ML072060138
| ML072060138 | |
| Person / Time | |
|---|---|
| Site: | Wolf Creek  |
| Issue date: | 07/17/2007 |
| From: | Broussard J, Collin J, Klug M, White G Dominion Engineering |
| To: | Office of Nuclear Reactor Regulation |
| References | |
| Download: ML072060138 (45) | |
Text
11730 Plaza America Dr. #310 Reston, VA 20190 703.437.1155 www.domeng.com Advanced FEA Crack Growth Calculations for Evaluation of PWR Pressurizer Nozzle Dissimilar Metal Weld Circumferential PWSCC Sponsored by: EPRI Materials Reliability Program Presented To:
Expert Review Panel for Advanced FEA Crack Growth Calculations Presented By:
Glenn White John Broussard Jean Collin Matthew Klug Dominion Engineering, Inc.
Tuesday, July 17, 2007 Meeting on Implications of Wolf Creek Dissimilar Metal Weld Inspections DEI Offices, Reston, Virginia and via Webcast
Project Review Meeting: Advanced FEA Crack Growth Evaluations 2
July 17, 2007, Reston, VA, and via Webcast Agenda Introductions / Opening Remarks Results Missing from Draft A Report New Case S9b to Further Address Effect of Multiple Flaws Validation Evaluation Criteria Final Industry Report July 12 NRC Comments Remaining Work Meeting Summary and Conclusions
Project Review Meeting: Advanced FEA Crack Growth Evaluations 3
July 17, 2007, Reston, VA, and via Webcast Principal Meeting Participants EPRI Project Management / Support
- Craig Harrington, EPRI
- Christine King, EPRI
- Tim Gilman, Structural Integrity Associates Project Team
- Glenn White, DEI
- John Broussard, DEI
- Jean Collin, DEI
- Matthew Klug, DEI Expert Review Panel
- Ted Anderson, Quest Reliability, LLC
- Warren Bamford, Westinghouse
- David Harris, Structural Integrity Associates
- Doug Killian, AREVA
- Pete Riccardella, Structural Integrity Associates
- Ken Yoon, AREVA NRC Participants
- Al Csontos, NRC Research
- Tim Lupold, NRC NRR
- Dave Rudland, EMC2
- Simon Sheng, NRC NRR
- Ted Sullivan, NRC NRR
Project Review Meeting: Advanced FEA Crack Growth Evaluations 4
July 17, 2007, Reston, VA, and via Webcast Results Missing from Draft A Report Summary Stable arrest was still be confirmed for 6 matrix cases
- Now confirmed using FEACrack Five matrix cases were still in progress
- Case 23c multiple repair case: Results on next two slides
- Case 28b surge nozzle case: Results on next slide
- Case 36c stress redistribution case: Results expected July 17
- Cases 52 and 53 with nozzle-to-safe-end geometry: Results expected ~July 23 Leak rate was still to be reported in Table 7-6 for Case 48b at time load margin factor reaches 1.2
- Missing leak rate is 70.1 gpm Leak rate and stability time plots were requested for all cases with a load margin factor of ~1.7 or lower when leak rate is 1 gpm
- Results not in Draft A are provided below for all cases with factor 1.75
Project Review Meeting: Advanced FEA Crack Growth Evaluations 5
July 17, 2007, Reston, VA, and via Webcast Results Missing from Draft A Report Case 23c and 28b Results Nozzle Type Geometry Configuration Ri (in) t (in)
Time to TW (yrs)
Fraction Xsection Cracked Crack Face F (kips)
Max tot Faxial (kips)
Max Pm Based on CF (ksi)
Stability Margin Factor Support.
Pm (ksi)
Support.
Pb (thick)
(ksi) 23 c
S&R Config 2a/2b 2.810 1.065 0.5 0.298 14.91 75.33 3.37 3.55 12.0 27.1 28 b
surge bounding 5.920 1.580 3.4 0.518 77.12 331.23 4.97 1.25 6.2 12.5 Geometry Case WRS Subcase Case Surface Crack Stability Results (Press + DW + NT loads and Z-factor for Critical Size)
Case and Step Fraction Xsection Cracked Crack Face Force (kips)
Max tot Faxial (kips)
Max Pm Based on CF (ksi)
Support.
Pm (ksi)
Support.
Pb (thick)
(ksi)
Stability Margin Factor Time since TW (hrs)
Time since TW (days)
Leak Rate (gpm @
70°F)
C23cS16 0.369 18.44 78.87 3.53 10.53 22.76 2.99 1799 75 1.003 C28bS00 0.355 52.84 306.96 4.61 8.49 15.47 1.84 0
0 2.426 Case and Step Fraction Xsection Cracked Crack Face Force (kips)
Max tot Faxial (kips)
Max Pm Based on CF (ksi)
Support.
