L-20-277, Response to Request for Additional Information Regarding Steam Generator Tube Sleeve License Amendment Request

Response to Request for Additional Information Regarding Steam Generator Tube Sleeve License Amendment Request

ML21022A133
Person / Time
Site:	Beaver Valley
Issue date:	01/22/2021
From:	Grabnar J Energy Harbor Nuclear Corp
To:	Document Control Desk, Office of Nuclear Reactor Regulation
Shared Package
ML21025A366	List: L-20-277, Response to Request for Additional Information Regarding Steam Generator Tube Sleeve License Amendment Request
References
EPID L-2020-LLA-0140, L-20-277
Download: ML21022A133 (71)

Text

WITHHOLD FROM PUBLIC DISCLOSURE UNDER 10 CFR 2.390 WHEN SEPARATED FROM ENCLOSURES B AND D, THIS DOCUMENT CAN BE DECONTROLLED energy Energy Harbor Nuclear Corp.

• - - harbor Beaver Valley Power Station P.O. Box 4 Shippingport, PA 15077 John J. Grabnar 724-682-5234 Site Vice President, Beaver Valley Nuclear January 22, 2021 L-20-277 ATTN: Document Control Desk U.S. Nuclear Regulatory Commission Washington, DC 20555-0001

Subject:

Beaver Valley Power Station, Unit No. 2 Docket No. 50-412, License No. NPF-73 Response to Request for Additional Information Regarding Steam Generator Tube Sleeve License Amendment Request (EPID L-2020-LLA-0140)

By correspondence dated June 25, 2020 (Accession No. ML20177A272), Energy Harbor Nuclear Corp. submitted to the Nuclear Regulatory Commission (NRC) a request to amend Beaver Valley Power Station, Unit No. 2 Technical Specification requirements related to methods of inspection and service life for Alloy 800 steam generator tubesheet sleeves. By email dated October 22, 2020 (Accession No. ML20297A322), the NRC requested additional information to complete its review. The response to items 1 through 6 in the request for additional information is provided in enclosures A and B (except for Item 1.c that is due later), and the response to Item 7 is provided in the attachment.

In response to the request for additional information, proprietary and non-proprietary versions of document number SG-CDMP-19-19, Probability of Flaw Detection in the Alloy 800 Mechanical Sleeve Lower Tubesheet Joint Using the Ghent Version 2 Eddy Current Probe have been updated. Therefore, enclosures C and D to this letter replace, in their entirety, the previously-submitted enclosures C and E provided in the June 25, 2020 submittal.

Enclosure E contains Affidavit CAW-21-5140 signed by Westinghouse Electric Company LLC (Westinghouse). The affidavit sets forth the basis on which proprietary information owned by Westinghouse that is contained in enclosures B and D may be withheld from public disclosure by the Nuclear Regulatory Commission (Commission)

Beaver Valley Power Station, Unit No. 2 L-20-277 Page 2 and addresses with specificity the considerations listed in paragraph (b)(4) of Section 2.390 of the Commissions regulations.

Accordingly, it is respectively requested that the information that is proprietary to Westinghouse be withheld from public disclosure in accordance with 10 CFR Section 2.390 of the Commissions regulations.

Correspondence with respect to the copyright or proprietary aspects of the items listed above or the supporting Westinghouse Affidavit should reference CAW-21-5140 and should be addressed to Zachary S. Harper, Manager, Licensing Engineering, Westinghouse Electric Company, 1000 Westinghouse Drive, Suite 165, Cranberry Township, Pennsylvania 16066.

The information provided by this submittal does not invalidate the significant hazards consideration analysis provided in the June 25, 2020 submittal.

There are no regulatory commitments contained in this submittal. If there are any questions or if additional information is required, please contact Mr. Phil H. Lashley, Manager - Fleet Licensing, at (330) 696-7208.

I declare under penalty of perjury that the foregoing is true and correct. Executed on January ___, 22 2021.

Sincerely, Gmbnur,lohn 1')()72 Site Vice !'resident, Beaver Valley I am approving this document Gra.lma.r,John 19072 Jan 22 20 21 10:33 AM John J. Grabnar

Attachment:

Energy Harbor Nuclear Corp. Response to October 22, 2020 Request for Additional Information Item 7 Related to Steam Generator Tube Sleeve License Amendment Request

Enclosures:

A. Responses to Request for Additional Information Regarding Beaver Valley Power Station Unit No. 2 Steam Generator Tube Inspection and Repair (Non-proprietary)

B. Responses to Request for Additional Information Regarding Beaver Valley Power Station Unit No. 2 Steam Generator Tube Inspection and Repair (Proprietary)

Beaver Valley Power Station, Unit No. 2 L-20-277 Page 3 C. Document Number SG-CDMP-19-19 NP-Attachment, Revision 2, "Probability of Flaw Detection in the Alloy 800 Mechanical Sleeve Lower Tubesheet Joint Using the Ghent Version 2 Eddy Current Probe, dated January 2021 (Non-proprietary)

D. Document Number SG-CDMP-19-19 P-Attachment, Revision 2, "Probability of Flaw Detection in the Alloy 800 Mechanical Sleeve Lower Tubesheet Joint Using the Ghent Version 2 Eddy Current Probe, dated January 2021 (Proprietary)

E. Affidavit for Withholding Proprietary Information cc: NRC Region I Administrator NRC Resident Inspector NRR Project Manager Director BRP/DEP Site BRP/DEP Representative

Attachment L-20-277 Energy Harbor Nuclear Corp. Response to October 22, 2020 Request for Additional Information Item 7 Related to Steam Generator Tube Sleeve License Amendment Request Page 1 of 1 The Nuclear Regulatory Commission request for additional information is provided below in bold text and followed by the Energy Harbor Nuclear Corp. response.

7. Section 3.8, Tube Sleeve Inspections, of Enclosure A (Evaluation of the Proposed Amendment) states that the Ghent V2 probe will be used to inspect the Alloy 800 tubesheet sleeves each outage. Please provide the following information about these inspections:

a) Clarify which probe (+PointTM or Ghent V2) will be used for the call of record when inspecting different portions of the sleeve/parent tube assembly.

Response

The Ghent V2 probe includes a +POINTTM probe mounted slightly above and 180 degrees out from the Ghent transmit and receive coils. This allows for a single pull of the probe through the sleeve to complete all required examinations. Up until the time of Ghent probe approval, the +POINTTM probe is the call of record for the entire sleeve assembly. Once the license amendment request for the Ghent V2 probe is approved by the NRC, the Ghent V2 probe will be used as the call of record for the half-inch long nickel band, with the +POINTTM probe being the call of record for the remaining portion of the tube, sleeve, and expansion assembly.

b) Please confirm that each in-service sleeve will be inspected each refueling outage throughout the sleeve service life.

Response

Each inservice sleeve will be inspected each refueling outage throughout the sleeve service life.

Enclosure A L-20-277 Responses to Request for Additional Information Regarding Beaver Valley Power Station Unit No. 2 Steam Generator Tube Inspection and Repair (Non-proprietary)

(15 pages follow)

Westinghouse Non-Proprietary Class 3 Westinghouse Electric Company LTR-CDMP-20-38 NP-Attachment Revision 0 Responses to Request for Additional Information Regarding Beaver Valley Power Station Unit No. 2 Steam Generator Tube Inspection and Repair January 7, 2021 Authors:

Jay R. Smith* - (RAIs 2, 3 and 6)

OSG/RSG Engineering & Chemistry Ronald J. Pocratsky* - (RAI 4)

SG Inspection & Repair Solutions Jeffrey M. Raschiatore* - (RAIs 1a, 1b, 5)

US Operations Verifiers:

David S. Suddaby* - (RAIs 2, 3 & 6)

OSG/RSG Engineering and Chemistry Christopher M. Belville (RAIs 1a, 1b, 4 and 5)*

SG Inspection & Repair Solutions Reviewer:

Gary W. Whiteman*

Licensing Engineering Approved:

Michael E. Bradley, Manager Component Design & Management Programs

©2021 Westinghouse Electric Company LLC All Rights Reserved

• Electronically approved records are authenticated in the Electronic Document Management System.

LTR-CDMP-20-38 NP-Attachment Page 1 of 15 Rev. 0

• • • This record was final approved on 1/11/2021 7:55:57 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 Responses to Request for Additional Information Regarding Beaver Valley Power Station Unit No. 2 Steam Generator Tube Inspection and Repair

Background

By letter dated June 25, 2020, Energy Harbor Nuclear Corporation submitted a License Amendment Request (LAR) to revise Beaver Valley Power Station Unit 2 Technical Specifications (TS) related to Inspection Method and Service Life for Alloy 800 Steam Generator Tubesheet Sleeves. The requested changes revise TS 5.5.5.2.d, Provisions for SG [Steam Generator] Tube Inspection, and TS 5.5.5.2.f.3, Provisions for SG Tube Repair Methods, requirements related to methods of inspection and service life for Alloy 800 steam generator tubesheet sleeves.

The Nuclear Regulatory Commission (NRC) Staff has determined that additional information is needed to complete its review of the request (Reference 1). As a result, Westinghouse was contracted by Energy Harbor Nuclear Corporation to respond to the NRC request for additional information (RAI). The Westinghouse response to each information request is provided below.

Responses to Request for Additional Information

1. Enclosure D, Section 5, Probe Design, Developments and Testing Results, discusses the stages of Ghent probe development for sleeve inspections. In addition to other probe changes, the positioning of coils in Ghent Version 1 probe was modified to improve detection of circumferential flaws by Ghent Version 2 (V2) probe. Given that a [ ]b through-wall (TW) inner diameter (ID) circumferential electro-discharge machined notch was not detected in the sample set used for site specific technique qualification:

a) Is the difference in detection capability of the probe for axial and circumferential cracks associated with the probe design and/or the source of noise in the nickel band region in the lower transition sleeve roll joint?

Response

Differences in detection capability of the probe for axial and circumferential cracks are due to the [ ]a,c,e. [ A

]a,c,e. While the elimination of the nickel response was improved, it did not totally remove all the nickel.

This can be observed in Figure 1-1 below.

Figure 1-1 shows the nickel band region of Sample S5 when tested with the Ghent V2 probe. The upper Lissajous and C-scan plot show the axial detection response, while the lower Lissajous and C-scan plot show the circumferential detection response. These responses are for the test frequency of [

]a,c,e at Beaver Valley Unit 2. In the lower Lissajous is [ ]a,c,e.

LTR-CDMP-20-38 NP-Attachment Page 2 of 15 Rev. 0

• • • This record was final approved on 1/11/2021 7:55:57 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3

[

]a,c,e. The peer review team used for technique qualification [

]a,c,e.

[

]a,c,e.

a,c,e Figure 1-1. Nickel Band Region of Sample S5 when Tested with the Ghent V2 Probe LTR-CDMP-20-38 NP-Attachment Page 3 of 15 Rev. 0

• • • This record was final approved on 1/11/2021 7:55:57 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 b) What modification was made to the coil positioning (i.e., increase or decrease in spacing) to improve the detection capability of Ghent V2 for circumferential cracks?

Response

Spacing was [ ]a.c.e in the Ghent Version 2 probe for detection of circumferential flaws. This change to probe design was suggested by the probe designer. The Ghent Version 2 probe had [

]a.c.e. It is not known if [

]a.c.e.

c) Enclosure D, Page 5-1, discusses unexpected results during probe development believed to be attributed to the sleeves having a nickel band thickness at the highest tolerance limit compared to the sleeves used during the feasibility study. Please discuss if differences in nickel band thickness for installed sleeves would affect the probability of detection of cracks using the Ghent V2 probe.

Response

The response to Item 1c will be provided under separate Westinghouse letter at a later date.

2. Table 3-2 in Enclosure E contains the results of destructive examination for axial outer diameter stress corrosion cracking (ODSCC) specimens but provides only the depth of each crack.

Please provide the length of each crack in that table.

Response

The axial length of each ODSCC specimen is provided in Table 2-1. The measurements were obtained from destructive examination (DE) of each flaw and includes the flaw depth in percent through-wall (%TW) and total length in inches.

