4-6  !

5 l E

E E

E l l

1 R l L,DETEGRTFY g

' / ASSOCIAIMINC

LIST OF FIGURES l

Pagg FIGURE 2-1 Isometric of Recirculation System - Lcap A Note: Abbreviated ISI Weld Numbers . . . . . . . . 2-5 -

FIGURE 2-2 Isometric of Recirculation System - Loop B Note: Abbreviated ISI Weld Numbers . . . . . . . . 2-6 FIGURE 3-1 Recirculation System Piping Model - Loop A (Reference 3) . . . . . . . . . . . . . . . . . . . 3-13 FIGURE 3-2 Recirculation System Piping Model - Loop B

( Re f e re nce 3) . . . . . . . . . . . . . . . . . . . 3-14 FIGURE 3-3(a) Assumed Through-Wall Welding Residual Stress Distribution in Small-Diameter Weldments

(< 12 in.) (Reference 6) . . .. . . . . .... . 3-15 FIGURE 3-3(b) Through-Wall Distribution of Axial Residual Stress in a Large-Diameter Weldment (Reference 4) . . . . 3-15 FIGURE 3-4 Through-Wall Residual Stress Profile for a Welded and IHSI Treated 26-Inch Schedule 80 Pipe at a Cross-Section 0.12 Inch (0.3 cm) from the Weld Centerline (Reference 5) . . . . . . . . . . . . . 3-16 FIGURE 3-5 Through-Wall Residual Stresses Computed at a Cross-Section in the Sensit ized Zone, 0.12 Inch (0.3 cm) from Weld Centerline, of a Welded and IHS! Treated 10-Inch Schedule 80 Pipe (Reference 5) . . . . . . 3-17 FIGURE 3-6 Magnification Factors of Circumferential Crack in a Cylinder (a/t = 0.1) . .x. . . . . . . . .. . . 3-18 FIGURE 3-7 Stress Corrosion Crack Growth Data for Sensitized Stainless Steel in BWR Environment (Reference 6) . 3-19 FIGURE 3-8 Common Assumptions Used to Estimate Circumferential Crack Growth ...................

3-20 FIGURE -9 Average Ef fective Circumferential Crack Growth Rate as Function of Operation Periods Used in Calculati of Time Between Inspections . . . . . . . . . . . on . 3-21 FIGURE 3-10 Stress Intensity Factor Versus Crack Growth Depth for J. A. Fitzpatrick Recircula Weld 12-4 . . . . . . . . . . . tion. System l . . . . . . . . . 3-22 iii g (N accwaum

FIGURE 3-11 Predicted Stress Corrosion Crack Growth for Observed Ultrasonic Flaw Indication . ....... 3-23 FIGURE 3-12 Comparison of Predicted Crack Growth With Allowable Flaw Size Limits - Weld 12-4 ...... 3-24 FIGURE 3-13 Stress Intensity Factor Versus Crack -Depth for J. A. Fitzpatrick Recirculation System -

Weld 28-112 . . . . . . . . . . . . . . . . . . . . 3-25 FIGURE 3-14 Predicted Stress Corrosion Crack Growth for Observed Ultrasonic Flaw Indication . ....... 3-26 FIGURE 3-15 Comparison of Predicted Crack Growth With Allowable Flaw Size Limits - Weld 28-112 ..... 3-27 FIGURE 4-1 Weld Overlay Design for Weld 22-2-22 ....... 4-7 E

E E

E E

E E

E E

E g

O(

f -

INTEGRFFY ASSOCIAIESINC

~

1.0 INTRODUCTION

During the spring,1985 outage at the James A. Fitzpatrick Nuclear Power Plant, ultrasonic examination (UT) was performed on seven recirculation system piping welds which were previously reported to contain IGSCC-like indications (one reported during the fall,1983 outage and six during fall, 19S4). These welds had previously been shown to be acceptable by fracture mechanics analysis and returned to service without repair (1,2), in addition, each of these joints was treated by Induction Heat ing Stress Improvement (IHSI), followed by a UT examination, during the f all, 1984 outage.

The current inspection results on the seven weld joints yielded some dif ferences from the previous inspections. This report evaluates those apparent differences, to determine whether the new inspection results significantly alter the conclusions of the prior evaluations. Where appropriate, the report documents revised analyses, demonstrat ing that design basis safety margins are still maintained, considerir.g worst case interpretation of the new UT results. Weld overlay design calculations are liso presented for the one weld in which the new inspection results mandate that such a repair be performed.

Section 2 of this report summarizes the inspection results, comparing the spring, 1935 results to prior inspection results and illustrating the differences and similarities between the inspection results. Sec t ion 3 presents the flaw evaluations for the welds which were not repaired, comparing the design margins to previous flaw evaluation results, and Section 4 describes the weld overlay design analysis for the weld which was repaired.

Section 5 discusses potential safety implications of the evaluations and repairs and conclusions.

1 i

N I

rcIsTaucTunat i INTEGRITY

,, / ASSOCIATESINC $

L -

. _ _ _ = _ _ . . _ _

2.0

SUMMARY

OF INSPECTION RESULTS Figures 2-1 and 2-2 contain isometric drawings (Loop A and Loop B, '

respectively) of the recirculation system at the J. A. Fitzpatrick Nuclear Power Plant. Note that an abbreviated weld numbering scheme is used in these figures, and in the balance of this report, in which the system identifi-cation has been deleted from the weld numbers since all of the welds under discussion are in the recirculation system.

During the current outage, a thorough in-service inspection was performed of stainless steel piping system welds previously reported to contain IGSCC-like indications. The inspections included seven weld joints which were evaluated as acceptable for continued service without repair (two riser welds, one end-cap weld, and four 28-inch pipe welds) (1,2). Indications were found in two 12 inch diameter riser welds, in two of four 28 inch diameter welds and in a 22 inch end cap weld. Table 2-1 provides a weld-by-weld sumary of these indications including length and depth from the current inspections. Table 2-2 provides a weld-by-weld sumary of the indications in these welds as sized during the previous fall, 1984 outage. Note that two weld joints, which were believed to contain IGSCC-like indications during the fall, 1984 inspection were found not to have reportable indications during the current inspections and two of the other welds contained indications which were below the recording level. Another weld, 22-22 (end cap), was originally identified as containing an IGSCC-like indication during the 1983 inspection, having length and depth as presented in Table 2-

3. However, an in-service inspection performed during the fall,1984 outage by another UT team, observing the same UT signal, identified the indications as geometry. During this spring,1985, outage, the same UT team has reversed its call, now identifying weld 22-22 as containing IGSCC-like indications.

Despite the apparent significant differences in IGSCC detection and sizing R when comparing the information in Tables 2-1, 2-2 and 2-3, the UT results do not necessarily imply that any IGSCC crack propagation has occurred in-service, subsequent to the fall, 1984 IHSI treatment. The differences are I

more likely due to the use of different UT techniques and interpretations.

Furthermore, with the exception of the end cap weld, the new inspection resultsdonotsignificantlyaltertheevaluationresults.l E 71 ASSOCIATESINC

E'~EEEEEEEEEEEEEEENME . .

