ML12306A291

Transmittal Email and WCAP-17128-NP, Revision 1, Flaw Evaluation of CE Design RCP Suction and Discharge Nozzle Dissimilar Metal Welds, Phase III Study, May 2010

ML12306A291
Person / Time
Site:	Arkansas Nuclear
Issue date:	11/01/2012
From:	Clark R Entergy Operations
To:	Kalyanam N Plant Licensing Branch IV
Kalyanam N
References
TAC ME7646
Download: ML12306A291 (120)

Text

From: CLARK, ROBERT W To: Kalyanam, Kaly

Subject:

WCAP-17128-NP Date: Thursday, November 01, 2012 4:36:42 PM

Kaly, I sent an electronic copy of the subject report to you and Jay Wallace on July 19, 2012. The subject report was too large to send in one email so I had to break it up into six pieces. I sent three emails to you that transmitted two pieces each of the report. Therefore if you combine the three emails you should have the complete WCAP report.

If you have any questions concerning this message, please let me know.

Thank you Bob

Westinghouse Non-Proprietary Class 3 WCAP-17128-NP May 2010 Revision 1 Flaw Evaluation of CE Design RCP Suction and Discharge Nozzle Dissimilar Metal Welds, Phase III Study

WESTINGHOUSE NON-PROPRIETARY CLASS 3 WCAP-17128-NP Revision 1 Flaw Evaluation of CE Design RCP Suction and Discharge Nozzle Dissimilar Metal Welds, Phase III Study B. Reddy Ganta*

Gordon Z. Hall*

Patrick J. Bachant*, **

MRCDA-I David J. Ayres*

Steam Generator Management Program Matthew H. Kelley*

Decision Analysis Science & Technology Department Brandon F. Good*

MRCDA-2 Warren H. Bamford*

Primary Systems Design and Repair May 2010 Reviewer: David F. Baisley*, **

MRCDA-I Approved: Carl Gimbrone*

Manager, MRCDA-I This work was performed under PWROG Project Number PA-MSC-0525.

• Electronically approved records are authenticated in the electronic document management system.

• • Carl Gimbrone for Patrick J. Bachant and David F. Baisley.

Westinghouse Electric Company LLC P.O. Box 355 Pittsburgh, PA 15230-0355

© 2009 Westinghouse Electric Company LLC All Rights Reserved

ii WESTINGHOUSE NON-PROPRIETARY CLASS 3 LEGAL NOTICE This report was prepared as an account of work performed by Westinghouse Electric Company LLC.

Neither Westinghouse Electric Company LLC, nor any person acting on its behalf:

A. Makes any warranty or representation, expressed or implied, including the warranties of fitness for a particular purpose or merchantability, with respect to the accuracy, completeness, or usefulness of the information contained in this report, or that the use of any information, apparatus, method, or process disclosed in this report may not infringe upon privately owned rights; or B. Assumes any liabilities with respect to the use of, or for damages resulting from the use of, any information, apparatus, method, or process disclosed in this report.

COPYRIGHT NOTICE This report has been prepared by Westinghouse Electric Company LLC and bears a Westinghouse Electric Company copyright notice. Information in this report is the property of and contains copyright material owned by Westinghouse Electric Company LLC and/or its subcontractors and suppliers. It is transmitted to you in confidence and trust, and you agree to treat this document and the material contained therein in strict accordance with the terms and conditions of the agreement under which it was provided to you.

As a participating member of this task, you are permitted to make the number of copies of the information contained in this report that are necessary for your internal use in connection with your implementation of the report results for your plant(s) in your normal conduct of business. Should implementation of this report involve a third party, you are permitted to make the number of copies of the information contained in this report that are necessary for the third partys use in supporting your implementation at your plant(s) in your normal conduct of business if you have received the prior, written consent of Westinghouse Electric Company LLC to transmit this information to a third party or parties. All copies made by you must include the copyright notice in all instances and the proprietary notice if the original was identified as proprietary.

DISTRIBUTION NOTICE This report was prepared for the PWR Owners Group. This Distribution Notice is intended to establish guidance for access to this information. This report (including proprietary and non-proprietary versions) is not to be provided to any individual or organization outside of the PWR Owners Group program participants without prior written approval of the PWR Owners Group Program Management Office.

However, prior written approval is not required for program participants to provide copies of Class 3 Non-Proprietary reports to third parties that are supporting implementation at their plant, and for submittals to the NRC.

REVISION 1 Revision 1 of this report was prepared to update the residual stresses used for the flaw tolerance evaluation of Section 6 of this report. The revised residual stresses are discussed in Section 3.2.

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 iii PWR OWNERS GROUP MEMBER PARTICIPATION LIST* FOR PWROG PROJECT PA-MSC-0525 Participant Utility Member Plant Site(s) Yes No AmerenUE Callaway (W) X American Electric Power D.C. Cook 1 & 2 (W) X Arizona Public Service Palo Verde 1, 2, & 3 (CE) X Constellation Energy Group Calvert Cliffs 1 & 2 (CE) X Constellation Energy Group Ginna (W) X Dominion Connecticut Millstone 2 (CE) X Dominion Connecticut Millstone 3 (W) X Dominion Kewaunee Kewaunee (W) X Dominion VA North Anna 1 & 2; Surry 1 & 2 (W) X Duke Energy Catawba 1 & 2; McGuire 1 & 2 (W); X Oconee 1, 2, & 3 (B&W)

Entergy Palisades (CE) X Entergy Nuclear Northeast Indian Point 2 & 3 (W) X Entergy Operations South Arkansas 2, Waterford 3 (CE) X Entergy Operations South Arkansas 1 (B&W) X Exelon Generation Co. LLC Braidwood 1 & 2; Byron 1 & 2 (W); X TMI 1 (B&W)

FirstEnergy Nuclear Operating Corp. Beaver Valley 1 & 2 (W); X Davis-Besse (B&W)

Florida Power & Light Group St. Lucie 1 & 2 (CE) X Florida Power & Light Group Turkey Point 3 & 4; Seabrook (W) X Florida Power & Light Group Point Beach 1 & 2 (W) X Luminant Power Comanche Peak 1 & 2 (W) X Xcel Energy Prairie Island 1 & 2 (W) X Omaha Public Power District Fort Calhoun (CE) X Pacific Gas & Electric Diablo Canyon 1 & 2 (W) X Progress Energy Robinson 2; Shearon Harris (W); X Crystal River 3 (B&W)

PSEG - Nuclear Salem 1 & 2 (W) X Southern California Edison SONGS 2 & 3 (CE) X South Carolina Electric & Gas V.C. Summer (W) X WCAP-17128-NP May 2010 Revision 1

iv WESTINGHOUSE NON-PROPRIETARY CLASS 3 PWR OWNERS GROUP MEMBER PARTICIPATION LIST* FOR PWROG PROJECT PA-MSC-0525 (cont.)

Participant Utility Member Plant Site(s) Yes No South Texas Project Nuclear Operating Co. South Texas Project 1 & 2 (W) X Southern Nuclear Operating Co. Farley 1 & 2; Vogtle 1 & 2 (W) X Tennessee Valley Authority Sequoyah 1 & 2; Watts Bar (W) X Wolf Creek Nuclear Operating Co. Wolf Creek (W) X

• This is a list of participants in this project as of the date the final deliverable was completed. On occasion, additional members will join a project. Please contact the PWROG Program Management Office to verify participation before sending documents to participants not listed above.

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 v PWR OWNERS GROUP INTERNATIONAL MEMBER PARTICIPATION LIST* FOR PWROG PROJECT PA-MSC-0525 Participant Utility Member Plant Site(s) Yes No British Energy Sizewell B X Electrabel (Belgian Utilities) Doel 1, 2, & 4; Tihange 1 & 3 X Hokkaido Tomari 1 & 2 (MHI) X Japan Atomic Power Company Tsuruga 2 (MHI) X Mihama 1, 2, & 3; Ohi 1, 2, 3, & 4; X Kansai Electric Co., LTD Takahama 1, 2, 3, & 4 (W & MHI)

Korea Hydro & Nuclear Power Corp. Kori 1, 2, 3, & 4; X Yonggwang 1 & 2 (W)

Korea Hydro & Nuclear Power Corp. Yonggwang 3, 4, 5, & 6; X Ulchin 3, 4, 5, & 6 (CE)

Kyushu Genkai 1, 2, 3, & 4; X Sendai 1 & 2 (MHI)

Nuklearna Elecktrarna Krko Krko (W) X Nordostschweizerische Kraftwerke AG Beznau 1 & 2 (W) X (NOK)

Ringhals AB Ringhals 2, 3, & 4 (W) X Spanish Utilities Asco 1 & 2; Vandellos 2; X Almaraz 1 & 2 (W)

Taiwan Power Co. Maanshan 1 & 2 (W) X

Électricité de France 54 Units X

• This is a list of participants in this project as of the date the final deliverable was completed. On occasion, additional members will join a project. Please contact the PWROG Program Management Office to verify participation before sending documents to participants not listed above.

WCAP-17128-NP May 2010 Revision 1

vi WESTINGHOUSE NON-PROPRIETARY CLASS 3 TABLE OF CONTENTS 1 INTRODUCTION ........................................................................................................................1-1 2

SUMMARY

OF RESULTS AND CONCLUSIONS ...................................................................2-1 3 SUCTION AND DISCHARGE NOZZLE LOADING AND RESIDUAL STRESSES..............3-1 3.1 NOZZLE LOADINGS ....................................................................................................3-1 3.2 RESIDUAL STRESSES..................................................................................................3-3 3.3 VALIDATION OF RESIDUAL STRESS MODELING ................................................3-5 4 SURVEY OF OBSTRUCTIONS FOR INSERVICE INSPECTIONS ........................................4-1

4.1 INTRODUCTION

...........................................................................................................4-1 4.2

SUMMARY

OF PLANT OBSTRUCTIONS FOR INSPECTION DATA.....................4-1 4.3 ANALYTICAL ESTIMATION OF OBSTRUCTIONS .................................................4-3 5 JUSTIFICATION FOR DEVIATION FROM INSPECTION COVERAGE REQUIREMENTS:

DEFENSE IN DEPTH..................................................................................................................5-1 5.1 LEAK DETECTION CAPABILITY...............................................................................5-1 5.2 LEAK RATE METHODOLOGY ...................................................................................5-2 5.3 CIRCUMFERENTIAL THROUGH-WALL CRITICAL FLAW SIZES - ASME SECTION XI, APPENDIX C ..........................................................................................5-4 5.3.1 Through-wall Circumferential Flaw Stress Intensity Factor Calculation ...........5-5 5.4 RESULTS ........................................................................................................................5-6 5.5 POTENTIAL FOR BORIC ACID CORROSION DAMAGE ........................................5-7 6 FLAW TOLERANCE PER ASME SECTION XI .......................................................................6-1 6.1 TRANSIENT ANALYSIS FOR THROUGH-WALL AXIAL STRESS DISTRIBUTION FOR USE IN FCG ...........................................................................................................6-1 6.2 PWSCC GROWTH CALCULATIONS..........................................................................6-2 6.3 FATIGUE CRACK GROWTH CALCULATIONS........................................................6-3 6.4 COMBINED PWSCC AND FATIGUE CRACK GROWTH EVALUATION..............6-5 6.5 ASME SECTION XI FLAW TOLERANCE CALCULATIONS...................................6-6 7 ADVANCED PWSCC GROWTH BY FEA ................................................................................7-1 7.1 INITIAL FLAW SIZE .....................................................................................................7-1 7.2 STRESS INTENSITY FACTOR CALCULATION .......................................................7-1 7.3 FINITE ELEMENT FRACTURE MECHANICS MODEL............................................7-2 7.4 BOUNDARY CONDITIONS .........................................................................................7-2 7.5 NOZZLE END AXIAL LOADS .....................................................................................7-3 7.6 PWSCC CRACK GROWTH WITH FEACRACK PROGRAM ....................................7-3 8 PROBABILITY OF CRACKS .....................................................................................................8-1 8.1 PURPOSE........................................................................................................................8-1

8.2 DESCRIPTION

OF CALCULATION METHODOLOGY ............................................8-1 8.3 IMPORTANT ASSUMPTIONS .....................................................................................8-2 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 vii 8.4 RESULTS ........................................................................................................................8-3 9 PROPOSED CODE CHANGE.....................................................................................................9-1 10 REFERENCES ...........................................................................................................................10-1 APPENDIX A: ASME CODE CASE N-770 .......................................................................................... A-1 WCAP-17128-NP May 2010 Revision 1

viii WESTINGHOUSE NON-PROPRIETARY CLASS 3 LIST OF TABLES Table 2-1: Results of Advanced Finite Element Crack Growth Analyses for Circumferential Flaws ......2-3 Table 3-1: Nominal Dimensions Used for Flaw Evaluation.....................................................................3-2 Table 4-1: Summary of Obstructions for Inspection of CE Fleet RCP Nozzles from Drawings..............4-4 Table 4-2: Obstruction Region Estimated based on Enveloped Plant RCP Nozzles ................................4-8 Table 5-1: Summary of Leak Detection Capability, Operating Temperatures, and Inspection Data .....5-11 Table 5-2: Initial Total Flaw Lengths for Various Leak Rates ...............................................................5-12 Table 5-3: Critical Circumferential Flaw Lengths Using the ASME XI Appendix C Approach ...........5-12 Table 7-1: Initial Flaw Dimensions for Three-Dimensional FEA PWSCC Analyses ..............................7-4 Table 8-1: Summary Results Table...........................................................................................................8-3 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 ix LIST OF FIGURES Figure 1-1: Example of Built-in Obstructions for an RCP Discharge Nozzle DM Weld .........................1-2 Figure 2-1: Cumulative Probability of a Flaw with a Depth of 7% of the Wall Thickness ......................2-2 Figure 2-2: Time from Leakage to Critical Circumferential Flaw Length (No Residual Stress Case) for a Through-wall Flaw ..........................................................................................................2-4 Figure 3-1: Finite Element Models of the Three Repair Configurations Modeled for the Pipe to Safe-end Weld.................................................................................................................................3-7 Figure 3-2: Axial Stress Results for All Cases Considered ......................................................................3-8 Figure 3-3: Hoop Stress Results for All Cases Considered ......................................................................3-8 Figure 4-1: Nozzle Circumferential Location Convention .......................................................................4-5 Figure 4-2: Sample Safety Injection Nozzle Uninspectable and Obstruction Dimensions.......................4-6 Figure 4-3: Sample Charging and Spray Nozzles Uninspectable and Obstruction Dimensions...............4-6 Figure 4-4: Sample RTD Nozzle Uninspectable and Obstruction Dimensions ........................................4-7 Figure 5-1: Analytical Predictions of Critical Flow Rates of Steam-Water Mixtures ..............................5-8 Figure 5-2: Critical or Choked Pressure Ratio as a Function of L/D........................................................5-9 Figure 5-3: Idealized Pressure Drop Profile through a Postulated Crack .................................................5-9 Figure 5-4: Circumferential Flaw Geometry ..........................................................................................5-10 Figure 5-5: Time from Leakage to Critical Circumferential Flaw Length (No Residual Stress Case)...5-10 Figure 6-1: Axisymmetric FEA Model for Transient Stress Analysis......................................................6-9 Figure 6-2: Alloy 82/182 Weld Fatigue Crack Growth Rate Properties in a PWR Environment ..........6-10 Figure 6-3: Axial Residual Stresses for RCP Suction and Discharge Nozzles [6] .................................6-11 Figure 6-4: Crack Tip Stress Intensity versus Circumferential Through-wall Crack Length Used for PWSCC Growth Evaluation ..........................................................................................6-12 Figure 6-5: PWSCC Only Growth of Circumferential Through-wall Flaws with Maximum Normal Operating Nozzle Axial Loads for Various Initial Lengths...........................................6-13 Figure 6-6: Maximum and Minimum Through-wall Crack Tip Stress Intensity Factors during a Heatup Transient as a Function of Circumferential Crack Length.............................................6-14 Figure 6-7: Fatigue Only Growth of Circumferential Through-wall Flaws with Maximum Normal Operating Nozzle Axial Loads for Various Initial Lengths...........................................6-15 Figure 6-8: Combined PWSCC and Fatigue Growth of Circumferential Through-wall Flaws with Maximum Normal Operating Nozzle Axial Loads for Various Initial Lengths ............6-16 Figure 6-9: Circumferential ID Surface FCG for Maximum Pipe Load with No Residual Stress .........6-17 Figure 6-10: Circumferential ID Surface FCG for Minimum Pipe Load with No Residual Stress ........6-17 WCAP-17128-NP May 2010 Revision 1

