IR 05000313/1997014

Insp Repts 50-313/97-14 & 50-368/97-14 on 970527-0807.No Violations Noted.Major Areas Inspected:Maint & Engineering

ML20216H346
Person / Time
Site:	Arkansas Nuclear
Issue date:	09/12/1997
From:	NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV)
To:
Shared Package
ML20216H336	List: IR 05000313/1997014 ML20216H330
References
50-313-97-14, 50-368-97-14, NUDOCS 9709160206
Download: ML20216H346 (27)

Text

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

_ _ ...__ __

d

.

. l LNCLASML1 U.S. NUCLEAR REGULATORY COMMISSION -!

i

. REGION IV  :

- Docket Nos.: 50 313 [

50 368 j License Nos.: DPR 51 i

-

NPF 6  ;

Report No.: 50 313/97 14 '

50 368/97 14

"

Licensee: Entergy Operations, In ;

,

Facility:- Arkansas Nuclear One, Units 1 and 2-t Location: Junction of Hwy. 64W and Hwy 333 South -

.

Russellville, Arkansas y

'

Dates: May 27 through August 7,1997 inspector: 1. Barnes, Technical Assistant

,

Accompanied By: Dr. C. V. Dodd. NRC Consultant ,

Approved By: Arthur T. Howell lit, Director Division of Reactor Safety i

,

v i

i iTTACHMENTS:

- Attachment 1: Supplemental Information . -

Attachment 2: Entergy Operations Arkansas Nuclaar One Unit 2 Position Paper on

"

" Missed" Indications I b

,

O k

-

'

l P

..

.

n .

m

'

, e.n,,,-w+>--.r- s,.w,.-- r. , v v yr-- y,- y- .e- %=vw ,.-rv~ w y v- --%-+ y y e-+ww-y '-wwy g y- 7- y-y- v y -w-r-, y-----+y--- --

g*-ru+w-~7*

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

.

. 2-EMC1)TIVE SUMMARY Arkansas Nuclear One, Units 1 and 2 NRC Inspection Report 50 313/97 14;50 368/97-14 This inspection was performed to assess Unit 2 steam generator degradation status and the circurnstances pertaining to the issue on April 11,1997, of a Notice of Enforcement Discretion (and subsequent Amendment 180 to Facility Operating License DPR 51) for the Unit 1 steam generator Technical Specification requirements, h1R!n19nangg

• No significant disagreement with either production analyst " calls" or resolution determinations occurred during review by the NRC consultant of a sample of Refueling Outage 2R12 eddy current data, it was considered that some of the axial tube flaws present at eggcrate locations would not be routinely identified if analysis was cursory, with analysis difficulty increased by the presence of mix residuals of the order of 1 volt and some of the data being noisy (Section M1.1).

Ena!DEliDG

• The licensee use of a majority of data from Crystal River 3 in the Unit 1 bobbin coil sizing technique qualification, which had a different intergranular attack morphology and tubing degradation location to that observed at Arkansas Nuclear One, Unit 1, was considered technically questionable. An apparent violation was identified in regard to the f ailure to comply with the data sample and morphology requirements of the selected qualification method (Section E1.1).

• The licensee f ailure to correct regression analysis predictions of flaw depths (to reflect effects of data scatter) was considered nonconservative and a direct contributor to under prediction of true depth of intergranular attack (Section E1.1).

• The acceptance limit in Appendix H of Electric Power Research Institute Report TR 106589-V1 of 25 percent root mean square error (for flaw depth sizing techniques) was considered a potential contributor to errors in eddy current depth siring predictions (Section E1.1).

• The overall Unit 2 Refueling Outage 2R12 examination scope was considered to be comprehensive and appropriately responsive to emerging degradation modes in Combustion Engineering steam generators Section E1.2).

• The degradation data differences between the Unit 2 steam generators indicated that differences in thermal hydraulic characteristics potentially existed (Section E1,2).

__

. _ ______________ __________ -

.

.

3-

• An apparent violation was identified pertaining to the f ailure to comply with the requirements of the Unit 2 Technical Specifications 4.4.5.4a 7 and 4.4.5.4b during Outage 2F961. A total of 210 tubes were documented during a Refueling Outage 2R12 review of Outage 2F961 eddy current data history, as containing flaw indications in excess of the 40 percent plugging or repair limit of the Technical Specifications (Section E1.3).

• The licensee appropriately implemented Unit 2 oddy current program commitments made in a March 25,1997, meeting with the NRC staff. Of particular note was the use of an *Entergy Review Group" for improved oversight of eddy current data analysis activities, measures implemented to track and trend analyst performance, and actions taken to improve training and testing of analysts (Section E1.3),

• The licensee responded promptly to the notification of the identification of eggerate erosion damage in a Combustion Engineering steam generator at another licensee's f acility. No evidence of similar degradation was found in the Unit 2 steam generators (Section E1.4 * An apparent violation was identified pertaining to the failure to take prompt corrective action (i.e., acceptance for continued service, without a demonstrated technical basis) of sleeved tubes exhibiting weld zone eddy current indications. An inspection followup item was also identified regarding the future examination criteria adopted for two sleeved tubes, for which a nonconformance report disposition had identified the potential for early primary water stress corrosion cracking (Section E1.5).

• The microstructure present in both Tubes R16C56 and R70C98 (i.e., fine grain size and intragranular carbides) was characteristic of that resulting from use of a low final annealing temperature. This type of microstructure no;mally results in enhanced susceptibility to inte granular stress corrosion cracking, compared to a coarser grain structure with predominantly intergranular carbides present (Section E8.1).

• Lead and sulfur (as sulfate) were present on the crack surfaces of Tubes R16C56 and R70C98, indicating their invulvement in the degradation process (Section E8.1).

• Pitting or erosion of grain faces was observed at tube depths at least up to mid wall in Tube R70C98. The grain pitting appeared to be indicative of a crack that had been in existence for a considerable period. The metallurgicalinformation suggested that tube through wall degradation in excess of tube mid wall was

necessary at eggerate support locations prior to bobbin coil tube examinations being capable of its detection (Section E8.1).

• Burst test values of 3200 psig and 3250 psig, respectively, were obtained for Tubes R16C56 and R70C98. These values exceeded the main steam line break accident analysis pressure of 2950 psig, but did not meet the Regulatory Guide 1.121 structural integrity criteria of 4750 psig (Section E81).

.