TABLE 2-1 JAMES A. FITZPATRICK - RESULTS OF SPRING, 1985 IGSCC INSPECTIONS Weld No. Loop Weld Location Crack Length Thru Depth IHSI Discovery Previous Resolution Remarks Type Wall Method Inspection TT~4 A ~P Tpe to ~C 6.~6% No 7.T% Fis~~ ~tJT ~TO'/64' Re ~ Analysis sweep-o-let 9/84 12-17 A Pipe to C Not Recordable Yes UT 10/84 Prior Analysis safe end -

9/84 Still Applicable 28-48 A Pipe to N'one N/A N[A N/A Yes N/A 10/84 Prior Analysis safe end 3/84 Still Applicable 78-53 A Elbow to C Notlecordable Yes (JT 10/84 Prior Analysis valve 9/84 Still Applicable 28-112 8 Elbow to C 5.4% No 17% Yes UT 10/84 Re-Analysis valve 9/84 M3 8 Valve to pipe None N/A N/A N/A Yes N/A 10/84 Prior Analysis o 9/84 _

Still Applicable b 22-22 B End Cap C 76% No Max. Yes UT 10/84 Weld to Manifold 27% 3/84 Overlay ll P5 .

ENMEEEEEE E E E E EE R E K .E.

TABLE 2-2 JAMES A. FITZPATRICK - RESUI.TS Of Tall, 1984 IGSCC INSPECTIONS Weld No. Loop Weld Location Crack Length Thru Depth IHS1 Discovery Previous Corrective Remarks Type Wall Var. Method Inspection Action 7.~5f YFs~~UTFreT

~

17T4 A Pipe to C ~~~T.~0%" ~ tio 6~/83 ITISI &

sweep-o-let 9/84 Post IHSI Analysis 12-12 A Pipe to C lTiDY~~~Yes Avg. 50% Yes PT, Visual, Weld See Note safe end 1 Max. 100% 9/84 UT, Post IHSI Overlay 1 Min. 5%

12-17 A Pipe to C 4 M. No Max. lW Yes UT, Post IHSI 12/81 IHSI & Indica-safe end 3.0% 9/84 Analysis tions 90%

apart 12-23 A Pipe to C 1 011% No Avg. 40% Yes OT, Post IHSI 6/8'3 Weld See Note safe end int. Max. 75% 9/84 Overlay 1 Min. 5%

y T2T64 B Pipe to C l'Oli% Yes Ave. 30 F Yes YT-~ VTsu a l , Weld See Note L safe end int. 2 Max. 100% 9/84 UT, Post IHSI Overlay 1 Min. 5%

12-69 B Pipe to C 100% Yes Not avail Yes PT, Visual 12/81 Weld See Note safe end 2 9/84 UT, post IHSI Overlay 1 12-70 B (Thow to pipe C 12.6% No 4T% Yes UT, Pre-IIITI 12781 Weld 9/84 Overlay 78-48 A Pipe to C 1.1% No 15% Yes UT, Pre-lHSI IHSI &

safe end 3/84 sis 28-53 A ETbow to C .3% No 5% Yes UT, Pre-IllTf- 6783 Analy&

IHSI valve 9/84 Analysis 78 8 ETbow to C .6% No IT% Yes UT-~P re-IliSI 6/83 IIISI &

valve 9/84 Analysis h_-112 '13

/

B Valve to pipe C . 5f No 16% Yes 9/84 UT, Pre-IH51 6/83 IIISI &

Analysis i

1 - No pre-lHSI UT examination performed.

s .

l l

TABLE 2-3 JAMES A. TITZPATRICK - RESULTS Of FALL, 1983 IGSCC INSPECTIONS Weld No. Loop Weld Location Crack Length Thru Depth IHSI Discovery Previous Corrective Remarks Type Wall Method Inspection Action 22-22 B End Cap C 2~.7% No TOY No- di Analysis to Manifold m

b i

e h

F> >

i_-.-

i gg. i to?a f

2. \

it 16

) g.2 6

• kI , . . , , it.n nu c

.n s t, s it-sr N M 23*S i..as st gg.1N N i g.11 X it 18 g.66 ,,,,,

s.2,... 22 19 {

it. a it Sh\ a*

,,a A ,,,.

n. ,

it.60 p

., ,gg , , 2s. s,

,s. ,,

22-se 2s-ios I g.t A 1

,z... o 4 35

,a y

4.iot

..,s g 4-**

as..os 2**** t w ss **

naQ)2s-io' i> as-ito I l- Welds with UT indications

' 1. Spring, 1985

2. Fall, 1984

3. Fall, 1983 2s-tie l R - Weld overlay repairs ts-iss n 2S-ssoA

[zasist yas.in Las-ti 2,3

' FIGURE 2-1. Isometric of Recirculation System - Loop A Note: Abbre,iated ISI Weld Numbers STRUCTURAL DITEGRITY 2-5 - AssocipanilNC i

i

. s 81= $

04. p 58 12*s f

M 2s*SE 3

~

M*2C+$f (g.g g1 2- a u s<

i , s. 2- ,,.7 W

2.

fu .

,s...

h nf 3 1, 1. u. ,

,,g 11-13 l 0 1 -19 13 84 g 23 5o A 12 18 l

I 22*ll i

1%*24

• I g2*19 22 2s 0 24 El 12 15 II'"

I

!!'JS N 4ada 22*le 2s.jl 22-31

- Welds with UT indication

1. Spring, 1985

2. Fa11, 1984 s

3. Fall, 1983 I

] - Weld overlay repairs l rs-sa ,,,,

I 9 2s-s2

)

h0 rs s32,3 64s 2s s4 23.u 88 34 2s-s7 4 44 as s6 N 44s s-u 4 42 4.se 26 17 ****

• • H 4 4a

{

FIGURE 2-2. Isometric of Recirculation System - Loop 8 Note: Abbreviated 151 Weld Numbers 1 mmorry b 2-6

,.r-' , , , - ~ ___ --

ii M' .,

s.

l 3.0 FLAW EVALUATION Two of the welds inspected in the spring, 1985 outage contained UT ,

indications which apparently increased in length,12-4, and 28-112. Re-evaluations of these welds, in accordance with ASME Section XI, IWB-3600 (4),

supplemented by the recomendations of NRC Generic Letter 84-11 are documented below. Where prior evaluation results (1) are still applicable (four of the welds), since the indications remained the same or decreased in size, they are not repeated here.

3.1 Methodology 3.1.1 Input Stresses Piping stresses in the J. A. Fitzpatrick recirculation system were obtained from the General Electric Co. stress report (3). Figures 3-1 and 3-2 illustrate the piping model used by GE in that report. Stresses taken from (3) at the two weld joints under re-evaluation are summarized in Table 3-1.

It is noteworthy that, although not required by the ASME Section XI, thermal stresses have been included in the calculations to account for possible effects of low toughness weld materials. Referring to Table 3-1, the stresses due to (pressure + deadload + thermal + expansion) were used for crack growth evaluations and the stresses due to (pressure + deadload +

thermal + expansion + seismic) were used for critical flaw size evaluations.

A wide body of experimental and analytical data exist on the residual stesses in austenitic pipe welds. A sumary of such data for axial residual stresses in large diameter pipes (20 inch nominal diameter and greater) taken from (6) is given in Figure 3-3(b). Note that there is considerable variability, but that the majority of the data show a U-shape trend varying from tension on the inside surf ace to compression in the central region of the pipe, and back to tension again near the outside surf ace. Also shown in this figure is an

" evaluation" curve recommended in (6) for use in IGSCC crack growth evaluations of large pipe. This curve is seen to provide a reasonably conservative representation of the available data for large diameter l STRUCTURAL INTEGRITY 3-1

r t

1 weldments. Thus, this " evaluation" curve was used in the present IGSCC crack l growth analyses for weld 28-112. This curve can be expressed as follows:

&=

I 30 - 242.05(x/t) + 394.20(x/t)2 - 174.30(x/t)3 (1) where e is axial stress in the units of Ksi, and x and t, are the distance from the inside surface and the wall thickness, respectively.

