ML24261B913
| ML24261B913 | |
| Person / Time | |
|---|---|
| Issue date: | 09/20/2024 |
| From: | David Dijamco, John Tsao Licensing Processes Branch |
| To: | Advisory Committee on Reactor Safeguards, Licensing Processes Branch |
| References | |
| Download: ML24261B913 (1) | |
Text
NRC Staff Evaluation of PWROG-18068-NP, Revision 1 Use of Direct Fracture Toughness for Evaluation of RPV Integrity
David Dijamco John Tsao Office of Nuclear Reactor Regulation Division of New and Renewed Licenses Vessels and Internals Branch
Presentation to the ACRS Fuels, Materials, & Structures Subcommittee September 20, 2024 Content
- Staff Evaluation - key aspects we were looking for
- T0 testing is performed in accordance with an accepted standard
- Uncertainties in T0 are adequately addressed
- Adequacy of the proposed method for embrittlement prediction (i.e., use of ASTM E900)
- When data from material test reactors are used, flux effect is adequately addressed
- Regulatory exemptions for using methodology are clear
- Data Adjustments and Uncertainties - key requests for additional information (RAIs)
- T30-T 0 correlation
- Material variability
- Observations
- How much margin/relaxation in RPV integrity is gained?
- How does TR methodology fit in the broader sense with respect to use of T0?
- Regulatory implications
- Conclusion
- Why we find the methodology acceptable
2 Approved Use of T0 is Not New
- ASME Code,Section XI, G-2110
- ASME Code,Section III, NB-2331 Materials for Vessels
- ASME Code Cases N-830, N-629
- Kewaunee circumferential RPV weld (ML011210180)
- Approved topical reports BAW-2308, Revisions 1 and 2, for Linde 80 welds (ML052070408 and ML080770349)
3 Staff Evaluation
T0 Testing
- T0 must be determined in accordance with ASTM E1921
- Latest version incorporated by reference in the regulations is 2017 (via IBR of ASME Code,Section XI)
- 2020 version is used in the methodology because it has procedures for screening and evaluating inhomogeneous datasets (adds conservatism in the T0 if dataset is inhomogeneous)
- 2 key aspects
- 1) Test specimen configuration
- Specifies use of standard-shaped or disk-shaped compact tension, C(T) or DC(T), and single-edge notch bars, SE(B), specimens.
- Specifies additional 18°F if SE(B) specimens are used
- 2) Has provisions for inhomogeneous screening and procedures for determining additional value to T 0 to address inhomogeneous datasets.
4 Staff Evaluation
T0 Testing (continued)
- The TR methodology provides basis for use of mini-C(T) specimens.
- The mini-C(T) geometry is C(T) fracture specimen geometry scaled to a smaller size such that specimens can be machined from a broken Charpy half and is designed to be compliant with ASTM E1921.
- Up to four of these mini-C(T) specimens (per half Charpy specimen) can be machined from a broken Charpy specimen.
- Japanese lab sponsored an international round robin with seven test labs participating to compare T0 values using mini-C(T) and larger C(T) specimens.
- Dataset comprised of RPV base metals and welds.
- Constraint condition difference between mini-C(T) and larger C(T ) was shown to be negligibly small.
- The results show no significant difference between the T0 values determined by the different labs, nor is there a significant difference regarding T0 values between the mini-C(T) and the larger-size C(T) specimens (avg. difference = 0.7 °C, standard dev. of difference = 7.2° C)
- If SE(B), i.e., 3-point Charpy bend, specimens are used, adding a bias of 18°F to the T 0 test value is required.
5 Staff Evaluation
T0 Testing (continued)
- Datasets are screened for inhomogeneity.
- If screening fails, i.e., the dataset is potentially inhomogeneous and will be evaluated in accordance with Appendix X5 of ASTM E1921.
- If dataset determined to be inhomogeneous, T0 value is increased to a larger value, T0IN, because the dataset is inhomogeneous.
- Thus, staff finds T0 testing used in the methodology acceptable.
