ML20091M320

Recirculation,Rhr & Reactor Water Cleanup Weld Joint Insp Program - 1984 Maint/Refueling Outage

ML20091M320
Person / Time
Site:	Hatch
Issue date:	05/31/1984
From:	GEORGIA POWER CO.
To:
Shared Package
ML20091M315	List: ML20091M307 ML20091M320
References
TAC-55236, NUDOCS 8406110240
Download: ML20091M320 (29)

Text

. _ - .-. - - . . __

l 1

ENCIDSURE 1 Hatch Unit 1 Recirculation, RHR, and RWCU Weld Joint Inspection Progran -

[ 1984 Maintenance /Retueling Outage i

mee e

t 8406i10240 840531 PDR ADOCK 05000321 PDR 0

14AY 31 1984-1

TABLE OF CONTENTS 1.0 Review of Hatch Unit 1 Status as of Last Refueling Outage (Fall 1982) 1.1 Inspections Performed 1.2 Inspection Results and Actions Taken 1.3 Inspection Adequacy 2.0 Current Outage Inspection Plans 2.1 Proposed Inspection Scope 2.2 Inspection Procedure and Personnel Qualification 2.3 Inspection of Overlays 2.4 Criteria for Flaw Evaluation and Weld Overlay Repair 2.5 Leak Detection and Leakage Limits for Next Fuel Cycle 3.0 Justification for Continued Operation with Existing Weld Overlays 3.1 Current Overlay Design Details 3.2 Weld Overlay Residual Stress Data 3.3 Weld Overlay Materials Considerations 3.4 Weld Overlay Inspectability 3.5 On-going Progrms 4.0 Conclusions 5.0 References i

l l

WAY 311984 l

- ~ i- -

j

1.0 REVIEW OF HA'ICH UNIT 1 STATUS AS OF IAST REFUELING OUTAGE (FALL 1982) 1.1 Inspections Performed During the Fall 1982 maintenance / refueling outage, nineteen (19) circtinferential and branch connection welds in the recirculation systs were ultrasonically examined by Southern Cmpany Services (SCS), its contractor, Lambert-MacGill-Thmas, Inc. (IMT) , and Southwest Research Institute (SWRI) personnel. As a result of observing crack-like indications in the original scope of exminations, additional welds in the subject systen were exanined pursuant to ASME Section XI Code requirenents. Ultimately, a total of fifty-one (51) circumferential and branch connection welds were exanined to the requirenents of NRC I&E Bulletin 82-03, Rev. 1. In addition, eleven (11) welds in the RHR systen and six (6) welds in the FWCU systen were exanined. The welds exanined were picked on the basis of ASME Section XI Code, NURB3-0313, Rev. 1 guidance (high stresc welds with Sn > 2.4),

cmmitments to NRC resulting frm the Spring 1982 chloride intrusion, high stress rule index numbers, and/or high carbon content.

Procedures utilized during the examinations were provided by SCS and SWRI.

hhile IMT personnel performed exaninations, they were under contract to SCS and were subssjuently tested to and used the SCS procedure. The following is a stmnary description of procedures and techniques used during the exanination of the recirculation, RHR, and 1NCU piping at Hatch Unit 1.

A. Procedures The procer3ures r(quired a 3/8" dimeter,1.5 MHz 450 angle bean transducer to be used dt. ring the ultrasonic examinatien of the subject austenitic stainless steel piping.

B. Calibration Standards Stainless steel curved calibration blocks incorpcrating 1/S" cr 3/16" dianeter sido drilled holes at 3/4T,1/2T, and 3/4T depths were utilized.

C. Scanning Sensitivity For manual ultrasonic (Uf) exaninations, scanning was performed at 14dB above the Primary Reference Level. However, upon concurrence of a NDE Level III, this level ceuld be reduced if baseline screen noise hindered the interpretation of the UT scope CRT. 'Ihis scanning level was always at least 6dB above the Primary Reference Level. For mechanized examinations, scanning was performed at 6dB above the Primary Reference Level.

D. Recording Criteria All indications which produced a response greater than 20% of the Distance Anplitude Curve (Primary Reference Level) which -were not determined to be i

caused by outside dianeter gemetry were recorded. In addition, the 1 examiners were instructed that any indication interpreted to be significant l should be recorded.

- M AY 31 1984 _i.

s

l

[ Pursuant to the raguirenents of NRC I&E Bulletin 82-03, Rev.1, SCS and SwRI personnel ultrasonically examined at Battelle-Colmbus Laboratories (BCL)

[' five Nine Mile Point recirculation systen piping weld specimens containing l j intergranular stress corrosion cracking (IGSCC) indications. The purpose of i

.the exaninations was to evaluate crack detection capabilities of UT procedures and techniques of the various utilities and/or exanination

, agencies. SCS and SWRI exanination personnel denonstrated to the  !

satisfaction of the NRC inspectors in attendance at the qualification their

ability to adequately detect and evaluate IGSCC. As noted above, IMT personnel were tested to and utilized the SCS procedures qualified at BCL  ;

for exanination at Plant Hatch. All final reviews were performed by

BCL-qualified Ievel IIIs. %ese exaninations at BCL and the tR exanination

- procedures and techniques were representative of exaninations that were l performed at Hatch Unit 1. While the SwRI renotely operated mechanized (R

! equipnent procedures and techniques were not denonstrated at BCL, review of j' their- procedures and techniques by SCS personnel indicated they were adequate in the detection of IGSCC.

l.2 Inspection Results and Actions Taken 1.2.1'.End Cap-to-Manifold Welds l Ultrasonic exanination results indicated IGSCC _in the vicinity of four end

[ cap-to-manifold welds:

1-

1) 1B31-lRC-22AM-1 with a maximm depth of 63% of wall

} 2) 1B31-lRC-22AM-4 with a maximm depth of 72% of wall l: 3) -1B31-lRC-22BM-1 with a maximm depth of 64% of wall

4) 1B31 lRC-22EM-4 with a maximum depth of 67% of wall

.All'of the indications w & e axial and all had lengths of approximately 1/2".  ;

l E .

, %ese four welds were repaireo using structural weld overlays. The design '

overlay thickness was 0.25".. %e actual minimm thickness of the cmpleted 1 overlays was 0.275"i ne overlays were required to extend a minimm of 3.0" '

1 fra the original joint on the pipe side of the repair and a minimm of 3.5" i j on the end cap side of the repair.

P j_ One of the end cap-to-4nanifold welds was later found to have. a through-wall

crack .as evidenced by leakage observed during the weld overlay application.

i 1

1 Porosity was observed during . overlay welding of the welds 22AM-4, 22BM-1, i

' and 122BM-4. The regions with porosity were later repaired by grinding to the base ' metal and then filling the cavities with shielding _ metal arc

. welding.

d

'1.2.2 Elbow-to-Pipe Welds 7

, Ultrasonic ' exanination results indicated IGSCC in elbow-to-pipe weld lEll-lRHR-20BD-3. Five ' axially oriented flaws and ~ two iciremferentially .