LTR-CDMP-20-38 NP-Attachment Page 4 of 15 Rev. 0

• • • This record was final approved on 1/11/2021 7:55:57 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 Table 2-1. Destructive Examination Results of Axial ODSCC Specimens Max.

Radial DE Flaw DE Flaw Max. Radial Depth of DE Flaw Axial Axial Tube Flaw Flaw Flaw Depth Flaw Depth Length, Length, Sample No. Location (µm) (mils) (%TW) (µm) (inch)

- - b 1 31 deg 2 71 deg J-2-3 3 102 deg 4 285 deg 1 91 deg 2 91 deg J-3 3 168 deg 4 278 deg 5 354 deg 1 26 deg J-8 2 228 deg J-12 1 340 deg

3. With respect to the Ghent V2 probe nickel band sleeve test program results:

a) Section 4.2.4 in Enclosure E compares simple probability of detection (POD) to noise-based model assisted POD (MAPOD) and concludes that the simple POD distribution is a better representation of actual flaw performance since the effect of the nickel band material is inherent in the detection result. The Ghent V2 signal amplitude (behind the nickel band) for Flaw #1 in Sample J-8 in Table 4-2 is listed as having a 0.09-volt signal. In view of the relatively large variability in the level of noise among sleeve/tube/collar assemblies, please provide a justification for considering a signal with such a low signal to noise (S/N) as a hit in developing the simple POD curve shown in Figure 4-1.

Response

The Westinghouse investigation into Flaw #1 in Sample J-8 flaw size and signal-to-noise ratio revealed that the flaw should have been measured at 0.23-volt and not at 0.09-volt as originally reported. This result was confirmed by four Level III eddy current data analysts.

LTR-CDMP-20-38 NP-Attachment Page 5 of 15 Rev. 0

• • • This record was final approved on 1/11/2021 7:55:57 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 The original analysis of the flaw did not measure the largest signal during a line-by-line review. The signal-to-noise ratio increased from [ ]b when using the 0.23 -volt flaw signal on the minimum noise component. Figure 3-1 provides the Ghent V2 probe graphic of the flaw showing the flaw signal and voltage amplitude measurement.

The revised signal-to-noise ratios for Flaw #1 for Sample J-8 are more consistent with the remaining flaw samples. Therefore, a flaw of this voltage amplitude and signal-to-noise ratio, detection is reasonable and expected. Table 3-1 provides the updated flaw voltage amplitude and signal-to-noise ratios for Flaw #1 in Sample J-8 and for the remaining flaws that are unchanged from Table 4-2 of Reference 2.

The remaining flaw voltage amplitudes were re-reviewed, and no other changes were required.

Updating the voltage amplitude of Flaw #1 for Sample J-1 results in changes to the Ahat distributions and signal-to-noise voltage thresholds used to calculate the voltage-based probability of detection distributions using the MAPOD methodology. Refer to Revision 2 of Reference 2 for discussions on the full impact of the change.

Table 3-1. Test Sample Signal-to-Noise Adjacent to Detected Flaws Ghent V2 Noise S/N S/N S/N Flaw Beside Using Using Using EPRI Voltage Flaw Average Minimum Maximum Sample Flaw No. Vpp Vpp Noise Noise Noise b a,c,e J-2-3 2 0.35 J-8 1 0.23 f--- -

J-12A 1 0.39 J-3 3 3.76 f--- -

S/N Average Excluding Sample J-3:

---

= -

Note: Values updated from those listed in Table 4-2 of Reference 2 are identified in bold font.

LTR-CDMP-20-38 NP-Attachment Page 6 of 15 Rev. 0

• • • This record was final approved on 1/11/2021 7:55:57 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 I SCREEM SE TUP IIUTO __ AMALYSIS STAMOIILOH E h _ __ di s ~ UTIL APF TOOL ftiiiii---Tifl>iiiTI I REPO RT 2 f""'iPi'"J---Ti4TiSfi6~1 SlfflllRY I t1 ESSIIGE I SCR[EMI MG f""'SCiiii"2---TiAT I H[ II REPORT f""'iiITT IS 70Kh z 8 SCILE*200.8 cfTSET"ll'l.367 5ib<Z&lxe.

CIG4 Ff' !: 34 ,9/ll.1129 1~=2 .7~9 e. I TICKS/SC~

< lap 16% > AUTORPC SETI.I' Figure 3-1. Ghent V2 Probe Graphic for Flaw #1 in Sample J-8, Revised Voltage Amplitude b) Please discuss how the simple POD curve would change if the Flaw #1 in Sample J-8 was not included as a hit by providing the 50th and 95th percentile POD values if a new POD curve did not include that flaw.

Response

A simple hit-miss POD curve is generated using a generalized linear model (GLM) binary log-logistic or logistic function. For this binary function to produce a distribution, there must be data points for detections and non-detections with overlapping flaw depths, meaning that there must be smaller flaw depths for detections than the larger flaw depths for non-detections. The flaw depth data points that are with the same depth span that contains detections and non-detections are referred to as overlapping flaws.

Sample J-8 Flaw #1 is the only overlapping flaw between detections (hits) and non-detections (misses). Consequently, a GLM binary log-logistic or logistic POD curve cannot be generated without overlapping of hits and misses. Without overlapping data, the POD is simply a step change between hits and misses. All flaws are detected over a certain flaw depth and all flaws are not detected below a certain flaw depth. The largest flaw not detected was [ ]a,c,e and the smallest flaw detected is [ ]a,c,e excluding the detected Flaw #1 in Sample J-8. Therefore, all flaws less than [ ]a,c,e would a,c,e not be detected and all flaws [ ] and larger would be detected. Figure 3-2 provides the simple hit-miss POD curves with and without the Sample J-8 Flaw #1 data point.

LTR-CDMP-20-38 NP-Attachment Page 7 of 15 Rev. 0

• • • This record was final approved on 1/11/2021 7:55:57 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 a,c,e Figure 3-2. Ghent V2 Probe Simple POD Curve Without Sample J8 Flaw #1 c) According to post-test parent tube inspection results shown in Table 3-1, Flaws #1 and #2 in Sample J-3 with comparable destructive examination depths (~[ ]a,c,e % TW and

~[ ]a,c,e % TW) have significantly different signal amplitudes (0.28 v and 0.63 v) based on Ghent V2 data. The signal amplitudes based on the +POINTTM1 data for these flaws have the same amplitude (0.44 v). In comparison with +POINT, a weaker correlation between Ghent V2 signal amplitude and maximum depth is also observed for other flaws listed in Tables 3-1 and 3-2. Please discuss the possible sources (i.e., probe design, noise interference, etc.) that may contribute to the differences between Ghent V2 probe and

+POINT results. The Staff notes that although the Ghent V2 probe has been qualified according to App. H for detection only (i.e., not sizing), the variability in signal amplitude directly affects the S/N values that are used in POD evaluations.

Response

The post-test +POINT probe inspection of Sample J-3 Flaw #1 and Flaw #2 could not differentiate the two flaws separately due to being in close proximity. The signals were 1

+POINT is a trademark or registered trademark of Zetec, Inc. Other names may be trademarks of their respective owners.

LTR-CDMP-20-38 NP-Attachment Page 8 of 15 Rev. 0

• • • This record was final approved on 1/11/2021 7:55:57 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 merged into one flaw with a voltage of 0.44 volt. The same +POINT probe voltage was assigned for both flaws in Table 3-1 of Reference 2. An unmodified Ghent RG-3/4 probe was able to distinguish each flaw for separate voltage measurements and were reported separately in Table 3-1. Therefore, comparison and correlation of the merged flaw +POINT probe voltage to the separate flaws from the unmodified Ghent probe cannot be made.

Table 3-1 of Reference 2 provides the post-test parent tube inspection with unmodified Ghent RG-3/4 probe voltages with no sleeve installed. An unmodified Ghent RG-3/4 probe was used for the post-test examination of the parent tube with no sleeve because the Ghent V2 probe diameter is reduced to accommodate inspection of the smaller diameter sleeve and cannot be used to inspect the larger diameter parent tube. The unmodified Ghent RG-3/4 probe design is different than the Ghent V2 probe in that the Ghent V2 probe was designed to suppress the effects of the nickel through use of stronger magnets. The

+POINT and unmodified Ghent RG-3/4 probes were used only for determining if ligament tearing occurred or other signal changes resulted from the testing process.

While the maximum depths and total lengths of the two flaws were similar, the unmodified Ghent RG-3/4 voltage amplitudes were 0.28 volt for Flaw #1 and 0.63 volt for Flaw #2.

Review of the fractographic images of Flaw #1 and Flaw #2 in Sample J3 revealed differing flaw shapes that may describe the differences in the unmodified Ghent RG-3/4 voltage amplitudes of Flaw #1 and Flaw #2 in Sample J-3. The area and depth profile of corrosion for Flaw #1 varied over its length while the corrosion area and depth profile of Flaw #2 was more uniform, leading to a larger volume under the probe field of view and thus a larger voltage amplitude response. Refer to Figure 3-3 and Figure 3-4 for the fractographic images of Flaw #1 and Flaw #2 of Sample J-3.

Figure 3-3. Sample J-3 Flaw #1 Fractographic Image LTR-CDMP-20-38 NP-Attachment Page 9 of 15 Rev. 0

• • • This record was final approved on 1/11/2021 7:55:57 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 Figure 3-4. Sample J-3 Flaw #2 Fractographic Image

4. Enclosure E, Section 3.3, describes the test program with the simulated tubesheet collar assembly. Page 14 of 38 states: As the Ghent probe is surface riding, not achieving full expansion to the full diameter of a prototypic in-generator tubesheet bore would have negligible effects on the eddy current results due to the small amount of tube wall thinning.

This statement appears to consider the effects of eddy current probe lift-off. Please discuss if the amount of manual hard rolling could also affect the probe eddy current response due to potential variability in effective conductivity across the parent tube-collar interface.

Response

This hypothesis was considered and investigated during the development of the Version 2 Ghent probe detection technique. It was determined that a hard roll torque of [

]a,c,e.

To determine if higher expansion torques would alter the conductivity/magnetic properties,

[

]a,c,e.

LTR-CDMP-20-38 NP-Attachment Page 10 of 15 Rev. 0

• • • This record was final approved on 1/11/2021 7:55:57 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 Table 4-1 Sample S10 with [ ]a,c,e ACTUAL MEASUREMENT DIMENSION inch OD FBH at [ ]a,c,e Degrees at Center of Ni FBH FBH Edge to Center Area Diameter Depth of FBH b

Pre-Expansion (at [ ]b IP)

Post-Expansion (at [ ]b IP)

Difference Percent Difference Table 4-2

[ ] a,c,e

[ ] IP b

[ ]b IP [ ]b IP inch Expansion Expansion Expansion OD FBH

[ ]a,c,e Volts Volts Volts Phase Phase Phase kHz Vp-p Vp-p Vp-p b

Axial Circ -_

Based on these observations, it was concluded that the higher roll expansion process does alter the geometry of the tubing and the OD FBH but does not affect the electromagnetic properties of the Ni in a manner that would reduce detection beyond the Ni in the parent tube with eddy current. In addition, although some cold working of the sleeve/joint assembly was observed during the expansion process, it has been demonstrated that this cold working does not influence the Ghent V2 probe detectability of degradation behind the nickel band in the parent tube.

5. Enclosure E, Page 34 of 38 discusses that the NDE test program and POD curves were based on ODSCC flaw samples due to the unavailability of primary water stress corrosion cracking samples. The primary justification for applying [

]a.c.e.

While the statement above regarding the reduction of magnetic flux density (current density) as a function of depth is correct, the inference regarding lower detection capability for outer diameter (OD) flaws would be true if the effect of noise is negligible. Since the detection probability is a function of S/N, in the presence of ID originated noise (e.g., the presence of nickel band at the sleeve OD-tube ID interface in the tubesheet transition joint), deep OD cracks in a parent tube could plausibly have a higher detection probability than ID flaws of comparable LTR-CDMP-20-38 NP-Attachment Page 11 of 15 Rev. 0

• • • This record was final approved on 1/11/2021 7:55:57 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 depth. This could be attributed to the smaller influence of the ID originated noise on OD flaws (i.e., larger phase separation between signal and noise). Please discuss any additional supporting information that the POD curves based on OD originating flaws conservatively bound the POD for ID originating flaws in the presence of a nickel band.