Pm (ksi)
Support.
Pb (thick)
(ksi)
Stability Margin Factor Time since TW (hrs)
Time since TW (days)
Time since 1 gpm (hrs)
Time since 1 gpm (days)
Leak Rate (gpm @
70°F)
C23cS20 0.387 19.34 79.76 3.57 10.08 21.57 2.83 2296 96 497
>>21 1.272 C28bS30 0.424 63.08 317.20 4.76 5.71 12.00 1.20 655 27 655 27 28.752
Project Review Meeting: Advanced FEA Crack Growth Evaluations 6
July 17, 2007, Reston, VA, and via Webcast Results Missing from Draft A Report Case 23c Crack Mesh for TW Step 20
Project Review Meeting: Advanced FEA Crack Growth Evaluations 7
July 17, 2007, Reston, VA, and via Webcast Results Missing from Draft A Report Leak Rate and Load Margin Factor vs. TimeCase 17b 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 0
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 Time after Initial Through-Wall Crack (days)
Leak Rate (gpm @ 70°F) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 Stability Margin on Load C17b Leakage C17b Stability Margin Load Factor = 1.2
Project Review Meeting: Advanced FEA Crack Growth Evaluations 8
July 17, 2007, Reston, VA, and via Webcast Results Missing from Draft A Report Leak Rate and Load Margin Factor vs. TimeCase 27b 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 0
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 Time after Initial Through-Wall Crack (days)
Leak Rate (gpm @ 70°F) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 Stability Margin on Load C27b Leakage C27b Stability Margin Load Factor = 1.2
Project Review Meeting: Advanced FEA Crack Growth Evaluations 9
July 17, 2007, Reston, VA, and via Webcast Results Missing from Draft A Report Leak Rate and Load Margin Factor vs. TimeCase 44c 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 0
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 Time after Initial Through-Wall Crack (days)
Leak Rate (gpm @ 70°F) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 Stability Margin on Load C44c Leakage C44c Stability Margin Load Factor = 1.2 Time at 1 gpm
Project Review Meeting: Advanced FEA Crack Growth Evaluations 10 July 17, 2007, Reston, VA, and via Webcast Results Missing from Draft A Report Leak Rate and Load Margin Factor vs. TimeCase 46b 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 0
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 Time after Initial Through-Wall Crack (days)
Leak Rate (gpm @ 70°F) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 Stability Margin on Load C46b Leakage C46b Stability Margin Load Factor = 1.2
Project Review Meeting: Advanced FEA Crack Growth Evaluations 11 July 17, 2007, Reston, VA, and via Webcast Results Missing from Draft A Report Leak Rate and Load Margin Factor vs. TimeCase 48b 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 0
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 Time after Initial Through-Wall Crack (days)
Leak Rate (gpm @ 70°F) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 Stability Margin on Load C48b Leakage C48b Stability Margin Load Factor = 1.2
Project Review Meeting: Advanced FEA Crack Growth Evaluations 12 July 17, 2007, Reston, VA, and via Webcast Results Missing from Draft A Report Leak Rate and Load Margin Factor vs. TimeCase S1b 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 0
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 Time after Initial Through-Wall Crack (days)
Leak Rate (gpm @ 70°F) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 Stability Margin on Load S1b Leakage S1b Stability Margin Load Factor = 1.2
Project Review Meeting: Advanced FEA Crack Growth Evaluations 13 July 17, 2007, Reston, VA, and via Webcast Results Missing from Draft A Report Leak Rate and Load Margin Factor vs. TimeCase S2b 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 0
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 Time after Initial Through-Wall Crack (days)
Leak Rate (gpm @ 70°F) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 Stability Margin on Load S2b Leakage S2b Stability Margin Load Factor = 1.2
Project Review Meeting: Advanced FEA Crack Growth Evaluations 14 July 17, 2007, Reston, VA, and via Webcast Evaluation Case Matrix Effect of Multiple Crack Initiation Sites Sensitivity cases investigate the effect of multiple crack initiation (e.g., Wolf Creek surge nozzle NDE results)
