ML20211D236

Steam Generator Run Time Analysis Cycle 9

ML20211D236
Person / Time
Site:	San Onofre
Issue date:	09/25/1997
From:	SOUTHERN CALIFORNIA EDISON CO.
To:
Shared Package
ML20211D231	List: ML20211D229 ML20211D236 ML20211D252
References
NUDOCS 9709290081
Download: ML20211D236 (37)

Text

e---m- 5 mwa_e s .-J__m -- 6 L ,m----.--,-m, a-,us.A-- ei.,.,m,e- - a- _ am _ w.

,LO'5 SOUTHERN CALIFORNIA EDISON COMPANY SAN ONOFRE UNIT 2 STEAM GENERATOR RUN TIME ANALYSIS CYCLE 9

???"28n8n38361 PDR-G

. r J ,

i TABLE OF CONTENTS I Table of Contents . . . . . . . . . . . . . . . . . . . . . . .-. . . . . Page 1 1.0 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . Page 2 2.0 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . Page 3 Table'1 - SONGS Unit 2 - Tubes Plugged Cycle 9 Outage - 1997 . . Page 6 3.0 Degradation Assessment . . . . . . . . . . . . . . . . . . . . . Page 8 4.0 In-Situ Pressure Testing ................... Page 10 Table 2 - San Onofre Unit 2 In-situ Pressure Test Results . . . Page 11 5.0 Tube Removal, Laboratory Burst Testing, and Supplemental Eddy Current Performance Demonstration ................... Page 12 Table 3.- Laboratory Leak and Burst Test Results ....... Page 14 Figure 1 - Probability of Detection (POD) Curve ........ Page 15 Figure 2 - Profile of Axial Degradation ............ Page 17 Figure 3 - Profile of Axial Degradation ............ Page 1( ,

6.0 Condition Monitoring Assessment ................ Pagr 19 7.0 Probabilistic Operational Assessment Summary ......... Page 21 .

Table 4 - San Onofre Unit 2 Cycle 9 - Probability of Burst .. Page 24 Table 5 - San Onofre Unit 2 Cycle 9 - Accident Leakage Evaluation Page 25 ,

8.0 Inspection and Operational Enhancements . . . . . . . . . . . . Page 25 9.0 Conclusions . . . . . . . . . . .-. . . . . . . . . . . . . . . Page 26 ,

l 10.0 References .......................... Page 27 Appendix 1 - History of ECT Inspections and Tube Plugging . . . . . . . Page 28 Appendix 2 - Supplemental Performance Demonstration ......... Page 21 Appendix 3 - NRC Generic Letter 95-05 Evaluation ........... Page 31 Appendix 4.- APTECH Engineering Services Report . . . . . . . . . . . . Page 36 9

.+

-1

i

1.0 INTRODUCTION

The purpose of this report is to describe Southern California Edison's (SCE's) condition monitoring and operational assessment of steam generator (SG) tubing at San Onofre Nuclear Generating Station (SONGS) Unit 2. Condition monitoring is applicable to the previous Cycle 8 of operation, and operational assessment applies to the current Cycle 9 of operation. Steam generator tube performance has been evaluated based on the restilts of extensive Eddy Current (ECT) examinations, pressure testing, and analyses described in this report. e r

This run time analysis uses guidance and criteria in NRC Regulatory Guide 1.121 " Bases for Plugging Degraded Pressurized Water Reactor Steam Generator Tubes,"Rev0, August 1976(Reference 1),NRCDraftRegulatoryGuideX.XX

' Steam Generator Tube Integrity," November 1996 (Reference 2), and applicable portionsofGenericLetter(GL)95-05"VoltageBasedRepairCriteriafor Westinghouse Steam Generator Tubes Affected by Outside Diameter Stress i CorrosionCracking," August 3,1995(Reference 3). The objective of the run time analysis is t . determine the duration which the San Onofre Unit 2 SGs can be safely operated within the criteria in References 1, 2 and 3.

The next planned inspection of tubes for degradation at the top of the tubesheet will occur during the next refueling outage. The structural integrity performant;. criteria and the accident leakage criteria of the above described regulatory documents can be met throughout a full cycle of operation without this particular inspection during a mid-cycle outage.

The next planned inspection of the full length of all tubes with a bobbin probe will occur during a mid-cycle outage. The purpose of this is to project meeting of the structural integrity criteria for axially oriented primary water stress corrosion cracking at dented eggcrate intersections. The probability of burst for this degradation is projected to be within the structural integrity criteria until the mid-cycle inspection. The mid-cycle outage provides an opportunity to identify tubes effected by this degradation and remove then: from service. With this action taken, the probability of burst for this degradation is projected to be within the structural integrity criteria at the e'id of the cycle.

The next planned inspection of the full length of all tubes with a bobbin probe during the mid-cycle outige will also address axial cracking in tubing freespans and axial cracking ct undented eggcrate supports. Affected tubes will be identified and removed from service. This will provide added 2

assurance of safe operation. However, this degradation is considered to be l within the critef a that would not indicate a need for a mid-cycle inspection.  ;

The primary applicability of GL 95 05 is to voltage based repair criteria for Westinghouse SGs. San Onofre does not apply this repair criteria to it's Asea BrownBoveri/CombustionEngineering(ABB/CE)SGs. Nonetheless, the operational assessment employs the guidance provided in GL 95-05 and a -

physically based, statistical model. 7 A mid-cycle inspection outage is planned so that the inspection interval for the bobbin probe will be less than 0.92 effective full power years (EFPY).

The conditional probability of tube burst for all degradation mechanisms, ,

1.45x10, is lower than the draf t RG X.XX criteria of 5x10. The maximum ,

-conditional probability of tube burst for an individual mechanism is 0.96x10

and is lower than the GL 95-05 and draf t RG X.XX criteria of 1x10.

Postulated accident leakage is acceptable.

Therefore, Cycle 9 operation for the San Onofre Unit 2 steam generators with a mid cycle outage is acceptable. Acceptability has been determined using guidance contained in RG 1.121, GL 95-05, and draft RG X.XX.

2.0 BACKGROUND

2.1 Commitment On February 6,1997 Southern California Edison (SCE) provided a Special Report to the U.S. Nuclear Regulatory Commission regarding inservice inspection of steam generator tubes that was completed at San Onofre Unit 2 on January 24,1997(Reference 4). Technical Specifications 5.7.2.C required this report. in this report SCE made the following commitment:

"The initial inspection results indicate that an additional inspection should be performed after about 12 months of operation. Edison will submit a final assessment of the enclosed inspection results within 180 days of the end of the Unit 2 Cycle 9 refueling outage. Because Unit 2 Cycle 9 is scheduled for about 24 months of operation, Edison is planning a mid-cycle outage to complete this inspection."

i The purpose of this evaluation is to fulfill the connitment to submit a final assessment of the steam generator inspection results for the Urlt 2 Cycle 9 refueling outage. Unit 2 entered Mode 2 at the end of the refueling outage on March 29, 1997. Thus, this assessment is due to be submitted to the NRC by 40tember 25, 1997.

3

2.2 Steam Generator Description The San Onofre Unit 2 design includes two recirculating ASEA Brown Boveri Combustion Engineering (CE) designed and manufactured Steam Generators (SG) ,

which are vertical U-tube and shell heat exchangers. _The SGs are designed and fabricated per ASME Code, Section 111 through Summer 1971 Addenda. Each SG 3 contains 9,350 Alloy 600, high temperature, mill annealed tubes which are 3/4inchODandhaveanominalwallthicknessof0.048 inches. The tubes are explosively expanded into the tubesheet for the entire tubesheet thickness.

