ML12299A498

Amend. Revise Technical Specification 6.8.4.g, Steam Generator Program, and Technical Specification 6.9.1.7, Steam Generator Tube Inspection Report, for a Permanent Alternate Repair Criteria

ML12299A498
Person / Time
Site:	Millstone
Issue date:	12/06/2012
From:	James Kim Plant Licensing Branch 1
To:	Heacock D Dominion Nuclear Connecticut
Kim J NRR/DORL/LPL1-1 301-415-4125
References
TAC ME8455
Download: ML12299A498 (32)

Text

UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20555-0001 December 6, 2012 Mr. David A. Heacock President and Chief Nuclear Officer Dominion Nuclear Innsbrook Technical Center 5000 Dominion Boulevard Glen Allen, VA 23060-6711

SUBJECT:

MILLSTONE POWER STATION UNIT NO.3-ISSUANCE OF AMENDMENT RE: REVISE TECHNICAL SPECIFICATION 6.8.4.g, "STEAM GENERATOR (SG)

PROGRAM," AND TECHNICAL SPECIFICATIONS 6.9.1.7, "STEAM GENERATOR TUBE INSPECTION REPORT," FOR A PERMANENT ALTERNATE REPAIR CRITERIA (TAC NO. ME8455)

Dear Mr. Heacock:

The Commission has issued the enclosed Amendment No. 255 to Renewed Facility Operating License No. NPF-49 for the Millstone Power Station, Unit No.3, in response to your application dated April 12, 2012.

The amendment revises Technical Specification (TS) 6.8.4.g, "Steam Generator (SG) Program,"

to exclude a portion of the steam generator tubes below the top of the steam generator tubesheet from periodic inspections. Inclusion of the permanent alternate repair criteria (PARC) in TS 6.8.4.g permits deletion of the previous temporary alternate repair criteria (TARC) for Cycle 15. In addition, this amendment request also proposes to revise the reporting criteria in TS 6.9.1.7, "Steam Generator Tube Inspection Report," to remove reference to the previous Cycle 15 TARC, and add reporting requirements specific to the PARCo A copy of the related Safety Evaluation is also enclosed. Notice of Issuance will be included in the Commission's biweekly Federal Register notice.

Sincerely,

~~

James Kim, Project Manager Plant Licensing Branch 1-1 Division of Operating Reactor Licensing Office of Nuclear Reactor Regulation Docket No. 50-423

Enclosures:

1. Amendment No. 255 to NPF-49

2. Safety Evaluation cc w/encls: Distribution via Listserv

UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20555-0001 DOMINION NUCLEAR CONNECTICUT, INC.

DOCKET NO. 50-423 MILLSTONE POWER STATION, UNIT NO.3 AMENDMENT TO RENEWED FACILITY OPERATING LICENSE Amendment No. 255 Renewed License No. NPF-49

1.

The Nuclear Regulatory Commission (the Commission) has found that:

A.

The application for amendment by the applicant dated April 12, 2012, complies with the standards and requirements of the Atomic Energy Act of 1954, as amended (the Act), and the Commission's rules and regulations set forth in 10 CFR Chapter I; B.

The facility will operate in conformity with the application, the provisions of the Act, and the rules and regulations of the Commission; C.

There is reasonable assurance (i) that the activities authorized by this amendment can be conducted without endangering the health and safety of the public, and (ii) that such activities will be conducted in compliance with the Commission's regulations; D.

The issuance of this amendment will not be inimical to the common defense and security or to the health and safety of the public; and E.

The issuance of this amendment is in accordance with 10 CFR Part 51 of the Commission's regulations and all applicable requirements have been satisfied.

-2

2.

Accordingly, the license is amended by changes to the Technical Specifications as indicated in the attachment to this license amendment, and paragraph 2.C.(2) of Renewed Facility Operating License No. NPF-49 is hereby amended to read as follows:

(2)

Technical Specifications The Technical Specifications contained in Appendix A, revised through Amendment No. 255 and the Environmental Protection Plan contained in Appendix B, both of which are attached hereto are hereby incorporated into the license. DNC shall operate the facility in accordance with the Technical Specifications and the Environmental Protection Plan.

3.

This license amendment is effective as of the date of issuance, and shall be implemented within 60 days of issuance.

FOR THE NUCLEAR REGULATORY COMMISSION

.Jo <l.. t-l~

George A. Wilson, Chief Plant Licensing Branch 1-1 Division of Operating Reactor Licensing Office of Nuclear Reactor Regulation

Attachment:

Changes to the License and Technical Specifications Date of Issuance: December 6, 2012

ATTACHMENT TO LICENSE AMENDMENT NO. 255 RENEWED FACILITY OPERATING LICENSE NO. NPF-49 DOCKET NO. 50-423 Replace the following page of the Renewed Facility Operating License with the attached revised page. The revised page is identified by amendment number and contains marginal lines indicating the areas of change.

Remove Insert Page 4 Page 4 Replace the following pages of the Appendix A Technical Specifications, with the attached revised pages. The revised pages are identified by amendment number and contain marginal lines indicating the areas of change.

Remove Insert 6-17b 6-17b 6-17c 6-17c 6-21 6-21 6-21a 6-21a

ADMINISTRATIVE CONTROLS PROCEDURES AND PROGRAMS (Continued)

Leakage is not to exceed 500 gpd per SG.

3. The operational LEAKAGE perfonnance criterion is specified in RCS LCO 3.4.6.2, "Operational LEAKAGE."

c.

Provisions for SG tube repair criteria: Tubes found by in service inspection to contain flaws with a depth equal to or exceeding 40%

of the nominal tube wall thickness shall be plugged.

The following alternate tube repair criteria shall be applied as an alternative to the 40% depth-based criteria:

1. Tubes with service-induced flaws located greater than 15.2 inches below the top ofthe tubesheet do not require plugging.

Tubes with service-induced flaws located in the portion ofthe tube from the top ofthe tube sheet to 15.2 inches below the top ofthe tubesheet shall be plugged upon detection.

MILLSTONE - UNIT 3 6-17b Amendment No. 238, 245, 249, ~ 255

ADMINISTRATIVE CONTROLS PROCEDURES AND PROGRAMS (Continued)

d.

Provisions for SG tube inspections: Periodic SG tube inspections shall be performed. The number and portions ofthe tubes inspected and methods of inspection shall be performed with the objective of detecting flaws of any type (e.g., volumetric flaws, axial and circumferential cracks) that may be present along the length ofthe tube, from the tube-to-tubesheet weld at the tube inlet to the tube-to-tubesheet weld at the tube outlet, and that may satisfy the applicable tube repair criteria. Portions ofthe tube below 15.2 inches below the top of the tubesheet are excluded from this requirement. The tube-to-tubesheet weld is not part ofthe tube. In addition to meeting the requirements ofd.l, d.2, and d.3 below, the inspection scope, inspection methods, and inspection intervals shall be such as to ensure that SG tube integrity is maintained until the next SG inspection. An assessment of degradation shall be performed to determine the type and location of flaws to which the tubes may be susceptible and, based on this assessment, to determine which inspection methods need to be employed and at what locations.

1. Inspect 100% ofthe tubes in each SG during the first refueling outage following SG replacement.

2. Inspect 100% ofthe tubes at sequential periods of 120, 90, and, thereafter, 60 effective full power months. The first sequential period shall be considered to begin after the first inservice inspection ofthe SGs. In addition, inspect 50% of the tubes by the refueling outage nearest the midpoint ofthe period and the remaining 50% by the refueling outage nearest the end of the period. No SG shall operate for more than 48 effective full power months or two refueling outages (whichever is less) without being inspected.

3. If crack indications are found in portions ofthe SG tube not excluded above, then the next inspection for each SG for the degradation mechanism that caused the crack indication shall not exceed 24 effective full power months or one refueling outage (whichever is less). If definitive information such as from examination ofa pulled tube, diagnostic non-destructive testing, or engineering evaluation indicates that a crack-like indication is not MILLSTONE - UNIT 3 6-l7c Amendment No. 69,..J:.86, ~, ~

243,245,249,~ 255

ADMINISTRATIVE CONTROLS 6.9.1.6.c The core operating limits shall be determined so that all applicable limits (e.g. fuel thermal-mechanical limits, core thermal-hydraulic limits, ECCS limits, nuclear limits such as SHUTDOWN MARGIN, and transient and accident analysis limits) of the safety analysis are met.

6.9.1.6.d The CORE OPERATING LIMITS REPORT, including any mid-cycle revisions or supplements thereto, shall be provided upon issuance, for each reload cycle, to the NRC Document Control Desk with copies to the Regional Administrator and Resident Inspector.

STEAM GENERATOR TUBE INSPECTION REPORT 6.9.1.7 A report shall be submitted within 180 days after the initial entry into MODE 4 following completion ofan inspection performed in accordance with TS (i.8.4.g, Steam Generator (SG)

Program. The report shall include:

a.

The scope of inspections performed on each SG,

b.

Active degradation mechanisms found,

c.

Nondestructive examination techniques utilized for each degradation mechanism,

d.

Location, orientation (if linear), and measured sizes (if available) of service induced indications,

e.

Number oftubes plugged during the inspection outage for each active degradation mechanism,

f.

Total number and percentage oftubes plugged to date,

g.

The results of condition monitoring, including the results oftube pulls and in-situ

testing,

h.

The effective plugging percentage for all plugging in each SG,

1.