Also shown in Figure 3-3(a) is a linear through-wall residual distribution which is widely accepted as a reasonable representation of the residual stresses in moderate to small diameter piping (12 inch nominal diameter and less). It should be noted that residual stress distributions in smaller diameter piping show more sensitivity to specific weld parameters, and as a result, much more variability from weld to weld as well as azimuthal variability in a given weld can be expected. The small pipe residual stress pattern of Figure 3-3(a) generally results in more rapid crack propagation than the large pipe distribution discussed previously.

The linear curve of Figure 3-3(a) can be expressed as v = 30 - 60(x/t) (2) where e, x, and t are defined in the same way as in Equation (l). This linear curve used in the re-evaluation of weld 12-4.

A large body of laboratory data and analytical solutions also exist on post-IHSI residual stresses in austenitic pipe welds. These data are summarized in (5). Typical post-lHSI axial stress distributions are illustrated in Figures 3-4 and 3-5 for large and small diameter pipes, respectively. These stress distributions were also curvefit by third order polynomials for use in the analysis, and the resulting equations are given in Figures 3-4 and 3-5.

These post-IHS! residual stress distributions are also used in the re-evaluation.

I

^

. STRUCTUIULL N

3-2 ASSOCIATESINC

L 3.1.2 Stress Intensity Factors I

Pipe dimensions used in this analysis are as follows (3): '

28-inch pipe 12-inch pipe Outside Diameter (in.) 28.363 12.662 Inside Diameter (in.) 25.867 11.442 dipe_Wal_1 Thickness (in.)

_ 1.248 0.61 An analytical model of a 3600 circumferential crack in a cylinder of radius to thickness ratio of 10:1 (7) was used for the fracture mechanics evaluation. For the pre-IHSI case, applied loading is the superposition of piping stresses tabulated in Table 3-1, and the as-welded residual stress defined by Equation (1) for 28-inch pipe or Equation (2) for 12-inch pipe.

For the post-!HS! case, applied loading is the sum of the same piping ,

stresses frem Table 3-1 and the post-IHS! residual stress distributions given in Figure 3-4 for 28-inch pipe or Figure 3-5 for 12-inch pipe.

For purposes of the fracture mechanics analysis, the axial stress dis-tributions of piping stress, pre-IHSI residual stress, and post-IHSI residual stress have all been expressed in terms of third degree polynominals of the form:

e = Ao + Ai x + A2x2 + A 3x3 (3) where cr and x are defined the same as in Equation (1) and AO-A3 are the coefficients resulting from the curvefit.

The stress intensity factor for a circumferential crack in a cylinder of i radius to thickness ratio of 10:1 can be expressed as follows (7):

1 Kg = [nT (A f0l+n AF a Af 4 a3 34 12+T 2 3 + Tn Af) (4) l where F 1 , F 2 . f ,3 and F4 are magnification factors and a is crack depth as shown in Figure 3-6.

STRUCTURAL INTEGRITY

/ MEOCIATESINC 3-3

1 3.1.3 Crack Growth A large body of laboratory data also exist on stress corrosion crack growth -

rates for sensitized stainless steels in simulated BWR environments, these data are surraarized in Figure 3-7, also taken from (6). These data were obtained using fracture mechanics type specimens with different crack sizes and loadings, which canbe characterized by the crack tip stress intensity factor K. The data represent a wide variation in material sensitization, as well as levels of dissolved oxygen in the water. While subject to some criticism because the sirrulated water chemistry in these tests did not contain levels of impurities (chlorides, sulfates, etc.) that could exist in operating BWRs, the widely used power law "best estimate" curve of Figure 3-7, is beliesed to provide a representative crack propagation rate for plant crack growth assessments. The "best estimate" curve can be described by a power law representation of the form:

da/dt = 2.27 x 10-8(K)2.26 (5) where a is the crack depth in units of inches, t is time in units of hours, and K is the stress intensity f3ctor iq units of Ksi % .

Crack growth analyses typically make use of one of the two assumptions illustrated in Figure 3-8 regarding crack length extension, self-similar crack growth or constant aspect ratio crack growth, the former assumes that the incremental crack extension is the same at all points on the crack front, while the latter assumes that the ratio of depth to length remains constant during crack extension. Considering field and laboratory experience with circumferential crack extension, it appears that the self-similar assumption may underpredict crack length versus time, while the constant aspect ratio E assumption overpredicts.

Recent work by Gerber (10), under contract to EPRI, provides a new approach for addressing circumferential crack extension which is more technically defensible than the above self-similar or constant aspect ratio approaches.

This approach utilizes data generated in a laboratory stress corrosion test NM g 34 n ====

- mcmuum ,

I of a 26-inch diameter welded pipe specimen at Battelle Pacific Northwest ,

I Laboratories (11). IGSCC was induced in this pipe through loading to a high I applied stress in a simulated BWR environment, which was accelerated by toe use of graphite wool to create an artificial crevice. Crack growth occurrred and was monitored both during operation and at several scheduled shutdown intervals for the test. A number of small cracks initiated early in the test, the length of which was periodically measured and the initiation of new cracks was noted and their lengths susbsequently tracked as well. At the completion of the test, there were a total of 63 cracks with a combined length of 32.57 inches.

The average ef fective circumferential crack extension observed in this test g is presented in Figure 3-9. This rate includes both growth of existing E cracks as well as new defects initiating and contributing to the effective crack growth rate in each inspection interval. Examination of Figure 3-9 suggests that an average effective circumferential crack growth rate of 0.5 mils / hour should give a reasonably conservative estimate. Thus, 0.5 mils / hours was used as the crack length growth rate in this report. It should be pointed out, however, that although this is an average effective rate, it is based on a laboratory test in which the local environment, load and cycles were all intentionally modified to accelerate IGSCC relative to actual plant conditions. Test and analytical data (20) have also shown that the IHS! will suppress not only crack initiation but also crack propagation for small cracks in both the length and depth directions.

3.1.4 Allowable Flaw Size Based on detailed calculations presented in (8), allowable flaw sizes for various levels of primary applied loading (Pm + Pb ) have been specified in E ASME Section XI, IWB-3640 (4). A tabulation of allowable flaw sizes as a function of applied load is given in Table 3-2, which is taken directly from Section XI, IWB-3640. Note that this table permits very large defects in some cases (as great as 75% of pipe wall) and does not include consideration of any stress other than primary, notably secondary and peak stresses from the design stress report as well as any weld residual stresses or mis-alignment / fit-up stresses which might exist from construction.

The ag y(nm: carry

) 3-5

for this exclusion is that, given the extremely high ductility of austenitic stainless steel, these strain controlled effects will self relieve after a small amount of plastic deformation and/or stable crack extension, and will have little or no impact on the loads and flaw sizes needed to cause unstable crack propagation or pipe rupture.

However, some recent fracture toughness data may invalidate the above argument, at least for some classes of austenitic weld metal (9). To account for possibility of low ductility weld metal, secondary stress from stress report were also included in this report although it is not required by the ASME code.

It is important to note that the very low measured toughness occured only in a small percentage of the materials addressed in (9), and may be of only limited concern from a probabilistic viewpoint. Indeed, most IGSCC observed to date has been restricted to weld heat affected zones, which should exhibit the high toughness attributed to base material. Also, the low toughness data to date has been limited to flux types of weldments (submerged arc or shielded metal arc), which are not used in current construction practice nor in weld overlay repairs of pipe cracks. fievertheless, to address these possible concerns, the analysis procedure used throughout this report includes thermal expansion and weld overlay shrinkage effects as a primary stress condition in determining allowable flaw size from Table 3-2.