- Enveloping of multiple initial flaws with one modeled flaw
- Modeling of a part-depth 360° flaw
- Growing multiple individual flaws and then combining on a single weld cross section for stability calculation
Project Review Meeting: Advanced FEA Crack Growth Evaluations 15 July 17, 2007, Reston, VA, and via Webcast New Case S9b Further Addresses Effect of Multiple Flaws Case S9b added to further address this concern for limiting surge nozzles
- Case 9b is based on Case 17b, but with 21:1 26%tw initial flaw placed at top and bottom of weld cross section
- Crack interaction effects are insignificant for this case based on distance between flaws and Quest Reliability, LLC experience with interaction effects
- Thus, leak rate and stability margin trends can be based on separate growth of flaws and then combination of flaws in crack stability calculation
- Two 21:1 26%tw initial flaws represent 46% (167°) of the ID circumference Results vs. Case 17b
- 1.22 years to go through-wall is unaffected
- Leak rate trend with time of Case 17b is unaffected
- Stability margin factor trend is lowered by between 0.10 and 0.12
- Time from 1 gpm to load margin factor of 1.2 is reduced from 35 to 29 days
Project Review Meeting: Advanced FEA Crack Growth Evaluations 16 July 17, 2007, Reston, VA, and via Webcast New Case S9b Crack Profiles vs. Time: Cartesian Coordinates
-8.0
-6.0
-4.0
-2.0 0.0 2.0 4.0 6.0 8.0
-8.0
-6.0
-4.0
-2.0 0.0 Case S9b after breaking through-wall Case S9b 7 days after through-wall Case S9b 12 days after through-wall Case S9b 25 days after through-wall Case S9b 35 days after through-wall Case S9b 42 days after through-wall
Project Review Meeting: Advanced FEA Crack Growth Evaluations 17 July 17, 2007, Reston, VA, and via Webcast New Case S9b Crack Profiles vs. Time: Polar Coordinates 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0
30 60 90 120 150 180 Circumferential Position, (deg)
Nondimensional Crack Depth, y /t Case S9b after breaking through-wall Case S9b 7 days after through-wall Case S9b 12 days after through-wall Case S9b 25 days after through-wall Case S9b 35 days after through-wall Case S9b 42 days after through-wall
Project Review Meeting: Advanced FEA Crack Growth Evaluations 18 July 17, 2007, Reston, VA, and via Webcast New Case S9b Leak Rate and Load Margin Factor vs. Time 42, 1.00 42, 1.00 42, 1.00 42, 1.00 42, 1.00 25, 1.28 42, 0.90 42, 0.90 42, 0.90 42, 0.90 42, 0.90 0, 1.73 25, 1.40 12, 1.60 35, 1.20 0, 1.71 35, 1.09 0, 1.60 7, 1.52 12, 1.48 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 160.0 0
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 Time after Initial Through-Wall Crack (days)
Leak Rate (gpm @ 70°F) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 Stability Margin on Load Case 17b and Case S9b Leakage Case 17b Stability Margin Case S9b Stability Margin Load Factor = 1.2
Project Review Meeting: Advanced FEA Crack Growth Evaluations 19 July 17, 2007, Reston, VA, and via Webcast Validation Topics Duane Arnold EU Mockup MRP-107
Project Review Meeting: Advanced FEA Crack Growth Evaluations 20 July 17, 2007, Reston, VA, and via Webcast Validation Duane Arnold Circumferential Crack The Duane Arnold crack was applied as a validation case
From MRP-113: Crack initiation and growth were attributed to the presence of a fully circumferential crevice that led to development of an acidic environment because of the oxygen in the normal BWR water chemistry, combined with high residual and applied stresses as a result of the geometry and nearby welds. The water chemistry conditions that contributed to cracking at Duane Arnold do not exist for the case of Alloy 82/182 butt welds in PWR plants.
Project Review Meeting: Advanced FEA Crack Growth Evaluations 21 July 17, 2007, Reston, VA, and via Webcast Validation Duane Arnold Circumferential Crack (contd)
Duane Arnold WRS (ksi) Profile Fit y = 2091.284192x 4 - 4024.030339x 3 + 2171.322441x 2 - 279.638139x + 4.697888
-50
-40
-30
-20
-10 0
10 20 30 40 50 60 70 80 0.0 0.2 0.4 0.6 0.8 1.0 Dist from Thermal Sleeve ID (in.)
