ML20300A551

TLR-RES_DE_CIB-2020-09 - RG-1.99r2 Update Favor Scoping Study

ML20300A551
Person / Time
Issue date:	10/26/2020
From:	Patrick Raynaud NRC/RES/DE/CIB
To:
Raynaud P
Shared Package
ML20300A564	List: ML20300A551 ML21126A326
References
TLR-RES/DE/CIB-2020-09, UNR NRR-2017-007
Download: ML20300A551 (140)

Text

Technical Letter Report TLR-RES/DE/CIB-2020-09 RG 1.99 Revision 2 Update FAVOR Scoping Study Date:

October 26, 2020 Prepared in support of the RG 1.99 Rev. 2 revision project by:

Patrick Raynaud Senior Materials Engineer Component Integrity Branch Division of Engineering Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washington, DC 20555-0001

DISCLAIMER This report was prepared as an account of work sponsored by an agency of the U.S. Government.

Neither the U.S. Government nor any agency thereof, nor any employee, makes any warranty, expressed or implied, or assumes any legal liability or responsibility for any third party's use, or the results of such use, of any information, apparatus, product, or process disclosed in this publication, or represents that its use by such third party complies with applicable law.

i

This report does not contain or imply legally binding requirements. Nor does this report establish or modify any regulatory guidance or positions of the U.S. Nuclear Regulatory Commission and is not binding on the Commission.

ii

Contents Executive Summary ....................................................................................................................................... 1 1 Introduction .......................................................................................................................................... 2 2 PWR Scoping Study ............................................................................................................................... 3 2.1 Vessel and Transient Characteristics ............................................................................................ 3 2.1.1 Vessel Characteristics and Embrittlement Map .................................................................... 3 2.1.2 Transient Definitions ............................................................................................................. 6 2.1.3 Other FAVOR Simulation Input Information ....................................................................... 14 2.2 RTNDT Shift Distribution................................................................................................................ 15 2.3 Results ......................................................................................................................................... 18 2.3.1 PT100-100 Cooldown .......................................................................................................... 18 2.3.2 PT100-50 Cooldown ............................................................................................................ 24 2.3.3 PT50-50 Cooldown .............................................................................................................. 30 2.3.4 C50P Cooldown ................................................................................................................... 36 2.3.5 Sensitivity Study on Actual PWR Cooldowns for Shallow Flaw ........................................... 41 3 BWR Scoping Study ............................................................................................................................. 44 3.1 Vessel and Transient Characteristics .......................................................................................... 44 3.1.1 Vessel Characteristics and Embrittlement Map .................................................................. 44 3.1.2 Transient Definitions ........................................................................................................... 46 3.1.3 Other FAVOR Simulation Input Information ....................................................................... 54 3.2 RTNDT Shift Distribution................................................................................................................ 55 3.3 Results ......................................................................................................................................... 58 3.3.1 PT100-100 Cooldown .......................................................................................................... 58 3.3.2 PT100-50 Cooldown ............................................................................................................ 65 3.3.3 PT50-50 Cooldown .............................................................................................................. 70 3.3.4 S50 Cooldown ..................................................................................................................... 76 3.3.5 LT40-40 Leak Test................................................................................................................ 79 3.3.6 LT40-100 Leak Test.............................................................................................................. 85 3.3.7 LTA Leak Test....................................................................................................................... 90 4 Summary ............................................................................................................................................. 95 5 References ........................................................................................................................................ 104 6 Appendix A: List of FAVOR Analyses ................................................................................................. 105 6.1 PWR Analyses ............................................................................................................................ 105 6.2 BWR Analyses............................................................................................................................ 112 iii

7 Appendix B: Analysis of Other PWR Transients ................................................................................ 118 7.1 Transient P7C1: Partial Cooldown ............................................................................................ 118 7.2 Transient P7C5: High-Pressure Cooldown ................................................................................ 121 7.3 Transient P7C6: Unusual Cooldown .......................................................................................... 124 7.4 Transient P7C8: Partial Cooldown Followed by Heatup ........................................................... 127 7.5 Transient P8C3: Cooldown Followed by Partial Heatup ........................................................... 130 7.6 Transient PBC1: Cooldown with Pressure Hold and Low Temperature Pressure Spike ........... 133 iv

Executive Summary This report documents the probabilistic fracture mechanics analyses performed between October 2019 and May 2020 in support of the RG 1.99 Rev.2 revision effort. The primary goal of the analyses was to evaluate the analytical impact of a proposed change in embrittlement correlation, from the one currently used in RG 1.99 Rev.2 to the one used in the ASTM E900-15 standard. The analyses were performed with the NRCs FAVOR v16.1 code.

The study specifically analyzed the change in conditional probability of crack growth initiation (CPI) and in conditional probability of vessel failure (CPF) as a function of the shift in embrittlement that might result from changing the embrittlement correlation in RG 1.99. The potential embrittlement shift deltas associated with switching from the RG 1.99, Rev. 2 to the E900-15 embrittlement correlation were predicted for the entire US fleet. The range of embrittlement shift deltas in this study were chosen to capture the variability of the entire fleet of PWRs and BWRs.

Several PWR and BWR transients deemed relevant for the purpose of this assessment were modeled assuming both a 1/4 thickness flaw depth (as required by ASME code) and a shallow inner-surface flaw penetrating through the stainless steel cladding and barely into the reactor pressure vessel base metal (ferritic carbon steel).

With the modeling assumptions used in this study, the transients resulting in the highest through-wall cracking frequencies (TWCF) were PWR cooldowns along the pressure-temperature (P-T) limit for 1/4 thickness flaws and for high values of embrittlement shift delta. However, the highest calculated TWCF still remained below 1x10-6 per year, which is the acceptance limit used in 10 CFR 50.61a.

For BWRs, the transients resulting in the highest TWCF were leak tests, with TWCF reaching close to 1x10-7 for an actual BWR leak test and TWCF reaching close to 1x10-8 for leak tests at the P-T limit.

However, once again, even the highest calculated TWCF still remained below 1x10-6 per year.

Realistic cooldown transients resulted in the lowest TWCF values, with most being zero, and the highest reaching around 1x10-10 per year for one actual PWR cooldown modeled.

1

1 Introduction Beginning in 2019, the NRC evaluated whether to update Regulatory Guide 1.99, Revision 2 [1]

(hereafter RG 1.99, Rev. 2), in particular with regards to the embrittlement correlations to be used in reactor pressure vessel safety analyses. Specifically, the NRC was proposing to replace the embrittlement correlation in RG 1.99, Rev. 2 with the embrittlement correlation from ASTM E900-15 [2]

(hereafter E900-15). This report documents the analyses performed in support of the RG 1.99 Rev.2 revision effort between October 2019 and May 2020.

The primary purpose of this study is to evaluate the analytical impact of a change in embrittlement correlation from the one currently in RG 1.99, Rev. 2 to the one from E900-15. Specifically, this study focuses on the change in conditional probability of crack initiation (CPI) and conditional probability of vessel failure (CPF) at 72 Effective Full Power Years (EFPY), as a result of the change in embrittlement correlation. This report first presents the PWR analysis results, and then the BWR analysis results.

The study was performed at 72 EFPY because this bounds the US Fleet 80 year subsequent license renewal periods. The change in CPI and CPF as a result of the change in embrittlement correlation was studied for a postulated high embrittlement PWR and for a postulated high embrittlement BWR (high embrittlement plants are those with a high Adjusted Reference Temperature [ART]). Studies were performed for both a quarter-thickness ID surface flaw (1/4 T flaw or 0.25 T flaw) and for the shallowest ID surface flaw whose crack tip penetrated through the cladding into the base metal (0.03 T for the PWR case and 0.04 T for the BWR case). These flaws were chosen based on NRC regulatory guidance and previous results concerning CPI and CPF sensitivity to flaw types [3].

The RTNDT shifts associated with switching from the RG 1.99, Rev. 2 to the E900-15 embrittlement correlation were predicted using the embrittlement trend correlations. The values of the shift used in this study were chosen to capture the variability of the entire fleet of PWRs and BWRs. Calculations were performed for shifts of -40°F, 0°F (baseline), and then the 50th, 75th, 95th, and 99th percentiles of the predicted RTNDT shift populations for the PWR and BWR fleets.

Finally, for each combination of reactor type, flaw depth, and RTNDT shift, a variety of transients were studied, ranging from licensed maximum cooldown transients and leak tests to more realistic transients.

The more realistic transients were studied to better understand the actual risk of vessel failure when considering actual practice in the US fleet. The FAVOR v16.1 code [4] [5] was used in all analyses.

2

2 PWR Scoping Study 2.1 Vessel and Transient Characteristics 2.1.1 Vessel Characteristics and Embrittlement Map The PWR vessel modeled had the following characteristics:

• Inner radius 86 inches

• Cladding Thickness 0.25 inches

• Base metal thickness 8.5 inches

• Total wall thickness 8.75 inches The material properties used in the FAVOR model are given in Table 2-1 and Table 2-2. The stress free temperature was assumed to be equal to 488°F. Pressure loading was applied since the calculations presented herein are for an ID surface flaw. Finally, weld residual stress (WRS) was included in calculations for axial and circumferential welds. These modeling parameters were chosen to match the analysis parameters historically used by NRC for regulatory analyses (e.g. 10 CFR 50.61a technical bases),

and are the ones recommended in the FAVOR theory manual.

Table 2-1: PWR Base Metal Properties Temperature Thermal Specific Heat Thermal Temperature Young's Modulus Conductivity Expansion Coefficient

°F BTU/(hr.ft.°F) BTU/(lbm.°F) 1/°F °F ksi 70 24.8 0.1052 70 29200 100 25 0.1072 0.00000706 200 28500 150 25.1 0.1101 0.00000716 300 28000 200 25.2 0.1135 0.00000725 400 27400 250 25.2 0.1166 0.00000734 500 27000 300 25.1 0.1194 0.00000743 600 26400 350 25 0.1223 0.0000075 700 25300 400 25.1 0.1267 0.00000758 800 23900 450 24.6 0.1277 0.00000763 Temperature Poisson's Ratio 500 24.3 0.1304 0.0000077 °F 550 24 0.1326 0.00000777 0 0.3 600 23.7 0.135 0.00000783 1000 0.3 650 23.4 0.1375 0.0000079 700 23 0.1404 0.00000794 750 22.6 0.1435 0.000008 800 22.2 0.1474 0.00000805 3

Table 2-2: PWR Cladding Properties Temperature Thermal Specific Heat Thermal Temperature Young's Modulus Conductivity Expansion Coefficient

°F BTU/(hr.ft.°F) BTU/(lbm.°F) 1/°F °F ksi 70 8.1 0.1158 68 22045.7 100 8.4 0.1185 0.00000855 302 20160.2 150 8.6 0.1196 0.00000867 482 18419.8 200 8.8 0.1208 0.00000879 Temperature Poisson's Ratio 250 9.1 0.1232 0.0000089 °F 300 9.4 0.1256 0.000009 0 0.3 350 9.6 0.1258 0.0000091 1000 0.3 400 9.9 0.1281 0.00000919 450 10.1 0.1291 0.00000928 500 10.4 0.1305 0.00000937 550 10.6 0.1306 0.00000945 600 10.9 0.1327 0.00000953 650 11.1 0.1335 0.00000961 700 11.4 0.1348 0.00000969 750 11.6 0.1356 0.00000976 800 11.9 0.1367 0.00000982 The PWR vessel rollout diagram and corresponding embrittlement information from the RVID database are provided in Figure 2-1 and Table 2-3 respectively.