x WESTINGHOUSE NON-PROPRIETARY CLASS 3 Figure 6-11: Circumferential ID Surface FCG for Maximum Pipe Load with Residual Stress, No ID Repair.............................................................................................................................6-18 Figure 6-12: Circumferential ID Surface FCG for Minimum Pipe Load with Residual Stress, No ID Repair.............................................................................................................................6-18 Figure 6-13: Maximum Acceptable Initial Axial Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 10% Inner Diameter Weld Repair, with No PWHT .............................6-19 Figure 6-14: Maximum Acceptable Initial Axial Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 10% Inner Diameter Weld Repair, with PWHT ...................................6-20 Figure 6-15: Maximum Acceptable Initial Axial Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 25% Inner Diameter Weld Repair, with PWHT ...................................6-22 Figure 6-16: Maximum Acceptable Initial Axial Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 50% Inner Diameter Weld Repair, with PWHT6-22 Figure 6-17: Maximum Acceptable Initial Circumferential Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 10% Inner Diameter Weld Repair, with No PWHT...6-23 Figure 6-18: Maximum Acceptable Initial Circumferential Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 10% Inner Diameter Weld Repair, with PWHT.6-24 Figure 6-19: Maximum Acceptable Initial Circumferential Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 25% Inner Diameter Weld Repair, with PWHT.6-25 Figure 6-20: Maximum Acceptable Initial Circumferential Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 50% Inner Diameter Weld Repair, with PWHT.6-26 Figure 7-1: Finite Element Fracture Mechanics Model ............................................................................7-5 Figure 7-2: Crack-face End View of Applied Crack Face Pressures........................................................7-6 Figure 7-3: Applied Free-end Pressures (for Moment plus Axial Force) .................................................7-7 Figure 7-4: Rotated View of Applied Free-end Pressures (for Moment plus Axial Force)......................7-8 Figure 7-5: PWSCC Flaw Growth with Initial ID Surface Flaw of 14% Circumferential, 20% Depth, Case 1...............................................................................................................................7-9 Figure 7-6: SIFs along Crack Front for ID Surface Flaws during PWSCC Growth with Initial Flaw of 14% Circumferential, 20% Depth, Case 1 .....................................................................7-10 Figure 7-7: PWSCC Flaw Growth with Initial ID Surface Flaw of 14% Circumferential, 30% Depth, Case 2.............................................................................................................................7-11 Figure 7-8: SIFs along Crack Front for ID Surface Flaws during PWSCC Growth with Initial Flaw of 14% Circumferential, 30% Depth, Case 2 .....................................................................7-12 Figure 7-9: PWSCC Flaw Growth with Initial Through-wall Flaw of 14% Circumferential, Case 3 ....7-13 Figure 7-10: SIFs along Crack Front during PWSCC Flaw Growth with Initial Through-wall Flaw of 14% Circumferential, Case 3 .........................................................................................7-14 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 xi Figure 7-11: ID Surface PWSCC Flaw Growth with Initial Flaw Size of 23% Circumferential, 20%

Depth, Case 4.................................................................................................................7-15 Figure 7-12: SIFs along Crack Front for ID Surface Flaws during PWSCC Growth with Initial Flaw of 23% Circumferential, 20% Depth, Case 4 .....................................................................7-16 Figure 7-13: ID Surface PWSCC Flaw Growth with Initial Flaw of 23% Circumferential, 30% Depth, Case 5.............................................................................................................................7-17 Figure 7-14: SIFs along Crack Front for ID Surface Flaws during PWSCC Growth with Initial Flaw of 23% Circumferential, 30% Depth, Case 5 .....................................................................7-18 Figure 8-1: All Available Large DM Weld Inspection Results (7% Through-wall) - Case 1..................8-4 Figure 8-2: All Available Large DM Weld Inspection Results (7% Through-wall) - Case 2..................8-4 Figure 8-3: All Available Large DM Weld Inspection Results (7% Through-wall) - Case 3..................8-5 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 1-1 1 INTRODUCTION All Alloy 82/182 butt welds in Combustion Engineering (CE) plants are required to be inspected by the ASME Code Section XI [1]. In addition to this requirement, all of these nozzle regions must be volumetrically inspected by December 2010, in accordance with industry report, MRP-139 [2]. These inspections are required to be carried out using the performance demonstration requirements of Section XI Appendix VIII [1], and Supplement 10 of Appendix VIII. CE plants have a number of dissimilar metal (DM) butt welds in the cold leg. In particular, the large diameter cold leg reactor coolant pump (RCP) suction and discharge nozzle Alloy 82/182 butt welds have an as-built configuration that is not conducive to meeting the 90% inspection coverage requirements of MRP-139 [2] and ASME Code Appendix VIII [1]. In addition, the cast stainless steel material at the safe-end of these nozzles is not addressed by Appendix VIII or Supplement 10, and therefore, would only allow for a one-sided examination.

The large-diameter pump nozzle dissimilar metal welds are exposed to nominal cold leg temperatures of nominally 550°F, and therefore, are less susceptible to primary water stress corrosion cracking (PWSCC) initiation than nozzles in the hot leg. PWSCC initiations, as well as the rate of cracking, and overall susceptibility are a strong function of temperature. Therefore, the probability of crack initiation, as well as the crack growth rate in the cold leg, is significantly less than that of a similar crack in the hot leg.

Required inspection coverage is often difficult to obtain because of additional nozzles which penetrate the pipe and obstruct the weld region. Figure 1-1 illustrates this type of obstruction.

These obstructions could also make mitigation difficult, creating the need for strong technical arguments to demonstrate the integrity of these nozzles, so realistic inspection plans can be carried out.

This document is a follow-up to the initial assessment of the flaw tolerance of these regions, using the rules of ASME Code, Section XI [1] and supersedes it for the RCP nozzle. The calculations in an earlier WCAP [22] present the maximum allowable initial flaw sizes in the DM welds, accounting for PWSCC growth, for the temperatures and loadings of interest and furthermore demonstrating the existence of a favorable flaw tolerance in these regions. This report updates those calculations by considering longer periods of operation and adding the consideration of fatigue crack growth, as well as a more detailed treatment of the residual stresses in the region.

The technical arguments documented in this report can be used for several purposes. First, they support the argument that frequent (every few outages), high-percentage (90%) coverage inspections are not necessary because crack initiation in these regions is highly unlikely. The results presented in this document support less frequent and lower-percentage coverage inspection.

Second a very large margin exists between the size flaw from which detectable leakage can be observed, and the size flaw which could cause the pipe to fail in the region of interest. This margin can be quantified in terms of relative flaw lengths of through-wall flaws or in the time required for a leaking flaw to grow to a critical flaw. This time will then be compared with the action time required for all plants detecting a leak. This action could be triggered as early as one 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> period, or as long as seven days, as a result of a change in the seven day moving average.

This argument provides for defense in depth for this region.

WCAP-17128-NP May 2010 Revision 1

1-2 WESTINGHOUSE NON-PROPRIETARY CLASS 3 This report also provides documented flaw evaluations of the regions of interest, in the case an indication is discovered during a routine ultrasonic testing (UT) examination. Specifically, the work presented herein covers the RCP suction and discharge nozzles for all CE designs with DM welds in the region, for both axial and circumferential flaw orientations. Crack growth due to both fatigue and PWSCC has been considered. These very high flaw tolerance results also support the argument that frequent, high-percentage (90%) coverage inspections are not necessary.

Figure 1-1: Example of Built-in Obstructions for an RCP Discharge Nozzle DM Weld WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 2-1 2

SUMMARY

OF RESULTS AND CONCLUSIONS An extensive series of evaluations have been performed on the Alloy 82/182 dissimilar metal butt welds located at the safe-end regions of the CE designed reactor coolant pump suction and discharge nozzles. These nozzles present inspection coverage challenges, which hinder the likelihood of obtaining the required inspection coverage of MRP-139 [2], and the successor document, ASME Code Case N-770 (see Appendix A). Furthermore, the geometry of the region also contributes to the difficulty of performing standard mitigation techniques.

There are two primary goals of this work:

Provide a technical basis for revision of the inspection requirements for this region, to account for the access limitations. Specifically, changes to ASME Section XI Code Case N-770 are proposed in Section 9 of this report.

Provide flaw evaluations which could be used to allow further operation without repair, in accordance with the rules of Section XI of the ASME code. The results of these flaw tolerance evaluations are provided in Section 6 of the report.

The first step of the project was to document the extent of the obstruction for inspection coverage.

This was done by surveying the plants involved. Results showed obstructions ranged from 11% to 23% of the circumference, but by the time the work described in this WCAP was completed, progress had been made in the inspectability area, and the largest region of obstruction is now 14% of the circumference. Although the inner 33 percent of the pipe may be obstructed over this length, typically inspections do allow some limited examination of the remaining 66% of the thickness.

However, these nozzle regions operate at cold leg temperatures, nominally 550ºF and have a very high resistance to the potential for PWSCC, and a low predicted crack growth rate, if such a flaw were to exist in the region. This leads to the suggestion the required inspection regimen may be too strong for these regions, and the study described here was structured to investigate that possibility and develop a technical basis for proposing changes to inspection requirements consistent with the flaw tolerance of the region. The technical basis for these changes is described in the remainder of this report. The technical basis rests on three complementary findings:

1. The probability of a flaw existing or initiating in this region is very low;

2. There is a significant margin between the size flaw which would leak at a detectable rate, and the size flaw which would cause the pipe to fail. This provides a significant level of defense in depth for the region; and

3. The flaw tolerance of the region, for both axial and circumferential flaws, has been documented as measured by the size flaw which could grow to the ASME Code Section XI [1] allowable flaw size for either flaw type.

Probability of Cracking A compilation of all cracking experienced in these Alloy 82/182 welds was completed, and the information used to develop a Weibull Model of cracking probability as a function of time. The full range of pump operating temperatures was considered for all affected units, and the probability of cracking was extremely low, as seen in Figure 2-1.

WCAP-17128-NP May 2010 Revision 1

2-2 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Defense in Depth All CE designed plants with this pump design were surveyed, and their leakage action levels were obtained. The utilities have all committed to initiate a condition report and follow up on the source of the leak, up to and including containment entry, after identifying a leak or change in the long term trend in an unidentified leakage. Calculations of the time to grow a crack from a through-wall length resulting in the actionable leak rate of 0.1 gpm to the critical length of a through-wall flaw showed that at least 14 years are required, an extremely large margin over the one to seven day action time. These times are shown for a range of leak rates in Figure 2-2.

Figure 2-1: Cumulative Probability of a Flaw with a Depth of 7% of the Wall Thickness Flaw Tolerance A series of calculations were carried out to determine the time required for a postulated surface flaw to reach the ASME Section XI [1] allowable flaw size. Both fatigue crack growth and stress corrosion cracking were considered, and the results are presented in terms of the allowable service time for a range of flaw sizes and shapes. Results show the range of flaws which are acceptable for service periods from two to four years, for example. These results include the required Section XI [1] flaw evaluation margins and are presented for both axial and circumferentially oriented flaws. The revised design-specific residual stresses were found to be lower for circumferential flaws, and higher for axial flaws, than the stresses used in the earlier work. Circumferential flaw results are shown in Figures 6-17 through 6-20, and show that very large flaws can be tolerated in this region. Residual stress effects were found to retard flaw growth for circumferential flaws. The results for axial flaws are shown in Figures 6-13 through 6-

16. While the axial flaw results are not as beneficial as the circumferential flaw results, the WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 2-3 limited length of the flaw causes the aspect ratios to also be limited. The results for a deep axial flaw, which would have an aspect ratio a/l = 0.50, are also very acceptable, as seen in Figures 6-13 through 6-16.

The flaw tolerance work was supplemented with advanced finite element analyses, wherein the postulated flaw was allowed to grow in a natural shape, dictated by the stresses present. These results are shown in Table 2-1 and are based on a postulated surface flaw in the region which cannot be inspected, with length equal to 14% of the circumference. The depth of the flaw was varied from 20% to 30% of the wall, to bracket the range of uninspectable materials. These depths were chosen based on very conservative aspect ratios of 0.04 and 0.03, respectively.

These are significantly larger than the aspect ratio of 0.1667 observed in service experience, and it is highly likely that any flaws deeper than this would have tails which would be detected in the inspected region. Results show that the postulated flaw will remain within the ASME Code [1]

acceptable depth for 7.5 to over 11 years, depending on its depth, and requires between 9.3 and 13 years to reach a through-wall condition. These results do not account for the impact of the stainless steel closure weld, which induces a region of compressive stress in the mid wall region of the pipe and would further retard the crack growth.

Conclusions This work has demonstrated that the pump safe-end to nozzle weld regions have significant margins, and therefore do not require the inspection frequency specified in [28]. The flaw tolerance option similar to that included in [28] has been used to demonstrate this within this report.

The three approaches used to support this conclusion have been consistent in their findings. There is only a very small probability of having a flaw in the cold leg region, and if it existed, the evaluations showed that more than 14 years would be required from the time a leak is discovered to the point when the integrity of the pipe would be challenged. Finally, the flaw tolerance of the weld region was examined using both classical and advanced finite element analysis techniques.

It was shown that a circumferential flaw postulated in the region would require between 7.5 and 11 years to reach the ASME Code [1] limiting depth of 75% of the wall thickness. This supplementary analysis discussed in detail in Section 7 did not take advantage of the impact of the safe-end to pump closure weld, which would surely increase the times calculated.

Table 2-1: Results of Advanced Finite Element Crack Growth Analyses for Circumferential Flaws Initial Initial Time to Time to Depth/Thickness Length/Circumference a/t = .75 a/t = 1.0 (a/t) 0.20 0.14 10.68 years 12.52 years 0.20 0.23 9.6 years 11.1 years 0.30 0.14 7.44 years 9.34 years 0.30 0.23 6.45 years 7.85 years WCAP-17128-NP May 2010 Revision 1

2-4 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Figure 2-2: Time from Leakage to Critical Circumferential Flaw Length (No Residual Stress Case) for a Through-wall Flaw WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 3-1 3 SUCTION AND DISCHARGE NOZZLE LOADING AND RESIDUAL STRESSES 3.1 NOZZLE LOADINGS The first step in the analytical evaluations reported herein is to determine the appropriate loadings for the service conditions which apply to the pump nozzle DM welds. Both the maximum allowable end-of-evaluation-period flaw sizes and stress intensity factors are functions of the piping stresses, crack geometry, and material properties. Loadings for normal, upset, and test conditions are required, as well as those for emergency and faulted conditions.

The RCP suction and discharge nozzle DM weld regions are subject to piping reaction loads resulting from pressure, thermal expansion, self-weight, seismic, and accident loading conditions.

The self-weight is generally small, often not available separately, and included with normal operating conditions. Therefore, it is not included in the detailed flaw evaluations performed here. Upset, emergency, and faulted load conditions, such as operating or design basis seismic, safe shutdown seismic, loss of coolant accident (LOCA), branch line pipe break (BLPB), and accident conditions were obtained from the engineering specifications and summarized in [22] for the RCP suction and discharge nozzles. Load combinations are plant specific. For this analysis, all load conditions were classified as:

1. Normal operation (NOP) represents thermal loading;

2. Normal operation plus operating basis earthquake (NOP + OBE), representing the upset load level;

3. Normal operation plus safe shutdown earthquake (NOP + SSE), representing the emergency load level; and

4. Normal operation plus accident (NOP + SSE + LOCA, NOP + SSE + BPLB, NOP +

accident), representing the faulted load level.

The normal operation loading condition pipe forces and bending moments, along with the internal pressure loads, were used for the PWSCC flaw growth estimation.

Load condition number 2, listed above, was used for the maximum allowable end-of-evaluation period flaw size for the normal and upset load conditions, as well as conditions 3 and 4 of the corresponding flaw size for the emergency and faulted load conditions. Normal operation loads (without pressure) were used as secondary thermal stresses. The internal pressure load and additional loads beyond the normal operation are assumed to be due to additional pipe mechanical loads (seismic, LOCA, BLPB, and accident) and are used for the primary membrane and bending stresses.

Piping stresses for all the plants were calculated using the corresponding RCP weld geometries are provided in [22]. The nominal dimensions used for this evaluation are shown in Table 3-1.

These dimensions are designed to be best estimates for the weld region of interest here. These stresses are bounded first within each plant; then bounded again to obtain overall maximum values to be used as a generic candidate for the flaw evaluation. Nominal dimensions were then used in the actual calculation of the PWSCC crack growth, fatigue crack growth, and maximum end-of-evaluation-period flaw sizes.