The primary to secondary LEAKAGE rate observed in each SG (ifit is not practical to assign the LEAKAGE to an individual SG, the entire primary to secondary LEAKAGE should be conservatively assumed to be from one SG) during the cycle preceding the inspection which is the subject ofthe report, MILLSTONE - UNIT 3 6-21 Amendment No. ~, 40, W, 69, -l-04,

+73,~,~,~,~,245,249.~ 255

ADMINISTRATIVE CONTROLS STEAM GENERATOR TUBE INSPECTION REPORT (Continued)

J.

The calculated accident induced leakage rate from the portion of the tubes below 15.2 inches from the top ofthe tubesheet for the most limiting accident in the most limiting Sa. In addition, ifthe calculated accident induced leakage rate from the most limiting accident is less than 2.49 times the maximum operational primary to secondary leakage rate, the report should describe how it was determined; and

k.

The results ofmonitoring for tube axial displacement (slippage). If slippage is discovered, the implications ofthe discovery and corrective action shall be provided.

SPECIAL REPORTS 6.9.2 Special reports shall be submitted to the U.S. Nuclear Regulatory Commission, Document Control Desk, Washington, D.C. 20555, one copy to the Regional Administrator Region I, and one copy to the NRC Resident Inspector, within the time period specified for each report.

6.10 Deleted.

6.11 RADIATION PROTECTION PROGRAM 6.11.1 Procedures for personnel radiation protection shall be prepared consistent with the requirements of 10 CFR Part 20 and shall be approved, maintained, and adhered to for all operations involving personnel radiation exposure.

6.12 HIGH RADIATION AREA As provided in paragraph 20.1601 (c) of 10 CFR Part 20, the following controls shall be applied to high radiation areas in place ofthe controls required by paragraph 20.1601 (a) and (b) of 10 CFR Part 20:

MILLSTONE - UNIT 3 6-21a AmendmentNo.~,245,249,~ 255

UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20555-0001 SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATED TO AMENDMENT NO. 255 TO RENEWED FACILITY OPERATING LICENSE NO. NPF-49 DOMINION NUCLEAR CONNECTICUT, INC.

MILLSTONE POWER STATION, UNIT NO.3 DOCKET NO. 50-423

1.0 INTRODUCTION

By letter dated April 12, 2012 (Reference 1), Dominion Nuclear Connecticut, Inc. (the licensee) submitted a request for a license amendment to change the Technical Specifications {TS} for Millstone Power Station, Unit 3 (MPS3). The request proposed changes to TS Section 6.6.4.g, "Steam Generator (SG) Program," and TS 6.9.1.7, "Steam Generator Tube Inspection Report."

The proposed changes would establish permanent SG tube alternate repair criteria for tubing flaws located in the lower region of the tubesheet. The proposed changes would replace similar criteria for Millstone that was approved on an interim basis during Refueling Outage 14 (RF014) for the subsequent operating cycle.

1.1.

BACKGROUND MPS3 has four Model F SGs, which were designed and fabricated by Westinghouse. There are 5,626 Alloy 600 thermally treated (Alloy 600TT) tubes in each SG, each with a nominal outside diameter of 0.666 inches and a nominal wall thickness of 0.040 inches. The tubes are hydraulically expanded for the full depth of the 21-inch tubesheet and are welded to the tubesheet at each tube end. Until the fall of 2004, no instances of stress corrosion cracking affecting the tubesheet region of Alloy 600TT tubing had been reported at any nuclear power plants in the United States.

In the fall of 2004, crack-like indications were found in tubes in the tubesheet region of Catawba Nuclear Station (Catawba) Unit 2. These crack-like indications were found in a tube overexpansion (OXP) that was approximately seven inches below the hot-leg tubesheet in one tube, and just above the tube-to-tubesheet (TITS) weld in a region of the tube known as the tack expansion region in several other tubes. Indications were also reported near the TITS welds, which join the tube to the tubesheet. An OXP is created when the tube is expanded into a tubesheet bore hole that is not perfectly round. These out-of-round conditions were created during the tubesheet drilling process by conditions such as drill bit wandering or chip gouging.

The tack expansion is an approximately 1-inch long expansion at each tube end. The purpose of the tack expansion is to facilitate performing the TITS weld, which is made prior to the hydraulic expansion of the tube over the full tubesheet depth.

- 2 Since the initial findings at Catawba Unit 2 in the fall of 2004, other nuclear plants with Alloy 600TT tubing have found crack-like indications in tubes within the tubesheet as well; most of these indications were in the tack expansion region near the tube-end welds and were a mixture of axial and circumferential primary water stress corrosion cracking.

Over time, these cracks can be expected to become more and more extensive, necessitating more extensive inspections of the lower tubesheet region and more extensive tube plugging or repair, with attendant increased cost and the potential for shortening the useful lifetime of the SGs. To avoid these impacts, the affected licensees and their contractor, Westinghouse Electric Company, LLC (Westinghouse), have developed proposed alternative inspection and repair criteria applicable to the tubes in the lowermost region of the tubesheets. These criteria are referred to as the "H*" criteria. H* is the minimum engagement distance between the tube and tubesheet, measured downward from the top of the tubesheet, that is proposed as needed to ensure the structural and leakage integrity of the TITS joints. The proposed H* amendment would exclude the portions of tubing below the H* distance from inspection and plugging requirements on the basis that flaws below the H* distance are not detrimental to the structural and leakage integrity of the TITS joints.

Requests for permanent H* amendments were proposed for a number of plants as early as 2005. The U.S. Nuclear Regulatory Commission (NRC) staff identified a number of issues with these early proposals and in subsequent proposals made in 2009, and was unable to approve H* amendments on a permanent basis pending resolution of these issues. The NRC staff found it did have a sufficient basis to approve H* amendments on a interim (temporary) basis, based on the relatively limited extent of cracking existing in the lower tubesheet region at the time the interim amendments were approved. The technical basis for approving the interim amendments was provided in detail in the NRC staffs safety evaluations accompanying issuance of those amendments. An Interim H* amendment was approved most recently in 2011 for Millstone Unit 3 (Reference 2).

2.0 REGULATORY EVALUATION

The SG tubes are part of the reactor coolant pressure boundary (RCPB) and isolate fission products in the primary coolant from the secondary coolant and the environment. For the purposes of this safety evaluation, SG tube integrity means that the tubes are capable of performing this safety function in accordance with the plant design and licensing basis. The General Oesign Criteria (GOC) in Appendix A to Title 10 of the Code of Federal Regulations (10 CFR), Part 50 provide regulatory requirements that are applicable to Millstone and state that the RCPB shall have "an extremely low probability of abnormal leakage, of rapidly propagating failure, and of gross rupture" (GOC 14), "shall be designed with sufficient margin" (GOC 15 and 31), shall be of "the highest quality standards practical" (GOC 30), and shall be designed to permit "periodic inspection and testing... to assess... structural and leaktight integrity" (GOC 32).

The licensee discusses compliance with each of these GOC for Millstone Unit 3 in Section 3.1 of the Updated Final Safety Analysis Report and does not identify any deviations from these GOC for SG tube related issues.

In 10 CFR Part 50, Section 55a (50.55a) specifies that components which are part of the RCPB must meet the requirements for Class 1 components in Section III of the American Society of Mechanical Engineers Boiler and Pressure Vessel Code (ASME Code), except as provided in

- 3 10 CFR 50.55a(c)(2), (3), and (4). Section 50.55a further requires that throughout the service life of pressurized water reactor (PWR) facilities (like Millstone Unit 3), ASME Code Class 1 components meet the Section XI requirements of the ASME Code to the extent practical, except for design and access provisions, and pre-service examination requirements. This requirement includes the inspection and repair criteria of Section XI of the ASME Code. The Section XI requirements pertaining to in-service inspection of SG tubing are augmented by additional requirements in the TS.

In 10 CFR 50.36, "Technical specifications," the requirements related to the content of the TS are established. Pursuant to 10 CFR 50.36, TS are required to include items in the following five categories related to station operation: (1) safety limits, limiting safety system settings, and limiting control settings; (2) limiting conditions for operation (LCO); (3) surveillance requirements; (4) design features; and (5) administrative controls. The rule does not specify the particular requirements to be included in a plant's TS.

In 10 CFR 50.36(c)(2), LCO are stated to be "the lowest functional capability or performance levels of equipment required for safe operation of the facility." For Millstone Unit 3, the LCO pertaining to the subject license amendment request (LAR) is in TS 3.4.6.2, "Reactor Coolant System Operational LEAKAGE."

In 10 CFR 50.36(c)(5), "Administrative Controls" are, "the provisions relating to organization and management, procedures, recordkeeping, review and audit, and reporting necessary to assure operation of the facility in a safe manner." Programs established by the licensee, including the SG program, are listed in the administrative controls section of the TS to operate the facility in a safe manner. For Millstone Unit 3, the SG program requirements, including requirements for SG tube inspection and repair, are in TS 6.8.4.g, while the reporting requirements for the SG Program are in TS 6.9.1.7.