6 Staff Evaluation
Uncertainties in T0
Sources of uncertainty in using the methodology
- T0 uncertainty
- Adjustment uncertainty (next set of slides on ASTM E900)
- Uncertainty due to radiation exposure parameters, fluence and temperature
- Uncertainties due to material variability and T 0-T 30 correlation (John)
7 Staff Evaluation
Uncertainties in T0 (continued)
Margin term for irradiated data Margin = 2 (E19212 + adjustment2 + tempspecimen2 +
fluencespecimen2 + tempRPV2 + fluenceRPV2)
Margin term for unirradiated data Margin = [(k1S)2 + (2adjustment)2 + (2tempRPV)2 +
(2fluenceRPV)2]
8 Staff Evaluation
Uncertainties in T0 (continued)
Uncertainty Source Margin term Margin term Irradiated Unirradiated Explanation T0 specimen testing E1921 n/a As defined in ASTM E1921
Unirradiated T0 data n/a k1S NUREG-1475 95/95 statistical bound parameters Adjustment term adjustment adjustment Equation 12 of TR Specimen temperature tempspecimen n/a Specimen fluence fluencespecimen n/a RPV temperature tempRPV tempRPV RPV fluence fluenceRPV fluenceRPV
9 Staff Evaluation
Uncertainties in T0 (continued)
- Equation 12 of TR
- adjustment is the standard deviation of T 30 (i.e., RT NDT) of RPV as defined in ASTM E900 scaled to the ratio of the adjustment term to RPV T 30.
- Multiplier of 1.1 for base metal
- Minimum 9°C (based on statistics of embrittlement shift data of welds and base metals)
- Double-counts fluence and temperature uncertainties; there are already separate fluence and temperature margin terms, but since Equation 12 is based on E900, which is a function of fluence and temperature (among four other parameters), some fluence and temperature uncertainties are included in adjustment.
- Thus, staff finds that sources of uncertainties in T0 adequately addressed.
10 Staff Evaluation
Adequacy of ASTM E900 Embrittlement Trend Curve
- Improved accuracy of E900 compared to RG 1.99 Rev. 2, especially at high fluence
- Exposure parameters are fluence and temperature (compared to only fluence RG 1.99 Rev. 2); temperature parameter adds robustness
- Similar scatter in predicted RTNDT relative to measured RTNDT but more falling within +/-2 (next two slides)
- Documented in work by staff ( ML21314A228) referenced in proposed embrittlement rulemaking SECY (ML21314A215)
11 Staff Evaluation Adequacy of ASTM E900 Embrittlement Trend Curve (continued)
E900 compared to RG 1.99 Rev. 2, predicted RTNDT relative to measured RTNDT BASE METALS
E900
RG 1.99 Rev. 2
12 Staff Evaluation Adequacy of ASTM E900 Embrittlement Trend Curve (continued)
E900 compared to RG 1.99 Rev. 2, predicted RTNDT relative to measured RTNDT WELD METALS
E900
RG 1.99 Rev. 2
13 Staff Evaluation
Adequacy of ASTM E900 Embrittlement Trend Curve (continued)
- In our evaluation, we focused how margin (in the use of E900) is addressed (slide 10)
- As previously discussed, minimum of 9°C and double-counting of uncertainties by the adjustment term and separate terms for fluence and temperature uncertainties for both the test specimens and RPV of interest.
- Thus, the staff finds the use of ASTM E900 trend curve acceptable for the methodology.
14 Staff Evaluation Flux Effect from Material Test Reactors When material test reactor (MTR) irradiations are used, such as in an MTR campaign, the MTR irradiated data must be validated against PWR irradiation data.
- The MTR irradiated data must be:
- from same Cu grouping as the PWR validation material (since flux effect on embrittlement is dependent on Cu level)
- from same heat as the PWR validation material
- within +/-50% of PWR fluence
- Compare the adjusted MTR irradiation T0 with the PWR validation material T0 (Equation 9), and if former is a larger value, no increase is necessary.
- If former is a smaller value, increase MTR irradiation T0 per Equation 10.
- The staff finds the above a reasonable approach for handling the potential flux effect from MTR irradiations.
15 Staff Evaluation Regulatory Exemptions
- As stated in draft SE, exemption is needed because the definition of RT PTS in methodology is different than the definition in 10 CFR 50.61.