WCAP-17128-NP May 2010 Revision 1

3-2 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Table 3-1: Nominal Dimensions Used for Flaw Evaluation Suction, Discharge Parameter (in)

Outside Diameter 36 Inside Diameter 30 Thickness 3 Operating pressure is 2,250 psi, and the temperature ranges between 543F and 553F. The design pressure of 2,500 psi and temperature of 553F were used in all flaw evaluations to provide some conservatism in the evaluations. High pressure results in higher stress, and higher temperature results in higher crack growth rates.

The stresses at the DM welds for normal, upset, emergency, and faulted conditions were determined using the following equations in the evaluation:

Fa tot m tot Equation 3-1 A

M b tot b tot Equation 3-2 Z

Fa nop M b nop e Equation 3-3 A Z where:

m-tot = primary membrane stress due to total load b-tot = primary bending stress due to total load e = total secondary stress due to normal operation loads Fa-tot = axial force due to pressure and mechanical loads Fa-nop = axial force due to thermal loads Mb-tot = bending moment across the pipe cross-section due to mechanical loads Mb-nop = bending moment across the pipe cross-section due to thermal loads A = pipe cross-sectional area Z = pipe cross-sectional modulus The piping loads are tabulated in [22]. For the PWSCC analysis, only the steady-state operating loads (due to pressure, self-weight, and thermal) are used. Along with the operating loads, the hoop and axial residual stress distributions discussed in Sections 3.2 and 3.3 were used to calculate both the fatigue and PWSCC crack growth. External loads, such as seismic and accident conditions and take place for only a short duration, would not have any significant impact on the overall crack growth.

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 3-3 3.2 RESIDUAL STRESSES The dissimilar metal weld of interest in this report attaches the stainless steel safe-end to the carbon steel piping segment leading to the RC pump. The piping segment and nozzle are fabricated in the shop and can be seen in Figure 1-1Figure 1-1. The portion of the pipe segment where the safe-end will be attached is buttered, and then the entire segment is stress relieved.

After the stress relief, the stainless steel safe-end is attached to the pipe segment with the dissimilar metal weld, and no further stress relief is applied and not required. The segment can then be welded to the pump suction or discharge nozzle in the field, with a stainless steel to stainless steel weld.

The residual stresses do not affect the allowable flaw size, as determined for these ductile materials per Appendix C of Section XI [1]; yet both the fatigue and PWSCC crack growth calculations are affected. The effect on fatigue crack growth is not large because the residual stress exists for both the maximum and minimum points of each transient. The effect on PWSCC is important because the residual stresses make up a significant portion of the total stress.

The residual stresses from the fabrication of the dissimilar metal weld were obtained from finite element modeling, and the model is shown in Figure 3-1Figure 3-1. Note that an axial flaw is more or less self limiting, by the width of the dissimilar metal weld.

The methodology used for the thermal solution is described in some detail below. The temperature constraint method was used, where the weld beads are held to a near-melt temperature, and then allowed to cool.

Each weld bead was held at temperature for 10 seconds in the thermal solution, to capture the effect of heat input on the weld simulation. This analysis was used to obtain the residual stress results for the loop piping - pump nozzle connection after assembly for four cases. The cases are as follows:

1. A 10% inner diameter weld repair with heat treatment after the loop piping butter, but with no heat treatment after the weld repair. Note that this condition is similar to the original condition of this region with no repair, as the weld is back chipped.

2. A 10% inner diameter weld repair with heat treatment after the loop piping butter and with heat treatment after the weld repair.

3. A 25% inner diameter weld repair with heat treatment after the loop piping butter and with heat treatment after the weld repair.

4. A 50% inner diameter weld repair with heat treatment after the loop piping butter and with heat treatment after the weld repair.

The inner diameter repair was simulated as part of this analysis. The residual stresses resulting from the assembly process and inner diameter repair were calculated using an ANSYS finite element two-dimensional axisymmetric model.

ANSYS, ANSYS Workbench, Ansoft, AUTODYN, CFX, EKM, Engineering Knowledge Manager, FLUENT, HFSS and any and all ANSYS, Inc. brand, product, service and feature names, logos and slogans are trademarks or registered trademarks of ANSYS, Inc. or its subsidiaries located in the United States or other countries. ICEM CFD is a trademark used by ANSYS, Inc. under license. CFX is a trademark of Sony Corporation in Japan. All other brand, product, service and feature names or trademarks are the property of their respective owners.

WCAP-17128-NP May 2010 Revision 1

3-4 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Two ANSYS library element types, PLANE55 and PLANE42 were used to create the finite element model. PLANE55 elements were used for the thermal analysis and PLANE42 elements were used for the structural analysis. PLANE55 is a 4-node two-dimensional thermal solid element with a single degree of freedom, temperature, at each node. PLANE42 is a 4-node two-dimensional structural solid element having two degrees of freedom per node: translation in the nodal x and y directions. These element types are appropriate for an axisymmetric evaluation using ANSYS. The same finite element mesh was used to evaluate both the thermal and structural loadings. Note that the global y-axis was oriented along the nozzle centerline and the global x-axis was in radial direction oriented axially 90 degrees° clockwise from the y-axis (required by ANSYS for axisymmetric evaluations).

All of the elements were included in the initial model and brought in and out of the solution using the birth and death capabilities in ANSYS. Temperature-dependent, nonlinear material properties along with the multi-linear kinematic strain hardening model were used in the analysis.

The full length of the stainless steel safe-end and a sufficient length of the stainless steel pump nozzle and carbon steel loop piping were included in the finite element model to ensure end effects have no impact on the regions of interest. The models are shown in Figure 3-1.

The residual stress modeling was designed to match the actual welding process followed in the fabrication shop in Chattanooga, TN, as closely as possible. This information was obtained from the drawings as well as from interviews with personnel who worked there at the time, and were involved in the process. The piping segment was first buttered with Alloy 182, the nozzles welded in, and then the entire piece was heat treated. Following this process, the stainless steel safe-end which is approximately 5.125 inches long, was attached with Alloy 182 weld, to produce a single V weld. After the weld was completed, the inner portion of the weld was removed by grinding, to a depth of approximately 10% of the wall, and then the weld was completed from the ID. Note that this original or un-repaired configuration corresponds to a repair of 10% of the wall. Any repairs to this configuration would have been recorded, as they would have meant an interruption in the shop traveler schedule.

The finite element analysis consisted of a thermal solution followed by an elastic-plastic structural solution. The thermal solution was used to calculate the temperature response of the region of interest. The structural solution calculated the residual stress due to the temperature cycling from the assembly process. After each step of the assembly process the finite element model was allowed to cool to a uniform temperature of 70°F. After the loop piping buttering was simulated, a heat treatment was simulated in accordance with the temperatures required by the ASME Code, Section III Table NB-4622.1-1. The loop piping and attached buttering was raised to a temperature of 1,100°F, and then cooled to 70°F. This same process was repeated after the safe-end to loop piping inner diameter weld repair was simulated for cases 2 through 4.

Hydrostatic test conditions were simulated after the assembly process was completed. A shakedown analysis was then conducted to demonstrate that the nozzle with weld repair do not continue to plastically deform after being cycled from ambient to operating conditions. The shakedown analysis consisted of four cycles of the assembly changing from ambient to operating conditions. Steady state operating conditions included a uniform temperature and a pressure loading of 2,235 psi on the internal surfaces. Steady state ambient conditions included a uniform thermal loading of 70°F and no pressure loading on the inside surfaces of the model.

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 3-5 The finite element model was created in ANSYS Workbench to take advantage of the modeling and meshing capabilities of Workbench. Workbench was then used to write an ANSYS input file to transfer the mesh to ANSYS, where the thermal and structural solutions were completed.

The results for the cases studied are summarized in Figures 3-2 and 3-3. The axial stresses shown in Figure 3-2 show a very similar pattern for all the cases, with stresses rather low at the inside surface, and then rising slightly over the depth of the assumed repair. Then, some distance into the wall beyond the repair, the stresses drop significantly to 15 to 20 ksi in compression. In the outer 20 percent of the wall, the stresses are very similar, rising gradually. Overall, the axial stresses are rather low.

The hoop stresses follow a similar pattern to that shown for the axial stresses, but they are generally significantly higher. The stresses are all positive at the inside surface, and then rise further with distance into the wall, before dropping off significantly at a distance somewhat beyond the depth of the repair. The 25% and 50% repairs drop the most, but in all cases the stresses remain positive.

3.3 VALIDATION OF RESIDUAL STRESS MODELING The finite element modeling of the welding process was validated by comparison of calculated and measured residual stresses from a fabricated pressurizer safety nozzle. Although the pipe size is somewhat smaller, the methodology is the same.

Finite element analysis (FEA) of the weld residual stresses in a pressurizer safety nozzle to safe-end weld was completed, for two cases, before and after application of a structural weld overlay

[29]. The results before the overlay was applied are more appropriate for presentation here, and they are provided in Figures 3-4 and 3-5, for axial and hoop residual stresses, respectively. The finite element analysis was completed prior to the experimental measurements; that is the experimental residual stress measurements were not used to develop the finite element analysis.

An elastic-plastic two-dimensional axisymmetric model was utilized to calculate the residual stresses through-wall at the centerline of the DM weld. The model utilized kinematic strain hardening and the temperature constraint method which greatly simplified the simulation as compared to detailed heat source modeling methods. The temperature constraint method holds the weld beads at near-melt temperature for a range of heat inputs where the range of heat inputs are controlled by the time at which the weld beads are held at temperature. Specifically, five different hold times, i.e., 0.1, 0.5, 1.0, 5.0 and 10.0 seconds, were utilized in the thermal solution to capture the effect of heat input on the simulation.

Figure 3-6 illustrates the FEA model used for the evaluation along with the stress path used for reporting results. For the simulations, the global y-axis was along the safety/relief nozzle centerline and the global x-axis was in the radial direction oriented axially 90º clockwise from the y-axis as is required by ANSYS for axisymmetric evaluations.

Residual stresses in the seven positions selected were measured through-wall with deep hole drilling (DHD) residual stress measurement techniques. Note that all measurements were performed starting from the mockup outer surfaces and progressed through the wall thickness to completion at the inner surface.

From Figures 3-4 and 3-5, it is evident that near the ID and OD surfaces of the mockup, there is good agreement between the measured and modeled results with excellent agreement throughout WCAP-17128-NP May 2010 Revision 1

3-6 WESTINGHOUSE NON-PROPRIETARY CLASS 3 a majority of the mid-wall thickness. Note that near the ID and OD surfaces, the measured residual stresses are slightly more compressive than the modeled values.

While the residual stresses for this smaller thickness case compare very well with the measured values, the results for the thicker section of interest here are somewhat different due to the larger thickness and diameter. These differences are expected, which is the reason this additional work was performed.

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 3-7 10% Weld Repair 25% Weld Repair 50% Weld Repair Figure 3-1: Finite Element Models of the Three Repair Configurations Modeled for the Pipe to Safe-end Weld WCAP-17128-NP May 2010 Revision 1

3-8 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Axial Stress at Operating Conditions 50% Repair 25% Repair 10% Repair With Heat Treat 10% Repair No Heat Treat 40 30 20 10 Stress (ksi) 0

-10

-20

-30 0.00 0.20 0.40 0.60 0.80 1.00 r/t Figure 3-2: Axial Stress Results for All Cases Considered Hoop Stress at Operating Conditions 50% Repair 25% Repair 10% Repair With Heat Treat 10% Repair No Heat Treat 60 55 50 45 Stress (ksi) 40 35 30 25 20 15 0.00 0.20 0.40 0.60 0.80 1.00 r/t Figure 3-3: Hoop Stress Results for All Cases Considered WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 3-9 Figure 3-4: Axial Residual Stress Validation Results for the Pressurizer Safety Nozzle [29]

Figure 3-5: Hoop Residual Stress Validation Results for the Pressurizer Safety Nozzle [29]

WCAP-17128-NP May 2010 Revision 1

3-10 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Alloy 82/182 Weld Cut Figure 3-6: Finite Element Model Geometry for Pressurizer Safety Nozzle Validation [29]

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 4-1 4 SURVEY OF OBSTRUCTIONS FOR INSERVICE INSPECTIONS

4.1 INTRODUCTION

As described in the project authorization in [3], a letter request [4] was made to all participating utilities regarding information concerning obstructions to in-service inspections, as well as leak detection capabilities in the RCP suction and discharge nozzle regions. This request specifically consisted of the following:

Plant leak detection capability as used in licensing activities, Plant leakage detection action levels, Obstructions to inspection, including the fillet radii, namely at small nozzle locations:

o Circumferential as well as axial direction, o Location with respect to the Alloy 182 weld along the axial direction, o Location (angle) around the circumference of the cold leg with respect to the 12 oclock position, and o "Permanent obstructions," such as piping branch connections, elbow intrados, instrument nozzles, etc.

Operating temperatures, including changes over service history, for the:

o Reactor vessel inlet nozzle, o Reactor vessel outlet nozzle, and o Reactor coolant pump suction and discharge nozzles, Inspection information, including the date of the latest UT inspection, and whether it was PDI qualified, for the:

o Reactor vessel inlet nozzle, o Reactor vessel outlet nozzle, and o Reactor coolant pump suction and discharge nozzles.

The characterization of the uninspectable region with permanent obstructions should consider the inspection technique used its requirements, transducer widths, nozzle fillet radii, drawing tolerances, differences between the as-built configurations and the as-designed, weld contours, pipe whip restraints, and any other limitations that prevent the inspection.

This actual data is sought on the uninspectable area around the small nozzles near the RCP suction and discharge nozzle Alloy 182 weld locations on the cold leg. All the data obtained is described and summarized in Section 4.2.

4.2

SUMMARY

OF PLANT OBSTRUCTIONS FOR INSPECTION DATA Information on obstructions for in-service inspection has been summarized on a plant by plant basis. Estimates made by Westinghouse engineers were based on as-built drawings and supplemented the information obtained from surveys given to the plants.

Percentages presented are in terms of the percentage of the inside circumference.

WCAP-17128-NP May 2010 Revision 1

4-2 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Calvert Cliffs The spray nozzle obstructs two of the RCP nozzles. While a customer report claims only 4.44%

of the circumference is obstructed, Westinghouse estimates 11% obstruction.

ANO Unit 2 The spray nozzle obstructs two of the RCPs. An inspection completed in Fall 2009 achieved over 90% coverage.

Waterford Unit 3 A spray nozzle and one RTD cause obstruction. While a customer report provides no data regarding this obstruction, Westinghouse estimated a total of 10.7% + 1% = 11.7% obstruction.

These are potentially connected depending on the RTD weld pad size. Consequently, the space between the spray nozzle and RTD need to be included.

St. Lucie Units 1 and 2 St. Lucie RCP nozzles have already been studied through direct sponsorship of a project from Florida Power and Light (FPL). Some photographs of St. Lucie nozzles are also available, but lack dimensional information. Previous estimates by FPL resulted in a maximum obstruction length of 23%, which includes the spray nozzle and RTDs.

Studies by WESDYNE were conducted to quantify the inspectable and non-inspectable regions.

Spray nozzle obstruction and RTD nozzle reinforcement pads are not located far enough away from the DM welds and are, therefore, considered as obstruction for inspection. FPL is considering grinding the RTD pads to reduce the obstruction.

Millstone Unit 2 Millstone Unit 2, in their recent relief request submitted to the Nuclear Regulatory Commission (NRC), has identified a total volumetric coverage ranging between 73.1% to 80% for all 8 DM welds in their RCP nozzles.

SONGS Units 2 and 3 The four RCP discharge nozzles in the two SONGS units have different obstructions. One discharge nozzle has three RTDs (at 0, 45, and 315 degrees) only. Two RCP discharge nozzles have one charging or spray line attached at the 90-degree location, in addition to the three RTDs.

The fourth pump has both a spray and charging nozzle at the 90-degree and 270-degree locations, in addition to the three RTDs. This results in a total of 24% circumferential obstruction. The UT limitation for each of the spray and charging nozzles is roughly estimated to be 11% of the circumference. These blind zones are separated from the RTD blind zone by an inspectable band approximately 24-degrees of pipe circumference. This was estimated from the photographs obtained from SONGS. Spray and charging nozzles are 180 degrees apart, so they do not need to be combined in the obstruction evaluation.

All Plants A summary of obstruction estimates for the participating utilities is provided in Table 4-1 and summarized generically in Table 4-2.

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 4-3 Based on data available to date, it appears the SONGS plant has the most limiting case in percent coverage obstruction at one pump. There seems to be adequate space between the big nozzles and the RTD pads to consider these obstructions separate for SONGS. If one of the other plants (e.g. Waterford 3) has a large RTD nozzle pad, there may not be adequate space between the big nozzle and the RTD, then it might become the governing plant.