The tubes are arranged in rows, with all tubes in a given row having the same length. The rows are staggered, forming a triangular pitch arrangement. The  !

shorter tubes, which have 180' bends, are at the center of the tube bundle in ,

the first 18 rows. All subsequent rows have double 90' bends. The vertical  :

tube lengths are supported by seven full diameter eggerates (lattice bars),

and one to three partial eggerates for the longer tubes. The bends and horizontal lengths are supported by batwings and vertical lattice supports, respectively. ,

2.3 History of Tube Degradation Two tube degradation mechanisms were identified during the first refueling outage. One of these mechanisms was related to inadequate heat treatment of portions of certain tubes. The inadeouate heat treatment could be identified by eddy current bobbin testing; thus, the full length of all (100%) of the tubes was tested. The identified tubes were removed from service. The other degradation mechanism was wear of the tubing at tube supports (batwings and verticallatticesupports). The degradation mechanism of tubing wear has continued to be active throughout the operating lifetime of this unit, in 1989 tube denting and cracking was identified in a single tube. This mechanism was related to tie rod corrosion. This mechanism is possible on a small number of tubes. This mechanism is managed by focused inspections each refueling outage, in 1993 the following tubing degradation mechanisms were identified on the inlet (hot leg) side of the tubes:

o circumferentially oriented cracking 't the top of the tubesheet.

• axially ~ oriented cracking in the vicinity of the top of the .

tubesheet.

• axially oriented cracking at eggcrate tube supports.

These degradation nicchanisms continue to occur.

-4

. . , + -r,-,w- ~-<w,--,,--wrwe,-,v-m--w- e--- - - - ~ - -, , -,,n,-,,,.,,,.,,,...,,.g.,., ,,.,-.,ve-

,-,.--u,. - , . ,,q.e -, .-.. ,

7 In 1997 the following tubing degradation meuhanisms were identified on the  !

inlet (hotleg)sideofthetubes:

e axially oriented cracking in freespan regions (not at tube supports).

e axially oriented cracking at dented eggcrate tube supports.

Also, in 1997, systematic trending of denting indicated that the number of dents in tubing at hot leg .upports was increasing.

The new degradation mechanisms identified in 1997 and the continuance of axial cracking at eggerate supports prompted this detailed assessment.

2.4 Water Chemistry The following actions have been taken in the last year to improve the secondary side water chemistry endronment for the steam generator tubing.

These actions have been reviewed by a panel of industry experts for application at San Onofre. The panel concurs with the following measures:

• Chemical Cleaning of the Entire Tube Bundle (full Bundle)

• Addition of an inhibitor (Titanium dioxide) for Inter Granular Attack /StressCorrosionCracking(IGA / SCC)immediatelyafterthe ,

Chemical Cleaning, for maximum crevice penetration potential

• Usc of Efbanolamine to reduce deposition on the tubing

• Planning for Plant Modifications for Boric Acid Addition to the secondary side to help reduce denting and stress corrosion cracking of tubing San Onofre Unit 2 has operated with all volatile chemistry since startup in August 1983 and accumulated approximately 10.1 effective full power years of operation through the end of Cycle 8 operation on November 30, 1996. The condensers are seawater cooled. The feedwater system has a state-of-the-art, full-flow, deep bed, condensate polisher system that was placed in service during the second cycle of operation. The condensate polisher system is operated continuously for maximum protection of the steam generators from cooling water in-leakage, improvements in secondary chemistry control have paralleled industry-developments with implementation of the EPRI PWR Secondary WaterChemistryGuidelines(Reference 5)andsubsequentrevisions, i

l 5

L i a

2.5 Steam Generator Tube Inspections A history of ECT inspections and tube plugging is provided in Appendix 1.

The following time line is provided to summarize significant historical innovations in San Onofre Unit 2 Steam Generator Tube Inspections.

80BR13 PROBE 1987 - Full Length sampling increased from 3% to 20's 1995 - Implemented examination of the full Length of 100% of the tubing ROTATING PROBE 1984 - During the first refueling outage, used as a characterization tool for indications of significance 1989 - More wide usage as a characterization tool, and used in all tubes adjacent to tie rods 1993 - 100% examination of hot leg top-of-tubesheet expansion transitions with "3-coil" probe, including the 0.115 inch diameter pancake coil 1997 - Use of Plus-Point

• coil probe for rotating probe applications.

The most recent SG inspection was completed at the end of Cycle 8 in early 1997. The detailed scope and the results of that inspection were previously summarized in a Special Report to the NRC (Reference 4). The tube plugging table from that Special Report is repeated here for completeness (Table 1}.

2.6 Conservative Tube Plugging The only tubing degradation mechanism for which SCE uses an eddy current technique for sizing is mechanically induced wear of tubing at tube supports.

Tubes are plugged for wear based on sizing results. Other imlications of tubing degradation are plugged upon detection, without the use of a sizing technique to justify leaving affected tubes inservice. Tube plugging was performed by FRAMATOME Technologies, Inc. Corrosion resistant, thermally treated alloy 690 material is used for plugging. Plugs are installed mechanically using a qualified rolling process. Welded alloy 690 plugs are used in a few special cases, such as when a tube is removed.

6

0 8 Table 1 - SONGS UNIT 2 - Tubes Plugged CYCLE 9 OUTAGE - 1997 Steam Generator Indication Orientation and Location E-088 E-089-Mixed mode (circumferential and axial oriented 0 1 -

indications) that intersect in the tube and are near the expansion transition at the hot leg top-of-tubesheet Both circumferential and axial oriented indications that 5 3 do not intersect and are near the expansion transition at the hot leg top-of-tubesheet Circumferential1y oriented indications near the 88 49 expansion transition at the hot leg top-of-tubesheet Axlally oriented indications in the upper bundle 0 3 (typically above 7H), and not associated with a tube support (Freespan) 0 1 As above, but located at a tube support Axially oriented indications in the lower bundle 5 21 (typically below elevation 07H), and not assulated with a tube support (Freespan)

Axially oriented indications at a hot leg eggcrate tube 2 6 support location Axially oriented indications at a dented hot leg 4 4 eggcrate tube support location Axially oriented indications near the expansion 58 50 transition at the hot leg top-of-tubesheet Preventive plugging of-tubes adjacent to a tic rod 6 0 Indications of wear at a tube support location 2 7 Volumetric indication at a miscellaneous location in a 3 4 tube Prevenilve plugging based on presence of a foreign 0 3 object Highest growth rate dent 1 0 Miscellaneous preventiva plugging 4 2 Total 178 154 7

2.7 Operational Leakage Monitoring Primary-to-secondary leakage monitoring at San Onofre has been reviewed against available guidance to ensure that leakage monitoring measures are effective at detecting leakage and reducing the potential for tube rupture.

Detection of low level leakage is accomplished primarily by chemical sampling and the condenser air ejector radiation alarm. Nitrogen-16 monitoring is available to assist in diagnosis of low-level leakage. Action levels and leakage limits in San Onofre Abnormal Operating Instructions are consistent with the "EPRI PWR Primary-to-Secondary leak Guidelines" (Reference 6).

The potential for tube rupture, detection, measurement, and assessment of rapidly increasing leakage is ? quired by San Onofre procedures. San Onofre procedures require unit shutdown for rates of change of steam generator tube leakage equal to 60 gpd in any one hour period. Operator responses to leakage and rates of change of leakage are specified in San Gnofre procedures. These procedures and responses have been reviewed against available guidance and historical leakage events.

3.0 DEGRADATION ASSESSMENT 3.1 Circumferential Indications Circumferential stress corrosion cracking initiated at both the outside and inside surface of the SG tubing was first detected at the tube expansion transition in the hot legs of the San Onofre Unit 2 SGs during the 1993 inspection. Rotating coil inspection has been routinely performed for this region since that time. SCE data analysis guidelines have been cantinually updated to incorporate industry experience and evolving .echnology. No indications of this type were identified in a similar examination of a sample of tubes on the cold leg side of the steam generators. SCE has used EPRI sizing techniques that are available and appropriate for San Onofre to determine that indications have met the structural criteria of RG 1.121, 3.2 Free Span Axial Indications Subsequent to a tube rupture in March of 1993 at Palo Verde Unit 2 (CE System 80) it was determined that upper bt.ndle deposits may act as a precursor to frea span axial cracking. Thermal hydraulic evaluations of steam generators, completed since the tube rupture indicate that certain upper bunole regions have higher potential for deposit accumulation. Axial indications have recently been reported in the free span region of ABB/CE 1' designed units with more operation time than San Onofre Unit 2.