Axial WRS (ksi)
Tip of Crevice Thermal Sleeve Safe-End From 30% TW 360° Surface Flaw Actual Crack Profile Simulated Crack Profile
Project Review Meeting: Advanced FEA Crack Growth Evaluations 22 July 17, 2007, Reston, VA, and via Webcast Validation EU MockupDEI Hoop Stress
Project Review Meeting: Advanced FEA Crack Growth Evaluations 23 July 17, 2007, Reston, VA, and via Webcast Validation EU MockupDEI Axial Stress
Project Review Meeting: Advanced FEA Crack Growth Evaluations 24 July 17, 2007, Reston, VA, and via Webcast Validation EU MockupDEI Butter Hoop Stress
Project Review Meeting: Advanced FEA Crack Growth Evaluations 25 July 17, 2007, Reston, VA, and via Webcast Validation EU MockupDEI Butter Axial Stress
Project Review Meeting: Advanced FEA Crack Growth Evaluations 26 July 17, 2007, Reston, VA, and via Webcast Validation MRP-107 Lab Study of PWSCC in Alloy 182 The report summary for MRP-107 (EPRI 1009399, 2004) includes the following:
- Abstract: Detailed examinations of Alloy 182 capsule samples containing PWSCC established the relationship between crack initiation sites and the microstructure of the weld metal. These examinations also identified microstructural features that facilitate or arrest PWSCC propagation. Crack initiation only occurred at high angle, high energy, dendrite packet grain boundaries, and growth apparently arrested at low energy boundaries due to low angular misorientation or coincidence of lattice sites. The work also revealed important findings with regard to crack geometries, in particular what aspect ratios may develop during PWSCC of nickel-base (Ni-base) weld metals.
- The cracks exhibited an unusual aspect ratio in that they never showed a large lateral surface extent, even when they extended through the wall thickness. This is a very different feature compared to PWSCC in Ni-base alloys such as Alloy 600. The aspect ratio is thought to relate to indications of crack arrest observed at low energy grain boundaries in Alloy 182.
Project Review Meeting: Advanced FEA Crack Growth Evaluations 27 July 17, 2007, Reston, VA, and via Webcast Evaluation Criteria Figure 7-1 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 0
10 20 30 40 50 60 70 80 90 Time after Initial Through-Wall Crack (days)
Leak Rate (gpm at 70°F) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Stability Margin on Load 7 days x4 on 0.25 gpm leak rate stability margin leak rate begin time load factor of 1.2 Does this point fall below the stability margin line?
Illustration of Approach for Hypothetical Leak Rate and Crack Stability Results
Project Review Meeting: Advanced FEA Crack Growth Evaluations 28 July 17, 2007, Reston, VA, and via Webcast Final Industry Report Topics Preliminary Results Preliminary Conclusions Schedule
- Draft B
- Industry and NRC Review
- Main report
- Appendix A on probabilistic assessments
- Schedule for Release of Rev. 0 Missing items
- EPRI report summary, abstract, and full list of acronyms
- EU Mockup WRS simulation
- Discussion of Implications of MRP-107
- Duane Arnold crack growth case
- Missing matrix results and discussion, including new Case S9b
- Move references to Section 9
Project Review Meeting: Advanced FEA Crack Growth Evaluations 29 July 17, 2007, Reston, VA, and via Webcast Final Industry Report Preliminary Results All 105 completed cases in the main sensitivity matrix showed either
- stable crack arrest (59 cases), or
- crack leakage and crack stability results satisfying the evaluation criteria (46 cases)
- generally considerable margins beyond evaluation criteria 10 supplemental cases further investigated effect of multiple flaws on limiting surge nozzle cases
- Conservative application of the three indications found in the Wolf Creek surge nozzle weld to limiting surge nozzles (fill-in weld and relatively high moment load) gives results meeting the evaluation criteria with additional margin
- Multiple flaw case based on Case 17b with two 21:1 26%tw initial flaws at opposite sides of model shows modest effect on crack stability, with reduction of only 6 days in time interval from 1 gpm leak rate to 1.2 load margin factor (35 to 29 days)
- On this basis, it is concluded that the concern for multiple flaws in the limiting surge nozzles is adequately addressed by cases that satisfy the evaluation criteria with additional margin
Project Review Meeting: Advanced FEA Crack Growth Evaluations 30 July 17, 2007, Reston, VA, and via Webcast Final Industry Report Preliminary Conclusions Assumption of semi-elliptical flaw shape shown to result in large unnecessary overconservatism All 51 subject welds are adequately covered by crack growth sensitivity cases that satisfy the evaluation criteria Results show tendency of circumferential surface cracks to show stable arrest
- Axisymmetric welding residual stress profile must self-balance
- Consistent with Wolf Creek experience given unlikeliness that four indications found in narrow depth band were growing rapidly at that time Sensitivity cases indicate a large beneficial effect of relaxation of secondary loads upon through-wall penetration
- Detailed evaluations tend to support such a relaxation effect
- Not credited in main cases
Project Review Meeting: Advanced FEA Crack Growth Evaluations 31 July 17, 2007, Reston, VA, and via Webcast July 12 NRC Comments Welding Residual Stress Uncertainty The WRS profile applied in Case 17b is conservative with respect to:
- DEI WRS FEA result for Type 8 surge nozzle, including SS weld simulation
- EMC2 WRS FEA result for Type 8 surge nozzle, including SS weld simulation