Per Westinghouses survey of design drawings of RCP nozzles, the RTDs of many of the plants are more than 11 inches from the weld centerline, which is greater than two times the wall thicknesses plus the weld width, so the RTDs should not interfere.

4.3 ANALYTICAL ESTIMATION OF OBSTRUCTIONS An analytical estimate of obstructions is obtained from design drawings, then compiled, and summarized in Table 4-1. This table lists various nozzles in the DM weld regions for all plants considered, and includes nozzle outside diameters, axial and circumferential lengths of the nozzle attachments, and the distance of the nozzle centerlines from the edge of the DM weld. When information was circulated to all participating utilities, the obstruction dimensions were increased by the size of the inspection transducer width of approximately 1 inch on either side of the nozzle.

This information was used as a starting point for collection of obstruction data from participating plants in this study.

According to the analytical estimation, the largest circumferential obstruction angle occurs due to the safety injection nozzle attachment. Including the fillet radii on either side of the nozzle, a total of approximately 80° circumferential angle, or 22% of the circumference, is obstructed from in-service inspection (ISI). The next largest obstruction occurs due to charging and sprays nozzle attachments with approximately 40° or 11% of the circumference.

For the flaw evaluation, the largest obstruction assumed was 14% of the circumference, which is based on improvements planned or implemented by several participating utilities during the PWROG project.

WCAP-17128-NP May 2010 Revision 1

4-4 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Table 4-1: Summary of Obstructions for Inspection of CE Fleet RCP Nozzles from Drawings Pipe Nozzle Axial Circumferential Pump Weld Circumferential Plant Suction/

OD Nozzle OD Length Length Angle Axial(2) Location(1)

Name Discharge (in) (in) (in) (in) (°) (in) (°)

SI 21.063 25.5 22.06 78.2 34.59 0 Charging 7.375 10.88 10.88 36.2 50.56 270 Constellation Discharge 35 Calvert Spray 7.375 10.88 10.88 36.2 2.56 0 Cliffs 1 and 2 RTD 7.125 7.125 7.125 23.5 4.44 45, 315 Suction 35 Drain Drain nozzle is far away from the DM weld and is not an obstruction.

SI 21.063 25.5 22.06 78.2 34.78 0 Charging 7.375 10.88 10.88 36.2 24.81 90 Dominion Discharge 35 CT Millstone Spray 7.375 10.88 10.88 36.2 2.75 0 2

RTD 7.125 7.125 7.125 23.5 4.63 45, 315 Suction 35 Drain Drain nozzle is far away from the DM weld and is not an obstruction.

SI 21.063 25.5 22.06 75.6 30.81 0 Charging 7.375 10.88 10.88 35.2 51.88 270 Entergy Discharge 36 Spray 7.375 10.88 10.88 35.2 2.78 0 ANO2 RTD 0.993 0.993 0.993 3.2 7.72 45, 315 Suction 36 Drain Drain nozzle is far away from the DM weld and is not an obstruction.

SI 21.063 25.5 22.06 75.6 30.63 0 Charging 7.375 10.88 10.88 35.2 46.56 270 Entergy Discharge 36 Spray 7.375 10.88 10.88 35.2 3.56 0 Waterford 3 RTD 0.993 0.993 0.993 3.2 3.50 45, 315 Suction 36 Drain Drain nozzle is far away from the DM weld and is not an obstruction.

SI 21.063 25.5 22.06 78.2 34.59 0 35 Charging 7.375 10.88 10.88 36.2 50.56 270 FPL St. Discharge Spray 7.375 10.88 10.88 36.2 2.56 0 Lucie 1 RTD 7.125 7.125 7.125 23.5 4.44 45, 315 Suction 35 Drain Drain nozzle is far away from the DM weld and is not an obstruction.

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 4-5 Table 4-1: Summary of Obstructions for Inspection of CE Fleet RCP Nozzles from Drawings (continued)

Pipe Nozzle Axial Circumferential Pump Weld Circumferential Plant Suction/

OD Nozzle OD Length Length Angle Axial(2) Location(1)

Name Discharge (in) (in) (in) (in) (°) (in) (°)

SI 21.063 25.5 22.06 78.2 34.59 0 Charging 7.375 10.88 10.88 36.2 50.56 270 FPL St. Discharge 35 Spray 7.375 10.88 10.88 36.2 2.56 0 Lucie 2 RTD 7.125 7.125 7.125 23.5 4.44 45, 315 Suction 35 Drain See Note SI 21.063 25.5 22.06 75.6 17.31 0 Charging 7.375 10.88 10.88 35.2 2.56 90, 270 SCE Discharge 36 SONGS 2 Spray 7.375 10.88 10.88 35.2 2.56 90 and 3 RTD 0.993 0.993 0.993 3.2 2.50 0, 45, 315 Suction 36 Drain Drain nozzle is far away from the DM weld and is not an obstruction.

Notes:

SI = safety injection nozzle, RTD = resistance thermocouple detector (1)

Convention: standing on the ground, looking from the pump towards the pipe. 0° is at the 12 o'clock position; 90° is at the 9 o'clock, i.e., counter clockwise. See Figure 4-1. Also, see Figure 4-1 through Figure 4-4 for sample dimension conventions used for this table.

(2)

Axial distance is measured from nozzle fillet edge to weld edge.

Figure 4-1: Nozzle Circumferential Location Convention All dimensions are nominal. The width of the DM weld is approximated from the drawings.

Figure 4-3 shows the dimension convention for each nozzle type.

WCAP-17128-NP May 2010 Revision 1

4-6 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Figure 4-2: Sample Safety Injection Nozzle Uninspectable and Obstruction Dimensions Figure 4-3: Sample Charging and Spray Nozzles Uninspectable and Obstruction Dimensions WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 4-7 Figure 4-4: Sample RTD Nozzle Uninspectable and Obstruction Dimensions Notes:

1. For the RTD nozzle without a pad, the uninspectable "Axial Length" and "Circumferential Length" is the outside diameter.

2. The "Pump Weld Axial" is the outside diameter edge to the DM weld edge.

3. For the RTD nozzle with a pad, the uninspectable "Axial Length" and "Circumferential Length" are the pad diameter plus two times the fillet radius.

4. The Pump Weld Axial is the edge of the pad fillet to the DM weld edge.

WCAP-17128-NP May 2010 Revision 1

4-8 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Table 4-2: Obstruction Region Estimated based on Enveloped Plant RCP Nozzles Pipe Axial Suction/ Nozzle Circumferential OD Length Discharge Type Angle (°)

(in) (in)

SI 26 79 Charging 12 40 Discharge 35 Spray 12 40 RTD 7.1 23 Suction 35 No Obstruction WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 5-1 5 JUSTIFICATION FOR DEVIATION FROM INSPECTION COVERAGE REQUIREMENTS: DEFENSE IN DEPTH 5.1 LEAK DETECTION CAPABILITY After a number of recent operating events, the industry imposed an NEI-03-08 needed requirement, to improve their leak detection capability. As a result, virtually all pressurized water reactors (PWRs) in the US have a leak detection capability of less than or equal to 0.1 gpm. All plants also monitor seven day moving averages of reactor coolant system leak rates.

Action levels have been standardized for all PWRs, and are based on deviations from:

The seven day rolling average, Specific values, and The baseline mean.

Action response times following a leak detection vary, based on the action level exceeded and range up to containment entry to identify the source of the leak. Utilities take the commitment of shutdowns due to unidentified leakage seriously. This is exemplified with utility shutdowns in July 2009, due to a 0.2 gpm leakage, and another in August 2009, with 0.09 gpm leakage. This improvement in leak detection sensitivity is due to multiple measures being monitored.

Leak rate action levels are identified in PWROG report, WCAP-16465 [24], and are below:

Each PWR utility is required to implement the following standard action levels for RCS inventory balance in their RCS leakage monitoring program.

A. Action levels on the absolute value of unidentified RCS inventory balance (from surveillance data):

Level 1 - One seven day rolling average of unidentified RCS inventory balance values greater than 0.1 gpm.

Level 2 - Two consecutive unidentified RCS inventory balance values greater than 0.15 gpm.

Level 3 - One unidentified RCS inventory balance value greater than 0.3 gpm.

Note: Calculation of the absolute RCS inventory balance values must include the rules for the treatment of negative values and missing observations.

B. Action levels on the deviation from the baseline mean:

Level 1 - Nine consecutive unidentified RCS inventory balance values greater than the baseline mean [] value.

Level 2 - Two of three consecutive unidentified RCS inventory balance values greater than [ + 2], where is the baseline standard deviation.

Level 3 - One unidentified RCS inventory balance value greater than [ +3].

Information obtained about leak detection capabilities, detection levels, inspection obstruction regions, operating temperatures, and the latest inspection type and year regarding applicable plants is listed in Table 5-1.

WCAP-17128-NP May 2010 Revision 1

5-2 WESTINGHOUSE NON-PROPRIETARY CLASS 3 5.2 LEAK RATE METHODOLOGY As discussed earlier, the CE cold leg RCP nozzles have permanent obstructions that preclude the required, ultrasonic inspection coverage for circumferential flaws in the Alloy 82/182 welds. The combined calculated missed circumferential examination coverage ranges from 11% to 14% of the circumference. Since the action levels now employed by all the PWR utilities allow the early detection of small leakages, it is necessary to identify the extent of defense in depth this new sensitivity allows.

Quantifying the margins between leakage detection and the time required for the flaw to reach a critical length provides another measure of the flaw tolerance which exists in the RCP nozzle region.

Postulation of the initial through-wall circumferential flaws is determined based on leakage calculations consistent with current Nuclear Regulatory Commission (NRC) approved leak-before-break methodology [25]. Circumferential flaws yielding a leak rate of 0.1, 0.25, 0.5, 1.0, and 2.0 gpm were postulated as initial flaws for the current analysis. These leak rates are within typical nuclear power plant leakage detection capabilities, as discussed above.

The basic method used in the leak rate calculations was developed by Fauske [7] for the two-phase choked flow. To this, pressure loss due to friction upstream of the choked exit plane was added.

The flow rate through a crack was calculated in the following manner. Figure 5-1 [8] was used to estimate the critical pressure, Pc, for the primary loop enthalpy condition and an assumed flow.

Once Pc was found for a given mass flow, the stagnation pressure upstream of the choked plane is obtained from Figure 5-2, which is taken from [8]. For all cases considered, the length to diameter ratio, L/DH > 40, Pc/Po, is equal to 0.55. Therefore, this method will yield a two-phase pressure drop due to momentum effects, as illustrated in Figure 5-3, where Po is the operating pressure. Using the assumed flow rate, G, can be calculated as:

(L / DH - 40)G 2 Pf = f , Equation 5-1 2gc (144 )

where f = friction factor,

= density of the fluid, G = assumed flow rate, L/DH = length to diameter ratio of the pipe, and gc = acceleration due to gravity.

Here, f is determined using the Moody diagram. The crack relative roughness () was obtained from fatigue crack data on stainless steel samples. The relative roughness value used in these calculations was 300 micro-inches root-mean-square (RMS). The frictional pressure drop using Equation 5-1 is then calculated for the assumed flow rate and added to the momentum pressure drop calculated using the Fauske model to obtain the total pressure drop from the primary system to the atmosphere for a given assumed flow rate, G.

Absolute Pressure - 14.7 = PT = (Pf + P2 choked flow) Equation 5-2 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 5-3 If the right-hand side of Equation 5-2 does not agree with the pressure difference between the primary loop and the atmosphere, then the procedure is repeated until Equation 5-2 is satisfied to within an acceptable tolerance, which in turn leads to a flow rate value for a given crack size.

Leak rate calculations were made as a function of crack length using the normal operating (NOP) loads provided in [5]. The NOP loads consist of the deadweight, thermal expansions, and pressure loads. Seismic loading is not included since it is an upset condition and also because it will result in a larger leakage flaw size for a given flow rate. The NOP loads for leak rate predictions are calculated by the following equations:

F = FDW + FTH + FP MX = (MX)DW + (MX)TH MY = (MY)DW + (MY)TH MZ = (MZ)DW + (MZ)TH where, DW = deadweight, TH = normal thermal expansion, and P = load due to internal pressure.

The stresses due to axial loads and bending moments in the leakage flaw size determination are calculated by the following equation:

F M Equation 5-3 A Z where,

= stress, F = axial load, M = moment, A = pipe cross-sectional area, and Z = section modulus.

The moments for the desired loading combinations are calculated by the following equation:

M M X2 M Y2 M Z2 Equation 5-4 where, Mx = X-component of the moment, torsion, My = Y-component of the bending moment, and Mz = Z-component of the bending moment.

The crack opening areas were estimated using the method of [9], and the leak rates were calculated using the two-phase flow formulation described above. The material properties at NOP temperature of 550°F were used for these calculations. The flaw sizes to yield a leak rate of 0.25, 0.50, 1.0, and 2.0 gpm were calculated using the computer code FHG [10, 11]. Crack opening areas to determine the leakage rates are calculated using the MPBK [10, 11] computer WCAP-17128-NP May 2010 Revision 1

5-4 WESTINGHOUSE NON-PROPRIETARY CLASS 3 program. To account for the PWSCC crack morphology for the Alloy 82/182 weld leak rate calculation, a factor of 1.69 was applied to the leakage flaw size calculated for the fatigue crack morphology [12]. The results of the leakage flaw lengths for various leak rates are provided in Section 5.4.

5.3 CIRCUMFERENTIAL THROUGH-WALL CRITICAL FLAW SIZES - ASME SECTION XI, APPENDIX C The critical through-wall circumferential flaw size determination is based on limit load methodology: the critical flaw size calculated is the circumferential flaw length required to cause pipe failure due to plastic collapse. The critical flaw lengths for through-wall circumferential flaws are also calculated based on Appendix C of ASME Section XI [1]. For flaws with circumferential angle (+) as shown in Figure 5-4, the relation between the applied loads and flaw size at net plastic collapse is given by:

2 f a bc = ( 2 sin - sin ) Equation 5-5 t

1 a

( - - m ) Equation 5-6 2 t f where, bc = bending stress at incipient plastic collapse,

= one-half of the final flaw angle,

= angle to neutral axis of flawed pipe, a/t = set to unity for through-wall circumferential flaws based on Code Case N-513-2 [1],

S y + Su f = flow stress = , and 2

m = applied membrane stress.

The allowable bending stress, Sc, used to calculate the maximum allowable end-of-evaluation period flaw sizes for the DM welds, is computed using:

1 bc 1 Sc e m 1 Equation 5-7 (SFb ) Z Z(SFm )

where Sc = allowable bending stress for circumferentially flawed pipe, cb = applied bending stress at incipient plastic collapse, m = applied membrane stress, e = thermal expansion stress, SFm = safety factor for membrane stress (for Service Level A, B, C, and D, SFm= 2.7, 2.4, 1.8, and 1.3, respectively),

SFb = safety factor for bending stress (for Service Level A, B, C, and D, WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 5-5 SFb= 2.3, 2.0, 1.6, and 1.4, respectively),

Z = 0.0000022(NPS)3 - 0.0002(NPS)2 + 0.0064(NPS) + 1.1355 , and NPS = nominal pipe size.

The critical flaw length can then be determined by equating the applied bending moment at the nozzle to the allowable bending stress (Sc) in the above equation. It should be noted the Z correction factor from [1] is used, since it is representative of the Alloy 182 dissimilar metal weld of concern here. The results for the ASME limit load calculations are given in Section 5.4 for the pump suction and discharge nozzle DM welds.

5.3.1 Through-wall Circumferential Flaw Stress Intensity Factor Calculation The axial stresses due to the normal operating loads from [5] (deadweight and thermal expansion) are combined with the residual stresses from [6] (illustrated in Figure 6-3) at the DM welds to determine the stress intensity factors for the through-wall circumferential flaw configuration.

Once the stress intensity factors are determined, stress corrosion crack growth calculations can be performed using a PWSCC crack growth rate model developed in [13].

The bounding total stress (piping plus residual stresses) from the enveloped CE fleet RCP nozzle case were used to calculate the stress intensity factor (SIF) at the pump inlet and outlet nozzles.

Recent literature solutions from Zangs paper in [14] for SIF expressions were used. These solutions provide representation of the through-wall stress distribution profile at the DM weld using a 4th order polynomial fit.

The stress intensity factors solutions from [14] were determined from a three-dimensional finite element model for through-wall cracks in cylinders. The axial stress distribution to calculate SIF can be determined by a 4th degree polynomial as follows:

for through-wall stress distribution,

( x) A0 A1 x A2 x 2 A3 x 3 A4 x 4 Equation 5-8 and for a global pipe bending moment, z

x gb Equation 5-9 Ro where, A0, A1, A2, A3, and A4 = the stress profile curve fitting coefficients to be determined, x = distance from the wall surface where the crack initiates, z = radial distance to the point in the pipe wall thickness, Ro = outer radius of the pipe, gb = maximum global bending stress at the outside surface of pipe, and

= axial stress.

