ML19158A265

From kanterella
Jump to navigation Jump to search

Draft Safety Evaluation by the Office of Nuclear Reactor Regulation for Pressurized Water Reactor Owners Group Topical Report Pwrog-15109-NP, PWR Pressure Vessel Nozzle Appendix G Evaluation, Revision 0 (Pa-MSC-1091, Revision 4) February 20
ML19158A265
Person / Time
Site: 99902037
Issue date: 06/25/2019
From: James Drake
NRC/NRR/DLP/PLPB
To:
J. Drake, NRR/DLP, 301-415-8378
Shared Package
ML19158A230 -Pkg. List:
References
EPID L-2018-TOP-0009
Download: ML19158A265 (20)
v • d • e


Text

1 DRAFT SAFETY EVALUATION 2 BY THE OFFICE OF NUCLEAR REACTOR REGULATION 3 FOR PRESSURIZED WATER REACTOR OWNERS GROUP TOPICAL REPORT 4 PWROG-15109-NP, REVISION 0, PWR PRESSURE VESSEL 5 NOZZLE APPENDIX G EVALUATION 6 EPID L-2018-TOP-0009 7

8

1.0 INTRODUCTION

9 10 By letter dated March 5, 2018 (Agencywide Documents Access and Management System 11 (ADAMS) Accession No. ML18067A228), as supplemented by letter dated March 27, 2019 12 (ADAMS Accession No. ML19091A089), the Pressurized Water Reactor (PWR) Owners Group 13 (PWROG) submitted to the U.S. Nuclear Regulatory Commission (NRC) topical report (TR) 14 PWROG-15109-NP, Revision 0, PWR Pressure Vessel Nozzle Appendix G Evaluation, 15 (ADAMS Accession No. ML18067A229) for review and approval.

16 17 The TR addresses the potential for pressure-temperature (P-T) limit curves (or P-T limits) for 18 inlet or outlet nozzle corners of pressurized water reactors (PWRs) to be more limiting than 19 those of the shell (and associated welds) of the traditional beltline region of the reactor 20 pressure vessel (RPV). The PWROG developed the TR to demonstrate that the RPV nozzle 21 corner P-T limits are bounded by the NRC-approved P-T limits of the shell (and associated 22 welds) in the RPV traditional beltline region for a 60-year license for U.S. PWRs. Specifically, 23 the TR presented generic PWR fracture mechanics analyses of RPV inlet and outlet nozzle 24 corners to show that P-T limits for nozzles corners, developed in accordance with the 25 requirements of Appendix G, Fracture Toughness Requirements, to Title 10 of the Code of 26 Federal Regulations (10 CFR) Part 50, are bounded by the P-T limits of the shell (and 27 associated welds) in the RPV traditional beltline region.

28 29

2.0 REGULATORY EVALUATION

30 31 The NRC has established requirements in 10 CFR Part 50 to protect the integrity of the reactor 32 coolant pressure boundary in nuclear power plants. The NRC staff (the staff) evaluates the 33 acceptability of a facility's proposed P-T limits based on the following NRC regulations and 34 guidance:

35 36

  • Section 50.60 of 10 CFR, Acceptance criteria for fracture prevention measures for 37 lightwater nuclear power reactors for normal operation, imposes fracture toughness and 38 material surveillance program requirements, which are set forth in 10 CFR Part 50, 39 Appendices G and H, Reactor Vessel Material Surveillance Program Requirements.

40 41

  • Appendix G to 10 CFR Part 50 requires that a facilitys P-T limits for the RPV be at least as 42 conservative as those obtained by following the methods of analysis and the margins of 43 safety in Appendix G to Section XI of the American Society of Mechanical Engineers Boiler 44 and Pressure Vessel Code (ASME Code).

45 Enclosure

1 The most recent version of Appendix G to Section XI of the ASME Code which has been 2 endorsed in 10 CFR 50.55a, and therefore by reference in 10 CFR Part 50, Appendix G, is the 3 2013 Edition of the ASME Code. Calculations of P-T limits are based, in part, on the nil-ductility 4 reference temperature (RTNDT) for the material, as specified in the ASME Code,Section XI, 5 Appendix G. The RTNDT is the critical parameter for determining the critical or reference stress 6 intensity factor (fracture toughness, KIC) for the material. As required by 10 CFR Part 50, 7 Appendix G, RTNDT values for materials in the RPV beltline region shall be adjusted to account 8 for the effects of neutron irradiation. Regulatory Guide (RG) 1.99, Revision 2, Radiation 9 Embrittlement of Reactor Vessel Materials, contains methodologies for calculating the adjusted 10 RTNDT (ART) due to neutron irradiation.

11 12 Appendix G to 10 CFR Part 50 defines the beltline or beltline region of the reactor vessel as the 13 region of the RPV (shell material including welds, heat affected zones, and plates or forgings) 14 that directly surrounds the effective height of the active core and adjacent regions of the RPV 15 that are predicted to experience sufficient neutron irradiation damage to be considered in the 16 selection of the most limiting material with regard to radiation damage.

17 18 Determination of the P-T limits for a plant in accordance with the requirements of Appendix G to 19 10 CFR Part 50 considers several factors, which include the initial properties and chemical 20 composition of the RPV materials, the accumulated neutron fluence for each material, the stress 21 levels applied to the materials resulting from heatup and cooldown transients (which include 22 internal pressure and thermal gradient loads), and structural discontinuities such as nozzles.

23 Development of P-T limits for the beltline region of the RPV considers not only the RPV shell 24 material but also other RPV materials with structural discontinuities such as nozzles.

25 26 3.0

SUMMARY

OF THE TR 27 28 The TR is organized as follows:

29 30 Section 1, Background - provides a background of why nozzle corners must be considered in 31 evaluations of P-T limits, a summary of the NRC-approved methodologies for development of P-T 32 limits for Westinghouse Electric Company (Westinghouse), Combustion Engineering, Inc., and 33 Babcock & Wilcox Company (B&W) PWR designs, a summary of reports that inform nozzle 34 corner analyses, and a summary of low-temperature overpressure protection.

35 36 Section 2, Flaw Size - describes the basis for postulating a smaller than quarter-thickness 37 (1/4T) flaw and describes the small flaw size models postulated in the inlet and outlet nozzles.

38 39 Section 3, Fracture Toughness - provides details of the determination of generic nozzle 40 fracture toughness using the master curve approach and generic embrittlement trend curve.

41 42 Section 4, Stress Intensity Factor Calculation - provides details of the determining stress 43 intensity factors (SIFs) using the finite element method for the small flaw size models in 44 Section 2 of the TR.

45 46 Section 5, Pressure-Temperature Limit Curves - describes determination of the P-T limits for 47 nozzle corners with a small flaw (using information from Sections 3 and 4 of the TR) and 1/4T 48 flaw; compares P-T limits for nozzles corners with those from NRC-approved P-T limits for shell 49 (and associated welds) in the RPV beltline region.

50 51 Section 6, Conclusion - concludes that the generic P-T limits for nozzles corners developed in 52 the TR in accordance with the requirements of Appendix G to 10 CFR Part 50 are bounded by

1 the P-T limits of the shell (and associated welds) in the RPV beltline region in the U.S. PWR 2 fleet.

3 4

4.0 TECHNICAL EVALUATION

5 6 The staff reviewed the TR to determine whether the PWROGs evaluation to demonstrate the 7 P-T limits of shell (and associated welds) in the RPV traditional beltline region bound those of 8 inlet and outlet nozzle corners is acceptable. The staff also reviewed the TR to determine that 9 the technical bases are consistent with the requirements of 10 CFR 50.60.

10 11 The staff evaluated eleven major topics of the TR. Each topic is addressed in the subsections of 12 the safety evaluation (SE) that follow. Within each major topic, the staff summarized the 13 relevant content of a subsection of the TR or described the relevant information in the 14 subsection that falls under each major topic. Then the staff provided its findings or 15 determinations on the TR subsection or on the major topic.

16 17 4.1 Postulated Flaw Size 18 19 In Section 2 of the TR, the PWROG explained that a traditionally postulated 1/4T flaw in the 20 nozzle corner region can result in a depth of approximately 4 to 5 inches as measured at a 21 45-degree angle from the nozzle corner to the RPV outside surface since the nozzle and RPV 22 are thicker in the vicinity of nozzles per the ASME Code design requirements. Crack driving 23 forces for a postulated 1/4T flaw could lead to overly conservative P-T limits. Therefore, the 24 PWROG opted to postulate smaller flaws as allowed in the 2008 edition of the ASME Code, 25 Section XI, Appendix G, Subarticle G-2120, Maximum Postulated Defect, which states that 26 flaws less than 1/4T may be used on an individual case basis if a smaller size of maximum 27 postulated defect can be ensured. Additionally, the 2008 edition of the ASME Code,Section XI, 28 Appendix G, Paragraph G-2223(a), Toughness Requirements for Nozzles, states that 29 examination methods shall be sufficiently reliable and sensitive to detect these smaller 30 defects. The PWROG created finite element models (FEMs) with a small postulated flaw in the 31 nozzle corner (the flaw penetrates 0.5 inch into the low alloy steel (LAS) from the clad-to-LAS 32 interface) to determine SIFs since closed-form SIF solutions for nozzle corners are typically for 33 1/4T flaws. Additionally, the PWROG created FEMs with a postulated flaw that penetrates 34 0.05 inch into the LAS to address the effect of the difference in coefficient of thermal expansion 35 (CTE) between the clad and the LAS.