- ASME profile as modified by EMC2 Because the WRS profile applied in Case 17b is shifted significantly in the conservative direction versus all three of these profiles, it appropriately addresses the effect of WRS uncertainty Consistent with the most likely Wolf Creek behavior, the EMC2 WRS FEA result for Type 8 surge nozzle (including SS weld simulation) leads to crack arrest in the growth model
Project Review Meeting: Advanced FEA Crack Growth Evaluations 32 July 17, 2007, Reston, VA, and via Webcast July 12 NRC Comments Surge Nozzle Axial WRS at NOT
-80,000
-60,000
-40,000
-20,000 0
20,000 40,000 60,000 80,000 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Nondimensional Distance from ID, x /t Axial Stress (psi)
Type 8-1 (base case)
ASME Modified per EMC2 EMC2 NoRepair-WithSS (Left-Right)
Project Review Meeting: Advanced FEA Crack Growth Evaluations 33 July 17, 2007, Reston, VA, and via Webcast July 12 NRC Comments Effect of Multiple Through-Wall Crack Segments Bill Shack of ACRS inquired on July 11 regarding the effect of multiple through-wall flaw segments on the leak rate The effect of multiple through-wall flaw segments to reduce the leak rate (in comparison to a single through-wall flaw) is mitigated by the increased resistance to rupture provided by the ligaments between the flaw segments
- Significant axial offsets between crack segments are perhaps likely because of the relatively long axial region of susceptible material The effect of the tight intergranular SCC type morphology is generally addressed by the leak rate prediction methodology
Project Review Meeting: Advanced FEA Crack Growth Evaluations 34 July 17, 2007, Reston, VA, and via Webcast July 12 NRC Comments Effect of Multiple Through-Wall Crack Segments (contd)
Substantial margin beyond the evaluation criteria exists for nearly all cases in main matrix
- Applying a leak rate margin factor of 10 rather than 4 on the 0.25 gpm detectability limit results in all 14 of the most limiting cases* in the main matrix satisfying the evaluation criteria with one exception (Case 44c)
- A leak rate margin factor of about 9 does satisfy the evaluation criteria for Case 44c
- All other cases in the main matrix very likely satisfy the evaluation criteria with a leak rate margin factor of 10 based on the compiled leak rate and stability data
- A leak rate margin factor of 10 has historically been applied in long-term regulatory LBB assessments
- The most limiting surge nozzle case (Case 17b) is predicted to have an initial through-wall leak rate of 2.6 gpm, with the leak rate increasing to 69 gpm when the load margin factor decreases to 1.2, indicating robustness with respect to the value of the leak rate margin factor
- The 14 most limiting cases are defined here as those cases for which the load margin factor is 1.75 or less at the time the leak rate is calculated to be 1 gpm.
Project Review Meeting: Advanced FEA Crack Growth Evaluations 35 July 17, 2007, Reston, VA, and via Webcast July 12 NRC Comments Effect of Multiple Through-Wall Crack Segments (contd)
Given the above points, the matrix results show sufficient margin to address modeling uncertainties such as those associated with the potential for multiple through-wall crack segments More detailed evaluation of the effect of multiple through-wall crack segments may be considered in the context of longer-term evaluations
- More detailed evaluations will require significant additional developmental effort
- Current study has had benefit of significantly refining crack growth evaluation tools, but explicit evaluation of multiple flaws is an emerging area
Project Review Meeting: Advanced FEA Crack Growth Evaluations 36 July 17, 2007, Reston, VA, and via Webcast Remaining Work Remaining DEI Work
- Nozzle-to-safe end geometry crack growth cases Final Industry Report August 9 Meeting at North Bethesda Marriott NRC Safety Assessment
Project Review Meeting: Advanced FEA Crack Growth Evaluations 37 July 17, 2007, Reston, VA, and via Webcast Nozzle-to-safe-end Geometry Cases Example Cracked Model
Project Review Meeting: Advanced FEA Crack Growth Evaluations 38 July 17, 2007, Reston, VA, and via Webcast Meeting Summary and Conclusions Industry NRC
NRC Questions and Comments on the NRC Questions and Comments on the Industry Advanced FEA Draft Report Industry Advanced FEA Draft Report Ted Sullivan & Al Csontos July 17, 2007
2 U.S. Nuclear Regulatory Commission General Comments General Comments
- NRC staff has reviewed the industry draft report and will provide our comments and questions today
- NRC recognizes the significant effort to develop, benchmark, verify, and evaluate the advanced FEA program and the validation of weld residual stresses
- Overall, the industry developed a groundbreaking and technically sound research program
- The products from this research program will be essential in resolving the regulatory issues at hand
3 U.S. Nuclear Regulatory Commission General Comments General Comments
- NRC comments are for clarification & completeness
- For many, the industrys advanced FEA report will be sole source of information on this issue
- Report needs some additional information provided in the industrys presentations at public meetings
- Interested parties outside the project deliberations may need the additional information for clarity
- NRC needs to review the references, supporting sections, and appendices as soon as they become available
4 U.S. Nuclear Regulatory Commission General Questions General Questions
- When will the supplementary analyses discussed in the draft industry report be available to the NRC?