The TS for all PWR plants require that an SG program be established and implemented to ensure that SG tube integrity is maintained. For Millstone Unit 3, SG tube integrity is maintained by meeting the performance criteria specified in TS 6.8A.g.b for structural and leakage integrity, consistent with the plant design and licensing basis. Technical specification 6.8.4.g.a requires that a condition monitoring assessment be performed during each outage in which the SG tubes are inspected, to confirm that the performance criteria are being met. TS 6.8.4.g.d includes provisions regarding the scope, frequency, and methods of SG tube inspections. These provisions require that the inspections be performed with the objective of detecting flaws of any type that may be present along the length of a tube, from the TITS weld at the tube inlet to the TITS weld at the tube outlet, and that may satisfy the applicable tube repair criteria. The applicable tube repair criteria, specified in TS 6.8.4.g.c., are that tubes found by inservice inspection to contain flaws with a depth equal to or exceeding 40 percent of the nominal wall thickness shall be plugged, unless the tubes are permitted to remain in service through application of the alternate repair criteria provided in TS 6.8.4.g.c. Section 3.4.6.2 of the plant TSs includes a limit on operational primary-to-secondary leakage of 150 gallons per day, which if exceeded, requires the plant to be promptly shut down. Should a flaw exceeding the tube repair limit not be detected during the periodic tube surveillance required by the plant technical specifications, the operational leakage limit provides added assurance of timely plant shutdown before tube structural and leakage integrity, consistent with the design and licensing bases, are impaired.

- 4 As part of the plant's licensing bases, applicants for PWR licenses are required to analyze the consequences of postulated design-basis accidents (DBA), such as a SG tube rupture and a main steam line break (MSLB). These analyses consider primary-to-secondary leakage that may occur during these events and must show that the offsite radiological consequences do not exceed the applicable limits of the 10 CFR Part 50.67 accident source term, GDC 19 for control room operator doses (or some fraction thereof as appropriate to the accident), or the NRC-approved licensing basis (e.g., a small fraction of these limits). No accident analyses for Millstone Unit 3 are being changed because of the proposed amendment and, thus, no radiological consequences of any accident analysis are being changed. The use of the proposed alternate repair criteria does not impact the integrity of the SG tubes; therefore, the SG tubes still meet the requirements of the GDC in Appendix A to 10 CFR Part 50, and the requirements for Class 1 components in Section III of the ASME Code. The proposed changes maintain the accident analyses and consequences that the NRC has reviewed and approved for the postulated DBAs for SG tubes.

License amendment 252 is currently approved for Millstone Unit 3. This amendment modified TS 6.8.4.g, "Steam Generator (SG) Program," and TS 6.9.1.7, "Steam Generator Inspection Report," incorporating interim alternate repair criteria and associated tube inspection and reporting requirements that were applicable during RF014, and the subsequent operating cycle.

The proposed permanent amendment uses the same tube inspection and reporting requirements that were approved for the interim amendment, but allows these requirements to be used on a permanent basis. The alternate repair criteria (Le., the H* distance) associated with the permanent amendment is the same criteria used in the prior interim amendment.

3.0 TECHNICAL EVALUATION

3.1 Proposed Changes to the TS The datum for the indicated changes below are the current technical specifications, including the currently approved interim alternate repair criteria and associated tube inspection and reporting requirements. The proposed changes are shown in markup form for clarity.

TS 6.8.4.g.c. would be revised as follows:

c.

Provisions for SG tube repair criteria. Tubes found by inservice inspection to contain flaws with a depth equal to or exceeding 40% of the nominal tube wall thickness shall be plugged.

The following alternate tube repair criteria shall be applied as an alternative to the 40% depth based criteria:

Tubes with service-induced flaws located greater than 15.2 inches below the top of the tubesheet do not require plugging. Tubes with service-induced flaws located in the portion of the tube from the top of the tubesheet to 15.2 inches below the top of the tubesheet shall be plugged upon detection.

~ 5 ~

TS 6.8.4.g.d. would be revised as follows:

d.

Provisions for SG tube inspections. Periodic SG tube inspections shall be performed. The number and portions of the tubes inspected and methods of inspection shall be performed with the objective of detecting flaws of any type (e.g., volumetric flaws, axial and circumferential cracks) that may be present~

along the length of the tube, from the tube~to-tubesheet weld at the tube inlet to the tube~to~tubesheet weld at the tube outlet, and that may satisfy the applicable tube repair criteria. Portions of the tube below 15.2 inches below the top of the tubesheet are excluded from this requirement. The tube~to-tubesheet weld is not part of the tube. In addition to meeting the requirements of d.1, d.2, and d.3 below, the inspection scope, inspection methods, and inspection intervals shall be such as to ensure that SG tube integrity is maintained until the next SG inspection. An assessment of degradation shall be performed to determine the type and location of flaws to which the tubes may be susceptible and, based on this assessment, to determine which inspection methods need to be employed and at what locations.

1.

[No change/Not shown]

2.

[No change/Not shown]

3.

[No change/Not shown]

TS 6.9.1.7. Would be revised as follows:

A report shall be submitted within 180 days after the initial entry into MODE 4 following completion of an inspection performed in accordance with the Specification 6.8.4.g, Steam Generator (SG) Program. The report shall include:

a. - h. [No change/Not shown]

i.

The primary to secondary LEAKAGE rate observed in each SG (if it is not practical to assign the LEAKAGE to an individual SG, the entire primary to secondary LEAKAGE should be conservatively assumed to be from one SG) during the cycle preceding the inspection which is the subject of the report;

j.

The calculated accident induced leakage rate from the portion of the tubes below 15.2 inches from the top of the tubesheet for the most limiting accident in the most limiting SG. In addition, if the calculated accident induced leakage rate from the most limiting accident is less than 2.49 times the maximum operational primary to secondary leakage rate, the report should describe how it was determined; and

k.

The results of monitoring for tube axial displacement (slippage). If slippage is discovered, the implications of the discovery and corrective action shall be provided.

- 6 3.2 Staff Technical Evaluation The TITS joints are part of the pressure boundary between the primary and secondary systems. Each TITS joint consists of the tube, which is hydraulically expanded against the bore of the tubesheet, the TITS weld located at the tube end, and the tubesheet. The joints were designed in accordance with the ASME Code,Section III, as welded joints, not as friction joints. The TITS welds were designed to transmit the tube end cap pressure loads, during normal operating and DBA conditions, from the tubes to the tubesheet with no credit taken for the friction developed between the hydraulically expanded tube and the tubesheet. In addition, the welds serve to make the joints leak tight.

This design basis is a conservative representation of how the TITS joints actually work, since it conservatively ignores the role of friction between the tube and tubesheet in reducing the tube end cap load that is transmitted to the TITS weld. The initial hydraulic expansion of the tubes against the tubesheet produces an "interference fit" between the tubes and the tubesheet; thus, producing a residual contact pressure (RCP) between the tubes and tubesheet, which acts normally to the outer surface of the tubes and the inner surface of the tubesheet bore holes.

Additional contact pressure between the tubes and tubesheet is induced by operational conditions as will be discussed in detail below. The amount of friction force that can be developed between the outer tube surface and the inner surface of the tubesheet bore is a direct function of the contact pressure between the tube and tubesheet multiplied by the applicable coefficient of friction.

To support the proposed TS changes, the licensee's contractor, Westinghouse, has defined a parameter called H* to be that distance below the top of the tubesheet over which sufficient frictional force, with acceptable safety margins, can be developed between each tube and the tubesheet under tube end cap pressure loads associated with normal operating and design basis accident conditions to prevent significant slippage or pullout of the tube from the tubesheet, assuming the tube is fully severed at the H* distance below the top of the tubesheet.

For Millstone Unit 3, the licensee has proposed an H* distance of 15.2 inches. Given that the frictional force developed in the TITS joint over the H* distance is sufficient to resist the tube end cap pressure loads, it is the licensee's and Westinghouse's position that the length of tubing between the H* distance and the TITS weld is not needed to resist any portion of the tube end cap pressure loads. Thus, the licensee is proposing to change the TS to not require inspection of the tubes below the H* distance and to exclude tube flaws located below the H* distance (including flaws in the TITS weld) from the application of the TS tube repair criteria. Under these changes, the TITS joint would now be treated as a friction joint extending from the top of the tubesheet to a distance below the top of the tubesheet equal to H* for purposes of evaluating the structural and leakage integrity of the joint.

The regulatory standard by which the NRC staff has evaluated the subject license amendment is that the amended technical specifications should continue to ensure that tube integrity will be maintained, consistent with the current design and licensing basis. This includes maintaining structural safety margins consistent with the structural performance criteria in TS 6.8.4.g.b.1 and the design basis, as is discussed in Section 3.2.1.1 below. In addition, this includes limiting the potential for accident-induced primary-to-secondary leakage to values not exceeding the accident-induced leakage performance criteria in TS 6.8.4.g.b.2, which are consistent with

- 7 values assumed in the licensing basis accident analyses. Maintaining tube integrity in this manner ensures that the amended TS are in compliance with all applicable regulations. The NRC staff's evaluation of jOint structural integrity and accident-induced leakage integrity is discussed in Sections 3.2.1 and 3.2.2 of this safety evaluation, respectively.

3.2.1 Joint Structural Integrity 3.2.1.1 Acceptance Criteria Westinghouse has conducted extensive analyses to establish the necessary H* distance to resist pullout under normal operating and DBA conditions. Based on the physical geometry of the SG tubesheet, the NRC staff finds that pullout is the structural failure mode of interest, since the tubes are radially constrained against axial rupture by the presence of the tubesheet. The axial force which could produce pullout derives from the pressure end cap loads due to the primary-to-secondary pressure differentials associated with normal operating and DBA conditions. Westinghouse determined the needed H* distance on the basis of maintaining a factor of three against pullout under normal operating conditions and a factor of 1.4 against pullout under DBA conditions. The NRC staff finds that these are the appropriate safety factors to apply to demonstrate structural integrity, because they are consistent with the safety factors embodied in the structural integrity performance criteria in TS 6.8.4.g.b.1 and with the design basis; namely the stress limit criteria in the ASME Code,Section III.