. 4 figport Qq1p l Sgnrnar.y of Pla_n1Slalug Unit 1 was at 100 percent power dunng the onsito portion of the inspection. Unit 2 was in Refueling Outage 2R12 at the beginning of the inspection penod, with Mode 1 reached following the outage on June 9,199 Il<Molnignmitt M1 Conduct c' Maintenance M 1.1 fluview of Tube Examination Data jnpoection Sqqng The inspector selected a limited sample of Unit 2 eddy current data from Refueling Outogo 2R12 for independent assessment by the NRC consultant. Included in the assessment scope were tubes that had been selected for in situ pressure testing; tubes containing tungsten inert gas and kinetic welded sleeves; defectivo calls by the primary and/or secondary analyst, which were overruled by the resolution j

'

analysts; and calls initially identified by the Entergy Review Grou Observations and Findinns The NRC consultant had no significant disagreement with either production analyst

" calls" or the resolution process, as a result of his review of a sample of Refueling Outage 2R12 eddy current data. During review of kinetic welded sleeve data, the NRC consultant noted that the data indicated that significant probe lift off occurred during the examinations and questioned the inherent ability of the selected plus point probe to fit down into the wold region ond give a sensitive examination. The NRC consultant also considered that the plus point coil examination techniques (used for both the kinetic welded and tungsten inert gas wolded sleeves) had been optimized for detection of tubo outside diameter indications, For tungsten inert gas wolded sleeves, the inspector did not consider this examination approach to be the most sensitivo, in that the preponderance of defects would be expected to initiate at the interf ace between the sleeve and the tube inside diameter surf ac DurMg the inspection, the NRC consultant performed a limited review of three coil (i.e.,0.115 inch pancake, axial sensitivo, and circumferential sensitive) motorized rotating pancake coil and bobbin coil tube data from eggerate support location In this review, which included a small amount of data from both Refueling '

Outage 2R12 and Outage 2F961, it was noted _that some of the axial tube flaws present at eggcrato locations would not be routinely identified by cursory analysi The NRC consultant noted that a mix was set up using a drilled support plate to eliminate the effects of the eggcrate support. Af ter mixing, however, a residual of the order of 1 volt remained, which made defect identification more dif ficult. Some of the motorized rotating pancake coil data was also noisy, therefore, tending to mask smaller signal _ _ _ _ _ _ - _

.

, 5- C_0ECLVH19El No significant disagreement with either production analyst " calls" or resolution determinations occurred as a result of review by the NRC consultant of a sample of Refuehng Outage 2R12 eddy current data, it was considered that some of the axial tube flaws present at eggerate locations would not be routinely identified if analysis was cursory, with analysis difficulty increased by the presence of mix residuals of the order of 1 volt and some of the data being nois AllJnainM1h19 E1 Conduct of Engineering E 1.1 Unit 1 Notice of Enfo1 cement Discretion During Refueling Outage 1R13 in 1996, an oddy current technique was used to depth size intergranular attack that was present in Unit 1 steam generator tubing in the tube region located within the upper tube sheet. The licensee utilized the criteria contained in Supplement H2, " Qualification Requirements for Examination of Steam Generator Tubing," to Appendix H of Electric Power Research Institute Report TR 106589 V1, "PWR Steam Generator Examination Guidelines: Revision 4,"

Volume 1, for performance of the sizing molification. On April 8,1997, the licensee discovered that the destructive examination results (i.e., metallography)

from three tube samples removed during Refueling Outage 1R13 indicated nonconservative bias of up to 50 percent through wall in the eddy current measurements of 1 packages of intergranular attack, thus, creating the possibility that tubes could have been left in service with flaws which exceeded the plugging or repair limit of the Technical Specification The staff made a determination on April 11,1997, to exercise discretion not to enforce compliance with Technical Specification 4.18.5 b until May 7,1997, or the date of issuance of an amendment to Technical Specification 4.18.5.b. The amendment, which was issued on May 7,1997, authorized operation to the next refueling outage with tubes containing intergranular corrosion indications in excess of the Technical Specifications plugging limit, [pJngetion Scone (50002)

The inspector reviewed the sizing qualification report, including the methodology used for the sizing qualification process; the criteria contained in Appendix H to Electric Power Research Instituto Report TR 106589 V1; the circumstances described in the request for Notice of Enforcement discretion; and the information contained in Licensee Event Report 50 313/97-001 01 and its supplement dated July 15,199 _ _ _ _ - - -

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

,

,

, 6-

.

I Qbservations and Findinas During review of Supplement H2 to Appendix H of Electric Power Research Instituto Report TR 106589 V1, the inspector ascertained that Report TR 106589 V1 (which has not been formally endorsed by the NRC) considered a technique qualified for sizing if a root mean square error of less than or equal to 25 percent through wall was demonstrated. The inspector considered that if this root mean aquare error acceptance limit was approached in technique qualification, it would create the potential for errors in eddy current depth sizing predictions. Of particular concern to the inspector was information provided by licensee personnel, which indicated that the regression equation (derived from the relationship between eddy current flaw ,

depth predictions and metallographic measurements of flaw depths) was applied without correction for the offects of observed eddy current data scatter on the flaw ,

depth sizing prediction. The inspector viewed this information as a probable direct contributor to the observed under prediction of actual flaw depths in the removed tube samples.

"

Review by the inspector of the data used to establish the sizing qualification identified that the majority of the data originated from Crystal River 3. The inspector questioned the basic validity of this approach, in that it did not appear technically appropriate to use tubing free span intergranular attack, from a plant with a different operating history to Arkansas Nuclear One, Unit 1, to predict depths of intergranular attack in tubing in the upper tube sheet. The NRC consultant also informed the inspector regarding British work that was performed in the 1980s by McIntyre on conductivity of intergranular attack. This work, which was published as Electric Power Research Institute Report NP-4478, identified that tubing intergranular attack located in a tube sheet could exhibit significantly greater conductivity (i.e., up to a factor of six) than intergranular attack located in the free span of tubing. This information both illustrated a potential pitfallin the licensee sizing approach and suggested to the inspector that eddy current detection capabikties would be lower for intergranular attack located in a crevice region compared to degradation present in the tubing free spa ,

Review of the root cause analysis contained in the supplement to Licensee Event Report 50 313/97 001 identified two root causes for the under prediction of flaw depths. The first identified root cause was that the data sets used in the sizing qualification did not fully conform to the requirements of Appendix H to Electric Power Research Institute Report TR 106589-V1, with respect to data distribution (i.e., only 7 of 53 data points were greater than or equal to 60 percent flaw depth versus a requirement for at least two thirds of the data points to be in that range).

The second identified root cause was that there were morphology differences between the flaws used to develop the sizing technique and the flaws in tubes removed during Refueling Outage 1R13 for laboratory examiution. The intergranular attack in the Crystal Rivet 3 tube samples was symmetrical and pit like in configuration, whereas the intergranular attack in the licensee pulled tube samples (removed from the steam generators in 1982 and 1984) was patch like

'

with no consistent physical shape. The licensee reported that the intergranular attack in the 1996 pulled tube samples was also patch like in configuration, but contained sho3t deep intergranular penetrations. The licensee concluded that the

_ - _ _ _ _ _ _ , _ _- __ _ ,_ . _ _ _ _

l

.

'!

, -7  ;

morphology differences between the data sets used for the sizing qualification and

- the observed flaws in the tube samples removed during 1996, resulted in the data sets not being appropriate for sizing the 1996 morphology flaws. The inspector did x

not fully concur with the licensee position, in that the bobbin coil examination technicue inherently averages the depth of the flaw present. Accordingly, patch-like flaws without consistent physical shape would inherently tend to be sized with potential vanance from true maximum dept The performance nf the sizing qualification, using data that did not conform to the requirements of the selected qualification method, was identified to licensee personnel on August 8,1997, as an apparent violation of Criterion IX of Appendix B te 10 CFR Part 50 (50 313/9714-01).