WCAP-17128-NP May 2010 Revision 1

5-6 WESTINGHOUSE NON-PROPRIETARY CLASS 3 The SIF for through-wall circumferential cracks due to the stresses defined above can be expressed as:

4 K I = c Ai Fi gb F5 Equation 5-10 i=0 where, Fi, i = 1 through 4 are the normalized SIF influence coefficients for the polynomial stress fit coefficients, F5 = the influence coefficient for the global bending stress, and c = the average half crack length around the circumference.

The normalized SIFs for through-wall stress distributions, Fi, i equals 1 to 4, have been further determined at the inside surface, intermediary locations, and outside surface of the cylinder. The normalized SIF have been calculated for the case of t/Rin = 0.2 (thickness to inside radius ratio),

which most closely represents the pump inlet and outlet nozzle geometries. The SIFs were calculated as a function of crack length. These results will be used to generate PWSCC crack growth for various initial crack lengths in this section. The stress intensity factors for part-through flaws were determined from the work of Raju and Mehtu [19, 20].

5.4 RESULTS Circumferential through-wall flaw lengths for various leak rates, ranging from 0.1 gpm to 2 gpm were calculated, for two cases, one for the minimum normal operating loads, and a second for the maximum normal operating loads. This is to cover the total leak rate crack lengths for the entire range of the RCP nozzles. The minimum normal operating load case results in a larger initial crack length and reaches the critical flaw length sooner, compared to the maximum normal operating case. This time period for a leakage flaw to reach critical crack size also depends on the other emergency and faulted loads as the latter determines the maximum critical crack lengths.

Table 5-2 lists initial total circumferential flaw lengths with various leak rates for the minimum and maximum normal operating loads. This table shows the leak rate flaws range from as small as 1.37 inches for a 0.1 gpm leak rate with maximum normal operating loads, to as long as 6.72 inches for 1.0 gpm leak rate with minimum normal operating loads. As all the CE plants listed in Table 5-1 have a leak detection capability of 0.1 gpm, initial crack sizes as small as 1.4 inch are of interest for the flaw growth.

Critical circumferential through-wall flaw sizes are computed for all the CE fleet RCP nozzles.

As the normal, upset, emergency, and faulted loads vary considerably between various plants.

Plant specific critical flaw sizes were computed for each plant as the enveloping load will be too restrictive for the rest of the plants. Table 5-3 shows the total circumferential crack lengths for the end-of-evaluation period. Any initial leak rate or assumed obstruction flaw propagation to these maximum lengths show the total time period available for inspection. This is discussed in the Section 6.

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 5-7 Calculations of the time to grow a crack from a length resulting in the actionable leak rate of 0.1 gpm to the critical length of a through-wall flaw showed that at least 14 years are required, an extremely large margin over the 12 hour1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> maximum action time. This margin is shown graphically in Figure 2-2.

5.5 POTENTIAL FOR BORIC ACID CORROSION DAMAGE The effect of potential reactor coolant leakage in this region was also assessed; although it seemed apparent that no such damage would occur. To complete the evaluation, it was assumed that a leak of 0.15 gpm occurred in the dissimilar weld of interest here. The reactor coolant temperature is assumed to be 560°F with a pressure of 2,235 psia. The maximum level of boric acid in the system would occur at the beginning of the fuel cycle and would be approximately 2,000 ppm boron.

There are a number of components and materials in close proximity to this weld:

The pump body and safe-end materials (stainless steel at 550°F - 560°F)

The reactor coolant piping (clad carbon steel at 550°F - 560°F)

The supports for the pump (carbon steel at ~120°F)

The concrete holding the supports (120°F)

Leakage through a crack in the weld of interest would result in the reactor coolant flashing to steam, but there is a potential for some liquid to remain in the mixture. Because the temperature of the pipe is 550°F - 560°F, the remaining liquid will quickly boil off, leaving dry boric acid.

Therefore, there is concern for steam to escape and potentially condense on nearby equipment.

The other hot locations would simply boil off any liquid that might land on them, but there is potential for damage to the cooler locations. Each location in question will be discussed below.

Although the period of time over which the utility would take action is likely to never exceed seven days, a period of two months will be assumed here.

Stainless Steel: There is no impact because it is hot and resistant to damage.

Carbon Steel Piping: For this location, the only exposure would be to dry boron crystals.

Reference [26] indicates no measurable corrosion at this temperature range (550°F - 560°F).

Carbon Steel Supports: The corrosion rate for carbon steel regions operating at 210°F is given in [26] as 4.8 inches/year for dripping boric acid. Since the supports are kept at 120°F or less by the Heating, Ventilation, and Air Conditioning (HVAC) system, this rate needs to be corrected for this lower temperature (120°F). Assuming the corrosion rate doubles for every 10°F, the resulting rate at 120°F would be < 0.010 inches/year. Therefore, the degradation of a support would be insignificant over the time of interest here.

Concrete: In most cases the concrete is coated, and so there is no direct contact with boric acid.

For conservatism, this evaluation will consider the concrete to be in contact with the boric acid.

Reference [27] indicates the depth of degradation may be modeled by:

Depth = Cot0.5 With Co = 0.00812 inches/ day0.5 WCAP-17128-NP May 2010 Revision 1

5-8 WESTINGHOUSE NON-PROPRIETARY CLASS 3 For the 60 days of exposure assumed, the depth of the attack is 0.063 inches, which is insignificant.

Therefore, there is no concern for the degradation of any of the components which might be affected by a leak in the region of interest.

Figure 5-1: Analytical Predictions of Critical Flow Rates of Steam-Water Mixtures WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 5-9 Figure 5-2: Critical or Choked Pressure Ratio as a Function of L/D Figure 5-3: Idealized Pressure Drop Profile through a Postulated Crack WCAP-17128-NP May 2010 Revision 1

5-10 WESTINGHOUSE NON-PROPRIETARY CLASS 3 a

t 2

t R

Neutra l Axis Figure 5-4: Circumferential Flaw Geometry Figure 5-5: Time from Leakage to Critical Circumferential Flaw Length (No Residual Stress Case)

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE PROPRIETARY CLASS 3 5-11 Table 5-1: Summary of Leak Detection Capability, Operating Temperatures, and Inspection Data Operating Temperatures Leak Detection Inspectable Leak Detection RV RV WEC Capability /Obstruction RCP RCP Latest Plant Name Level Inlet Outlet Identified Licensed Region Suction Discharge Inspection, (gpm) Nozzle Nozzle Year and Obstruction(1)

(gpm) (%) (°F) (°F)

(°F) (°F) Type (%)

1 gpm for 12.9% obstruction 0.1 gpm for 7 day Constellation unidentified, 10 for ASME Section rolling average Calvert Cliffs gpm for XI and 22.7% 11%

unidentified 1 and 2 identified obstruction per leakage leakage MRP-139 1 gpm for 0.1 gpm for 7 day unidentified, 10 Fall 2009 Entergy ANO rolling average Over 90% 10.7% by gpm for 545.4°F to 553°F with PDI 2 unidentified coverage achieved spray nozzle identified UT leakage leakage 0.1 gpm for 7 day information on Entergy rolling average 544°F average for all 11.7% by 1 gpm obstruction is not 610°F Waterford 3 unidentified four cold legs spray nozzle clearly identified leakage 0.1 gpm for 7 day FPL St. Lucie rolling average 11% by spray 1 gpm 548.5°F to 550°F 1 and 2 unidentified nozzle leakage PDI UT in 8% each by 93% for RTD for 594°F 2002, non- RTD and 11%

SCE SONGS 0.10 gpm all cold legs. 84%

1 gpm to 540°F to 553°F PDI UT in each by spray 2 and 3 unidentified source charging inlet 605°F 1996 and and charging based on 1 cold leg 1999 nozzles WCAP-17128-NP May 2010 Revision 1

5-12 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Table 5-2: Initial Total Flaw Lengths for Various Leak Rates Maximum Normal Minimum Normal Operating Load Operating Load Leak Rate Crack Length Crack Length (gpm) (in) (in) 0.1 1.37 2.71 0.25 1.98 3.90 0.5 2.62 5.13 1.0 3.45 6.72 2.0 4.53 8.75 Table 5-3: Critical Circumferential Flaw Lengths Using the ASME XI Appendix C Approach Limiting Limiting 2crit 2Ccrit Plant (°) (in)

FP&L SL1 and 2 114.4 32.9 DC M2 86.5 24.9 CEG CC1 and 2 92.4 26.6 ANO2 104.8 30.2 W3 81.6 23.5 SONGS 2 and 3 71.7 20.7 Enveloped 71.5 20.6 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 6-1 6 FLAW TOLERANCE PER ASME SECTION XI 6.1 TRANSIENT ANALYSIS FOR THROUGH-WALL AXIAL STRESS DISTRIBUTION FOR USE IN FCG The through-wall transient stresses for the RCP pipe to safe-end Alloy 82/182 DM weld were calculated using WESTEMS'. WESTEMS is a Westinghouse proprietary computer code, verified and configured for this type of analysis per [15]. WESTEMS permits the calculation of detailed stresses from pressure and thermal loads, as well as from externally applied forces and moments. Linear scaling of unit load finite element runs obtain stresses for mechanical cases (pressure, force, and moment). Time-dependent temperature profiles generate thermal loads using function integration. These temperature profiles utilize transfer function databases created with unit load (1°F) thermal analyses.

The stresses for the unit loading cases are calculated using ANSYS. ANSYS is a commercially available general-purpose finite element computer code, verified and controlled in the Westinghouse computer system [16]. ANSYS generates the transfer functions using non-temperature dependent material properties and constant values of heat transfer coefficients.

Therefore, the WESTEMS results must be benchmarked. This benchmark compares generic transient results generated by WESTEMS with ANSYS-generated results with standard temperature dependent material properties. An adjustment factor from the comparison was used in the WESTEMS transient stress calculation.

The axisymmetric ANSYS Finite Element Model (FEM) conservatively models a typical dissimilar metal weld geometry with 30-inch inner diameter and 3 inch wall thickness. Physical properties [17] of the SA-516 Gr70 material were assigned to the carbon steel pipe; Alloy 82/182 properties were assigned for the dissimilar metal weld. The FEM and the ANSYS path, referred to as Analysis Section Number (ASN) in WESTEMS, is shown in Figure 6-1. WESTEMS' provides the through-wall transient stresses in a format that can be used in the fatigue crack growth analysis.

As Figure 6-1 shows, the bottom end of the ANSYS model is constrained in the Y-direction for the pressure and thermal/mechanical analyses. Blow off pressure is applied at the top end of the model for pressure analysis to simulate the rest of the piping system. For the thermal/mechanical analysis, nodes at the top end of the model are coupled in the Y-direction to simulate a long pipe.

The cut defined at the middle of the Alloy 182 weld was divided into ten equally spaced sections and contains eleven nodes through the cut. Figure 6-1 shows the ASN location on the model. For the heat transfer analysis, a conservative film coefficient (4,384 BTU/hr-ft2-°F) was applied to the inside surface of the pipe. The outside surface was conservatively assumed to be insulated. The temperature of the inside surface is increased by 1°F in one second. The case is then run to 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br />, where the model reaches equilibrium. The postulated temperature time-history transients are applied in WESTEMS. The thermal stresses are calculated using the transfer function method.

WESTEMS is a trademark of Westinghouse Electric Company, LLC.

WCAP-17128-NP May 2010 Revision 1

6-2 WESTINGHOUSE NON-PROPRIETARY CLASS 3 The mechanical pipe loads used in the WESTEMS analysis are provided in [5]. Both the maximum and the minimum applied loads are considered in this analysis. For piping loads, only the axial force and bending moment are considered, since the effect of shear stress on crack growth is insignificant.

6.2 PWSCC GROWTH CALCULATIONS The CE design pump nozzle to safe-end dissimilar metal weld region is made of nickel based alloys. This nickel based alloy material (Alloy 82/182) is susceptible to the PWSCC growth mechanism. Once the stress intensity factors are determined, PWSCC crack growth can be calculated based on the applicable ASME Code recommended crack growth curves for PWSCC

[13]. The recommended PWSCC growth curve for Alloy 182 material is as follows:

da Qg exp (1/T 1/Tref) (K) Equation 6-1 dt R where:

da

= crack growth rate in m/sec, dt Qg = thermal activation energy for crack growth = 130 kJ/mole (31.0 kcal/mole),

R = universal gas constant = 8.314 x 10-3 kJ/mole-K (1.103 x 10-3 kcal/mole-

°R),

T = absolute operating temperature at the location of crack, °K (°R),

Tref = absolute reference temperature used to normalize data = 598.15°K (1,076.67°R),

= crack growth amplitude = 1.50 x 10-12 at 325°C (617F),

= exponent = 1.6, and K = crack tip stress intensity factor (MPam).

The pump outlet nozzle nominal operating temperature was taken as 550F [22]. This temperature is used in the fracture mechanics analyses. The stresses used for PWSCC evaluations included normal operating condition piping stresses and pressure. The PWSCC growth rate was determined as shown below, where K is in units of psiin and the resulting growth rate is in units of inches per hour.

da 6.925 10 -13 (K)1.6 Equation 6-2 dt Typical crack tip stress intensity factors across the nozzle thickness for various circumferential through-wall crack lengths are plotted in Figure 6-4. The figure also consists of enveloping the maximum, as well as an averaged SIF across the nozzle wall thickness. These represent the maximum and average crack driving forces occurring in the wall for circumferential crack WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 6-3 propagation. Here, after comparison with the detailed three-dimensional FEACrack analysis of crack propagation under PWSCC conditions, described in Section 7, the average SIF was chosen for the PWSCC growth evaluation. Results of the evaluation for the maximum normal operating loads with various initial crack sizes are shown in Figure 6-5. These initial crack sizes represent different leak rates, as well as the average maximum obstruction of 11% of the nozzle outside circumference.

6.3 FATIGUE CRACK GROWTH CALCULATIONS The through-wall stress distributions used in the crack tip SIF calculation were determined by combining the stresses from the plant operating transients with the residual stresses. The axial and hoop residual stresses used in this evaluation are from MRP-113 [6]. The residual stresses at ambient temperature were conservatively assumed for both ambient and normal operating conditions. It is assumed the residual stresses remain unchanged for the entire duration of plant life.

At each time step, crack tip SIFs were computed for each transient. Full-circumferential part-through-wall flaws were considered in the evaluation. To compute the SIFs for axial and circumferential flaws, Raju-Newman and NASA solutions from [19 and 20] were used.

Once the SIFs were computed for each transient, the maximum and minimum SIFs for various flaw depths were determined. Then, these minimum and maximum SIFs were curve-fit separately into a 6th-order polynomial as a function of flaw depth. Finally, the resulting polynomials were used in the fatigue crack growth (FCG) evaluation.

The FCG analysis procedure involves postulating an initial flaw at the region of concern.

Postulated flaws are subjected to cyclic loads due to transients. The input required for an FCG analysis is the range of crack tip SIFs, K. K depends on the crack size, crack shape, geometry of the structural component where a crack is postulated, and the applied cyclic stresses. Also, load ratio, R = Kmin/Kmax, is required for the scaling parameter in the crack growth model.

Once R and K are calculated, the crack growth due to any given stress cycle can be calculated.

Then, this increment of crack growth is added to the original crack size, and the analysis proceeds to the next transient. The procedure is continued in this manner until all the transients known to occur in the period of evaluation have been analyzed. The design transient load cycles were based on a 40-year plant design life. The crack growth for each transient for a given time interval can be computed using the following equation:

ai 1 ai a Equation 6-3 The incremental crack depth is given by:

da a N Equation 6-4 dN env

FEACrack software is a trademark of Quest Reliability, LLC.

WCAP-17128-NP May 2010 Revision 1

6-4 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Fatigue crack growth was calculated based on the through-wall Kmax and Kmin polynomials and the design transient cycles.

The general crack growth rate for Alloy 182 materials in PWR environments are given by:

da Fweld Fenv C (T ) S ( R )(K )

n Equation 6-5 dN env where, C(T) = scaling factor for temperature effects, S(R) = scaling factor for load ratio effects, Fweld = factor for weld material, Fenv = factor for environment, K = SIF range = Kmax - Kmin, MPam (ksiin),

R = load ratio Kmin / Kmax, Kmax = maximum SIF, MPam (ksiin),

Kmin = minimum SIF, MPam (ksiin),

da

= crack growth rate in environment, m/cycle (inch/cycle), and dN env n = crack growth law exponent.