The above approach equates tube pullout to gross structural failure which is conservative.

Should the pullout load be exceeded, tube slippage would be limited by the presence of adjacent tubes and support structures such that the tube would not be expected to pull out of the tubesheet.

The licensee has committed in Reference 1 to monitor for tube slippage during each refueling outage as part of the SG inspection program. Under the proposed license amendment, TS 6.9.1.7.k will require that the results of slippage monitoring be included as part of the required 180-day report, which is required by TS 6.9.1.7. In addition, TS 6.9.1.7.k requires that should slippage be discovered, the implications of the discovery and corrective action shall be included in the report. The NRC staff finds that slippage is not expected to occur for the reasons discussed in this safety evaluation. However, in the unexpected event it should occur, it will be important to understand why it occurred so that the need for corrective action can be evaluated.

The NRC staff concludes the commitment to monitor for slippage and the accompanying reporting requirements are acceptable.

3.2.1.2 3-D Finite Element Analysis A detailed 3-D finite element analysis (FEA) of the lower SG assembly (consisting of the lower portion of the SG shell, the tubesheet, the channel head, and the divider plate that separates the hot-and cold-leg inlet plenums inside the channel head) was performed to calculate tubesheet displacements due to primary pressure acting on the primary face of the tubesheet and SG channel head; secondary pressure acting on the secondary face of the tubesheet and SG shell; and the temperature distribution throughout the entire lower SG assembly. The calculated tubesheet displacements were used as input to the TITS interaction analysis evaluated in Section 3.2.1.3 below.

- 8 The tubesheet bore holes were not explicitly modeled. Instead, the tubesheet was modeled as a solid structure with equivalent material property values selected such that the solid model exhibited the same stiffness properties as the actual perforated tubesheet. This is a classical approach for analyzing perforated plates that the NRC staff finds acceptable.

Two versions of the 3-D FEA model were used to support the subject license amendment request, a "reference model" documented in Reference 3, which was submitted to support a previous request for a permanent H* amendment for Millstone Unit 3 and a "revised model" described in the technical support document, Reference 4. The revised model was submitted with a permanent H* amendment request from Duke Energy Carolinas, LLC for Catawba Nuclear Power Station Unit 2. The revised model described in Reference 4 is applicable to the Millstone Unit 3 SGs. The reference 3-D FEA model was used to provide displacement input to the thick shell TITS interaction model described in Section 3.2.1.3.1 below. The revised 3-D FEA model was used to provide displacement input to the square cell TITS interaction model described in Section 3.2.1.3.2 below. The revised 3-D model employs a revised mesh near the plane of symmetry (perpendicular to the divider plate) to be consistent with the geometry of the square cell model such that the displacement output from the 3-D model can be applied directly to the edges of the square cell model.

Some non-U.S. units have experienced cracks in the weld between the divider plate and the stub runner attachment on the bottom of the tubesheet. Should such cracks ultimately cause the divider plate to become disconnected from the tubesheet, tubesheet vertical and radial displacements under operational conditions could be significantly increased relative to those for an intact divider plate weld. Although the industry believes that there is little likelihood that cracks such as those seen abroad could cause a failure of the divider plate weld, the 3-D FEA conservatively considered both the case of an intact divider plate weld and a detached divider plate weld to ensure a conservative analysis. The case of a detached divider plate weld was found to produce the most limiting H* values. In the reference model (Reference 3), a factor was applied to the 3-D FEA results to account for a non-functional divider plate, based on earlier sensitivity studies. The revised 3-D FEA model in Reference 4 assumes the upper 5 inches of the divider plate to be non-existent. The NRC staff finds this further improves the accuracy of the 3-D FEA for the assumed condition of a non-functional divider plate and therefore finds this acceptable.

3.2.1.3 TITS Interaction Model 3.2.1.3.1 Thick Shell Model The licensee stated in its LAR that resistance to tube pullout is the axial friction force developed between the expanded tube and the tubesheet over the H* distance. The friction force is a.

function of the radial contact pressure between the expanded tube and the tubesheet. In the reference analysis (Reference 3), Westinghouse used classical thick-shell equations to model the interaction effects between the tubes and tubesheet under various pressure and temperature conditions for purposes of calculating contact pressure (TITS interaction model). Calculated displacements from the 3-D FEA of the lower tubesheet assembly (see Section 3.2.1.2 above) were applied to the thick shell model as input to account for the increment of tubesheet bore diameter change caused by the primary pressure acting on the primary face of the tubesheet and SG channel head, secondary pressure acting on the secondary face of the tubesheet and

- 9 SG shell, and the temperature distribution throughout the entire lower SG assembly. However, the tubesheet bore diameter change from the 3-D FEA tended to be non-uniform (eccentric) around the bore circumference. The thick shell equations used in the TITS interaction model are axisymmetric. Thus, the non-uniform diameter change from the 3-D finite element analyses had to be adjusted to an equivalent uniform value before it could be used as input to the TITS interaction analysis. A 2-D plane stress finite element model was used to define a relationship for determining a uniform diameter change that would produce the same change to average TITS contact pressure as would the actual non-uniform diameter changes from the 3-D finite element analyses.

In Reference 3, Westinghouse identified a difficultly in applying this relationship to Model 05 SGs under MSLB conditions. In reviewing the reasons for this difficulty, the NRC staff developed questions relating to the conservatism of the relationship and whether the tubesheet bore displacement eccentricities are sufficiently limited such as to ensure that TITS contact is maintained around the entire tube circumference. This concern was applicable to all SG models with Alloy 600TT tubing. In Reference 5, the NRC staff documented a list of questions that would need to be addressed satisfactorily before the NRC staff would be able to approve a permanent H* amendment. These questions related to the technical justification for the eccentricity adjustment, the distribution of contact pressure around the tube circumference, and a new model under development by Westinghouse to address the aforementioned issue encountered with the Model 05 SGs.

On June 14 and 15,2010, the NRC staff conducted an audit at the Westinghouse Waltz Mill Site (Reference 6). The purpose of the audit was to gain a better understanding of the H* analysis pertaining to eccentricity, to review draft responses to the NRC staff's questions in Reference 5, and to determine which documents would need to be provided on the docket to support any future requests for a permanent H* amendment. Based on the audit, including review of pertinent draft responses to the questions in Reference 5, the NRC staff concluded that eccentricity does not appear to be a significant variable affecting either average tube-to tubesheet contact pressure at a given elevation or calculated values of H*. The NRC staff found that average contact pressure at a given elevation is primarily a function of average bore diameter change at that elevation associated with the pressure and temperature loading of the tubesheet. Accordingly, the NRC staff concluded that no adjustment of computed average bore diameter change considered in the thick shell model is needed to account for eccentricities computed by the 3-D FEA. The material reviewed during the audit revealed that computed H*

values from the reference analyses continued to be conservative when the eccentricity adjustment factor is not applied.

3.2.1.3.2 Square Cell Model Documentation for the square cell model is contained in Reference 4, and the licensee's response to the NRC staff request for additional information regarding Reference 4 is included with the subject amendment request for Millstone Unit 3 (Reference 1). The square cell model is a 2-D plane stress FEA model of a single square cell of the tubesheet with a bore hole in the middle and each of the four sides of the cell measuring one tube pitch in length. Displacement boundary conditions are applied at the edges of the cell, based on the displacement data from the revised 3-D FEA model. The model also includes the tube cross-section inside the bore.

- 10 Displacement compatibility between the tube outer surface and bore inner surface is enforced except at locations where a gap between the tube and bore tries to occur.

This model was originally developed in response to the above-mentioned difficulty encountered when applying the eccentricity adjustment to Model D5 SGs TITS interaction analysis under MSLB conditions using the thick shell model. Early results with this model indicated significant differences compared to the thick shell model, irrespective of whether the eccentricity adjustment was applied to the thick shell model. The square cell model revealed a fundamental problem with how the results of the 3-D FEA model of the lower SG assembly were being applied to the tubesheet bore surfaces in the thick shell model. As discussed in Section 3.2.1.2 above, the perforated tubesheet is modeled in the 3-D FEA model as a solid plate whose material properties were selected such that the gross stiffness of the solid plate is equivalent to that of a perforated plate under the primary-to-secondary pressure acting across the thickness of the plate. This approach tends to smooth out the distribution of tubesheet displacements as a function of radial and circumferential location in the tubesheet, and ignores local variations of the displacements at the actual bore locations. These smoothed out displacements from the 3-D FEA results were the displacements applied to the bore surface locations in the thick shell model. The square cell model provides a means for post-processing the 3-D FEA results such as to account for localized variations of tubesheet displacement at the bore locations as part of TITS interaction analysis. Based on these findings, square cell models were developed for each of the SG model types, including the Model F SGs at Millstone Unit 3.

The square cell model is applied to nine different elevations, from the top to the bottom of the tubesheet, for each tube and loading case analyzed. The square cell slices at each elevation are assumed to act independently of one another. TITS contact pressure results from each of the nine slices are used to define the contact pressure distribution from the top to the bottom of the tubesheet.