36 37 The PWROG showed a probability of detection (POD) plot for vessels that indicates a POD of 38 100 percent for a 0.5-inch flaw into the LAS and stated that crack growth analyses have been 39 performed for postulated flaws smaller than 0.5 inch based on their high POD. Although a 40 similar POD for nozzle corners does not exist, the PWROG qualitatively concluded that the POD 41 for nozzle corners would be high because pre-service examination through ultrasonic testing 42 (UT) was performed from the inside surface and presented a conclusion by the Electric Power 43 Research Institute (EPRI) Nondestructive Examination Center that detecting flaws as small as 44 0.25 inch by UT located in RPV nozzles is excellent.

45 46 The staff reviewed the POD information for vessels in the TR, which the PWROG obtained from 47 Performance Demonstration Initiative (PDI) data from UT performed in accordance with ASME 48 Code,Section XI, Appendix VIII. The staff also reviewed the ASME Code,Section XI, 49 examination requirements for nozzle corners, which requires nozzle corners to be examined by 50 UT through PDI in accordance with ASME Code,Section XI, Appendix VIII. Accordingly, the 51 staff determined that the PWROGs qualitative evaluation of high POD for nozzle corners to be 52 reasonable. Thus, the staff determined that the postulated flaw size of 0.5 inch into the LAS

1 meets the detectability criterion of Paragraph G-2223(a) of the 2013 edition of ASME Code, 2 Section XI, which is the latest NRC-approved version of the ASME Code.

3 4 The staff noted that the ASME Code,Section XI, examination volume requirement for nozzle 5 corners specifies a maximum depth of 0.5 inch into the LAS. Thus, the staff determined that the 6 postulated flaw size of 0.5 inch into the LAS meets the required examination volume.

7 8 The staff noted that Paragraph G-2223(a) in the 2008 edition is different than in the 2013 9 edition. In addition to the detectability criterion described in Section 4.1 of this SE, 10 Paragraph G-2223(a) of the 2013 edition states that the postulated smaller flaw must 11 appropriately consider the combined effects of internal pressure, external loading, thermal 12 stresses, and flaw shape, and the postulated smaller flaw shall be no smaller than the 13 applicable inservice inspection criteria in Table IWB-3410-1 of ASME Code,Section XI. The 14 staff reviewed Section 4.7, Loads, of the TR and determined that the PWROG applied the 15 appropriate loads and flaw shape (evaluated in Section 4.6 of this SE). The staff also reviewed 16 the flaw size requirements in Table IWB-3410-1 of the ASME Code,Section XI, and determined 17 that the flaw size of 0.5 inch into the LAS region that the PWROG postulated meets the 18 requirements of the table. Based on this and the POD information the PWROG provided, the 19 staff finds that a postulated flaw of 0.5 inch into the LAS is acceptable and meets the criteria of 20 Subarticle G-2120 and Paragraph G-2223(a) of the 2013 edition of ASME Code,Section XI.

21 22 4.2 Fracture Toughness 23 24 Generic Nozzle Forging Master Curve Reference Temperature 25 26 In Section 3.1, Generic Nozzle Forging Master Curve Reference Temperature of the TR, the 27 PWROG stated that the use of the lower bound plane-strain, static fracture toughness (KIC) 28 curve has inherent margin since RTNDT is a conservative method for locating the KIC curve.

29 RTNDT is based on drop weight testing, which is a crack arrest transition temperature 30 measurement, and the Charpy impact test, which is a blunt notch impact test. These data are 31 conservatively bounded by the KIC curve, which is a lower bound crack initiation fracture 32 toughness curve.

33 34 In contrast, the PWROG stated that the master curve method is based on an initiation transition 35 temperature true fracture toughness test technique and the master curve index temperature (T0) 36 provides a much more accurate measure of the material fracture toughness. The PWROG 37 explained that existing master curve fracture toughness data for A-508 Class 2 type forgings 38 was gathered to establish a generic mean and standard deviation for alternate RTNDT for the 39 U.S. PWR inlet and outlet nozzles. Specifically, the master curve fracture toughness data is 40 used with ASME Section XI Code Case N-629, Use of Fracture Toughness Test Data to 41 Establish Reference Temperature for Pressure Retaining Materials,Section XI, Division 1, 42 which is endorsed by RG 1.147 and incorporated by reference in 10 CFR 50.55a as an 43 alternative to RTNDT.

44 45 The staff noted that the 2013 edition of ASME Code Section XI (i.e., the latest edition endorsed 46 by 10 CFR 50.55a) permits the use of an alternate RTNDT, which is consistent with Code Case 47 N-629. Specifically, Subarticle G-2110 in the 2013 edition of ASME Code Section XI states, in 48 part, that if material-specific temperature value, T0, for ferritic steels in the transition range is 49 available then a reference temperature, RTT0, may be used in place of RTNDT.

50 51 Since Code Case N-629 is incorporated by reference (e.g., RG 1.147) in 10 CFR 50.55a, and 52 the use of RTT0 in lieu of RTNDT is permitted by ASME Code Section XI, the staff finds the use of

1 a fracture-toughness-based reference temperature, RTT0, acceptable and that an exemption to 2 Appendix G to 10 CFR Part 50 by the licensees is not required.

3 4 Master Curve Data Search 5

6 In Section 3.1.1, Master Curve Data Search, of the TR, the PWROG described the approach it 7 used for searching and gathering master curve data relevant to RPV nozzle forgings in U.S.

8 PWRs. The PWROG explained that relevant data was gathered from open literature, the 9 Electric Power Research Institute (EPRI) fracture toughness database, and internal 10 Westinghouse references. Specifically, the PWROG considered thick sections of A-508 Class 2 11 or similar forgings that were used in RPV fabrication or are representative of the materials used 12 to construct U.S. PWR inlet and outlet nozzles. The purpose was to capture all available 13 transition temperature fracture toughness data to establish a generic master curve transition 14 reference temperature for A-508 Class 2 type forgings. The PWROG explained in its 15 supplement that representative means that the forging heats from which master curve data 16 were obtained had material specifications similar to A-508 Class 2 forgings used in U.S. PWR 17 inlet and outlet nozzles. Specifically, the staff noted that the PWROGs selection included 18 alternate forging alloys22NiMoCr37, 20NiMoCr26, and SFVQ2A (the staffs review regarding 19 the applicability of these alternate forging alloys to U.S. PWR inlet and outlet nozzles is 20 discussed below). The PWROG also stated that the meaning of bounding is explained in 21 Section 3.1.2.2 of the TR. The staff noted that the master curve data is considered bounding 22 because it included irradiated materials, fracture toughness data based on KIC, one material with 23 RTNDT greater than 60°F, and a diversity of relevant forgings (as evidenced by the large 24 standard deviation presented in Section 3.1.2.2 of the TR), all of which conservatively impact 25 fracture toughness. Based on its review, the staff finds the scope of materials that the PWROG 26 considered and included into the master curve data is representative and reasonably bounds 27 the fracture toughness of RPV inlet and outlet nozzle forgings in U.S. PWRs.

28 29 The PWROG explained that the nozzle forgings used in U.S. PWRs are all ASME SA-508 30 Class 2 or ASTM A-508 Class 2 with the following exceptions: Prairie Island Nuclear 31 Generating Station Units 1 and 2 nozzles, which are SA-508 Class 3, Palo Verde Nuclear 32 Generating Station Units 2 and 3 nozzles (which are a combination of SA-508 Classes 2 and 3),

33 and R.E. Ginna Nuclear Power Plant nozzles (which are SA-336). The PWROG noted that the 34 Ginna nozzles meet the A-508 Class 2 specification requirements per the Ginna Certified 35 Material Test Reports (CMTRs). With regard to the nozzles forgings that are SA-508 Class 3, 36 the PWROG stated that master curve data was assessed for A-508 Class 3 and showed that 37 the fracture toughness properties were better than A-508 Class 2. Based on its review of the 38 master curve data for A-508 Class 3 materials referenced by the PWROG, the staff finds it 39 reasonable that the A-508 Class 2 generic RTT0 developed in this TR is conservative compared 40 to A-508 Class 3 forgings. In addition, the staff finds that the A-508 Class 2 generic RTT0 41 developed in the TR is appropriate for the SA-336 forgings because plant-specific CMTRs 42 demonstrate that these forgings meet the A-508 Class 2 specification requirements.

43 44 The PWROG provided a description of the materials relevant to the U.S. PWR nozzle forgings 45 that were included in its master curve data search. Specifically, the PWROG included in its 46 supplement available master curve data, chemical composition, and mechanical properties of 47 the following materials: 22NiMoCr37, ASTM A-508-64 Class 2, SA-508 Class 2 (1971), SA-508 48 Grade 2 Class 1 (2007), 20NiMoCr26, and SFVQ2A. The staff reviewed the chemical 49 composition and mechanical properties listed for the different forgings and noted that the 50 differences in the chemical composition limits and mechanical properties between all the 51 different alloys are very minor when compared to the alloys used in U.S. PWR nozzle forgings.