Figure 2-1: PWR Vessel Rollout Diagram 4

Table 2-3: PWR Vessel Embrittlement Map Information and Corresponding FAVOR Input max neutron 1/4T RTNDT @

ID RTndt (u)

Major fluence @72 72 EFPY(2)

Description Heat ID Plate or Weld ID %Cu %Ni %P [Initial Region EFPY(1) w/o Margin RTndt]

1019 n/cm2 o F

1 Axial Weld W5214 LOWER SHELL AXIAL WELDS 3-112A/C 0.213 1.010 0.019 -56.000 2.96 210.7 2 Axial Weld W5214 LOWER SHELL AXIAL WELDS 3-112A/C 0.213 1.010 0.019 -56.000 3.81 225.9 3 Axial Weld W5214 LOWER SHELL AXIAL WELDS 3-112A/C 0.213 1.010 0.019 -56.000 3.81 225.9 4 Axial Weld W5214 INTERMEDIATE SHELL AXIAL WELDS 2-112 A/C 0.213 1.010 0.019 -56.000 3.86 226.6 5 Axial Weld W5214 INTERMEDIATE SHELL AXIAL WELDS 2-112 A/C 0.213 1.010 0.019 -56.000 2.99 211.3 6 Axial Weld W5214 INTERMEDIATE SHELL AXIAL WELDS 2-112 A/C 0.213 1.010 0.019 -56.000 3.86 226.6 7 CircWeld 27204 CIRC. WELD 9-112 0.203 1.018 0.013 -56.000 4.85 234.3 8 Plate C-1308A D-3804-1 0.190 0.480 0.016 0.000 4.85 164.8 9 Plate C-1308B D-3804-2 0.190 0.500 0.015 -30.000 4.85 137.7 10 Plate B-5294 D-3804-3 0.120 0.550 0.010 -25.000 4.85 79.9 11 Plate C-1279 D-3803-1 0.240 0.510 0.009 -5.000 4.90 198.9 12 Plate A-0313 D-3803-2 0.240 0.520 0.010 -30.000 4.90 175.7 13 Plate C-1279 D-3803-3 0.240 0.500 0.011 -5.000 4.90 197.0 The resulting FAVOR PWR embrittlement map input was as follows:

• ========================================================================================================*

• 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

• 1 8 9 1 2.9625 0.213 1.01 0.019 1 1 0 -56 17 11 0.958 60.71 516 1 0.000 0.000 2 9 10 2 3.8089 0.213 1.01 0.019 1 1 0 -56 17 11 0.958 60.71 516 1 0.000 0.000 3 10 8 3 3.8089 0.213 1.01 0.019 1 1 0 -56 17 11 0.958 60.71 516 1 0.000 0.000 4 11 12 4 3.8554 0.213 1.01 0.019 1 1 0 -56 17 11 1.158 93.98 799 1 0.000 0.000 5 12 13 5 2.9892 0.213 1.01 0.019 1 1 0 -56 17 11 1.158 93.98 799 1 0.000 0.000 6 13 11 6 3.8554 0.213 1.01 0.019 1 1 0 -56 17 11 1.158 93.98 799 1 0.000 0.000 7 8 11 7 4.8502 0.203 1.018 0.013 1 2 0 -56 17 11 360 1.31 4833 2 0.000 0.000 8 8 8 8 4.8456 0.19 0.48 0.016 1 0 0 0 17 21 119.0 60.71 0.000 0 0.000 0.000 9 9 9 9 4.8456 0.19 0.5 0.015 1 0 0 -30 17 21 119.0 60.71 0.000 0 0.000 0.000 10 10 10 10 4.8456 0.12 0.55 0.01 1 0 0 -25 17 21 119.0 60.71 0.000 0 0.000 0.000 11 11 11 11 4.9015 0.24 0.51 0.009 1 0 0 -5 17 21 118.8 93.98 0.000 0 0.000 0.000 12 12 12 12 4.9015 0.24 0.52 0.01 1 0 0 -30 17 21 118.8 93.98 0.000 0 0.000 0.000 13 13 13 13 4.9015 0.24 0.5 0.011 1 0 0 -5 17 21 118.8 93.98 0.000 0 0.000 0.000 Of note, RG 1.99, Rev. 2 was used to calculate the RTNDT at the 0.25 T location so as to calculate the P-T limit for a 100°F/hour cooldown (last column in Table 2-3). The maximum RTNDT at the 0.25 T location at 72 EFPY was determined to be 234.3°F, which is the value that was used for the P-T limit calculation.

5

2.1.2 Transient Definitions 2.1.2.1 PT100-100: 100°F/Hour Cooldown at Maximum Allowed Pressure for 100°F/hour Cooldown In order to determine the allowable pressure for the P-T limit cooldown at a rate of 100°F/hour, ASME Section XI, Appendix G, Paragraph G-2215 [6] was used. The risk-informed alternative was NOT used.

Assuming the limiting case of an axial flaw and a wall thickness between 4 and 12 inches, the allowable pressure is given by:

=

2 with

= 33.2 + 20.734 [0.02( )]

and

= 0.926 In order to determine the applied stress intensity factor due to thermal stresses (KIt), FAVOR was used to simulate a 100°F/hour cooldown from 550°F to 70°F with no applied pressure, and using a stress free temperature of 550°F (such that the initial thermal stresses are zero). The resulting K calculated by FAVOR for a 0.25 T flaw and an aspect ratio of 6 was used to determine the allowable pressure using the equation above. The value of crack tip RTNDT used was 234.3°F, as described in 2.1.1.

The resulting P-T limit transient for a 100°F/hour cooldown from 550°F to 70°F is shown in Figure 2-2.

This transient is referred to as the PT100-100 transient in this document, and its frequency of occurrence is 6x10-6 events per year (based on the BTP 5-3 closure memo assessment [7]).

Figure 2-2: PWR PT100-100 Cooldown Transient Characteristics 6

2.1.2.2 PT100-50: 50°F/Hour Cooldown at Maximum Allowed Pressure for 100°F/hour Cooldown The same P-T limit pressure history based on a 100°F/hour cooldown (see section 2.1.2.1) was used along with a 50°F/hour cooldown temperature history. The resulting P-T limit transient for a 50°F/hour cooldown from 550°F to 70°F is shown in Figure 2-3 (same pressure history as for the PT100-100 cooldown, but with a slower cooldown of 50°F/hour). This transient is referred to as the PT100-50 transient in this document, and its frequency of occurrence is estimated at 6x10-6 events per year (based on the BTP 5-3 closure memo assessment [7]).

Figure 2-3: PWR PT100-50 Cooldown Transient Characteristics 7

2.1.2.3 PT50-50: 50°F/Hour Cooldown at Maximum Allowed Pressure for 50°F/hour Cooldown In order to determine the allowable pressure for the P-T limit cooldown at a rate of 50°F/hour, ASME Section XI, Appendix G, Paragraph G-2215 [6] was used. The risk-informed alternative was NOT used.

Assuming the limiting case of an axial flaw and a wall thickness between 4 and 12 inches, the allowable pressure is given by:

=

2 with

= 33.2 + 20.734 [0.02( )]

and

= 0.926 In order to determine the applied stress intensity factor due to thermal stresses (KIt), FAVOR was used to simulate a 50°F/hour cooldown from 550°F to 70°F with no applied pressure, and using a stress free temperature of 550°F (such that the initial thermal stresses are zero). The resulting K calculated by FAVOR for a 0.25 T flaw and an aspect ratio of 6 was used to determine the allowable pressure using the equation above. The value of crack tip RTNDT used was 234.3°F, as described in 2.1.1.

The resulting P-T limit transient for a 50°F/hour cooldown from 550°F to 70°F is shown in Figure 2-4.

This transient is referred to as the PT50-50 transient in this document, and its frequency of occurrence is estimated at 6x10-6 events per year (based on the BTP 5-3 closure memo assessment [7]).

Figure 2-4: PWR PT50-50 Cooldown Transient Characteristics 8

2.1.2.4 C50P: 50th Percentile (Median) of Actual Cooldown Transients A combined 42 actual normal operation plant cooldown histories were obtained from 17 different PWRs. These cooldown histories were analyzed to determine the pressure and temperature histories corresponding to various percentiles of the pressure and temperature history distributions based on the population considered.

It is important to note that no information was available regarding how representative these transients are of actual PWR fleet operations. These transients were simply those that were made available to NRC by Westinghouse in 2013 through an informal exchange of data, and they may or may not be statistically representative of normal PWR cooldowns. Nonetheless, they all appeared to represent successfully terminated cooldowns, and were treated as such.

Figure 2-5 and Figure 2-7 respectively show the actual 42 temperature and pressure histories obtained for a total of 17 different PWRs. Figure 2-6 and Figure 2-8 respectively show the temperature and pressure histories corresponding to various percentiles of the distribution of the temperature and pressure histories. Figure 2-9 and Figure 2-10 show the P-T curves for the actual plant cooldowns as well as the P-T histories corresponding to various percentiles of the population studied. The values of pressure and temperature corresponding to the various percentiles were calculated at each time step based on the distribution of the data for the 42 transients analyzed. The 50th percentile (median) histories are highlighted in bold red lines. Importantly, the time is given here in hours instead of minutes.

When comparing the actual cooldowns with the three P-T limit cooldowns studied (PT100-100, PT100-50, and PT50-50), the actual cooldown transients generally occur over a much longer period of time.

Looking at the median histories (50th percentile), the cooldown rate is roughly ~50°F/hour from 550°F to 350°F, then ~20°F/hour from 350°F to 200°F, then ~2.5°F/hour from 200°F to 100°F. The corresponding pressure reduction rate is roughly 400psig/hour to 500psig/hour for the first 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />, such that in 4 to 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />, the pressure is reduced to about 350psig, then the pressure is maintained at ~350psig until the temperature reaches ~100°F, then rapidly dropped to zero (0psig).

The analysis results for the 50th percentile of actual cooldown transients are presented in Section2.3.4.

This transient is referred to as the 50th percentile cooldown or C50P transient in this document, and its frequency of occurrence is estimated at 1.0 events per year (based on the BTP 5-3 closure memo assessment [7]).

9

Figure 2-5: Temperature histories for 42 PWR cooldown transients Figure 2-6: Temperature histories corresponding to various percentiles of the population of PWR cooldowns considered 10

Figure 2-7: Pressure histories for 42 PWR cooldown transients Figure 2-8: Pressure histories corresponding to various percentiles of the population of PWR cooldowns considered 11

Figure 2-9: P-T histories for 42 PWR cooldown transients Figure 2-10: P-T histories corresponding to various percentiles of the population of PWR cooldowns considered 12

2.1.2.5 PWR Transient PT Curves Figure 2-11 shows the PT curves for the 4 cooldown transients modeled for the PWR RTNDT shift sensitivity studies. Overall, the PT100-100 and PT50-50 transients are similar on a PT curve, and one would thus expect relatively similar trends for stress intensity factor histories, as can be seen by comparing Figure 2-14 with Figure 2-25 and Figure 2-17 with Figure 2-28. Furthermore, based on the relative positions of the PT curves, one would expect that for a 1/4 T flaw (which is the postulated flaw for the creation of PT curves), the CPI and CPF values for the transients shown would be ranked as follows from highest to lowest: PT50-50, PT100-100, PT100-50, and finally C50P. This is indeed generally the trend for 1/4 T flaws, as described in Section 2.3. However, for shallow flaws, the trends are different, and the PT curves are not a good indication of the relative values of CPI and CPF for the cooldown transients.