- EU validation writeup?
- Westinghouse fabrication writeups?
- David Harris leak rate writeup?
- References?
- What, if any, additional cases will the industry run?
- Will the industry respond to Bill Shacks comments related to leak rate modeling with multiple TWCs?
- What are industrys plans for a peer review?
Page #
Line #
NRC 1-2 27-29 The group of nine PWRs should be planning to accelerate outages or take mid-cycle outages based on commitments made.
Revise to read...the group of nine PWRs planning to accelerate outages or take mid-cycle outages based on commitments made. Should this study demonstrate flaw stability via sufficient time from initial detectable leakage until pipe rupture, as demonstrated to the NRC, these plants could then resume plans to perform PDI inspection or mitigation during the spring 2008 outage season.
NRC 1-6 6-7 The images of each example pressurizer nozzle should contain identification markers for all major fabrication components Add identification markers NRC 1-7 1
The image of the CE pressurizer nozzle should contain identification markers for all major fabrication components Add identification markers NRC 2-General In general, this section needs to be augmented with the information provided in previous public meetings to include more figures detailing the nozzle geometries, dimensions, and typical fabrication procedures for the three types of plants; CE, W, W with CE-like fabrication procedures Augment this section to include information previously provided in public meetings by DEI (Glenn White) and Westinghouse (Cameron Martin)
NRC 2-2 14 It would be helpful to explain why the definition of Pm is given as PDo/4t in this section, but the pressure is used differently in the crack growth portion of the work Add explanation NRC 2-3 9
With regards to the back welding, this should clearly state that certain amounts of the ID material are removed and weld material is reapplied to the ID.
Revise, make similar to Type 8 Surge Nozzle description on pate 3-5, lines 23-31.
NRC 2-3 16 In addition to no fill-in welds, include no back chipping/backwelding completed Make revision NRC 2-5 3
In this table labels such as land thickness and fill-in weld are a bit confusing. It is suggested that figures in the WRS section be placed here to help better explain these geometric details.
Provide figures with the major fabrication details of the nozzles identified.
NRC 2-6 Question for Plant H spray line PDI results.
Recheck Plant H results.
NRC 3-1 16-30 Some dimensions are needed for completeness of the report, i.e., SS weld location, fill-in dimensions, etc. (see general comment)
Add dimensions, figures, or references as needed.
NRC 3-2 6
Note that the 5/16 inch repair simulates the back chipping and weld buildup Make revision NRC 3-2 22 How long is the repair?
Add length of repair to explanation NRC 3-2 24 this line notes that minor simplifications are made List the minor simplifications NRC 3-2 26 In the WRS analyses long slender weld beads were used. Since these beads differ from actual weld beads, please explain how the approximation resulting from this approach was assessed and reflected in the study.
Add explanation NRC 3-2 35 How was piping system compliance treated in the WRS analysis? How long was the stainless steel pipe in the analysis?
Add explanation NRC 3-3 16-31 Why was the elastic limit defined as the yield strength for the base metal but the flow strength for the weld material? The base metal near the weld may also melt and solidify as would the weld.
Add explanation NRC 3-3 34 The stainless steel yield strength of 28.9 ksi at 600F seems to be a high value; in looking at 550F tensile data from Vol. 8 of Degraded Piping semi-annual reports, the yield strengths for 5 stainless steel pipes ranged from 20.1 ksi to 26.1 ksi with an average value of 22.7 ksi. How would this affect the WRS modeling and critical crack size calculations?
Add explanation NRC 3-4 4
Different than what, the base material?
Add explanation NRC 3-5 4-10 The stress improvement observed from a hydrotest will be limited for low R/t pipes such as on the pressurizer None NRC 3-5 8
hydrostatic testing referred to as a form of mechanical stress improvement Delete mechanical and just refer to as a form of stress improvement NRC 3-5 13 Why was 653F listed as the operating temperature when the surge line will be at a temperature closer to 644F?
Add explanation NRC 3-5 34 Refers to an amount of the ID that was "ground out" Was this ground or machined? If machined, revise.
NRC 3-5 36 Statement that the inside surface is 0.25 inch smaller should read, "the inside diameter is 0.25 inches smaller replace surface with diameter NRC 3-6 1
"removed back" Replace with machined, if actually machined NRC 3-6 5
States the welds were v-weld. Should be U-groove Verify this is suppose to be U-groove NRC 3-6 5
v-weld Refer to as v-groove weld NRC 3-6 36 Was it verified that a stress path perpendicular to the axial direction represents the maximum stress path for PWSCC growth?