, Toriented ~ flaws . were indicated. The deepest axial flaw 'had an indicated

. depth _ of L 94% of. wall. Se axial- length - of this flaw was approximately

~3/8". We ciremferentially oriented flaws were each approximately 1 1/2" length with a maximum depth of 33% of wall.

. _ MAY 31 i984 ~ -

p E

A structural weld overlay was applied to this weld. The overlay had design thickness of 0.4" and was required to extend to either side of the original joint by at least 3.5".

1.2.3 Pipe-to-Pipe Weld Pipe-to-pipe weld lEll-lRHR-24B-R-13 was indicated by UT inspection to have an axially oriented flaw with a depth of 47% of wall and a length of 1/2".

%is weld was selected for a structural overlay weld repair with the everlay design thickness being 0.3". He overlay was required to extend a minimum of 4.0" to either side of the original pipe joint.

1.2.4 Sweepolet-to-Manifold Weld Sweepolet-to-manifold weld 1B31-lRC-22AM-lEC-1 was found to have seven ultrasonic indications suspected to be IGSCC. The indications were snall flaws and all were transverse to the weld (similar to an axial flaw in a piping girth weld). ne largest flaw was determined to be approximately 12%

of the wall in depth and approximately 1/2" long.

This weld was not repaired since analysis showed that the weld would continue to meet all code requirenents for at least five more years of operation.

1.3 Inspection Adequacy Nhile the exmination ot the Hatch Unit 1 austenitic stainless steel piping welds was not perforn.ed to the more stringent. exmination requirenents of the later-issued NRC I&E Bulletin 83-02, the exminations conoucted were adequate and met the intent of the afore entioned bulletin. The following table simnarizes the NRC I&E Bulletin 83-02 exmination requirenents versus the exminations performed during the Hatch Unit 1 Fall 1982 maintenance / refueling outage.

NRC I&E Eulletin 83-02 Hatch Unit 1 Fall

, Exmination Requirements _

1982 Examinations

. Minimum of 10 welds 26 welds 120" dia-

) 20" dimeter meter examined in recirculation and RHR systens Minimum of 10 welds 23 recirculation riser in 12" risers and safe piping welds (includ-ends ing safe ends)~ examined Minimum of 2 sweep- 8 recirculation sweep-olet-to manifold welds olet-to-manifold welds examined Increase scope of exmi- As a result of observing nations pursuant to ASME crack-like indications in Code Section XI, IWB-2430 the original scope of MAY _31 1934

NRC I&E Bulletin 83-02 Hatch Unit 1 Fall Exmination Requirments 1982 Exminations if crack-like indications exminations, the scope observed was expanded pursuant to ASME Code,Section XI, IWB-2430 In addition to the quantity and size / type of welds examined, similar personnel, procedures, techniques, and equipnent as that used during the Hatch Unit 1 exminations were later used in the successful detection and interpretation of IGSCC during the ISE Bulletin 83-02-rsquired exminations at Hatch Unit 2. Therefore, it can be reasonably concluded that the Hatch Unit 1 exminations were adequate and met the intent of NRC I&E Bulletin 83-02.

2.0 CURRENT OUTAGE INSPECTION PIANS 2.1 Proposed Inspection Scope Georgia Power Cmpany intends to include thirty-seven (37) welds of the recirculation, RHR, and RWCU systens in the initial smple selected for examination during the 1984 maintenance / refueling outage at Hatch Unit 1.

'Ihe smple size has been determined based upon the following table.

Available Percent in Number of Resultant Condition Poculation Initial Sa ple Examinations Overlays of 6 100 6 repaired welds with IGSCC Unre;.oirei welds 1 100 1 with suspected IC3CC Previously inspected 60 20 15 welds (in recircu-lation, RHR, and RWCU systes) with no indications of IGSCC Previously unin- 71 20 15 spected (in recir-culation, RHR, and RWCU systes)

'IUTAL 37 Etrther distribution of the saple set will include . assurance that the different pipe sizes are represented in the smple.

MAY 311984 The welds proposed to be exmined will be chosen based on crack experience, where available. Where such infctmation is not available, high stress rule index number and/or high carbon content will be used for selecting the welds to be exmined.

If additional IGSCC is detected in the samples representing the welds not previously inspected or the previously inspected welds not found to contain IGSCC, the sample size will be increased in accordance with ASME Section XI, IWB-2430 as defined in NRC I&E Bulletin 83-02.

The proposed inspection of thirty seven welds in the recirculation, RHR, and RWCU systes meets the intent of proposed exmination criteria in the following documents:

e Hatch Unit 1 Safety Evaluation Report gdated February 11, 1983),

e NRC I&E Bulletin 83-02, e NRC letter SECY-83-267C, and e NRC Generic Letter 84-11.

2.2 Inspection Procedure and Personnel Qualification The stainless steel weld exminations to be performed during the 1984 maintenance / refueling outage will be conducted using a SCS procedure similar to those discussed above which were previously qualified at BCL. Since the inspection of weld; in 1902, tne procedure has been revised slightly and was successful in the detection of IGSCC at Hatch Unit 2 durir.g itu 1933 outage. Char.ges made to the procedure were editorial in nature. The latest approved revision of the procedure technically meets or exceeds the originally BCL-qualified procedure, e.g., calibration requirments, recording requirments, etc.

With regard to personnel qualification, all exmination persennel will be qualified to basic levels I, II, and III of American Society for Uondestructive Testing dccument SNT 'IC-1A as applicable. Additionally, all examination personnel will be dmonstrated to be qualified to a level of canpetence comensurate with their functions. Ex miners involved in-equipnent setup or scanning operations will be trained and dmonstrated proficient to assure both their technical ability and their ability to perform activities consistent with the principles of keeping radiation exposure as low as reasonably achievable (AIARA) .

l The process of qualification of all personnel who will perform evaluations will be that process presently in effect at the EPRI NCE Center. Currently, j there are five (5) Levels II and III SCS personnel who have qualified at l that facility by practical dmonstration. NDE personnel under contract to l

SCS who have qualified at the EPRI NDE Center in the detection and I interpretation of IGSCC may also perform exminations and evaluations, as appropriate.

WAY 311984

2.3 Inspection of Overlays The existing weld overlays in the inspection program as well as any necessary new overlays, will be ultrasonically ex mined to verify the integrity of both the weld metal and its bond to the pipe base material, in a manner consistent with ASME Code,Section V, Paragraph T550. In addition, a liquid penetrant exanination will be conducted on the weld overlay and 1" of base material on either side of the weld overlay.