Figure 2-11: PT curves for the 4 PWR cooldown transients modeled for the RTNDT shift sensitivity studies 13

2.1.3 Other FAVOR Simulation Input Information In addition to the vessel and transient information, the following FAVOR simulation inputs were used:

Simulation Parameters Irradiation shift correlation choice Regulatory Guide 1.99 Rev. 2 Normal operating coolant temperature 550°F Plant operating time 72 EFPY Ductile tearing model in IGA submodel off Ductile tearing initiation checks and reporting off Additional reports on CPI/CPF for major beltline regions off Flow stress for net-section plastic collapse failure model 80 ksi Maximum value for KIc or KIa 200 ksiin Crack arrest model High-constraint KIa (CCA specimens)

Resampling when advancing into new weld layer on Normalized crack depth corresponding to vessel failure 0.9 Fluence Sampling Multiplier for standard deviation of mean local fluence 0.118 Multiplier for standard deviation of local fluence 0.056 Weld Chemistry Sampling Standard deviation for Copper sampling in welds 0.167 wt%

Standard deviation for Nickel sampling in welds 0.162 wt%

Standard deviation for Phosphorus sampling in welds 0.0013 wt%

Plate Chemistry Sampling Standard deviation for Copper sampling in plates 0.0073 wt%

Standard deviation for Nickel sampling in plates 0.0244 wt%

Standard deviation for Phosphorus sampling in plates 0.0013 wt%

To produce FAVOR runs corresponding to a single ID surface flaw of depth 0.25 T and aspect ratio 6, the VFLAW input format was used to specify the flaw population as follows:

• The surface flaw density was set to result in a single flaw of depth 0.25 T and aspect ratio 6, with zeros everywhere else

• The embedded weld flaw density was set to zero for all flaw depths and aspect ratios

• The embedded plate flaw density was set to zero for all flaw depths and aspect ratios To produce FAVOR runs corresponding to a single ID surface flaw of depth 0.03 T and aspect ratio 6, the VFLAW input format was used to specify the flaw population as follows:

• The surface flaw density was set to result in a single flaw of depth 0.03 T and aspect ratio 6, with zeros everywhere else

• The embedded weld flaw density was set to zero for all flaw depths and aspect ratios

• The embedded plate flaw density was set to zero for all flaw depths and aspect ratios 14

2.2 RTNDT Shift Distribution In order to simulate the effects of changing the embrittlement correlation from RG 1.99, Rev. 2 to E90015 without changing the FAVOR code, the initial RTNDT value was changed via input to the FAVOR embrittlement map. The RTNDT shift amount was defined as follows:

1.992900 ( 900 ( 1.992 (

= 72 ) = 72 ) 72 )

Furthermore, the adjusted reference temperature (ART) as a result of using the E90015 embrittlement correlation is defined by:

0 900 (

(72 ) = + 72 )

leading to:

0 1.992 (

(72 ) = + + 72 )

SHIFT By shifting the initial RTNDT by RTNDT , FAVOR can thus simulate the effects of changing the embrittlement correlation from RG 1.99, Rev. 2 to E90015 while still using the RG 1.99, Rev. 2 embrittlement correlation, and thus without needing to change the FAVOR code.

SHIFT A compilation of RTNDT values for the entire PWR fleet, based on all the available embrittlement capsule data, was performed as part of the RG 1.99, Rev. 2 update effort. Figure 2-12 shows the distribution of values across the US PWR fleet considering this large set of embrittlement information, and Table 2-4 shows the values of corresponding to key percentiles (and vice versa).

Figure 2-12 and Table 2-4 show that when considering all available capsule data for PWRs at 72 EFPY, changing the embrittlement correlation from RG 1.99, Rev. 2 to E90015 would increase the adjusted reference temperature in most cases, since a 0°F shift (i.e. no shift) corresponds to the 23rd percentile for . Furthermore, the distribution is quite large, such that the median is 40°F and the upper 99 percentile is 193°F, which represents a very large shift.

th 15

Figure 2-12: Probability Distribution of the PWR RTNDT Shift as a Result of the Change from RG 1.99, Rev.

2 to E90015 Embrittlement Correlations (all capsule data)

Table 2-4: PWR RTNDT Shift as a Result of the Change from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations at Percentiles of Interest for this Study (all capsule data)

RTNDT Shift from RG 1.99, Rev. 2 to Percentiles E90015 0.01 -74 0.05 -49 0.07 -40 0.1 -30 0.23 0 0.25 2 0.5 40 0.75 78 0.9 112 0.95 131 0.99 193 It is important to point out that considering all available capsule does not provide the most accurate picture of the actual material embrittlement in the US fleet, because some capsules pulled early in the reactor life and for which data has been extrapolated can result in large errors in the actual embrittlement of reactor materials. A more accurate method may consist of filtering the capsule data to only keep the fluence capsule data for the longest actual irradiation time for each reactor in the fleet, as shown in Figure 2-13 and Table 2-5. This reduces extrapolation error and results in a far more accurate picture of the material embrittlement for the vessels in the US fleet.

16

Figure 2-13: Probability Distribution of the PWR RTNDT Shift as a Result of the Change from RG 1.99, Rev.

2 to E90015 Embrittlement Correlations (filtered capsule data)

Table 2-5: PWR RTNDT Shift as a Result of the Change from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations at Percentiles of Interest for this Study (filtered capsule data)

RTNDT Shift from RG 1.99, Rev. 2 to Percentiles E90015 0.01 -86 0.05 -64 0.16 -40 0.1 -53 0.25 -24 0.5 -8 0.6 0 0.75 17 0.9 31 0.95 37 0.99 52 Figure 2-5 and Table 2-12 show that when considering only the highest fluence capsule for each reactor, the distribution is much narrower and the 99th percentile is only a little over 50°F.

17

2.3 Results Probabilistic FAVOR runs were performed for the highlighted values from Table 2-4. The values chosen were -40°F, 0°F, followed by the 50 , 75 , 95 , and 99 percentile values, so as to cover the th th th th wide range of possible values (see Section 2.2). This section presents results for a PWR vessel containing a single flaw of aspect ratio equal to 6 and depths of 0.25 T and 0.03 T, for the transients described in section 2.1.2.

2.3.1 PT100-100 Cooldown 2.3.1.1 1/4-T Flaw Results The stress intensity factor history for a 0.25 T ID surface flaw subjected to the PT100-100 cooldown transient described in 2.1.2.1 is shown in Figure 2-14 for flaws with aspect ratios of 2, 6, 10 and infinity.

The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. It is important to note that all probabilistic analyses were performed with a flaw of aspect ratio equal to 6. Figure 2-14 clearly shows that K increases initially while the pressure is held constant and the temperature decreases linearly at a rate of 100°F/hour, reaching a maximum between 140 and 160 minutes. Then, at 164 minutes, the sudden pressure drop is accompanied by a sudden drop in K values. The rate of decrease in K is initially high but diminishes as time goes by. Finally, at 288 minutes, when the temperature reaches 70°F and remains constant, the rate of decrease in K values accelerates and creates an inflection point in the K versus time curves. After this inflection point, the rate of decrease in K values slowly diminishes until the end of the calculation.

Figure 2-14: Stress Intensity Factor for the PWR PT100-100 Cooldown for a 0.25 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity 18

A probabilistic calculation assuming a 0.25 T flaw of aspect ratio equal to 6 and the PT100-100 cooldown resulted in a CPI of 9.4E-07 and a CPF of 6.2E-08 using the RG 1.99, Rev. 2 embrittlement correlation.

This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient. Figure 2-15 shows the calculated CPI and CPF as a function of the number of RPV trials in FAVOR for this reference case, and one can see that the FAVOR calculations are statistically converged after about 80,000 trials. Similar plots were generated for all the runs performed in this study, so as to assess statistical convergence of the results.

Figure 2-15: CPI and CPF versus Number of Trials for a 0.25 T Flaw in a PWR with 0°F RTNDT Shift for the PT100-100 cooldown The same PT100-100 cooldown determined using the RG 1.99, Rev. 2 embrittlement correlation (shown in Figure 2-2) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. Of course, in actual safety analyses, the P-T limit would be adjusted as a function of the maximum ART in the vessel being analyzed, so a higher ART would reduce the allowed pressure, which In turn would result in lower CPI and CPF than presented here for cases where is greater than 0°F.

Table 2-6 and Figure 2-16 show the results of the probabilistic analyses for the 0.25 T flaw for the PT100-100 cooldown, as a function of . As expected, the CPI and CPF increase as increases. It should be noted that the analysis for = 40° did not statistically converge because of the very low calculated probabilities, but the CPI and CPF trend over the number of trials was beginning to stabilize, so the results are believed to be in the correct order of magnitude.

19

Table 2-6: Calculated CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a PWR Subjected to the PT100-100 Cooldown RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 6.0759E-09 NO 9.4160E-12 NO 5.6496E-17 0 9.4399E-07 YES 6.2145E-08 YES 3.7287E-13 40 3.4276E-05 YES 1.9320E-05 YES 1.1592E-10 78 6.7942E-04 YES 5.8762E-04 YES 3.5257E-09 131 1.1926E-02 YES 1.1499E-02 YES 6.8994E-08 193 9.1868E-02 YES 9.0872E-02 YES 5.4523E-07 Figure 2-16: CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a PWR Subjected to the PT100-100 Cooldown, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 20

2.3.1.2 Shallow Flaw Results The stress intensity factor history for a 0.03 T ID surface flaw subjected to the PT100-100 cooldown transient described in 2.1.2.1 is shown in Figure 2-17 for flaws with aspect ratios of 2, 6, 10 and infinity.

The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. Figure 2-17 clearly shows that K increases initially while the pressure is held constant and the temperature decreases linearly at a rate of 100°F/hour. The maximum K values are reached at 163 minutes, just before the sudden pressure drop that results in a sudden drop in K values. During the subsequent pressure and temperature decrease, the K values reach a minimum sometime between 230 and 240 minutes, and then increase again between 240 minutes and 288 minutes, when a second peak is observed. This second peak is characteristic of shallow surface flaws that barely penetrate the base metal because these cracks are highly affected by the stresses due to the mismatch in thermal expansion coefficient between the cladding and the base metal. After 288 minutes, the temperature is held constant at 70°F, and thus the thermal stresses and associated component of K decrease rapidly, hence there is a decrease in overall K values despite the increase in pressure.

Figure 2-17: Stress Intensity Factor for the PWR PT100-100 Cooldown for a 0.03 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity 21

Figure 2-18: CPI and CPF versus Number of Trials for a 0.03 T Flaw in a PWR with 0°F RTNDT Shift for the PT100-100 cooldown A probabilistic calculation assuming a 0.03 T flaw of aspect ratio equal to 6 and the PT100-100 cooldown resulted in a CPI of 1.4E-07 and a CPF of 1.6E-08 using the RG 1.99, Rev. 2 embrittlement correlation.

This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient. Figure 2-18 shows the calculated CPI and CPF as a function of the number of RPV trials in FAVOR, and one can see that the FAVOR calculations are statistically converged after about 80,000 trials. Similar plots were generated for all the runs performed in this study, so as to assess statistical convergence of the results.

The same PT100-100 cooldown determined using the RG 1.99, Rev. 2 embrittlement correlation for a 0.25 T flaw (shown in Figure 2-2) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. That is, a new P-T limit was not calculated for a 0.03 T shallow flaw, but was instead calculated using current practice based on ASME Section XI, Appendix G, which requires analysis based on a 0.25 T flaw. Furthermore, as for the 0.25 T case, in actual safety analyses, the P-T limit would be adjusted as a function of the maximum ART in the vessel being analyzed, which would result in a different CPI and CPF than presented here. If the P-T limit was recalculated for each value, the shallow flaw effect makes it difficult to predict a priori whether the CPI and CPF values would move up or down.

Table 2-7 and Figure 2-19 show the results of the probabilistic analyses for the 0.03 T flaw for the PT100-100 cooldown, as a function of . As expected, the CPI and CPF increase as increases. It should be noted that the analysis for = 40° did not statistically converge because of the very low calculated probabilities, but the CPI and CPF trend over the number of trials was beginning to stabilize, so the results are believed to be in the correct order of magnitude.