}

- _QR DI R M L7

.

.

The licenseb use in the sizing qualification of a majority of data from Crystal River 3, whLi had a different morphology and tubing location to the observed Unit 1 intergranular attack, was considered questionable. The ac eptance limit in Appendix H of Electric Pawer R9 search Institute Report TR-106589 V1 of 25 percen;ioot mean square error (for flaw depth sizing techniques) was considered a potential contributor to errois in eddy current depth sizing predictions. The licensee f ailure to correct regression analysis predictions of flaw depths, to reflect effects of data scatter, was considered a direct contributor to the observed under prediction of trae depth of intergranular attack in the pulled tube samples. An opparent violation was identified in regard to the f ailure to compiy with the data sample and-morphoiogy requirements of the selected qualification nietho E1.2 Review of Refuelina Outaae 2R12 Tube Examination Scoce and Results Insoection Scone (GOOO2J The inspector reviewed the tube examinatin scope and methods that were csed in Refueling Outage 2R12 with respect to Technical Specification requirements,

industry guidance, and as a result of emerging degradation modes. A review was also performed of tube plugging and repair history for the Unit 2 steam generators, the specific plugging data for R:!1eling Outage 2R12, and the reasons for tube

_ plugging, i Observations and Findinas The licensee Unit 2 final examination scope for Refueling Outage 2R12 was: a full

'

length bobbin coil examination of_ all active tetos in both steam generators, with the exception of tne aiea below sleeves (where 20 percent of the tubes were examined with the'cobbin coil); a rotating pancake coil examination of all active hot-leg side expansion transition regions in both steam generators; a rotating pancake coil examination of 20 percent of the cold-leg side expansion transition regions in the

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

.

.

area bounded by the sludge pile; a plus point probe examination of all active Babcock and WHcox kinetic welded sleeves: a plus point probe examination of a 20 percent sample (plus 28 previously identified indications) of active Combustion Engineering tungsten inert gas welded sleeves (which was expanded to 100 percent

) of the sleeves in Steam Generator B); a rotating pancake coil examination of a 40 g percent sample of Row 1 low radius U bends; and rotating pancake coil examination of a 20 percent sample of tubes at hot leg side drilled support locations (which was expanded to 100 percent for Steam i3cnerator B). The referenced rotating pancake coil probe examinations utilized a 0.115-inch diameter pancake coil, an axici

] direction sensitive coil. and a circumferential direction sensitive coil. This probe is commonly referred to as a Delta probe. The licensee has been able to demonstrate, by in-situ pressure testing, that its selection of the Delta probe (over the more sensitive plus point probe) for examination of hot-leg side expansion transitions has not resulted in a decrease in demonstrated structural integrity margin with respect to Regulatory Guide 1.121 requirements. The inspector considered the overall examination scope to be comprehensive and appropriately responsive to emerging degradation modes in Combustion Engineering steam generators, b Table 1 lists the cumulative repair history for Unit 2 Steam Generators A and B.

3 A total of 350 tubes and 366 tubes, respectively, were removed from service in f Steam Generators A and B, which resulted in respective reductions of 78 and 18 y( sleeved tubes in service. Both steam generators remained below current approved

• ~ effective repair limits, f Tables 2 and 3 show, respectively, a cumulative compilation for Steam Generators A and B of degradation by repair location. These tabulations con"nue to exhibit what was first observed, during the initial steam generator tube integrity inspection performed in 1992, regarding differences between the steam generators in incidence of degradatio As of Refueling Outage 2R12, the total repairs in Steam Generator A related to the top of tube sheet and sludge pile regions (i.e., degiadation helieved to be related prirnarily to outside diameter stress corrosion cracking in the tube expansion transition region and sludge pile intergranular attack) were 1542 tubes, versus a corresanding total of 520 tubes in Steam Generator B. Attempts to explain this difference in incidence of degradation between the steam generators focused on the observec' greater historical sludge quantities in Steam Generator A and the original return of the partial domineralizer output to the condensate train for Steam Generator B. The corresponding degradation rates at eggerate support locations (believed to be primarily axial stress corrosion cracking) were, however, reversed, with a total of 280 tubes repaired in Steam Generator A, versus a total of 604 tubes in Steam Generator B. Wear at batwing locatiuns in the tube bundle also varied significantly between the steam generator 3 as of Refueling Outage 2R12, with respective repair totals of 19 tubes in Steam Generator A and 121 tubes in Steam Generator I 1,

- _ _ _ _ _ _ _ _ _ _ _ _ _ _-- _

.__

.

Table 1 UNIT 2 GTEAM GENERATORS (SGs) A AND 8 TUBE REPAIR HISTORY Time of Repair Operational . SG A Repairs SG B Repairs Outage N Time (EFPYs')

Plugs SI' Plugs Sl2 Pre Commercial 0.00 15 0 29 0 2R1 (1981) 0.89 0 0 0 0 2R2 (1982) 1.69 0 0 1 0 2R3 (1982) 2.?3 0 0 0 0 2R4 (1985) 3.31 0 0 0 0 2RS (1986) 4.16 0 0 0 0 2R6 (1988) 5.38 0 0 6 0 2R7 (1989) 6.52 0 0 0 0 2R8 (1991) 7.67 0 0 73 0 2F92 (1992) 8.51 29 392 11 56 2R9 (1992) 8.85 67 ( 4) 132 0 2P93 (1993) 9.36 47 0 3 0 ,

,

2R10 (1994) 10.20 170 (-1 ) 77 (-1 ) ___

2P95 (1995) 10.85 215 0 85 0

-

2R11 (1995) 11.49 200 412 150 175 2F961 (1996) 12.43 73 ( 6) 144 (-7)

'

2R12 (1997) 12.80 350 (-78) 366 (- 18) d

• '

\ M

' ' '

'

_

Total Repairs 1881 1282 ]

Effective Repairs (Plug Equivabnt)' 1192 1083

% Repairs (Effective. Total) 14.18, 22.36 12.88,15.24 1 - Effective full power years of operation 2 - Sleeves 3 - Based on 18 sleeves for kinetic welded and 44 sleeves for tungsten inert gas welded being considered thermally equivalent to 1 plug

,

I

.