The crack growth rate reference curves for the Alloy 82/182 weld have not been developed for Section XI in the ASME Code; therefore, information available from the literature was used.

Based on the results reported in [21], the parameters for the crack growth model for Alloy 82/182 material are:

CA600 = 4.835 x 10-14 + (1.622 x 10-16)T - (1.490 x 10-18)T2 + (4.355 x 10-21)T3 Equation 6-6 S= (1 - 0.82R)-2.2 Equation 6-7 Fenv = 1 + A [CSKn]m-1TR1-m Equation 6-8 Fweld= 10 where, T = temperature (C),

K = SIF range, MPam (ksiin),

Kmax = maximum SIF, MPam (ksiin),

Kmin = minimum SIF, MPam (ksiin),

n = crack growth law exponent (= 4.1),

A = constant in crack growth law for Alloy 82/182 weld (= 4.4 x 10-7),

m = exponent in crack growth law for Alloy 82/182 weld (= 0.33),

TR = rise time, seconds, and Fweld = factor for weld.

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 6-5 The values for A and m in Equation 6-8 are provided in [21] through a least-square curve fitting of the FCG data on Alloy 82/182 material in high-purity water with 300 ppb dissolved oxygen.

For the Alloy 82/182 material, Fweld = 10 is used to determine the FCG. The basis for the crack growth rate (CGR) curves from [21] is shown in Figure 6-2.

The transient stresses from the WESTEMS analysis discussed previously were used in the fatigue crack growth calculations. The fabrication weld residual stresses from [6] are then added to the transient stresses. Then, each of the transient stress was evaluated for through-wall crack tip SIFs at various transient time steps and cyclic minimum and maximum values- captured for different flaw lengths. Typical values for the heatup transient are shown in Figure 6-6. This procedure was followed for all the transients. Then the fatigue crack growth evaluation was performed, and results are summarized in Figure 6-7. This figure shows the results for a through-wall circumferential flaw for various initial crack sizes. It can be seen from this figure that fatigue crack growth is considerably slower than the PWSCC growth, indicating the later to be the predominant mechanism.

Additionally, a fatigue crack growth analysis was performed for an ID surface flaw, using WES_FRAMES [18]. The residual stresses from [6] were used. Initial flaw depths ranging from 50% to 100% of the wall thickness were evaluated. A total of six cases were considered:

1. Maximum pipe load with no residual stress,

2. Minimum pipe load with no residual stress,

3. Maximum pipe load with residual stress, no ID weld repair,

4. Minimum pipe load with residual stress, no ID weld repair,

5. Maximum pipe load with residual stress and ID weld repair, and

6. Minimum pipe load with residual stress and ID weld repair.

As shown in Figure 6-9 through Figure 6-12, the results of fatigue crack growth is negligible for surface flaws with initial flaw depths below 60% wall thickness. For initial flaw depths greater than 60% wall thickness, the effect of FCG is small, but measurable. Therefore, for the surface flaws which are of interest to the evaluations discussed in this report, fatigue crack growth can be ignored.

6.4 COMBINED PWSCC AND FATIGUE CRACK GROWTH EVALUATION Since fatigue crack growth for through-wall flaws was found to make a meaningful contribution to the total growth, a methodology was developed to allow calculation of the combined growth from both fatigue and PWSCC. (Note this was not necessary for surface flaws, since growth was negligible.)

While PWSCC occurs throughout the operating period between the outages, fatigue crack growth occurs only when the transient cycle is being applied during operation between the outages.

Also, the actual timing of the transient occurrence is not known in advance and may vary from outage to outage and plant to plant. To start the analysis, a sequential flaw growth with PWSCC was assumed to occur continuously for one year. This was followed by fatigue crack growth for all the cycles over the course of a one-year period. First, Equation 6-1 was applied for the WCAP-17128-NP May 2010 Revision 1

6-6 WESTINGHOUSE NON-PROPRIETARY CLASS 3 PWSCC growth for one year. Then, the FCG was evaluated using Equation 6-5 for all the transient cycles per year. The process was then continued, and total crack growth was then plotted on a yearly basis.

Typical results for the combined crack growth are shown in Figure 6-8. The combined crack growth indicates, for example, an initial 11% circumferential length flaw grows to approximately a total crack length of 20.6 inches in 4.7 years compared to about seven years if only PWSCC growth was considered. The most limiting critical circumferential flaw length for the CE fleet with maximum applied piping loads is 20.6 inches. For the least severely loaded plant, the critical length is as high as 33 inches. For the latter case, an initial flaw of 11% circumferential through-wall reaches the critical length in approximately seven years under the combined PWSCC and fatigue crack growth mechanism.

6.5 ASME SECTION XI FLAW TOLERANCE CALCULATIONS The flaw evaluation performed in Phase I of the PWROG study [22] revealed that these nozzles operating at cold leg temperatures have considerable flaw tolerance, but the results were limited to a two-year service period. This was because only PWSCC growth was considered, and for longer time periods, it was thought fatigue crack growth could play a role. With the present study, both fatigue and PWSCC growth have been evaluated. Therefore, the flaw tolerance evaluation can be extended to longer service periods.

As discussed in Section 5.3 of this report, the allowable flaw depth has been determined from the governing loads, as a function of the flaw shape. Fatigue crack growth has been determined to be negligible, so the PWSCC results will govern the flaw tolerance. Both axial and circumferential flaws were evaluated, and the results are presented in terms of the largest initial flaw, which is acceptable for a range of time periods. The results presented here are for periods of 24, 36, and 48 months, but the evaluations could be easily extended to justify the acceptability of a smaller flaw, should one be discovered during an in-service inspection.

The maximum allowable flaw size, per Appendix C of Section XI [1], is not affected by residual stresses, since the material is ductile. However, since PWSCC is the dominant mechanism of growth for flaws in this region, the residual stresses will affect the growth. A design-specific finite element analysis was completed, and is discussed in detail in Section 3 of this report. Four cases were studied:

Fabrication plus a 10% ID repair, Fabrication plus a 10% ID repair, with post weld heat treatment (PWHT),

Fabrication plus a 25% ID repair, with PWHT, and Fabrication plus a 50% ID repair, with PWHT Repair induces compressive axial residual stresses in the mid-wall region, just beyond the repair.

The closure weld is therefore effectively a mitigation, causing compressive axial stresses at the pipes ID, thus essentially preventing crack initiation. The hoop stresses are depressed as well, as result of the closure weld, but not as severely as the axial stresses. This is consistent with the results on closure welds in smaller diameter pipes [23].

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 6-7 The allowable flaw depths for both the suction and discharge nozzles are very large. For axial flaws, the allowable depth ranges from 60 to 75 percent of the pipe wall thickness, depending on the flaw shape. For circumferential flaws, the allowable depth ranges from 73 to 75 percent of the thickness, depending on the flaw shape.

Flaw evaluation charts were developed for the region of interest, using the design-specific residual stresses described in Section 3, and a series of figures was prepared to cover a range of repair scenarios. These charts all have the same character, and are designed to allow quick evaluation of indications which may be identified during inspection. The curves in the charts were determined fron PWSCC calculations, and include the effects of fatigue crack growth, which was found to be negligible.

Once an indication is identified, it must be characterized as to its location, length (l) and depth dimension (a). This characterization is discussed in further detail in Article IWA 3000 of Section XI[1].

The following parameters must be calculated from the above dimensions to use the charts (see Figure 6-13 for example):

Flaw Shape Parameter, a/l Flaw Depth Parameter, a/t where t = wall thickness of region where indication is located l = length of indication a = depth of surface flaw; or half depth of embedded flaw in the width direction Once the above parameters have been calculated, these two parameters for each indication allow a point to be plotted directly on the appropriate evaluation chart. Their location on the chart determines the acceptability immediately, through the end of the evaluation period identified.

Eight flaw evaluation charts were prepared for the region of interest, four for axial flaws, and four for circumferential flaws. The cases covered are listed below:

Figure 6-13: Maximum Acceptable Initial Axial Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 10% Inner Diameter Weld Repair, with No PWHT Figure 6-14: Maximum Acceptable Initial Axial Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 10% Inner Diameter Weld Repair, with PWHT Figure 6-15: Maximum Acceptable Initial Axial Flaws, Accounting for PWSCC and Fatigue Crack Growth, With a 25% Inner Diameter Weld Repair, with PWHT Figure 6-16: Maximum Acceptable Initial Axial Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 50% Inner Diameter Weld Repair, with PWHT WCAP-17128-NP May 2010 Revision 1

6-8 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Figure 6-17: Maximum Acceptable Initial Circumferential Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 10% Inner Diameter Weld Repair, with No PWHT Figure 6-18: Maximum Acceptable Initial Circumferential Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 10% Inner Diameter Weld Repair, with PWHT Figure 6-19: Maximum Acceptable Initial Circumferential Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 25% Inner Diameter Weld Repair, with PWHT Figure 6-20: Maximum Acceptable Initial Circumferential Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 50% Inner Diameter Weld Repair, with PWHT In summary, results show very large flaws are acceptable for service periods up to four years.

These results include the required Section XI [1] flaw evaluation margins and were developed for both axial and circumferentially oriented flaws.

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 6-9 Figure 6-1: Axisymmetric FEA Model for Transient Stress Analysis WCAP-17128-NP May 2010 Revision 1

6-10 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Figure 6-2: Alloy 82/182 Weld Fatigue Crack Growth Rate Properties in a PWR Environment WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 6-11 60 40 20 Stress (ksi) 0

-20

-40 Axial without Weld Repair Axial with Weld Repair

-60 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Thru-wall Distance Ratio x/t Figure 6-3: Axial Residual Stresses for RCP Suction and Discharge Nozzles [6]

WCAP-17128-NP May 2010 Revision 1

6-12 WESTINGHOUSE NON-PROPRIETARY CLASS 3 KI vs. Crack Angle Ratio - Discharge Nozzle 700 Distance from ID x/t Ratio 600 0 0.25 0.5 500 0.75 1 y = 11895x3 - 7174.2x2 + 1823.2x + 48.763 Maximum 400 Average Poly. (Maximum)

KI (ksiin)

Poly. (Average) 300 200 3 2 y = 4770.8x - 1722.1x + 795.48x + 16.469 100 0

0.0 0.1 0.2 0.3 0.4 0.5 Crack Length Ratio (/ or c/Rm)

Figure 6-4: Crack Tip Stress Intensity versus Circumferential Through-wall Crack Length Used for PWSCC Growth Evaluation WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 6-13 30 Initial Total Crack Length 3.110 25 3.446 4.529 11.404 20 Half Crack Length (in.)

15 10 5

0 0 5 10 15 20 25 Time Period (years)

Figure 6-5: PWSCC Only Growth of Circumferential Through-wall Flaws with Maximum Normal Operating Nozzle Axial Loads for Various Initial Lengths WCAP-17128-NP May 2010 Revision 1

6-14 WESTINGHOUSE NON-PROPRIETARY CLASS 3 400 Distance from ID x/t Ratio 350 Heatup KI-min y = 3040.8x3 - 1147.8x2 + 640.54x + 17.776 Heatup KI-max 300 Poly. (Heatup KI-min)

Poly. (Heatup KI-max) 250 KI (ksiin) 200 150 100 y = 331.12x3 - 114.9x2 + 52.165x + 1.0806 50 0

0.0 0.1 0.2 0.3 0.4 0.5 Crack Length Ratio (/ or c/Rm)

Figure 6-6: Maximum and Minimum Through-wall Crack Tip Stress Intensity Factors during a Heatup Transient as a Function of Circumferential Crack Length WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 6-15 25 Initial Total Crack Length (in) 1.981 3.110 20 3.446 4.529 11.404 Half Crack Length (in.)

15 10 5

0 0 5 10 15 20 25 Time Period (years)

Figure 6-7: Fatigue Only Growth of Circumferential Through-wall Flaws with Maximum Normal Operating Nozzle Axial Loads for Various Initial Lengths WCAP-17128-NP May 2010 Revision 1

6-16 WESTINGHOUSE NON-PROPRIETARY CLASS 3 PWSCC + Fatigue Crack Growth - VB 25 Initial Total Crack Length (in.)

1.98 3.110 20 3.446 4.529 11.40 Half Crack Length (in.)

15 10 5

0 0 5 10 15 20 Time Period (years)

Figure 6-8: Combined PWSCC and Fatigue Growth of Circumferential Through-wall Flaws with Maximum Normal Operating Nozzle Axial Loads for Various Initial Lengths WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 6-17 1

0.9 0.8 0.7 0.6 a/t Ratio 0.5 0.4 initial a/t=0.4 0.3 initial a/t=0.45 initial a/t=0.5 0.2 initial a/t=0.55 initial a/t=0.6 initial a/t=0.65 0.1 initial a/t=0.7 0

0 5 10 15 20 25 30 35 40 Time (years)

Figure 6-9: Circumferential ID Surface FCG for Maximum Pipe Load with No Residual Stress 1

0.9 0.8 0.7 0.6 a/t Ratio 0.5 initial a/t=0.4 0.4 initial a/t=0.45 initial a/t=0.5 0.3 initial a/t=0.55 initial a/t=0.6 initial a/t=0.65 0.2 initial a/t=0.7 0.1 a = crack depth t = wall thickness 0

0 5 10 15 20 25 30 35 40 Time (year)

Figure 6-10: Circumferential ID Surface FCG for Minimum Pipe Load with No Residual Stress WCAP-17128-NP May 2010 Revision 1

6-18 WESTINGHOUSE NON-PROPRIETARY CLASS 3 1

0.9 0.8 0.7 0.6 a/t Ratio 0.5 0.4 0.3 initial a/t=0.5 initial a/t=0.55 0.2 initial a/t=0.6 initial a/t=0.65 0.1 initial a/t=0.7 0

0 5 10 15 20 25 30 35 40 Time (year)

Figure 6-11: Circumferential ID Surface FCG for Maximum Pipe Load with Residual Stress, No ID Repair 0.9 0.8 0.7 0.6 0.5 a/t Ratio 0.4 initial a/t=0.6 0.3 initial a/t=0.65 initial a/t=0.7 0.2 initial a/t=0.75 initial a/t=0.8 0.1 a = crack depth t = wall thickness 0

0 5 10 15 20 25 30 35 40 Time (years)

Figure 6-12: Circumferential ID Surface FCG for Minimum Pipe Load with Residual Stress, No ID Repair WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 6-19 1

Time (months) to Reach ASME Allowable Crack Depth 0.9 24 months 10% Repair with HT 0.8 36 months 10% Repair with HT Initial Crack Depth / Thickness Ratio, a/t 0.7 48 months 10% Repair with HT 0.6 0.5 0.4 0.3 0.2 0.1 0

0 0.1 0.2 0.3 0.4 0.5 Crack Depth / Length Ratio, a/

Figure 6-13: Maximum Acceptable Initial Axial Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 10% Inner Diameter Weld Repair, with No PWHT WCAP-17128-NP May 2010 Revision 1

6-20 WESTINGHOUSE NON-PROPRIETARY CLASS 3 1

Time (months) to Reach ASME Allowable Crack Depth 0.9 0.8 24 months 10% Repair no HT Initial Crack Depth / Thickness Ratio, a/t 36 months 10% Repair no HT 0.7 48 months 10% Repair no HT 0.6 0.5 0.4 0.3 0.2 0.1 0

0 0.1 0.2 0.3 0.4 0.5 Crack Depth / Length Ratio, a/

Figure 6-14: Maximum Acceptable Initial Axial Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 10% Inner Diameter Weld Repair, with PWHT WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 6-21 1

Time (months) to Reach ASME Allowable Crack Depth 0.9 0.8 24 months 25% Repair no HT Initial Crack Depth / Thickness Ratio, a/t 36 months 25% Repair no HT 0.7 48 months 25% Repair no HT 0.6 0.5 0.4 0.3 0.2 0.1 0

0 0.1 0.2 0.3 0.4 0.5 Crack Depth / Length Ratio, a/

Figure 6-15: Maximum Acceptable Initial Axial Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 25% Inner Diameter Weld Repair, with PWHT WCAP-17128-NP May 2010 Revision 1

6-22 WESTINGHOUSE NON-PROPRIETARY CLASS 3 1

Time (months) to Reach ASME Allowable Crack Depth 0.9 0.8 24 months 50% Repair no HT Initial Crack Depth / Thickness Ratio, a/t 36 months 50% Repair no HT 0.7 48 months 50% Repair no HT 0.6 0.5 0.4 0.3 0.2 0.1 0

0 0.1 0.2 0.3 0.4 0.5 Crack Depth / Length Ratio, a/

Figure 6-16: Maximum Acceptable Initial Axial Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 50% Inner Diameter Weld Repair, with PWHT WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 6-23 1