22

Table 2-7: Calculated CPI and CPF for a 0.03 T Flaw of Aspect Ratio 6 in a PWR Subjected to the PT100-100 Cooldown RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 5.2829E-09 NO 1.1418E-10 NO 6.8508E-16 0 1.4259E-07 YES 1.5513E-08 YES 9.3078E-14 40 1.4816E-06 YES 4.6824E-07 YES 2.8094E-12 78 9.3457E-06 YES 5.5843E-06 YES 3.3506E-11 131 7.9849E-05 YES 6.9312E-05 YES 4.1587E-10 193 4.7912E-04 YES 4.6376E-04 YES 2.7826E-09 Figure 2-19: CPI and CPF for a 0.03 T Flaw of Aspect Ratio 6 in a PWR Subjected to the PT100-100 Cooldown, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 23

2.3.2 PT100-50 Cooldown 2.3.2.1 1/4-T Flaw Results The stress intensity factor history for a 0.25 T ID surface flaw subjected to the PT100-50 cooldown transient described in 2.1.2.2 is shown in Figure 2-20 for flaws with aspect ratios of 2, 6, 10 and infinity.

The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. It is important to note that all probabilistic analyses were performed with a flaw of aspect ratio equal to 6. Figure 2-20 clearly shows that K increases initially while the pressure is held constant and the temperature decreases linearly at a rate of 50°F/hour, reaching a maximum between 140 and 160 minutes. Then, at 164 minutes, the sudden pressure drop is accompanied by a sudden drop in K values. The rate of decrease in K is initially high but diminishes as time goes by. At 288 minutes, when the pressure increases again, a corresponding increase in K is observed. Finally, at 576 minutes, when the temperature reaches 70°F and remains constant, K begins to decrease again as the portion of K due to thermal stresses decreases. After this point, the rate of decrease in K values slowly diminishes until the end of the calculation.

Figure 2-20: Stress Intensity Factor for the PWR PT100-50 Cooldown for a 0.25 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity A probabilistic calculation assuming a 0.25 T flaw of aspect ratio equal to 6 and the PT100-50 cooldown resulted in a CPI and a CPF of zero using the RG 1.99, Rev. 2 embrittlement correlation. This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient.

The same PT100-50 cooldown (shown in Figure 2-3) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. Of course, in actual safety analyses, the P-T limit would be adjusted as a function of the maximum ART in the vessel being analyzed, so a higher 24

ART would reduce the allowed pressure, which In turn would result in lower CPI and CPF than presented here for cases where is greater than 0°F.

Table 2-8 and Figure 2-21 show the results of the probabilistic analyses for the 0.25 T flaw for the PT100-50 cooldown, as a function of . As expected, the CPI and CPF increase as increases. It should be noted that the analysis for < 78° resulted in CPI and CPF predictions equal to zero.

Furthermore, for = 78°, the CPF was zero and the CPI did not statistically converge because of the very low calculated probability, but the CPI trend over the number of trials was beginning to stabilize, so the results are believed to be in the correct order of magnitude.

Table 2-8: Calculated CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a PWR Subjected to the PT100-50 Cooldown RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 0 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 78 4.8419E-09 NO 0.0000E+00 YES 0.0000E+00 131 1.6005E-06 YES 1.9420E-07 YES 1.1652E-12 193 2.1283E-04 YES 1.6256E-04 YES 9.7536E-10 25

Figure 2-21: CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a PWR Subjected to the PT100-50 Cooldown, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 2.3.2.2 Shallow Flaw Results The stress intensity factor history for a 0.03 T ID surface flaw subjected to the PT100-50 cooldown transient described in 2.1.2.2 is shown in Figure 2-22 for flaws with aspect ratios of 2, 6, 10 and infinity.

The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. Figure 2-22 clearly shows that K increases initially while the pressure is held constant and the temperature decreases linearly at a rate of 50°F/hour. The maximum K values are reached at 163 minutes, just before the sudden pressure drop that results in a sudden drop in K values. During the subsequent temperature decrease (with pressure initially going down to about 250psi and then coming back up to stabilize around 575psi), the K values reach a minimum sometime between 230 and 288 minutes, and then increase again between 288 minutes and 576 minutes, when a 26

second peak is observed at 576 minutes. This second peak is characteristic of shallow surface flaws that barely penetrate the base metal because these cracks are highly affected by the stresses due to the mismatch in thermal expansion coefficient between the cladding and the base metal. After 576 minutes, the temperature is held constant at 70°F, and thus the thermal stresses and associated component of K decrease rapidly, hence there is a decrease in overall K values despite the constant pressure.

Figure 2-22: Stress Intensity Factor for the PWR PT100-50 Cooldown for a 0.03 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity 27

Figure 2-23: CPI and CPF versus Number of Trials for a 0.03 T Flaw in a PWR with 0°F RTNDT Shift for the PT100-50 cooldown A probabilistic calculation assuming a 0.03 T flaw of aspect ratio equal to 6 and the PT100-50 cooldown resulted in a CPI of 3.8E-05 and a CPF of 2.5E-06 using the RG 1.99, Rev. 2 embrittlement correlation.

This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient. Figure 2-23 shows the calculated CPI and CPF as a function of the number of RPV trials in FAVOR, and one can see that the FAVOR calculations are statistically converged after about 20,000 trials. Similar plots were generated for all the runs performed in this study, so as to assess statistical convergence of the results.

The same PT100-50 cooldown (shown in Figure 2-3) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. That is, a new P-T limit was not calculated for a 0.03 T shallow flaw, but was instead calculated using current practice based on ASME Section XI, Appendix G, which requires analysis based on a 0.25 T flaw. Furthermore, as for the 0.25 T case, in actual safety analyses, the P-T limit would be adjusted as a function of the maximum ART in the vessel being analyzed, which would result in a different CPI and CPF than presented here. If the P-T limit was recalculated for each value, the shallow flaw effect makes it difficult to predict a priori whether the CPI and CPF values would move up or down.

Table 2-9 and Figure 2-24 show the results of the probabilistic analyses for the 0.03 T flaw for the PT100-50 cooldown, as a function of . As expected, the CPI and CPF increase as increases.

Because of the relatively high CPI and CPF values calculated and the relatively large (100,000) number of runs, all calculated conditional probabilities achieved statistical convergence.

28

Table 2-9: Calculated CPI and CPF for a 0.03 T Flaw of Aspect Ratio 6 in a PWR Subjected to the PT100-50 Cooldown RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 1.6642E-05 YES 4.9767E-07 YES 2.9860E-12 0 3.8229E-05 YES 2.5280E-06 YES 1.5168E-11 40 7.4338E-05 YES 8.1860E-06 YES 4.9116E-11 78 1.2374E-04 YES 1.9628E-05 YES 1.1777E-10 131 2.1537E-04 YES 5.2804E-05 YES 3.1682E-10 193 3.4458E-04 YES 1.3171E-04 YES 7.9026E-10 Figure 2-24: CPI and CPF for a 0.03 T Flaw of Aspect Ratio 6 in a PWR Subjected to the PT100-50 Cooldown, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 29

2.3.3 PT50-50 Cooldown 2.3.3.1 1/4-T Flaw Results The stress intensity factor history for a 0.25 T ID surface flaw subjected to the PT50-50 cooldown transient described in 2.1.2.3 is shown in Figure 2-25 for flaws with aspect ratios of 2, 6, 10 and infinity.

The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. It is important to note that all probabilistic analyses were performed with a flaw of aspect ratio equal to 6. Figure 2-25 clearly shows that K increases initially while the pressure is held constant and the temperature decreases linearly at a rate of 50°F/hour, reaching a maximum between 300 and 320 minutes. Then, at 310 minutes, the sudden pressure drop is accompanied by a sudden drop in K values. The rate of decrease in K is initially high but diminishes as time goes by. Finally, at 576 minutes, when the temperature reaches 70°F and remains constant, the rate of decrease in K values accelerates and creates an inflection point in the K versus time curves. After this inflection point, the rate of decrease in K values slowly diminishes until the end of the calculation.

Figure 2-25: Stress Intensity Factor for the PWR PT50-50 Cooldown for a 0.25 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity 30

Figure 2-26: CPI and CPF versus Number of Trials for a 0.25 T Flaw in a PWR with 0°F RTNDT Shift for the PT50-50 cooldown A probabilistic calculation assuming a 0.25 T flaw of aspect ratio equal to 6 and the PT50-50 cooldown resulted in a CPI of 3.1E-08 and a CPF of 5.9E-11 using the RG 1.99, Rev. 2 embrittlement correlation.

This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient. Figure 2-26 shows the calculated CPI and CPF as a function of the number of RPV trials in FAVOR, and one can see that the FAVOR calculations are not quite statistically converged after 100,000 trials. However, the order of magnitude of CPI and CPF is believed to be sufficiently accurate for the purposes of this study. Similar plots were generated for all the runs performed in this study, so as to assess statistical convergence of the results.

The same PT50-50 cooldown (shown in Figure 2-4) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. Of course, in actual safety analyses, the P-T limit would be adjusted as a function of the maximum ART in the vessel being analyzed, so a higher ART would reduce the allowed pressure, which In turn would result in lower CPI and CPF than presented here for cases where is greater than 0°F.

Table 2-10 and Figure 2-27 show the results of the probabilistic analyses for the 0.25 T flaw for the PT50-50 cooldown, as a function of . As expected, the CPI and CPF increase as increases. It should be noted that the analysis for < 0° resulted in CPI and CPF predictions equal to zero.

Furthermore, for = 0°, the CPF and the CPI did not statistically converge because of the very low calculated probability, but the CPI and CPF trend over the number of trials was beginning to stabilize, so the results are believed to be in the correct order of magnitude.

31

Table 2-10: Calculated CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a PWR Subjected to the PT50-50 Cooldown RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 0 3.0806E-08 NO 5.9409E-11 NO 3.5645E-16 40 1.7008E-06 YES 2.3462E-07 YES 1.4077E-12 78 3.9427E-05 NO 2.3607E-05 YES 1.4164E-10 131 1.2837E-03 YES 1.1396E-03 YES 6.8376E-09 193 1.6079E-02 YES 1.5494E-02 YES 9.2964E-08 Figure 2-27: CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a PWR Subjected to the PT50-50 Cooldown, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 32

2.3.3.2 Shallow Flaw Results The stress intensity factor history for a 0.03 T ID surface flaw subjected to the PT50-50 transient described in 2.1.2.3 is shown in Figure 2-28 for flaws with aspect ratios of 2, 6, 10 and infinity. The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. Figure 2-28 clearly shows that K increases initially while the pressure is held constant and the temperature decreases linearly at a rate of 50°F/hour. The maximum K values are reached at 310 minutes, just before the sudden pressure drop that results in a sudden drop in K values.

During the subsequent temperature and temperature decrease (with pressure going down to about 400psi), the K values reach a minimum sometime between 430 and 450 minutes, and then increase again between 450 minutes and 576 minutes, when a second peak is observed at 576 minutes. This second peak is characteristic of shallow surface flaws that barely penetrate the base metal because these cracks are highly affected by the stresses due to the mismatch in thermal expansion coefficient between the cladding and the base metal. After 576 minutes, the temperature is held constant at 70°F, and thus the thermal stresses and associated component of K decrease rapidly, hence there is a decrease in overall K values despite the pressure coming back up to stabilize around 575psi.

Figure 2-28: Stress Intensity Factor for the PWR PT50-50 Cooldown for a 0.03 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity 33

Figure 2-29: CPI and CPF versus Number of Trials for a 0.03 T Flaw in a PWR with 0°F RTNDT Shift for the PT50-50 cooldown A probabilistic calculation assuming a 0.03 T flaw of aspect ratio equal to 6 and the PT50-50 cooldown resulted in a CPI of 2.3E-05 and a CPF of 2.4E-07 using the RG 1.99, Rev. 2 embrittlement correlation.

This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient. Figure 2-29 shows the calculated CPI and CPF as a function of the number of RPV trials in FAVOR, and one can see that the FAVOR calculations are statistically converged after about 20,000 trials. Similar plots were generated for all the runs performed in this study, so as to assess statistical convergence of the results.