- Appendix G to 10 CFR 50
- As stated in draft SE, exemption is needed because Section IV. Fracture Toughness Requirements of this regulation states:
For the reactor vessel beltline materials, including welds, plates and forgings, the values of RTNDT and Charpy upper-shelf energy must account for the effects of neutron radiation, including the results of the surveillance program of Appendix H of this part.
- Appendix H to 10 CFR 50
- As stated in draft SE, exemption is needed when Charpy specimens in a surveillance program are modified and tested for T0 in a three-point bending configuration (or another configuration).
- As stated in draft SE, n o exemption is needed when small fracture toughness specimens from broken Charpy specimens or untested heat-affected zone specimens are used for T 0 testing.
16 Data Management
- How to manage data to ensure data integrity in prediction
- Prediction of RPV Embrittlement involves various parameters
- Unirradiated and irradiated T0
- T30 data vs. T0 data
- Data from Test specimens vs. RPV materials (same heat vs different heat)
- Fluence and flux in pressurized water reactor vs material test reactor
- Uncertainties in measurements
- Uncertainty in prediction method
17 ASTM Standards
- E900- 15
- provides guidance on predicting transition temperature shift (i.e., embrittlement) of RPV material based on surveillance data.
- Contains statistical methods to evaluate uncertainty in shift prediction.
- E1921-20
- provide guidance on the test method using compact or bend bar specimens to determine T0 in the transition range.
- contains master curve.
- Provides statistical analysis of test data,
- consideration of margin adjustments, material inhomogeneity vs.
homogeneity, evaluation of uncertainty.
- Staff has not approved E900- 15 and E1921-20. Approving the TR does not imply or infer generic approval of these standards.
18 Adjustment Term in TR Methodology
- TR equations # 1, 2, and 3 include an adjustment term to adjust test specimen data to irradiation level of RPV material of interest; this adjustment term depends on :
- Chemistry
- Temperature
- Fluence
- Correlation between T30 and T0 ; Will focus on T30 and T0 correlation
19 Adjustment Term - Correlation between T30 and T0
- Background
- Correlation is needed to address the difference between T30 and T0.
- No industry accepted embrittlement trend curve model is based on T0
- TR uses T30 for embrittlement shift because ASTM E900-15 data are based on T 30 data.
- Use correlation between T30 and T0 to justify the use of T 30 data in the adjustment term.
- Staff found that an adjustment for model uncertainty (uncertainty in correlation between T30 and T0) was not included in the methodology
- TR Methodology
- Measured difference exists between the embrittlement shift T30 and T0
- Difference in embrittlement shift T30 and T0 data are correlated and adjusted.
20 Adjustments - Correlation between T30 and T0
- TR performed statistical analysis of embrittlement shift between T30 and T0 data for base metal and welds.
- TR method bounds greater than 95% of the measured T0 values and fracture toughness values.
- TR derives a linear correlation slope of 1.1 for base metal and 1.0 for welds T30 and T0 data which is applied in the margin term.
- TR methodology includes 9°C (16°F) as part of adjustment uncertainty (adjustment) in the margin term.
- Staff Evaluation
- Staff finds acceptable that no model uncertainty associated with correlation of T30 and T0 data needs to be included because
- statistical analysis of correlation of T 30 and T0 data demonstrates linear correlation with a slope of 1.1 for base metal and 1.0 for welds.
- TR includes 9° C (16°F) as part of adjustment uncertainty (adjustment) in the margin term.
21 Margin in Existing Regulation
- A margin term is used to ensure that the evaluation is conservative.
- The margin term in 10 CFR 50.61 and Reg Guide 1.99 Rev 2
- RTPTS = RTNDT(U) + RTNDT + Margin; U =unirradiated
- Adjusted Reference Temperature (ART) = Initial RTNDT + RTNDT + Margin
- Margin = 2 I 2 + 2 where I is the standard deviation of the initial (or unirradiated) RT NDT and is the standard deviation of shift RTNDT
- If a measured data for initial RTNDT is used, use the precision of the test method. If a generic mean value is used, I is the standard deviation from the set of data used to obtain the mean.