52 The PWROG confirmed that each of these forgings in the master curve dataset was quenched

1 and tempered steel for pressure vessels, and that a similar heat treatment was used to produce 2 the required properties. The PWROG also confirmed that the master curve data was produced 3 from specimens taken from thick section forgings except for the 20NiMoCr26 forging, which was 4 thinner. For this particular forging that was thinner, the PWROG indicated that consideration of 5 the forging in the dataset is conservative (i.e., increases the average generic RTT0 in the TR).

6 Based on the impact of the 20NiMoCr26 forging to the average generic RTT0 determined in the 7 TR, the staff find its inclusion into the master curve dataset to be conservative.

8 9 Based on its review, the staff considers A-508 and SA-508, Class 2, 22NiMoCr37, 20NiMoCr26, 10 and SFVQ2A forgings are essentially the same alloy because of the minor differences in the 11 chemical composition and mechanical properties, and the PWROGs confirmation regarding the 12 methods used to produce these forgings. Thus, the staff finds the PWROGs inclusion of A-508 13 and SA-508, Class 2, 22NiMoCr37, 20NiMoCr26, and SFVQ2A materials in its master curve 14 dataset to be acceptable and representative of U.S. PWR nozzle forgings.

15 16 Based on its review, the staff finds the scope of materials considered by the PWROG and 17 included into the master curve data is representative and reasonably bounds the fracture 18 toughness of the RPV inlet and outlet nozzle forgings in U.S. PWRs.

19 20 Results from Master Curve Data Search 21 22 Section 3.1.2, Results from Master Curve Data Search, of the TR states that master curve 23 data for 22 distinct forgings were identified, and in all cases the heats selected are 24 representative of the forgings used in commercial PWRs and boiling water reactors (BWRs) 25 from Japanese, Swedish, German, and U.S. RPVs. The PWROG confirmed that the references 26 were checked to ensure that all the data collected for the TR was from unique forgings.

27 28 The PWROG explained that in some cases the references only reported KIC values; 29 nevertheless, the master curve reference temperature can conservatively be developed from 30 these KIC values. The PWROG stated that the KIC values are always the same or lower than the 31 cleavage-onset fracture toughness (KJC) values from the same test; thus, the T0 value 32 developed from these KIC values would be conservative. The staff noted that where KIC is used 33 instead of KJC, KIC is defined by ASTM E399, Standard Test Method for Linear-Elastic Plane-34 Strain Fracture Toughness KIC of Metallic Materials, and is the applied SIF (K) where the load 35 displacement trace deviates from linearity by 5 percent. Whereas in ASTM E1921, Standard 36 Test Method for Determination of Reference Temperature, T0, for Ferritic Steels in the Transition 37 Region, KJC is K converted from the applied J-integral at cleavage. The staff noted that the KIC 38 curve was established using only data deemed to be valid by linear elastic fracture mechanics 39 criteria per ASTM E399; thus, only the lower range of cleavage fracture toughness values were 40 used, whereas KJC is determined from data from specimens in a temperature range where either 41 cleavage cracking or crack pop-in develops during the loading of specimens and is not limited to 42 the lower range values. Thus, the staff finds it acceptable and conservative that the PWROG 43 included relevant KIC values in determining the master curve reference temperature because 44 these values only include the lower range of cleavage fracture toughness data.

45 46 Table 3-2 All Available Master Curve Data on A-508 Class 2 Type Forgings, of the TR 47 presents the results from the master curve data search performed by the PWROG. The 48 PWROG stated that the range of RTNDT values in Table 3-2 of the TR exceeds the range (i.e.,

49 more conservative) of the RTNDT values generally observed in U.S. PWR nozzle forgings utilizing 50 the criteria in NB-2300 of Section III of ASME Code, which typically fall between -34°C 51 and -12°C. Additionally, the PWROG stated that the average RTNDT (-11°C) of the 22 forgings in 52 Table 3-2 of the TR falls above (i.e., more conservative) this typical range of RTNDT values

1 observed for U.S. PWR nozzle forgings based on measured data and ASME Code NB-2300 2 criteria. The PWROG summarized in its supplement NB-2300-compliant measured RTNDT 3 values for U.S. PWR nozzle forgings developed from a review of original CMTRs. The staff 4 noted that this information is not intended to be a complete list of all U.S. PWR nozzle RTNDT 5 values but contains those readily available to the PWROG, which are representative of 6 approximately half of the U.S. PWR nozzle forgings.

7 8 The staff reviewed these NB-2300-compliant RTNDT values and noted an average value 9 of -10.5°F (-23.6°C). The staff noted the average value from the reported RTNDT values in the 10 master curve data search (i.e., Table 3-2 in the TR) is 12.2°F (-11°C). Based on the readily 11 available data from U.S. PWR nozzle forgings and the master data search in the TR, the staff 12 finds the forgings included in the master curve data used to develop the A-508 Class 2 generic 13 RTT0 in the TR, on average, is not as tough as the nozzle material in U.S. PWRs and therefore 14 conservatively represents the fracture toughness of U.S. PWR nozzle forgings.

15 16 Based on the discussion above, the staff finds the PWROG demonstrated that the master curve 17 data presented in the TR is conservatively representative with respect to fracture toughness of 18 U.S. PWR nozzle forgings. Specifically, since RTT0 is an acceptable alternate to RTNDT, the staff 19 finds the A-508 Class 2 generic RTT0 developed in this TR is also considered conservatively 20 representative of the U.S. PWR fleet of nozzle forgings.

21 22 Specimen Geometry Constraint Adjustment 23 24 Table 3-2 of the TR provides the details of the specimen geometry of the forgings that were 25 used to determine generic nozzle forging master curve reference temperature. Section 3.1.2.1, 26 Specimen Geometry Constraint Adjustment, of the TR indicates that, as observed by 27 Tregoning and Joyce (Ref. 45 of the TR), there is a systematic, non-conservative bias toward 28 the Single Edge Notched Bend (SE(B)) specimen of generally 5°C to 10°C relative to the 29 compact tension (CT) specimen geometry due to its lower constraint. Thus, the PWROG 30 elected to address this by adding a 10°C bias to the SE(B) T0 values to adjust for the lower 31 constraint SE(B) geometry, as shown in Table 3-2 of the TR.

32 33 By letter dated August 4, 2005 (ADAMS Accession No. ML052070408), the staff approved the 34 use of a 10°C bias for the lower constraint SE(B) geometry in its SE of BAW-2308, Revision 1.

35 In addition, the staff noted that recent editions of ASTM E1921 included an average difference 36 between the CT and SE(B) of 10°C.

37 38 The staff finds the PWROGs use of a 10°C bias to the SE(B) T0 values acceptable because it is 39 consistent with (1) the data and information available on the differences between SE(B) 40 specimen and CT specimen test result, and (2) the previous approval of a 10°C bias for the 41 lower constraint SE(B) geometry.

42 43 Surface Effect 44 45 Section 3.2, Surface Effect, of the TR describes improved toughness near the surface of a 46 forging material compared to a location deeper in the forging. The PWROG cited references 47 that illustrated the improved toughness near the surface and presented transition temperature 48 data for 24 longitudinal (LT) specimens and seven transverse (TL) specimens. The data 49 consisted of shifts in transition temperature at the surface relative to the 1/4T location and were 50 determined from Charpy V-Notch (CVN) or the master curve measurements. The PWROG 51 stated that specimens without a reported orientation were included in the LT data set.

52 Table 3-3, Summary of Transition Temperature Shifts for LT and TL Specimens, of the TR

1 showed the average and standard deviation of the transition temperature shifts for the LT and 2 TL data sets. The PWROG selected the conservative set of average and standard deviation 3 (i.e., the LT data set) to take credit for improved fracture toughness for the small flaw models 4 described in Section 4.1 of this SE.

5 6 The staff reviewed the information in Section 3.2 of the TR and verified the average and 7 standard deviation of the transition temperature shifts in Table 3-3 of the TR. The staff noted 8 these observations in the LT measurements: five of the measurements were taken at less than 9 the assumed flaw size of 0.5 inch and two measurements had only a small difference in the 10 depths that they were taken. The staff recalculated the average and standard deviation without 11 these LT measurements and determined that they caused a negligible change. Therefore, the 12 staff finds the average and standard deviation of the temperature shifts shown in Table 3-3 of 13 the TR to be acceptable. The staff noted that the inherent scatter in CVN measurements tend 14 to increase the standard deviation in the transition temperature shifts, which is conservative; 15 thus, the staff also finds including CVN measurements to be acceptable.