2.4 Criteria for Flaw Evaluation and Weld Overlay Repair Flaw evaluation and weld overlay repair, if rquired, will be perfomed to criteria consistent with those specified in Attachnent 2 of NRC Generic Letter 84-11 dated April 19, 1984.

Should any new unacceptable, crack-like indications or any significant growth of old IGSCC cracks be observed during the inspection, information concerning the following will be subnitted to NRC for its review and approval:

e Flaw evaluations, and re-evaluation, and e Weld overlay design and analyses techniques, if rquired.

2.5 Leak Detection and Leakago Limits for Next R2el Cycle By letters dated February 10 and 11,1983, Georgia Power Canpany proposed Technical Specification changes to cugnent then existing reactor coolant leakage detection requirenents. NRC reviewed and approved the Technical Snecification chagen es discussed in the Hatch Unit 1 Safety Evaluation Report dated February 11, 1993. %e changes meet the intent of the leak detection and leakage limits discussed in Attachnent 1 of NRC Generic Letter 84-11.

On unrepaired sweepolet-to-manifold weld 1B31-lRC-2 Wi-lEC-1, two local scoustic enission devices wee installed to monitor for any potential leakage fran that particular weld. Acoastic anission devices are very sensitive and capable of detecting very tiny steam leaks. In the unlikely event of a through-wall cmck, these devices will provide an early warning to operations personnel to initiate appropriate correction action.

In addition to the above actions, a visual examination for leakage of the reactor coolant piping will be performed during each plant outage in which

- the containment is deinerted. We examination will be performed consistent with the requirenents of INA-5241 and Ih%-5242 of the 1980 Edition of the -

ASME Section XI Code. We systen boundary subject to this examination will be in accordance with IWA-5221. This examination is consistent with that identified in Attachment 1 of NRC Generic Letter 84-11.

3.0 JUSTIFICATION FOR OONTINUED OPERATION WITH EXISTING hTLD OVERIAYS 3.1 Current Overlay Design Details l 1

WAY 311984 ,

1 l

I

l 3.1.1 Overlay Design Approach '

The design approach used at Hatch Unit 1 resulted in repaired welds which meet or exceed the margins of safety which are inherent in the ASME Section III Design Rules (Reference 2) . Also, since the repairs involved leaving known flaws in the welds, the design approach also reflected the requirments of ASME Section XI, Article IWB-3640 (Reference 3). All overlays were required to have a design life in excess of five years.

Structural load capacity of the overlay weld repairs with respect to internal pressure and applied mechanical loads was dmonstrated via an ASME Boiler and Pressure Vessel Code Section III, Class 1 analysis. The design pressure of 1325 psi was obtained frm Reference 4. The dead weight and seismic loads were obtained frm Reference 5.

%e designs also cmply with the secondary stress and fatigue limits of ASME Section III. The thermal expansion loads were obtained frm Reference 5 and were conservatively cmbined into three cmposite trruients. The first is ,

a startup/ shutdown transient with a heatup/cooldown rate of 100 F/ 0 hour0 days <br />0 hours <br />0 weeks <br />0 months <br />. I

%e second cmposite transient consists of a 500F step temperature change and the third is an energency event with a 4160F step temperature change.

Fatigue resistance for the five-year design life of the repairs was evaluated by conbining the stresses fran the above strength evaluation with the thermal and other secondary stresses and perfctming a conventional fatigue analysis per Reference 2. A fatigue strength reduction factor of 5 was applied to account for the existing cracks. The fatigue usage factor was conservatively calculated assuming 38 startups, 25 small temperature change cycles and 1 mergency cycle within the de3ign life of 5 years, and was found to be negligible for all overlays.

Crack growth due to both fatigue and IGSCC was determined using Linear  ;

Elastic Fracture Mechanics crack grmth analysis techniques. The beneficial effect of overlay weld induced residual stresses was incorporated ir,to thece calculations. An allwable crack length and depth was established for each repair based on the net section collapse criteria of Reference 3. Crack ,

growth due to fatigue during the 5-year cesign life was calculated to be less than 0.01" for all of the repaired fit.ws and for the unrepaired sweepolet.

IGSCC crack growth was calculated using a conservative, eupirically based i crack growth law. The calculated crack lengths and depths were cmpared to the previously established allowable (based on the net section collapse )

criteria of Reference 3) and determined to be acceptable.

3.1.2 Overlay Design Perspective A _ total of six joints were repaired using weld overlays. Five of these six contained only axially oriented flaws. Since these flaws are due to IGSCC and thus depend on the presence of sensitized material for continued growth, their growth in the axial direction is restricted to the weld heat affected zone. Wis means that axial flaws cannot grow axially and thus will never present a significant pipe break threat. ,

1 l

MAY 31 1984 L

The overlay welds consist of 308L weld metal with controlled ferrite which has been denonstrated to be highly resistant to IGSCC. With this barrier to continued IGSCC at the outer pipe surface and the resistant 304 SS base metal limiting the axial growth, the axial flaws are effectively contained.

The only indicated 'circumferential flaws were relatively short (1.5" each) with the deepest having a maximm depth of 33% of the unrepaired wall. With the beneficial effect of the overlay weld induced residual stresses, calculations predict that these circumferential cracks will have essentially no growth during the five-year design life. However, even if these calculations or the 33% through wall sizing are significantly in error, the overlay for this joint is substantially overdesigned and would accomodate much longer, deeper ciremferential flaws with no loss in safety margin.

(IhB-3640 calculations for weld lEll-lRHR-20-BD-3 indicate the acceptable ciremferential flaw size for the overlaid weld to be a through-the-original-pipe-wall flaw of length agual to 30% of the ciremference, or 18.8 inches.)

3.2 Weld Overlay Residual Stress Data A wide body of analytical and experimental overlay weld residual stress data exists and is continually growing. The analytical data are primarily from finite elenent calculations using the BCL-developed WEWS-II progrm.

Experimental data consist of surface as well as through thickness stress measurements fran such techniques as hole-drilling, chip renoval and layer renoval. All overlays for which data are presented below had water inside the pipe during the ov& lay application.

3.2.1 EPRI/J.A. Jones 24" Overlay Mock-Up A 24" pipe with a 1.48" wall thickness was weld cverlaid at the J. A. Jones Applied Research Cmp =ay. 'Ihe overlay consisted of 5 weld layers for a total thickness of approximately 0.35". The overlay process was simultaneously modeled by NUTECH Engineers, Inc. using the WE!I;S-Il program (Reference 6). The results (both experimental and analytical) are shown in Figures 3.1 through 3.4. Wey dmoastrate that the axial and hoop stresses were cmpressive at the inner surface and reaained cmpressive for a depth of about 50% to 70% of the repaired wall thickness. The beneficial cmpressive hoop and axial stress was present after the first weld layer in much the sme form as after all five layers which shows that a thick overlay was not necessary in order to establish inner surface canpression.