8 I

The San Onofre Unit 2 and 3 SG tube bundle regions most susc ytible to deposit accumulation were estimated. This was done by review of industry experience t.nd computer modeling of thermal hudraulic conditions. (Reference 7) ,

in 1997 SCE implemented inspection plan changes to include rott. ting Plus Point" probe inspection of 20% of the tubes in selected regions in both San Onofre Unit 2 SGs. _ The region was defined based on thermal-hydraulic co W derations. No indications were found in SG 88. In response to axial 4 inogations in SG 89, the sampling program of 609 tubn as expanded to an approximate total of 1409 tubes. 1he results in the e ..',ded sample indicated no need for further sample expansion in the selected region of the tube bundles.

Axial cracking was identified in freespan parts of the tubing during the most recent examination. Indications were detected in 26 tubes in the 100% bobbin probe examination of the tubing. The location of this cracking could not be readily correlated with results of the thermal hydraulic evaluation of tube bundle regions susceptible to deposit accumulation.

3.3 Axial Indications at Eggerates and in the Sludge Pile SCE has utilized a bobbin coil technique for detection of outside-diameter-initiated,stresscorrosioncracking(ODSCC)indicationsattheselocations.

The technique used is EPRI qualified in accordance with Appendix H of the EPRI PWRSteamGeneratorExaminationGuidelines(Reference 6). SCE has only used the detection portian of this qualification. SCE has conservatively " plugged on detection," and the has not used the sizing portion of this qualification to leave tubes in service. However, the sizing portion of this qualification provides one tool in estimation of the structural integrity significance of such indications per RG 1.121.

Further, SCE has enhanced detection capatilitics in the sludge pile by performing rotating Plus-Point

• coil inspection of all tubing in the vicinity of hot leg expansion transitions. ,

3.4 Axial Indications at DenteJ Hot leg Eggerate Support Intersections SCE uses the full length bobbin exam of 100% of the tubes to identify dents at tube support intersections and dings in tubing freespan. SCE uses the voltaga normalization technique in widest industry usage, consistent with EPRI Guidelines and GL 95-05 (all bobbin frequency channel voltages normalized to 4 volts on the prime frequency for 20' 4 drill holes). For the 1997 inspection SCE implemented a practice of inspection with the rotating Plus-Point

• coil of all- hot leg dents and dings- that are greater than 5 volts by bobbin.

9 p

w g, ,,.,- - -

3.5 Tthing Wear San Onofre Unit 2 experiences tubing wear that is typical for this particular designofABB/CESGs. A technique that is qualified in accordance with Appendix'H of the EPRI PWR Steam Generator Examinatior Guidelines is used-for detection and sizing of indications.

4.0 IN-SITU PRESSURE TESTING In-situ pressure testing was conducted during the most recent inspection to verify that RG 1.121 structural margins were maintained. Twenty-one in-situ pressure tests were completed using a full tube hydrostatic or defect-specific hydrostatic test method. The full tube method was performed on most of these tubes. The defect specific method was used for 2 tubes in which the eddy current indications being tested were circumferential or axial indications at tube expansion transitions. A summary table of locations tested and tsults was-previously~ provided in a special report to the NRC (Reference 4). This information has been supplemented to include information on the eddy current indications at each location and is included here as Table 2.

4.1 Criteria Several criteria were used in screening candidate eddy current indications for in-situ pressure testing. The selection process included taview of bobbin and rotating probe data for estimated maximum through wall depth, maximum voltage, and length. Percent degraded area (PDA) was also calculated for candidate circumferential defects to assure that indications tested are bounding.

Finally, lead ECT data analyst recommendations-were considered.

In-situ testing target pressures include an adjustment for testing at ambient temperature rather than the maximum design temperature of 650*F. For circumferential indications, pressures are also adjusted to account for assumed, locked tube support conditions. Acceptance criteria are those specified by regulatory guidance.

4.2 Results One of the in-situ pressure tested tubes in Steam Generator 89 (Row 13 Column

~

117) experienced axial crachog at the sixth hot leg eggcrate support. The cddy current bobbin probe amplitude of the the crack was 4.49 volts and the space associated dt:a was 23 volts. The axial cracking at this dented location exhibited leakage of 0.24 gallons per minute (gpm) at a pressure comparable to a Main Stw4 lire Break (MSLB) accident, and exceeded the leakage capabilities of tM eqJpment at a pressure of 3150 psi. The Conditten lionitoring Assesstaent Section of this report addresses the significance of these results.

All-of the other twenty in-situ pressure -tested tubes demonstrated leaktightness and structural integrity at all test pressures.

wy, ,, ..-g . g . - p _y y7..y .q 7

Tabla 2 - San Onofre Unit 2 I;-Sita Pressure Test Results TueC Ano t007 Ctaatsi 1 Non=Arree 1:-SITv ttST stsan - anteat te=P, v_ ,

8E5104 - TUBE IsF0eMATIOe PLUS PoleT DATA 909815 Ceuments slMIP SP4P GNISBDro . &

80PO M5La Post M5ts .->

a0W COL Locatism tength ' volts PDA l Orientatten tetts Press m Steam Generator 88' EEGCRATE 44 56 07W - 0.07 1.14 / 4A ID SA! DaT/15v Nighest Plus Pt velts 0 .O m/A 4744 9 , m,.

044 + 28.94 0.95 C2 t 'A OD 5A1 100 0 Nig*est Plus Pt & Bet 4te I

11 169 048 + 4.53 0.26 OJ j AA 00 SA1 0.85 valtages; 2nd utghest 0 0 s/A 4746 044 + 5.01 0.2 0.~ NA 00 SA! 0.78 Bettte 4TW 52 10 754 + 0.10 1 0.33 52 CD SCI N00 Righest PDA 0 0 N/A 5292 US 5 E / T5m + 0.68 0.21 0.09 WA 00 5A1 0.28 M @ est plus Pt. Settle.

4746 p

53 111 754 + 1.41 0.39 2.15 nA 00 5AI 2.04 & Pav ake ts1tages I ' 'O Steam Generator 89 13 117 064 - 0.33 0.9 4.86 eA  ?? 5AI 4.49 2n0 Vettage: Cente4 0 l 0.24 0.051 #/A M RATE 02H - 0.30 0.66 0.7 MA 00 SA! 1.08 Caiculation 0 0 4746 92 128 5/A 94 110 064 + 0.22 0.15 0.27 nA 00 SAI EDO Calculatten 0 0 h/A 4746 04N + 21.97 23 0.29 NA 00 MAI 0.33 14 166 05A + 13.87 34 0.31 4A 00 MAI 0.28 toegest Length 8 0 W/A 4746 06W + 9.21 11 0.38 at 00 mat 0.51 05M + 9.3 9.5 0.83 4A 00 MAI 0.4 15 15 Calculatten 0 0 m/A 4746 05H + 33.74 2.3 0.36 mA GD MAI 0.34 m + 5.64 3.7 0.56 RA M MA! 0.4 23 15 Calculatten 0 0 nlA 4746 05M + 11.26 0.85 0.44 NA OD MAI 0.6

- 31 17 OEM + 6.4 0.3 0.23 nA 00 SAI 0.58 Calculatten 0 0 m/A 4744 06M + 17.79 16 0.23 aA 00 MAI 0.47 82 24 07N + 8.63 4.9 0.13 WA 00 MAI 0.44 Nighest Plus Pt Septh 0 0 n/A 4746 M1044M 08W + 9.03 4.3 0.16 tA 00 MAI 0.75 92 34 05H + 25.11 17 0.45 #A OD MAI 0.51 Calculatten 0 0 m/A 4?46 06H + 17.71 0.4 0.33 kA 0D SAI 0.24 106 38 06a + 21.01 1.4 0.21 nA DD SAI 0.16 Ntghest Plus Pt vettage 0 0 N/A 4746 06M + 27.82 4.3 0.33 NA 0D SA! 0.13 064 + 8.32 14 1.08 24 OD MAI 0.73 110 42 06d + 4.90 4.9 0.69 RA 00 MAI 0.9 Nighest Pav ate Volts 0 0 N/A 4746 06N + 16.78 0.4 0.18 nA 00 SA! 0.74 pighest Bobbin Weltage II 06M + 20.23 4.5 0.57 aA 00 SA! 1.34 & Nighest %TM 0 t/A 4746 122 112 Den + 6.21 0.67 0.35 mA 00 SA! #DO U-Bene 0 n/A 4746 .:

DBd + 13.27 1.2 0.32 mA 00 SAI uD0  ;

UPPER But0LE 122 120 08N + 5.17 0.59 0.27 nA 00 SA! 250 U-Bend e  : 0 N/A 4746 DSM + 7.34 0.89 0.26 hA 00 5A1 300 I D!d + 9.67 2.4 0.44 NA 0D SA! NDO _.