The resisting force to the applied end cap load, which is developed over each incremental axial distance from the top of the tubesheet, is the average contact pressure (over that incremental distance) multiplied by both the tubesheet bore surface area (equal to the tube outer diameter surface area over the incremental axial distance) and the coefficient of friction. The NRC staff reviewed the coefficient of friction used in the analysis and judges it to be a reasonable lower bound (conservative) estimate. The H* distance for each tube was determined by integrating the incremental friction forces from the top of the tubesheet to the distance below the top of the tubesheet where the friction force integral equaled the applied end cap load times the appropriate safety factor as discussed in Section 3.2.1.1.

The square cell model assumes as an initial condition that each tube is fully expanded against the tubesheet bore such that the outer tube surface is in contact with the inner surface of the tubesheet bore under room temperature, atmospheric pressure conditions, with zero residual contact pressure associated with the hydraulic expansion process. The NRC staff finds the assumption of zero residual contact pressure in all tubes to be a conservative assumption.

The limiting tube locations in terms of H* were determined during the reference analysis to lie along the plane of symmetry perpendicular to the divider plate. The outer edges of the square cell model conform to the revised mesh pattern along this plane of symmetry in the 3-D FEA model of the lower SG assembly, as discussed in Section 3.2.1.2. Because the tubesheet bore holes were not explicitly modeled in the 3-D FEA, only the average displacements along each

- 11 side of the square cell are known from the 3-D FEA. Three different assumptions for applying displacement boundary conditions to the edges of the square cell model were considered to allow for a range of possibilities about how local displacements might vary along the length of each side. The most conservative assumption, in terms of maximizing the calculated H*

distance, was to apply the average transverse displacement uniformly over the length of each edge of the square cell.

Primary pressure acting on the tube inside surface, and crevice pressure 1 acting on both the tube outside surface and tubesheet bore surface, are not modeled directly as in the case of the thick shell model. Instead, the primary side (inside) of the tube is assumed to have a pressure equal to the primary pressure minus the crevice pressure. Note the crevice pressure varies as a function of the elevation being analyzed, as discussed in Section 3.2.1.4.

The NRC staff concludes that the square cell model provides for improved compatibility between the 3-D FEA model of the lower SG assembly and the TITS interaction model, more realistic and accurate treatment of the TITS joint geometry, and added conservatism relative to the thick shell model used in the reference analyses.

3.2.1.4 Crevice Pressure Evaluation The H* analyses postulate that interstitial spaces exist between the hydraulically expanded tubes and tubesheet bore surfaces. These interstitial spaces are assumed to act as crevices between the tubes and the tubesheet bore surfaces. The NRC staff finds that the assumption of crevices is conservative since the pressure inside the crevices acts to push against both the tube and the tubesheet bore surfaces, thus reducing contact pressure between the tubes and tubesheet.

For tubes which do not contain through-wall flaws within the thickness of the tubesheet, the pressure inside the crevice is assumed to be equal to the secondary system pressure. For tubes that contain through-wall flaws within the thickness of the tubesheet, a leak path is assumed to exist, from the primary coolant inside the tube, through the flaw, and up the crevice to the secondary system. Hydraulic tests were performed on several tube specimens that were hydraulically expanded against tubesheet collar specimens to evaluate the distribution of the crevice pressure from a location where through-wall holes had been drilled into the tubes to the top of the crevice location. The TITS collar specimens were instrumented at several axial locations to permit direct measurement of the crevice pressures. Tests were run for both normal operating and MSLB pressure and temperature conditions.

The NRC staff finds that the use of the drilled holes, rather than through-wall cracks, is conservative since it eliminates any pressure drop between the inside of the tube and the crevice at the hole location. This maximizes the pressure in the crevice at all elevations, thus reducing contact pressure between the tubes and tubesheet.

1 Although the tubes are in tight contact with the tubesheet bore surfaces, surface roughness effects are conservatively assumed to create interstitial spaces, which are effectively crevices, between these surfaces. See Section 3.2.1.4 for more information.

- 12 The crevice pressure data 'from these tests were used to develop a crevice pressure distribution as a function of normalized distance between the top of the tubesheet and the H* distance below the top of the tubesheet where the tube is assumed to be severed. These distributions were used to determine the appropriate crevice pressure at each axial location of the TITS interaction model and are concluded to be acceptable for this purpose by the NRC staff.

Because the crevice pressure distribution is assumed to extend from the H* location, where crevice pressure is assumed to equal primary pressure, to the top of the tubesheet, where crevice pressure equals secondary pressure, an initial guess as to the H* location must be made before solving for H* using the TITS interaction model and 3-D finite element model.

The resulting new H* estimate becomes initial estimate for the next H* iteration.

3.2.1.5 H* Calculation Process The calculation of H consists of the following steps for each loading case considered:

1. Perform initial H* estimate (mean H* estimate) using the TITS interaction model and 3-D FEA models, assuming nominal geometric and material properties, and assuming that the tube is severed at the bottom of the tubesheet for purposes of defining the contact pressure distribution over the length of the TITS crevice. Two sets of mean H* estimates are pertinent to the proposed H* value, mean H* estimates calculated with the reference TITS interaction and 3-D FEA models (Reference 3) and mean H* estimates calculated with the square cell TITS interaction and revised 3-D FEA models (Reference 4). The highest calculate mean H* estimate (for the most limiting tube) from the reference analysis is 5.23 inches, for the most limiting case of normal operating conditions (with the associated factor of safety of 3 as evaluated in Section 3.2.1.1). This estimate includes the adjustments in items 2 and 3 below. The highest calculated mean H* estimate with the square cell TITS interaction model, in conjunction with the revised 3-D lower SG FEA model, is 8.66 inches. The most limiting loading case for this revised analysis is still the normal operating condition (with its associated factor of safety of 3.0). The NRC staff finds that the difference in mean H* estimates between the reference analysis and the revised analysis is dominantly due to the improved post-processing of the 3-D FEA model displacements for application to the TITS interaction model.

2. In the reference analysis (Reference 3), a O.3-inch adjustment was added to the initial H*

estimate to account for uncertainty in the bottom of the tube expansion transition (BET) location relative to the top of the tubesheet, based on an uncertainty analysis on the BET for Model F SGs conducted by Westinghouse. This adjustment is not included in the revised H* analysis accompanying the subject amendment request. as discussed and evaluated in Section 3.2.1.5.1 of this safety evaluation.

3. In the reference analysis (Reference 3). for normal operating conditions only, an additional adjustment was added to the initial H* estimate to correct for the actual temperature distribution in the tubesheet compared to the linear distribution assumed in the reference 3-D FEA analysis. This adjustment is no longer necessary, as discussed in Section 3.2.1.2, since the temperature distributions throughout the tubesheet were calculated directly in the revised 3-D FEA supporting the current request for a permanent H* amendment.

- 13

4. Steps 1 through 3 yield a so-called "mean" estimate of H*, which is deterministically based. Step 4 involves a probabilistic analysis of the potential variability of H*, relative to the mean estimate, associated with the potential variability of key input parameters for the H* analyses. This leads to a "probabilistic" estimate of H*, which includes the mean estimate. The NRC staff's evaluation of the probabilistic analysis is provided in Sections 0 and 0 of this safety evaluation.

5. Add a crevice pressure adjustment to the probabilistic estimate of H* to account for the crevice pressure distribution which results from the tube being severed at the final H*

value, rather than at the bottom of the tubesheet. This step is discussed and evaluated in Section 3.2.1.5.2 of this safety evaluation.

6. A new step, step 6, has been added to the H* calculation process since the Reference 3 analysis to support the subject permanent amendment request. This step involves adding an additional adjustment to the probabilistic estimate of H* to account for the Poisson contraction of the tube radius due to the axial end cap load acting on each tube.

This step is discussed and evaluated in Section 3.2.1.5.3 of this safety evaluation.

3.2.1.5.1 BET Considerations The diameter of each tube transitions from its fully expanded value to its unexpanded value near the top of the tubesheet (TTS). The BET region is located a short distance below the top of tubesheet so as to avoid any potential for over-expanding the tube above the TTS. In the reference H* analysis (Reference 3), a O.3-inch adjustment was added to the mean H* estimate to account for the BET location being below the TTS, based on an earlier survey of BET distances conducted by Westinghouse. This adjustment was necessary since the reference analysis did not explicitly account for the lack of contact between the tube and tubesheet over the BET distance.

BET measurements, based on eddy current testing, have subsequently been performed for all tubes at Millstone Unit 3. These measurements showed that seven tubes in SG A had a maximum BET measurement of greater than one inch and were subsequently removed from service by plugging during RF014 (Reference 1).

However, the most recent H* analyses using the square cell TfTS interaction model (Reference 4) has made the need for a BET adjustment unnecessary, as the square cell model shows a loss of contact pressure at the TTS that is greater than the possible variation in the BET location. The loss of contact pressure at the TTS shown in the square cell model (which is unrelated to BET location) is compensated for by a steeper contact pressure gradient than was shown previously in the thick shell model H* analysis. The NRC staff concludes that the proposed H* value adequately accounts for the range of BET values at Millstone Unit 3.

3.2.1.5.2 Crevice Pressure Adjustment As discussed in Section 3.2.1.5, steps 1 through 4 of the H* calculation process leading to a probabilistic H* estimate are performed with the assumption that the tube is severed at the bottom of the tubesheet for purposes of calculating the distribution of crevice pressure as a

- 14 function of elevation. If the tube is assumed to be severed at the initially computed H* distance and steps 1 through 4 repeated, a new H* may be calculated which will be incrementally larger than the first estimate. This process may be repeated until the change in H* becomes small (convergence). Sensitivity analyses conducted with the thick shell model showed that the delta between the initial H* estimate and final (converged) estimate is a function of the initial estimate for the tube in question. This delta (i.e., the crevice pressure adjustment referred to in step 5 of Section 3.2.1.5) was plotted as a function of the initial H* estimate for the limiting loading case and tube radial location. Although the sensitivity study was conducted with the thick shell model, the deltas from this study were used in the Reference 4 (square cell model) analysis to make the crevice pressure adjustment to H*.