16 17 The PWROG addressed in its supplement the specimens without a reported orientation being 18 included in the LT data set in two aspects. First, the PWROG confirmed that the ten B&W 19 forgings, the Westinghouse Four-Loop Inlet Nozzle, Westinghouse Four-Loop Outlet Nozzle #1, 20 and Westinghouse Four-Loop Outlet Nozzle #2 have CMTRs dated from 1969 and 1970. The 21 PWROG stated that testing of TL specimens was not required until after the issuance of the 22 Summer 1972 Addenda of the 1971 Edition of the ASME Code Section III. Although the staff 23 does not find it reasonable that these forgings produced prior to 1972 were tested in the LT 24 direction, the second aspect of how the PWROG addressed specimens with unknown 25 orientation is reasonable. The PWROG stated that in addition to the forgings discussed above, 26 the orientation was not reported for the BethForge forging ID, BethForge forging OD, Forging 27 M1, Forging I, and the French forging. For all of the forgings with unknown orientation identified 28 above, the PWROG illustrated the breakdown of the LT dataset measured transition 29 temperature surface shift between those with a reported LT orientation and those with an 30 assumed LT orientation. The PWROG explained that the addition of the assumed LT 31 orientation data biases the average shift value in the conservative direction compared to the 32 dataset with only known LT orientation. Specifically, the staff noted that the known LT dataset 33 provides an average shift of 44.7°F; whereas, the unknown LT dataset in this second category 34 would only provide an average shift of 33.8°F. Thus, when the known and unknown LT 35 datasets are both included, the average shift and standard deviation values in the TR (36.5°F 36 and 28.9°F, respectively) result in a more conservative ART compared to the ART value based 37 only on known LT data.

38 39 In summary, the staff finds that the PWROG adequately addressed the forgings without a 40 reported orientation and their inclusion with the known LT data is appropriate and conservative, 41 as described above. Thus, the staff finds that the PWROG has selected a conservative dataset 42 set to determine the improved fracture toughness near the surface of a forging material and 43 finds it acceptable when addressing the small flaw models described in the TR.

44 45 Underclad Heat-Affected Zone Toughness 46 47 Section 3.3, Underclad HAZ Toughness, of the TR states that a significant portion of the small 48 postulated flaw in this TR would be in the underclad heat-affected zone (HAZ); therefore, the 49 properties of the HAZ relative to the adjoining base metal must be considered. The staffs 50 evaluation of the small postulated flaw is documented in Section 2 of this SE.

51

1 The PWROG provided information from Oak Ridge National Laboratory (ORNL), in which ORNL 2 conducted Charpy impact testing on a stainless steel cladded plate to determine the effect of 3 clad on the propagation of small surface flaws. This plate was specifically heat treated to 4 produce a high transition temperature but was not quenched and only slightly tempered. The 5 testing performed by ORNL showed that the clad HAZ had significantly better properties 6 (i.e., lower transition temperature) than the 1/4T location in the plate. The PWROG stated that 7 since the plate was not quenched, the improved HAZ transition temperature would not be due to 8 a faster cooling rate from quenching, but the tempering of the cladding operation.

9 10 The staff noted that HAZ test results from surveillance specimens have revealed the 11 inhomogeneous nature of the HAZ material, which also resulted in significant scatter of the HAZ 12 Charpy test data. As discussed in Irradiation Embrittlement of Reactor Pressure Vessels 13 (RPVs) in Nuclear Power Plants (Soneda, N. ed., 2015), the weld HAZ has been shown to 14 exhibit superior fracture toughness compared to the plate or forging. In addition, the staff also 15 noted that the continued need to include HAZ material in RPV material surveillance programs 16 was more recently investigated in a paper by Koichi Masaki, Jinya Katsuyama, and Kunio 17 Onizawa, Study on the Structural Integrity of RPV Using PFM (Probabilistic Fracture 18 Mechanics) Analysis Concerning Inhomogeneity of the Heat-Affected Zone. This paper 19 investigated the features of HAZ inhomogeneity in RPV steels to determine the need for 20 surveillance test specimens of HAZ materials in Japan. The authors examined the 21 inhomogeneous distribution of fracture toughness for HAZ materials using a PFM code and 22 determined that the high-toughness coarse grain HAZ caused arrest of postulated cracks. This 23 outcome is expected metallurgically, because the HAZ is a tempered version of the plate or 24 forging and, as such, it should exhibit superior fracture toughness compared to the plate or 25 forging.

26 27 Thus, the staff finds that the PWROG has adequately addressed the properties of the HAZ 28 relative to the adjoining base metal and finds the PWROGs conclusion that underclad HAZ in 29 nozzles is as tough, or tougher than, the adjacent forging base metal to be acceptable.

30 31 4.3 Neutron Embrittlement 32 33 Section 3.4, Neutron Embrittlement, of the TR states that the copper (Cu) content was not 34 measured for all the nozzles manufactured for the U.S. PWR fleet, however it was measured for 35 a substantial number covering nearly the full range of manufacturing dates and all major U.S.

36 RPV fabricators. Cu measurements were averaged for 178 inlet and outlet nozzles yielding an 37 average of 0.0947 percent with a standard deviation of 0.0319 percent, yielding a best-estimate 38 value, as defined by RG 1.99, Revision 2, of average plus one standard deviation of 39 0.127 percent. The PWROG explained that for nickel (Ni) content, the upper limit of the SA-508 40 Class 2 specification during the fabrication time period is used, which was 0.90 percent. The 41 PWROG stated that the Cu and Ni contents discussed above are appropriate for the U.S. PWR 42 nozzles, since the database was established from Cu measurements from PWR nozzle forgings 43 only.

44 45 The staff reviewed the information in Section 3.4 of the TR and RG 1.99, Revision 2, regarding 46 the Cu and Ni content. The staff finds the PWROG appropriately determined Cu and Ni 47 contents that are representative of U.S. PWR nozzle forgings consistent with the guidance in 48 RG 1.99, Revision 2.

49 50 Using the Cu and Ni contents discussed above and RG 1.99, Revision 2, the PWROG 51 developed an embrittlement trend curve (ETC) that shows the shift in RTNDT (RTNDT) as a 52 function of neutron fluence, applicable to U.S. PWR nozzles. The PWROG then determined the

1 fluence value of 4.28 x 1017 n/cm2 for a RTNDT of 25°F. The PWROG cited NRC Technical 2 Letter Report TLR-RES/DE/CIB-2013-01, Evaluation of the Beltline Region for Nuclear Reactor 3 Pressure Vessels, Office of Nuclear Regulatory Research (RES), dated November 14, 2014 4 (ADAMS Accession No. ML14318A177), as a basis for not considering the shift due to 5 irradiation of RPV beltline materials (including nozzles) if RTNDT is less than 25°F. The 6 PWROG used the fluence value at RTNDT of 25°F as a screening threshold below which 7 embrittlement due to irradiation may be neglected in the calculation of ART (as discussed in 8 Section 4.4 of this SE). Section 3.4.5, Future Increased Nozzle Fluence Projections, of the TR 9 indicates that as long as the nozzle fluence projections are less than the fluence screening 10 threshold, the nozzle P-T limits developed in the TR is applicable (if the new fluence is greater 11 than the threshold, a plant-specific RTNDT or ART shall be calculated).

12 13 The PWROG provided additional justification in its supplement that supports the 14 recommendation in TLR-RES/DE/CIB-2013-01 related to RTNDT of 25°F. The PWROG stated 15 that predictions of RTNDT have inherent scatter due to uncertainty in RTNDT data 16 measurement and uncertainty in RTNDT prediction models. The PWROGs premise is that a 17 RTNDT of 25 °F does not have to be considered because 25°F is a reasonable value that 18 represents the scatter in RTNDT due to these uncertainties. To demonstrate this, the PWROG 19 compared the standard deviation (a measure of scatter) of the RTNDT data in TLR-20 RES/DE/CIB-2013-01 and from the embrittlement database used to develop the ASTM E900 21 ETC, which included data from welds, plates, and forgings from tested surveillance capsules.

22 The PWROG determined a standard deviation of RTNDT of 23 °F from the data in 23 TLR-RES/DE/CIB-2013-01 and a standard deviation of RTNDT of 18.6 °F from the ASTM E900 24 ETC data. The standard deviation from the ASTM E900 ETC data included fluence levels up to 25 4.28 x 1017 n/cm2, which is the fluence corresponding to a RTNDT of 25 °F and the fluence 26 threshold the PWROG is proposing in the TR below which embrittlement shifts for nozzles do 27 not have to be considered. The PWROG noted that 18.6 °F is slightly less than 23°F but is 28 consistent with the standard deviation of RTNDT from other ETCs, which included ETCs based 29 on RG 1.99, Revision 2 and 10 CFR 50.61a.

30 31 To further demonstrate that 25°F is a reasonable value below which embrittlement shifts do not 32 have to be considered, the PWROG, using the ETC from RG 1.99, Revision 2, determined 33 RTNDT values of 24.5°F, 25.4°F, and 29.6°Fall comparable to 25°F-for RPV materials that 34 have hypothetically high Cu content (i.e., highly embrittled) at a fluence level of 0.99 x 1017 35 n/cm2. This fluence level is slightly less than the 1 x 1017 n/cm2 threshold established in 36 Appendix H to 10 CFR Part 50 for monitoring changes in the fracture toughness properties of 37 ferritic materials in the reactor vessel beltline region. Since 0.99 x 1017 n/cm2 is less than 38 1 x 1017 n/cm2, these RTNDT values for RPV materials having hypothetically high Cu content 39 would not have been considered. Based on the discussion above, the PWROG concluded that 40 25°F is a reasonable value below which embrittlement shifts do not have to be considered.