0.9 0.8 Initial Crack Depth / Thickness Ratio, a/t 0.7 0.6 0.5 0.4 Time (months) to Reach ASME Allowable Crack Depth 0.3 0.2 24 36 0.1 48 0

0 0.1 0.2 0.3 0.4 0.5 Crack Depth / Length Ratio, a/

Figure 6-17: Maximum Acceptable Initial Circumferential Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 10% Inner Diameter Weld Repair, with No PWHT WCAP-17128-NP May 2010 Revision 1

6-24 WESTINGHOUSE NON-PROPRIETARY CLASS 3 1

0.9 0.8 Initial Crack Depth / Thickness Ratio, a/t 0.7 0.6 0.5 0.4 Time (months) to Reach ASME Allowable Crack Depth 0.3 0.2 24 36 0.1 48 0

0 0.1 0.2 0.3 0.4 0.5 Crack Depth / Length Ratio, a/

Figure 6-18: Maximum Acceptable Initial Circumferential Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 10% Inner Diameter Weld Repair, with PWHT WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 6-25 1

0.9 0.8 Initial Crack Depth / Thickness Ratio, a/t 0.7 0.6 0.5 0.4 Time (months) to Reach ASME Allowable Crack Depth 0.3 0.2 24 36 0.1 48 0

0 0.1 0.2 0.3 0.4 0.5 Crack Depth / Length Ratio, a/

Figure 6-19: Maximum Acceptable Initial Circumferential Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 25% Inner Diameter Weld Repair, with PWHT WCAP-17128-NP May 2010 Revision 1

6-26 WESTINGHOUSE NON-PROPRIETARY CLASS 3 1

0.9 0.8 Initial Crack Depth / Thickness Ratio, a/t 0.7 0.6 0.5 0.4 Time (months) to Reach ASME Allowable Crack Depth 0.3 0.2 24 36 0.1 48 0

0 0.1 0.2 0.3 0.4 0.5 Crack Depth / Length Ratio, a/

Figure 6-20: Maximum Acceptable Initial Circumferential Flaws, Accounting for PWSCC and Fatigue Crack Growth, with a 50% Inner Diameter Weld Repair, with PWHT WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 7-1 7 ADVANCED PWSCC GROWTH BY FEA Flaw evaluations of CE design RCP outlet nozzle Alloy 82/182 dissimilar metal butt welds were also performed using a rigorous three-dimensional finite element model containing a circumferential flaw. Both, finite depth and length inside surface circumferential flaws, as well as through-wall flaws, were considered in the evaluation. The purpose of the inside surface flaws was to assess the time period for the flaw to grow to an acceptable depth per Section XI [1], and also to grow through the wall and reach the outside surface. The through-wall circumferential case was to compute the time period required for a maximum obstruction flaw length to grow in the circumferential direction and reach a critical flaw length. Only the PWSCC growth mechanism was considered in this three-dimensional finite element analysis (FEA) flaw evaluation. In all the evaluations, the goal was to generate a realistic crack growth assessment, for comparison with the traditional methods used elsewhere in the project, and reported in Sections 5 and 6.

7.1 INITIAL FLAW SIZE The initial inside surface finite depth flaw lengths considered in this evaluation are a 14% and 23% of circumference of the nozzle. The 14% flaw represents, conservatively, the largest single obstruction from the charging or spray nozzle and accounts for the inspection transducer width on either side of the nozzle. Surface flaw depths of 20% and 30% were assumed. These depths were chosen based on very conservative aspect ratios of 0.04 and 0.03, respectively. These are significantly larger than the aspect ratio of 0.1667 observed in service experience, and it is highly likely that any flaws deeper than this would have tails which would be detected in the inspected region. Since the finite element analysis software does not allow surface flaws to grow to through-wall as a continuous crack growth process due to mesh changing restrictions, the surface flaw was first allowed to grow through the wall and almost reach the outside surface with depths exceeding 90% of the wall thickness. Then, the flaw was assumed to be through the wall.

Subsequent three-dimensional FEA evaluations consisted of a through-wall flaw with different inside and outside lengths that simulated the end of surface flaw growth, which was then allowed to propagate around the circumference to reach the critical flaw length. Total service life was then obtained by addition of the time periods from the ID surface flaw to reach the outside surface and then propagate in the circumferential direction.

7.2 STRESS INTENSITY FACTOR CALCULATION SIF calculations in the FEACrack software program are performed by using crack tip finite elements and strain-energy contour integrals around the crack front. The fracture mechanics model geometry is generated using FEACrack. The five different cases completed are listed in Table 7-1. The model geometry, model external loads, and initial flaw sizes are defined in FEACrack software input parameters. Using this information, the software generates three-dimensional FEA models with surface or through-wall cracks for crack growth with a continuously moving crack front and prepares an input mesh to ANSYS for the finite element solution.

WCAP-17128-NP May 2010 Revision 1

7-2 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Once the FEA model with flaws is analyzed by ANSYS, the FEACrack program processes the results for crack tip SIFs along the crack front. The SIFs are obtained using linear elastic J-integral and KI relationships.

Initial surface flaw depths considered a range between 20% and 30% of the wall thickness. This will address the range in depth of the ID flaws that may have been missed during an in-service inspection. These initial flaw depths were analyzed to determine the time period for the flaw to reach the allowable depth per Section XI [1], and then to penetrate the nozzle wall. A semi-elliptical surface flaw in the circumferential direction was assumed for the crack front profile and allowed to grow based on the crack front KI values. The four surface flaw cases were:

1. 14% length 20% depth inside surface circumferential flaw,

2. 14% length, 30% depth inside surface circumferential flaw,

3. 23% length 20% depth inside surface circumferential flaw, and

4. 23% length, 30% depth inside surface circumferential flaw.

The fifth flaw considered in this analysis was a through-wall flaw with an initial flaw length of 14% of the nozzle circumference, resembling the shape of the last step of the flaw shape from flaw Case 2, mentioned above. Details of the flaw depths and lengths are listed in Table 7-1.

7.3 FINITE ELEMENT FRACTURE MECHANICS MODEL FEACrack was used to generate all the finite element fracture mechanics models analyzed. A typical FEA model is shown in Figure 7-1. All surface flaw cases evaluated in this study were based on the same set of parameters, for ANSYS eight-noded solid element type SOLID45. The FEA mesh parameters for the through-wall case vary slightly from those of the surface flaw cases to accommodate the differences in the flaw shapes. An appropriate axial length of the piping was included in the model to minimize the boundary effects on the dissimilar metal weld location.

7.4 BOUNDARY CONDITIONS Each FEA model developed has a quarter-symmetry with the center of the dissimilar metal weld taken as a symmetry plane and the other one along the nozzle axis. The boundary conditions prescribed on the symmetry planes are shown in Figure 7-2. The DM weld crack symmetry plane is fixed along the axial x-direction of the nozzle. The nozzle axial symmetry plane has a fixed boundary condition along the circumferential z-direction. These boundary conditions are automatically assigned within FEACrack by specifying a quarter symmetric pipe model.

FEACrack automatically applies a fixed boundary condition at an appropriate node in the y-direction to prevent rigid body motion.

Fabrication welding residual stresses from MRP-113 [6] show the large diameter pipes typical of the CE fleet cold leg nozzles are compressive in nature in the 15% to 40% through-wall distance from the inside surface. As the magnitude of this compressive residual stress is high, in the range of 20 to 50 ksi, any crack growth in the radial direction in the FEACrack program is prevented, and the crack propagation stops. Since the intent of this study on the propagation of inside WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 7-3 surface cracks is to determine the time period to reach through-wall thickness, fabrication residual stresses were ignored. Only the crack face pressure of 2.5 ksi due to internal pressure loads were applied. The initial fabrication residual stresses were also ignored for the through-wall Case 3, for consistency.

7.5 NOZZLE END AXIAL LOADS An axial force and bending moment loading on the nozzle free-end surface were applied in the FEACrack model as shown in Figure 7-3 and Figure 7-4. These loadings were applied to the FEA models through the element face pressures. Based on the free-end surface element orientation, element face pressures are automatically calculated by FEACrack and applied to the appropriate elements.

7.6 PWSCC CRACK GROWTH WITH FEACRACK PROGRAM All evaluation cases considered in this study are summarized in Table 7-1. Figure 7-5, Figure 7-7, Figure 7-9, Figure 7-11, and Figure 7-13 show the crack front shape plots for each case as a function of time. As the fabrication weld residual stresses were ignored due to their compressive nature near the ID surface, the crack front shapes maintain their shape close to the initial elliptical shape. In all the surface flaw cases, the crack fronts grow significantly in the radial direction with minimal growth occurring in the circumferential direction. The presence of residual stress may change this trend, but for deeper cracks only, as the shallower ones have compressive residual stresses.

The total amount of time to reach the critical flaw size is determined by adding the amount of time shown for an internal surface flaw to reach through-wall and the amount of time for the through-wall crack to reach the critical flaw size.

According to Cases 1 and 3, the amount of time it takes for an internal surface flaw with a length equal to 14% of the circumference and depth equal to 20% of the wall to reach through the thickness is 12.5 years (Case 1). An additional 8.5 years is required for the flaw to grow circumferentially to reach the critical crack length (Case 3) with total time equaling 21 years.

Times for various initial flaw sizes can be inferred from Figure 7-5 and Figure 7-7.

An inside surface flaw with 23% circumferential length and 20% through the wall depth, shown in Figure 7-11, takes approximately 11 years to reach through-wall. Once this flaw reaches the outside surface, the resulting through-wall flaw propagates circumferentially to reach the critical flaw length within a very short time, so the total time to critical length is about equal to the time to penetrate the wall.

Crack tip stress intensity factors are plotted in Figure 7-6, Figure 7-8, Figure 7-10, Figure 7-12, and Figure 7-14 for Cases 1 through 5, respectively. These plots show a variation of SIFs during the crack growth period at various time steps. Trends in these plots show the SIFs are low near the ID surface, and hence causes very slow growth along the circumferential direction. For the through-wall Case 3 shown in Figure 7-10, the SIF distribution is high at the ID surface, due to the flaw shape assumption and quickly evens out, indicating that the flaw shape will approach radial through-wall shape and then grows more uniformly. This is seen by the approximately WCAP-17128-NP May 2010 Revision 1

7-4 WESTINGHOUSE NON-PROPRIETARY CLASS 3 parallel crack fronts in Figure 7-9. All the FEACrack analyses assumed crack growth based on local SIFs.

Table 7-1: Initial Flaw Dimensions for Three-Dimensional FEA PWSCC Analyses Flaw Length Flaw Depth Length Depth Case

(% Circumference) (% Wall Thickness) (in) (in) 1 14 20 13.2 0.6 2 14 30 13.2 0.9 3 14 Through-wall 14.5 Through-wall 4 23 20 21.7 0.6 5 23 30 21.7 0.9 Notes:

All three-dimensional FEAs were performed with RCP discharge nozzle geometry with a nominal pipe geometry having an inside radius of 15 inches and a wall thickness of 3 inches.

The through-wall flaw length on the inside surface for Case 3 was assumed to be the same as that at the end of the flaw growth for surface flaw Case 2.

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 7-5 Crack Face Free-End Axial Force and Moment Load Symmetry Boundary Conditions Figure 7-1: Finite Element Fracture Mechanics Model WCAP-17128-NP May 2010 Revision 1

7-6 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Applied Crack Face Pressure Loading Figure 7-2: Crack-face End View of Applied Crack Face Pressures WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 7-7 Free-End Moment +

Axial Force Loading Figure 7-3: Applied Free-end Pressures (for Moment plus Axial Force)

WCAP-17128-NP May 2010 Revision 1

7-8 WESTINGHOUSE NON-PROPRIETARY CLASS 3 Figure 7-4: Rotated View of Applied Free-end Pressures (for Moment plus Axial Force)

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 7-9 Figure 7-5: PWSCC Flaw Growth with Initial ID Surface Flaw of 14% Circumferential, 20% Depth, Case 1 Note: EFPY = Effective Full Power Years WCAP-17128-NP May 2010 Revision 1

7-10 WESTINGHOUSE NON-PROPRIETARY CLASS 3 90 0 EFPY 3.3 EFPY 80 5.8 EFPY 7.7 EFPY 70 9.3 EFPY 10.6 EFPY 60 11.7 EFPY 12.5 EFPY 50 K (ksiin) 40 30 20 10 0

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Normaized Crack Front Position Figure 7-6: SIFs along Crack Front for ID Surface Flaws during PWSCC Growth with Initial Flaw of 14% Circumferential, 20% Depth, Case 1 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 7-11 Figure 7-7: PWSCC Flaw Growth with Initial ID Surface Flaw of 14% Circumferential, 30% Depth, Case 2 WCAP-17128-NP May 2010 Revision 1

7-12 WESTINGHOUSE NON-PROPRIETARY CLASS 3 100 0 EFPY 90 2.7 EFPY 4.8 EFPY 80 6.4 EFPY 70 7.8 EFPY 8.9 EFPY 60 9.4 EFPY K (ksiin) 50 40 30 20 10 0

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Normaized Crack Front Position Figure 7-8: SIFs along Crack Front for ID Surface Flaws during PWSCC Growth with Initial Flaw of 14% Circumferential, 30% Depth, Case 2 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 7-13 Figure 7-9: PWSCC Flaw Growth with Initial Through-wall Flaw of 14% Circumferential, Case 3 WCAP-17128-NP May 2010 Revision 1

7-14 WESTINGHOUSE NON-PROPRIETARY CLASS 3 300 250 200 K (ksiin) 150 100 0 EFPY 1.5 EFPY 3.2 EFPY 50 4.9 EFPY 6.3 EFPY 7.5 EFPY 8.4 EFPY 0

0 0.5 1 1.5 2 2.5 3 Crack Depth from ID (inches)

Figure 7-10: SIFs along Crack Front during PWSCC Flaw Growth with Initial Through-wall Flaw of 14% Circumferential, Case 3 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 7-15 Figure 7-11: ID Surface PWSCC Flaw Growth with Initial Flaw Size of 23% Circumferential, 20% Depth, Case 4 WCAP-17128-NP May 2010 Revision 1

7-16 WESTINGHOUSE NON-PROPRIETARY CLASS 3 120 0 Years 3.1 EFPY 100 5.4 EFPY 7.1 EFPY 8.4 EFPY 80 9.5 EFPY 10.4 EFPY 11.1 EFPY K (ksiin) 60 40 20 0

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Normaized Crack Front Position Figure 7-12: SIFs along Crack Front for ID Surface Flaws during PWSCC Growth with Initial Flaw of 23% Circumferential, 20% Depth, Case 4 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 7-17 Figure 7-13: ID Surface PWSCC Flaw Growth with Initial Flaw of 23% Circumferential, 30% Depth, Case 5 WCAP-17128-NP May 2010 Revision 1

7-18 WESTINGHOUSE NON-PROPRIETARY CLASS 3 120 0 EFPY 2.5 EFPY 100 4.3 EFPY 5.6 EFPY 6.7 EFPY 80 7.6 EFPY 7.9 EFPY K (ksiin) 60 40 20 0

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Normaized Crack Front Position Figure 7-14: SIFs along Crack Front for ID Surface Flaws during PWSCC Growth with Initial Flaw of 23% Circumferential, 30% Depth, Case 5 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 8-1 8 PROBABILITY OF CRACKS 8.1 PURPOSE The purpose of the probabilistic analysis was to assess the susceptibility of CE reactor coolant pump suction and discharge nozzles to PWSCC. The analysis considers available industry experience with the locations of Alloy 82/182 DM welds. More specifically, information included in the analysis included Alloy 82/182 DM welds that were nominally 28 inches in diameter or larger at the:

1. Reactor vessel inlet and outlet nozzles,

2. Steam generator inlet and outlet nozzles,

3. Reactor coolant pump suction and discharge nozzles, and

4. Pressurizer surge nozzle.

8.2 DESCRIPTION

OF CALCULATION METHODOLOGY The following process was used to calculate Weibull parameters and the corresponding probabilities of flaw indications.

Locations utilized in this analysis where large (greater than ~28 in diameter with ~3 wall thickness). These locations included plants with relevant Alloy 600 and Alloy 82/182 DM welds but varied by availability of data.

Locations where adjusted to effective full power years (EFPY), based on the plants capacity factor.

EFPY Agecalendar years Capacity Factor To further reduce the variation between locations, the EFPYs were transformed to effective degradation years (EDY) using the following formula:

Q1 1 1 EDY EFPY exp R ActualTemp ( F ) 459.7 Reference Temp(F) 459 .7 where, R = 1.103E-03 kcal/mole-R, and Q1 = 50 kcal/mole.