Response

A large phase separation between [ ]a.c.e does not exist. While the Ghent V2 test coil rides on the ID of the sleeve, influence of the nickel band occurs [ ]a.c.e. [ ]a.c.e a.c.e when testing with rotating probe coils are in the range of [ ] . But in this case, [

]a.c.e.

While the [ ]a.c.e in the Ghent V2 probe removes most of the nickel, the response of the remaining nickel occurs [

]a.c.e. Flaw response for the OD cracks used for POD determination fall into a range of [ ]a.c.e. Orientation of all OD cracks is axial. For comparison, the [ ]a.c.e used in the technique development had phase angles ranging from [ ]a.c.e. These phase angles are like the OD cracks reported in the POD study. Phase separation between signals (ID notches and OD cracks) and noise (nickel) are similar, with a small phase separation.

Any noise generated on the ID of the sleeve is inconsequential for detection of OD parent tube flaws. The higher test frequencies [ ]a.c.e can be used to distinguish the parent tube from sleeve indications due to the increased phase response and reduction in sensitivity to parent tube indications. Lower test frequencies [ ]a.c.e are a.c.e used for evaluation of the parent tube region adjacent to the [ ] nickel band.

6. Page 26 of 38 in Enclosure E states that the noise distribution used in the MAPOD process was developed from tube/sleeve collar assemblies related to the Ghent V2 probe EPRI Appendix H qualification program. This was attributed to no field data (at that time) from which to build an in-generator noise distribution. During the Beaver Valley Ghent V2 Probe LAR pre-submittal meeting on May 26, 2020 (ADAMS Accession No. ML20143A035), Slide 11 states that all 567 sleeves in service were examined with the Ghent V2 probe during the April 2020 refueling outage inspections. Please discuss if the noise data from those inspections confirms that the noise distribution used in the MAPOD process is either representative or conservative with respect to the field data.

Response

A total of 567 in-service nickel banded sleeves were inspected with the Ghent V2 probe during the Beaver Valley spring 2020 refueling outage (2R21), including 88 newly installed sleeves.

Noise measurements at the center of the nickel band region were taken for 50 sleeves in each LTR-CDMP-20-38 NP-Attachment Page 12 of 15 Rev. 0

• • • This record was final approved on 1/11/2021 7:55:57 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 of the three SGs. The sleeves were randomly selected amongst the various calibration groups from each SG. All sleeves that were selected had been in service for one or more cycles of operation. The 88 sleeves installed during 2R21 were not included in the noise evaluation. The method of noise measurement was the same as the method for the EPRI Appendix H qualification noise program as discussed in Reference 2 (SG-CDMP-19-19). The noise measurements were taken at the reporting frequency of 70 kHz with a window width of 60 degrees for axial oriented degradation for both peak-to-peak voltage amplitudes (Vpp) and for vertical maximum voltage amplitudes (Vvm).

Figure 6-1 provides the noise distributions for the 150 sleeve samples from the BVPS 2R21 field testing and for the noise distribution taken from the Electric Power Research Institute (EPRI) Appendix H qualification samples for both Vpp and Vvm voltage amplitudes. The Vpp noise distribution for the field data is nearly identical to the noise distribution assumed from the EPRI Appendix H qualification program up to a probability of about 80%. The field data distribution displays a larger distribution tail at higher voltage amplitudes, mainly resulting from about 12 outlying data points of higher noise. Review of these outlying data points showed that the larger voltage amplitudes were caused by a larger horizontal noise component that is outside the flaw phase angle window. Probe performance, such as probe wear, probe centering issues, and motor unit anomalies were determined not to be the source of the horizontal noise. The likely cause of the noise is local conditions within the tubesheet bore and parent tube or from hardrolling the sleeve into the tubesheet. The Vvm noise measurements provide the vertical component of the noise which lie within the flaw phase angle window and are less affected by horizontal noise components for flaw detection. Figure 6-1 also provides a comparison of the BVPS 2R21 field and EPRI Appendix H qualification sample Vvm distributions. The field Vvm distribution closely approximates the EPRI Appendix H qualification sample Vvm noise distribution without the large tail on the distribution as observed in the Vpp distribution.

Revision 2 of Reference 2 evaluates the effects of the BVPS 2R21 field noise distributions on the MAPOD distributions.

LTR-CDMP-20-38 NP-Attachment Page 13 of 15 Rev. 0

• • • This record was final approved on 1/11/2021 7:55:57 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 BVPS 2R21 Ghent V2 Nickel Band Noise, All SGs 70kHz, Axial, 60 Deg Window 0.9 0.8 0.7 0.6 LL 0 0.5

• BVPS 2R21 Field Noise, All SGS, Vpp u

• BVPS 2R21 Field Noise, All SGs, Vvm 0.4

• POD Study App H Qual, Vpp 0.3 o POD Study App H Qual, Vvm 0.2 0.1 0

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 Ghent V2 Voltage, 70 kHz Figure 6-1. Beaver Valley 2R21 Nickel Band Sleeve Ghent V2 Probe Noise Distribution

7. Section 3.8, Tube Sleeve Inspections, of Enclosure A (Evaluation of the Proposed Amendment) states that the Ghent V2 probe will be used to inspect the Alloy 800 tubesheet sleeves each outage. Please provide the following information about these inspections:

a. Clarify which probe (+POINT or Ghent V2) will be used for the call of record when inspecting different portions of the sleeve/parent tube assembly.

Response

To be provided by Energy Harbor.

b. Please confirm that each in-service sleeve will be inspected each refueling outage throughout the sleeve service life.

Response

To be provided by Energy Harbor.

The responses to Items 7a and 7b are the responsibility of Energy Harbor.

LTR-CDMP-20-38 NP-Attachment Page 14 of 15 Rev. 0

• • • This record was final approved on 1/11/2021 7:55:57 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 References

1. Email from Timothy L. Saibena, Energy Harbor to Jay R. Smith, Westinghouse, RE: [External]

Beaver Valley Units 1 and 2 - DRAFT Request for Additional Information - Steam Generator Tube Sleeve LAR (EPID L-2019 - LRA-0140), October 8, 2020. (Attached in EDMS)

2. Westinghouse Report SG-CDMP-19-19 P-Attachment, Revision 2, Probability of Flaw Detection in the Alloy 800 Mechanical Sleeve Lower Tubesheet Joint Using the Ghent Version 2 Eddy Current Probe, January 2021.

LTR-CDMP-20-38 NP-Attachment Page 15 of 15 Rev. 0

• • • This record was final approved on 1/11/2021 7:55:57 AM. (This statement was added by the PRIME system upon its validation)

Enclosure C L-20-277 Document Number SG-CDMP-19-19 NP-Attachment, Revision 2, "Probability of Flaw Detection in the Alloy 800 Mechanical Sleeve Lower Tubesheet Joint Using the Ghent Version 2 Eddy Current Probe, dated January 2021 (Non-proprietary)

(46 pages follow)

Westinghouse Non-Proprietary Class 3 SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Probability of Flaw Detection in the Alloy 800 Mechanical Sleeve Lower Tubesheet Joint Using the Ghent Version 2 Eddy Current Probe

@ Westinghouse

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 SG-CDMP-19-19 NP-Attachment Revision 2 Probability of Flaw Detection in the Alloy 800 Mechanical Sleeve Lower Tubesheet Joint Using the Ghent Version 2 Eddy Current Probe January 2021 Author:

Electronically Approved*

Jay R. Smith*

RSG/OSG Engineering and Chemistry Verifier: Reviewer:

Electronically Approved* Electronically Approved*

David A. Suddaby* Ronald J. Pocratsky Component Design and Management Programs Steam Generator Inspection and Repair Solutions Approved:

Electronically Approved*

Michael E. Bradley*, Manager Component Design and Management Programs

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

Westinghouse Electric Company LLC 1000 Westinghouse Drive Cranberry Township, PA 16066, USA

© 2021 Westinghouse Electric Company LLC All Rights Reserved SG-CDMP-19-19 NP-Attachment_Revision_2.docx

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 EXECUTIVE

SUMMARY

In October of 2018, FirstEnergy Nuclear Operating Company (FENOC) submitted a license amendment request (LAR) to extend the service life of Westinghouse Alloy 800 steam generator (SG) tube sleeves from five fuel cycles as stated in the plant Technical Specifications to eight fuel cycles. By Nuclear Regulatory Commission (NRC) letter dated February 25, 2019, Amendment 193 to Facility Operating License No.

NPF-73 was issued that extended the allowable service life of Alloy 800 sleeves as requested from five fuel cycles to eight cycles. This service life restriction was related to perceived limitations of the non-destructive examination (NDE) methods introduced by the nickel band applied to the outer diameter (OD) of the sleeve within the lower joint mechanical roll expansion region. Specifically, the NRC staff had questions regarding the capabilities of the current NDE inspection technique to detect flaws within the lower nickel banded joint of the tubesheet sleeve. The NRC staff noted in their approval for the extended sleeve service life that a qualified inspection technique would be needed to approve the tubesheet sleeve design on a permanent basis.

In response to the NRC staff questions regarding the NDE flaw detection capabilities, an improved inspection technique was developed specifically for the nickel banded area in the lower sleeve-to-tube joint within the tubesheet. The new eddy current inspection technique uses a magnetically biased transmit-receive rotating probe referred to as a Ghent Version 2 probe. Prior sleeve inspections used a motorized rotating +POINT' 1 probe coil. The Ghent Version 2 probe demonstrated improved inspection capabilities 0F as compared to the +POINT probe through reducing the interfering effects of the nickel band material. As part of another effort, the Ghent Version 2 probe inspection technique was qualified to the Electric Power Research Institute (EPRI) SG Examination Guideline Appendix H requirements. This document addresses the flaw detection capabilities through establishing defined probability of detection (POD) distributions for detecting stress corrosion cracking (SCC) within the parent tubing behind the nickel banded lower tubesheet sleeve joint using the Ghent Version 2 probe.

Tubing samples containing SCC flaws were obtained from EPRI and were used to determine the POD performance for the Ghent Version 2 probe. The crack samples were installed into tubesheet simulant collars, nickel banded sleeves were installed, and inspections were performed with the Ghent Version 2 probe. The crack samples were destructively examined to determine the physical flaw sizes. Engineering analysis was performed using the crack sample NDE inspection and destructive examination results to develop POD distributions. The engineering analysis methodology followed industry accepted modeling techniques endorsed by EPRI and is referred to the model assisted probability of detection (MAPOD) method. The MAPOD method was also described within the FENOC sleeve life extension License Amendment Request (LAR) and NRC staff approval documents described above. The exclusive use of crack samples to generate the POD distributions described within this report reduces uncertainty of combining different techniques as discussed in the prior LAR submittal and NRC staff safety evaluation 1

+POINT' is a trademark or registered trademark of Zetec, Inc. Other names may be trademarks of their respective owners.

SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 3 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 report that supported the LAR approval. Using the developed POD distributions and industry accepted operational assessment methodologies resulted in satisfaction of all SG performance criteria with considerable margin after one fuel cycle of operation, thus demonstrating the acceptability of the Ghent Version 2 probe to identify parent tube flaws behind the nickel band at the lower Alloy 800 tubesheet sleeve joint.

Revision 1 of this document was issued to update Reference 8 to Revision 1.

Revision 2 of this document is being issued to update a change in eddy current sizing of Flaw 1 in crack sample J-8. This change modifies the Ahat distribution, the signal-to-noise threshold values, and resultant probability of detection distributions discussed in Section 4.2. Additionally, this document is being updated to include the noise distribution obtained during the Ghent Version 2 probe field inspection of installed Alloy 800 nickel banded tubesheet sleeves from the spring 2020 Beaver Valley Unit 2 outage.

SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 4 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 TABLE OF CONTENTS EXECUTIVE

SUMMARY

........................................................................................................................... 3 LIST OF TABLES ........................................................................................................................................ 6 LIST OF FIGURES ...................................................................................................................................... 7 1 BACKGROUND ............................................................................................................................. 8 2 GHENT VERSION 2 PROBE DESIGN ....................................................................................... 11 3 GHENT VERSION 2 TEST PROGRAM ...................................................................................... 12 3.1 GENERAL OVERVIEW OF ANALYSIS METHOD ...................................................... 12 3.2 PARENT TUBE CRACK SAMPLES .............................................................................. 12 3.3 GHENT VERSION 2 PROBE INSPECTION OF NICKEL BAND TZ SLEEVES ........ 15 3.4 DESTRUCTIVE EXAMINATION .................................................................................. 20 4 PROBABILITY OF DETECTION MODELING AND RESULTS .............................................. 22 4.1 SIMPLE HIT-MISS POD MODEL .................................................................................. 22 4.2 NOISE-BASED MAPOD MODEL .................................................................................. 23 4.2.1 Development of the Ahat Distribution.............................................................. 23 4.2.2 Development of the Noise Distribution ............................................................ 28 4.2.3 S/N Threshold Value Development .................................................................. 30 4.2.4 MAPOD Simulations........................................................................................ 36 4.3 APPLICABILITY OF OD FLAW DETECTION TO ID FLAW ..................................... 41 4.4 EXAMPLE OPERATIONAL ASSESSMENT USING DEVELOPED POD CURVES .. 41 5 CONCLUSIONS ........................................................................................................................... 44 6 REFERENCES .............................................................................................................................. 45 SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 5 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 LIST OF TABLES Table 3-1. Ghent Version 2 Ni Band Sleeve Test Program Results............................................................ 13 Table 3-2. Destructive Examination Results .............................................................................................. 21 Table 4-1. Ghent Version 2 Probe Test Sample Noise ............................................................................... 33 Table 4-2a. Test Sample Signal-to-Noise Adjacent to Detected Flaws, Vpp ............................................. 33 Table 4-2b. Test Sample Signal-to-Noise Adjacent to Detected Flaws, Vvm ............................................ 34 Table 4-3a. Estimated Signal-to-Noise Ratio of Non-Detected Flaws, Vpp .............................................. 35 Table 4-3b. Estimated Signal-to-Noise Ratio of Non-Detected Flaws, Vvm ............................................ 36 Table 4-4. Test Sample J-2-3 Flaw Voltage Inside and Outside of Ni Band .............................................. 38 Table 4-5. Ghent V2 Probe MAPOD Simulation Results .......................................................................... 38 Table 4-6. Fully Probabilistic Operational Assessment Results................................................................. 43 SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 6 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 LIST OF FIGURES Figure 1-1. Alloy 800 Leak Limiting Mechanical Tubesheet Sleeve......................................................... 10 Figure 3-1. Parent Tube Crack Sample J-2-3, Post Sleeve Test Standard Ghent Probe C-Scan ................ 13 Figure 3-2. Parent Tube Crack Sample J-3, Post Sleeve Test Standard Ghent Probe C-Scan ................... 14 Figure 3-3. Parent Tube Crack Sample J-8, Post Sleeve Test Standard Ghent Probe C-Scan ................... 14 Figure 3-4. Parent Tube Crack Sample J-12, Post Sleeve Test Standard Ghent Probe C-Scan ................. 15 Figure 3-5. Sample J-2-3 Sleeve/Tube Assembly Ghent V2, Nickel Over Flaw #1 C-Scan ..................... 17 Figure 3-6. Sample J-2-3 Sleeve/Tube Assembly Ghent V2, Nickel Over Flaw #2/Flaw #4 C-Scan ....... 17 Figure 3-7. Sample J-2-3 Sleeve/Tube Assembly Ghent V2, Nickel Over Flaw #3 C-Scan ..................... 18 Figure 3-8. Sample J-3 Sleeve/Tube Assembly Ghent V2, Nickel Over All Flaws C-Scan ...................... 18 Figure 3-9. Sample J-8 Sleeve/Tube Assembly Ghent V2, Nickel Over All Flaws C-Scan ...................... 19 Figure 3-10. Sample J-12 Sleeve/Tube Assembly Ghent V2, Nickel Over Flaw C-Scan.......................... 19 Figure 3-11. Parent Tube Crack Samples Following Burst Testing ........................................................... 20 Figure 3-12. Typical Crack Morphology from LOM Surface Examination of Sample J-2-3 .................... 20 Figure 3-13. Typical SEM Fractograph for Sample J-2-3 Flaw #2 ............................................................ 21 Figure 4-1. Ghent Version 2 Probe Simple POD for Parent Tube SCC Behind Sleeve Nickel Band ........ 23 Figure 4-2a. Maximum Depth to Vpp Voltage Amplitude for Detected Flaws.......................................... 25 Figure 4-2b. Maximum Depth to Vvm Voltage Amplitude for Detected Flaws ........................................ 25 Figure 4-3a. Ghent Version 2 Probe Linear Ahat Function Correlation, Vpp ............................................ 26 Figure 4-3b. Ghent Version 2 Probe Linear Ahat Function Correlation, Vvm .......................................... 26 Figure 4-4a. Ghent Version 2 Probe Logarithm Ahat Function Correlation, Vpp ..................................... 27 Figure 4-4b. Ghent Version 2 Probe Logarithm Ahat Function Correlation, Vvm .................................... 27 Figure 4-5. Ghent Version 2 Probe Axial Noise Distribution, 70kHz 60 Degree Window ........................ 29 Figure 4-6a. Voltage-to-Depth Correlation Uncertainty, Vpp .................................................................... 34 Figure 4-6a. Voltage-to-Depth Correlation Uncertainty, Vpp .................................................................... 35 Figure 4-7a. Ghent Version 2 Probe Noise Based MAPOD for Parent Tube SCC Behind Sleeve Nickel Band, Vpp ......................................................................................................................... 39 Figure 4-7b. Ghent Version 2 Probe Noise Based MAPOD for Parent Tube SCC Behind Sleeve Nickel Band, Vvm ........................................................................................................................ 40 SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 7 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 1 BACKGROUND In 2008, First Energy Nuclear Operating Company (FENOC) submitted a License Amendment Request (LAR) to allow repair of Beaver Valley Unit 2 steam generator (SG) tubes with Alloy 800 mechanical tube sleeves (Reference 1). The request provided two sleeve designs; one to repair tubes at tube support plate locations (TS) and one to repair tubes at the top of the tubesheet transition zone locations (TZ). The tubesheet TZ sleeve design contains a nickel band at the lower sleeve roll joint to improve the leak limiting capabilities of the joint. Figure 1-1 provides the configuration of the tubesheet sleeve design. The upper joint of the TZ sleeve and the TS sleeve hydraulic expansion joints do not contain nickel bands. The presence of the nickel band material at the lower joint provided challenges and limited data demonstrating reliable detection of flaws within the parent tubing behind the nickel band region. As such, FENOC requested a limited service life of five-fuel cycles for the Alloy 800 sleeve. The NRC staff found the five-fuel cycle service life limitation to be acceptable and approved the LAR for the TZ and TS sleeves (Reference 2).

The inspection technique and program submitted to the NRC staff in support of the 2008 LAR relied upon

+POINT probe technology. The qualification and demonstration of the parent tube inspection behind the nickel band consisted of electro-discharge machine (EDM) notches and a limited number of samples with cracking in the parent tube. The four crack samples penetrated through or nearly through the thickness of the tube. The qualification and demonstration concluded that the +POINT probe could detect outer diameter (OD) EDM notch flaws with depths greater than 40% through-wall (TW) in the parent tube behind the nickel band region (Reference 2). It was concluded that the outer diameter stress corrosion cracking (ODSCC) flaws that are 100% TW or approaching 100% TW could also be detected with the +POINT probe. The NRC staff found the inspection capabilities of the +POINT probe to be acceptable for the limited service life of the nickel band TZ sleeve. The limitation on the service life of the Alloy 800 sleeve limits the amount of time that degradation of the sleeve joint could occur as discussed in Reference 2.

In 2018, FENOC submitted a LAR to the NRC to extend the nickel banded TZ sleeve service life from five fuel cycles to eight fuel cycles (Reference 3). The LAR also provided a clarification that the limited service life of five cycles does not apply to the TS sleeve as the TS sleeve joint design does not include a nickel band. The submittal as supplemented by Reference 4 and Reference 5 provided additional and updated analyses that demonstrated that the NDE capabilities were not as restrictive as initially perceived for detection of degradation behind the nickel band region at the lower sleeve joint. The NDE technique presented within the updated analyses was still the +POINT probe technology. Probability of detection (POD) curves were developed using the industry methodology referred to as Model Assisted Probability of Detection (MAPOD) that showed reliable detection capabilities for parent tube flaws behind the nickel band using the +POINT probe technology. The NRC staff continued to have questions regarding uncertainties with the MAPOD analysis. However, the NRC staff approved the LAR as requested to extend to nickel banded sleeve service life to eight fuel cycles (Reference 6). The NRC staff noted in the license amendment approval document that a qualified inspection technique would be needed for approval of the SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 8 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 leak-limiting Alloy 800 TZ sleeves on a permanent basis. The NRC staff also noted in Reference 6 that the technical basis for the MAPOD analysis (Reference 4) did not provide a direct correlation between inspection frequencies of 300 kHz for the parent tube with no sleeves and 70 kHz for the tube/sleeve configuration. These correlations were inherent to the MAPOD methodology and results. However, the MAPOD conclusions were in general agreement with the results from limited experimental assessments using nondestructive evaluation specimens with laboratory produced outer diameter stress corrosion cracks (ODSCC). The final NRC staff position as described in Reference 6 is that a time-limited service life of eight fuel cycles is justified and acceptable with the current +POINT probe inspection technology for the nickel banded region of the lower TZ sleeve joint. The NRC staff also approved the clarification that the TS sleeve does not have a time-limited service life.

In response to the NRC staff questions (Reference 6) regarding the NDE flaw detection capabilities, an improved inspection technique was developed specifically for the nickel banded area in the lower sleeve-to-tube joint within the tubesheet through a joint FENOC, Westinghouse and Zetec effort. The improved eddy current inspection technique uses a magnetically biased transmit-receive rotating probe referred to as a Ghent Version 2 probe.

This document addresses the flaw detection capabilities through development of probability of detection distributions for detecting stress corrosion cracking (SCC) in the parent tubing behind the nickel banded lower tubesheet joint using the Ghent Version 2 probe. Revision 2 of this document is being issued to update a change in eddy current sizing of Flaw 1 in crack sample J-8. This change modifies the Ahat distribution, the signal-to-noise threshold values, and resultant probability of detection distributions discussed in Section 4.2. Additionally, this document is being updated to include the noise distribution obtained during the Ghent Version 2 probe field inspection of installed Alloy 800 nickel banded tubesheet sleeves from the spring 2020 Beaver Valley Unit 2 outage.

SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 9 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 a,c Figure 1-1. Alloy 800 Leak Limiting Mechanical Tubesheet Sleeve SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 10 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 2 GHENT VERSION 2 PROBE DESIGN The eddy current probe initially qualified to inspect the Alloy 800 sleeve pressure boundary is the +POINT probe. The +POINT probe is the current regulatory basis for performing baseline and in-service inspections of installed sleeves at the upper and lower joints, the sleeve between the joints and the parent tube pressure boundary behind the TS and TZ sleeve (Reference 7). With the exception of behind the nickel band for the TZ sleeve, the +POINT probe is qualified to be adequate for flaw detection capability for flaws greater than 45% through-wall (TW) for the sleeve and 50% through-wall for the parent tube in accordance with Appendix H of the EPRI SG Examination Guidelines (Reference 10). As a result, the TZ sleeve is limited to a service life of eight fuel cycles due to the NRC staff questions surrounding inspection of the parent tube behind the nickel band at the lower tubesheet joint.