3.2.2 hvIECH/ Georgia Power Canpany 12" Weld Overlay Mock-Ups Georgia Power Canpany, in conjunction with hVIECH Engineers (Reference 17) fabricated three weld overlay test specimens in conjunction with recent repair activities at Plant Hatch. A total of three specimens were fabricated, one each for a 0.20" overlay, a 0.23" overlay and a Last Pass Heat Sink Weld (LPHSW). 'Ihe weld overlay lengths were 4". The weld overlays were applied to butt welds in short sections of 12-inch, Schedule 100, Type 304 stainless steel pipe using the same procedures, operaters and sluipnent as were used for the in-plant repair work. The calculated axial residual stresses are shown in Figure 3.5 and the measured values (for a

-N 31 19g4 b

~

representative ' measurement location) are shown in Figure 3.6. Both the calculated and the measured results indicate that the inner half of the 4-repaired section is . in . axial cmpression. The calculated residual stress

l magnitudes are, however, significantly smaller than the measured values.

3.2.3 Structural Integrity /IVA Sweepolet Overlay Design Report

"Four sweepolets of . the 'IVA Browns Ferry Unit 1 recirculation systs were

repaired with weld overlays during the July 1983 outage. An inteoral part of the design analysis (Reference 7) was the simulation of the weld overlay repair with the finite elment progran WEIDS-II.

A sweepolet mock-up was welded by Welding Services, Inc. and used for both surface and through thickness residual stress measursents at J. A. Jones e Applied Research Cm pany. 'Ibe unrepaired wall was approximately 1.125" thick and the overlay was 0.25" thick by 4.0" long.

- Cmputed and measured transverse (to the weld) residual stresses are plotted in Figure 3.7 fr. a section at the top of the manifold and on the sweepolet side of the weld. . 'Ibe sweepolet mock-up had a - free edge near to the overlay

at this section and thus the model was exercised for two cases. The first case was for a. sweepolet attached to a long section of manifold and the second case was for a sweepolet with a very short manifold (without hoop constraint) . It 'is seen that the measurments are generally between the two

, calculated curves, and indicate cmpressive stresses on the inner portion of

!- the pipe wall.

5 Cmputed and measured transverse residual stresses are plotted in Figure 3.8 +

l for a section ' located at 90 degress to _tne section of Figure 3.7 (i.e., on the side of the manifold). Unfortunately, the measurments at this section

~ contain considerable uncertainty due to the . lack of strain data for the

. first two stress relaxation cuts. However, there is still qualitative agrement between the calculations and the . measurements and both suggest that cmpressive stresses exist within the inner half wall thickness for this section as well.- ,

3.2.4 Overlay Residual Stress Perspective

- All of the above evidence . indicates that the application of weld overlays (with water inside the pipe to act as a heat sink during welding) results in <

favorable residual stress patterns in terms of preventing. IGSCC initiation I

-and in terms of slowing or l arresting the . growth : of any existing IGSCC.

Results are available 'for- a variety of pipe sizes and schedules and for a i variety :of special- configurations -including sweepolets, endcaps, and safe I ends and all show favorable residual stress patterns. l

'3.3 Weld Overlay Materials Considerations - -1 3.3.1 Weld Metal Resistance to IGSCC- 'l lUntil recently, types 308 and 308L weld - nietals have generally been -l considered as imune to IGSC. This' review has been supported by the fact that no field leakage bas:ever been observed in any WR welds due to cracks

._9_

<=..

MY ? 3.1 1934

[&_.

i growing through the weld metal even though the residual stresses are generally higher in the weld than in the weld heat affected zone in which the cracking has generally been experienced.

Sm.e experimental data is now becming available which suggests that under sme circumstances, 308 welds and to a lesser extent 308L welds may be susceptible to envirorsnentally assisted crack propagation. The degree of susceptibility appears to correlate with the levels of ferrite and carbon in the weld metals as will be illustrated in several of the following exmples.

General Electric Test Procram As part of the GE experimental study to evaluate the structural stability of large dimeter pipes containing intergranular stress corrosion cracks, fracture mechanics (IT-WOL) specimens were fabricated frm Type 304 SS plates welded using either Type 308 or Type 308L containing varying ferrite levels (Reference 8) . The specimens were cycled in high tenperature water and 6 ppn 02 with an initial crack tip stress intensity factor range of 26 ksi (in)1/2 and an R of 0.05, where R is the ratio of minimum to maximum cyclic load. The specimens were tested for 5448 hours0.0631 days <br />1.513 hours <br />0.00901 weeks <br />0.00207 months <br />.

It was found that intergranular stress corrosion cracks which had initiated in the base metal had grown into the weld metal in six of the seven specimens. For the 308L weld metal specimens, all of the weld metal cracks arrested within .031 inches of the point of entry into the weld metal. (Branches of the cracks in the base metal continued to grow). These 308L welds contained frm 5.5 to 11.5% ferrite and 0.025 wt% carbon.

'Iwo of the 308 weld specimens had ferrite levels of less than 9%. For the specimen with the 1mest ferrite (1.9 - 3.3%) the crack arrested after growing 0.104 inches into the weld. For the specimen with ferrite in the range of 7.0 to 8.5% the crack grew to 0.101 inches into the weld and showed no signs of arresting.

The single 308 weld specimen with ferrite greater than 9% (9.5 - 11.5%)

had the crack grow .045 inches into the weld and then arrest. The carbon level for all the 308 weld metal was 0.053 wt%.

Weld Metal Cracking in Inverse IHSI Smole j Several 12 inch pipe smples of 304 SS were fabricated by Ishikawajima l Harima Heavy Industries using girth welds which were induction heated so l as to produce IGSCC when exposed to high purity, oxygenated, 5500F water (Reference 9) . One of these samples developed an intergranular strees corrosion crack which grew into the weld several millimeters before arresting. 'Ihe weld was examined metallurgically to determine the level of sensitization and the ferrite content. The weld metal, which was 308 was found to be highly sensitized, due probably to a 5000C/24 hour sensitization treatment. The crack entered the weld metal at a point with approximately 5% ferrite and appeared to arrest at a point with approximately 9% ferrite.

N 31

~

Igg i

k_ _- m. -

Constant Extension Rate Tests (CERTS)

Test results frm CERTs (References 10 and 11) for low carbon Types 308 and 308L are shown in Table 3.1. The CERT is similar to a tensile tests, performed at slow strain rates (constant extension rate) in an aggressive envirement (5500 F , 8 ppu oxygen water) to force fracture.