Table 2

_

UNIT 2 STEAM GENERATOR A REPAIR HISTORY BY REPAIR LOCATION Outage EFPYs' RW2 TSP) SP' TTS5 FS5 OT'

t Pre Commercial 0.00 0 0 0 ,

0 15 2R1 (1981) 0.89 0 0 0 0 0 0 2R2 (1982) 1.69 0 0 0 0 0 0 2R3 (1983) 2.33 0 0 0 0 0 0 2R4 (1985) 3.31 0 0 0 0 0 0 2RS (1986) 4.16 0 0 0 0 0 0 2R6 (1988) 5.38 0 0 0 0 0 0 2R7 !1989) 6.52 0 0 0 0 0 0 2R8 (1991) 7.07 0 0 0 0 0 0 l

,

2F92 (3/1992) 8.51 1 1 166 253 0 0 2R9 (10/1992: 8.85 9 30 11 17 0 0 2P93 (1993) 9.36 0 0 2 45 0 0

.,

2R10 (1994) 10.16 3 7 13 147 0 0 2P95 (1/1995) 10.86 0 0 12 203 0 0 2R11 (9/1995) 11.46 2 43 38 519 16 21 2F96-1 (1996) 12.43 0 29 6 13 4 21 2R12 (1997) 12.80 4 170 31 66 38 41 m + m Total Repairs 19 280 279 1263 58 98 1 Effective fu:1 power years of operation 2 - Batwing 3 - Tube support plate 4 - Sludge pile 5 - Top of ~;ube sheet ti Free span 7 - Other

!

-

0-11

.

Table 3 UNIT 2 STEAM GENERATOR B REPAIR HISTORY BY REPAIR LOCATION Outage EFPYs' BW' TSP' SP* TTS' FS* OT'

Pre-Commercial O.00 O O O O O 29 2R1 (1981) 0.89 0 0 0 0 0 0 2R2 (1982) 1,69 1 O O O O O 2R3 (1983) 2.33 0 0 0 0 0 0

_

2R4 (1985) 3.31 O O O O O O 2H5 (1986) 4.10 0 0 0 0 0 0 2R6 (1988) 5.38 6 0 0 0 0 0 2R7 (1989) 6.52 O O O O O O 2R8 (1991) 7.67 16 52 5 0 0 0 2F92 (3/1992) 8.51 2 3 4 58 0 0 2R9 (10/1992) 8.85 25 94 5 8 O O 2P93 (1993) 9.36 0 0 0 3 O O'

2R10 (1994) 10.16 17 32 5 23 0 1 2P95 (1/1995) 10.86 0 0 5 80 0 0 2R11 (9/1995) 11.46 19 73 19 215 0 4 2F961 (1996) 12.43 11 100 3 13 0 17 2R12 (1997) 12.80 24 250 21 53 4 14 m

Total Repairs 121 604 67 453 4 65 1 - Effective full power years of operation 2 - BatwinD 3 - Tube support plate 4 - Sludge pile 5 Top of tube sheet 6 - Free span 7 - Other

-_

l

.

.

-12-

. Conclusions The overall Unit 2 Refueling Outage 2R12 examination scope was comprehensive and appropriately responsive to emerging degradation modes in Combustion Engineering steam generator The degradation data differences between the A and B steam generators indicated that differences in thermal-hydraulic characteristics potentially existe E1.3 Review of Steara Generator Tube Examination Proaram Reauirements inspectipfi Scoce (50002)

The inspector compared the steam generator eddy current examination program requirements or Refueling Outage 2R12 against specific commitments made by the licensee during a March 25,1997, meeting with the NRC's Office of Nuclear Reactor Regulatio The inspector reviewed: Document ANO 2-OTH ESP-SGMAN, " Arkansas Nuclear One Unit 2 Steam Generator Eddy Current Training Manual," Revision 1; Engineering Standard HES-14, " Steam Generator Integrity Program Unit 2," Revision 1:

Engineering Standard HES-26, " Steam Generator Tube Plug Defect Management,"

Revision 7: Engineering Standard HES-28, "ANO-2 Steam Generator ECT Data Analysis Guidelines," Revision 6 and Standard Change Notice (s) 01 through 04; Engineering Standard HES-29, "ANO-2 Steam Generator ECT Perfo; nance Demonstration Guidelines," Revision 3; Engineering Standard HES 31, "ANO 2 ECT Data Management and Control," Revision 0; and HES-42, " Steam Generator Tube Sleeve Management," Revision 1. In addition, the inspector observed in-process eddy current analysis activities, discussed program criteria with licensee and contractor staff, and made a specific review of the Refueling Outage 2R12 program requirements with respect to the March 25,1997, meeting commitments, Observations and Findinas The inspector determined that the licensee had appropriately implemented the commitments made in the March 25,1997, meeting. The inspector considered the most significant process change maae was the creation of an "Entergy Review Group" to strengthen the oversight of eddy current analysis activities. This group, which was made up of experienced Level 111 eddy current analysts, reported directly to licensee supervision and provided an independent evaluation of the analysis process, included in the group charter was a review of data discarded by the resolution process, sampling of tubes with "no_ detectable degradation" calls by

___- _____-

.

.

-13-primary and secondary analysts, review of historical comparisons performed by the resolution analysis team, review of guideline change forms, and ensuring that all repairable codes that had been dispositioned as non-repairable by resolution analysis had been. appropriately reviewed and documente The inspector also noted that the data analysis guidelines had been simplified to make them more user friendly, training of analysts was more structured with direct licensee personnel involvement, and testing of analysts was more focused on known Unit 2 degradation modes. All bobbin coil indications at eggerate supports, which were confirmed by a rotating pancake probe, and bobbin coil indications in the free span that exhibited growth, were required to be plugged. The licensee also implemented, for the first time, use of new Zetec analyst efficiency tracking software. This toftware provided the capability u track and trend individuai analyst 4 performance. 't he performance improvements that resulted from use of this process were notabl Review of the program criteria identified that acquisition of eddy current data was required to conform to the Electric Power Research Institute PWR Examination Guidelines. The inspector noted, however, no similar expectation with respect to performance of eddy current data analysis. The inspector found this status to be somewhat unusua While observing in process eddy current analysis activities, the inspector noted that eddy current analysts had documented (for indications that were confirmed s

to be flaws during Refueling Outage 2R12 and accordingly removed from service)

that re-review of Outage 2F96-1 data for these tubes indicated that flaw indications in excess of the 40 percent plugging or repair limit (Unit 2 Technical Specifications 4.4.5Aa.7) were present at that time. The inspector ascertained that a total of 78 tubes in Steam Generator A and 132 tubes in Steam Generator B were documented, from the re-review of Outage 2F961 data, to contain flaw indications in excess of the Technical E pecification limit. Unit 2 Technical Specification 4.4.5.4b requires steam generator tubes, which exceed the plugging or repair limit, to be plugged or repaired before determining the steam generator to be operable. The failure to plug or repair the 210 affected tubes in Outage 2F96-1, prior to return of the Unit 2 steam generators to service, was identified to licensee personnel on August 8,1997, as an apparent violation of the Unit 2 Technical Specifications 4.4.5.4a 7 and 4.4.5.4b (50 368/9714 02).

The licensee subsequently provided a position papes regarding these indications,

-which has been included as Attachment 2 to the inspection report. The inspector concurred with the licensee's comments that the leaker outage, Outage 2F96-1, was not caused by a missed call, that the flaws detected in Refueling Outage 2R12 were typically of small amplitude, that the in situ pressure test results were adequate, and that the enhanced performance resulted from the program improvements made prior to Refueling Outage 2R12. The position paper did not

- address, however, that it was the licensee's records, which indicated that the

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

_ - - - -

_ - . - - - . - -- -

o

.