Updating this sensitivity study would have been very resource intensive, requiring many new 2-D FEA square cell runs. In response to an NRC staff question as to whether it is conservative to rely on the existing sensitivity study, as opposed to updating it to reflect the square cell model, Westinghouse submitted an analysis (Reference 1) demonstrating that if the sensitivity study were updated, it would show that the crevice pressure adjustment H* is negative, not positive as is shown by the existing study. This is because the square cell model predicts a much longer zone (approximately 4 inches) of no TITS contact below the top of the tubesheet than does the thick shell model. Therefore, the crevice pressure must reduce from primary side pressure (at the iterative H* location) to secondary side pressure approximately 4 inches below the TTS.

This leads to higher predicted pressure differentials across the tube wall over the iterative H*

distance than exists during the initial iteration, when crevice pressure is initially assumed to vary from primary pressure at the very bottom of the tubesheet to secondary pressure at the very top of the tubesheet. Based on its review of the Westinghouse analysis, the NRC staff concludes that the positive crevice pressure adjustment to H* in the Reference 4 analysis, which is based on the existing sensitivity study, is conservative and that an updated sensitivity analysis based on use of the square cell model would show that a negative adjustment can actually be justified.

Thus, the NRC staff concludes the crevice pressure adjustment performed in support of the proposed H* amendment is conservative and acceptable.

3.2.1.5.3 Poisson Contraction Effect The axial end cap load acting on each tube is equal to the primary-to-secondary pressure difference multiplied by the tube's cross-sectional area. For purposes of resisting tube pullout under normal and accident conditions, the end cap loads used in the H* analyses are based on the tubesheet bore diameter, which the NRC staff finds to be a conservative assumption. The axial end cap load tends to stretch the tube in the axial direction, but causes a slight contraction in the tube radius due to the Poisson's Ratio effect. This effect, by itself, tends to reduce the TITS contact pressure and, thus, to increase the H* distance. The axial end cap force is resisted by the axial friction force developed at the TITS joint. Thus, the axial end cap force begins to decrease with increasing distance into the tubesheet, reaching zero at a location before the H*

distance is reached. This is because the H* distances are intended to resist pullout under the end cap loads with the appropriate factors of safety applied as discussed in Section 3.2.1.1.

A simplified approach was taken to account for the Poisson radial contraction effect. First, thick shell equations were used to estimate the reduction in contact pressure associated with application of the full end cap load, assuming none of this end cap load has been reduced by

- 15 friction between the tube and the tubesheet. The T fTS contact pressure distributions determined in Step 4 of the H* calculation process in Section 3.2.1.5 were reduced by this amount. Second, the friction force associated with these reduced TfTS contact pressures were integrated with distance into the tubesheet, and the length of engagement necessary to react one times the end cap loading (Le., no safety factor applied) was determined. At this distance (termed attenuation distance by Westinghouse), the entire end cap loading was assumed to have been reacted by friction with the tubesheet, and the axial load in the tube below the attenuation distance was assumed to be zero. Thus, the T fTS contact pressures below the attenuation distance were assumed to be unaffected by the Poisson radial contraction effect.

Finally, a revised H* distance was calculated, where the TfTS contact pressures from Step 4 of Section 3.2.1.5 were reduced only over the attenuation distance.

The NRC staff finds the simplified approach for calculating the H* adjustment for the Poisson contraction effect to contain significant conservatism relative to a more detailed approach.

Regarding the safety factor of unity assumption, Westinghouse states that it is unrealistic to apply a safety factor to a physical effect such as Poisson's ratio. The NRC staff has not reached a conclusion on this point. However, irrespective of whether a safety factor is applied to the Poisson's contraction effect (consistent with Section 4.2.1.1 above), the NRC staff concludes there is ample conservatism embodied in the proposed H* distance to accommodate the difference.

3.2.1.6 Acceptance Standard - Probabilistic Analysis The purpose of the probabilistic analysis is to develop an H* distance that ensures with a probability of 95 percent that the population of tubes will retain margins against pullout consistent with criteria evaluated in Section 3.2.1.1 of this safety evaluation, assuming all tubes to be completely severed at their H* distance. The NRC staff finds this probabilistic acceptance standard is consistent with what the NRC staff has approved previously and is acceptable. For example, the upper voltage limit for the voltage based tube repair criteria in NRC Generic Letter 95-05 (Reference 7) employs a consistent criterion. The NRC staff also notes that use of the 95 percent probability criterion ensures that the probability of pullout of one or more tubes under normal operating conditions and conditional probability of pullout under accident conditions is well within tube rupture probabilities that have been considered in probabilistic risk assessments (References 8 and 9).

In terms of the confidence level that should be attached to the 95 percent probability acceptance standard, it is industry practice for SG tube integrity evaluations, as embodied in industry guidelines, to calculate such probabilities at a 50 percent confidence level. The NRC staff has been encouraging the industry to revise its guidelines to call for calculating such probabilities at a 95 percent confidence level when performing operational assessments and a 50 percent confidence level when performing condition monitoring (Reference 10). In the meantime, the calculated H* distances supporting the subject amendment request have been evaluated at the 95 percent confidence level, as recommended by the NRC staff.

Another issue relating to the acceptance standard for the probabilistic analysis is determining what population of tubes needs to be analyzed. For accidents such as MSLB or Feedline Break (FLB), the NRC staff and licensee agree that the tube population in the faulted SG is of interest, since it is the only SG that experiences a large increase in the primary-to-secondary pressure

- 16 differential. However, normal operating conditions were found to be the most limiting in terms of meeting the tube pullout margins in Section 3.2.1.1. For normal operating conditions, tubes in all SGs at the plant are subject to the same pressures and temperatures. Although there is not a consensus between the NRC staff and industry on which population needs to be considered in the probabilistic analysis for normal operating conditions, the calculated H* distances for normal operating conditions supporting the subject amendment request are 95 percent probability 195 percent confidence estimates based on the entire tube population for the plant, consistent with the NRC staffs recommendation.

Based on the above, the NRC staff concludes that the proposed H* distance in the subject LAR is based on acceptable probabilistic acceptance standards evaluated at acceptable confidence levels.

3.2.1.7 Probabilistic Analyses 3.2.1.7.1 Reference Analyses Sensitivity studies were conducted during the reference analyses (Reference 3) and demonstrated that H* was highly sensitive to the potential variability of the coefficients of thermal expansion (CTE) for the Alloy 600 tubing material and the SA-50 aClass 2a tubesheet material.

Given that no credit was taken in the reference H* analyses (Reference 3) for residual contact pressure associated with the tube hydraulic expansion process2, the sensitivity of H* to other geometry and material input parameters was judged by Westinghouse to be inconsequential and were ignored, with the exception of Young's modulus of elasticity (a measure of stiffness) for the tube and tubesheet materials. Although the Young's modulus parameters were included in the reference H* analyses sensitivity studies, these parameters were found to have a weak effect on the computed H*. Based on its review of the analysis models and its engineering judgment, the NRC staff finds that the sensitivity studies adequately capture the input parameters which may significantly affect the value of H*. This conclusion is based, in part, on no credit being taken for RCP during the reference H* analyses.

These sensitivity studies were used to develop influence curves describing the change in H*,

relative to the mean H* value estimate (see Section 3.2.1.5), as a function of the variability of each CTE parameter and Young's modulus parameter, relative to the mean values of CTE and Young's Modulus. Separate influence curves were developed for each of the four input parameters. The sensitivity studies showed that of the four input parameters, only the CTE parameters for the tube and tubesheet material had any interaction with one another. A combined set of influence curves containing this interaction effect were also created.

Two types of probabilistic analyses were performed independently in the reference analyses.

One was a simplified statistical approach utilizing a "square root of the sum of the squares" method and the other was a detailed statistical (Monte Carlo) sampling approach. The NRC staffs review of the reference analysis relied on the Monte Carlo analysis, which provides the most realistic treatment of uncertainties. The NRC staff reviewed the implementation of probabilistic analyses in the reference analyses and questioned whether the H* influence curves 2 Residual contact pressures are sensitive to variability of other input parameters.

- 17 had been conservatively treated. To address this concern, new H* analyses were performed as documented in References 11 and 12. These analyses made direct use of the H* influence curves in a manner the NRC staff finds to be acceptable.

The revised reference analyses in Reference 11 divided the tubes by sector location within the tube bundle and all tubes were assumed to be at the location in their respective sectors where the initial value of H* (based on nominal values of material and geometric input parameters) was at its maximum value for that sector. The H* influence curves discussed above, developed for the most limiting tube location in the tube bundle, were conservatively used for all sectors. The revised reference analyses also addressed a question posed by the NRC staff concerning the appropriate way to sample material properties for the tubesheet, whose properties are unknown but do not vary significantly for a given SG, in contrast to the tubes whose properties tend to vary much more randomly from tube to tube in a given SG. This issue was addressed by a staged sampling process where the tubesheet properties were sampled once and then held fixed, while the tube properties were sampled a number of times equal to the SG tube population. This process was repeated 10,000 times, and the maximum H* value from each repetition was rank ordered. The final H* value was selected from the rank ordering to reflect a 95 percent probability value at the desired level of confidence for a single SG tube population or all SG population, as appropriate. The NRC staff concludes that this approach addresses the NRC staffs question in a realistic fashion and is acceptable.