To situate the locations as like-kind inputs, the final step is to adjust each flaws percent through-wall to the same depth, which was chosen as 7% of the wall thickness (7% tw). This depth was more or less arbitrary, but does correspond to the smallest depth of PWSCC flaw discovered in-service. To make this adjustment, an estimate of the time from 7% tw to the discovered depth in each component in the database was calculated. This time, in EFPY, was then Arrhenius temperature adjusted for the temperature of the component and subtracted from EDY at WCAP-17128-NP May 2010 Revision 1

8-2 WESTINGHOUSE NON-PROPRIETARY CLASS 3 discovery. Resulting EDY value are seen as the best estimate as to when the flaw might have been at 7% tw.

Once the database was established and corrected for the fixed depth, the Weibull model was complete. It was then used to predict the probability of a flaw existing at the 7% tw depth. Three temperatures were selected for the analysis with the intent of covering the range of temperatures on the cold nozzle DM weld locations (548°F to 556°F), as well as a representative hot nozzle DM weld location (615°F). Results are presented in terms of the cumulative probability of a flaw with depth equal to 7% of the wall thickness, as a function of time, in EFPY, up to 60 EFPY.

The Weibull shape and scale parameters were generated using the Maximum Likelihood Estimation:

1 x i Equation 8-1 i

r where,

= scale,

= shape, r = number of failures, and X = EDY of the ith location.

Since both the shape and scale are unknown, a goal seek method is used to estimate the shape parameter. The method calculated the shape parameter when given a range of values for the scale parameter until they collectively best fit the input data. This method includes a reduced bias adjustment on the shape parameter.

Given the resulting Weibull shape and scale parameters, cumulative probabilities can be calculated using:

X 1 exp where, X = EFPY 8.3 IMPORTANT ASSUMPTIONS Serving the intent of the project, certain conservative assumptions have been made, such that portions of the analysis are not considered to be best estimate assumptions. The major assumption is the cracking data inputs from all the large DM weld locations are part of the same family with regard to cracking susceptibility. Therefore, all are relevant to be incorporated into the generation of Weibull shape and scale parameters. A reasonable counter argument can be made to this assumption, in that the different nozzle DM weld locations differ in one or more WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 8-3 characteristics, including: the manufacturer, stress profile, surface finishing, and any applied mitigations, such as zinc addition. These differences were ignored, so all large nozzle DM weld indications could be utilized in the analysis. The judgment was made due to the additional confidence obtained by using this larger database outweighed the uncertainties resulting from the differences discussed above. These assumptions were verified by running separate cases, including multiple sets of data. The results showed independent of which inputs were included, the results for the cold leg temperature nozzles were not significantly changed.

A 7% through-wall flaw was assumed to be the smallest detectable flaw by performance demonstration initiative (PDI) qualified inspections. The accompanying figures show the probability of finding an indication at a 7% through-wall flaw. Multiple through-wall flaws of approximately 6% to 7% were found in the steam generator inlet nozzle DM welds in Japan.

8.4 RESULTS The results summarized in Table 8-1, and shown graphically in Figure 8-1, Figure 8-2, and Figure 8-3, correspond to the different combinations of data discussed above. Figure 8-1 shows the probability of cracking for the pump nozzle DM welds, based on all the available inspection results, for reactor vessel nozzles, steam generator nozzles, pump nozzles, and pressurizer surge nozzles; this has been called Case 1. The next case, Case 2, includes all the nozzles except the pressurizer nozzles, and Case 3 includes only the reactor vessel and RCP nozzles.

The results show there is no discernable difference between the cases, with the probability of cracking for the pump nozzle DM welds being extremely low, even at 60 EFPY. Results indicate that even though DM welds have had many flaws at hot temperature locations, none have been found at cold temperature butt weld locations, and this gives a very low probability of flaws existing in cold temperature locations. Results in Table 8-1 show the highest probability of an indication was only 1.42%, at 60 EFPY (Case 1 at 556°F). A 60 EFPY value is well beyond a plants licensed life, even with a 20-year life extension.

Table 8-1: Summary Results Table At EFPY Case 1 Case 2 Case 3 Temperature 548°F 20 0.25% 0.00% 0.01%

40 0.57% 0.03% 0.05%

60 0.93% 0.12% 0.15%

Temperature 556°F 20 0.38% 0.01% 0.02%

40 0.88% 0.10% 0.13%

60 1.42% 0.35% 0.35%

Temperature 615°F 20 6.98% 20.92% 9.84%

40 15.32% 86.63% 44.34%

60 23.71% 99.92% 80.10%

WCAP-17128-NP May 2010 Revision 1

8-4 WESTINGHOUSE NON-PROPRIETARY CLASS 3 All Avaliable Large DM Weld Inspection Results (@7% tw) 100%

90%

Weibull Parameters 80% Shape: 1.2 Scale: 324 EDY Weibull Cumulative Probability 70%

60%

50%

of a 7% tw flaw 615 40% 556 548 30%

20%

10%

0%

0 10 20 30 40 50 60 Effective Full Power Years (EFPY)

Figure 8-1: All Available Large DM Weld Inspection Results (7% Through-wall) - Case 1 RV, RCP, and SG Large DM Weld Inspection Results (@7% tw) 100%

90%

Weibull Parameters 80% Shape: 3.1 Scale: 58 EDY Weibull Cumulative Probability 70%

60%

50%

of a 7% tw flaw 615 40% 556 548 30%

20%

10%

0%

0 10 20 30 40 50 60 Effective Full Power Years (EFPY)

Figure 8-2: All Available Large DM Weld Inspection Results (7% Through-wall) - Case 2 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 8-5 RV and RCP Large DM Weld Inspection Results (@7% tw) 100%

90%

Weibull Parameters 80% Shape: 2.5 Scale: 90 EDY Weibull Cumulative Probability 70%

60%

50%

of a 7% tw flaw 615 40% 556 548 30%

20%

10%

0%

0 10 20 30 40 50 60 Effective Full Power Years (EFPY)

Figure 8-3: All Available Large DM Weld Inspection Results (7% Through-wall) - Case 3 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 9-1 9 PROPOSED CODE CHANGE The inspection of Alloy 182/82 DM welds is presently being performed to the requirements of report MRP-139, Revision 1 [2]. These inspection requirements will be replaced by those of Code Case N-770 (see Appendix A), beginning in fall 2010, or shortly thereafter. MRP-139 [2]

contains a provision which allows for a flaw tolerance calculation to justify the acceptability of inspection coverage less than the required 90%.

It is essential to revise Code Case N-770 to include a similar provision, and the work documented in this report forms the technical basis for such a revision. It is important to understand the locations for mitigation are practical and have, for the most part, already been mitigated, or are planned to be mitigated. A few regions, such as the pump nozzles of the CE fleet, do not lend themselves to mitigation, and a reasonable solution is to continue inspections at a frequency determined by the flaw tolerance of the region. In this case, the nozzles operate at cold leg temperatures and the probability of flaws is small. Any propagation from an existing flaw is also very slow, so the flaw tolerance is high. The results in this report suggest a ten year inspection frequency is justifiable for these regions.

The proposed change is shown below. There will be an additional sub-paragraph added under paragraph 2500 of the Code Case. The existing Code Case is reproduced as Appendix A of this report.

Proposed Revision to N-770 for Cold Leg Locations Add Para -2500 (d):

For piping with diameters greater than or equal to 14 inches (355 mm), in locations with operating temperatures of less than 570ºF (299ºC), and where inspection coverage is limited by permanent obstructions, the following inspection coverage requirements of this case may be used in place of -2500(c):

(a) For axially oriented flaws, achieve the maximum coverage possible, and document any limitations, provided 90% coverage of the circumference is achieved.

(b) For circumferentially oriented flaws, achieve the maximum coverage possible, and document any limitations.

(c) If the coverage achieved in either (a) or (b) is less than 90%, perform the following flaw tolerance evaluations:

a. Postulate a through-wall flaw in the region where inspection coverage is obstructed, with length equal to that which would yield the minimum detectable leakage for the plant.

Calculate the critical through-wall length using IWB-3640, and show the time for the postulated flaw to reach a critical length is longer than the time to the next inspection, and WCAP-17128-NP May 2010 Revision 1

9-2 WESTINGHOUSE NON-PROPRIETARY CLASS 3

b. Postulate a part-through semi-elliptic surface flaw in the region where inspection coverage is obstructed, with depth equal to 20% of the wall thickness, and length equal to the length of the largest obstruction. Calculate the Section XI allowable flaw depth using IWB- 3640, and show the time for the postulated flaw to reach the allowable size is longer than the time to the next inspection.

(d) If 90% coverage is not achieved for either axial or circumferential flaws, VT-2 examinations of the region are required during each refueling outage.

(e) If 90% coverage is not achieved for either axial or circumferential flaws, document the likelihood of leakage occurring at the location of interest between inspections, and document leakage monitoring action levels.

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 10-1 10 REFERENCES

1. ASME Boiler and Pressure Vessel Code Section XI, Rules for Inservice Inspection of Nuclear Power Plant Components, 2007 Edition with 2009 Addenda, July 1, 2009.

2. Material Reliability Program: Primary System Piping Butt Weld Inspection and Evaluation Guideline (MRP-139), EPRI, Palo Alto, CA: 2005. 1010087.

3. Westinghouse Project Plan, PP-MRCDA-09-2, Rev. 0, PWROG PA-MSC-00525 Fracture Mechanics Evaluation of CE Design Pump DM Welds - Phase 2, July 30, 2009.

4. Westinghouse Letter, OG-09-154, Rev. 0, PWR Owners Group Information Needed to Support Fracture Mechanics Evaluation of CE Design Pump DM Welds - Phase 2, PA-MSC-0525, April 14, 2009.

5. Westinghouse Calculation Note, CN-MRCDA-08-37, Rev. 0, Reactor Coolant Pump and Safety Injection Nozzle Design Information for Fracture Mechanics Evaluation of CE-Designed DM Welds, June 16, 2008.

6. Materials Reliability Program: Alloy 82/182 Pipe Butt Weld Safety Assessment for US PWR Plant Designs (MRP-113), EPRI, Palo Alto, CA: 2005. 1009549.

7. Fauske, H. K., Critical Two Phase, Steam Water Flows, Proceedings of the Heat Transfer and Fluid Mechanics Institute, Stanford, California, Stanford University Press, 1961.

8. M. M, El-Wakil, Nuclear Heat Transport, International Textbook Company, New York, N.Y, 1971.

9. U.S. Nuclear Regulatory Commission, The Application of Fracture Proof Design Methods Using Tearing Instability Theory to Nuclear Piping Postulating Circumferential Through Wall Cracks, Section II-1, The Effects of Shell Corrections on Stress Intensity Factors and the Crack Opening Area of Circumferential and a Longitudinal Through-Crack in a Pipe. NUREG/CR 3464. September 1983.

10. Westinghouse Letter, NSASD-STD-94-020, Rev. 0, Configuration Control of Programs FHG and MPBK January 5, 1994.

11. Westinghouse Letter, LTR-SST-03-44, Rev. 0, FHG Version 1, LIMIT3 Version 1, MPBK Version 1, PC-SIGEP Version 1.00 and WECRACK Version 1.2, Installation Testing on the Windows XP System State, July 29, 2003.

12. U.S. Nuclear Regulatory Commission, Barrier Integrity Research Program, Section 3, Review of RCS Leakage Experiments and Leak-Rate Models. NUREG/CR-6861. December 2004.

13. Materials Reliability Program (MRP) Crack Growth Rates for Evaluating Primary Water Stress Corrosion Cracking (PWSCC) of Alloy 82, 182, and 132 Welds (MRP-115), EPRI, Palo Alto, CA:

2004. 1006696.

14. Zang, W., Stress Intensity Factor Solutions for Axial and Circumferential Through-wall Cracks in Cylinders, SAQ Report SINTAP/SAQ/02, SAQ Kontroll AB, 1997.

15. Westinghouse Letter, LTR-PAFM-08-17, Rev. 0, Software Release Letter for WESTEMS' 4.5.1, February 1, 2008.

WCAP-17128-NP May 2010 Revision 1

10-2 WESTINGHOUSE NON-PROPRIETARY CLASS 3

16. Westinghouse Letter, LTR-SST-06-21, Rev. 0, Release of ANSYS 10 for XP, HPUX 11.0, and HPUX 11.23 and ANSYS Error Reports, July 12, 2006.

17. ASME Boiler and Pressure Vessel Code Section II - Materials, Part D - Properties, 2001 Edition through 2003 Addenda.

18. Westinghouse Calculation Note, CN-PAFM-08-24, Rev. 0, WES_FRAMES Version 4.1 Change Specification and Validation, April 22, 2008.

19. Raju, I. S. and Newman, Jr., J. C., Stress Intensity Factor Influence Coefficients for Internal and External Surface Cracks in Cylindrical Vessels, ASME Pressure Vessels and Piping, Volume 58, 1982, pages 37-48.

20. NASA Lyndon B. Johnson Space Center Report, NASA-TM-111707, Stress Intensity Factors for Part-Through Surface Cracks in Hollow Cylinders, July 1992.

21. U.S. Nuclear Regulatory Commission, Effects of Alloy Chemistry, Cold Work, and Water Chemistry on Corrosion Fatigue and Stress Corrosion Cracking of Nickel Alloys and Welds. NUREG/CR-6721, ANL-01/07. April 2001.

22. Westinghouse Report, WCAP-16925-NP, Rev. 1, Flaw Evaluation of CE Design RCP Suction and Discharge, and Safety Injection Nozzle Dissimilar-Metal Welds," August 25, 2009.

23. Ogawa, N. et al, "Residual Stress Evaluation of Dissimilar Weld Joint Using Reactor Vessel Outlet Nozzle Mock-up Model (Report-2)" Paper # PVP2009-77269 in Proceedings of ASME Pressure Vessels & Piping Conference, July 26-30, 2009, Prague, Czech Republic.

24. Westinghouse Report, WCAP-16456-NP, Rev. 0, Pressurized Water Reactor Owners Group Standard RCS Leakage Action Levels and Response Guidelines for Pressurized Water Reactors, October 2, 2006.

25. U.S. Nuclear Regulatory Commission, Standard Review Plan 3.6.3, Rev. 1, "Leak-before-Break Evaluation Procedures". NUREG-0800. March 2007.

26. Boric Acid Corrosion Guidebook, Revision 1: Managing Boric Acid Corrosion Issues at PWR Power Stations. Electric Power Research Institute, Palo Alto, CA: November 2001. 1000975.

27. Boric Acid Attack of Concrete and Reinforcing Steel in PWR Fuel Handling Buildings. Electric Power Research Institute, Palo Alto, CA: June 2009. 1019168.

28. Materials Reliability Program: Advanced FEA Evaluation of Growth of Postulated Circumferential PWSCC Flaws in Pressurizer Nozzle Dissimilar Metal Welds (MRP-216, Rev. 1). Electric Power Research Institute, Palo Alto, CA: August 2007. 1015383.

29. Marlette, S. et. al., Simulation and Measurement of Through-wall Residual Stresses in a Structural Weld, Overlaid Pressurizer Nozzle, to be published in Proceedings ASME Pressure Vessels and Piping Conference, July 2010.

WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 A-1 APPENDIX A: ASME CODE CASE N-770 WCAP-17128-NP May 2010 Revision 1

A-2 WESTINGHOUSE NON-PROPRIETARY CLASS 3 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 A-3 WCAP-17128-NP May 2010 Revision 1

A-4 WESTINGHOUSE NON-PROPRIETARY CLASS 3 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 A-5 WCAP-17128-NP May 2010 Revision 1

A-6 WESTINGHOUSE NON-PROPRIETARY CLASS 3 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 A-7 WCAP-17128-NP May 2010 Revision 1

A-8 WESTINGHOUSE NON-PROPRIETARY CLASS 3 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 A-9 WCAP-17128-NP May 2010 Revision 1

A-10 WESTINGHOUSE NON-PROPRIETARY CLASS 3 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 A-11 WCAP-17128-NP May 2010 Revision 1

A-12 WESTINGHOUSE NON-PROPRIETARY CLASS 3 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 A-13 WCAP-17128-NP May 2010 Revision 1

A-14 WESTINGHOUSE NON-PROPRIETARY CLASS 3 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 A-15 WCAP-17128-NP May 2010 Revision 1

A-16 WESTINGHOUSE NON-PROPRIETARY CLASS 3 WCAP-17128-NP May 2010 Revision 1

WESTINGHOUSE NON-PROPRIETARY CLASS 3 A-17 WCAP-17128-NP May 2010 Revision 1

A-18 WESTINGHOUSE NON-PROPRIETARY CLASS 3 WCAP-17128-NP May 2010 Revision 1