-14-Unit 2 Technical Specification plugging or repair limit was exceeded for the tubes in Outage 2F961, without plugging or repair being accomplished. The inspector also noted that bobbin amplitudes for indications were similar between Outage 2F961 and Refueling Outage 2R12, and accordingly concluded that the defect indications were present in Outage 2F901. The difference in result appeared to the inspector to relate primarily to the increased use in Refueling Outage 2R12 of confirmatory rotating probe examination , Conclusions An apparent violation was identified pertaining to the failure to comply with the requirements of the Unit 2 Technical Specifications 4.4.5.4a.7 and 4.4.5.4b during Outage 2F961. A total of 210 tubes were documented during a Refueling Outage 2R12 review of Outage 2F961 eddy current data history, as containing flaw indications in excess of the 40 percent plugging or repair limit of the Technical Specifications. The licensee appropriately implemented Unit 2 eddy current program cornmitments made in a March 25,1997, meeting with the NRC staff. Of particular note was the use of an "Entergy Review Group" for improved oversight of eddy current data analysis activities, measures implemented to track and trend analyst performance, and actions taken to improve training and testing of analyst E1.4 VisualInsoeetion of Unit 2 Steam Generator Eancrate Supports n Lnmaction Scone (50002)

The inspector revwwed video tape of remote visual examinations that were performed, using a Welch-Allen camera, of the periphery of the eggerate supports in Steam Generator A. Visual examination of the eggcrate peripheries in the Unit 2 steam generators was performed by the licensee following notification of the identification of erosion damage at this location in the San Onofre Nuclear Generating Station, Unit 3, steam generators. The specific tape segment reviewed was obtained from the West side of the steam generator hot leg, Qbservations and Findinas The video tape reviewed indicated that the peripheral qgcrate isttice in all of the supports was in excellent condition, with no visual evidence of any erosion damage or thinning, Conclugigng The licensee responded promptly to the notification of the identification of eggerate erosion damage in a Combustion Engineering steam generator at another licensee's facility. No evidence of similar degradation was found in the Unit 2 steam generator __ _ . _ . - - . -

_ _ _ _ __ _ _ _ _ . . _ _

.

.

-15-E1.5 Review of Unit 2 ABB-Combustion Enaineerina Tunasten inert Gas Welded Sleeves Inspection Scone (50002)

A followup was performed of problems associated with the installation of welded sleeves by ABB Combustion Engineering during Unit 2 Refueling Outage 2R11 in November 1995. The inspection scope included a review of sleeve welding procedure and welding operator qualifications, procedural requirements, and sleeve nonconformance c.ieory and disposition Observations and Findinas The inspector reviewed automatic gas tungsten arc Welding Procedure Specification GTAA-43.43.992, Revision 0, and its supporting procedure qualification record, GTAA 43.43 01503, Revision 01. In addition, a review was performed of the qualifications of six welding operators who had performed automatic gas tungsten arc sleeve welds during Refueling Outage 2R11. No discrepancies were noted during this revie The inspector reviewed the technical requirements contained in contract procedures for sleeve installation pertaining to tube cleanliness preparation; sleeve insertion, expansion, and welding; post-weld heat treatment; and nondestructive examination of the siseve, tube parent metal, and weld, included in this review were:

(1) Procedure STD-400157, " Visual Examination of Steam Generator Tubs Sleeve and Sleeve Plug Welds," Revision 0; Procedure 00000 NSS-062, " Procedure for Ultrasonic Examination of Steam Generator Tube to Sleeve Upper Welds, Revision 10; Procedure STD 400-151, "Linearity Verification of Ultrasonic Instrumentation Used for the Examination of Sleeved Tube Welds, Revision 0; Procedure STD-400158, " Technical Operating Procedure for the Transition Zone Sleeve Rolling Tool and Controls for Steam Generator Tubes With 3/3 inch .043 inch or .048 inch Wall Thickness Using the Computer Controlled Sleeving

- System," Revision 0; Procedure STD 500-020, " Technical Operating Procedure for the Post Weld Heat Treatment and Heater Dryout of Above the Tubesheet (ATS)

Welds Using Resistance Heating for Steam Generator Tubes With .048 inch Wall Thickness," approved July 31,1992; and Traveler 2005331-001, " Installing Tube Sleeves in the Steam Generator Tubesheet Expansion Transition Zone Using the Genesis 2000 Control System," Revision O. The inspector observed during this review, that no provisions for inspection of the cleaned tube surface (prior to insertion of the sleeve) had been included in the process controls. This procedural guidance omission was subsequently corrected by ABB Combustion Engineering following the identification of the presence of oxide inclusions in Prairie Island pulled tube sample L . _ - . _ _ _ _

.

.

- 16-The inspector ascertained from review of Contract NAS00176, effective date-August 14,1995, that the contractor's quality assurance program was invoked for the sleeving activity and that the licensee had required its review and approval of disposition recommendations for nonconforming conditions. The inspector examined the nonconfortronce report log for the contract (ABB Project 2005331)

and noted a total of 28 reports existed. The inspector selected a sample of nonconformance reports for review, including those issued for sleeve over expansions and eddy current exarnination indications. All nonconformances reviewed were noted to have been signed by either one or two licensee employee During review of Nonconformance Reports 2005331-14, 2005331-17, and 2005331 18, the inspector noted that ABB Combustion Engineering dispositioned tubes, which exhibited eddy current wald zone indications and suspected geometric anomalies, as "use-as is" based solely on review of ultrasonic examination and visual inspection results. Samples of sleeved tubes exhibiting eddy current indications viere not removed from the steam generators for laboratory examination to allow identification of the reason (s) for the eddy current signals and verification that other nondestructive examination methods could be relied upon for assurance of structural integrity. Laboratory examination in 1996 of samples of Prairie Island sleeved steam generator tubes, exhibiting similar eddy current indications, identified that the eddy current indications were the result of circumferentially oriented oxide inclusions and/or weld suckback on the sleeve outside diameter. These indications t were contrary to the sleeve license amendment, in that the minimum weld height was not being met and the oxide inclusions provided potential leak path The inspector questioned the licensee regarding the technical bases for disregarding the results of the eddy current examinations, in that it had not been demonstrated that ultrasonic examination and visual inspection could be relied upon to determine whether the eddy current indications were (or were not) indicative of the presence of flews Licensee personnel indicated in response that eddy current examin. ?iens were considered, at the time of installation of the Unit 2 sleeves, to not be applicable to acceptance of the sleeve installation. Rather, the examinations were performed as a baseline for subsequent inservice eddy current examinations. The inspector reviewed the topical report (marked as containing proprietary information) {

that had been approved by NRC staff, " Arkansas Nuclear One Unit 2 Steam f Generator Tube Repair Using Leak Tight Sleeves," dated July 1992, in order to establish governing requirements. Section 3.0, Acceptance Criteria, indicated with respect to Criterior 6 (pertaining to nondestructive examination for levels of detectability requirad to show structura! adequacy) that an eddy current technique had been developed that could detect a 20 percent through-wall defec Section 5.0, Sleeve Examination Programs, stated that a multi-frequency eddy current method would be used to provide a volumetric examination of the pressure boundary,_ including the sleeve, parent tube, and weld. The inspector also noted in Section 5.1.2 that the only defined expectation for ultrasonic examinations was to confirm the presence of satisfactory sleeve to tube weld fusio i

._ _____ _ - ___ - -

I

.