41 The staff reviewed the PWROGs justification for using the recommendation in 42 TLR-RES/DE/CIB-2013-01 for not having to consider a RTNDT of 25°F. The staff noted that 43 the RTNDT data from the ASTM E900 embrittlement trend curve (Figure 1 in the supplement) 44 have more positive shifts than negative shifts and shifts could be up to 60°F. However, the 45 staff recognizes that the effect of embrittlement is difficult to distinguish from the data scatter for 46 shifts less than 25°F. Therefore, given the safety significance of RPV components, the staff 47 does not find the justification sufficient to demonstrate generically that embrittlement shifts less 48 than 25°F do not have to be considered. In order to determine whether the recommendation in 49 TLR-RES/DE/CIB-2013-01 of excluding 25°F embrittlement is acceptable specifically for this 50 TR, the staff evaluated the safety significance of the recommendation by identifying if there are 51 any U.S. PWRs in which the nozzles are the limiting material for P-T limits when accounting for 52 an embrittlement shift of 25°F.

1 For this independent assessment to be focused on those U.S. PWRs in which the nozzle 2 material is more limiting than the traditional beltline for P-T limits, the following criteria were 3 used to screen out U.S. PWRs as not needing any additional review:

4 5

  • Plants that are already shutdown or not pursuing a renewed operating license 6
  • Plant-specific license amendment requests have been reviewed and approved by the 7 NRC to address irradiation embrittlement of the nozzles 8
  • Plant-specific Pressure-Temperature Limits Report (PTLR) demonstrates that the NRC-9 approved P-T limit curves are limiting 10
  • Neutron fluence at the nozzle region is less than 1 x 1017 n/cm2 (E > 1 MeV) at the end 11 of 60-years of plant operation (neutron fluence information is publicly available in plant-12 specific license renewal applications, license amendment requests, or PTLR) 13
  • Reactors with traditional beltline materials with Cu 0.2 wt. % (information is publicly 14 available in Reactor Vessel Integrity Database (RVID) Version 2.0.1) 15 16 The staff determined that reactors with a neutron fluence at the nozzle region less than 1 x 1017 17 n/cm2 (E > 1 MeV) at the end of 60 years of plant operation are screened out consistent with the 18 threshold established in Appendix H to 10 CFR Part 50 for monitoring changes in the fracture 19 toughness properties of ferritic materials in the RPV beltline region. Furthermore, the staff 20 determined that it is reasonable that U.S. PWRs with traditional beltline materials with Cu 0.2 21 wt. % are screened out because this level of Cu content would cause a significant shift due to 22 embrittlement in the P-T limits such that it will continue being the limiting material through the 23 license renewal period (i.e., 40 to 60 years of operation).

24 25 Following this initial screening, the staff reviewed the information available in RVID 2.0.1 to 26 identify candidate U.S. PWRs based on the following criteria:

27 28

  • Reactors with a traditional beltline material with low Cu content (i.e., 0.03 wt. %)

29

  • Reactors with NRC-approved P-T limits based on a limiting material with low Cu content 30
  • Nozzle material information (e.g., initial RTNDT, Cu, Ni, and neutron fluence) is available 31 in ADAMS to generate P-T limit curves 32 33 The staff noted that reactors meeting these criteria, particularly those reactors with good beltline 34 material properties (i.e., low initial RTNDT), have the highest likelihood that a shift due to 35 embrittlement of the nozzle could lead to nozzle P-T limits being more limiting than the 36 NRC-approved P-T limits based on a traditional beltline material. Since a data search was 37 being performed for nozzle material property information for the candidate reactors, the staff 38 opted to also include any additional reactors at the site since the information was already 39 available in the source documents (e.g., license renewal application). This resulted in a total of 40 nine U.S. PWRs that the staff further investigated by generating P-T limit curves for the limiting 41 nozzle forging using ART values based on an effective full power year (EFPY) that was 42 available from the appropriate source document or data. These nozzle P-T limit curves are 43 based on a 100°F per hour cooldown rate and a postulated inside corner flaw of depth 1/4T.

44 45 For the independent assessment, the staff determined applied SIFs for nozzles due to pressure 46 loading (KIP) and thermal gradients (KIT) consistent with those published in the ORNL study, 47 ORNL/TM-2010/246, Stress and Fracture Mechanics Analyses of Boiling Water Reactor and 48 Pressurized Water Reactor Pressure Vessel Nozzles -Revision 1, June 2012. The staff noted 49 that these SIF solutions are also consistent with those in the 2013 edition of the ASME Code, 50 Section XI, Paragraph G-2223(c), which are applicable to postulated nozzle corner flaws, 51 regardless of plant design. The staff used the limiting nozzle location from ORNL/TM-2010/246

1 (i.e., the nozzle location with the highest stresses) in its independent assessment. As such, the 2 staff finds that the use of the SIF solutions in ORNL/TM-2010/246 for calculating the KIP and KIT 3 values for the nozzles are acceptable and appropriate for use in its independent assessment.

4 5 The nozzle P-T limit curves generated by the staff for these candidate U.S. PWRs were then 6 compared to their respective NRC-approved P-T limit curves, both of which were based on ART 7 values calculated at the same EFPY. Based on this comparison, the staff determined that for 8 these nine candidate reactors, the limiting traditional beltline NRC-approved P-T limit curves 9 were bounding compared to the nozzle P-T limit curves generated by the staff.

10 11 For the remaining U.S. PWRs, the staff noted that nozzle material information (e.g., initial 12 RTNDT, Cu, Ni, and neutron fluence) was not readily available. Thus, for the staff to determine if 13 the nozzle P-T limit curve is limiting, a generic screening ART value for the nozzle was 14 calculated and then compared against the ART values from the traditional beltline. The staff 15 noted that if the ART value for the traditional beltline materials (information available in RVID 16 2.0.1) is less than this screening generic nozzle ART value, there is a potential that the nozzle 17 P-T limit curve may be more limiting. Since the plant-specific nozzle information was not 18 available, the staff used the generic mean alternate RTNDT value determined in the TR for U.S.

19 PWR nozzle forgings. As discussed in Sections 4.2 and 4.4 of this document, the staff 20 determined the generic mean alternate RTNDT value (i.e., RTT0) in the TR is relevant and 21 conservatively representative of U.S. PWR nozzle forgings.

22 23 The generic screening nozzle ART value was determined in the following manner:

24 25

  • marginEmbrittle = 2(i2+2) 1/2 - Per RG 1.99, Revision 2 27
  • RTT0 Inital = -66.4°F - Per Section 3.5 of the TR 28
  • RTNDT Stress = 25°F - Bounding shift due to stress based on review of P-T limit curves 29
  • RTNDT Embrittle = 25°F - Maximum shift due to embrittlement 30
  • i = 54.5°F - Per Section 3.5 of the TR 31 * = 12.5°F - Per RG 1.99, Revision 2, cannot be more than 1/2 of RTNDT Embrittle 32
  • Generic Nozzle ARTscreening = 95.4°F 33 34 As noted above, RTNDT Stress represents the shift of the nozzle P-T limit curve resulting from the 35 stress levels due to the structural discontinuities in the nozzle region as compared to the P-T 36 limits curve generated for the traditional beltline. Based on its observations and previous 37 reviews of License Amendment Requests for P-T limits curves, the staff noted that a value of 38 25°F is appropriate and bounding to account for the increased stress levels due to the structural 39 discontinuity in the nozzle. Based on this screening generic nozzle ART value, the staff 40 identified four U.S. PWRs that needed a detailed assessment. The staff determined that two of 41 these U.S. PWRs are governed by the P-T limit curve from the bounding unit at the site, which 42 was previously screened out because the ART value for a traditional beltline material was 43 greater than the screening generic nozzle ART value. For the remaining two PWRs, the staff 44 generated P-T limit curves for a generic nozzle ART value, consistent with the methods 45 described above, for comparison with the traditional beltline NRC-approved P-T limit curves.

46 However, to generate these nozzle P-T limit curves, RTNDT Stress = 25°F was not included in the 47 ART value because RTNDT Stress was only for the purpose of screening in PWRs for 48 assessment. The resulting generic nozzle ART value used for generating the P-T limit curves is 49 70.4°F.

1 The staff generated nozzle P-T limit curves for the two remaining PWRs using the generic ART 2 value of 70.4 °F and compared them to the NRC-approved P-T limit curves. Based on this 3 comparison for the two U.S. PWRs, the staff noted the following:

4 5

  • The NRC-approved P-T limit curve was limiting for one reactor 6
  • The NRC-approved P-T limit curve coincided with the P-T limit curve generated with the 7 generic nozzle ART value of 70.4 °F for the other reactor 8

9 For the case in which the two curves coincided, the staff noted that the NRC-approved P-T limit 10 curve was based on 36 EFPY (i.e., 40 years of plant operation); whereas, the generic nozzle 11 ART value is based on a neutron fluence in the nozzle region that is conservatively expected 12 after 60-years of plant operation. The staff noted that if the comparison of the NRC-approved P-13 T limit curve and the P-T limit curve generated with the generic nozzle ART value for the subject 14 reactor was at the same EFPY, the nozzle material would not be limiting. In addition, as 15 discussed in Section 4.2 of this document, the generic nozzle ART value, which is based on RTT0 16 and i developed in the TR, is conservatively representative of the U.S. PWR nozzle forgings.