40 56 T5M + 1.31 0.54 1.4 NA OD SAI  !.12 Nighest Pancake volts 0 0 s/A 4746 MBE5H W / 59 79 TSN + 0.13 0.51 0.48 19 OD SCI NDD Mine4 Antal & Circ 0 0 5292 SN mE n/A T54 + 0.22 0.12 0.26 aA CD 5AI RDO 94 86 T5M + 1.16 0.23 0.3 NA 00 SAI WD0 Calculatten 0 l 0 N/A 4746

'5.0 TUBE REMOVAL, LABORATORY BURST TESTING, AND SUPPLEMENTAL EDDY CURRENT PERFORMANCE DEMONSTRATION Three tubes were removed from San Onofre Unit 2 SG 89 during the 1997 outage.

Tubes were removed-to characterize degradation mechanisms, demonstrate eddy current bobbin detection capabilities for axial cracking in freespan regions, determinepressuretestleak/burstbehavior,anddeterminethep'aysical-and metallurgical characteristics of the tubes. The select'a criteria for the removed tubes considered the worst case indications at tube supports. The -

degradation mechanism-specific details are provided below. One of the removed tubes (Row 94 Column 32) had indications of two types of degradation-mechanisms at different areas of the tube.

Detailed examination'of the physical, chemical, and metallurgical properties of these three tubes led to the conclusion that they had microstructures that were not consistent with industry expectations for High Temperature Mill Annealed (HTMA) Alloy 600 tubing. The requirements of ASME SB-163 were supplemented by ABB/CE specification requirements for a maximum yield strength

( f 55,000 psi (which required a relatively high annealing temperature). The microstructeral characterization indicated a microstructure with relatively small grains (ASTM grain sizes 10 to 12). rather than the expected grain sizes of ASTM 5 to 6. The carbon content was higher than anticipated, but within the tubing manufacturing specifications. The microstructure of the pulled tubes revealed random carbide precipitation within the grains and on the grain boundaries. It should be noted that a microstructure acceptance criteria, in terms of grain size and carbide precipitation, was not contained in the tubing specification when it was developed.

5.1 Axial Cracking in Freespan Regions of Tubing Two of these tubes were removed to gain information on this degradation mechanism. One of the removed tubes (Row 16 Column 164) was affected by this degradation mechanism from below the fourth hot leg eggcrate support to above the sixth hot leg support. Another removed tube (Row 94 Column 32) was affected by this degradation mechanism from above the fifth hot leg eggcrate support to above the sixth hot leg support.

Laboratory Leakage and Burst Testing were performed on removed tubing sections with eddy current indications of this mechanism. The same testing was done on additional adjacent removed sections to demonstrate bounding of the affected region. Table 3 provides the results of this testing. In summary, for this degradation mechanism there was no leakage at pressures comparable to accident

conditions, and tubes significantly exceeded the structural integrity margins addressed in RG 1.121. In fact, tubing structural margin was demonstrated similar to that of unflawed tubing, i

This freespan axial cracking consists of relatively small (typically less than )

35% through-wall) outside-diameter-initiated short axial microcracks that are I distributed around the tube. This characteristic gives this freespan axial l cracking a different visual appearance than that observed in other affected tubesremovedfromABB/CEdesignedsteamgenerators.  !

A " supplemental performance demonstration," described in draf t RG X.XX, was performed to demonstrate the eddy current bobbin probe detection capabilities for this freespan axial cracking. This demonstration used eddy current indications (nondestructive examination data) with corresponding metallurgical laboratory data (destructive examination data) in sections of pulled tubes.

The most significant result of this demonstration, a Probability of Detection (POD) Curve, is illustrated in Figure 1. This figure shows a reference P0D which is the technique 1;mit and individual team performances which reflect human factors considerations. More detail is provided in Appendix 2, which is a summary of the purposes and protocol of this demonstration. P0D results are considered in the Probabilistic Operational Assessment.

Indications of the secondary side chemical environment were limited due to the secondary side cleaning of the steam generators just prior to tube removal.

The degradation environment may have been formed wiihin deposits on the tubing surface. Heat transfer performance degradation and the amount of material removed during chemical cleaning corroborated the existence of significant deposits. Deposits have been suspected of creating an environment that allows deleterious chemical species to concentrate near the tube in a manner contributing to 0DSCC.

L Table 3 - Laboratory Leak and Burst Test Results Unit 2 Steam Generator E-089 Tube Leak and Burst Test Results GPM0 Max Pressure Connent s 0rientation Row Col Location MSLB Achieved (Note 1)- Pressure (Note 2) 7H-6H FSA 0 12,600 6H FSA 0 11,900 SH FSA 0 11,800 16 164 5H-4H FSA 0 12,500 4H FSA 0 13,200 3H NDD 0 13,450 2H 3H NDD 0 13,400 6H FSA 0 13,050 6H-SH FSA 0 13,200 5H NDD 0 13,200 5H-4H NDD 0 13,800 94 32 5,300 4H EA 0 (6 Volt Dent)

No leakage till endpoint &

None at 5 min. hold at 4800 psig. ,

3H NDD 0 13,700 7H EA 1.35 3.515 7H-6H NDD 0 10,650 Endpoints occurred in the area welded to the tube for sealing with the test gear 61 111 6H NDD 0 11,900 (not in the tubing).

Testing complete with the objective favorably niet.

5H NDD 0 13,500 4H NDD 0 14,000 Notes:

1. Abbreviations:

FSA Axial cracking in the freespan EA Axial cracking at an eggcrate NDD No detectable degradation

2. Lab testing pressure that is equivalent to SONGS 2 normal operation differential pressure multiplied by a safety factor of 3 is 4746 psi.

1 Figure 1 Probability of Detection (P0D) Curve - Freespan Axial Cracking 580 Bobbin Probe I- g , e. e- ,

d I '

l ,'P REF

• 0,8 TeamA /' e'

-+-

,g '

T amBx 0.6 TeamC ,'0,'

4 g

0.4 j TeamD

+ , J <*

TeamE e,'

~* 0.2 A}/

l 0 60 80 10t 0 20 40 X

Metallographic Depth, %TW t

5.2 Axial Cracking of Tubing at a Dented Eggerate Support One tube was removed to gain information on this degradation mechanism. One ofthethreeremovedtubes(Row 94 Column 32)wasaffectedbyinside-diameter-initiated axial cracking at the fourth hot leg eggcrate support. This cracking had been detected by the eddy current bobbin probe, and was confirmed and further characterized with a rotating probe with a Plus-Point

• coil and a high-frequency' shielded pancake coil.

The eddy current bobbin probe amplitude of the associated dent was 6 volts.

The dent was dimensionally measured in the laboratory with a dial calipers as an 11 mil dent (0.011 inches calculated as the average freespan tubing outside diameter minus the minimum-dented tubing outside diameter).

Thelength/depthprofileofthisdegradation,determinedbyscanningelectron microscopy, is illustrated in figure 2.

Laboratory Leakage and Burst Testing was performed at this location. The results at this location are shcwn in Table 3. The axial cracking at this dented location did not exhibit leakage at a pressure comparable to MSLB accident conditions, and met the structural integrity margins addressed in RG 1.121.

5.3 Axial Cracking of Tubing at an Undented Eggerate Support One tube was removed to gain information on this degradation mechanism. One of the three removed tubes (Row 61 Column 111) was affected by inside-diameter-initiated axial cracking at the seventh hot leg eggcrate support. This cracking had been detected by the eddy current bobbin probe, and was confirmed and further characterized with a rotating probe with a Plus-Point

• coil and a high-frequency shielded pancake coil.