The fracture is then exmined for IGSCC characteristics. No indications of IGSCC were found for the Types 308 and 308L specimens, even for the cases where the weld metal was given a severe furnace sensitizing treatment.

Constant Ioad Tests Constant load tensile test results at 5500F in 0.2 to 100 ppu oxygen content water for 308 and 308L are given in Reference 12. Ioads are as high as 125% of the yield strength at 5500F. Results include as-deposited and furnace sensitized conditions. No failures resulted for 308L specimens regardless of ferrite content and no failures resulted for Type 308 specimens with ferrite levels greater than 8%.

Other constant load tests are reported in Reference 12 for 308L in chloride environments. Type 308L weld overlays on Type 304 stainless steel were tested at 125% of the 750F yield strength in an queous enviroment of 100 ppu Cl- at 2000F. No cracking or attack was found after test times of 178 and 138 hr., even for a specimen given a sensitizing treatment of 10 hr. at 11500F.

Ferrite Effect on 308 and 308L Sensitivity to IGSCC A laboratory study investigating the int e action effects among carbon level, territe level and ferrite distribution cn the IGSCC susceptibility and sensitization immunity to Types 308 and 308L weld metal is descriced in Reference 13. It was observed that although chraaium carbide precipitation occurs int.ergeanularly during heat treatment - (at sensitization tenperatures) of Types 308 and 308L weld metal, the precipitation occurs -solely along the austenitic ferrite  ;

grain boundaries. Since the ferrite is rich in chranim , and the diffusion of chroniun in ferrite is faster than in austenite by approximately a factor of 1000, the chraniun for this precipitation is  ;

supplied principally fran the ferrite phase. The modest chroniun  !

depletion in the austenite is replenished in time by diffusion of chranium fran within the austenite grain. After this " healing", the material is thought to be imune to IGSCC.

1

%e model developed in Reference 13 predicts the required amount and l distribution of ferrite for the above described immunization as a l function of the carbon content. For as-depositeo Type 308L SS containing 0.03wt% carbon, the model suggests that approximately 3%

ferrite is rq uired for inrunity to IGSCC; whereas for Type 308L containing 0.015wt% carbon, essentially no ferrite is reluired for imunity to IGSCC.

It is seen fran the above data that the 308L weld metal specified for the Hatch Unit 1 overlay repairs has been subjected to many severe tests. %e results of these tests have either been that no IGSCC occurred in the weld MAY - 31 f984 e.

~

I

metal or that lIGSCC (which initiated in the base metal and propagated into the weld) arrested within a short distance frm the cracks entry into the weld metal.

Se above data indicates that 308 weld metal is more susceptible to IGSCC than 308L and that ferrite levels in excess of 8-9% may be required for 308 .

to have IGSCC innunity cmparable to 308L. Since many of the welds in existing.-piping systens are 308 with ferrite levels significantly below 9%

and since no leaks have resulted fra IGSCC in these welds, this suagests that the Hatch Unit 1 overlays, made with 308L weld metal and greater than 8% ferrite possess a high degree of confidence in their ability to arrest any IGSCC.

3.3.2 Weld Metal Fracture Toughness

%e overlay design report for Hatch Unit 1 (Reference 1) includes analyses to determine the failure loads for the repaired sections using a tearing modulus approach. These analyses involved calculating the fracture mechanics parameter J as a function of axial crack length for the gemetry of interest -and for a nunber of load levels using elastic-plastic constitutive relations. With this information it is then possible to '

construct a curve of J versus T (the material tearing modulus as defined in Reference '14) in which each point on the curve represents a different level of -applied. load (increasing loads correspond to increasing values of J) .

We' straight-lines in Figure 3.9 which intersect at the origin are applied J/r curves for the weld repairs at Hatch Unit 1 taken frm heference 1.

The failure load for a given crack configuration . is determined by the o intersection of the appropriate applied J versus ? (J/T) curve with a material J/T curve. . Material J/T curves are plotted in Figure 3.9 for a variety of stainless steels (Reference 15) . Base metals as well as weld metals are included. Se material curve of Reference 16 which was used in

-the Hatch Unit 1 overlay design report (Reference 1) is also included.

~

It is seen from Figure 3.9 that, the material. J/T curves represent a wide range of material toughnesses. We materials with curves closest to the

. origin are the least tough and these curves are: generally curves for . weld

metals rather than base metals. It is clear that sme of these welds metals-have been found to be .less tough than the base -material curve of Reference 18.

We least tough materials in Figure 3.9 are the CF8A welds tested by 'Gudas and the subnerged arc welds' on 304 SS _ base metal tested by Westinghouse.

hbile more toughness testing of welds is needed to clarify .this weld metal -

toughness issue, the data of Figure 3.9 seens to' indicate that the subnerged arc . weld process results L in - less - tough welds than the GTAW/TIG welding process used in the Hatch Unit 1 overlay repairs

~

-If one considers only' the weld data'of Figure 3.9 for 308L and TIG or GTAW welding processes, and extrapolates the lowest toughness curve as indicated (in Figure L 3.9 lone obviously. get factors of safety which are smaller - than were reported Lin - the : Hatch c Unit : 1 design report using the curve ~ from i

"" 3 AY"31j984 '

1 C -

u .- _ -

--_=. . - - - -.

Reference 16. However, the decrease in these factors is not as large as one might expect due to the very nonlinear relationship between applied load and J. For example, the text of Reference 1 states that the factors of safety for the end cap overlay designs (as given by the J/T analysis) are in excess of 4. This is still true if the extrapolated 308L/GTAW material toughness curve of Figure 3.9 is used. Similarly, the factor of safety for the unrepaired sweepolet was stated as being greater than 3.3 and is still greater than 3.3 with the lower material curve.

The factor of safety for the pipe-to-pipe overlay design was greater than 4 ,

with the material data of Reference 16 and is slightly less than 4 for the lower material toughness curve of Figure 3.9. Of all the J/T analyses reported in Reference 1, the elbow-to-pipe overlay design had the smallest calculated factor of safety of about 3. With the lower toughness curve, this is reduced to a factor of safety slightly greater than 2.5. Becaus.

these are axial flaws which must grow through the much tougher 304 base metal before posing a serious threat to the integrity of the pipe, and because of the conservatism of the postulated 0.8" semicircular flaw used in the J/T analysis, this factor of safety is considered adequate.

3.4 Weld overlay Inspectability One concern which has been raised regarding long-term operability of EWR pipe welds with weld overlay repairs is the relative difficulty of conducting confirmatory, non-destructive exminations on such welds.

Conventional ultrasonic inspection techniques have had only limited success at inspecting through weld overlays to identify and size the underlying IGSCC in the original pipe joint.