3.2 Evaluations and Results 3.2.1 Weld 12-4 Input to the flaw evaluation for this weld was as follows:

Indication Length - 2.5 inches (versus 0.4 inches in prior evalur. tion)

Indication Depth - 0.046 inch Pipe 0.D. - 12.662 inches i

Pipe I.D. - 11.442 inches Pipe Wall Thickness - 0.61 inch

, _NTEGRITY I

5 3-6 / ASSOCIATMINC l

s Applied Stresses (Table 3-1)

Pressure + DW + Thermal - 14030 PSI Pressure + DW + Thermal + Seismic - 17245 PSI Residual Stresses Pre-IHSI - Figure 3-3(a)

Post-!HSI - Figure 3-5 Figure 3-10 provides applied stress intensity factor versus crack depth i data for the three load cases used in the evaluation. Assuming the indication to be IGSCC, these stress intensity curves were used to perform F !GSCC crack growth estimates for both as-welded and post-IHSI residual stress conditions. The resulting crack growth prediction is illustrated in I Figure 3-11 for the as-welded case. The post-IHSI case is also shown, and results in no predicted crack growth for the balance of plant life.

The allowable end-of-cycle flaw size was determined in acccrdance with ASME Section XI, Article IWB-3640, and is illustrated in Figure 3-12 in terms of allowable flaw depth versus length. Note that although not required by IWB-3640, thermal expansion stresses have been included in the evaluation to account for possible effects of low toughness weldment material. Also, in g accordance with the recommendations of NRC Generic letter E4-11, a maximum sus allowable flaw size of 2/3 of the IWB-3640 limit (shown as a dashed line in Figure 3-12) is used to allow for uncertainty in flaw depth sizing.

L Referring to Figure 3-12, it is seen that the 2/3 of IWB-3640 limit is exceeded in approximately nine months in the as-welded case, but is p

satisfied indefinitely in the post-IHS! case, since no crack propagation is predicted. To add further assurance in the post-IHSI case, the IGSCC crack

! growth analysis has been repeated assuming various initial flaw sizes ranging upward from the observed UT depth. No crack propagation is predicted in the post-IHSI condition for initial crack depths up to 0.488 inch, or 50% of the pipe wall. It is also noteworthy that, given the relatively short length of the observed indication (6.24% of circum-ference), it would not lead to rupture of the pipe joint even if the above c crack growth or initial flaw size estimates are.rry significantly~

b 3-7

/ ASSOCIATESINC

1-Leak before break is clearly the expected, hypothetical f ailure mode for this indication.

f On the basis of the above evaluation, it is concluded that the new flaw sizing information does not significantly effect the prior evaluation results for this weld. Continued operation of the plant with this weld, considering the observed indication and the IHSI treatment which has been applied, will not lead to a reduction in plant safety margins, or a plant operational concern.

3.2.2 Weld 28-112 Input to the flaw evaluation for this weld was as follows:

Indication length - 4.72 inch (versus 0.5 inches in prior evaluation)

Indica' ion Depth - 0.221 inch l

Pipe 0.D. - 28.363 inches Pipe I.D. - 25.867 inches Pipe Wall Thickness - 1.248 inches Applied Stresses (Table 3-1)

Pressure + DW + Thermal - 6899 PSI Pressure + DW + Thermal + Seismic - 7935 PS!

E Residual Stresses Pre-lHSI - Figure 3-3(b)

Post-lHS! - Figure 3-4 Figure 3-13 provides applied stress intensity factor versus crack depth data for the three load cases used in the evaluation. Assuming the indication to be IGSCC, these stress intensity curves were used to perform IGSCC crack growth estimates for both as-welded and post-lHSI residual stress conditions. The resulting crack growth prediction is illustrated in Figure 3-14 for the as-welded case. The post-IHSI case is also shown, and results in no predicted crack growth for the balance of plant life.

- STRUCTURAI.

INTEGRITY g g

/ ASSOCIRIEINC

W The applied stresses (pressure + DW + thermal + seismic) were also used, in l conjunction with ASME Section XI, Article IWB-3640, to determine allowable indication size.

~

This result is illustrated in Figure 3-15. Also, in accordance with the recommendations of NRC Generic Letter 84-11, a maximum allowable crack size of 2/3 of the IWB-3640 limit is used to allow for uncertainty in crack depth sizing. Referring to Figure 3-15, it is seen that this limit is not predicted to be exceeded for greater than 36 months in the subject weld, even considering as-welded residual stress conditions.

If IHSI residual stress benefits are accounted for, the flaw will remain at its present size indefinitely, and thus satisfy the allowable flaw size limit by a large margin for the balance of plant life.

On the basis of the above evaluation, it is concluded that the new flaw izing information does not significantly effect the prior evaluation results for this weld. Continued operation of the plant with this weld, considering the observed indication, will not lead to a reduction in plant safety margins, or a plant operational concern. The application of IHSI to the joint further reinforces this conclusion.

l 2.2.3 Welds 12-17, 28-48, 28-53,28-113 l The current ultrasonic examination data on these four welds, as detailed in Table 2-1, indicate either no reportable indications (28-48 and 28-113) or indications which are below the recording level (12-17 and 28-53). It is therefore assumed, for purposes of evaluation, that the ultrasonic flaw sizes have not changed from the Fall, 1984 examination, and that the previous evaluation results (1) are still directly applicable.

An additional point of f act should be made regarding weld 12-17. The previous flaw evaluation for this weld (1) demonstrated that if the as-welded residual stresses were used in the crack growth analysis, the allowable crack depth limit of 2/3 of IWB-3640 would be exceeded following approximately 5000 additional hours of operation. However, use of the post-IHSI residual stress ditt"hution results in a prediction of no additional crack propagation in this joint (1). Besed upon the current UT results (Table 2-1), compared to the Fall,1984, UT results (Table 2-2), the lack of 7[ STRUCTURAL INTEGRITY

' ASSOCIA'IEilNC 3-9

any measurable crack growth confirms the use of the post-Ills! residual stress distribution in the flaw analysis of weld 12-17. Using this ,

l

• distribution of residual stress, no additional crack propagation is pre-dicted in this joint. It is also noteworthy that, given the relatively short length of the indication (7% of circumference from the Fall, 1984, inspection), it would not lead to rupture of the pipe joint even if the above I crack growth or initial flaw size estimates are significantly in error.

Leak before break is clearly the expected, hypothetical failure mode for this indication.

E E

E E

N

'? STRUCTURAL lINTEGRITY

/ ASSOCIAl'ESINC 3-10

1; TABLE 3-1

SUMMARY

OF PIPING STRESSES l WELD N0 J0 INT N0* STRESS, KS!** STRESS, KSI***

l 12-4 258 14.03 17.245 28-112 175 6.899 7.935

}

Nodal numbers in Ref. (3)

• • Pressure + Deadicad + Thermal + Expansion

• • • Pressure + Deadioad + Thermal + Expansion + Seismic 1

1 1

1 1

1 i

n i

E I

N; l

E:

1 I

k i

E! .

l l

l

.~_

l N

' ASSOCIAl'ESINC

3-11

TABLE 3-2 ALLOWABLE END-0F-EVALUATION PERIOD FLAW OEPTH2 TO THICKNESS RATIO FOR CIRCUMFERENTIAL FLAWS - NORMAL OPERATING (INCLUDING UPSET AND TEST) CONDITIONS F, + P, Rate of Fla Length. /,. to Poe Circumferme (Note 01) 0.5 5 or More (Note (21] 00 01 0.2 03 04 (4) (46 (41 15 .4) (41 (4) 0 15 (di (41 14 0 75 0 40 0 21 0 39 0 27 0 22 0 19