- The value of to be used is 28 0F for welds and 17 0F for based metal. The value of need not exceed the on-half of RTNDT
- If surveillance data are used, may be reduced by half.
- As a comparison, TR methodology for RTPTS equation includes an adjustment term and a margin term
22 Margin Term in TR Methodology
- TR equations # 1, 2, and 3 include a margin term to account for uncertainty which includes (as previously discussed):
- Testing Uncertainty, E1921
- Adjustment Uncertainty, adjustment
- Temperature Uncertainty, tempspecimen or tempRPV
- Neutron Fluence Uncertainty, fluencespecimen or fluenceRPV
- Material Variability Uncertainty not explicitlyincluded in margin term; Will focus on material variability
23 Material Variability
- Background -
- RPV material embrittlement relies on prediction model which is based on shift (embrittlement) measurement of materials
- Empirical Embrittlement Trend Curve in ASTM E900-15 reflects uncertainties of material variation and chemistry uncertainties in the prediction.
- Chemistry variation and initial fracture toughness variation need to be considered
- Material variability is addressed in ASTM E1921-20 which describes homogeneity screening procedures to detect if dataset can be representative of macroscopically inhomogeneous material.
- ASTM E1921-20 prescribes evaluation guidance for datasets that failed homogeneity screening criterion.
- Inhomogeneity in fracture toughness could be caused by the initial material properties and embrittlement effects.
- Measurement of T0 based fracture toughness reduces uncertainties associated with RTNDT correlated/based fracture toughness.
24 Material Variability
- TR Methodology
- TR does not change safety factors in 10 CFR 50, Appendix G nor 10 CFR 50.61
- TR does not explicitly include material variability uncertainty as part of the margin term aside from homogeneity screening in E1921-20.
- TR considers that (a) measurement of direct fracture toughness reduces uncertainty associated with the correlation of RTNDT to fracture toughness, and (b) measurement of irradiated fracture toughness reduces uncertainty associated with embrittlement prediction.
- Staff Evaluation
- Initially, staff had concerns that material variability uncertainty was not considered in the margin term. Staff requested for additional information.
- Based on the PWROGs response, staff finds that (a) the TR methodology includes 9 °C (16
°F) as part of adjustment uncertainty ( adjustment) in the margin term, (b) the margins currently required by the SAME Code,Section III are sufficient to account for material variability uncertainty, (c) evaluating embrittlement at the 1/4T location of the RPV shell provides structural margin as compared to the RPV shell surface because driving force at the 1/4 T location is higher than at the surface, and (d) the TR methodology uses large amount of T0 data which reduces uncertainty as compared to small datasets.
- Based on the adjustment uncertainty and other associated margins, the staff finds that the TR methodology addresses material variability uncertainty satisfactorily.
25 Regulatory Implication
- The proposed TR methodology is not a new regulatory requirement; it is an option that licensee could use.
- Licensee could use the TR methodology if its PTS value will exceed the screening criteria of 270 °F for axial welds, and forgings and 300 °F for circumferential welds at the end of license.
- Licensee could use the TR methodology to develop its Pressure Temperature limit curves so that it could operate its reactor efficiently.
- The TR specifies that licensees who plan to use the proposed methodology needs to submit exemptions from the appropriate regulations.
- Rulemaking
- Rulemaking would eliminate the need for exemptions.
- Staff did not consider rulemaking during its review.
- Staff would like to review an actual licensee submittal to determine the feasibility of the methodology; achieve regulatory stability prior to exercise rulemaking.
- Staff would not rule out rulemaking in the near future.
26 Observations
- Additional staff considerations on the proposed methodology
- The proposed methodology is consistent with NRC approved precedent.
- The proposed method is adequate and ensures the integrity of the RPV material.
- The TR uses industry standards ASTM E1921-20 and E900- 15 consistently although staff has not officially (i.e., not generically) approved these two standards.
27 Observations - continued
- The proposed methodology provides fracture toughness that is closer to actual transition temperature fracture toughness of reactor vessel shell materials than fracture toughness determined by the existing method.