The same PT50-50 cooldown (shown in Figure 2-4) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. That is, a new P-T limit was not calculated for a 0.03 T shallow flaw, but was instead calculated using current practice based on ASME Section XI, Appendix G, which requires analysis based on a 0.25 T flaw. Furthermore, as for the 0.25 T case, in actual safety analyses, the P-T limit would be adjusted as a function of the maximum ART in the vessel being analyzed, which would result in a different CPI and CPF than presented here. If the P-T limit was recalculated for each value, the shallow flaw effect makes it difficult to predict a priori whether the CPI and CPF values would move up or down.

Table 2-11 and Figure 2-30 show the results of the probabilistic analyses for the 0.03 T flaw for the PT50-50 cooldown, as a function of . As expected, the CPI and CPF increase as increases.

Because of the relatively high CPI and CPF values calculated and the relatively large (100,000) number of runs, all calculated conditional probabilities achieved statistical convergence.

34

Table 2-11: Calculated CPI and CPF for a 0.03 T Flaw of Aspect Ratio 6 in a PWR Subjected to the PT50-50 Cooldown RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 1.0140E-05 YES 3.4793E-08 YES 2.0876E-13 0 2.2891E-05 YES 2.4388E-07 YES 1.4633E-12 40 4.3563E-05 YES 9.0718E-07 YES 5.4431E-12 78 7.1213E-05 YES 2.3264E-06 YES 1.3958E-11 131 1.2150E-04 YES 7.5567E-06 YES 4.5340E-11 193 1.9106E-04 YES 2.8034E-05 YES 1.6820E-10 Figure 2-30: CPI and CPF for a 0.03 T Flaw of Aspect Ratio 6 in a PWR Subjected to the PT50-50 Cooldown, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 35

2.3.4 C50P Cooldown 2.3.4.1 1/4-T Flaw Results The stress intensity factor history for a 0.25 T ID surface flaw subjected to the 50th percentile cooldown transient described in 2.1.2.4 is shown in Figure 2-31 for flaws with aspect ratios of 2, 6, 10 and infinity.

The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. It is important to note that all probabilistic analyses were performed with a flaw of aspect ratio equal to 6. Figure 2-31 shows that for this transient, K continuously decreases as temperature and pressure decrease over time. The initial decrease is very sharp, dominated by the sharp pressure drop until about 300 minutes. After the initial pressure drop in the system, the value of K is driven by thermal stresses and basically correlate to the cooldown rate. In fact, one can observe a first plateau from ~300 minutes to ~600 minutes, a second plateau from ~600 minutes to ~2800 minutes, and finally a step down at ~2800 minutes with final values of K essentially flat after 2900 minutes.

Figure 2-31: Stress Intensity Factor for PWR 50th Percentile Cooldown for 0.25 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity A probabilistic calculation assuming a 0.25 T flaw of aspect ratio equal to 6 and the 50th percentile cooldown resulted in a CPI and a CPF of zero using the RG 1.99, Rev. 2 embrittlement correlation. This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient. The same 50th percentile cooldown (shown in Figure 2-6 and Figure 2-8) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient.

Table 2-12 and Figure 2-32 show the results of the probabilistic analyses for the 0.25 T flaw for the 50th percentile cooldown, as a function of . All values of CPI and CPF calculated were zero using 36

100,000 vessel simulations, except the CPI value for a value of 193°F (resulting in a very high ART of around 427°F).

Table 2-12: Calculated CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a PWR Subjected to the 50th Percentile Cooldown RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 0 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 78 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 131 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 193 2.5504E-08 NO 0.0000E+00 YES 0.0000E+00 37

Figure 2-32: CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a PWR Subjected to the 50th Percentile Cooldown, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged). Note that all CPF values were zero, and thus are not shown, and the only non-zero value of CPI is shown as an open symbol.

2.3.4.2 Shallow Flaw Results The stress intensity factor history for a 0.03 T ID surface flaw subjected to the 50th percentile cooldown transient described in 2.1.2.4 is shown in Figure 2-33 for flaws with aspect ratios of 2, 6, 10 and infinity.

The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. Figure 2-33 shows that in the early part of the cooldown, K increases as a result of the temperature drop, but the simultaneous pressure reduction results in a large decrease in K between the peak around 60 minutes and the minimum reached at about 360 minutes.

Beyond 360 minutes, the pressure remains essentially constant until the final pressure drop at ~2800 38

minutes, and the K increase between 360 minutes and 2800 minutes is driven by the decrease in temperature. After 2800 minutes, K remains almost constant as both the pressure and temperature remain roughly constant.

Figure 2-33: Stress Intensity Factor for PWR 50°F/hour P-T Limit Cooldown for 0.03 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity Figure 2-34: CPI and CPF versus Number of Trials for the 0.03 T Flaw in a PWR with 0°F RTNDT Shift for the 50th Percentile Cooldown 39

A probabilistic calculation assuming a 0.03 T flaw of aspect ratio equal to 6 and the 50th percentile cooldown resulted in a CPI of 1.0E-09 and a CPF of zero using the RG 1.99, Rev. 2 embrittlement correlation. This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient. Figure 2-34 shows the calculated CPI as a function of the number of RPV trials in FAVOR, and one can see that the FAVOR calculations are statistically converged after about 40,000 trials. Similar plots were generated for all the runs performed in this study, so as to assess statistical convergence of the results.

The same 50th percentile cooldown (shown in Figure 2-6 and Figure 2-8) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. Table 2-13 and Figure 2-35 show the results of the probabilistic analyses for the 0.03 T flaw for the 50th percentile cooldown, as a function of . As expected, the CPI increases as increases, but CPF remained zero in all cases. All calculated conditional probabilities achieved statistical convergence.

Table 2-13: Calculated CPI and CPF for a 0.03 T Flaw of Aspect Ratio 6 in a PWR Subjected to a 50°F/hour Cooldown at the P-T Limit RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 1.6441E-10 YES 0.0000E+00 YES 0.0000E+00 0 1.0556E-09 YES 0.0000E+00 YES 0.0000E+00 40 4.0286E-09 YES 0.0000E+00 YES 0.0000E+00 78 1.0078E-08 YES 0.0000E+00 YES 0.0000E+00 131 2.4315E-08 YES 0.0000E+00 YES 0.0000E+00 193 4.7276E-08 YES 0.0000E+00 YES 0.0000E+00 40

Figure 2-35: CPI and CPF for a 0.03 T Flaw of Aspect Ratio 6 in a PWR Subjected to a 50°F/hour Cooldown at the P-T Limit, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 2.3.5 Sensitivity Study on Actual PWR Cooldowns for Shallow Flaw In order to better understand the possible CPI and CPF variability (i.e. the range of risk) associated with normal cooldowns, the 42 actual cooldown transients were simulated for a single value of RTNDT shift equal to 100°F. Since the risk of crack growth initiation (CPI) associated with shallow flaws is greater than that associated with 1/4 T flaws for the 50th percentile of actual transients, the simulations of the 42 actual cooldowns were performed only for the case of a 0.03 T shallow flaw. It is important to note that calculated CPF values for this transient were equal to zero in all the cases analyzed (for both the 1/4 T and 0.03 T flaw).

41

Figure 2-36 show the distribution of CPI and CPF values for the 42 actual cooldowns simulated. The CPI values range from 4.9E-12 to 2.0E-05, with a mean of 1.5E-06, a median of 1.3E-07, and a 99th percentile of 1.6E-05. 6 of the 42 actual cooldowns resulted in a CPI of zero. The CPF values range from 4.2E-15 to 1.1E-10, with a mean of 2.6E-12, a median of zero, and a 99th percentile of 6.4E-11. 38 of the 42 actual cooldown transients resulted in a CPF of zero. In summary, none of the 42 actual cooldown transients presented any significant vessel failure risk, and all were at least roughly 4 orders of magnitude below 1.0E-06 events per year.

Figure 2-36: CPI (top) and CPF (bottom) Distribution for the 42 Actual PWR Cooldown Transients Modeled, for RTNDT Shift of 100°F Figure 2-37 shows the calculated CPI and CPF values for the 42 actual PWR cooldowns in comparison with the CPI and CPF calculated for the 50th percentile cooldown as a function of RTNDT shift. One can see that the value of CPI for the 50th percentile cooldown for a RTNDT shift value of 100°F is about 2x10-8, which corresponds roughly to the 35th percentile of the calculated CPI values for the 42 actual 42

cooldowns. As described above, none of the CPF values calculated were of any concern from a vessel integrity and safety point of view. In fact, for the actual cooldowns simulated, the transient frequency can be set equal to 1. As a result, for these transients, CPF = TWCF (through-wall cracking frequency).

The highest TWCF value predicted for the 42 actual cooldowns is thus 1.1E-10, which is almost 4 orders of magnitude below 1.0E-06 events per year.

Figure 2-37: Comparison of the 0.03 T Flaw CPI and CPF at a RTNDT Shift of 100°F for the 42 Actual PWR Cooldowns with the CPI as a Function of RTNDT Shift for the 50th Percentile Cooldown (CPF Values Were all Zero for the 50th Percentile Cooldown) 43

3 BWR Scoping Study 3.1 Vessel and Transient Characteristics 3.1.1 Vessel Characteristics and Embrittlement Map The BWR vessel modeled had the following characteristics:

• Inner radius 110.37 inches

• Cladding Thickness 0.21 inches

• Base metal thickness 5.62 inches

• Total wall thickness 5.83 inches The material properties used in the FAVOR model are the same as for the PWR cases, and are given in Table 2-1 and Table 2-2. The stress free temperature was again assumed to be equal to 488°F. Pressure loading was applied since the calculations presented herein are for an ID surface flaw. Finally, WRS was included in calculations for axial and circumferential welds. These modeling parameters were chosen to match the analysis parameters historically used by NRC for regulatory analyses (e.g. 10 CFR 50.61a technical bases), and are the ones recommended in the FAVOR theory manual.

The BWR vessel rollout diagram and corresponding basic embrittlement information from the RVID database are provided in Figure 3-1 and Table 3-1 respectively.

Figure 3-1: BWR Vessel Rollout Diagram 44

Table 3-1: BWR Vessel Embrittlement Map Information and Corresponding FAVOR Input max neutron 1/4T RTNDT ID RTndt (u)

Major fluence @72 @ 72 EFPY Description Heat ID Plate or Weld ID %Cu %Ni %P [Initial Region EFPY w/o Margin RTndt]

1019 n/cm2 o F

1 Axial Weld 1P2809 LOWER INTERMEDIATE AXIAL WELDS 1-308 0.270 0.735 0.013 -50.000 0.47 92.7 2 Axial Weld 1P2809 LOWER INTERMEDIATE AXIAL WELDS 1-308 0.270 0.735 0.013 -50.000 0.47 92.7 3 Axial Weld 1P2809 LOWER INTERMEDIATE AXIAL WELDS 1-308 0.270 0.735 0.013 -50.000 0.47 92.7 4 Axial Weld 13253 LOWER SHELL AXIAL WELDS 1-307A/C 0.221 0.732 0.023 -50.000 0.32 63.1 5 Axial Weld 13253 LOWER SHELL AXIAL WELDS 1-307A/C 0.221 0.732 0.023 -50.000 0.32 63.1 6 Axial Weld 13253 LOWER SHELL AXIAL WELDS 1-307A/C 0.221 0.732 0.023 -50.000 0.32 63.1 7 CircWeld 33A277 CIRC WELD 1-313B 0.258 0.165 0.017 -50.000 0.32 25.6 8 Plate C4337-1 LOWER INT. SHELL G-4803-7 0.170 0.620 0.011 -20.000 0.47 68.5 9 Plate C3985-2 LOWER INT. SHELL G-4804-1 0.130 0.580 0.015 -20.000 0.47 42.8 10 Plate C4114-2 LOWER INT. SHELL G-4804-2 0.130 0.700 0.010 -20.000 0.47 44.9 11 Plate C4112-1 LOWER SHELL G-4805-1 0.130 0.640 0.011 8.000 0.32 63.1 12 Plate C4112-2 LOWER SHELL G-4805-2 0.130 0.640 0.011 10.000 0.32 65.1 13 Plate C4149-1 LOWER SHELL G-4805-3 0.140 0.570 0.009 -10.000 0.32 49.0 The resulting FAVOR BWR embrittlement map input was as follows:

• ========================================================================================================*

• 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

• 1 8 9 1 0.4666 0.27 0.735 0.013 1 2 0 -50 10 11 0.714 116.7 656 1 0.000 0.000 2 10 8 2 0.4666 0.27 0.735 0.013 1 2 0 -50 10 11 0.714 116.7 656 1 0.000 0.000 3 9 10 3 0.4666 0.27 0.735 0.013 1 2 0 -50 10 11 0.714 116.7 656 1 0.000 0.000 4 11 12 4 0.32 0.221 0.732 0.023 1 2 0 -50 10 11 0.714 53.75 302 1 0.000 0.000 5 13 11 5 0.32 0.221 0.732 0.023 1 2 0 -50 10 11 0.714 53.75 302 1 0.000 0.000 6 12 13 6 0.32 0.221 0.732 0.023 1 2 0 -50 10 11 0.714 53.75 302 1 0.000 0.000 7 8 11 7 0.32 0.258 0.165 0.017 1 2 0 -50 10 11 360 1.25 4005 2 0.000 0.000 8 8 8 8 0.4666 0.17 0.62 0.011 1 0 0 -20 10 21 119.2 116.7 0.000 0 0.000 0.000 9 9 9 9 0.4666 0.13 0.58 0.015 1 0 0 -20 10 21 119.2 116.7 0.000 0 0.000 0.000 10 10 10 10 0.4666 0.13 0.7 0.01 1 0 0 -20 10 21 119.2 116.7 0.000 0 0.000 0.000 11 11 11 11 0.32 0.13 0.64 0.011 1 0 0 8 10 21 119.2 53.75 0.000 0 0.000 0.000 12 12 12 12 0.32 0.13 0.64 0.011 1 0 0 10 10 21 119.2 53.75 0.000 0 0.000 0.000 13 13 13 13 0.32 0.14 0.57 0.009 1 0 0 -10 10 21 119.2 53.75 0.000 0 0.000 0.000 Of note, RG 1.99, Rev. 2 was used to calculate the RTNDT at the 0.25 T location so as to calculate the P-T limit for a 100°F/hour cooldown (last column in Table 3-1). The maximum RTNDT at the 0.25 T location at 72 EFPY was determined to be 92.7°F, which is the value that was used for the P-T limit calculation.

45

3.1.2 Transient Definitions 3.1.2.1 PT100-100: 100°F/Hour Cooldown at Maximum Allowed Pressure for 100°F/hour Cooldown As was done for the PWR analysis, in order to determine the allowable pressure for the P-T limit cooldown at a rate of 100°F/hour, ASME Section XI, Appendix G, Paragraph G-2215 [6] was used. The risk-informed alternative was NOT used. Assuming the limiting case of an axial flaw and a wall thickness between 4 and 12 inches, the allowable pressure is given by:

=

2 with

= 33.2 + 20.734 [0.02( )]

and

= 0.926 In order to determine the applied stress intensity factor due to thermal stresses (KIt), FAVOR was used to simulate a 100°F/hour cooldown from 550°F to 70°F with no applied pressure, and using a stress free temperature of 550°F (such that the initial thermal stresses are zero). The resulting K calculated by FAVOR for a 0.25 T flaw and an aspect ratio of 6 was used to determine the allowable pressure using the equation above. The value of crack tip RTNDT used was92.7°F, as described in 3.1.1.

The resulting P-T limit transient for a 100°F/hour cooldown from 550°F to 70°F is shown in Figure 3-2.

This transient is referred to as the PT100-100 transient in this document, and its frequency of occurrence is 1x10-7 events per year (based on the BTP 5-3 closure memo assessment [7]).

Figure 3-2: BWR PT100-100 Cooldown Transient Characteristics 46

3.1.2.2 PT100-50: 50°F/Hour Cooldown at Maximum Allowed Pressure for 100°F/hour Cooldown The same P-T limit pressure history based on a 100°F/hour cooldown (see section3.1.2.1) was used along with a 50°F/hour cooldown temperature history. The resulting P-T limit transient for a 50°F/hour cooldown from 550°F to 70°F is shown in Figure 3-3 (same pressure history as for the PT100-100 cooldown, but with a slower cooldown of 50°F/hour). This transient is referred to as the PT100-50 transient in this document, and its frequency of occurrence is 1x10-7 events per year (based on the BTP 5-3 closure memo assessment [7]).

Figure 3-3: BWR PT100-50 Cooldown Transient Characteristics 47

3.1.2.3 PT50-50: 50°F/Hour Cooldown at Maximum Allowed Pressure for 50°F/hour Cooldown As was done for the PWR analysis, in order to determine the allowable pressure for the P-T limit cooldown at a rate of 50°F/hour, ASME Section XI, Appendix G, Paragraph G-2215 [6] was used. The risk-informed alternative was NOT used. Assuming the limiting case of an axial flaw and a wall thickness between 4 and 12 inches, the allowable pressure is given by:

=

2 with

= 33.2 + 20.734 [0.02( )]

and

= 0.926 In order to determine the applied stress intensity factor due to thermal stresses (KIt), FAVOR was used to simulate a 50°F/hour cooldown from 550°F to 70°F with no applied pressure, and using a stress free temperature of 550°F (such that the initial thermal stresses are zero). The resulting K calculated by FAVOR for a 0.25 T flaw and an aspect ratio of 6 was used to determine the allowable pressure using the equation above. The value of crack tip RTNDT used was 92.7°F, as described in 3.1.1.

The resulting P-T limit transient for a 50°F/hour cooldown from 550°F to 70°F is shown in Figure 3-2.

This transient is referred to as the PT50-50 transient in this document, and its frequency of occurrence is 1x10-7 events per year (based on the BTP 5-3 closure memo assessment [7]).

Figure 3-4: BWR PT50-50 Cooldown Transient Characteristics 48

3.1.2.4 S50: 50°F/Hour Saturation Pressure Cooldown Figure 3-5 Shows the pressure and temperature histories for a 50°F/hour cooldown at saturation pressure and temperature. This transient is hereafter referred to as the 50°F/hour saturation cooldown or S50 transient in this document, and its frequency of occurrence is 1.0 events per year (based on the BTP 5-3 closure memo assessment [7]).

Figure 3-5: BWR Saturation Cooldown at 50°F/hour Transient Characteristics 49

3.1.2.5 LT40-40: Leak Test with 40°F/Hour Heatup and 40°F/Hour Cooldown Figure 3-6 shows the pressure and temperature histories used for a BWR leak test transient with a 40°F/hour heatup rate and a 40°F/hour cooldown rate. The allowable pressure and temperature were derived from the ASME Section XI Appendix G, Paragraph G-2215 rules [6], as was done for the P-T limit cooldowns. For this transient, the leak test pressure and temperature were 1.1 ksi and 167°F, respectively.

This transient is referred to as the LT40-40 transient in this document, and its frequency of occurrence is estimated at 1x10-3 events per year (based on the BTP 5-3 closure memo assessment [7]).

Figure 3-6: BWR LT40-40 Leak Test Transient Characteristics 50

3.1.2.6 LT40-100: 40°F/Hour Heatup and 100°F/Hour Cooldown Leak Test Figure 3-7 shows the pressure and temperature histories used for a BWR leak test transient with a 40°F/hour heatup rate and a 100°F/hour cooldown rate. The allowable pressure and temperature were derived from the ASME Section XI Appendix G, Paragraph G-2215 rules [6], as was done for the P-T limit cooldowns. For this transient, the test pressure and temperature were 1.1 ksi and 167°F, respectively.

This transient is referred to as the LT40-100 transient in this document, and its frequency of occurrence is estimated at 1x10-3 events per year (based on the BTP 5-3 closure memo assessment [7]).

Figure 3-7: BWR LT40-100 Leak Test Transient Characteristics 51

3.1.2.7 LTA: Actual Leak Test Figure 3-8 shows the pressure and temperature histories used for the actual leak test transient obtained from EPRI. For this transient, the leak test pressure and temperature were 1.0 ksi and 208.5°F, respectively.

This transient is referred to as the LTA transient in this document, and its frequency of occurrence is estimated at 1.0 events per year.

Figure 3-8: BWR LTA Leak Test Transient Characteristics 52

3.1.2.8 BWR Transient PT Curves Figure 3-9 shows the PT curves for the 4 cooldown transients and 3 leak test transients modeled for the BWR RTNDT shift sensitivity studies. Overall, the PT100-100 and PT50-50 transients are similar on a PT curve, and one would thus expect relatively similar trends for stress intensity factor histories, as can be seen by comparing Figure 3-12 with Figure 3-22 and Figure 3-15 with Figure 3-25. Furthermore, based on the relative positions of the PT curves, one would expect that for a 1/4 T flaw (which is the postulated flaw for the creation of PT curves), the CPI and CPF values for the cooldown transients shown would be ranked as follows from highest to lowest: PT100-100, PT50-50, PT100-50, and finally S50. This is indeed generally the trend for 1/4 T flaws, as described in Section 3.3. However, for shallow flaws, the trends are different, and the PT curves are not a good indication of the relative values of CPI and CPF for the cooldown transients.

Figure 3-9: PT curves for the BWR cooldown and leak test transients modeled for the RTNDT shift sensitivity studies 53

3.1.3 Other FAVOR Simulation Input Information Aside from the vessel and transient information, the following FAVOR simulation inputs were used:

Simulation Parameters Irradiation shift correlation choice Regulatory Guide 1.99 Rev. 2 Normal operating coolant temperature 550°F Plant operating time 72 EFPY Ductile tearing model in IGA submodel off Ductile tearing initiation checks and reporting off Additional reports on CPI/CPF for major beltline regions off Flow stress for net-section plastic collapse failure model 80 ksi Maximum value for KIc or KIa 200 ksiin Crack arrest model High-constraint KIa (CCA specimens)

Resampling when advancing into new weld layer on Normalized crack depth corresponding to vessel failure 0.9 Fluence Sampling Multiplier for standard deviation of mean local fluence 0.118 Multiplier for standard deviation of local fluence 0.056 Weld Chemistry Sampling Standard deviation for Copper sampling in welds 0.167 wt%

Standard deviation for Nickel sampling in welds 0.162 wt%

Standard deviation for Phosphorus sampling in welds 0.0013 wt%

Plate Chemistry Sampling Standard deviation for Copper sampling in plates 0.0073 wt%

Standard deviation for Nickel sampling in plates 0.0244 wt%

Standard deviation for Phosphorus sampling in plates 0.0013 wt%

54

3.2 RTNDT Shift Distribution As was done for PWR analyses, the initial RTNDT value was changed via input to the FAVOR embrittlement map to simulate the effects of changing the embrittlement correlation from RG 1.99, SHIFT Rev. 2 to E90015 without changing the FAVOR code. A compilation of RTNDT values for the entire BWR fleet, based on embrittlement capsule data, was performed as part of the RG 1.99, Rev. 2 update effort. Figure 3-10 shows the distribution of values across the US BWR fleet, and Table 3-2 shows the values of corresponding to key percentiles (and vice versa).

Figure 3-10: Probability Distribution of the BWR RTNDT Shift as a Result of the Change from RG 1.99, Rev.