The length / depth profile of this degradation, determined by scanning electron microscope, is illustrated in Figure 3. The eddy current bobbin probe indicated there was no denting. Dimensional measurements in the laboratory corroborated that there was no detectable denting.

Laboratory leakage and burst testing results are shown in Table 3. These results and associated calculations indicate that the axial cracking at this location exhibited leakage of 1.35 gpm at a pressure comparable to MSLB accident _ conditions. Calculation of this leakage was necessary because it exceeded the capabilities of test equipment at a pressure of 2100 psi.

Structural integrity at MSLB accident conditions was demonstrated. Maximum achievable test pressure was 3515 psi, which is below the target pressure of 4746 psi. The pressure of 4746 psi would have demonstrated structural integrity at a pressure simulating normal operating differential pressure multiplied by a factor of 3.

~

~ $

g SONGS-2 Tube R94L32 Section 7 Burst Opening Depth Profile

' f 100 %

95% l t

90 % -

pg - - - -

" 85 % -- -

E j ~7-0= m E 80% '

-- ---}-- - - - -j -- --

I 3 , j b 2 75% i 'T- i 4

~

a w ,

m

• m 70%

i

( I L,

y h 65% -

f'

~

g e 60%  ? --

[* 55 %

e $ 50% ,f, C*

o. '

= 45% o

• e 3 40% - >

-E.

. .c o  ;

e, E

o

'35%

r o 30% '

.E e-n & '

Q

• g 25 % -- - -

5 $ 20% -

o E 15%

0 10% - -

o o 5%

I 0%

cI , 20 202 20.4 20.6 20.8 21 21 2 21.4 21.6 21.8 22 EO Distance From Bottom of Section (Inches)

S=

c=

9 9

g SONGS-2 Tube RG1L111 Section 9 Burst Opening Depth Profile Sa 100 % -

l .

g g 95% -

{- b

{- 90 % >-

3- E SS% t

--} L_ - - - -

] 80% -

,, p - - - - - -

< {--

' -- I . -

--- -- f- , ,- - - - -- -

g 2 75 % --

{

?-- }

E N 70% "  ! -

I l E I 7 o 65% j { -j-'

] ' I e c 60 %

7: 7 55%- n h'  !

4, l

o 2 } 4, e o 50 % -

7 o O ,,

=

a 45% <

.o 3 j 40% ,,

5 E 35%

8 e c 30%

~ >

{ g 25%

• 20%

e

• Q- 15%

o O

10% -

2 5%

0%-

"w 12.5 12.7 12.9 13.1 13.3 13.5 13.7 13.9 14.1 14.3 14.5 5U Distance From Bottom of Section (Inches) 9

. .c.

6.0 CONDITION MONITORING ASSESSMENT The purpose of this " backward looking" assessment is to confirm that adequate tube integrity has been maintained.during the operating period before the inspection. Qualified " depth-sizing" techniques were only available for two degradation mechanisms (insidediameter-initiated,stresscorrosioncracking .

(IDSCC) at expansion transitions, and mechanically induced wear at tube supports). All available eddy current information was considered for other degradation mechanisms.

6.1 Methodology The first step was review of eddy current depth-sizing information for mechanically induced wear at tube supports. All these indications met criteria for structural and leakage integrity.

The next step for other degradation mechanisms was a review of bobbin and rotating probe data for estimated maximum through wall depth, maximum voltage, and length. Percent degraded area (PDA) was also calculated for applicable circumferential indications. Finally, lead ECT data analyst recommendations were considered. The output of this review consisted of a general ranking of themostsignificantindications,basedonthedegradationmechanism/ location.

This general ranking was then used as a leading input into the tube selection processes for in-situ pressure testing and tube removal. Leakage and burst testing was used to obtain condition monitoring information on both the most significant eddy current indications and also some less significant indications to demonstrate that the results were bounding for other indications. Results of in-situ pressure testing and laboratory leak / burst testing (on removed tubes) were consistent with expectations formed during the tube selection processes. This is considered an indication of the effectiveness of the selection process.

6.2 Test Results In-situpressuretestingidentifiedonedentedeggcrate/tubingintersection for which structural integrity (at a pressure simulating normal operating differential pressure multiplied by a factor of 3) could not be demonstrated due to leakage that exceeded the makeup capabilities of the test equipment.

Structural integrity at MSLB accident conditions was demonstrated. The axial cracking at this location exhibited leakage of 0.24 gpm at a pressure comparable to MSLB accident conditions.

LaboratoryLeak/BurstTestingidentifiedoneundentedeggcrate/ tubing intersection for which structural integrity (at a pressure simulating normal operating differential pressure multiplied by a factor of 3) was not demonstrated. This intersection lost structural integrity at 3515 psi, which was below the target pressure of _4746 psi. Structural integrity at MSLB i-l-_.-_:-. - - r

'l accident conditions was demonstrated. The axial cracking at this location exhibited leakage of 1.35 gpm at a pressure comparable to MSLB accident conditions.

The inspection and operational enhancements that address these test results are provided in Section 8.0, 6.3 Analysis The criteria for Condition Monitoring Assessment postulated accident leakage performance are that the calculated potential primary-to-secondary leak rate during limiting postulated events should:

(1) not exceed the total charging pump capacity of the primary coolant system (2) be such that the offsite radiological dose consequences do not exceed 10CFR Part 100 guidelines and radiological consequences to control room personnel are in accordance with General Design Criteria 19.

Theadditiveleakageof1.59gpmforonesteamgenerator(atMSLBpressure) for the 2 tubes exhibiting leakage during leak testing is t.onsidered to be the best available indicator of postulated MSLB accident leakage at the end of Cycle 8, just prior to inspection. Leakage solely in these 2 tubes was consistent with expectations formed during the selection of tubes for in-situ pressure testing and tube removal for laboratory leak and burst testing. The other tubes selected for these two leak testing processes did not leak. These other non-leaking tubes had been selected based on eddy current characteristics to demonstrate the leak integrity of the remainder of lesser magnitude indications that were not leak tested. There would be 0 gallons per minute primary-to-secondary leakage the other steam generator under MSLB accident conditions.

This leakage of 1.59 gpm (at MSLB pressure) is a very small fraction of the total charging pump capacity of the San Onofre Unit 2 primary system. The pre-trip MSLB analysis, with a concurrent single failure, was analyzed for the maximum break size outside of containment in keeping with NUREG-0800 Standard Review Plan (SRP) 3.6 and Branch Technical Positions MEB 3-1 and ASB 3-1. The assumed primary-to-secondary leak rate was 1.59 gpm to the unaffected steam generator and 0.0 gpm to the affected steam generator for the antire duration of the event. The primary-to-secondary leakage activity profile was consistent with 4.41% failed fuel and the maximum allowed Technical Specification primary coolant steady state activity limits (consistent with NUREG-0800 SRP 15.1.5 guidance for this event with fuel failures). The dose analysis assumed a 30 minute release from the main steam line break location to the environment with a steam generator iodine Partition Factor (PF) of 1.0

for botn steam generators. Initial secondary activity levels were assumed to be at the maximum allowed by Technical Specification limits and were assumed to be released out of the break during the first 30 minutes of the MSLB. No Emergency feedwater Actuation System (EFAS) actuation prior to 30 minutes was credited in the dose analysis. At the end of 30 minutes the operators were assurned to isolate the steam generators and initiate a controlled plant cooldown via the unaffected steam generator atmospheric dump value (ADV). The releases for the remainder of the event are from the unaffected steam generator ADV with an assumed steam generator iodine PF of 0.01. The noble gas PF is 1.0 for the entire event duration. The calculation of the resultant doses, presented below, were performed using RG 1.109 dose conversion factors.

PRE-TRIP MSLB-0UTSIDE CONTAINMENT DOSES D0st LOCATION REGULATORY D0st CRITERIA CAlfilLATED 00sts (Rem)

(Rem)

ontrol Room (event duration dose)

Thyroid Inhalation 30 23.3 Whole Body Ganna Imersion + shine 5 4.2 Beta-skin leersion 30 16.1

AB (2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> dose)

Thyroid Inhaletion 300 82.6 Whole Body Ganna Imersion 25 0.8 Beta-skin innersion no dose criterion 0.3 PZ (event duration dose)

Thyroid Inhalation 300 2.8 WMle Body Ganna innersion 25 < 0.1 Beta-skin innersion no dose criterion < 0.l _

The results of Condition Monitoring indicated that the significance of these results to future operation could best be assessed by a "Probabilistic Operational Assessment."