17 The staff noted that the generic screening nozzle ART value included the shift of 25°F due to 18 embrittlement and that the nozzle-specific shift can be less than this value. Thus, the staff noted 19 that if the plant-specific nozzle material properties for the subject reactor are used, it is 20 reasonable to expect that the nozzle would be tougher than the generic nozzle addressed in 21 this TR and would make the NRC-approved P-T limit curve more limiting than the nozzle P-T 22 limit curve.

23 24 In summary, based on its assessment of the PWROGs justification and the staffs independent 25 assessment, as described above, the staff finds that for a neutron fluence less than 4.28 x 1017 26 n/cm2 (E > 1 MeV) in the nozzle region, the NRC-approved P-T limit curves are limiting for 27 60 years of plant-operation when compared to the nozzle P-T limit curves. In addition, even 28 though the PWROG did not consider the shift due to irradiation of the nozzles if RTNDT is less 29 than 25°F, the staff demonstrated in its independent assessment that this assumption is unlikely 30 to cause the P-T limit curves for inlet or outlet nozzle corners of U.S. PWRs to be more limiting 31 than those of the shell (and associated welds) of the traditional beltline region of the RPV for a 32 neutron fluence less than 4.28 x 1017 n/cm2 (E > 1 MeV) in the nozzle region.

33 34 Fluence Location Relative to the Postulated Flaw Location and Fluence Methodology 35 36 Section 3.4.1, Calculated Fluence Location Relative to the Postulated Flaw Location, of the TR 37 states that nozzle fluence values are typically assumed to be equal to the RPV upper-shell-to-38 nozzle forging weld fluence value or the lowest extent of the nozzle forging, and thus the nozzle 39 fluence values are conservative. The PWROG stated that since the postulated flaws in the TR 40 are at the nozzle corners, which are at a higher elevation and therefore further away from the 41 active core, the fluence value is expected to be significantly lower than the fluence at the lowest 42 extent of the nozzle forging or weld. The staff reviewed the discussion of fluence location 43 relative to the postulated flaw location in Section 3.4.1 of the TR and finds it acceptable.

44 45 Section 3.4.2, Fluence Calculational Methodology, of the TR states the use of new fluence 46 evaluation methods can more accurately determine the nozzle fluence reducing the needed 47 conservatisms. The PWROG showed a comparison of nozzle fluence values between three 48 methods of fluence evaluations. The staff reviewed the information in Section 3.4.2 of the TR 49 and noted that the fluence methods approved by the NRC staff are unique to the individual 50 licensees current licensing basis. Thus, plant-specific fluence calculations performed by the 51 individual licensee in a manner consistent with the NRC-approved methodology will be 52 necessary to determine whether the use of the TR is applicable.

1 Neutron Streaming 2

3 Section 3.4.3, Neutron Streaming, of the TR states that neutron streaming up the cavity to the 4 nozzle region from the beltline region is an existing phenomenon. As such, the traditional 5 fluence attenuation equation used in the beltline (i.e., in RG 1.99, Revision 2) is not appropriate 6 in the nozzle region when only considering fluence calculated at the inside surface. The 7 PWROG indicated that the fluence at the outside diameter lowest extent of the nozzles can be 8 higher than the fluence at the lowest extent of the nozzle forging at the RPV inside surface due 9 to cavity neutron streaming. The PWROG investigated the stresses at the inlet and outlet 10 nozzles due to pressure and the thermal cooldown transient. The stresses are shown in 11 Section 4.8 Stresses at Limiting Locations of the TR, specifically in the figures from the 3D 12 finite element analysis. The PWROG states that these figures demonstrate that the stresses at 13 the lowest extent outside diameter of the nozzles are significantly lower than at the nozzle inside 14 corner, and when pressure stress and thermal stress are considered together, the combined 15 stress is likely compressive. As is discussed in Section 4.8 of the TR, the flaw is postulated at 16 the nozzle inside surface corner at a geometric discontinuity where the highest stresses exist.

17 As a result, the nozzle inside corner is the limiting location, and this location is where the 18 fluence is considered for embrittlement.

19 20 Based on its review, the staff finds the neutron streaming effect is applicable to the 3/4T 21 postulated flaw and that the PWROGs exclusion of the 3/4T postulated flaw in the development 22 of the P-T limits in the TR for the nozzles is appropriate, as described below. Specifically, the 23 staff noted that the pressure stress decreases as a function of distance from the inside corner 24 along the through-wall nozzle corner path, as shown in Figure 24 of ORNL/TM-2010/246.

25 Therefore, the applied SIF due to pressure for a 3/4T postulated flaw at the outside corner of the 26 nozzle would be lower than that for the 1/4T flaw postulated for the inside corner region. The 27 linear elastic fracture mechanics analyses in ORNL/TM-2010/246 do not address 3/4T 28 postulated flaws for this reason. It should be noted that, based on the analysis of the 1/4T 29 location and the smaller postulated flaw from the inside corner region, the nozzle P-T limits for a 30 heatup transient would be less restrictive than those calculated for a cooldown transient 31 because the thermal stresses for a postulated inside corner flaw are compressive for heatup.

32 Therefore, the staff determined that analyses of the 1/4T location and the smaller postulated 33 flaw at the nozzle inside corner during a cooldown transient generates the most bounding P-T 34 limits for the nozzles.

35 36 4.4 Calculation of ART 37 38 In Section 3.1.2.2, Calculation of Generic Mean Alternate RTNDT, and Section 3.5, Adjusted 39 Reference Temperature, of the TR, the PWROG calculated ART with and without the surface 40 effect using the RTT0 (evaluated in Section 4.2 of this SE) as the initial reference temperature.

41 The ART value with the surface effect is to be used for the postulated small flaw and the ART 42 value without the surface effect is to be used for the traditional, postulated 1/4T flaw in the 43 nozzle P-T limits developed in Section 5.1, Generation of Nozzle P-T Limit Curves, of the TR 44 (evaluated in Section 4.10 of this SE). Both ART calculations do not consider an embrittlement 45 shift of 25°F (which the staff evaluated in Section 4.3 of this SE) since the PWROG developed a 46 fluence threshold screening criterion of 4.28 x 1017 n/cm2. The staff noted that this fluence 47 threshold screening criterion corresponds to a RTNDT of 25°F below which embrittlement shifts 48 may be neglected.

49 50 The staff verified the ART calculations in Sections 3.1.2.2 and 3.5 of the TR consistent with the 51 guidance in RG 1.99, Revision 2, and finds the ART values of 43°F for the 1/4T flaw and 21°F 52 for the shallow flaw are acceptable for the A-508 Class 2 generic RTT0 developed in this TR.

53

1 4.5 Selection of Inlet and Outlet Nozzle Model Geometry 2

3 In Section 4.1 of the TR, the PWROG considered several geometric parameters that affect the 4 stress due to pressure and thermal transient in the nozzle corner region. The PWROG stated 5 that important characteristics that affect nozzle corner stress and SIF were assessed to ensure 6 representative or bounding models were chosen for the whole U.S. PWR fleet. The PWROG 7 considered nozzle radius-to-thickness (R/t) ratio, nozzle diameter, nozzle corner geometry, and 8 clad thickness as the important geometric parameters that affect the nozzle corner stress and 9 SIF. Table 4-1, Model Geometry Comparison, pictorially depicted in Figure 4-3, Diversity of 10 Nozzle Geometries Modeled, of the TR summarizes the inlet and outlet geometries that were 11 modeled.

12 13 The staff finds the PWROGs approach for selecting nozzle model geometries acceptable since 14 it is not practical to model the unique geometry of each inlet and outlet nozzle design in the U.S.

15 PWR fleet. It is reasonable to consider only the parameters that are most relevant, with respect 16 to the stress that can extend a postulated flaw in the nozzle corner. The staff considers that the 17 thickness of the nozzle section and the sharpness of the nozzle corner radius are the most 18 relevant parameters that can extend a postulated flaw in the nozzle corner. The staff finds that 19 the PWROG adequately addressed the effects of these two parameters by considering the 20 nozzle R/t ratio, nozzle diameter, nozzle corner geometry, and clad thickness. Considering the 21 nozzle R/t ratio (and nozzle diameter which accounts for the radius) addresses the section 22 thickness effect on stress. Considering the nozzle corner geometry addresses the effect of the 23 nozzle corner radius, which causes high stresses on the inside surface of the corner.

24 Considering the clad thickness addresses the effects of clad welding residual stress on the 25 small flaw models described in Section 2.

26 27 Furthermore, the staff determined that selecting a nozzle section thickness that bounds all U.S.

28 PWR fleet inlet and outlet nozzle is challenging for two reasons: (1) the effect of thickness on 29 stress due to internal pressure counterbalances the effect of thickness on stress due to thermal 30 transients: a thinner section would generate a higher stress due to internal pressure, but a 31 lower stress due to thermal transient; and (2) the time at which the maximum stress due to 32 internal pressure occurs does not occur at the same time the maximum stress due to thermal 33 transient occurs. Therefore, the staff determined that the PWROGs selection of a nozzle 34 geometry for modeling that is representative of the U.S. PWR fleet nozzle geometry is a 35 practical and reasonable approach.