In discussing this concern relative to Hatch Unit 1, however, it is important to note that the Hatch Unit 1 overlay designs, being full structural overlays, applied for the most part to axial cracks, do not rquire it.spection of the material underneath the overlay for assurance, oi

- structural integrity. ihe only rquir ment is to d monstrate that the et acks have not grown into the weld overlay material itself. A very careful, ultrasonic baseline exanination of the Hatch Unit 1 overlays was conducted following their application, and subsequent exanination during the upccming refueling outage will be conducted using similar proceduras and quipnent and qually qualified personnel. Such an examination is expected to produce a highly reliable inspection for any crack growth or other service induced degradation of the weld overlay itself.

With respect to inspection under the weld overlay, recent work at the EPRI EDE Center has dmonstrated considerable success with the use of refracted longitudinal ultrasonic waves, rather than the more usually applied shear wave propagation mode (Reference 17). The results obtained to date deonstrate the ability to verify overlay integrity and measure crack length. Additional work is underway to define crack depth sizing ability with this approach.

With respect to crack length measurment, the work at the EPRI NDE Center has dmonstrated that, if the overlay has suitable surface finish, and if NAY 31 1984 _13_

C

the crack is detectable, then the techniques for length measurment (with refracted L-waves) are it ical to the case where there is no overlay. The preliminary data show that e . ors in length measurment are on the order of the probe dimension, and that the length estimates are quite tr;peatable.

~ Since there were only two circumferentially oriented indications in Hatch Unit 1 at the last inspection (Weld lEll-lRHR-20-BD-3) and these indications were quite short (1.5 inches each), the length measurment procedure described above should be more than adequate to ensure the continued integrity of this weld. As noted above, the overlay design on Weld lEll-lRHR-20-BD-3 can accanodate a considerable increase in length, assuming a full through-wall crack depth, before violating IWB-3640. Also, as described above, there is no significant possibility of length crack growth on the axially oriented indications since they would very rapidly run out of sensitized heat affected zone into a material which is not susceptible to IGSCC in BWR enviroments.

3.5 On-going Progr ms In addition to the data cited above, EPRI, the Nuclear Regulatory Ccmission and the nuclear power industry are supporting a wide range of progras to further qualify the weld overlay as a long term repair. Included mong these are a pipe test progra and destructive examination of several weld overlaid joints reoved frm service. These progras, described in more detail below, will provide additional data to support continued operation of the Hatch Unit 1 weld overlays in a time frme consistent with the planned startuo of the plant following the upcming refueling outage.

3.5.1 General Electric Pipe Test Laboratory Overlay Test Plans EPRI is sponsoring a series of pipe tests in which the beneficial effect of overlay induced residual stresses on crack growth will be quantified experimentally (Reference 18) . 'Ihe first objective is to produce data to further support the use of overlay repairs for two fuel cycles (36 months).

This first objective will be met by the end of July 1984. The second objective is to define the effective life of r eedies applied to IGSCC cracked pipe, including weld overlay, induction heating stress improvment (IHSI) and last pass heat sink welding (LPHSW). The progrm will end in December 1985.

The main thrust of this progra will be the perfomance of pipe tests in the GE Pipe Test Laboratory (PrL) . A total of approximately 70 weldments will be tested under varying loads in a simulated EWR enviroment. The testing is similar to that used to statistically quantify the factor of improvment associated with the use of Type 316NG stainless steel as an alternate alloy for BWR piping. Test variables will include -initial IGSCC, weld overlay design ano applied stress level.

In addition to the pipe testing, several overlays which have been taken out of service will be examined under this progra. 'Iwo full thickness overlays reoved fra Monticello will be examined. Both overlays were applied to riser pipes. One was applied to a circumferential flaw and the other to an

-I'~

WAY 31 ggy L

axial flaw. 'Iwo mini-overlays from Hatch Unit 2 will also be examined.

Both of these overlays were applied to risers with circtznferential cracks.

The exmination will include determinations of sensitization due to the overlay weld, weld residual stress via MgCl 2, and crack depth by

- metallography.

3.5.2 Weld overlay 'Ibst Plans for Overlays frm Hatch Unit 2 A second program of destructive examination of overlaid welds reoved frm service is currently underway at Argonne National Laboratory. This progra (Reference 19) has the following objectives:

o Determine if the overlay process causes cracks to grow during application.

e Determine if additional crack growth has occurred during the approximately 12 months of plant operation after the overlays were applied, o Measure residual stress On the I.D. surface at several locations to determine general state of stress.

e Determine how much sensitization of pipe base metal occurred as a result of overlay application e Accurately determine ferrite content of the overlay material, e Determine if cracks will propagate into the weld overlay material.

'Ihe scheduled empletion date for this progran is the end cf 1984.

1 WAY; 31' 1984 L . .. . ,.

Table 3.1 1

Constant Extension Rate Test (CERT) Results '

For Types 308L and 308 Weld Metal (References 10 and 11) l Nominal Strain Heat No./ Sample Tvoe Heat Westment %C Rat e(min-1) IGSCC Notes M7616/26-B 308L As-Deposited .03 4.5 x 10-5 No. 1 M7616/27-B 303L Solution Heat .03 4.5 x 10-5 No 2 Wested L-B7 30S 13500C/l Hr. .04 1.0 x 10-3 No 3 L-B7 308 13500C/l Hr. .04 2.0 x 10-4 No 3 L-B7 305 13500C/l Hr. .04 2.0 x 10-5 No 3 L-B7 308 '13500C/1 Hr. + .04 1.5 x 10-4 No 2 4750CAO Hr.

L-B7 308 13500C/l Hr. + .04 1.3 x 10-4 No 2 4750C/100 Hr.

L-B7 308 ,13500C/l Hr. + .04 1.3 x 10-4 No 2 4750C/1000 Hr.

L-B7 308 13500C/l Hr + .04 1.0 x 10-5 No 2 6000C/l Hr.

L-B7 308 13500C/l Hr. + .04 1.3 x 10-5 No 2 6000C/2 Hr.

'L-B7 308 13500C/l Hr. + .04 1.3 x 10-5 No 2 6000C/10 Hr.

L-B7 308 13500C/l Hr. + .04 13 x 10-5 No 2 6000C/20 Hr.

L-B7 308 13500C/l Hr. + .04 1.6 x 10-5 No 2 6000C/20 Hr.

L-B7 305 13500C/l Hr. + .04 1.4 x 10-4 No 2 6000C/100 Hr.

IrB7 308 13500C/l Hr. + .04 1.3 x 10-4 No 2 7000C/l Hr.

Notes: 1 -- Corrosion resistant clad overlay; procedure specifientien requires a ferrite content of at least 8 FN.

2 -

No ferrite centent given.