The reference analyses in References 3 and 11 indicated normal operating conditions (with associated safety factor of 3) to be the limiting case for determining H* for Model F SGs. As discussed earlier in Section 3.2.1.5, subsequent analyses with the more accurate square cell model and revised 3-D FEA model (due to the improved displacement compatibility between the two models) show that normal operating conditions (with associated safety factor of 3) to still be the limiting case for the Model F SGs.

3.2.1.7.2 Revised Analyses to Reflect Square Cell and Revised 3-D FEA Models New Monte Carlo analyses using the square cell model to evaluate the statistical variability of H*

due to the CTE variability for the tube and tubesheet materials were not performed. This was because such an approach would have been extremely resource intensive and because a simpler approach involving good approximation was available. The Simplified approach involved using the results of the Monte Carlo analyses from the reference analysis, which are based on the thick shell TrrS interaction model, to identify CTE values for the tube and tubesheet associated with the probabilistic H* values near the desired rank ordering. Tube CTE values associated with the upper 10 percent rank order estimates are generally negative variations from the mean value whereas tubesheet CTE values associated with the higher ranking order estimates are generally positive variations from the mean value. For the upper 10 percent of the Monte Carlo results ranking order, a combined uncertainty parameter, "alpha," was defined as the square root of the sum of the squares of the associated tube and tubesheet CTE values for each Monte Carlo sample. Alpha was plotted as a function of the corresponding H* estimate and separately as a function of rank order. Each of these plots exhibited well defined "break lines," representing the locus of maximum H* estimates and maximum rank orders associated with a given value of alpha. From these plots, three paired sets of tube and tubesheet CTE values, located near the break line, were selected. One of these pairs was for the rank order corresponding to an upper 95 percent probability and 95 percent confidence value for H* on a whole plant basis, which the NRC staff finds is appropriate for normal operating conditions (see

- 18 Section 3.2.1.6). These CTE values were then input to the lower SG assembly 3-D FEA model and the square cell model to yield probabilistic H* estimates which approximate the H* values for these same rank orderings, had a full Monte Carlo been performed with the square cell and revised 3-D FEA models. These H* estimates were then plotted as a function of rank ordering, allowing the interpolation of H* values at the other rank orders. The resulting 95/95 upper bound H* estimate is 14.04 inches, which compares to the mean estimate of 8.66 inches as discussed in Section 3.2.1.5. With adjustments for Poisson's contraction (see Section 3.2.1.5.3) and crevice pressure (see Section 3.2.1.5.2), the final 95/95 upper bound H* estimate is 15.21 inches.

The NRC staff believes that the above break line approach to be a very good approximation of what an actual Monte Carlo would show. A perfect approximation would mean that if, hypothetically, one were to perform a square cell analysis for each paired set of tube and tubesheet CTE values associated with the top 10 percent of rank orders, and plot the resulting H* values versus the original rank ordering associated with the CTE couple, the calculated H*

values should monotonically increase from rank order to rank order. Westinghouse performed additional square cell analyses with CTE pairs for five consecutive rank orders for both Model D5 and Model F SGs. The results showed deviations from monotonically increasing values of H* with rank order to be on the order of only 0.3 inches for the Model D5 SGs and 0.1 inches for the Model F SGs. The NRC staff concludes that use of the break line approach adds little imprecision to the probabilistic H* estimates and is acceptable.

3.2.1.8 Coefficient of Thermal Expansion During operation, a large part of contact pressure in a SG TITS joint is derived from the difference in CTE between the tube and tubesheet. As discussed in Section 3.2.1.7, the calculated value of H* is highly sensitive to the assumed values of these CTE parameters.

However, CTE test data acquired by an NRC contractor, Argonne National Laboratory (ANL),

suggested that CTE values may vary substantially from values listed in the ASME Code for design purposes. In Reference 13, the NRC staff highlighted the need to develop a rigorous technical basis for the CTE values, and their potential variability, to be employed in future H*

analyses.

In response, Westinghouse had a subcontractor review the CTE data in question, determine the cause of the variance from the ASME Code CTE values, and provide a summary report (Reference 14). Analysis of the CTE data in question revealed that the CTE variation with temperature had been developed using a polynomial fit to the raw data, over the full temperature range from 75 of to 1300 of. The polynomial fit chosen resulted in mean CTE values that were significantly different from the ASME Code values from 75 of to about 300 of. When the raw data was reanalyzed using the locally weighted least squares regression method, the mean CTE values determined were in good agreement with the established ASME Code values.

Westinghouse also formed a panel of licensee experts to review the available CTE data in open literature, review the ANL provided CTE data, and perform an extensive CTE testing program on Alloy 600 and SA-508 steel material to supplement the existing data base. Two additional sets of CTE test data (different from those addressed in the previous paragraph) had CTE offsets at low temperature that were not expected. Review of the test data showed that the first test, conducted in a vacuum, had proceeded to a maximum temperature of approximately 1300 of,

- 19 which changed the microstructure and the CTE of the steel during decreasing temperature conditions. As a result of the altered microstructure, the CTE test data generated in the second test, conducted in air, was also invalidated. As a result of the large "dead band" region and the altered microstructure, both data sets were excluded from the final CTE values obtained from the CTE testing program. The test program included multiple material heats to analyze chemistry influence on CTE values and repeat tests on the same samples were performed to analyze for test apparatus influence. Because the tubes are strain hardened when they are expanded into the tubesheet, strain hardened samples were also measured to check for strain hardening influence on CTE values.

The data from the test program was combined with the ANL data that was found to be acceptable, and the data obtained from the open literature search. A statistical analysis of the data uncertainties was performed by comparing deviations to the mean values obtained at the applicable temperatures. The correlation coefficients obtained indicated a good fit to a normal distribution, as expected. Finally, an evaluation of within-heat variability was performed due to increased data scatter at low temperatures. The within-heat variability assessment determined that the increase in data scatter was a testing accuracy limitation that was only present at low temperature. The CTE report is included as Appendix A to Reference 3.

The testing showed that the nominal ASME Code values for Alloy 600 and SA-508 steel were both conservative relative to the mean values from all the available data. Specifically, the CTE mean value for Alloy 600 was greater than the ASME Code value and the CTE mean value for SA-508 steel was smaller than the ASME Code value. Thus, the H* analyses utilized the ASME Code values as mean values in the H* analyses. The NRC staff finds this to be conservative because it tends to lead to an over-prediction of the expansion of the tubesheet bore and an under-prediction of the expansion of the tube, thereby resulting in an increase in the calculated H* distance. The statistical variances of the CTE parameters from the combined data base were utilized in the H* probabilistic analysis.

Based on its review of Westinghouse CTE program, the NRC staff concludes that the CTE values used in the H* analyses are fully responsive to the concerns stated in Reference 13 and are acceptable.

3.2.2 Leakage Considerations Operational leakage integrity is assured by monitoring primary-to-secondary leakage relative to the applicable TS limiting condition for operation limits in TS 3.4.6.2, "Reactor Coolant System Operational LEAKAGE." However, it must also be demonstrated that the proposed TS changes do not create the potential for leakage during DBA to exceed the accident leakage performance criteria in TS 6.8.4.g.b.2, including the leakage values assumed in the plant licensing basis accident analyses.

If a tube is assumed to contain a 100 percent through-wall flaw some distance into the tubesheet, a potential leak path between the primary and secondary systems is introduced between the hydraulically expanded tubing and the tubesheet. The leakage path between the tube and tubesheet has been modeled by the licensee's contractor, Westinghouse, as a crevice consisting of a porous media. Using Darcy's model for flow through a porous media, leak rate is

- 20 proportional to differential pressure and inversely proportional to flow resistance. Flow resistance is a direct function of viscosity, loss coefficient, and crevice length.

Westinghouse performed leak tests of tube-to-tubesheet joint mockups to establish loss coefficient as a function of contact pressure. A I.arge amount of data scatter, however, precluded quantification of such a correlation. In the absence of such a correlation, Westinghouse has developed a leakage factor relationship between accident induced leak rate and operational leakage rate, where the source of leakage is from flaws located at or below the H* distance.

Using the Darcy model, the leakage factor for a given type accident is the product of four quantities. The first quantity is the ratio of the maximum primary-to-secondary pressure differential during the accident to the normal operating condition pressure differential. The second quantity is the ratio of primary coolant viscosity at normal operating temperature to primary coolant viscosity at accident condition temperature. The third quantity is the ratio of crevice length under normal operating conditions to crevice length under accident condition.

This third ratio equals 1, provided it can be shown that positive contact pressure is maintained along the entire H* distance for both normal operating and worst case accident condition. The fourth quantity is the ratio of loss coefficient under normal operating conditions to loss coefficient under the accident condition. Although the absolute value of these loss coefficients is not known, Westinghouse has assumed that the loss coefficient is constant with contact pressure, as thus the ratio is equal to 1. The NRC staff finds that this is a conservative assumption, provided there is a positive contact pressure for both conditions along the entire H* distance and provided that contact pressure increases at each axial location along the H* distance when going from normal operating to accident conditions. Both assumptions were confirmed to be valid in the H* analyses.