-17-The inspector did not agree with the licensee historical understanding of the expected role of eddy current examination, after review of the wording of the applicable topical report. The inspector also considered the view to not be relevant to the central concern; i.e., acceptance for continued service despite the lack of a demonstrated technical basis for disregarding the presence of eddy current examination weld zone indications. The failure to take prompt corrective action, prior to returning potentially defective sleeved tubes to service, was identified to licensee personnel on August 8,1997, as an apparent violation of Criterion XVI of Appendix B to 10 CFR Part 50 (50 368/9714-03).

During review of Nonconformance Report 2005331-25, which pertained to the sleeve lower joint torque roll exceeding the maximum specified limit in two tubes, the inspector noted that the basis for a "use-as-is" dispositica included a statement (addresning the potential for development of primary water stress corrosion cracking) that a minimum life of more than 1.5 effective full power years was expected. The inspector considered that future examination requirements for these two tubes should have been evaluated as part of the acceptance of a "use-as-is" disposition, based on the potential for a short remaining life. Review of licensee actions in this regard are considered an inspection followup item (50-368/9714-04).

The results of staff review of weld defects in ABB Combustion Engineering designed steam generator tube sleeves were transmitted to the licensee in a letter cated March 14,1997. This letter reflected a commitment by the licensee to perform a 100 percent plus point probe examination of welded sleeves with eddy current indications from the previous inspection, plus an examination of a 20 percent sample of the remaining previously installed welded ABB Combustion engineering sleeves, Conclusions An apparent violation was identified pertaining to the acceptance for continued service, without an appropriate demonstrated technical basis, of sleeved tubes exhibiting weld zone eddy current indications. An inspection followup item was also identified rcgarding the future examination criteria adopted for two sleeved tubes, for which a nonconformance report disposition had identified the potential for early primary water stress corrosion crackin E8 Miscellaneous Engineering issues E (Closed) inspection Followuo item 50-368/9628-03: Review of Laboratory Examination Results for Outaae 2F96-1 Tube Samolet Insoector Followuo The inspector reviewed the laboratory examination results for samples of two Steam Generator A tubes, R16C56 and R70C98, which were documented in ABB

_ _ _ _ _ _ _ _ _ _ _ _ .

.=

,

.

- 18-Combustion Engineering Final Report A-PENG-TR Oll, " Examination of Steam Generator Tubes Removed From ANO 2 in 1997," June 1997. Field motorized rotating pancake coil examination during Outage 2F96-1 identified the presence of single axial indications in these it.bes at the 01H eggcrate support. For Tube R16C56, the estimated through wall depth was 78 percent, the amplitude 8.42 volts, and the length 1.13 inches. For Tube R70C98, the estimated through wall depth was 85 percent, the amplitude 11.14 volts, and the length 1.15 inche The inspector noted from review of metallographic data for Tube R16C56 that approximately 10 other axialintergranular stress corrosion cracks were observed to be present in addition to the main crack. The average depth of the main crack was approximately 55 percent tnrough wall, with a maximum depth of 100 percent through wall. The largest of the other intergranular stress corrosion cracks was about 77 percent through wall. The inspector concluded from review of the metallographic data that the extent of degradation present was significantly under predicted by the Outage 2F961 eddy current examinations. The inspector also noted that significant intergranular attack surrounded the stress corrosion cracks, suggesting that environment and material susceptibility were more significant contributors to the observed degradatior then membrane stres The average depth of the main crack in Tube R70C98 was found by metallographic examination to be approximately 78 percent through wall, with the maximum depth ascertained to be (like Tube R16C56) 100 percent through wall. Approximately eight axialintergranular cracks were found to be present in addition to the main crack, with a maximum depth of 100 percent through wall. The cracking parallel to the main crack in Tube R70C98 was noted to be more severe than the corresponding parallel cracks in Tube R16C56. The inspector noted that, like Tube R16C56, the metallographic examination results for Tube R70C98 indicated that the extent of degradation was significantly under predicted by the eddy current examination The inspector noted the following additional information during review of ABB Combustion Engineering Report A PENG TR-Oll:

• All cracking in the two tubes initiated from the outside surface of the tubes, which is consistent with the degradation observed in previous tube pull

,

sample * Cracking was found to be present only at the points on the tube where the carbon steel eggcrate strips contacted or were closest to the tub * The carbon content of both Tube R16C56 and Tube R70C98 exceeded the 0.05 weight percent maximum of the tubing specification (i.e.,0.059 percent and 0.055 percent, respectively).

.

.

19-

• The microstructure of both tubes was fine grained (ASTM 7-11), with the distril ution of carbides predominantly intragranular and little present on the grairi joundaries. Relatively high yield strength values were measured for the h Des (Tube R16C56,55 ksi; R70C98,53 ksi). The observed micra ;tructure and yield strength properties suggested that the final anneal of the tubes was performed at a low temperature (i.e.,less than 1800 *F).

This inconel 600 microstructure normally results in enhanced susceptibility to intergranular stress corrosion cracking, compared to a coarser grain structure with predominantly intergranular carbides presen * Significant quantities of sulfur, as sulfate, were present on the fracture and outside diameter surfaces (i.e., fracture, up to 4.1 atomic percent; outside diameter, up to 1.6 atomic percent). The inspector concurred with the report assessment that the presence of such sulfur quantities suggested its involvement in the degradation process. It was unclear to the inspector on the intent of a report statement that conclusive evidence was lacking regarding whether the sulfur compounds accelerated the corrosion proces * Lead was also found associated with the fracture and outside diameter surfaces (i.e., fracture, up to 0.9 atomic percent; outside diameter, up to 0.25 atomic percent). The report made the same conclusion for the role of lead, as it did for sulfur compounds, regarding the lack of conclusive evidence regarding whether the lead accelerated the corrosion process. The inspector considered the quantities of lead present to be significantly higher than had been seen on other tube pull samples, and concluded that its presence was a fairly strong indicator of it being a contributor to the degradation process. No insights were developed from discussions with licensee personnel regarding the origins of the lead associated with the cracking in the steam generator tube * Scanning electron microscopy of intergranular cracking in Tube R70C98 showed evidence of pitting or erosion on grain faces at tube depths at least up to mid wall. The grain f acets near the crack tip did not show similar features. The inspector considered that the grain pitting was indicative of a crack that had been in existence for a considerable period.The inspector also viewed the metallurgical information as indicating thi: Nbe through-wall degradation in excess of tube mid wall was necessary at eggerate support locations prior to bobbin coil tube examinations being capable of its detectio * Burst testing of tube axial fiaws, which initiated at the 01H eggcrate support in Tube R16C56, developed a 0.8 inch long crack at the 340 location at a pressure of 2500 psig, which progressed to a 1.30 inch long fishmouth failure at the burst pressure of 3200 psig. The burst failure exceeded the main steam line break accident analysis pressure of 2950 psig, but did not meet the Regulatory Guide 1.121 structural integrity criteria of 4750 psi __-_______ _ __ _ _ _

.. . .. ..