3' 20% volume ferrite for this treatment.

MAY 31'f984 k - . . _ .

7-AXI AL STRESS AFTER FIFTH LAYER ll*ER SURFACE OF PfPC ~~~~ OUTER suRTACE OF PIPC to so- ,

40- /

t, ll x-  : '.  :

\

\ ..  :

j 20- ..- t

' ,: .- i ,

x m- ,.. ...- ...,

.f

., I

,./ .

., , .i m o- - -

-=

-n- \,,,/ y m -

g

-ro-

-u. '

' (x) Hole-tvill: M

- c- 3 cWoJ

-SC- .

EXTENT Cr OvCRLAY I e

=40 . , , ,

, , , i 10 4 e 4 2 0

• -2 -4 -4 -6 - 10 '

.AXI AL DISTANCE (INCH) l Figure 3.1 Calculated Inner -and Otiter Surface Axial Overlay Stresses for a 24 in. Pipe with 1.48 in. Wal^l.

(Overlay contains five weld layers for a total thickness of 0.35 in.) Experimental Results Also Shown for. Comparison.

l l

MAY 311984 _17_

a l

l l

l HOOP STRESS AFTER FIFTH LAYER l

IMOt SWtFACE OF PIPE l

---

CUTER SWIFACE OF PIPC so , ,

1 so-  !

.o- 2. i ,

x. \

l

, i, 20- '

I R ,i ' w) Hele Drilling McIhod ,

u m- ,

~

m e- -

m l

• w X

.e

-e. $ s s

" # 8 '

-2 . 2 X y

-m- , ,

-so- -

Exum or ovour i I

~40 .

.. . . i ,

m e . 4 2 o

-2 -. -..

-4 -o AXI AL DISTANCE (INCH)

Figure 3.2 Calculated Inner and Outer Surface Hoop Overlay Stresses for a 24 in. Pipe with 1.48 in. Wall.

(Overlay contains five weld layers for a total thickness of 0.35 in.) Experimental Results Also Shown for Comparison.

_. M AY . 31 1984 - - 18 .

.. .- -"rerNi

y -.. - - -

E C4, a

G STRESS THROUGli THICKNESS AFiER FIRST LAYER STRESS TilROUGil TillCKNESS AFTER SECONO LAYER AirtAt $1R($$ ----- HOOP STRCSS AMI AL Sir 4E55 ---- HOOP STRESS 37 37

• 12 37 ,

N Al Axl AL DISTANCE UIIIA AI ANI AL OlstAw"( l

SiRfACf

• SistACE O SCO INCL (5  ; y pgt O SD0 INCL (5 l g p, I I I

• l tt - ,

• 12- ,

I 3 t U U G -E E .

- i - I g l183- g it:3- l 3 3

.' z =o.5 d

l z = o.5 "

l '  : '

n7s -  : n 2s .  :

. .I lit (R j gggg simiAct  ; suntAct I I T PIPE I 1 M Part o ss , , ,-- , , , , , , , ,

-- m es- , , ,- , , , , , , , , 2

- 6ft 4 0 70 -n 0 D 20 30 to 50 6G SO 30 30 0 10 20 30 40 So so SIHESS (KSI) STRESS (KS1)

Figure 3.3 Calculated Through Wall Stresses Af ter the First and Second Weld Overlay Layers for a 24 in. Pipe with 1.48 in. Wall. (Overlay contains five weld layers for a total thickness of 0.35 in.)

4

E

'N ,

c.

5 STRESS TIIR00Gli iltlCKNESS AFTER FIFill LAYER Arsat sintss - --- same stetss

~

st 31

• WIIR Af AFI AL OlsIANCE .

0 Soo eHC4(s ,

1. I",

p 12 - ..-

.g q o u

• Z ,e O a g 1883 l a  :

ra Y

b4  :

i r -o 5" n se  :

,I i entR st9(S CL I '

I or PiK

• ' 4'O S'O 60 SO bo -20 10 0 s'D 20 3'O STRESS (KSI)

Figure 3.4 Calculated Through Wall Stresses After the Fifth and Final Weld Overlay Layer for a 24 in. Pipe with 1.48 in. Wall. (Overlay contains five weld layers for a total thickness of 0.35 in.)

1

l l

WELD OVERLAY TESTS l

OUTSIDE DIAMETER o 1 a.g .

.0.8 --

o 0.7 - - l

/ l

_ .. l E 0.s --

O ,

z g 0.4 - -

o

, 0.3 --

0.2 - -

0 0.1 - -

80 iO -60 60 40 3IO 2'O 1'O b 1'O 2'O 30 4'O 5'O 60 IO 80 STRESS (KSI)

INSIDE DIAMETER N PR E84.0611

.20" OVERLAY HATCH-1 (LONG) WELD PREP 20mm FROM CENTERLINE 1

Figure 3.5 Calculated Through-Wall Axial Residual Stress of 12" Schedule 100 Pipe l

WAY 311984

I WELD OVERLAY TESTS OUT3IDE DIAMETER 0.9 - -

0.8 - -

0.7 - -

O 5 0.8 -

E 0.5 - 0 O

O 0.4 '- -

o 0.3 - -

0 0.2 - -

0.1 -

6 . 6 - , 6 6 . . 6 , , , , ,

70 60 50 -40 30 20 10 0 10 20 30 40 50 60 70 STRESS (KSI)

INSIDE OfAMETER

.2tr* OVERLAY nmes m2 HATCH-1 (LONG) WELD PREP 20mm FROM CENTERLINE Figure 3.6 f4easured Through-Wall Axial Residual Stress of 12" Schedule 100 Pipe l

MAY 311984 .

i

. l

~

50 without hoop

\ s constraint 40 g y

\ ^

30 -

. t

/M t

20 -

\

\ j 10 -

i /

Cr \

g _

. 1 (ksi) \y p

O% yx I

-10 -

y ,

s

' l

+ ,

-20

),

y' N with hoop constraint

~l

-0 -

A overlav '

. N inner surface

-0 -

(e) hole-drillirg method

% ' TWcd4 WI.t., REST 0 GAL STR W M6ALv T M f?M

, t ,

f 9.5 10.0 10.5 11.0 11.5 12.0 Distance from manifold centerline, (in)

Figure 3.7 Comparison of Through-Wall Stress from TVA Sweepolet Longitudinal Section Models (with and without hoop constraint) and Experimental Results.