Leakage factors were calculated for design basis accidents exhibiting a significant increase in primary-to-secondary pressure differential, including MSLB, FLB, locked rotor, and control rod ejection. The design basis FLB heat-up transient was found to exhibit the highest leakage factor, 2.49, meaning that it is the transient expected to result in the largest increase in leakage relative to normal operating conditions.

The latest H* analyses by Westinghouse (Reference 4) continued to show an increasing TITS contact pressure when going from normal operating to MSLB conditions. The new analyses used the revised 3-D finite element model of the lower SG assembly and the new square cell model, discussed in Section 3.2.1.3.2 of this safety evaluation.

In Enclosure 8 to Reference 1, the licensee provided a commitment describing how the leakage factor will be used to satisfy TS 6.8.4.g.a for condition monitoring and TS 6.8.4.g.b.2 regarding performance criteria for accident induced leakage:

For the condition monitoring (eM) assessment, the component of operational leakage from the prior cycle from below the H* distance will be multiplied by a factor of 2.49 and added to the total leakage from any other source and compared to the allowable accident induced leakage limit. For the Operational Assessment (OA), the difference in the leakage between the allowable leakage and the accident induced leakage from sources other than the tubesheet expansion region will be divided by 2.49 and compared to the

- 21 observed operationa//eakage. An administrative limit will be established to not exceed the calculated value.

That this commitment is not part of the Millstone Unit 3 license is consistent with the fact that details of how condition monitoring and operational assessments are performed are generally not included as part of the operating license, including the technical specifications. Extensive industry guidance on conducting condition monitoring and operational assessments is available as part of the industry NEI 97-06 initiative (Reference 15). The above commitments ensure that plant procedures address the above leakage factor issue as they do industry guidelines.

The subject amendment request includes reporting requirements (TS 6.9.1.7.i and 6.9.1.7.j) relating to operational leakage existing during the cycle preceding each SG inspection and condition monitoring assessment, and the associated potential for accident induced leakage from the lower portion of the tubesheet below the H* distance. These reporting requirements will allow the NRC staff to monitor how the leakage factor is actually being used, and are acceptable.

The licensee stated that they will continue to monitor for tube slippage during each refueling outage as part of their SG Program. Any findings relative to tube slippage will be reported as part of TS 6.9.1.7.k. These reporting requirements will allow the NRC staff to monitor the results of the slippage inspections and are acceptable.

3.3 Summary of Technical Evaluation Since the initial proposal for a permanent H* amendment in 2005, the supporting technical analyses have undergone substantial revision and refinement to address NRC staff questions and issues. The current analyses supporting the proposed permanent amendment still embody uncertainties and issues (e.g., should a factor of safety be applied to the Poisson's contraction effect) as discussed throughout this safety evaluation. However, it is important to acknowledge that there are significant conservatisms in the analyses. Some examples, also discussed elsewhere in this safety evaluation, include taking no credit for residual contact pressures associated with the hydraulic tube expansion process, the assumed value of 0.2 for coefficient of friction between the tube and tubesheet, and taking no credit for constraint against pullout provided by adjacent tubes and support structures. The NRC staff has evaluated the potential impact of the uncertainties and concludes these uncertainties to be adequately bounded by the significant conservatism within the analyses and proposed H* distance.

The NRC staff finds the proposed changes to the technical specifications of Millstone Unit 3 ensure that tube structural and leakage integrity will be maintained, with structural safety margins consistent with the design basis and with leakage integrity within assumptions employed in the licensing basis accident analyses, without undue risk to public health and safety. Based on this finding, the NRC staff further concludes that the proposed amendment meets 10 CFR 50.36 and, thus, the proposed amendment is acceptable.

4.0 STATE CONSULTATION

In accordance with the Commission's regulations, the Connecticut State official was notified of the proposed issuance of the amendment. The State official had no comments.

- 22

5.0 ENVIRONMENTAL CONSIDERATION

The amendment changes a requirement with respect to installation or use of a facility component located within the restricted area as defined in 10 CFR Part 20. The NRC staff has determined that the amendment involves no significant increase in amounts, and no significant change in the types of any effluents that may be released offsite, and that there is no significant increase in individual or cumulative occupational radiation exposure. The Commission has previously issued a proposed finding that the amendment involves no significant hazards consideration, and there has been no public comment on such finding (76 FR 73730). Accordingly, the amendment meets the eligibility criteria for categorical exclusion set forth in 10 CFR 51.22(c)(9). Pursuant to 10 CFR 51.22(b), no environmental impact statement or environmental assessment need be prepared in connection with the issuance of the amendment.

6.0 CONCLUSION

The Commission has concluded, based on the considerations discussed above, that (1) there is reasonable assurance that the health and safety of the public will not be endangered by operation in the proposed manner, (2) there is reasonable assurance that such activities will be conducted in compliance with the Commission's regulations, and (3) the issuance of the amendment will not be inimical to the common defense and security or to the health and safety of the public.

7.0 REFERENCES

1.

Dominion Nuclear Connecticut, Inc. (DNC) letter 12-203, "License Amendment Request for Permanent Alternate Repair Criteria for Steam Generator Tube Inspection and Repair" dated April 12, 2012, NRC ADAMS Accession No. ML12110A114.

2.

NRC letter to Dominion Nuclear Connecticut, Inc., "Millstone Power Station, Unit 3*

Issuance of Amendment Re: Steam Generator Tube Inspection Alternate Repair Criteria" dated October 7,2011, NRC ADAMS Accession No. ML112580517.

3.

Westinghouse Electric Company LLC, WCAP-17071-P (Proprietary) and WCAP-17071 NP (Non-Proprietary), Rev. 0, "H*: Alternate Repair Criteria for the Tubesheet Expansion Region in Steam Generators with Hydraulically Expanded Tubes {Model F),"

April 2009, NRC ADAMS Accession No. ML093620086.

4.

Westinghouse Electric Company (WEC) report, WCAP-17330-P (Proprietary) and WCAP-17330-NP (Non-Proprietary), Rev. 1, "H*: Resolution of NRC Technical Issue Regarding Tubesheet Bore Eccentricity (Model F/Model D5)," November 2010; NRC ADAMS Accession Nos. ML11188A109 (Proprietary) and ML11188A108 (Non Proprietary). This report was enclosed with a permanent H* amendment request submitted by Duke Energy Carolinas, LLC for Catawba Nuclear Power Station Unit 2.

5.

NRC letter to Southern Nuclear Operating Company, "Vogtle Electric Generating Plant, Units 1 and 2, Transmittal of Unresolved Issues Regarding Permanent Alternate Repair Criteria for Steam Generators," November 23, 2009, NRC ADAMS Accession No. ML093030490.

- 23

6.

NRC memorandum, R. Taylor to G. Kulesa, "Vogtle Electric Generating Plant - Audit of Steam Generator H* Amendment Reference Documents," July 9, 2010, NRC ADAMS Accession No. 101900227.

7.

NRC Generic Letter 95-05, "Vogtle Based Alternate Repair Criteria for Westinghouse Steam Generator Tubes Affected by Outside Diameter Stress Corrosion Cracking,"

August 3,1995, NRC ADAMS Accession No. ML031070113.

8.

NUREG-0844, "NRC Integrated Program for the Resolution of Unresolved Safety Issues A-3, A-4, and A-5 Regarding Steam Generator Tube Integrity," September 1988.

9.

NUREG-1570, "Risk Assessment of Severe Accident-Induced Steam Generator Tube Rupture," March 1998.

10.

NRC Meeting minutes, "Summary of the January 8, 2009, Category 2 Public Meeting with the Nuclear Energy Institute (NEI) and Industry to Discuss Steam Generator Issues,"

February 6, 2009, NRC ADAMS Accession No. ML090370782.

11.

WEC letter, LTR-SGMP-09-100-P (Proprietary) and L TR-SGMP-09-100-NP (Non-Proprietary) "Response to NRC Request for Additional Information on H*; Model F and D5 Steam Generators," August 12, 2009, NRC ADAMS Accession Number ML101730391 (Non-Proprietary).

12.

SNC letter NL-09-1317, August 28,2009, transmitting WEC letter LTR-SGMP-09-104-P Attachment "White Paper on Probabilistic Assessment of H*" dated August 13, 2009, NRC ADAMS Accession No. ML092450029.

13.

NRC letter to Wolf Creek Nuclear Operating Corporation, "Wolf Creek Generating Station - Withdrawal of License Amendment Request on Steam Generator tube Inspections," February 28, 2008, NRC ADAMS Accession No. ML080450185.

14.

Nuclear Energy Institute (NEI) letter dated July 7,2008, NRC ADAMS Accession No. ML082100086, transmitting Babcock and Wilcox Limited Canada letter 2008-06-PK-001, "Re-assessment of PMIC measurements for the determination of CTE of SA 508 steel,"

dated June 6,2008, NRC ADAMS Accession No. ML082100097.

15.

NEI 97-06, Revision 3, "Steam Generator Program Guidelines," January 2011, NRC ADAMS Accession No. ML111310708.

Principal Contributor: Andrew Johnson Date: December 6, 2012

ML12299A498 "See memo dated October 19, 2012 Office LPL 1-1/PM LPL 1-1/LA STSB/BC ESGB/BC OGC - NLO LPL 1-1/BC Name JKim KGoldstein RElliott GKulesa" CKanatas GWilson Date 11/07/12 11/02/12 11/08/12 10/19/12 12/04/12 12/06/12