,

i

.

-20-Burst testing of tube axial flaws, which initiated at the 01H eggerate support in Tube R70C98, developed a 0.7 inch long crack at the 190 location at a pressure of 2000 psig. The crack was essentially through wall after pressurization to 2950 psig, with the crack held together by nine ligaments of non-corroded material. The tube burst at 3250 psig, with two axial cracks present, which had a total length of 1.65 inche Qgnglusion_g The micrc structure present in both Tubes R16C56 and R70C98 gi.e., fine grain size and intratranular carbides) was characteristic of that resulting from use of a low final annealing temperature. This type nf microstructure normally results in enhanced susceptibility to intergranular stress corrosion cracking, compared to a coarser grain structure with predominantly intergranular carbides present. Lead and sulfur (as sulfate) were present on the crack surfaces, indicating their involvement in the degradation proces *

Pitting or erosion of grain faces was observed at tube depths at least up to mid wall in Tube R70C98. The grain pitting was indicative of a crack that had been in existence for a considerable perio The metallurgical information indicated that tube through-wall degradation in excess of tube mid-wall was necessary at eggerate support locations prior to bobbin coil tube examinations being capable of its detectio Burst test values of 3200 psig and 3250 psig, respectively, were obtained for Tubes R16C56 and R70C98. These values exceeded the main steam line break accident analysis pressure of 2950 psig, but eu not meet the Regulatory Guide 1.121 structural integrity criteria of 4750 psi E8.2 Licensee Event Report 50-313/97-001: Inconsistencies Between Unit 1 Destructive Examination Results and Eddv Current Deoth Sizina Predictions (for Tube Flaws in Three Pulled Tube Samoles From the Uoper Tube Sheet Reaion) Created the Possibility of Flaws Beinn Present in Inservice Tubes Which Exceeded the Throuah-Wall Limit in Technical Specifica_1. ions The results of inspection followup of this licensee event report are documented in Section E1.1 abov E8.3 LClosedl Licensee Event Reoort 50-368/97-001: Steam Generator Tube Surveillances Were Performed Durina Refuelina Outaae 2R11 (September-November 1995) Which Failed to identifv Existina Tube Flaws for Further Evaluation as Reovired by the Steam Geerator Tube Surveillance Procram This licensee event report pertained to the identification during Outage 2F96-1 (November December 1996) that Refueling Outage 2R11 (September-November i

l

, . _ , . ..

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

_

o

.

-21-1995) analysis of bobbin coil examination results for Steam Generator A Tubes R16C56 and R70C98 had f ailed to identify the presence of distorted support indications, if identified, the surveillance program requirements would have required the performance of a motorized rotating pancake coil examination to f acilitate determination of whether the distorted support indications were indicative of the presence of an axial tube flaw. The circumstances pertaining to Steam Generator Tube R16C56 were evaluated and documented in NRC Inspection Report 50 313;368/96-28. A noncited violation was identified as a result of this evaluation. The inspector concluded that the identification of a second

-

tube with a missed distorted support indication would not have changed the determination to issue a noncited violatio V. Mananement Meetinas X1 Exit Meeting Summary The inspector presented the results of the onsite inspection to members of licensee management at the conclusion of the onsite inspection on June 20,1997. The licensee acknowledged the findings presented. Documents were reviewed during the inspection, which had been identified by ABB-Combustion Engineering as containing proprietary information. No information was included in the inspection report that was considered proprietary. Additional in-office review of the inspection findings was performed subsequent to the onsite inspection, with evaluations continuing until August 7,1997. A second exit meeting was conducted telephonically on August 8,1997, to inform the licensee that three apparent violations were identified during the additional revie '

-

,

.s

..

1 ATTACHMENT 1 SUPPLEMENTAL INFORMATION! -

~

PARTIAL LIST OF PERSONS CO_NTACTED

-

_

- . Licensee C. Anderson, Unit 2 Plant Manager G. Ashley, Supervisor, Licensing M. Cooper, Licensing Specialist

- D. Denton, Director, Support j D. Harrison, Supervisor, Engineering Programs

- R. Hutchinson, Vice President, Nuclear Operations B.-James, Unit 2 Outage Manager i R; Jones, Engineer, Engineering Programs R. Lane, Director, Design Engineering

= D.'Meatheany, Engineer, Engineering Programs W. McKelvy, Chemistry Superintendent ,

D. Mims,' Director, Nuclear Support S. Pyle,- Licensing Specialist .

R. Scheide,' Licensing Specialist j M. Smith, Manager, Engineering Programs C. Zimmerman, Unit 1 Plant Manager NBC

- K. Kennedy, Senior Resident inspector -

S. Burton, Resident inspector J. Melfi, Resident inspector j

.,

' INSPECTION PROCEDURES USED IP 50002 Steam Generators

'

ITEMS OPENED AND CLOSED Ooened ,

'50-313/9714 01 APV Apparent failure to comply with requirements of criteria used -

to qualify bobbin coil eddy current examination sizingi 1 _ technique

_

50-368/9714-0 . APV.:' Apparent failure during Outage 2F961 to remove tubes fro service, which contained flaws that exceeded Technical-

. Specification plugging or repair limit = j

.i r

j

. ?

.

______

e l'

I

.

50-368/9714-03 APV Lack of prompt corrective action in Novembu 199C 7rior to '

retuming potentially defective sleeved tubes to sers ee 50 368/9714 04 IFl Review of examination provisions for two sleeved tubes with identified potential for limited service life before initiation of primary water stress corrosion cracking C191251 50-368/9628-03 IFl Review of laboratory examination results for Outage 2F96-1 tube samples 50 313/97-001 LER inconsistencies between Unit 1 destructive examination results and eddy current depth sizing predictions for tube flaws in upper tube sheet region 50-368/97-001 LER Refueling Outage 2R11 steam generator tube surveillances failed to identify presence in two tubes of distorted support indications, thus precluding further evaluation (motorized rotating pancake coil) required by the surveillance program DOCUMENTS REVIEWFD Procedures / Documents HES-14, " Steam Generator integrity Program Unit 2," Revision 1 HES-26, " Steam Generator Tubo Plug Defect Management," Revision 7 HES-28, "ANO 2 Str.m Generator ECT Data Analysis Guidelines," Revision 6 and Standard Change Notice (s) 01 through 04 HES-29, "ANO-2 Steam Generator ECT Performance Demonstration Guidelines," Revision 3 HES 31, "ANO 2 ECT Data Management and Control," Revision 0 HES 42, " Steam Generator Tube Sleeve Management," Revision 1 ANO-2-OTH-ESP-SGMAN, " Arkansas Nuclear One - Unit 2 Steam Generator Eddy Current Training Manual," Revision 1 Acquisition Technique Specification Sheet ACTS 1, " Bobbin Examination," Revision 1 Acquisition Technique Specification Sheet ACTS 2, "O.115 Pancake, Axial and Cir Directed Coils," Revision 4 i

o

.