36 37 Based on the discussion above, the staff finds the four nozzle geometries listed in Table 4-1 of 38 the TR acceptable for representing the inlet and outlet nozzle designs in the U.S. PWR fleet.

39 40 4.6 Finite Element Model and Analyses 41 42 Model Creation 43 44 The PWROG described the FEMs of the inlet and outlet nozzles in Section 4.2, Model/Mesh, 45 Section 4.3, Flaw Modeling Methodology, and Section 4.6, Material Properties, of the TR.

46 The three-dimensional FEMs of the inlet and outlet nozzles included flaws in the nozzle corner 47 with depths of 0.05 inch and 0.5 inch into the LAS and with length-to-depth aspect ratios of 2:1 48 and 6:1. The mesh in the vicinity of the modeled flaws included very fine elements that have 49 features for handling the sharp edges around the flaw tip. The PWROG summarized the FEM 50 cases in Table 4-2, Flaw Case List, of the TR.

51 52 The staff reviewed the descriptions of the FEMs in Sections 4.2, 4.3, and 4.6 of the TR and finds 53 the methods (selection of element types, meshing, and definition of material properties) 54 acceptable.

1 Boundary Conditions 2

3 The PWROG described the boundary conditions applied to FEMs of the inlet and outlet nozzles 4 in Section 4.3, Thermal Boundary Conditions, and Section 4.4, Structural Boundary 5 Conditions, of the TR. Thermal boundary conditions included temperature coupling of the 6 coincident nodes of the modeled flaw, an assumption of infinite heat transfer coefficient on the 7 wetted surfaces, and insulated conditions on the surfaces of the FEMs where the models are 8 cut from the un-modeled structure. The structural boundary conditions included displacement 9 restraints, internal pressure on the wetted surface and on the crack face, end-cap pressure 10 loads on the modeled RPV shell and nozzle safe-end, and mechanical loads on the modeled 11 nozzle safe-end. Additionally, the temperature field from the thermal FEMs are applied to the 12 structural FEMs.

13 14 The staff reviewed the descriptions of the thermal and structural boundary conditions in 15 Sections 4.3 and 4.4 of the TR. One thermal boundary condition of the note is the assumption 16 of infinite heat transfer coefficient on the wetted surface. The staff determined that this 17 assumption produces a large temperature gradient across the nozzle section thickness due to a 18 cooldown transient, which generates conservative tensile stresses on the inside surface of the 19 nozzle corner. The staff, therefore, finds the assumption acceptable. The staff noted that the 20 application of the temperature field from the thermal FEMs into the structural FEMs is actually a 21 structural load in the structural FEMs and is therefore acceptable.

22 23 Based on the discussion above, the staff finds that the PWROG applied the proper boundary 24 conditions to the inlet and outlet nozzle FEMs, and therefore finds the boundary conditions 25 acceptable.

26 27 Loads 28 29 The PWROG described the loads applied to FEMs of the inlet and outlet nozzles in Section 4.7, 30 Loads, of the TR. The applied loads are the residual stress due to clad welding (clad residual 31 stress), mechanical piping loads, and cooldown transient. The internal pressure load is treated 32 as a boundary condition, which the staff evaluated in the Boundary Conditions section.

33 34 The staff reviewed the descriptions of the applied loads in Sections 4.7 of the TR. One applied 35 load of note is the clad residual stress. The staff reviewed the PWROGs modeling approach 36 that accounts for clad residual stress. The PWROG cited the review of the Sweden Nuclear 37 Power Inspectorate of programs that measured the effects of cladding on structural integrity of 38 cladded RPVs. Specifically, the PWROG referenced the residual stress profile measured 39 across the cladding of an RPV specimen. Then, using this residual stress profile as a reference 40 stress distribution and the FEM described in Section 4.2 of the TR, the PWROG determined, 41 through an iterative process, the average stress in the clad by adjusting the CTE reference 42 temperature of the clad material that would produce a similar effect at the flaw tip as the 43 measured clad residual stress profile. Given that the availability of residual stress 44 measurements due to clad welding is limited, the staff determined that this approach to address 45 the effect of clad residual stress is reasonable since the reference residual stress is based on 46 measured data. The staff also reviewed open literature and verified that the method of adjusting 47 the CTE reference temperature is a common approach to simulate a stress between two 48 adjacent materials. The staff, therefore finds the load due to clad residual stress acceptable.

49 50 The piping loads included those due to deadweight and thermal expansion loads at normal 51 operating conditions. The staff finds the piping loads acceptable. The cooldown transient 52 included one with composite rates (100°F/hour, then 50°F/hour, then 20°F/hour) and one with

1 100°F/hour for the limiting outlet nozzle FEM. The staff reviewed PWR systems manuals and 2 previous P-T limit curves for cooldown and determined that both cooldown transients are 3 acceptable.

4 5 Based on the discussion above, the staff finds the loads applied to the FEMs of the inlet and 6 outlet nozzle acceptable.

7 8 4.7 Stresses 9

10 The PWROG presented stresses for the inlet and outlet nozzle in Section 4.8, Stresses at 11 Limiting Locations, of the TR. The staff reviewed the stress contour plots due to internal 12 pressure and cooldown transient for the inlet and outlet nozzle FEMs and determined that the 13 stress values are within the expected values for these nozzles.

14 15 4.8 Stress Intensity Factors 16 17 The PWROG presented SIFs for the outlet nozzle in Section 4.9, Stress Intensity Factor 18 Results, of the TR and stated that it performed evaluations for both inlet and outlet nozzles, but 19 showed SIF results only for the outlet nozzle in the TR. The staff determined that showing SIF 20 results only for the outlet nozzle is sufficient for its review since the SIF results for the inlet 21 nozzle would show similar trends because it was subject to the same loads as the outlet nozzle.

22 The staff reviewed the SIF plots due to internal pressure and cooldown transient (which includes 23 the effect of clad residual stress) and determined that the SIF values are reasonable compared 24 to those calculated from a closed-form SIF solution for a nozzle corner crack.

25 26 4.9 Constraint and Cladding Effect 27 28 The staff reviewed the discussion of T-stress in Section 4.10.1, Constraint, of the TR, which is 29 commonly used as a measure of constraint and is correlated to toughness. The staff 30 determined that not taking credit for the increased toughness for a nozzle corner flaw (due to 31 lower constraint compared to the constraint on an SE(B) specimen, the data from which fracture 32 toughness is determined) is acceptable.

33 34 The staff also reviewed the discussion of cracking restraint due to cladding in Section 4.10.2, 35 Cladding, of the TR and determined that not taking credit for the ability of cladding to restrain 36 crack growth is acceptable.

37 38 4.10 Generic Nozzle P-T Limit Curves 39 40 The PWROG developed generic nozzle P-T limit curves in Section 5.1.1, Generation of Nozzle 41 P-T Limit Curves with Postulated Small Flaw, and Section 5.1.2, Generation of Nozzle P-T 42 Limit Curves with Postulated with 1/4T Beltline Thickness Size Flaw, of the TR based on the 43 methodology in Appendix G to Section XI of the ASME Code.

44 45 The PWROG presented the nozzle P-T limits for the postulated small flaws in Section 5.1.1 of 46 the TR with the ART determined in Section 3.5 of the TR, and the nozzle P-T limits for the 47 postulated 1/4T flaws in Section 5.1.2 of the TR with the ART determined in Section 3.1.2.2 of 48 the TR. The nozzle P-T limits for the small postulated flaws were based on SIFs developed in 49 Section 4 of the TR and included the effect of clad residual stress. The nozzle P-T limits for the 50 1/4T flaws were based on stresses determined from unflawed nozzle FEMs and SIFs from 51 ORNL/TM-2010/246 (Ref. 18 of the TR).

52

1 The staff reviewed the nozzle P-T limit curves in Sections 5.1.1 and 5.1.2 of the TR and 2 compared the limiting curves with known nozzle P-T limit curves. The staff finds the nozzle P-T 3 limit curves in the TR acceptable.

4 5 4.11 Comparison of Generic Nozzle P-T Limit Curves to RPV Shell P-T Limit Curves 6

7 In Section 5.2, Comparison of Nozzle to Traditional NRC Approved Pressure-Temperature 8 Limit Curves, the PWROG selected the limiting nozzle P-T limit curves developed in 9 Sections 5.1.1 and 5.1.2 of the TR and compared them to the NRC-approved P-T limits of the 10 shell (and associated welds) in the RPV beltline region. The PWROG determined that eleven 11 NRC-approved P-T limits of Westinghouse plants (identified in the TR as A through K) do not 12 bound the generic limiting nozzle P-T limits developed in this TR and evaluated them separately 13 in Figures 5-10 through 5-14 of the TR. For these eleven plants, plant-specific nozzle RTNDT 14 values were used instead of the generic nozzle RTT0 value developed in the TR, using the P-T 15 limit methodologies in Sections 4 and 5.1.2 of the TR. The staff reviewed the generic bounding 16 nozzle P-T limit curves compared to the NRC-approved P-T limit curves to determine whether 17 the PWROG adequately addressed that the NRC-approved P-T limit curves are bounding when 18 compared to bounding generic nozzle P-T limit curves. The PWROG provided additional 19 information in its supplement that aided the staffs review of the eleven plants in which the NRC-20 approved P-T limit curves did not bound the generic bounding nozzle P-T limits developed in the 21 TR.