2 to E90015 Embrittlement Correlations (all capsule data)

Table 3-2: BWR RTNDT Shift as a Result of the Change from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations at Percentiles of Interest for this Study (all capsule data)

RTNDT Shift from RG 1.99, Rev. 2 to Percentiles E90015 0.005 -40 0.01 -39 0.05 -27 0.1 -8 0.25 0 0.5 6 0.75 15 0.9 31 0.95 47 0.99 128 Figure 3-10 and Table 3-2 show that for BWRs at 72 EFPY, changing the embrittlement correlation from RG 1.99, Rev. 2 to E90015 would increase the adjusted reference temperature in most cases, since a 0°F 55

shift (i.e. no shift) corresponds to the 25th percentile for . However, the distribution for BWRs is narrower than that reported for PWRs in 2.2, such that the median is only 6°F and the upper 99th percentile is 128°F.

It is important to point out that considering all available capsule does not provide the most accurate picture of the actual material embrittlement in the US fleet, because some capsules pulled early in the reactor life and for which data has been extrapolated can result in large errors in the actual embrittlement of reactor materials. A more accurate method may consist of filtering the capsule data to only keep the fluence capsule data for the longest actual irradiation time for each reactor in the fleet.

This reduces extrapolation error and results in a far more accurate picture of the material embrittlement for the vessels in the US fleet.

Figure 3-11: Probability Distribution of the BWR RTNDT Shift as a Result of the Change from RG 1.99, Rev.

2 to E90015 Embrittlement Correlations (filtered capsule data) 56

Table 3-3: BWR RTNDT Shift as a Result of the Change from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations at Percentiles of Interest for this Study (filtered capsule data)

RTNDT Shift from RG 1.99, Rev. 2 to Percentiles E90015 0.01 -50 0.05 -44 0.06 -40 0.1 -14 0.17 0 0.25 3 0.5 19 0.75 31 0.9 48 0.95 62 0.99 63 Figure 3-11 and Table 3-3 show that when considering only the highest fluence capsule for each reactor, the distribution is narrower and shift downward, such that the 99th percentile is around 63°F.

57

3.3 Results Probabilistic FAVOR runs were performed for the highlighted values from Table 3-2. The values chosen were -40°F, 0°F, followed by the 50 , 75 , 95 , and 99 percentile values, so as to cover the th th th th wide range of possible values (see Section 3.2). This section presents results for a BWR vessel containing a single flaw of aspect ratio equal to 6 and depths of 0.25 T and 0.04 T, for the transients described in Section 3.1.2.

3.3.1 PT100-100 Cooldown 3.3.1.1 1/4-T Flaw Results The stress intensity factor history for a 0.25 T ID surface flaw subjected to the PT100-100 cooldown transient described in 3.1.2.1 is shown in Figure 3-12 for flaws with aspect ratios of 2, 6, 10 and infinity.

The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. It is important to note that all probabilistic analyses were performed with a flaw of aspect ratio equal to 6. Figure 3-12 clearly shows that K increases initially while the pressure is held constant and the temperature decreases linearly at a rate of 100°F/hour, reaching a maximum around 250 minutes. One exception is for the flaw aspect ratio of 2, where the maximum is reached around 80 minutes and then K is held almost constant until the pressure drops at 250 minutes. Then, at 250 minutes, the sudden pressure drop is accompanied by a sudden drop in K values. Finally, at 288 minutes, when the temperature reaches 70°F and remains constant, the rate of decrease in K values diminishes and K versus time curves plateau at their minimum value.

Figure 3-12: Stress Intensity Factor for the BWR PT100-100 Cooldown for a 0.25 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity 58

A probabilistic calculation assuming a 0.25 T flaw of aspect ratio equal to 6 and the PT100-100 cooldown resulted in a CPI of 1.6E-08 and a CPF of 2.3E-10 using the RG 1.99, Rev. 2 embrittlement correlation.

This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient. Figure 3-13 shows the calculated CPI and CPF as a function of the number of RPV trials in FAVOR. Although it looks like the CPF value may have sufficiently converged, the CPI value has not converged, but the order of magnitude of CPI is believed to be sufficiently accurate for the purposes of this study.

Figure 3-13: CPI and CPF versus Number of Trials for a 0.25 T Flaw in a BWR with 0°F RTNDT Shift for the PT100-100 cooldown To produce FAVOR runs corresponding to a single ID surface flaw of depth 0.25 T and aspect ratio 6, the VFLAW input format was used to specify the flaw population as follows:

• The surface flaw density was set to result in a single flaw of depth 0.25 T and aspect ratio 6, with zeros everywhere else

• The embedded weld flaw density was set to zero for all flaw depths and aspect ratios

• The embedded plate flaw density was set to zero for all flaw depths and aspect ratios The PT100-100 cooldown determined using the RG 1.99, Rev. 2 embrittlement correlation (shown in Figure 3-2) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. Of course, in actual safety analyses, the P-T limit would be adjusted as a function of the maximum ART in the vessel being analyzed, so a higher ART would reduce the allowed pressure, which In turn would result in significantly lower CPI and CPF than presented here for cases where is greater than 0°F.

59

Table 3-4: Calculated CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a BWR Subjected to the PT100-100 Cooldown RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 3.9308E-13 NO 0.0000E+00 YES 0.0000E+00 0 1.6133E-08 NO 2.3416E-10 NO 2.3416E-17 6 3.0774E-08 YES 1.3237E-09 YES 1.3237E-16 15 8.1789E-08 YES 1.1164E-08 YES 1.1164E-15 47 3.3045E-06 YES 2.2012E-06 YES 2.2012E-13 128 1.0648E-03 YES 1.0060E-03 YES 1.0060E-10 Figure 3-14: CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a BWR Subjected to the PT100-100 Cooldown, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 60

Table 3-4 and Figure 3-14 show the results of the probabilistic analyses for the 0.25 T flaw for the PT100-100 cooldown, as a function of . As expected, the CPI and CPF increase as increases. It should be noted that the analyses for 0° did not statistically converge because of the very low calculated probabilities, but the CPI and CPF trend over the number of trials was beginning to stabilize, so the results are believed to be in the correct order of magnitude. Overall, it can be said that CPI and CPF for BWRs with a 1/4 T flaw are about 1 or 2 orders of magnitude lower than for PWRs for the PT100-100 cooldown.

3.3.1.2 Shallow Flaw Results The stress intensity factor history for a 0.04 T ID surface flaw subjected to the PT100-100 cooldown transient described in 3.1.2.1 is shown in Figure 3-15 for flaws with aspect ratios of 2, 6, 10 and infinity.

The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. Figure 3-15 clearly shows that K increases initially while the pressure is held constant and the temperature decreases linearly at a rate of 100°F/hour. The maximum K values are reached at 250 minutes, just before the sudden pressure drop that results in a sudden drop in K values. During the subsequent pressure and temperature decrease, the K values reach a minimum sometime between 330 and 350 minutes, and then remain at this minimum value. It should be noted that a small inflection point can be observed in the decreasing K trend at 288 minutes, where the temperature reaches 70°F and subsequently remains there.

A probabilistic calculation assuming a 0.04 T flaw of aspect ratio equal to 6 and the PT100-100 cooldown resulted in a CPI of 1.5E-11 and a CPF of 4.0E-15 using the RG 1.99, Rev. 2 embrittlement correlation.

This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient. Figure 3-16 shows the calculated CPI and CPF as a function of the number of RPV trials in FAVOR, and one can see that the FAVOR calculations are not statistically converged after 100,000 trials. However, the values calculated are believed to have reached the correct order of magnitude.

61

Figure 3-15 Stress Intensity Factor for the BWR PT100-100 Cooldown for 0.04 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity Figure 3-16: CPI and CPF versus Number of Trials for a 0.04 T Flaw in a BWR with 0°F RTNDT Shift for the PT100-100 cooldown t To produce FAVOR runs corresponding to a single ID surface flaw of depth 0.04 T and aspect ratio 6, the VFLAW input format was used to specify the flaw population as follows:

62

• The surface flaw density was set to result in a single flaw of depth 0.04 T and aspect ratio 6, with zeros everywhere else

• The embedded weld flaw density was set to zero for all flaw depths and aspect ratios

• The embedded plate flaw density was set to zero for all flaw depths and aspect ratios The same PT100-100 cooldown determined using the RG 1.99, Rev. 2 embrittlement correlation for a 0.25 T flaw (shown in Figure 3-2) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. That is, a new P-T limit was not calculated for a 0.04 T shallow flaw, but was instead calculated using current practice based on ASME Section XI, Appendix G, which requires analysis based on a 0.25 T flaw. P-T Furthermore, as for the 0.25 T case, in actual safety analyses, the P-T limit would be adjusted as a function of the maximum ART in the vessel being analyzed, which would result in a different CPI and CPF than presented here. If the P-T limit was recalculated for each value, the shallow flaw effect makes it difficult to predict a priori whether the CPI and CPF values would move up or down.

Table 3-5 and Figure 3-17 show the results of the probabilistic analyses for the 0.04 T flaw for the PT100-100 cooldown, as a function of . As expected, the CPI and CPF increase as increases. It should be noted that the analyses for < 47° did not statistically converge because of the very low calculated probabilities, but the CPI and CPF trend over the number of trials was beginning to stabilize, so the results are believed to be in the correct order of magnitude. Overall, it can be said that CPI and CPF for BWRs with a 0.04 T flaw are at least 2 orders of magnitude lower than for PWRs for the PT100-100 cooldown.

Table 3-5: Calculated CPI and CPF for a 0.04 T Flaw of Aspect Ratio 6 in a BWR Subjected to the PT100-100 Cooldown RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 0.0000E+00 Maybe 0.0000E+00 Maybe 0.0000E+00 0 1.4841E-11 NO 3.9582E-15 NO 3.9582E-22 6 4.5757E-11 NO 2.2927E-13 NO 2.2927E-20 15 1.6055E-10 NO 2.3249E-12 NO 2.3249E-19 47 4.0525E-09 YES 2.3711E-10 YES 2.3711E-17 128 7.2475E-07 YES 1.8924E-07 YES 1.8924E-14 63

Figure 3-17: CPI and CPF for a 0.04 T Flaw of Aspect Ratio 6 in a BWR Subjected to the PT100-100 Cooldown, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 64

3.3.2 PT100-50 Cooldown 3.3.2.1 1/4-T Flaw Results The stress intensity factor history for a 0.25 T ID surface flaw subjected to the PT100-50 cooldown transient described in 3.1.2.2 is shown in Figure 3-18 for flaws with aspect ratios of 2, 6, 10 and infinity.

The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. It is important to note that all probabilistic analyses were performed with a flaw of aspect ratio equal to 6. Figure 3-18 clearly shows that K increases initially while the pressure is held constant and the temperature decreases linearly at a rate of 50°F/hour, reaching a maximum between around 250 minutes. Then, at 250 minutes, the sudden pressure drop is accompanied by a sudden drop in K values. The rate of decrease in K is high but at 288 minutes, when the pressure stabilizes, K begins to increase again as a result of the continued temperature decrease.

Finally, at 576 minutes, when the temperature reaches 70°F and remains constant, K begins to decrease again as the portion of K due to thermal stresses decreases. After this point, the rate of decrease in K values slowly diminishes until the end of the calculation.

Figure 3-18: Stress Intensity Factor for the BWR PT100-50 Cooldown for a 0.25 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity A probabilistic calculation assuming a 0.25 T flaw of aspect ratio equal to 6 and the PT100-50 cooldown resulted in a CPI and a CPF of zero using the RG 1.99, Rev. 2 embrittlement correlation. This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient.

The same PT100-50 cooldown (shown in Figure 3-3) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. Of course, in actual safety analyses, the P-T limit would be adjusted as a function of the maximum ART in the vessel being analyzed, so a higher 65

ART would reduce the allowed pressure, which In turn would result in lower CPI and CPF than presented here for cases where is greater than 0°F.

Table 3-6 shows the results of the probabilistic analyses for the 0.25 T flaw for the PT100-50 cooldown, as a function of . Interestingly, regardless of the value, the CPI and CPF predictions were equal to zero.