FENOC, Westinghouse, and Zetec Inc., collaborated to develop an improved eddy current probe to inspect the parent tube behind the nickel band at the lower tubesheet joint. Through feasibility studies and testing programs, a modified Ghent G3/G4 probe was selected for qualification. The Ghent probe is a magnetically biased rotating transmit-receive probe. The magnets within the probe serve to suppress the undesirable effects of the nickel. The final probe design is referred to as the Ghent Version 2 probe. This probe has an increased magnet strength of [ ]a,c,e from the prior prototype (Version 1) magnet strength of 50 MGO. The increased magnet strength improves the ability to saturate the nickel material and improves inspection capabilities of the parent tube behind the nickel band.

A subsequent prototype Ghent probe (Version 3) was tested that contained [ ]a,c,e but did not improve the inspection capability beyond that provided by the single [

]a,c,e contained in the Version 2 probe. No further improvements were judged as being obtainable with additional designs or modifications and therefore, the Version 2 probe was selected for qualification.

The Ghent Version 2 probe also contains a standard +POINT probe coil to improve field implementation efficiencies by allowing simultaneous data collection with both probe coils with a single probe tube insertion. The Ghent coils target inspection of the nickel band region at the lower hard roll joint and the

+POINT probe targets the remaining portions of the sleeve, including the hydraulic expansions at the upper joint.

The Appendix H qualification of the Ghent Version 2 probe was completed and documented in Reference 8.

This document provides the probability of detection development for stress corrosion cracking in the parent tube behind the nickel band at the lower TZ sleeve joint.

SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 11 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 3 GHENT VERSION 2 TEST PROGRAM 3.1 GENERAL OVERVIEW OF ANALYSIS METHOD The approach to developing the probability of detection curves for SCC in the parent tube behind the nickel band in the lower tubesheet sleeve joint involved analysis of parent tube crack samples through a detailed testing program. The parent tube crack samples were placed in tubesheet simulated collars and nickel banded Alloy 800 sleeves were installed in the tube/collar assemblies. The tube/sleeve/collar assemblies were inspected with the Ghent Version 2 probe and analyzed to determine which flaws were detected or not detected. The eddy current voltage amplitudes were recorded for the detected flaws. All parent tube crack samples were then destructively examined to determine the size of the cracks for both the detected and undetected flaws. The NDE and destructive examination (DE) results were evaluated to develop the crack-based POD curves using industry accepted methods.

3.2 PARENT TUBE CRACK SAMPLES Tubing samples containing SCC flaws were obtained from the EPRI to support this study. The samples were fabricated from Alloy 600 tubing that had a diameter of 0.875 inch and a nominal wall thickness of 0.050 inch. There were four tubing samples that contained a total of 12 laboratory produced axial ODSCC flaws of varying sizes. The parent tube samples were tested with the +POINT probe at a frequency of 300 kHz to obtain baseline data prior to beginning the Ghent probe sleeve testing program. Table 3-1 provides a listing of the samples and the baseline inspection results. This table also provides additional information discussed in later sections. The as-received baseline +POINT probe voltage amplitude recorded was in terms of peak-to-peak voltage (Vpp). The initial baseline +POINT probe voltages of the crack samples in the as-received condition ranged from 0.21 Vpp in Sample J-2-3 Flaw #4 to 18.07 Vpp in Sample J-3 Flaw

1. 3. As further discussed in Section 3.3, the depths of the flaws from destructive examination ranged from

[ ]b.

Each crack sample had locator marks to uniquely identify each flaw in terms of circumferential location around the tube from the locator mark as noted in Table 3-1. Figures 3-1 through 3-4 provide the standard Ghent probe C-scan graphic of each parent tube sample with identification of each flaw that were taken following the NDE test program and prior to the destructive examination program.

SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 12 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 Table 3-1. Ghent Version 2 Ni Band Sleeve Test Program Results As-Received Sleeve/Tube/Collar Post Test Parent Parent Tube Assembly Tube (No Sleeve)

Behind Behind Nickel Nickel Final Parent Tube Baseline Ghent Ghent Final Standard DE Exam Sample Flaw Location +POINT Version 2, Version 2, +POINT Ghent Depth Identification Number (degrees) Vpp Vpp Vvm Vpp Vpp %TW b

1 31 0.71 NDD NDD 0.74 1.20 2 71 1.49 0.35 0.31 1.49 2.62 J-2-3 3 102 0.73 NDD NDD 0.70 1.07 4 285 0.21 NDD NDD 0.21 0.27 1 91 NI NDD NDD 0.44 0.28 2 91 NI NDD NDD 0.44 0.63 J-3 3 168 18.07 3.76 3.45 20.49 21.60 4 278 0.26 NDD NDD 0.21 0.21 5 354 0.27 NDD NDD 0.26 0.59 1 46 0.86 0.23 0.19 0.86 1.38 J-8 2 228 0.40 NDD NDD 0.36 0.63 J-12A 1 312 1.02 0.39 0.38 1.41 2.55 NI - Not initially recorded NDD - No degradation detected Figure 3-1. Parent Tube Crack Sample J-2-3, Post Sleeve Test Standard Ghent Probe C-Scan SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 13 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 Figure 3-2. Parent Tube Crack Sample J-3, Post Sleeve Test Standard Ghent Probe C-Scan Figure 3-3. Parent Tube Crack Sample J-8, Post Sleeve Test Standard Ghent Probe C-Scan SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 14 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 Figure 3-4. Parent Tube Crack Sample J-12, Post Sleeve Test Standard Ghent Probe C-Scan 3.3 GHENT VERSION 2 PROBE INSPECTION OF NICKEL BAND TZ SLEEVES The test program for the Ghent Version 2 probe utilized test assemblies consisting of the parent tube crack sample, a nickel band TZ sleeve and a tubesheet simulant collar. Eddy current data was collected on each test assembly using the Ghent Version 2 probe.

The tubesheet simulant collar was a split collar fabricated from low alloy carbon steel. The split collar design provided the ability to use the assembled and then the disassembled for each of the parent tube crack samples. The tubesheet bore was machined to closely match the parent tube outer diameter. This was done to prevent the cracks from deforming or opening up due to ligament tearing when the parent tube was hard rolled into the tubesheet simulant collar. This was demonstrated through a previous test program which was discussed in the Question 5 response of Reference 5. Each parent tube crack sample was manually hard-rolled into the collar until firm contact was achieved to further prevent deformation of the cracks. As the Ghent probe is surface riding, not achieving full expansion to the full diameter of a prototypic in-generator tubesheet bore would have negligible effects on the eddy current results due to the small amount of tube wall thinning.

Eddy current data was collected for each parent tube/sleeve/collar test assembly using the Ghent Version 2 probe using the techniques and essential variables described in the Appendix H qualification documented in Reference 8. The data was collected at frequencies of 240/130/70/40 kHz. The 70kHz frequency channels are used to report degradation within the parent tube behind the nickel band at the lower sleeve SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 15 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 joint. All further references to the Ghent Version 2 eddy current results (i.e., voltage amplitude) herein are based upon the 70kHz reporting channel results.

Following analysis of the collected eddy current data, it was determined that four of the twelve parent tube flaws were detected through the nickel band region of the TZ sleeve using the Ghent Version 2 probe. From Section 3.4, the flaw depths from destructive examination depths of the detected flaws were [

]b. The depths of the flaws not detected ranged from [

]b. The eddy current graphics of the parent tube/sleeve/collar assemblies are shown in Figures 3-5 through 3-10. Because the sleeve nickel band could not completely encompass all of the multiple flaws contained in Sample J-2-3, it was necessary to re-position the sleeve and re-test the assembly two additional times. Figures 3-5, 3-6, and 3-7 show the results after each sleeve repositioning. Each figure clearly delineates the flaws that were within and not within the nickel band region.

Upon completion of the eddy current analysis of the test assemblies, the parent tube samples were removed from the tubesheet simulation split collar and the sleeves removed from the parent tubes. Additional eddy current was performed on each sample to obtain the final as-left condition prior to destructive examination.

Each parent tube sample (no sleeve) was tested with the +POINT probe and a non-modified standard Ghent G3/G4 probe. These results are shown in Table 3-1. Comparison of the pre-test and post-test +POINT probe results show little to no change in the voltage amplitudes. Therefore, the measures taken to avoid flaw ligament tearing during the test program were successful.

SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 16 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 Figure 3-5. Sample J-2-3 Sleeve/Tube Assembly Ghent V2, Nickel Over Flaw #1 C-Scan Figure 3-6. Sample J-2-3 Sleeve/Tube Assembly Ghent V2, Nickel Over Flaw #2/Flaw #4 C-Scan SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 17 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 Figure 3-7. Sample J-2-3 Sleeve/Tube Assembly Ghent V2, Nickel Over Flaw #3 C-Scan Figure 3-8. Sample J-3 Sleeve/Tube Assembly Ghent V2, Nickel Over All Flaws C-Scan SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 18 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 Figure 3-9. Sample J-8 Sleeve/Tube Assembly Ghent V2, Nickel Over All Flaws C-Scan Figure 3-10. Sample J-12 Sleeve/Tube Assembly Ghent V2, Nickel Over Flaw C-Scan SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 19 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 3.4 DESTRUCTIVE EXAMINATION The destructive examination of the parent tube crack samples was performed by Westinghouse at the Churchill laboratory facilities. Reference 9 provides the complete results of the destructive examination.

A summary of the key findings is provided below.

Each tube sample was internally pressure tested to facilitate opening of the tight cracks in preparation for the destructive examinations. The pressure tests resulted in fish-mouth openings of several of the deeper flaws and widened the crack openings of the shorter and shallow cracks. The pressure testing used internal bladders with copper foil covering the flaw to protect the bladder from the sharp edge of the crack should the tube burst. Figure 3-11 shows the four parent tube samples following the burst testing. Prior to the pressure test, each sample was examined using low magnification light optical microscopy (LOM) to assess the general surface condition and to further identify the presence of the cracking. This examination showed that the character of the flaws is typical of environmentally assisted axial stress corrosion originated on the tube outer diameter surface. Figure 3-12 shows a typical flaw morphology from the outer diameter surface as shown for Sample J-2-3.

Figure 3-11. Parent Tube Crack Samples Following Burst Testing Figure 3-12. Typical Crack Morphology from LOM Surface Examination of Sample J-2-3 SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 20 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 The destruction examination of the flaws consisted of scanning electron microscope (SEM) fractography to ascertain the radial flaw depths across the tube wall thickness. From the SEM fractography, the maximum flaw depths were recorded. Figure 3-13 provides a typical fractograph as shown for Sample J-2-3 Flaw #2. The recorded flaw depths from the destructive examination are provided in Table 3-2. The maximum depth of the twelve flaws range from [ ]b. The axial flaw lengths range from [ ]b.

Figure 3-13. Typical SEM Fractograph for Sample J-2-3 Flaw #2 Table 3-2. Destructive Examination Results Post Sleeve Max.

Standard Radial DE DE Ghent Max. Radial Depth of Flaw Flaw Tube Flaw Flaw Voltage Flaw Depth Flaw Depth Length Sample No. Location (Vpp) (µm) (mils) (%TW)(1) (inch) b 1 31 deg 1.2 2 71 deg 2.62 J-2-3 3 102 deg 1.07 4 285 deg 0.27 1 91 deg 0.28 2 91 deg 0.63 J-3 3 168 deg 21.6 4 278 deg 0.21 5 354 deg 0.59 1 26 deg 1.38 J-8 2 228 deg 0.63 J-12 1 340 deg 2.55 Notes:

(1) The percent through-wall flaw penetration is based on the nominal undegraded 50.0-mil wall thickness.

SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 21 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 4 PROBABILITY OF DETECTION MODELING AND RESULTS The results from the tube crack sample eddy current and destructive examinations performed and described in the previous sections provide the necessary information to develop the probability of detection parameters for SCC in the parent tube behind the nickel band at the lower sleeve joint. The methodology used to develop the POD distributions follow industry accepted methods, specifically binary (hit-miss)

Generalized Linear Modeling (GLM). The use of binary hit-miss GLM models for SG NDE applications is discussed in the EPRI SG Integrity Assessment Guidelines (Reference 11). For these applications NDE flaw detections, coded as 1 (hits), and flaw non-detections, coded as 0 (misses), are plotted using a structural parameter, typically flaw depth, and non-linearly regressed using a mathematical function to define the POD model. The mathematical functions used for SG NDE applications are typically the log-logistic and logistic functions, whereas the log-logistic function being more widely used. Binary hit-miss POD modeling can be performed for discrete data points directly from actual NDE results; herein referred to as the simple POD method. Binary POD modeling through probabilistic simulations by combining noise and structural parameter distributions to develop a large database of hits and misses for plotting is also used for SG applications. This method is known as the noise-based Model Assisted Probability of Detection (MAPOD) method. EPRI has developed a software package (Reference 13) to develop noise-based POD curves using the log-logistic function.

The simple hit-miss POD model and the noise-based MAPOD model are used to describe the capabilities of the Ghent Version 2 probe to detect SCC in the parent tube behind the nickel band at the lower TZ sleeve joint. These models and their results are discussed in the following sections.

4.1 SIMPLE HIT-MISS POD MODEL The simple hit-miss POD model uses the Ghent Version 2 probe eddy current data collected for each of the 12 flaws in the parent tube crack samples and their corresponding depths from destructive examination.

Table 3-1 provides the sleeve/tube/collar eddy current results for flaw detection or non-detection (NDD) using the Ghent Version 2 probe and the destructive examination depth. The Ghent Version 2 probe detected four of the twelve available flaws. The maximum depths of the detected flaws were [

]a,c,e. These detected flaws (hits) were coded as 1 for the binary modeling input. The remaining eight flaws were not detected. The non-detected flaw depths ranged from [

]a,c,e. These non-detected flaws (misses) were coded as 0 for the binary modeling input.

Figure 4-1 provides the resultant POD curves for the log-logistic and logistic functions for the binary GLM modeling. The POD performance for the log-logistic and logistic functions are essentially identical. The 95th percentile log-logistic POD, or POD(0.95), is 74.5% TW. The 50th percentile POD for both function types is 67.8% TW.

SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 22 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 a,c,e Figure 4-1. Ghent Version 2 Probe Simple POD for Parent Tube SCC Behind Sleeve Nickel Band 4.2 NOISE-BASED MAPOD MODEL The noise-based MAPOD model performs hit-miss simulations through probabilistic sampling of three distributions; a tube noise distribution, a flaw depth to flaw voltage amplitude distribution (Ahat), and a flaw signal-to-noise (S/N) threshold distribution for determining hits/misses. The POD model first samples a random depth and determines the corresponding flaw voltage amplitude from the Ahat distribution. A random noise voltage amplitude is sampled from the noise distribution. A S/N value is determined from this sampled data and compared to a sampled S/N threshold value. If the simulated flaw S/N value is greater than the sampled S/N threshold value, the sampled flaw depth is considered to be detected (a hit). Likewise, if the simulated flaw S/N is less than the sampled S/N threshold, then the sampled flaw depth is considered to be non-detected (a miss). This process is repeated 100,000 times to develop a robust hit-miss data set for input to the binary GLM model to produce the POD curve.

4.2.1 Development of the Ahat Distribution The flaw depth to flaw voltage amplitude correlation (Ahat) is derived from the Ghent Version 2 eddy current data described in Section 3.3 and the maximum depths determined through destructive examination SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 23 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 of the same flaws as described in Section 3.4. Ahat distributions can be used in terms of the peak-to-peak voltage amplitude (Vpp) or in terms of the vertical maximum amplitude (Vvm) to correlate the flaw voltage to flaw depth. Vpp Ahat distributions are typically used for flaws originating on the inner surface of the tube (ID) when the phase angles and/or the vertical component of the flaw are small or when there are no significant horizontal components of the noise signals. Vvm Ahat distributions are typically used when there are well-defined vertical components of the flaw signal. The noise distribution and signal-to-noise threshold values must be based on the same voltage type as the Ahat distribution (i.e., Vpp or Vvm). Figure 4-2a shows the correlation of the Ghent Version 2 probe raw Vpp voltage amplitude to the maximum depth from destructive examination for the flaws behind the nickel band that were detected by the Ghent Version 2 probe, while Figure 4-2b shows the Vvm voltage amplitude to the maximum flaw depth. The Ahat correlation used in the MAPOD model is in the form of the natural logarithm (Ln) of voltage amplitude on either depth (linear Ahat function) or the natural logarithm of depth (logarithm Ahat function). Figures 4-3a and 4-4a provide the Vpp Ahat correlations for the linear Ahat and logarithm Ahat functions, respectively. Figures 4-3b and 4-4b provide the Vvm Ahat correlations for the linear Ahat and logarithm Ahat functions. As shown on these figures, the linear Ahat (Ln(V) on depth) provides a slightly better correlation and slighter lower standard error. Therefore, the linear Ahat correlation will be used for the MAPOD simulations for both the Vpp and Vvm parameters.

SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 24 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 a,c,e Figure 4-2a. Maximum Depth to Vpp Voltage Amplitude for Detected Flaws a,c,e Figure 4-2b. Maximum Depth to Vvm Voltage Amplitude for Detected Flaws SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 25 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 a,c,e Figure 4-3a. Ghent Version 2 Probe Linear Ahat Function Correlation, Vpp a,c,e Figure 4-3b. Ghent Version 2 Probe Linear Ahat Function Correlation, Vvm SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 26 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 a,c,e Figure 4-4a. Ghent Version 2 Probe Logarithm Ahat Function Correlation, Vpp a,c,e Figure 4-4b. Ghent Version 2 Probe Logarithm Ahat Function Correlation, Vvm SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 27 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 4.2.2 Development of the Noise Distribution The Ghent Version 2 probe was developed and fabricated following the fall 2018 SG inspection outage at Beaver Valley Unit 2. At the time of the initial issuance of this document, there had been no field data collected with the Ghent Version 2 probe in which to build an in-generator noise distribution. Therefore, noise distributions for Vpp and Vvm voltage amplitudes were initially developed using the available data from the Ghent Version 2 probe EPRI Appendix H qualification program as described in Reference 8. This qualification program contained nine tube/sleeve collar test assemblies. The collected eddy current data was re-analyzed to measure the axial noise voltage amplitude at two locations per test assembly. The noise data was taken at the reporting frequency of 70kHz with a window width of 60 degrees that is appropriate for axially oriented degradation. Figure 4-5 provides the resultant noise distributions. This noise distribution from the Appendix H qualification test assemblies is a reasonable representation of in-generator noise as the effects of the nickel band provides the largest contributor to noise in the parent tube/sleeve configuration within the tubesheet.

The first field deployment of the Ghent Version 2 probe occurred in the spring of 2020 at BVPS Unit 2 (2R21). During this inspection, a total of 567 in service nickel banded sleeves were inspected with the Ghent V2 probe, including 88 newly installed sleeves. Noise measurements at the center of the nickel band region were taken for 50 sleeves in each of the three SGs. The sleeves were randomly selected amongst the various calibration groups from each SG. All sleeves that were selected had been in service for one or more cycles of operation. Sleeves installed during 2R21 were not included in this noise evaluation. The method of noise measurement was the same as utilized during the EPRI Appendix H qualification noise program as discussed above.

Figure 4-5 provides the noise distributions for the 150 sleeve sample from the BVPS 2R21 field testing and for the noise distribution taken from the EPRI Appendix H qualification samples for both Vpp and Vvm voltage amplitudes. The Vpp noise distribution for the field data is nearly identical to the noise distribution generated from the EPRI Appendix H qualification program up to a probability of about 80%. The field data distribution displays a larger distribution tail at higher voltage amplitudes, mainly resulting from about 12 outlying data points of higher noise. Review of these outlying data points showed that the larger voltage amplitudes were caused by a larger horizontal noise component that is outside the flaw phase angle window.

Probe performance, such as probe wear, probe centering issues, and motor unit anomalies were determined not to be the source of the horizontal noise. Additionally, the +POINT probe data of the parent tube prior to sleeve installation were also reviewed and no significant horizontal noise was noted. It is likely that the horizontal component of the noise may be attributed to hard rolling of the sleeve during installation or contained in the nickel band material from sleeve fabrication.

Noise-based POD distributions were developed using both the Vpp and Vvm noise distributions as discussed in the following sections.

SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 28 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 Figure 4-5. Ghent Version 2 Probe Axial Noise Distribution, 70kHz 60 Degree Window SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 29 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 4.2.3 S/N Threshold Value Development The noise based MAPOD model uses an upper and lower S/N threshold to determine if a simulated flaw S/N ratio results in flaw detection or non-detection. For each of the 100,000 simulations, a uniform distribution between the lower and upper S/N threshold values is randomly sampled to serve as the criteria for simulated flaw detection. Simulated flaw S/N values greater than the sampled S/N threshold are treated as flaw detections and likewise, flaw S/N values less than the threshold values are flaw non-detections. The upper and lower threshold values are determined by examining the results of the parent tube/sleeve/collar assembly eddy current data collected and described in Section 3.3. The lower S/N threshold value is determined by evaluating the detected flaw data, while the upper S/N threshold value is determined by evaluating the non-detected flaw data. S/N threshold values were determined separately for Vpp and Vvm voltage amplitude types.

S/N Threshold Value Development for Vpp

[

]a,c,e.

The Ghent Version 2 probe voltage-to-depth correlation shown in Figure 4-2a is used to estimate the Vpp voltage amplitude of undetected flaws. [

]a,c,e SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 30 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3

[

]a,c,e. This is illustrated in Figure 4-6a.

Table 4-3a provides the estimated S/N ratios of undetected flaws [

]a,c,e.

The S/N threshold values used in the MAPOD model when using Vpp voltage amplitudes were [

]a,c,e for the upper threshold.

S/N Threshold Value Development for Vvm The process applied for determining the upper and lower S/N threshold values for using the Vvm parameters is the same as applied for the Vpp parameters as discussed above. Table 4-1 provides the Vvm noise measurements from the parent tube/sleeve/collar assembly samples. Table 4-2b provides the S/N measurements for each detected flaw, excluding Flaw #3 in Sample J-3 which contained a 100% TW flaw.

The S/N measurements ranged from [

]a,c,e.

Similar to the process for determining the Vpp related upper S/N threshold, the Ghent Version 2 probe voltage-to-depth correlation shown in Figure 4-2b was used to estimate the Vvm voltage amplitude of undetected flaws along with application of a 10% voltage measurement uncertainty as described above.

Figure 4-6b shows the resultant depth-to-voltage uncertainty distributions. [

]a,c,e.

Table 4-3b provides the estimated S/N ratios of undetected flaws [

]a,c,e as discussed above. The estimated upper bound S/N values of undetected flaws range from

[ ]a,c,e. Therefore, an upper bound S/N threshold of [ ]a,c,e was applied.

SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 31 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 The S/N threshold values used in the MAPOD model when using Vvm voltage amplitudes were [

]a,c,e for the upper threshold.

SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 32 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 Table 4-1. Ghent Version 2 Probe Test Sample Noise Noise Sample Sample Sample Data Noise Noise I Average Minimum Maximum Sample Point Vpp Vvm- Noise Noise Noise b

~

1

- a,c,e 2

J-2-3 3

4 1

J-3 2 1

J-8 2 1

J-12 2

~

Table 4-2a. Test Sample Signal-to-Noise Adjacent to Detected Flaws, Vpp Ghent Noise S/N S/N S/N V2 Flaw Beside Using Using Using EPRI Flaw Voltage Flaw Average Minimum Maximum Sample No. Vpp Vpp Noise Noise Noise a,c,e b

J-2-3 2 0.35 J-8 1 0.23 J-12A 1 0.39 J-3 3 3.76 Average Excluding Sample J-3 Lower S/N Threshold Value with 10%

Uncertainty SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 33 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 Table 4-2b. Test Sample Signal-to-Noise Adjacent to Detected Flaws, Vvm Ghent Noise S/N S/N S/N V2 Flaw Beside Using Using Using EPRI Flaw Voltage Flaw Average Minimum Maximum Sample No. Vvm Vvm Noise Noise Noise a,c,e b

J-2-3 2 0.31 J-8 1 0.19 J-12A 1 0.38 J-3 3 3.45 Average Excluding Sample J-3 Lower S/N Threshold Value with 10%

Uncertainty a,c,e Figure 4-6a. Voltage-to-Depth Correlation Uncertainty, Vpp SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 34 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 a,c,e Figure 4-6b. Voltage-to-Depth Correlation Uncertainty, Vvm Table 4-3a. Estimated Signal-to-Noise Ratio of Non-Detected Flaws, Vpp Est. 95/50 Ghent V2 S/N Using Vpp using Standard EPRI Depth Minimum Error of 95/50 a,c,e Sample Flaw %TW b Noise Vpp 0.166 Vpp S/N 1

J-2-3 3 4

1 2

J-3 4

5 J-8 2 SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 35 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 Table 4-3b. Estimated Signal-to-Noise Ratio of Non-Detected Flaws, Vvm 95/50 S/N Using Est. using Standard EPRI Depth Ghent Minimum Error of 95/50 Sample Flaw %TW - bV2 Vvm Noise Vvm 0.155 Vvm S/N a,c,e 1 -

J-2-3 3 -- -

4 -- -

1 2

J-3 4

5 -- -

J-8 2 4.2.4 MAPOD Simulations POD simulations were initially performed using the noise-based MAPOD methodology with the EPRI Appendix H qualification sample noise distributions, linear Ahat correlation, and the S/N threshold values for Vpp and Vvm voltage amplitudes described above in Sections 4.2.1, 4.2.2, and 4.2.3. The resultant POD curves are shown in Figure 4-7a for the Vpp data set and Figure 4-7b for the Vvm data set. The 95th percentile POD values for the log-logistic functions are 98.9% TW and 92.6% TW for the Vpp and Vvm data sets respectively using the EPRI Appendix H qualification sample noise.

POD simulations were also performed using the BVPS 2R21 field noise distributions for both Vpp and Vvm and these distributions are also shown in Figures 4-7a and 4-7b. The field noise data resulted in 95th percentile POD values that are 0.9% TW to 2.6% TW higher than those using the EPRI Appendix H qualification sample noise.

Use of the logistic POD function produces POD(0.95) results that have 1.0% TW to 2.6% TW improvement over the log-logistic POD results. Table 4-5 provides a listing of all MAPOD simulation results, including the logistic POD function simulations.

Comparison of the simple POD (Figure 4-1) to the noise based MAPOD results (Figure 4-7a), initially shows an apparent disparity between the detection capabilities of each method. The simple POD results in a POD(0.95) of 74.5% TW whereas the noise-based MAPOD results in a POD(0.95) of 98.9% TW using the EPRI Appendix H qualification sample noise distribution. Further investigations discussed below have shown that the noise-based MAPOD method introduces unique conditions to the effect of the nickel band on the Ahat and noise distributions that are not present for applications with no nickel material. These effects account for noise both within the Ahat correlation and the noise distributions, thereby compounding the masking effects of the nickel material. The MAPOD methodology is based on the presumption that the full effects of noise and flaw masking are accounted for within the noise distribution. The nickel band SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 36 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 material in the sleeve/tube configuration reduces the flaw voltage amplitude in addition to introducing a noise component. This effect was demonstrated through the NDE program described in Section 3.3, with supplemental data analysis. The simple POD distribution (Figure 4-1) was generated with the flaw depths and the detection hit or miss result as the only inputs. The full effect of the nickel band material on detection is inherent in the detection result with no direct dependence on the developed noise and Ahat distributions.

Therefore, the simple POD distribution provides better representation of the actual flaw detection performance. To reduce the compounding masking effects of the nickel, an adjustment was applied to the noise-based MAPOD method that reduced the compounding masking effects of the nickel material to produce a POD distribution more comparable with the simple POD result as discussed below.

As discussed in Section 3.3, parent tube crack Sample J-2-3 contained multiple flaws that extended beyond the test sleeve nickel band. Therefore, the sleeve was re-positioned twice to ensure that each of parent tube flaws were eddy current tested within the center of the nickel band. The voltage amplitude of Flaw #2 was 0.36 Vpp when centered within the nickel band. When the sleeve was re-positioned, Flaw #2 was located on the edge of the nickel band. The effect of the nickel material on flaw signals at the edge of the band is

[

]a,c,e. A similar evaluation was performed for the Vvm voltage amplitudes of the Sample J-2-3 flaws with similar results that were bound by the Vpp results. The Vpp results are shown on Table 4-4.

A MAPOD simulation was performed using an optimized Ahat correlation to address the combined effect of nickel on the Ahat and noise distributions. To account for the compounding effects of the nickel material,

[

]a,c,e. The optimized linear Ahat correlation and the full noise distribution from the Appendix H qualification samples (Figures 4-5a and 4-5b) were used in the MAPOD simulations. Figures 4-7a and 4-7b provide the resultant POD curves for the Vpp and Vvm data sets. The optimized MAPOD simulations produce POD(0.95) improvements of 9.0% TW to 10.1% TW. The POD(0.95) values for the optimized MAPOD simulations are 88.8% TW and 83.7% TW for the Vpp and Vvm data sets respectively when using the EPRI Appendix H qualification sample noise distribution. The optimized MAPOD simulation result is more comparable to the simple POD result thus demonstrating a reasonable method to reduce compounding nickel material effects.

SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 37 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 Table 4-4. Test Sample J-2-3 Flaw Voltage Inside and Outside of Ni Band Edge of Edge of Center of Nickel Nickel Outside of Nickel Locn 1 Locn 2 Nickel Test Sample Vpp Vpp Vpp Vpp

---

=>- b J-2-3 Flaw # 1 NDD J-2-3 Flaw # 2 0.36 J-2-3 Flaw # 3 NDD -_

Table 4-5. Ghent V2 Probe MAPOD Simulation Results Voltage Noise S/N Ratio Model Type Distr. Thresholds POD(0.95) POD(0.50)

Log-Logistic POD Functions - - a,c,e Simple Hit-Miss N/A N/A EPRI App. H Qual Samples Vpp App. H EPRI App. H Qual Samples, Optimized Vpp App. H BVPS 2R21 Field Data Vpp Field -

BVPS 2R21 Field Data, Optimized Vpp Field EPRI App. H Qual Samples Vvm App. H -

EPRI App. H Qual Samples, Optimized Vvm App. H BVPS 2R21 Field Data Vvm Field BVPS 2R21 Field Data, Optimized Vvm App. H Logistic POD Functions Simple Hit-Miss N/A N/A -

EPRI App. H Qual Samples Vpp App. H -

EPRI App. H Qual Samples, Optimized Vpp App. H BVPS 2R21 Field Data Vpp Field BVPS 2R21 Field Data, Optimized Vpp Field EPRI App. H Qual Samples Vvm App. H -

EPRI App. H Qual Samples, Optimized Vvm App. H BVPS 2R21 Field Data Vvm Field -

BVPS 2R21 Field Data, Optimized Vvm App. H SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 38 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 a,c,e Figure 4-7a. Ghent Version 2 Probe Noise Based MAPOD for Parent Tube SCC Behind Sleeve Nickel Band, Vpp SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 39 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 a,c,e Figure 4-7b. Ghent Version 2 Probe Noise Based MAPOD for Parent Tube SCC Behind Sleeve Nickel Band, Vvm SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 40 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 4.3 APPLICABILITY OF OD FLAW DETECTION TO ID FLAW The principal degradation concern with a lower sleeve joint is the potential for primary water stress corrosion cracking (PWSCC) as a result of the stresses imparted to the tube due to the sleeve installation (Reference 7). Reference 4 describes the three elements necessary for SCC initiation; a corrosive environment, a susceptible material, and tensile residual stress. If any one of these elements are absent or reduced below some threshold, SCC cannot occur. The lower sleeve joint is located approximately in the middle of the full depth rolled tubesheet (neutral axis) thereby isolating the lower sleeve joint from the secondary side chemical environment. Therefore, ODSCC in the parent tube at the lower sleeve joint is not a credible or potential degradation mechanism and is substantiated by the lack of ODSCC within the tubesheet being reported by the industry. PWSCC has occurred within the industry in the expanded portion of the tubesheet in non-sleeved tubes. These occurrences were predominately located at bulges, overexpansions or other geometric discontinuities caused by tubesheet drilling anomalies during tubesheet fabrication. Even though Reference 4 concluded that the occurrence of PWSCC within the parent tube behind the nickel band at the lower sleeve joint is sufficiently low, the focus of NDE inspection techniques is detection of PWSCC.

The NDE test program and POD curves developed within this document were based on ODSCC flaw samples. This was due to the unavailability of PWSCC flaw samples. The detection of OD flaws provides a conservative estimate of ID flaws due to the nature of eddy current magnetic theory. Multiple test frequencies are used to concentrate the eddy current field at various depths of penetration through the material being tested. For the inspection of the parent tube/sleeve joint within the tubesheet, the test frequency of 70 kHz concentrates the eddy current field on the inner portion of the parent tube for optimum detection of ID originating flaws such as PWSCC. As the magnetic fields at 70 kHz frequency penetrate deeper into the parent tube material, the signal strength and the signal density decreases, thus decreasing the flaw detection capability. Consequently, detection of flaws originating at the outer parent tube surface is diminished as compared to flaws originating at the inside parent tube surface. This effect is more profound for shallower OD originating flaws which require greater depth of magnetic field penetration for flaw detection. Shallower flaws that may not be detected on the tube OD may be detected if they originated on the tube ID where the magnetic field signal strength and density is greater. Therefore, POD curves based on OD originating flaws conservatively bound the POD for ID originating flaws. The detection capability for ID originating flaws is improved over OD flaw detection at the same frequency.

4.4 EXAMPLE OPERATIONAL ASSESSMENT USING DEVELOPED POD CURVES Example operational assessments (OAs) were performed to demonstrate the acceptability of the Ghent Version 2 probe POD curves developed in Sections 4.1 and 4.2. The fully probabilistic OAs were performed in accordance with the guidance of Reference 11 for the cased listed in Table 4-6. The OAs assessed the SG performance criteria for probability of burst (POB) and probability of leakage (POL) over an inspection interval of one fuel cycle for the various POD and voltage type assumptions described SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 41 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 Sections 4-1 and 4-2. The SG performance criteria that must be satisfied for Beaver Valley Unit 2 are as follows:

• POB, 5%

• POL, 5%

• Upper 95th percentile steam line break leakage (SLB), 0.10 gpm (allotment for leakage not associated with an alternate repair criterion)

• Lower 5th percentile burst pressure, 4525 psi (i.e., 3 times normal operating differential pressure)

The OA evaluations assumed that the parent tube flaws were located in the straight leg freespan region of the tube with no sleeve or tubesheet present. The burst and leakage equations of Reference 14 provided the calculation methodologies. These are very conservative assumptions as the presence of the tubesheet behind the parent tube and the presence of the sleeve eliminates the burst potential and greatly reduces the postulated accident induced leakage compared to a freespan flaw.

The fully probabilistic OA begins with determining the undetected flaw size distribution in terms of depth and length. The undetected flaw depth distribution is generated using the developed POD curves. The flaw length distribution is assumed to be a uniform distribution from [ ]a,c inch in axial length.

The upper bound length extends to the full length of the nickel band and the lower bound length provides a conservative estimate for an undetected flaw. The number of undetected flaws is assumed to be three.

This is a conservative value as there have been no reports in the industry of flaws in the roll flat region of the parent tube at the lower sleeve joint. The industry maximum depth growth distribution provided in Reference 11 was applied to the beginning of cycle (BOC) flaw distribution. No length growth was applied as the BOC length distribution is a reasonable estimate of the flaw lengths at the end of the cycle. Each probabilistic OA run contained 100,000 simulations. The population of affected tubes is 567, which is the total number of TZ sleeves currently in service at Beaver Valley Unit 2 and all were assumed to be contained in one SG.