7.0 PROBABILISTIC OPERATIONAL ASSESSMENT

SUMMARY

The purpose of this " forward looking" assessment is to demonstrate reasonable assurance that tube integrity performance criteria will be met throughout the period prior to the next scheduled inspection. Since a significant amount of guidance for this type of assessment is provided in GL 95-05, an appendix to this assessment has been prepared that reformats this assessment to correlate with GL 95-05. This is attached as Appendiv 3. .

The significance of corrosion eegradation to the performance of steam generator tubing at San Onofre Unit 2 was evaluated in a Probabilistic Operational Assessrent, Probabilistic methods were applied to make projections throughout an assumed operating period of the structural and leakage integrity of the steam generator tubing, under postulated accident conditians(MSLB). A detailed report of this assessment is provided as Appendix 4. However, a brief overview of it is provided below.

7.1 Degradation Modes Testing of steam generator tubing at San Onofre Unit 2 has indicated corrosion degrariation. Eddy current inspection data and pulled tube examinations have been tools in characterizing the degradation. Five modes of corrosion were considered in this assessment:

(1) Circumferential degradation at the top of the tubesheet (2) Axial degradation at the top of the tubeshect (3) Axial freespan degradation (4) Axial 00 SCC /IGAatundentedeggcrateintersections (5) Axial PWSCC at dented eggcrate intersections 7.2 Metnods The basic calculational technique employed in the Probabilistic Operational Assessment is one of simulating the processes of crack initiation, crack growth, leakage, and detection via eddy current inspection. Monte Carlo simulation methods also provide an approach for accounting for the various sources of uncertainty.

Several probabilistic run time models were employed. Modeling included two operational periods, because there are 2 distinctly different types of inspections. It was necessary to determine the most appropriate operational period between inspections during the next cycle for these 2 different inspections:

(a) Inspection of the vicinity of expansion transitions at the top of the tubesheet is the principal detection tool for circumferential and axial degradation in this location. Results indicate that a full cycle of operation is an appropriate interval between these inspections.

. . l l

l l

(b) Inspection of the full ' length of the tubing with a standard differential bobbin probe is the principal detection tool for axial degradation in the remainder of the tubing. Results indicate that inspection during a mid-cycle outage is appropriate for these inspections.

7.3 Structural Integrity Assessment Description The Probabilistic Operational Assessment provides a projection of the structural integrity of the steam generator tubing. This projection is conrnonly referred to as " conditional probability of burst" (POB). It is generally defined us the probability that the burst pressures associated with one or more indications of degradation will be less than the maximum pressure differential across the tubing associated with a postulated MSLB that-is assumed to occur at the end of an assumed period of operation, just prior to a tubing inspection.

7.3.1 Criteria for Structural Integrity Performance The conditional P0B criteria that is applicable to any one degradation mechanism is 1 x 10. The conditional P0B criteria that is applicable to the total conditional P0B for all degradation mechanisms is 5 x 10.

7.3.2 Structural Integrity Assessment Results The dominant contributor to the conditional probability of burst is axial degradation at dented eggcrate supports. This is largely a function of the eddy current probability of detection (P00).

The Operational Assessment results for conditional P0B are shown in Table 4 and are well within the above P0B criteria.

7.4 Accident Leakage Assessment Description The Probabilistic Operational Assissment also provides a projection of the leakage integrity of the steam generator tubing. The basis of this projection is a probabilistic calculation for each degradation mechanism of its postulatedaccidentcondition(MSLB)leakratefortheendofanoperating period, just prior to inspection. Modeling projects maximum crack depths for this point in time, identifies those that have reached the full tube thickness (through wall), and finally models their corresponding leakege. Monte Carlo methods provide a way to account for uncertainties, and accordingly, the calculated leakage is an upper 95-percent probability at an upper 95-percent confidence bound.

Table 4-- San Onofre Unit 2 Cycle 9 - Probability of Burst Degradation Mechanism Projected Duration Probability of Burst (EFPY) at Postulated MSLB. (95%

Confidence)

Circumferential 2.0 0.0005 ODSCC/PWSCC at Expansion Transitions Axial ODSCC at Expansion 2.0 0.0033 Transitions .

Axial DDSCC at Undented 0.92 0.0008 Eggcrate Intersections Axial PWSCC at Dented- 0.92 0.0096 Eggerate Intersections Freespan Axial ODECC 0.92 0.0003 7.4.1 Criteria for Accident Leakage Assessment The criteria are that calculated potential primary-to-secondary leak rate during limiting postulated events should:

(1) not exceed the total charging pump capacity of the primary coolant system (2) be such that the offsite radiological dose consequences do not exceed 10CFR Part 100 guidelines and radiological consequences to control room personnel are in accordance with General Design Criteria 19.

7.4.2 Accident Leakage Assessment Results The projected MSLB accident leak rates are shown in Table 5.

t

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

4 4

Table 5 - San Onofre Unit 2 Cycle 9 - Accident Leakage Evaluation

-Degradation Mechanism Projected Duration 95/95 Leak Rate at (EFPY) Postulated MSLB (GPM at 600*F Circumferential 2.0 0.38 ODSCC/PWSCC at Expansion Trant'tions Axial 00 SCC at Expansion 2.0 0.009 Transitions Axial 00S00 at Undented 0.92 0 i Eggcrate Intersections Axial PWSCC at Dented 0.92 0.0044 Eggcrate Intersections Freespan Axial ODSCC 0.92 0 The maximum total projected 95%/95% probability / confidence leak rate is less

'than 0.40 gallons per minute at the end of a full cycle. Note that this total was computed for a cycle that includes a mid-cycle inspection outage. This projected-leak rate is a very small fraction.of the total charging pump capacity of the San Onofre Unit 2 primary system and is bounded by the primary-to-secondary leak rate assumption of 0.5 gallons per minute per steam generatorusedintheMSLBlicensingevent,asstatedintheSanOnofre2/3 Updated Final Safety Analysis Report (UFSAR) Section 15.1.3a.

8.0 INSPECTION AND OPERATIONAL ENHANCEMENTS Significantly shortening the operational period between inspections to address axial cracking in the tubing freespans and axial cracking of tubing at supports will provide assurance that the requirements of RG 1.121 will tue met throughout the present operating cycle.

Inspection practices at San Onofre have been routinely updated to incorporate industry experience and recommendations. One such example is the use of a rotating probe at hot leg dented eggerate tube support intersections for improved P0D for indications. Further, thorough in-situ pressure testing has been impitmented. These actions will serve to decrease the potential for a tube failing to meet criteria in regulatory guidance.

Three tubes-have been removed from.a steam generator and characterized in the laboratory.- The purposes of this was to characterize degradation mechanisms,.

demonstrate eddy current bobbin detection capabilities for axial cracking in

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

freespanregions,_determinepressuretestleakage/burstbehavior,and _

determine the physical and metallurgical characteristics of the tubes. ,

-Actions have been taken_to. improve the secondary side water _ chemistry environment for the steam generator tubing. These actions have been reviewed

-by a panel of i_ndustry experts for application at San Onofre. The expert panel concurs with these measures. These actions are: 1) Chemical cleaning of the entire tube bundle (Full Bundle), 2) Addition of an inhibitor (titanium q dioxide) for IGA / SCC immediately- after the chemical cleaning, for maximum >

crevice penetration potential, 3) Use of Ethanolamine, and 4) Planning for plant modifications for boric acid addition in the secondary side to help '

reduce denting of tube supports and stress corrosion cracking of tubing.-

Plant operators have procedur% consistent with EPRI guidelines to detect and 1

irespond to changes in steam generator primary to secondary leakage.

Specifically, they have guidance for shutdown of the unit prior to a significant leak or-tube rupture, should tube degradation exceed expected values.