1. 3 0 75 0 75 0 56 0 40 0 32 0 27 12 0 75 0 75 0.75 0 73 0 51 0 42 0 34 11 C 75 0 75 0 63 0 $1 0 41 10 0 'S 0 75 0 75 07) 0 59 C 47 09 0 75 0 75 0 75 0 75 0 75 0 68 0 53 08 C 75 f 0 75 0 75 0 75 0 58 07 0 ?S 0.75 0 75 0 75 0.75 0 'S 0 63 s00 0 75 l h0'E5 3 (11 Fla. eepte - a f a a s., dace e.

a 2J f 2r a s th'dase 8'an i = com*ea' th thmest Linear ete grat.c.n is pe e u t'e g

j (2) P, = gestaary re-traae gt ess P, = pr mary teWeg st ess

~

S. = alo*at e :es.99 st ess

• teas tr ' a atto Sance a th Sect.on Ill)

(3) Cert -8e eme tated v a:m et c ce 3 avete*

tel( AB 3514 ) sea:t tg se3 I ~?[rraucrmute

' ASSOCIAIEilNC INTEGRrrY 3-12

il1l11

!; i y

~

a t ,

2i b v d

3 8

s

/

(A i ,e4 r?*

e v*

i N'cn*. i.

w e

g

,)

s.

" , .. ,n. f, ;t e.o e

y g(\ ra. t i

g e.

A 1

(

,o(  ; n p N w%<o,g .

)

L o

o

- n.. -

l e

, d

.

• o

..

• M

\ 3 g q

n

- .s ,. . . ., i i.

.,~

> .e

.a

.y',.

N.,

i P

p N - .

. 8

.4 . , ,,

b,f',' //s,x w ys Q 2

3 . , m e  ;,

, . .. _ .~

.\

, t e

N  ;> .

s xx

• • y

'N v s a

S

~

dy//,/

.j , - ) ,34.

.'  % 6 : g - .

I' n) e ,* '.I 4 " *g io3 I

i s

O h> t e

,' l

- - ac

y. ", g 1

l n x'ys, f, "' d / r. ,

, - ' ' . ue

', cr

' re e i f

. 1k< ,.'~*/

ce

/ t s eR y .

• f,t,jc.h~M, t R(

. , '\

- ' . IJ. J1

. , ., ~ .

"s ,f

, 3 * .

1

, }g ~- 3 3 7, .

.ee d- E

p. - . R U

s

. . 8 e

m b..',

.~

. G I

' . F

.N

. 'L\ j 'c. '

/,.

.y 5

1

)lj O

rl

> i La 1i

MM M M M M I.' ).. ,;[ / ,2- u r.i eo.

<Ts\

)h../

q&

h.

. . x >=

- "* A ~' ~ =

f, un *< on

, ,dQy/

n,-

4

.y~ , h*

f-

.: , , . .y.

- e.s t an r,.. f. ~ ~ - -.., ..>. /g' -. ~ n

' = - /. .

r = = '-

• 1- :ll il

, n ra w .a "

dpi

, ,;,y G... N . ,/ {

4 n,/ n,

. ,{ ,.s

., u W" Flee'

. ,~

-n*

k ~' 'a,. . ,

• ** ' k., s}. ...

,O **"

• t'#*N.*s** vu,su .

=r=::

,z c,

.6-

~? L."s

1. p

'd '

x /C,,M po', ,[6 y s s.:

.i 'r J M..f(8h,,:.a

~

N o rs ,. u, .

sad

) ,,

r.1 w u, t.;)

..,... , ,g,g'- v.s sne. xw n ,

N /

f,.,...,....

• 4

,- ,n .o e- Nno>= .3 f, e,$ se, 0 o g

Ntttr ya en g yipig nigitiAct toasomATE L W.6Ja M2Ilu

.. v . I

d i _%,

ILE7, M- +ss -

l ,[a L.J a /

> .. ..f rJ

\

1,.'.:2. !

ra

. ,. . i.s,-s.e.o a rs- ci e

5 Recirculation System Piping Model - Loop B (Reference 3 )

FIGURE 3-2.

h

--- _ _ _ _ _ _ _ - _ _ - - _ _ _ _ - _ -~ _

! E i i i s i g g g g h oo - -

l M -

l m- - -

l ) 10 - -

o __ _--- -------

-20 - -

i

-30 -

, I , I , I , I i 0 a2 04 04 08 14 ett FIGURE 3-3(a). Assumed Through Wall Welding Residual Stress Distribution in Small-Diameter Weldments

(<l2 in.). (Reference 6) 1 l

1; u

E l

4C e .

R 20 -

li' , .

5 o M

$ M_

'/ .'"

,. y )

!: 3

~

~20 Evaluation -

l. g, Curve li

-40 -

u Q I

. e At 02 04 05 08 at DUTM/ThiCANt35 FIGURE 3-3(b). Through-Wall Distribution of l Axial Residual Stress in a large-Diameter Weldment.

(Reference 6) . STRUCTURAL i

<] I, INTEGRITY Assocmes mc 3-15

l \

I ,"

_ s )

b

[\dm Vt 4

0,

- - 3 En0 4 . WW E E 0

0 s L L t { DD -

0 + 2 f

l U

A -

L2 f' / I N

0 0

G ,

U R

E X

I A

A 0

Esyb@$r/ hw ,Egw Y1 S

I 0

8 0

3 4

L S

T 0 -

O S

- / ,

o_

ooP M

_ E S

,S R

N . 3

( TT 0rh er k 2 ,

b/ s o

o 3ao o0 l

t i cej v 2

)

mdh

' l'1 x 0 0

7W 4 -

f6a 0 r - l  %

of l X

mocR _'

)

- }

t he R h s eSi 1*I  !

cd 3 3 Whu ees 6 0e l l di 0w 43 -

b(3 O du 3 3 c 3o \_

_ l 45 t

lS w i Cet 0m 3 / 0.

a c 6

T E

e r -

2 c R S

n 8e 4 d)m S t 0s R E 0 '

j'[ m

) w s' U n e s S

- F r P F

[

8 i -

l iP D A 2 U C ip r A - n- E 0

/

neo 0 L2 e f 0

ai C aukOEgEM

( t l e

I R k$ sU. gg 0 8

/

R C 0 ea U 0 ,

f f M 5 M

. Co rr F

E 0

o. oPo 5o R . 3

) sa E s N E T

-W I A

s O

Se O el L2 .

cd S 0 te T id R I

/g k 2 o E N 0 N 0 na S E [

0d n .S 4 . R i

k 0 a S U

/) 3 0

1I P 0 2H F S A iI C 4 n E 0 c 0 h

l l

M Pa M Po

-300 -200 - soo r. ,( n ) /N k m) - 430 2no . sao o soo 200 300 06 .,

OUTER SIJRFACE os_

)

WELDING \ 05 2cm) b '*d 5

N.