- The TR methodology can be used by plants whose RT PTS values are approaching the PTS screening criterion.
- The proposed methodology may assist PWRs to generate P -T limit curves that are not as restrictive as P-T limit curves using the method in the current regulations.
28 Observations - continued
- Margins may be gained back from the use of the TR methodology.
- Based on the existing method, TR obtained a RT PTS of 253.8 °F, taken from Crystal River 60- year license renewal application.
- This is a 77.6 °F margin gain. This is only an example; plant-specific data may result in different margin gain.
- This specific result shows that the RPV embrittlement does not appear to be as severe as it is predicted using the existing method.
29 Conclusion
- The proposed TR methodology using irradiated or unirradiated T 0 data in lieu of current approach of using RT NDT for evaluating RPV structural integrity is acceptable because the TR method has provisions for determining adjustments and margins applied to T 0 data that ensure the final T 0 is adequate in maintaining RPV integrity.
30 BACKUP SLIDES ONLY
31 3 scenarios that calls for use of methodology
- You want to determine an irradiated T0 for use in PTS and PT limits for an RPV material of interest (MOI).
Typically, there is no irradiated T0 data for the RPV MOI, but there is irradiated T0 data elsewhere. What to do?
Three scenarios:
- Scenario 1: Have irradiated T0 data from specimens that are same heat as RPV MOI from other PWRs or;
- Scenario 2: Have irradiated T0 data from specimens that are same heat and same Cu grouping as RPV MOI from MTRs or;
- Scenario 3: Have unirradiated T0 data from material similar to RPV MOI (doesnt have to be same heat)
32 BACKUP Scenario 1:
Irradiated PWR Data
To determine T0 for RPV MOI if you have irradiated T 0 data from specimens that are same heat as RPV MOI from other PWRs:
For each specimen,
- Calculate Adjustment Term (Equation 6)
- Calculate Margin Term for irradiated data (Equation 11)
- Add both to T0 of each specimen dataset
- Compute weighted average to determine final T 0 for RPV MOI (more weight given to specimen closest to condition of RPV MOI)
33 BACKUP Scenario 2 Irradiated MTR Data To determine T0 for RPV MOI if you have irradiated T 0 data from specimens that are same heat as RPV MOI from MTRs:
Must first validate MTR data; for each MTR specimen,
- Must be in the same Cu grouping
- Fluence must be +/- 50% of PWR validation fluence
- Calculate Adjustment Term (Equation 6)
- Add to T0 of each MTR specimen validation dataset
- Compute weighted average to determine adjusted T 0 for the MTR specimens (more weight given to MTR specimen closest to condition of PWR validation material)
- If above weighted average adjusted MTR T0 is greater than PWR validation T0, MTR T0 is considered representative of or conservative relative to the RPV MOI.
- If not, increase MTR T0 such that MTR T0 used is representative of or conservative relative to the RPV MOI. 34 BACKUP Scenario 2 (continued)
Irradiated MTR Data After validation, for each MTR specimen,
- Calculate Adjustment Term (Equation 6)
- Calculate Margin Term for irradiated data (Equation 11)
- Add both to T0 of each MTR specimen
- Compute weighted average to determine final T 0 for RPV MOI (more weight given to MTR specimen closest to condition of RPV MOI)
35 BACKUP Scenario 3 Unirradiated Data
To determine T0 for RPV MOI if you have unirradiated T 0 data from specimens that are similar and that is similar to the RPV MOI:
- Must have at least 4 heats (doesnt have to be the same heat as RPV MOI)
- Calculate standard deviation (S) and one-sided 95% tolerance factor (k1) from NUREG-1475
- Is there irradiated T0 data from a heat in the generic grouping?
- If no, calculate average T0 value of the unirradiated data
- If yes, adjust generic unirradiated T0 value by adjustment term (Equ. 6) and unirradiated margin term (Equ. 5) the resulting adjusted generic value must bound 95% of the irradiated KJc1T data (if not bounded, cannot use unirradiated data)
- Calculate Adjustment Term (Equation 6)
- Calculate Margin Term for unirradiated data (Equation 5)
- Add both to generic unirradiated T0 to determine the final irradiatedT0
36