Add explanation NRC 3-7 3-16 Was the path changed for the case with the SS weld since the path of maximum stress may shift location??
Add explanation NRC 3-8 10 Section 3.3 should be referred to as Validation and Benchmarking Add title Cmt #
Comments and Proposed Changes: Due Date: July 20, 2007 Advanced FEA Evaluation of Growth of Postulated Circ PWSCC Flaws in PZR Nozzle DM Welds - Rev. A Comment Resolution Comment Location Comment Proposed Change Reviewer's Organization
Page #
Line #
Cmt #
Comments and Proposed Changes: Due Date: July 20, 2007 Advanced FEA Evaluation of Growth of Postulated Circ PWSCC Flaws in PZR Nozzle DM Welds - Rev. A Comment Resolution Comment Location Comment Proposed Change Reviewer's Organization NRC 3-8 14-17 Based on the available WRS validation and benchmarking activities to the EU report, this section should assess uncertainties related to WRS and how they will be addressed in the overall advanced FEA Phase II sensitivity matrix.
Add section NRC 3-15 5
Identify what is meant by 'backweld' Add clarification in text.
NRC 4-1 20 U-groove weld geometry stated here vs. V-groove geometry sated in section 3, page 6 and in the figures Revise to make consistent between sections NRC 4-1 27 What the rationale for using 8-noded brick elements in a computational fracture mechanics analysis? The crack tip singularity ahead of simulated sharp crack is approximated by collapsed 20-noded elements with the midpoint nodes moved to the quarter point location. If 8-noded elements are used, they must be of sufficient small size.
Explain and provide mesh sensitivity results for using the 8-noded elements NRC 4-2 16 Since ANSYS does not calculate fracture parameters, please explain the process on how it was done in this study Add explanation NRC 4-5 39 What is meant by ' remain self-similar'?
Add correction NRC 4-6 14 Why was the time between leakage and rupture not reported for the Phase I study?
Add Phase I leakage and margin results NRC 4-7 5
When you provide Duane Arnold crack validation, please provide what parameters were modified to obtain the final validation results.
Add explanation NRC 5-1 28 Explain CMTR Add explanation NRC 5-2 7
The study conducted by Riccardella and Anderson should be discussed in more detail since the report refers to the results several times. It would be nice if they could be included in detail as part of this report or in an appendix Add work from Riccardella and Anderson to report NRC 5-2 15 Along with radial differential thermal expansion, WRS are also not included in the stability calculations Modify sentence NRC 5-3 18 The DPZP may not be greater than unity for the cases of a complex crack where the apparent toughness is greatly reduced as compared to the C(T) toughness remove this sentence or modify NRC 5-4 8
It should be noted that in this study the DPZP was calculated using the C(T) toughness and not the apparent toughness for complex cracks This should be noted in this sentence NRC 6-1 14 Spell out COD Spell out as crack open displacement (COD)
NRC 6-1 25 Explain what (albeit over a longer length) means Add explanation NRC 6-2 29 List what the summary is provided in List Table 6-2 (?) as containing the summary of inputs NRC 6-3 10-11 Explain what is meant by the potentially important effect of moment bending Add explanation NRC 7-2 6-7 Further explain how the conversion is a conservative assumption given the complex crack envelopes the TW crack Add explanation NRC 7-2 15-16 Why were different axial stresses used in the stability and crack growth calculations?
Modify sentence NRC 7-2 35 Rather than state the acceptability involves licensing and regulatory issues, state the acceptability is dependent upon uncertainty of input parameters and the accuracy of the modeling methodology Revise as noted NRC 7-3 25 Recommend listing either six or seven days as being conservative.
Change to state "six days is conservatively required for the plant to shut down.
NRC 7-3 5
What is the basis for the 7 day criteria?
Provide the basis NRC 7-3 26-27 Would like to see discussion of where 0.25 gpm comes from, i.e. RCS leak rate monitoring committed to by licensees. Some mention of the baseline and that the margin factor of 4.0 also addresses leak rate changes that may occur that could affect the baseline but have not been incorporated in the baseline, i.e. a high baseline may have been measured in 1st seven days of operation that decreases over time or leaks may have been identified and repaired that effectively reduce the baseline. It would also be worthwhile to list the range of values used as baseline at the nine plants assessed in this evaluation.
Indicate where the 0.25 gpm is from. Include discussion of the baseline and how the margin factor of 4 encompasses baseline changes that may not have resulted in a baseline revision. Include range of baseline values being used.