/

- - wuoiwo os t lHSI E is-

}

2 U _

d [

f, o I \/ I '6 U*~

l T 3 -.s n : ,*

[,,

so/ - o 39 . h 0 *--so u (i Sacml IO 50cm) E f

/

) '

E - R .S 3 75.a 03 6Scm) p#~ [

'7 03- 7 w g 3,

~

N 6z 0:

a.

z h

X O E'-os '

k 02"-OS i

}

u M

5 U 08" $ 09-5  ; 5 INNER StMACE

-40 to O /0 40 -40 -to O 20 40 RESIDUAL AXI AL ST HESS, has RESsDUAL CIRCUMF ERDtTIAL STRESS, ksi IbQ

?l i-;

s--

b

~~

Thr ough-Wal l Res idua l ',tresse'. Computed at a Cross-Section in the FIGURE 3-5.

b ' ens it ized lone, 0.l? Inch (0.3 cm) from Weld Centerline, of a l Welded and IH5! Treated 10-Inch Schedule 80 Pipe (Ref, 5 ).

l l

E -

E I i i i a=Ag

• A,s . g e 2, ,3 l F1 E l B

~

g 3 l

N / p'.,

,% 2. - -

E " ar+- o F3 i +-t ,

! c 1. --

x , -f e . *,g.a,rpia,.rthig.7,8 3 ig o

0 .2 4 .e .6 F ra ctier.a 1 Dis ta nc e T r cug'.. V.' a :1 (al t

F;GURE 3-6. Magreification factors of Circumt:1:ential Crack in a Cylinder (a/t = 0.1)

.-C',

I STRUCTURAL INTEGIUTY ASSOCIRFESINC 3-18

l l

5 -

l 1

to'3 ,

~

Upper Bound (Furnace Sensitiza d) )

y da/dt = 5.65x10-9(K)3.07

\ /

0 10 -

g @ Best Estimate (Weld Sensitizet )

~

,A da/dt : 2.27 x 10-6(K )2.26 y : *u'irse . ers.-

f ., ia s , o a ".:#s . , , , . .

/ A St%1'itt: a' "b!'8 2

• C ; me

• N / / O sat a c164:

• I . t 1, e

• g o,...,'.....<....

io-3 g

_ /9 - o, . C . ,1. . -

N - 'y J 0 .";'. 'e','!!n1, 'Ju'. ' - '

_ v ou....,...,,,<.....

T T I

@ Cl . T'is '

@ wa%C Cues t . Cl%t

$ 506Cu2= . f stec

@ was a c a . . m.* a C a og .

10,4 ,

ge es;. 6a g )

@ eams.as cg go, gg O M*S'tef f 0 .' of LO.hG E96 af

" N2*f. Pe n O ssa C ogeg ess 39, l 3

l

' ' ' ' ' 8 10~7 U 30 M 30 40 6o 30 7o FIGURE 3-7. Stress Corrosion Crack Growth Data for Sensitized Stainless Steel in BWR Environment (Ref. 6)

[ :'LINTEGRITYSTRUCTURAL

' ASSOCIAl'ESINC

i

~

h

[ ~~

Y G

~

i WI N p f. -

i Y

I al self-similar Assun tion; l'- f = 2(a'-a) g\

l E

1 1

E -

~

~ ,.

l / f '

% a L- \

l f -- '

I fI {

f.

i b) Censtant Aspect Ratio Assugtion; f/a = l'/a' 1

u l

E FIGURE 3-8. n ssumptions Used to Estimate Circumferential i

1

'~

l  %

7.nmconrry 6

g 1-/o ASSOCIAHEINC

, ym. .

l .0 -

.9 '-

.8 _

1 ,.

5 5

5 .6 -

t-25 ups .5 -

O a4 -

o 5

ga .3 ~

1 I .2 -

O

.1 -

i I l i I i 1 I / 1 4 5 6 1 Mund 2cr of Operatson Periml. Ira Imfed in the

[ ', Average Crack Growth Hate t .ile islat son

&J l l I I I I I 7,000 4,000 6,000 11,000 10,000 12,000 14,000 Approximate Time Between inspection (Hours)

]

r=

M FIGURE 3-9. Average Effective Circumferential Crack Growth Rate As a Function 2 of Operation Periods Used in Calculation of Time Between Inspections O .

-=.

t _ _. . . _ -

1: PRESSURE +DW+TH 2:AS-WELDED K 3:IHSI c-:-

.i 4- -

1 L.i _

_2 -

' ~ -

1 /_.- _-

l :.

r, .-:,

- F_i  ; -- :  :  :  :  :  :  :  :  :

s _

I -1 _Ns.

. .ss J -2 -

N -3 -

s

.f -

,4 -4 -

_ , 3

,,__J1 _c: _

_~~

!_1 J . - .

i . I 1 i

, I , ,

c- -t, a* f' O 1 2 3 4 5 pe m i_:RAF_:K. DEF'TH

• 1H_ ':. I NF_:H ':'.

ai

+

MODEL C M Stn ss Intensi;y Fattor Versus Crack Growth Depth for FIGURE 3-10.

b J. A. F itzpatrick Recirculation System Weld 12-4

. t St.re55 Corr 05iOTi Cr.9.C k GP O Wt.h R.

6-M .

C

y. .

.s-l e

D i E 4 -

/

P -

/

PRE-IHSI T /

H 3 -

,/

/

. /

k$ .

2 -

/

1 I-1 1 -

I -

POST-IHSI b7)I g i i i . i u4 .

2 4

h H I.,O TIME / 1000 (HOUR)

Ei PASNY-01., WELD 12-4 9

FIGURE 3-11. Predicted Stress Corrosion Crack Growth for Observed Ultrasonic .

Flaw Indication

l ., l

.. l 3

l jl .

1 I

1.0 l

, l\ Pm + Pb + Thermal Sm

=

1.02

~8 ~

IWB-3640 Limit I =

y .6 -

5 E 2/3 of IWB-3640 Limit 1lI a

w 3 Mor s (ic -Welded? s

. S 's i

t **

_ O 's

-i 5

s.  %,

1 e s s -2 .e '~

c S '

s s

w  ;

s.7 v

.g

~

.? .' 2

?  ?'

i s 5 &

In1 di Flaw Size & Post-IHSI Case (No Growth) 0 '

O .1 .2 .3 .4 .5 i

FLAW LENGTH / PIPE CIRCUMFERENCE (f./nDm) k l' FIGURE 3-12. Coraparison of Predicted crack Growth with Allowable Flaw Size Limits - Weld 12 4

\

. : -m y m i

lINTEGRITY F '

' ASSOCIAIESINC J 3 94

-m - ^

- , - - , d

_we -

1: PRESSURE +DW+TH p'- 2:AS-WELDED 4 -3: IHSI .

l O. -

' 1 1

_ m I-1 J- ,< .

s s - -

p:. E1 . . . -,_ . . . . ,

O .

}

~~ _ _ ~ _

m I . x. -

= -

s -

N. ,

I 5

.s

_2_

N -

3 -

i

, ,.3 4

4 -

~

~

>:1 -

c- -

294

, i , i i,i , i s- -

~ 'i 0 2 4 6 S 10 r:

il g MODEL - -

De 2 FIGURE 3-13. Stress intensit y f ac tor Versus Crack Deptfi for J. A.

O Fitzpatrick Recirculation System - Weld 28-112

- ~ -

l,- '_;t.t-@_ =. .=

_ l_: n_ t t n_ .=_ i n_ ri l_:t .:n. _ le.

_ l it n_ t.J t. h R c Ui

.>l l

1 1:

V,. -

45 _

/

D -

r-E 4g _

P -

s

' PRE-lHSI T -

y H 3"- -

u .-

1 30 -

O -

/

b -ec-

_ ,/

. i

' POST-IH5I

( -

dVr i 'U t , i , i ujN F1 1 -

-) T IME ' HOUR::'

f g PASN'r'-01, 1F1F1F1F1 WELD 28-112 D!

y flGURE 3-14. Predicted Stress forrosion Crack Growth for Observed (5 Ultrasonic flaw Inditation .

I ,

1 1

1.0

\ Pm + Pb + Thermal Sm

= 0.47 0.8 -

IWB-3600 Limit

- 2 I i W 0.6 -

E

~ 2/3 of IWB-3600 Limit r Y

=

c.