22 23 The staff reviewed Figure 5-15 of the TR, which provided a comparison of the bounding nozzle 24 P-T limit curves compared to NRC-approved P-T limit curves for CE and B&W PWRs. Based 25 on this comparison, the staff finds the PWROG adequately addressed that the NRC-approved 26 P-T limit curves for CE and B&W PWRs bounds the generic nozzle P-T limit curves developed 27 in this TR, as shown in Figure 5-9 of the TR. The staff reviewed Figure 5-9 of the TR, which 28 provided a comparison of the Westinghouse bounding generic nozzle P-T limit curves compared 29 to the NRC-approved P-T limit curves for Westinghouse PWRs, except for the eleven plant-30 specific cases which are further discussed below (i.e., Plants A through K). Based on this 31 comparison, the staff finds the PWROG adequately addressed that the NRC-approved P-T limit 32 curves for Westinghouse PWRs (except for Plants A through K) bound the generic 33 Westinghouse nozzle P-T limit curves developed in this TR, as shown in Figure 5-9 of the TR.

34 The staffs review of Plants A through K identified in the TR is provided below.

35 36 For Plant A, the PWROG explained in its supplement that WCAP-18191-NP, which was 37 previously submitted to the NRC, contains a calculation of nozzle P-T limit curves using the 38 standard 1/4T nozzle corner flaw and the methods in ORNL/TM-2010/246, as well as the 39 determination of the initial RTNDT values for the nozzle forgings. The staff reviewed WCAP-40 18191-NP, Appendix B, and verified that the licensee performed confirmatory P-T limit curve 41 calculations of the RPV inlet and outlet nozzles. The staff noted that the Cu and Ni contents of 42 the nozzles were based on plant-specific CMTRs and that the unirradiated RTNDT values are 43 based on drop-weight data, TL CVN test data and NUREG-0800 Branch Technical Position 44 (BTP) 5-3, Fracture Toughness Requirements, Positions 1.1(3)(a) and (b), with the more 45 limiting unirradiated RTNDT value being selected. The staff noted that the methodology in BTP 46 5-3 paragraph 1.1(3)(b) was determined to be acceptable in closure memorandum dated April 47 2017 (ADAMS Accession No. ML16364A285). The staff noted that the licensee performed 48 these nozzle calculations solely to verify that the P-T limits for the RPV traditional beltline is 49 bounding compared to any P-T limit curves for the RPV inlet and outlet nozzles. The staff 50 verified that the licensee used the staff-developed methodology in ORNL/TM-2010/246 to 51 generate the P-T limits of the nozzles. Based on its review of the pertinent information in 52 WCAP-18191-NP, for the purposes of this TR, the staff finds the PWROG adequately

1 addressed that the NRC-approved P-T limit curve for Plant A bounds the nozzle P-T limit 2 curves, as shown in Figure 5-14 of the TR.

3 4 For Plant B, the PWROG confirmed that the NRC-approved P-T limit curves were previously 5 shown to not be impacted by the nozzle P-T limits curves using the 1/4T flaw, as documented in 6 letter dated January 22, 2015 (ADAMS Accession No. ML15029A417). The staffs review of this 7 comparison is documented in SE dated April 29, 2016 (ADAMS Accession No. ML16081A333).

8 The staff finds the PWROG has adequately addressed that the NRC-approved P-T limit curve 9 for Plant B is bounding.

10 11 For Plant C through Plant H, the staff noted the nozzle RTNDT values were measured to the 12 requirements of post-1973 ASME Subarticle NB-2300, and the uncertainty associated with an 13 RTNDT estimation method does not affect these RTNDT values. The staff also noted that the 14 PWROG used the FEM with postulated flaws in the TR to generate the nozzle P-T limit curves.

15 The staffs review of the FEM with postulated flaws in the TR are documented in Sections 4.5 16 through 4.9 of this SE. Based on its review, the staff finds it acceptable that the PWROG used 17 plant-specific nozzle RTNDT values instead of the A-508 Class 2 generic RTT0 developed in this 18 TR, along with the FEM with postulated flaws in the TR to generate the nozzle P-T limit curves.

19 Thus, the staff finds the PWROG has adequately addressed that the NRC-approved P-T limit 20 curves for Plant C though Plant H bound the nozzle P-T limit curve, as shown in 21 Figures 5-10, 5-11, and 5-12 of the TR.

22 23 For Plant I, the PWROG indicated in its supplement that the actual design dimension of the 24 cladding (5/8 inch after machining) was utilized with the postulated 0.5-inch deep LAS flaw, 25 which resulted in a flaw depth of 0.99 inch from the wetted surface. The SIFs for this 0.99-inch 26 flaw in the FEM were developed using the same methodologies that were used for the other 27 nozzle flaws in the TR. The staffs review of the FEM with postulated flaws in the TR are 28 documented in Sections 4.5 through 4.9 of this SE. For Plant I, the staff noted the nozzle 29 RTNDT values were measured to the requirements of post-1973 ASME Subarticle NB-2300, and 30 the uncertainty associated with an RTNDT estimation method does not affect these RTNDT values.

31 The staff finds it acceptable that the PWROG used plant-specific nozzle RTNDT values instead of 32 the A-508 Class 2 generic RTT0 developed in this TR, to generate the nozzle P-T limit curves.

33 Based on the use of the plant-specific nozzle RTNDT values and the PWROGs confirmation that 34 the flaw size for Plant I is based on the plant-specific design dimension of the cladding 35 thickness, the staff finds the PWROG adequately addressed that the NRC-approved P-T limits 36 curve for Plant I bounds the nozzle P-T limit curve, as shown in Figure 5-13 of the TR.

37 38 For Plants J and K, the PWROG confirmed in its supplement the initial RTNDT values for the 39 reactor vessel nozzle forging materials were determined using the methodology in BTP 5-3 40 paragraph 1.1(3)(b). The staff noted that the methodology in BTP 5-3 paragraph 1.1(3)(b) was 41 determined to be acceptable in closure memorandum dated April 2017 (ADAMS Accession No.

42 ML16364A285). Based on the PWROGs confirmation regarding the source of the initial RTNDT 43 values for Plants J and K, the staff finds the PWROG adequately addressed that the NRC-44 approved P-T limits curve for Plants J and K bounds their respective nozzle P-T limit curve, 45 as shown in Figure 5-11 of the TR.

46 47 Based on the staffs review of the comparison for the NRC-approved P-T limit curves and the 48 nozzle P-T limits developed in this TR, as described above, the staff finds that the PWROG has 49 adequately demonstrated that the nozzle P-T limit curves developed in the TR are bounded by 50 the NRC-approved P-T limit curves for U.S. PWRs.

51

1 5.0 USE AND REFERENCING OF THE TR 2

3 As addressed in the TR and in this SE, the use and referencing of this TR is only applicable to 4 U.S. PWR inlet and outlet nozzles with a projected nozzle corner neutron fluence, as calculated 5 by an NRC-approved method of fluence evaluation, of less than 4.28 x 1017 n/cm2 (E > 1 MeV).

6 As noted in the TR, if the nozzle fluence is greater than 4.28 x 1017 n/cm2 (E > 1 MeV), the shift 7 (RTNDT) may be calculated for those nozzles on a plant-specific basis using an NRC-approved 8 method; as long as the shift remains below 25°F or the plant-specific ART values remain below 9 the ART determined in the TR, the analysis in the TR is applicable.

10 11

6.0 CONCLUSION

12 13 The staff has reviewed the TR including the supplemental information, and based on the 14 evaluation in Section 4 of this SE, finds the TR as modified by this SE, provides an acceptable 15 means for addressing the potential for P-T limit curves for inlet or outlet nozzle corners of U.S.

16 PWRs to be more limiting than the current NRC-approved P-T limits (as of the time of issuance 17 of this SE) of the shell (and associated welds) in the traditional beltline region of the RPV. The 18 staffs independent safety assessment in Section 4.3 of this SE of the TRs use of the 19 recommendation in TLR-RES/DE/CIB-2013-01 of excluding 25°F embrittlement is specific only to 20 the TR, and as such, should not be construed as a generic safety assessment. Thus, other 21 applications that use the recommendation in TLR-RES/DE/CIB-2013-01 must be sufficiently 22 justified and shall be subject to NRC review and approval on a case-by-case basis. Accordingly, 23 PWROG-15109-NP, as modified by this SE, is acceptable for referencing to satisfy the fracture 24 toughness requirements in Appendix G to 10 CFR Part 50 for U.S. PWR inlet and outlet nozzles 25 only, which provide adequate margins of safety during any condition of normal operation, 26 including anticipated operational occurrences and system hydrostatic tests, to which the 27 pressure boundary may be subjected over its service lifetime.

28 29 Principal Contributor: On Yee 30 31 Date: May 28, 2019

Retrieved from "https://kanterella.com/w/index.php?title=ML19158A265&oldid=1908069"