MAY 31 1964

40 -

initial anneal stress

~

30 _ -

w' \

\

20 _

\

, \

l f%

~

\ 3 10 y, g g 4

g x hcMe-drilling \ w 0

raethod N 4 s  %-

(ksi) 0 y,

3  % \

/

O V'

• 2 g V-  %

g

-10 -

3 p

-20 -

s

. s

-30 - / ~

s,

- overlay l

. inner surface

-50 -

Y r}rscum % alt.- kcsituat EW SS W8th'f-6MEr#3

( s rgeW de4E s AP#u(D AN WO CMG)

I . r I 10.0 10.5 11 . 0

1. . (in)

Figure 3.8 Comparison of Through-Wall Stress from TVA Sweepolet Transverse Section Model After Each Overlay Layer and Experimental Results MAY ;31 1984 -.

lg . / TP.34 PIATE (53tfF)

/ / PAes' era.

/ / (sf esh th -

/l/

E34,

- \

l'l - ESM ,

/ *.ma EPRI-NP-1161

• g e(RCFin

? rt - =2

'8 M) APPLIEb T T C4WES. ,

N.c (/ f~o R. H ATC. MNLT'1. TP 3o4 Wei.D

.a 4 M e

/lf y

(r<EF f)3 s

3g. QT/,#

(27,StrdR w/ TP 4 bTE b, P

((x*

6 Or?.T s s 550*F

\

]

\

'j A 0

1. ' TP,5oV WELDI 3 (4# Tuln .

we$." \

i ~ CF8A WELSS TP.144 PL ATE I QuD S ETAL. (4B -J2,5So*f) g//7 41eds) W E ST.

t . , TP.24 WELD SoBL GTAW )

( IT,550*F) GUDAS erat Q ////

g ,f-TP.3o4/ W $LD U/W *A-4, f=50'F) WE37. .

O po see 3so  % goo Goo no T'

i Figure 3.9 Weld and Base Metal Toughness Data w.ith Applied J/T Curves for Hatch Unit 1 Weld Overlays l

l l

MAY 31 1984 1

l l

I

4.0 CONCLUSION

S e Inspections proposed for the recirculation, RHR, and IMCU systens during the Fall 1984 maintenance / refueling outage meet the intent of the Hatch Unit 1 SER, NRC I&E Bulletin 83-02, NRC letter SECY-83-267C, and NRC Generic Ietter 84-11.

e Continued operation with the existing six overlay repaired recirculation and RHR piping welds and the unrepaired recirculation sweepolet weld is acceptable for an additional cycle because:

A large anount of test data shows that Type 308L weld metal is not susceptible to IGSCC. Ferrite contents of 8% or greater provide additional margin against IGSCC.

Theoretical predictions support the test data and show that Type 308L with at least 8% ferrite is virtually immune to IGSCC.

- IGSCC cracks should not propagate into the 308L weld overlays applied to the six affected recirculation cnd RHR piping welds at Hatch Unit 1.

The unrepaired sweepolet-to-manifold weld was previously shown by analysis to continue to meet all code reluirenents for at least five years of operation.

Since the overlays at each weld joint provide structural adequacy, and the overlays are essentially imune to further crack growth, operation for an additional cycle will not reduce Safety Margins below those intended by the ASME Code,Section III.

Numerous tests of weld overlays are on-going; early results show favorable residual stress patterns.

1 MAY 311984

l

5.0 REFERENCES

1. " Design Report for Recirculation Systs and Residual Heat Removal Systm Weld Overlay Repairs and Flaw Evaluation at Hatch Unit 1", NUTECH Report GPC-04-104, Revision 1, March 1983.

2. ASME Boiler and Pressure Vessel Code Section III, Subsection NB, 1974 Edition with Addenda through Sumer 1975. t

3. ASME Boiler and Pressure Vessel Code Section XI, Paragraph IWB-3640 and Appendix C, " Acceptance Criteria for Austenitic Stainless Steel Piping, 1983 Edition with Winter 1983 Addendum.

4. General Electric Design Specification 22A1344, Revision 3.

5. General Electric letter G-GPC-2-Sll, "IE Bulletin 79-14 for E. I. Hatch Nuclear Plant Unit 1 - Transmittal of Preliminary Results of Recirculation Systs Analysis and Design Drawings," Dec e ber 17, 1982.

6. " Optimization of Weld Overlay Repair for ER Piping - Phase 1", NUTECH Report EPR-12-101, Rev. O, June 1983.

7. " Design Report for Recirculation Piping Sweep-o-lets Repair and Flaw Evaluation, Browns Ferry Nuclear Plant Unit 1", Structural Integrity Associates Report SIR-83-006, Revision 0, October 21, 1983.

8. "The Growth and Stability of Stress Ccrrosion Cracks in Large Dimeter BWR Piping", EPRI NP-2472, Vols. 1, 2, July 1982.

9. " Assessment of the Feasibility of Producing Pipe Smples With Tight

'Ihrough-Wall IGSCC, EPRI NP-2241-LD, February 1982.

10. N. R. Hughes Reedies, Volmeand 2",A. J. Giannuzzi, " Evaluation of Near-Term ER Piping EPRI NP-1222 Volme 2, Novmber 1979.

11. N. R. Hughes and A. J. Giannuzzi, " Evaluation of Hear-Term %R Piping Reedies, Volme 1", EPRI NP-1222 Volme 1, Novmber 1979.

12. W. L. Clarke and W. L. Walker, " Accelerated SCC Test Data Tabulation for 304 CF8, CF3, 308L and Wrought Austeno-Ferritic Alloys", General Electric Co., Unclassified, August 15, 1975.

13. N. R. Hughes Reedies, Volmeand3",A. J. Giannuzzi, " Evaluation of Near-Tenn &R Piping EPRI-NP-1222, Novmber 1979.

14. U. S. Nuclear Regulatory Ccmnission, "A Treatment of the Subject of Tearing Instability", U. S. NRC Report NURM-0311, July 1977.

15. J. F. Copeland, S. S. Tang, P. C. Riccardella, "An Assessnent of Weld Metal Toughness Influence on IWB-3640 Safety Margins", EPRI contract RP2457-3, presented at ASME Section XI Meeting, San Antonio, April 23, 1984.

MAY 31 1984

[..

r-

16. EPRI-NP-2261, " Application of Tearing Modulus Stability Concepts to Nuclear Piping", February 1982.

17. " Continued Service Justification for Weld Overlay Pipo Repairs",

Prepared by EPRI, General Electric, NUTECH Engineers, and Structural Integrity Associates for BWR Owners Group Pipe Cracking Progrm, May 25, 1984 Draft.

18. J. N. Kass, "EPRI/GE Degraded Pipe Reedies Progra , T-302-01, Presentation to IMR Owners Group Special Meeting, Chicago O' Hare Hilton, April 11, 1984.

19. L. J. Sobon, "NUTECH/ Georgia Power Co./NRC Weld Overlay Test Progra -

Phase II", Presentation to IMR Owners Group Special Meeting, Chicago O' Hare Hilton, April 11, 1984.

S

' W AY 31 1984