Seven fully probabilistic runs were performed. The only difference between the runs was the applied depth POD curve. All other inputs were the same. Case A used the simple hit/miss POD defined in Figure 4-1.

Case B and Case C used POD curves developed from the EPRI Appendix H qualification sample Vpp noise data set for both the linear Ahat and optimized Ahat functions. Cases D through G used the POD curves developed from the BVPS 2R21 field inspection of installed sleeves for the linear and optimized Ahat functions. Cases B, D, and F provide conservative estimates of POD due to the compounding effects of the nickel material on both the noise distribution and Ahat correlation.

Table 4-6 provides the results of the seven fully probabilistic OA simulations. For all cases, the SG performance criteria were satisfied following one fuel cycle of operation. Case A resulted in a POB of 1.15%, a POL of 0.034%, a burst pressure of 5355 psi and no SLB leakage with application of the simple hit/miss curve of Figure 4-1. In comparison, the POB for the unadjusted linear Ahat function cases ranged SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 42 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 from 3.75% to 4.08% for both the noise and voltage type assumptions. The field noise assumption cases produced slightly higher but similar burst and leakage parameter results. The voltage type assumption cases (Cases F and G) resulted in nearly identical results as their Vpp assumption counterparts (Cases D and E).

The optimized Ahat function Cases C, E, and G produced simulation results that were similar and were improved from the unadjusted linear Ahat assumption cases. The POB results for the optimized Ahat function cases ranged from 2.26% to 2.42%. This was an expected result as the associated POD curves were between the simple hit-miss and unadjusted linear POD curves.

All fully probabilistic simulations satisfied the SG performance criteria for structural and leakage integrity.

The results assumed parent tube flaws located in the freespan without the constraining effects of the tubesheet or presence of a tube sleeve. For the actual tube/sleeve/tubesheet configuration, parent tube burst is not possible, and leakage would be greatly diminished.

The results of these OA simulations demonstrate the acceptability of the Ghent Version 2 probe and the associated POD curves to detect parent tube flaws with SCC behind the nickel band region in the lower tubesheet sleeve joint.

Table 4-6. Fully Probabilistic Operational Assessment Results Lower 5th Upper 95th POD Percentile Percentile Log-Logistic POD Voltage Probability Probability of Burst SLB Leakage Case Model/Data Set Type of Burst (%) Leak (%) Pressure (psi) (gpm)

A Simple Hit-Miss N/A 1.15 0.034 5355 0 B EPRI App Qual Sample Vpp 3.75 0.165 4711 0 EPRI App Qual Sample, 2.26 0.103 5026 0 C Optimized Vpp D BVPS 2R21 Field Data Vpp 4.08 0.190 4656 0 BVPS 2R21 Field Data, 2.42 0.096 4976 0 E Optimized Vpp F BVPS 2R21 Field Data Vvm 3.84 0.183 4690 0 BVPS 2R21 Field Data, G Optimized Vvm 2.32 0.095 5005 0 Performance Criteria 5 5 4525 0.10 SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 43 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 5 CONCLUSIONS In response to the NRC staff questions regarding the NDE flaw detection capabilities, the Ghent Version 2 was developed specifically to improve inspection capabilities for the nickel banded area in the lower sleeve-to-tube joint within the tubesheet. The Ghent Version 2 probe demonstrated improved inspection capabilities as compared to the +POINT probe through reducing the interfering effects of the nickel band material. Through a testing program of axial ODSCC flaws, the Ghent Version 2 probe was able to detect cracks from [ ]b TW in simulated parent tube/sleeve/tubesheet collar assemblies. A POD curve was generated based upon the detection and non-detection of the test assembly flaws that resulted in a 74.5% TW at the 95th percentile and 67.8% at the 50th percentile. Noise based MAPOD distributions were developed that addressed the effect of nickel material on eddy current noise and flaw voltage amplitudes.

Noise distributions obtained from the EPRI Appendix H qualification samples and the BVPS 2R21 field deployment of the Ghent V2 probe for both voltage types (Vpp and Vvm) were used for the MAPOD distribution development. The OA process using these developed POD curves resulted in satisfaction of all SG performance criteria with margin after one fuel cycle of operation, thus demonstrating the acceptability of the Ghent Version 2 probe to identify parent tube flaws behind the nickel band at the lower Alloy 800 TZ sleeve joint. Through implementation of a 100% Ghent Version 2 probe inspection of all in-service Alloy 800 sleeves each outage, the eight-cycle service life restriction can be eliminated.

SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 44 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 6 REFERENCES

1. First Energy Letter No. L-08-307, Beaver Valley Power Station Unit No. 2 Docket No. 50-412, License No. NPF-73, License Amendment Request No.07-007 Alloy 800 Steam Generator Tube Sleeving, October 2008. (NRC ADAMS Accession No. ML082890823)

2. USNRC Letter, Beaver Valley Power Station, Unit 2 - Issuance of Amendment RE: The Use of Westinghouse Leak-limiting Alloy 800 Sleeves for Steam Generator Tubes Repair (TAC No.

MD9969), September 2009. (NRC ADAMS Accession No. ML092590189)

3. First Energy Letter No. L-18-081, Beaver Valley Power Station Unit No. 2 Docket No. 50-412, License No. NPF-73, Steam Generator Technical Specification Amendment Request, March 28, 2018. (NRC ADAMS Accession No. ML18087A293)

4. Westinghouse Letter LTR-SGMP-18-3 P-Attachment, Revision 0, Steam Generator Alloy 800 Nickel Band Tubesheet Sleeve Operating Cycle Length Extension License Amendment Request:

Technical Basis, March 2018.

5. Westinghouse Letter LTR-SGMP-18-40 P-Attachment, Revision 0, Responses to Request for Additional Information Regarding Beaver Valley Power Station Unit No. 2 Amendment Request for Use of Westinghouse Leak-Limiting Alloy 800 Sleeves in Steam Generators, October 2018.

6. USNRC Letter, Beaver Valley Power Station, Unit 2 - Issuance of Amendment 193 RE: Revise Steam Generator Technical Specifications (EPID L-2018-LLA-0075), February 2019. (NRC ADAMS Accession No. ML18348B206)

7. Westinghouse Report WCAP-15919-P, Revision 2, Steam Generator Tube Repair for Westinghouse Designed Plants with 7/8 Inch Inconel 600 Tubes Using Leak Limiting Alloy 800 Sleeves, January 2006.

8. Westinghouse Report SG-CDMP-19-17-P, Revision 1, Qualification of an Examination Technique to Inspect Parent Tube Flaws Adjacent to the Nickel Band of an Alloy 800 Sleeve at Beaver Valley Unit 2, April 2020.

9. Westinghouse Report RT-TR-19-43, Revision 0, Steam Generator Tube OD Flaw Sizing in Support of Ghent Probe POD for Beaver Valley Unit 2 Nickel Band Sleeving, November 2019.

10. EPRI Technical Report 3002007572, Steam Generator Management Program: Pressurized Water Reactor Steam Generator Examination Guidelines: Revision 8, June 2016.

11. EPRI Technical Report 3002007571, Steam Generator Management Program: Steam Generator Integrity Assessment Guidelines: Revision 4, June 2016.

12. EPRI Technical Report 3002007856, Steam Generator Management Program: Steam Generator In Situ Pressure Test Guidelines: Revision 5, June 2016.

13. EPRI Product 30020010334, Steam Generator Management Program: Model Assisted Probability of Detection Using R (MAPOD-R) Version 2.1, 2017.

SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 45 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3

14. EPRI Technical Report 3002005426, Steam Generator Management Program: Steam Generator Degradation Specific Management Flaw Handbook, Revision 2, October 2015.

SG-CDMP-19-19 NP-Attachment January 2021 Revision 2 Page 46 of 46

• • • This record was final approved on 1/11/2021 8:26:44 AM. (This statement was added by the PRIME system upon its validation)

Enclosure E L-20-277 Affidavit for Withholding Proprietary Information (3 pages follow)

Westinghouse Non-Proprietary Class 3 CAW-21-5140 Page 1 of 3 COMMONWEALTH OF PENNSYLVANIA:

COUNTY OF BUTLER:

(1) I, Zachary S. Harper, have been specifically delegated and authorized to apply for withholding and execute this Affidavit on behalf of Westinghouse Electric Company LLC (Westinghouse).

(2) I am requesting the proprietary portions of LTR-CDMP-20-38 P-Attachment, Revision 0, Responses to Request for Additional Information Regarding Beaver Valley Power Station Unit No. 2 Steam Generator Tube Inspection and Repair, and SG-CDMP-19-19 P-Attachment, Revision 2, Probability of Flaw Detection in the Alloy 800 Mechanical Sleeve Lower Tubesheet Joint Using the Ghent Version 2 Eddy Current Probe, be withheld from public disclosure under 10 CFR 2.390.

(3) I have personal knowledge of the criteria and procedures utilized by Westinghouse in designating information as a trade secret, privileged, or as confidential commercial or financial information.

(4) Pursuant to 10 CFR 2.390, the following is furnished for consideration by the Commission in determining whether the information sought to be withheld from public disclosure should be withheld.

(i) The information sought to be withheld from public disclosure is owned and has been held in confidence by Westinghouse and is not customarily disclosed to the public.

(ii) The information sought to be withheld is being transmitted to the Commission in confidence and, to Westinghouses knowledge, is not available in public sources.

(iii) Westinghouse notes that a showing of substantial harm is no longer an applicable criterion for analyzing whether a document should be withheld from public disclosure. Nevertheless, public disclosure of this proprietary information is likely to

• • • This record was final approved on 1/11/2021 11:22:31 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 CAW-21-5140 Page 2 of 3 cause substantial harm to the competitive position of Westinghouse because it would enhance the ability of competitors to provide similar technical evaluation justifications and licensing defense services for commercial power reactors without commensurate expenses. Also, public disclosure of the information would enable others to use the information to meet NRC requirements for licensing documentation without purchasing the right to use the information.

(5) Westinghouse has policies in place to identify proprietary information. Under that system, information is held in confidence if it falls in one or more of several types, the release of which might result in the loss of an existing or potential competitive advantage, as follows:

(a) The information reveals the distinguishing aspects of a process (or component, structure, tool, method, etc.) where prevention of its use by any of Westinghouse's competitors without license from Westinghouse constitutes a competitive economic advantage over other companies.

(b) It consists of supporting data, including test data, relative to a process (or component, structure, tool, method, etc.), the application of which data secures a competitive economic advantage (e.g., by optimization or improved marketability).

(c) Its use by a competitor would reduce his expenditure of resources or improve his competitive position in the design, manufacture, shipment, installation, assurance of quality, or licensing a similar product.

(d) It reveals cost or price information, production capacities, budget levels, or commercial strategies of Westinghouse, its customers or suppliers.

(e) It reveals aspects of past, present, or future Westinghouse or customer funded development plans and programs of potential commercial value to Westinghouse.

• • • This record was final approved on 1/11/2021 11:22:31 AM. (This statement was added by the PRIME system upon its validation)

Westinghouse Non-Proprietary Class 3 CAW-21-5140 Page 3 of3 (f) It contains patentable ideas, for which patent protection may be desirable.

( 6) The attached documents are bracketed and marked to indicate the bases for withholding. The justification for withholding is indicated in both versions by means oflower-case letters (a) through (f) located as a superscript immediately following the brackets enclosing each item of information being identified as proprietary or in the margin opposite such information. These lower-case letters refer to the types of information Westinghouse customarily holds in confidence identified in Sections (5)(a) through (f) of this Affidavit.

I declare that the averments of fact set forth in this Affidavit are true and correct to the best of my knowledge, information, and belief.

I declare under penalty of perjury that the foregoing is true and correct.

Executed on: / /<g_ / tl,c)#,/

I/

Licensing Engineering

• • • This record was final approved on 1/11/2021 11:22:31 AM. (This statement was added by the PRIME system upon its validation)