3-Acquisition Technique Specification Sheet ACTS 3, "Plus-Point for Welded Sleeves,"

Revision 5 Acquisition Technique Specification Sheet ACTS 4, *

Analysis Technique Specification Sheet ANTS 1, " Bobbin Coil Examination of Steam Generator Tubing," Revision 1 Analysis Technique Specification Sheet ANTS 2, "3 Coil RPC (0.115 Pancake, Axial &

Circ. Coils)," Revision 1 Analysis Technique Specification Sheet ANTS 3, "Plus Point for Kinetic Sleeves,"

Revision 1 Analysis Technique Specification Sheet ANTS 4, " Single Coil O.115 Pancake," Revision 3 5120.501, " Steam Generator integrity Program - Unit 2," Revision 6 5120.506, "C.E. Unit 2 Steam Generator Secondary Side inspection 2E24 A&B,"

Revision O ABB Combustion Engineering Report A-PENG-TR-Oll, " Examination of Steam Generator Tubes Removed From ANO-2 in 1997," June 1997 Procedure STD-400-157, " Visual Examination of Steam Generator Tube Sleeve and Sleeve Plug Welds," Revision O Procedure 00000-NSS 062, " Procedure for Ultrasonic Examination of Steam Generator Tube to Sleeve Upper Welds, Revision 10 Procedure STD 400-151, "Linearity Verification of Ultrasonic Instrumentation Used for the Examination of Sleeved Tube Welds, Revision 0 Procedure STD 400-158, " Technical Operating Procedure for the Transition Zone Sleeve Rolling Tool and Controls for Steam Generator Tubes With 3/3 Inch O.D. 043 Inch or .048 Inch Wall Thickness Using the Computer Controlled Sleeving System," Revision 0 Procedure STD-500 020, " Technical Operating Procedure for the Post-Weld Heat Treatment and Heater Dryout of Above the Tubesheet (ATS) Welds Using Resistance Heating for Steam Generator Tubes With .048 inch Wall Thickness," approved July 31, 1992 Traveler 2005331-001, " Installing Tube Sleeves in the Steam Generator Tubesheet Expansion Transition Zone Using the Genesis 2000 Control System," Revision 0

_

.

C

.

.

Welding Procedure Specification GTAA-43.43.992, Revision 0, and its supporting procedure qualification record, GTAA-43.43 01503, Revision 1

C e,

.

o ATTACHMENT 2 Entergy Operations Arkansas Nuclear One Unit Two Position Paper on Steam Generator " Missed" indications in the Fall of 1996, ANO 2 shutdown due to a steam generator tube leak. During the subsequent inspection, tubes were identified as having indications previously " missed" during the eddy current testing (ECT) analysis process in the prior outage. The definition of " missed" is difficult and presents a dilemma when one must make a reasonable determination of the condition of a tube in previous times. The tube which caused the t aker outage was not due to a " missed" call and thus the leaker was not preventable. An a 1 was initiated, and a non-cited violation was issued as a result of the " missed" calls, Because of the experience of the leaker outage, several changes were implemented as part of the ANO corrective actions from the root cause analysis which was performed. These changes included:

Entergy Level 111 Review Team Repair of all Eggcrate Cracks RPC of allindications Tertiary review using Computerized Data Screening Review of a sample of NDD calls by Resolution Analysts or Entergy Review Team Enhanced Analyst Site Performance Testing These changes resulted in significant improvements to the ECT Program and a high quality inspection in 2R12, our most recent outage. The changes were made by ANO as a result of the corrective actions from 2F96 based on the performance of the inspection in 2R1 In 2R12, none of the indications found f ailed the Regulatory Guide 1.121 requirements. All structural and leakage requirements were met. A review of the 2R12 indications disclosed a number of examples where there is some evidence of the flaw existing in the previous inspection. Review of flaw history is routinely performed to evaluate the growth rate of the degradation. The growth ra;c is used to make a "forwardlooking" evaluation, or operational assessment, of the unit and determine the safe operating interval. It is common for flaws to have not been called in the previous inspection, especially when the

" signal noise" ratio is small. It is a very qualitative assessment that determines if a given flaw was " missed," or if it was reasonable to assume the ECT signal was too difficult to have been expected to be calle The 2F96 indications of interest are for the most part " questionable" as to whether or not they should have been called. There were none similar to the tubes pulled in 2F96, which were more obvious misses. The process used in 2F96 was much improved over 2R1 There were, however, no real program changes made to increase analyst sensitivity, other than the knowledge of the cause of the leaker. Shortly after the inspection began, the ECT results for the leaking tube were reported to the analysts, and insight gained from that information improved the quality of the analysi The flaws detected in 2R12 were typically small amplitude, and posed no structural or leakage concerns. This was demonstrated by in situ pressure testing of the largest flaws, with all exceeding the RG 1.121 requirements. The indications detected in 2R12 that have been classed as " missed" are really not indicative of a " missed" indication. The changes

e e

.

e implemented in 2R12 provided for a drastic improvement in the quality of the inspectio That is evident by the results of the independent review of 10% of all NDD calls, which indicated that the analysts were making conservative calls, and were not " Missing" indications. This, along with the other changes implemented, yielded a high quality inspection. These changes were made following the review of the 2F96 circumstances, and this is the first opportunity we have had to implement them. A non cited violation was issued for the 2R11 " missed" indications identified in 2F96, and 2R12 was the first inspection af ter implementing corrective action During 2F96, two tubes were removed for analysis that contained axial cracks at the eggerates. The analysis provided information on the structural integrity of this type of flaw. The flaws burst at pressures above accident conditions but below Reg. Guide limits. It was at this time that ANO decided to remove all confirmed crack-like indications in the eggcrate region and make programmatic changes to enhance the probability of detection during 2R12. Tubes tested during 2R12 indicated the structural b'.egrity of the tubing is not challenged. This was proven by in situ testing the largest 11aws. A total of six tubes were tested and all had zero leakage at peak accident pressures and all were pressurized to 3 times the normal delta pressure limit specified in Reg. Guide 1.121. It is therefore believed that none of the 2R12 indications in question would have prevented the SGs from fulfilling their safety function, nor challenge the steam generator structural integrit A critical review of the ANO-2 program occurred after 2F96 and a comprehensive corrective action plan was implemented prior to 2R12. This corrective action plan was presented to the NRC during the March 1997 meeting. The benefits from those corrective actions were apparent in the number of indications identified, in summary, based on the following information:

A non-cited violation was issued following 2F96 Corrective actions were developed prior to 2R12 Tube pull destructive examination results were available just prior to 2R12 Comprehensive corrective actions administered during 2R12 Largest flaws passed Reg. Guide 1.121 with zero leakage at peak accident pressures ANO does not believe the recent review of 2F96 data identified tubes that were improperly dispositioned and should have been plugged during 2F96. Also, no additional corrective actions would be necessary if a violation was issued for the condition identified during 2R12.