State of the art probabilistic.models have been developed to demonstrate that a planned mid-cycle inspection outage for the steam generators will maintain the safety margins in regulatory guidance. A probabilistic leakage model has been developed to assess the end of inspection interval leakage that would result from postulated accident conditions. The leakage model demonstrates that the postulated leakage for Cycle 9 is bounded by the primary-to-secondary leakage assumption of 0.5 gallons per minute per steam generator used in the MSLB licensing event, UFSAR 15.1.3a.

9.0 CONCLUSION

S It is SCE's conclusion that tha San Onofre Unit 2 steam generators will fully support safe operation for the pt -iod until a mid-cycle inspection outage, and then further throughout the ful' Cycle 9 operating period.

The following enhancements are appropriate to address the iadications of tube ,

degradation that were found at San Onofre Unit 2 following Cycle 8 operation.

Significantly shortening the interval to the next inspection (a-midcycleinspectionoutage)

Improvements in inspection practices l In-situ pressure testing Tube ~ removal and-testing Improvements to the secondary side environment Tube leakage monitoring review Modeling and analysis

This report demonstrates that the operation, inspection, and repair program described herein constitutes a conservative approach which ensures that adequate structural ar.d leakage integrity is maintained for normal operations, transients, and postulated accident conditions for the period until a mid-cycle inspection outage, and then further throughout the full Cycle 9 operating period for the San Onofre Unit 2 steam generators.

10.0 REFERENCES

1. Regulatory Guide 1.121, " Bases for Plugging Degraded Pressurized Water Reactor Steam Generator Tubes," Rev 0, August 19;'6.

2. NRC Draft Regulatory Guide X.XX " Steam Generator Tube Integrity,"

Distributed November 1996.

3. NRC Generic Letter 95-05, Voltage Based Repair Criteria for Westinghouse Steam Generator Tubes Affected by Outside Diameter Stress Corrosion Cracking," August 3, 1995.

4. Letter from Southern California Edison to the US Nuclear Regulatory Commission "Special Report: " Inservice Inspection of Steam Generator Tubes, San Onofre Nuclear Generating Station, Unit 2," Dated February 6, 1997.

5. EPRI Guidelines, TR-106589-VI Rev. 4 190G, 'PWR Steam Generator Tube Examination Guidelines," dated June 1996 and the Final Draf t of Revision 5.

6. EPRI Guidelines, TR-104788 " Primary to Secondary Leak Guidelines;" dated May 1995.

7. ABB-CE Report, " Tube Bundle Thermal-Hydraulic Analysis to Assist the Development of Eddy Current Test Plans for the Songs Unit 2 Steam Generators," A-SONGS-9419-lll8, REV. 00; dated October 15, 1996.

8. EPRI Guidelines, TR-102134-R4, Revision 4 "PWR Secondary Water Chemistry Guidelines," dated November 1996.

-2 7-

Appendix 1 - History of ECT Inspections and Tube Pluggir.g SAN ONOFRE UNIT 2 STEAM GENERATOR EDDY CURkENT & TUBE PLUGGING HISTORY Exam DOBBIN ROTATING PROBE TUBES RESULTS & COM' DENTS Date EXAM EXAM PLUGGED mfg 100% -- 21 Preservice Shop Plugs 1984 <1% -- 1 Leal Outage, Improper Annealing 1985 100% -- 330 Batwing Wear Improper Annealing, Pulled 2 tubes 1986 SG 88: 6% -- l' Batwing Wear 1987 6% -- 142 Batwing Wear 1989 23% Approx. 84 HL Expan. 62 Batwing Wear Trans. (Tie Rod Related) Tie Rod Denting 1991 23% < 84 HL Expan. Trans. (Tie 41 Batwing Wear Rod Related) Tie Rod Denting 1993 66% 100% HL Expan. Trans. 32 Circ SCC - HL Expan. Trans.

Approx. 668 tubes for Freespan Axial Check Approx. 184 tubes for Bobbin follow up 1995 100% 100% HL Expan. Trans. 45 Circ SCC - HL Expan. Trans.

7% CL Expan. Tran:.

Approx 160 tubes for Bobbin Follow up 1997 100% 100% HL Expan. Trans. 332 Circ SCC - HL Expan. Trans.

7% CL Expan. Trans. Axial SCC - Sludge Pile & Supports &

20% Row 1&2 U-bends Freespan Approx. 2018 tubes for Pulled 3 tubes Freespan Axial Check 1916 HL Dents & Dings Approx 304 tubes for Bobbi,n Follow up Each = 9350 PLUG MARGIN -

TOTAL TUBES -

1000 Tubes each S/G TOTAL PLUGS -

S/G88= 47/ (5.1%)

S/G 89 = 545 (5.8%)

-28 1-- --T

APPENDIX 2 - SUPPLEMENTAL PERFORMANCE DEMONSTRATION OF TNE EDDY CURRENT BOBBIN

~ PROBE-FOR FREESPAN AXIAL CRACKING PURPOSE The purposes of this-demonstration were to:

1. Quantitatively assess flaw detection performance (using pulled tube metallographyandfractography"groundtruth")todemonstratethe reliability of.this inspection at San Onofre Units 2 and 3.

2.ProvideProbabilityofDetection(POD)performanceinputfortheCondition Monitoring and Operational Assessments.

PROTOCOL This is a summary of the pr-otocol of this demonstration.

A. APTECH Engineering, Inc. and FRAMATOME Technologies, Inc. provided consultation and review of the Destructive Examination Planning for the San Onofre Unit 2 Pulled Tubes. This maximized the applicability of fractography and metallography results to this effort.

B. The following tasks were performed by APTECH:

• Assembly of a written protocol for the demonstration

• Assembly of orientation and practice materials

• Selection of flawed and unflawed data for both practice and ;he demonstration

• Briefing of participants on the purpose and protocol

• Assisting participants during analysis of practice data, as rsquired.

• Proctoring the demonstration

• Reporting results for P0D input into assessments C. FRAMATOME Technologies, Inc. provided Qualified Data Analysts who served as

" Primary" data analysts.

D. ANATEC International, Inc. provided Qualified Data Analysts who served as

" Secondary" data analysts.

E. Use of these suppliers, in these roles, replicated historical actual practice. All analysts had completed San Onofre site specific performance demonstration and participated in an actual inspection in 1997.

F. The same data collection (or acquisition) technique was used for all data in the performance demonstration.

-29

. ( .

G. Data analysis was performed as a " blind" test. Tube identification numbers were changed-to enhance this aspect.

H.-Only eddy current indications that were quantified by metallography or

.fractography, after removal from the steam generator, were used as the

" grading unit flaws" for evaluating P00.

I. The data contained a sufficient number of flawed and unflawed grading units to permit POD to be evaluated at an appropriate level of confidence. The number of grading units exceeded that recommended in the EPRI PWR Steam Generator Examination Guidelines for personnel and technique qualification.

J. The latest revision of the existing San Onofre Unit 2 and 3 data analysis guidelines was used. This guideline specifically delineates analyst responsibilities and their roles within the production analysis and resolution analysis process.

K. Primary and Secondary Analysis was done in separate rooms. Communication between teams was not allowed, and analysts were not allowed to re-visit a tube after result: were reported, L.-Twenty data analysts participated in this demonstration. Five different analysis teams were used to demonstrate the repeatability of this technique. E:cli of the five analysis teams consisted of:

• a primary " production" analyst

• a secondary " production" analyst

• 'a primary " resolution" analyst

• a secondary " resolution" analyst

.n .

APPENDIX 3 - NRC GENERIC LETTER 95-05 EVALUATION NRCGenericLetter(GL)95-05, Voltage-GasedRepairCriteriaforWestinghouse

. Steam Generator Tubes Affected by Outside Diameter Stress Corrosion Cracking, was issued to give guidance to licensees who may wish to implement alternate steam generator tube repair criteria. Although San Onofre Unit 2 does not have Westinghouse steam-generators, does not use voltage-based repair criteria, and is not requesting a license amendment to implement such an alternate steam generator tube repair criteria, this document is used as guidance to evaluate steam generator tube performance and predict plant operating capabilities as discussed in the Probabilistic Operational Assessment of Appendix 4.