2 0.4 -

E u e 3 v 3' q# 36 Months

-3 h ,g-3W IS Months 0.< -

e initial Flaw Size & Post-!HS! Case (No Growth) ir g i i i i 0 0.1 0.2 0.3 0.4 0.5 s

FLAW LEtiGTH/ PIPE CIRCUMFEREtiCE (f.hrDm)

FIGURE 3- 15. Comparison of Predicted Crack Growth with

. Allcwable Flaw Size Limits - Weld 28-112 9 7 STRUCTURAL fLINTEGMTY ASSOCIKITEINC.

4 3-27

g. -

l .

4.0 WELD OVERLAY REPAIRS As a result of observing relatively 10ng and deep UT indications during the current inspection, weld 22-22 was repaired by depositing weld overlay -

g material 360 degrees around and to either side of the existing weld. Weld E overlay repairs of this type have been used in U.S. BWRs to increase pressure boundary thickness, restore the original design basis structural margins of the pipina, a1/ produce a favorable compressive residual stress pattern in cases where IGSCC-like indications have been found. IGSCC -

resistant, Type 308L weld metal with controlled ferrite is used for all such overlay repairs, to preclude the continued propagation of the IGSCC into the

\ repair itself. As indicated in figures 2-1 and 2-2, previous weld overlay repairs were performed on welds 12-12, 12-23, 12-64, 12-69, and 12-70.

} E Design calculations for these repairs are given in (1), and are not repeated here. Weld 22-22 thus represents the sixth weld overlay repair performed at

~.

the J. A. Fitzpatrick plant.

I 1 4.1 Methodology 1

g , As discussed at length in (6) and (12), there are three types of weld overlay design philosophies employed in operating BWRs:

Type I - In which the original pipe wall is assumed to be cracked (both s through-wall, and 3600 around the circumference), and the G

overlay is called upon to supply the total load-carrying capacity of the joint; l i

l Type II - In which the pipe wall is assumed to be through-wall over the entire observed length of the indication, with load carrying l capacity supplied by both the weld overlay and the remaining, uncracked portion of the pipe circumference; and l

Type III- In which both the depth and length of the crack are assumed to be known, and the entire uncracked portion of the pipe

~

wall is assumed to be participated in the load carrying capacity of the joint.

i  ?(STRUCTURAL INTEGRITY ASSOCIATESINC 4-1 l

l !_ .

I '.

](* Type I and/or Type II weld overlay designs have been adopted for the previous weld overlay repairs for the ultrasonic indications in the Fitzpatrick plant, and the same design philosophy is used here. '

l-Weld overlay thickness is determined iteratively using the applied stresses l and the allowable flaw size from ASME Section XI, IWB-3640 (Table 3-2).

Minimum weld overlay length is chosen to be 1.5Viit where R and t are outside radius and wall thickness of the pipe, respectively.

j1 J i

! A great deal of measurements and analyses of the residual stress patterns 1

produced by weld overlays have shown (12) that the weld overlay produces a

j;. highly favorable compressive stress field on the pipe ID, and through a substantial portion of the inner pipe wall. Such a compressive field on the pipe ID would arrest shallow cracks due to IGSCC. Besides this favorable

$ residual stress distribution, the weld overlay material itself, Type 308L stainless steel weld metal, is a low carbon and high ferrite material that hi. is highly resistant to IGSCC in BWR environments. Thus, with the high

^{ quality welding techniques employed (GTAW), and proper application pro-

J '

cedures (Section 4.3), the repaired weld joint is expected to be immune to any further degradation due to IGSCC.

i 4.2 Weld Overlay Designs a

$ 4.2.1 Weld 22-22 g

4 Indication Length / Pipe Circumference = 76%

Indication Depth / Wall Thickness = 27%

?

5 Applied Stresses j Pressure = 6994 psi

. Overlay Design (Figure 4-1)

I Thickness = 0.323 inch (see Table 4-1)

Length = 5.0 inches 5

i n m xarry 4-2 5v3ASSOCIATESINC

The minimum overlay length is 1.5 K = 4.87 inches '

where R and t were assumed to be 11 inches and 0.96 inch, respectively. An overlay length of 5 inches is used for this design.

4.3 Application Recommendations

\ The following application recommendations were employed in conjunction with the above weld overlay design:

As-deposited weld overlay ferrite levels were measured after each layer,

~ with a target minimum of 8 FN. The weld overlay material was also specified .

to a maximum carbon level of 0.02% to ensure no propagation of the observed IGSCC into the weld overlay.

The minimum overlay thickness was measured after the first PT clear layer.

This allows for the possibility of some propagation of the observed IGSCC into the ferrite diluted first layer, without reducing overlay design margins. '

'v The welding was performed using a controlled heat input process (<40 KJ/in) with water in the pipe to minimize any further sensitization of the underlying piping material, and to ensure a favorable residual stress pattern.

The as-installed designs were subjected to a thorough preservice baseline examination, including ultrasonic examination of the weld overlay material and its bond to the pipe, and best-effort examination of the underlying weldment to define the existing indications.

STRUCTURAI.

INTEGRITY 4-3 / ASSOCIATESINC

4.4 As-Built Measurements As-built dimensions, ferrite levels and shrinkages have been measured '

before, during and af ter the application of the overlay on weld 22-22.

Results of these measurements are summarized in Tat .e 4-2. It is seen that these as-built dimensions meet all design requirements of Section 4.2, with the exception of the 450 maximum end-angle. As-built end-angle measure-ments ranged from 490 to 640 However, considerably larger end-angles were justified in (1) for the other overlays at Fitzpatrick, and this justifi-cation is also applicable to weld 22-22.

4.5 Shrinkage Stresses

_ As illustrated in Table 4-2, axial shrinkages of 0.076 inch to 0.159 inch were observed af ter the weld overlay repair of weld 22-22. Since weld 22-22 is an end-cap weld, however, there is no external constraint, and these i

axial shrinkages can be accommodated with no additional stresses developed in the recirculation system.

1,

?

6

. STRUCTURAL

.. INTEGRITY 4,4 ASSOCIAl'ES,INC.

g .,

g ,

• O 1

1 .

TABLE 4-1 WELO OVERLAY SIZING FOR WELD 22-2-22 1 ,

' l

\ pcCRACE STRUCTURAL INTEGRITY ASSOCIATES, INC.

VERSION 1.O, AFRIL 1985 SAN JOSE, CA (408)978-8:00 I

WELD OVERLAY SIZING OVERLAY S:2ING FOR CIRCUMFERENTIAL CRACFI:-

5 PASNV-05, END CAP WELD 22-2-22, J.A.

FITZPATRICV NUCLEAR FCt4EF Ft.A N T

[k WALL THIC6 NESS = 0. 9600 STRESS A;.T10= 0,4140 )

h L/ CIRCUMFERENCE FINAL A- -

O.2 0.3 0.4 0,5.._.3".,

OVERL A', TH!c)r:Fqs ," ,~)

> 0. OO O.7500 0.74e0

, , . ~, f.f , - [u GO O . .]. 00 O . ;200 o , -- , --

a

]

"em J

1 J

J w

ammm 4-5 l

4 TABLE 4-2

..do WELO OVERLAY AS-8UILT MEASUREMENTS - WELD 22-22 PASNY-05 AS-BUILT WOL DIMENSION LOCATION O DEG 90 DEG 180 DEG 270 DEG AVERAGE -

_______________________________________________________________ 1 (t)o inches 1.027 .993 1.039 1.084 1.03575 (D)o inches 8.01 8 8 8 8.0025 (t)1 inches 1.115 1.093 1.118 1.226 1.138

, (t)w inches 1.565 1.492 1.653 1.621 1.58275 (A)o inches 1.2 1.003 .9 1.002 1.02625