NRC 7-4 3
New readers may not understand the stability margin factor.
Provide a definition of the stability margin factor NRC 7-4 5
Expand the explanation for why the factor of 1.2 is considered appropriate expand explanation NRC 7-4 12 The statement that there is no clear evidence that a purely limit load based approach is insufficient should be worded differently since there is no clear evidence that the purely limit load based approach is sufficient for cracks in A82/182 Modify the statement to say there is no experimental data on circumferential cracks in A82/182 that verify that limit load or elastic-plastic fracture conditions control.
NRC 7-6 19 Here the operating temperature is given as 650F. Table 6-1 and Page 3-5 give the operating temperature as 653F Correct temperature NRC 7-7 35 Were the stresses from Figure 3-19 used in the repair analyses?? Were the values interpolated from what's shown in Figure 3-19 to each circumferential position in the crack growth analyses?
Add explanation
Page #
Line #
Cmt #
Comments and Proposed Changes: Due Date: July 20, 2007 Advanced FEA Evaluation of Growth of Postulated Circ PWSCC Flaws in PZR Nozzle DM Welds - Rev. A Comment Resolution Comment Location Comment Proposed Change Reviewer's Organization NRC 7-10 28 This write-up always assumes that the dominate crack initiates on top of the pipe. What if it does not initiate on the top of the pipe?
Add explanation NRC 7-12 33 Same as previous Add explanation NRC 7-14 20-21 Under what conditions would the thermal loads be reduced during surface crack growth in such a way that the reduction would cause arrest?? If the surface crack is near critical, and much rotation occurs, some of the displacement-controlled loads may be reduced, but during subcritical crack growth, I'm not sure when the displacement controlled loads would be relieved Add explanation NRC 7-14 30-31 It is stated that using 360 deg flaws in surge nozzle analyses is not appropriate. However, 360 deg cracks were assumed in Case 18,26,29 and
- 30. Please clarify?
Add explanation NRC 7-15 1-2 Same comment as 7-10 line 28 Add explanation NRC 7-15 16 It may be helpful to have a table that shows the how the parameters varied in the sensitivity matrix affected the margin, i.e., 10% decrease in as built wall thickness can decrease margin by 30%
Add table if possible NRC 7-15 16 In addition, it may be worthwhile to comment on how the changes in margin due to the sensitivity parameters may be combined. For instance, a 10 %
decrease in wall thickness and the high growth rate can decrease the margin by more than 50% - Are these cases probable?
Add explanation NRC 7-15 24 Why was the 21:1 aspect ratio only used for the large bending moment cases?
Add explanation NRC 7-15 35 Same comment as 7-10 line 28 Add explanation NRC 7-16 7
Earlier it was stated that only two of the wolf creek surge flaws were enveloped with the 21:1 surge nozzle flaw. How are the three wolf creek flaws applied in the crack growth analyses?
Add explanation NRC 7-16 17 The wolf creek flaws may not have been growing rapidly in the depth direction, but may have been growing rapidly in the length direction, as indicated by the Phase 1 results.
Please modify sentence NRC 8
The same comments given above apply to the conclusion section NRC 8-2 31-33 Cases CS1b (SMF = 1.03 at TW leakage) and CS2b (3 days to SMF=1.2) do not support the statement that the results met the evaluation criteria with additional margin Revise to reflect actual results.
NRC A1-7 Table 2-1 There are some numbers in Table 2-1 of the probabilistic study in Appendix A that we believe are not correct. Calvert 2 had indications in the CL drain and the HL drain. Table 2-1has the CL drain listed as a circ indication, when it was actually an axial indication, as documented in LER 2005-001-00. The HL drain had 2 axial flaws attributed to PWSCC and a circumferential that was attributed to original construction once the original radiographs were digitized (this fact was not listed in the LER, but we have first hand information on this item). Also, the depths and lengths were not measured for these flaws, as the procedure used was not qualified for length or depth measurement for this size nozzle. It is unclear where the numbers for the flaw information come from.
We also checked Calvert 1 in Table 2-1 with the information Constellation sent to us in their flaw evaluation. The HL Drain thickness is 0.54, not 0.375; the surge nozzle thickness is 1.6", not 1.3, and the relief indication length was 0.6" not 0.000".
This is a high number of inaccuracies in 5 indications. It calls into question the accuracy of the remaining information in this table. We recommend that all the data in the table be verified for accuracy and a portion of section 2.1 be devoted to data accuracy and verification.
NRC General Supplementary analyses need to be provided for NRC review.
Provide supplementary analyses as soon as possible to expedite NRC's review.
NRC References References are needed to expedite NRC's review.
Provide references as soon as possible to expedite NRC's review.