To generally follow the guidance identified in GL 95-05, this evaluation used the format and criteria delineated in the generic letter. These criteria and the corresponding San Onofre Unit 2 applicable response are provided below (Note that GL Attachment 1 identified in the following criteria refers to Attachment 1 of GL 95-05):

Criterion 1) " Implementation of the applicability requirements discussed in Section 1 of GL Attachment 1. The applicability requirements ensure that the repair criteria are applied only to those intersections for which the voltage-based repair criteria were developed."

The only tubing degradation mechanism for which SCE uses an eddy current technique for sizing is mechanically induced wear of tubing at tube supports.

Other tubes which have degradation detected by eddy current testing are plugged upon detection, without the use of a sizing technique to justify leaving affected tubes inservice.

A " supplemental performance demonstration" was performed to demonstrate the-eddy current bobbin probe detection capabilities for freespan axial cracking. '

" Graded

• data in this demonstration consisted of eddy current indications (nondestructive examination data) with corresponding metallurgical laboratory data (destructiveexaminationdata)insectionsofpulledtubes. Appendix 2 provides a summary of the n'ethodology of this demonstration. The Probability of Detection (P0D) results are applied to the conditional probability of burst analyses and integrated leakage assessments as part of the Probabilistic Operational Assessment.

Criterion 2) " Implementation of the inspection guidance discussed in Section 3 of GL Attachment 1. The inspection guidance ensures-that the

. techniques used to_ inspect steam generator tubes are consistent with_the techniques used to develop the voltage-based repair criteria."

As stated in Item 1 a " supplemental performance demonstration" was performed to demonstrate the eddy current bobbin probe detection capabilities for freespan_ axial cracking. This demonstration was consistent with the Draft RG X.XX. " Graded" data in this demonstration consisted of eddy current indications (nondestructive examination data) with corresponding metallurgical laboratory. data (destructive examination data) in sections of pulled tubes.

_TheProbabilityofDetection(P00)resultsareappliedtotheconditional probability of burst analyses and integrated leakage assessments as part of the Probabilistic Operational Assessment.

Criterion 3) " Calculation of leakage according to the guidance discussed in Section 2.b of GL Attachment 1. This calculation, in conjunction with the use of licensing basis assumptions for calculating offsite and control room doses, enables licensees to demonstrate that the applicable limits of 10 CFR Part 100 and GDC 19 continue to be niet. This calculation is performed using the projected E0C voltage distribution for the next cycle of operation. If it is not practical to complete this calculation prior to returning the steam generators to service, the measured E0C voltage distribution can be used (from the previous cycle of operation) as an alternative (refer to Section 2.c of GL Attachment 1) for the purposes of determining whether the reporting criteria of Section 6.a.1 apply."

A Probabilistic Operational Assessment (a " forward looking" assessment) I1as been completed. One output of the analysis for this assessment is calculated leakage rate during a postulated accident condition (MSLB) that is assumed to occur at the end of an assumed period of operation, just prior to a tubing L inspection. Assessment output includes individual leakage rates for specific degradation mechanisms, which are also combined to provide a total calculated .

leakage rate.

Please refer to the discussion in Section 7.4 regarding this aspect of Probabilistic Operational ~ Assessment, including the Assessment Results. When all corrosion'niechanisms are considered and a mid-cycle inspection outage is planned, the projected value of leakage during a postulated MSLB accident remains well within availcble NRC criteria and guidance throughout the cycle.

g .. .-

i Criterion 4) " Calculation of conditional burst probability according to the-guidance discussed in Section 2.a of GL Attachment 1. This is a calculation to assess the voltage distribution for the next

- cycle of operation. The results are compared against a-threshold value. This calculation is performed using the projected voltage distribution for the next cycle of--operation.

If it is not practical to complete this calculation prior to returning the steam generators to service, the measured end of cycle (EOC) voltage distribution can be used (from the previous cycleofoperation)asanalternative(refertoSection2.c) for the purposes of determining whether the reporting criteria r of Section 6.a.3 apply."

The " conditional probability of burst" refers to the probability that the i

burst pressures associated with one or more indications will be less than the maximum pressure differential across the tubing associated with a postulated

-MSLB that is assumed to occur at the end of an assumed period of operation, just prior to a tubing irspection.

I Please refer to the discussion in Section 7.3 regarding this aspect of

. Probabilistic Operational Assessment, including the Assessment Results. When all corrosion mechanisms are considered and a mid-cycle inspection outage is planned, the projected conditional probability of burst remains well within available NRC criteria guidance throughout the cycle.

Criterion 5) " Implementation of the operational leakage monitoring program according to the guidance discussed in Section 5 of GL Attachment 1. The operational leak rate monitoring program is a defense-in-depth measure that provides a means for identifying leaks during operation to enable repair before such leaks result in tube failure."

Primary.to-secondary leakage monitoring at San Onofre has been reviewed

against available guidance to ensure that leakage monitoring measures are effective at detecting leakage and reducing the potential for tube rupture.

Detection of low level leakage is accomplished primarily by chemical sampling and the condenser air ejector radiation alarm. Nitrogen-16 monitoring is available to assist in diagnosis of low-level leakage. Action-levels and leakage limits in San Onofre Abnormal Operating Instructions are consistent with the "EPRI PWR Primary-to-Secondary Leak Guidelines."

The potential'for tube rupture is addressed by San Onofre procedures. San Onofre procedures require unit shutdown for rates of change of steam generator tube leakage equal to 60 gpd in any one hour period. Operator responses to increases in steady-state leakage and rapid rates of change of leakage are

-33

, e:.

specified in.the-San Onofre procedures. These procedures and responses.have

.been reviewed against-available guidance and historical leakage events.

Criterion 6) '" Acquisition of-tube pull data-according to the guidance discussed in Section 4 of GL Attachment 1."

Three tubes were removed (pulled) from San Onofre Unit 2 to increase '

understanding of the following degradation mechanisms:

e Axial Cracking in Freespan Regions of Tubing

• Axial Cracking of Tubing at a-Dented Eggcrate Support e- Axial. Cracking of Tubing at an Undented Eggcrate Support The following objectives, similar to those discussed in Section 4 of GL 05_ Attachment 1, were achieved during nondestructive and destructive-analysis of this tubing:

e Obtain information on the morphology of the degradation e Obtain data nn leakage and structural integrity for Condition Monitoring e Obtain information for comparison with pulled tube information from other similar units, and for comparison with any pulled tube data that may be obtained in the future e Assess inspection capability The selection critcria for those tubes to be removed included the following from Section 4.b of GL 95-05 Attachment 1:

• An emphasis on removing tube intersections with large voltage indications (notably, tubes removed incluged the axially oriented indications with a large eddy current voltage at both a dented and undentedeggcrateintersection) e The removed tubing covered the available range of eddy current voltages Labo_ratory Examination and Testing of the removed tubing included the following activities discussed in Section 4.c of GL 95-05 Attachment 1:

e Removed tubing was subjected to leak and burst tests under simulated MSLB conditions, and 3 times normal operating differential pressure to obtain leakage information at simulated

.MSLB conditions and to confirm the_ orientation of degradation. The 1eak rate and burst data were normalized to reflect the appropriate

~

pressure and temperature assumptions for a postulated MSLB,

s:

Specifically, testing pressure was increased 13% above postulated ,

MSLB pressure to account for the testing being cone at room temperature.

e- Subsequent to burst testing, removed tubing was destructively examincd to obtain information oa degradation morphology. The destructive examination included the techniques of metallography and scanning electron microscope (SEM) fractography to characterize the degradation morphology and to characterize the largest crack-ne'. works with regard to their orientation, length, depth, and ligaments.

Criterion 7) " Reporting of results according to the guidance discussed in Section 6 of GL Attachment 1."

This. report provides similar information. Notably, inspection during a planned mid-cycle outage has been shown by the Probabilistic Operational Acsessment to provide a calculated conditional probability of burst (P08) snat does not exceed 1 x 10 under. postulated accident conditirns (MSLB) for a single degradation mechanism at the end of the operating period.

-3 5-

e, * 's

+

APPENDIX 4 APTECH ENGINEERING SERVICES REPORT

-36 <

-__ - _ _ _ _ _ _ _ _ _ _ -