\ (D)w inches 7.934 7.882 7.888 7.841 7.88625

, (E)w inches 5.007 5.006 5.095 5.14 5.062 g

(F)w inches 5.534 5.876 6.088 5.837 5.83375 g FERRITE 4 IST 10 10 10 10 10 FERRITE # 2ND 12 12 12 12 12 FERRITE # 3RD 15 15 15 15 15 FERRITE ! 4TH 15 15 15 15 L 15 FERRITE i STH 17.5 17.5 17.5 17.5 17.5 SHRINKAGE inches .076 .118 .112 .159 .11625 1 (h)1 inches

.088 .1 .079 .142 .10225 (h)w inches .45 .399 .535 .395 .44475

$ (h)t inches .538 499 .614 .537 .547 THETA degree 63.90545 48.91992 51.03989 57.01741 55.22067 D. , Dm

'I A, L 7, , pm E* ' b -j l= _

Ae 3t e

• mst m.. e.Ax h, +

k t

i /

INTEGRITY I 4-6

m -4 g-.- g

~

WELD k

2.5" 2.5" 0.323 AFTER FIRST LAYE 450 MAX j AE0 W2 i \

0.96" l

MANIFOLD l

, END CAP 4

f NOT TO SCALE l

! Note:

1. Put first pass from toe to toe of the overlay.

2. Put second pass on the pipe side to equalize and smooth out overlay surface.

3. Items 1 and 2 together represent the first layer to be excluded in overlay effective thickness (per Generic Letter 84-11).

E M

FIGURE 4-1. Weld Overlay Design for Weld 22-22

I. .

5.0 CONCLUSION

S This report updates the fracture mechanics flaw evaluations and weld '

overlay repair design details for seven welds in the J. A. Fitzpatrick recirculation piping system which were re-inspected during the recent Spring, 1985 outage, and which showed some changes relative to prior examinations.

ia Six welds with relatively small, crack-like indications were previously g evaluated (1) in accordance with ASME Section XI, IWB-3640 and the g recommendations of NRC Generic Letter 84-11. In all cases, the indications iy we[eshowntobeacceptable. In addition, each weld was treated by IHSI, which should inhibit further IGSCC propagation in these welds. As a result of the current inspections, re-evaluations were performed on two welds h because of apparent increases in the length of the indications. These changes were shown to have an insignificant effect on the previous y analytical results. The previous evaluation results (1) for the other four welds remain directly applicable, considering the new inspection data.

L l

One new weld overlay repair was applied to a 22-inch end cap weld, in which the new UT results yielded relatively long, crack-like indications. The l repair was a full structural overlay, in which the design basis was to assurre no structural credit for any remaining pipe wall over the entire I length of the observed indication. It is also noteworthy that the first layer of weld overlay material was discounted in determining overlay M

thickness to account for possible crack propagation into diluted ferrite weld metal. These conservatisms were applied in addition to the standard weld overlay practices of controlling weldment ferrite and carbon levels and applying the overlays with low heat input and water in the pipe. Design and application of this weld overlay, as well as of the previous five weld overlays applied at Fitzpatrick, thus exceed the standard practices used in the large majority of weld overlays applied in U.S. BWRs, in terms of their ability to restore design basis structural margins and to arrest any future IGSCC propagation in the weldments.

STRUCTURAI.

INTEGRITY

- ASSOCIATESINC 5-1

\

I On the basis of these factors, it is concluded that the new (spring,1985) inipection results and corrective actions taken should not result in any

,_ reduction in design basis safety margins or increase in the probability of '

!*~-

a pipe rupture at the plant.

I C

I si lt I /

i 1

i

,[ INTEGRITY -

/ ASSOCIMEEINC 5-2

o l e

6.0 REFERENCES

1. Structural Integrity Associates Report SIR-84-038, Rev. O, " Fracture - '

Mechanics Evaluation and Weld Overlay Design for Recirculation System Piping in the James A. Fitzpatrick Nuclear Power Plant", February 4, 1985.

2. Structural Integrity Associates Report SI-83-002, Rev. 2, " Fracture Mechanics Evaluation of Observed Flaw Indication in 22-Inch Pipe to 1 End Cap Weld 22-02-2 James A. Fitzpatrick Nuclear Power Plant",

_ August, 1983,

3. GE Report -22A2622, Rev.1, " Design Report, Recirculation System for

' James A. Fitzpatrick Nuclear Power Station, ANSI-831.1 Calculations",

Dec. 6, 1976.

I i 4 ASME Soiler and Pressure Vessel Code,Section XI,1983.

i 5. EPRI Report NP-2662-LO, " Computational Residual Stress Analysis for l Induction Heating of Welded BWR Pipes", EPRI Project T113-8, Final g Report, Dec. 1982.

l 6. NUREG 1061, " Investigation and Evaluation of Stress Corrosion Cracking In Piping of Boiling Water Reactor Plants", U. S. Nuclear Regulatory

._ Commission, March, 1984.

7. Suchalet, C. B., and Bamford, W. H., " ASTM 8th National Spposium on Fracture Mechanics,1974", ASTM STP-590, pp. 385-402,1975.

8. Ranganath, S., Mehta, H. S., and Norris, D. M., " Structural Evaluation

+

t of Flaws in Power Plant Piping", ASME PVP-Vol. 94, Circumferential Cracks in Pressure Vessels and Piping - Vol. I, pp.91-116, 1984.

9. ASME Section XI Meeting Minutes, May 25, 1984 9

j I 10.

" Guidelines for Flaw Evaluation and Remedial Actions for Stainless Steel Piping Susceptible to IGSCC", Final Report for EPRI Project T303-1, Report No. SIR-84-005, April 13, 1984.

11. Bickford, R. L., et al, " Nondestructive Evaluation Instrument Sur-g veillance Test on 26-Inch Pipe", EPRI NP-3393, January, 1984 12.

_3 1 " Continued Service Justification for Weld Overlay Pipe Repairs", Final Draf t Report to BWR Owners Group, Prepared by Electric Power Research Institute, General Electric Company, NUTECH Engineers, and Structural j Integrity Associates, May 25, 1984.

e {

13. Peterson, R. E., " Stress Concentration Design Factors", John Wiley and Son, Inc, NY, 1953.

14. Heywood, R. B., " Design by Photoelasticity", Chapman and Hall, London 1952.

g7 hSTRUCTURAL omTT 6-1 3- ' ASSOCIATESIN

l ji. ' r References (continued) f

,}

i 15. Wichman, K. R., Hopper, A. G., and Mershon, J. L., " Local Stress in Spherical and Cylindrical Shells Due to External Loadings", Appendix B, Welding Research Council Bulletin 107, June 1977.

16. GE Report 22A2615, " Design Report for Recirculation System of Vermont

} Yankee Nuclear Power Station, ANSI-831.1 Calculation", June 10, 1970.

17. GE Report, "Results of Seismic Evaluation: As-Built Recirculation

.c Piping Inducing Replacement Actuator for F031 Discharge Valve for E.

I. Hatch Unit #1 Nuclear Power Station", Sept., 1984.

I 18. GE Report 22A4264, " Stress Report, E. I. Hatch, Unit #2 Nuclear Power 1

l

. ;-- - Station, Recirculation Piping System, Vol. I", March 23, 976.

[ 19. Structural Integrity Associates Report No. SIR-83-005, "A Damage Model y Based Cost / Benefit Evaluation of IGSCC Remedy / Repair Alternatives at the J. A. Fitzpatrick Nuclear Power Plant", Revision 0, Nov.,1983.

r 20. EPRI Report NP-81-4-LD, " Residual Stress Improvement by Means of Induction Heating", March 1981.

i e

W l.

^ sTaucrunn

)INTEGPITY ASSOCIATESlbC 6-2 I

_______ ________ _