Table 3-6: Calculated CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a BWR Subjected to the PT100-50 Cooldown RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 0 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 6 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 15 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 47 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 128 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 3.3.2.2 Shallow Flaw Results The stress intensity factor history for a 0.04 T ID surface flaw subjected to the PT100-50 cooldown transient described in 3.1.2.2 is shown in Figure 3-19 for flaws with aspect ratios of 2, 6, 10 and infinity.

The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. Figure 3-19 clearly shows that K increases initially while the pressure is held constant and the temperature decreases linearly at a rate of 50°F/hour. The maximum K values are reached at 250 minutes, just before the sudden pressure drop that results in a sudden drop in K values. The rate of decrease in K is high but at 288 minutes, when the pressure stabilizes, K begins to increase again as a result of the continued temperature decrease. Finally, at 576 minutes, when the temperature reaches 70°F and remains constant, K begins to decrease again as the portion of K due to thermal stresses decreases, resulting in a second peak in K values. After this point, the rate of decrease in K values slowly diminishes until the end of the calculation. The second peak is characteristic of shallow surface flaws that barely penetrate the base metal because these cracks are highly affected by the stresses due to the mismatch in thermal expansion coefficient between the cladding and the base metal.

66

Figure 3-19: Stress Intensity Factor for the BWR PT100-50 Cooldown for a 0.04 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity Figure 3-20: CPI and CPF versus Number of Trials for a 0.04 T Flaw in a BWR with 0°F RTNDT Shift for the PT100-50 cooldown 67

A probabilistic calculation assuming a 0.04 T flaw of aspect ratio equal to 6 and the PT100-50 cooldown resulted in a CPI of 1.5E-15 and a CPF of zero using the RG 1.99, Rev. 2 embrittlement correlation. This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient. Figure 3-20 shows the calculated CPI and CPF (note that CPF is zero, thus not visible on the plot) as a function of the number of RPV trials in FAVOR, and one can see that the FAVOR calculations are not statistically converged after 100,000 trials.

However, the values calculated are believed to have reached the correct order of magnitude. Similar plots were generated for all the runs performed in this study, so as to assess statistical convergence of the results.

The same PT100-50 cooldown (shown in Figure 3-3) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. That is, a new P-T limit was not calculated for a 0.04 T shallow flaw, but was instead calculated using current practice based on ASME Section XI, Appendix G, which requires analysis based on a 0.25 T flaw. P-T Furthermore, as for the 0.25 T case, in actual safety analyses, the P-T limit would be adjusted as a function of the maximum ART in the vessel being analyzed, which would result in a different CPI and CPF than presented here. If the P-T limit was recalculated for each value, the shallow flaw effect makes it difficult to predict a priori whether the CPI and CPF values would move up or down.

Table 3-7 and Figure 3-21 show the results of the probabilistic analyses for the 0.04 T flaw for the PT100-50 cooldown, as a function of . As expected, the CPI increases as increases. CPF was always zero except for the highest value of . It should be noted that the analyses for <

15° did not statistically converge for CPI because of the very low calculated probabilities, but the CPI trend over the number of trials was beginning to stabilize, so the results are believed to be in the correct order of magnitude. Overall, it can be said that CPI and CPF for BWRs with a 0.04 T flaw are many orders of magnitude lower than for PWRs for the PT100-50 cooldown.

Table 3-7: Calculated CPI and CPF for a 0.04 T Flaw of Aspect Ratio 6 in a BWR Subjected to the PT100-50 Cooldown RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 0 1.5472E-15 NO 0.0000E+00 YES 0.0000E+00 6 1.8957E-14 NO 0.0000E+00 YES 0.0000E+00 15 1.4315E-13 NO 0.0000E+00 YES 0.0000E+00 47 5.8837E-12 YES 0.0000E+00 YES 0.0000E+00 128 1.5554E-09 YES 3.5658E-12 YES 3.5658E-19 68

Figure 3-21: CPI and CPF for a 0.04 T Flaw of Aspect Ratio 6 in a BWR Subjected to the PT100-50 Cooldown, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 69

3.3.3 PT50-50 Cooldown 3.3.3.1 1/4-T Flaw Results The stress intensity factor history for a 0.25 T ID surface flaw subjected to the PT50-50 cooldown transient described in 3.1.2.3 is shown in Figure 3-22 for flaws with aspect ratios of 2, 6, 10 and infinity.

The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. It is important to note that all probabilistic analyses were performed with a flaw of aspect ratio equal to 6. Figure 3-22 clearly shows that K increases initially while the pressure is held constant and the temperature decreases linearly at a rate of 50°F/hour, reaching a maximum around 488 minutes. Then, at 488 minutes, the sudden pressure drop is accompanied by a sudden drop in K values. Finally, at 576 minutes, when the temperature reaches 70°F and remains constant, the rate of decrease in K values diminishes and K versus time curves plateau at their minimum value.

Figure 3-22: Stress Intensity Factor for the BWR PT50-50 Cooldown for a 0.25 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity A probabilistic calculation assuming a 0.25 T flaw of aspect ratio equal to 6 and the PT50-50 cooldown resulted in a CPI of 2.3E-09 and a CPF of 2.0E-14 using the RG 1.99, Rev. 2 embrittlement correlation.

This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient. Figure 3-23 shows the calculated CPI and CPF as a function of the number of RPV trials in FAVOR. The CPI and CPF values have not converged, but the order of magnitude is believed to be sufficiently accurate for the purposes of this study.

70

Figure 3-23: CPI and CPF versus Number of Trials for a 0.25 T Flaw in a BWR with 0°F RTNDT Shift for the PT50-50 cooldown To produce FAVOR runs corresponding to a single ID surface flaw of depth 0.25 T and aspect ratio 6, the VFLAW input format was used to specify the flaw population as follows:

• The surface flaw density was set to result in a single flaw of depth 0.25 T and aspect ratio 6, with zeros everywhere else

• The embedded weld flaw density was set to zero for all flaw depths and aspect ratios

• The embedded plate flaw density was set to zero for all flaw depths and aspect ratios The same PT50-50 cooldown following the P-T limit determined using the RG 1.99, Rev. 2 embrittlement correlation (shown in Figure 3-4) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. Of course, in actual safety analyses, the P-T limit would be adjusted as a function of the maximum ART in the vessel being analyzed, so a higher ART would reduce the allowed pressure, which In turn would result in significantly lower CPI and CPF than presented here for cases where is greater than 0°F.

71

Table 3-8: Calculated CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a BWR Subjected to the PT50-50 Cooldown RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 0 2.3475E-09 NO 1.9693E-14 NO 1.9693E-21 6 4.7785E-09 NO 1.2400E-12 NO 1.2400E-19 15 1.2570E-08 NO 1.0564E-10 YES 1.0564E-17 47 3.6727E-07 YES 1.2292E-07 YES 1.2292E-14 128 2.3713E-04 YES 2.1340E-04 YES 2.1340E-11 Figure 3-24: CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a BWR Subjected to the PT50-50 Cooldown at the P-T Limit, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 72

Table 3-8 and Figure 3-24 show the results of the probabilistic analyses for the 0.25 T flaw for the PT50-50 cooldown, as a function of . As expected, the CPI and CPF increase as increases. It should be noted that the analyses with 15° and with 6° did not statistically converge, respectively for CPI and CPF, because of the very low calculated probabilities. However, the CPI and CPF trend over the number of trials was beginning to stabilize, so the results are believed to be in the correct order of magnitude. Overall, it can be said that CPI and CPF for BWRs with a 1/4 T flaw are at least 1 order of magnitude lower than for PWRs for the PT50-50 cooldown.

3.3.3.2 Shallow Flaw Results The stress intensity factor history for a 0.04 T ID surface flaw subjected to the PT50-50 cooldown transient described in 3.1.2.3 is shown in Figure 3-25 for flaws with aspect ratios of 2, 6, 10 and infinity.

The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. Figure 3-25 clearly shows that K increases initially while the pressure is held constant and the temperature decreases linearly at a rate of 50°F/hour. The maximum K values are reached at 488 minutes, just before the sudden pressure drop that results in a sudden drop in K values. During the subsequent pressure and temperature decrease, the K values reach a minimum sometime between 640 and 660 minutes, and then remain at this minimum value. It should be noted that a small inflection point can be observed in the decreasing K trend at 576 minutes, where the temperature reaches 70°F and subsequently remains there.

A probabilistic calculation assuming a 0.04 T flaw of aspect ratio equal to 6 and the 100°F/hour P-T limit cooldown resulted in a CPI and CPF of zero using the RG 1.99, Rev. 2 embrittlement correlation. This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient.

Figure 3-25 Stress Intensity Factor for the BWR PT50-50 Cooldown for a 0.04 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity 73

To produce FAVOR runs corresponding to a single ID surface flaw of depth 0.04 T and aspect ratio 6, the VFLAW input format was used to specify the flaw population as follows:

• The surface flaw density was set to result in a single flaw of depth 0.04 T and aspect ratio 6, with zeros everywhere else

• The embedded weld flaw density was set to zero for all flaw depths and aspect ratios

• The embedded plate flaw density was set to zero for all flaw depths and aspect ratios The same PT50-50 cooldown following the P-T limit determined using the RG 1.99, Rev. 2 embrittlement correlation for a 0.25 T flaw (shown in Figure 3-4) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. That is, a new P-T limit was not calculated for a 0.04 T shallow flaw, but was instead calculated using current practice based on ASME Section XI, Appendix G, which requires analysis based on a 0.25 T flaw. P-T Furthermore, as for the 0.25 T case, in actual safety analyses, the P-T limit would be adjusted as a function of the maximum ART in the vessel being analyzed, which would result in a different CPI and CPF than presented here. If the P-T limit was recalculated for each value, the shallow flaw effect makes it difficult to predict a priori whether the CPI and CPF values would move up or down.

Table 3-9 and Figure 3-26 show the results of the probabilistic analyses for the 0.04 T flaw for the PT50-50 cooldown, as a function of . As expected, the CPI and CPF increase as increases. It should be noted that the analyses with < 47° did not statistically converge because of the very low calculated probabilities. However, the CPI and CPF trend over the number of trials was beginning to stabilize, so the results are believed to be in the correct order of magnitude. Overall, it can be said that CPI and CPF for BWRs with a 0.04 T flaw are several orders of magnitude lower than for PWRs for the PT50-50 cooldown.

Table 3-9: Calculated CPI and CPF for a 0.04 T Flaw of Aspect Ratio 6 in a BWR Subjected to the PT50-50 Cooldown RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 0 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 6 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 15 9.8045E-14 NO 0.0000E+00 YES 0.0000E+00 47 8.6349E-11 NO 1.6261E-12 NO 1.6261E-19 128 5.5078E-08 YES 8.2949E-09 YES 8.2949E-16 74

Figure 3-26: CPI and CPF for a 0.04 T Flaw of Aspect Ratio 6 in a BWR Subjected to the PT50-50 Cooldown, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 75

3.3.4 S50 Cooldown 3.3.4.1 1/4-T Flaw Results The stress intensity factor history for a 0.25 T ID surface flaw subjected to the 50°F/hour saturation cooldown transient described in 3.1.2.4 is shown in Figure 3-27 for flaws with aspect ratios of 2, 6, 10 and infinity. The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. It is important to note that all probabilistic analyses were performed with a flaw of aspect ratio equal to 6. Figure 3-27 shows that K decreases as both the pressure and the temperature decrease along the saturation curve. A local minimum is reached between 400 and 500 minutes, with a very small subsequent increase in K values between the time of the minimum and 576 minutes, at which point the temperature reaches 70°F and remains constant.

After 576 minutes, K begins to decrease again as the portion of K due to thermal stresses decreases.

After this point, the rate of decrease in K values slowly diminishes until the end of the calculation.

Figure 3-27: Stress Intensity Factor for BWR Saturation Cooldown for 0.25 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity A probabilistic calculation assuming a 0.25 T flaw of aspect ratio equal to 6 and the 50°F/hour saturation cooldown resulted in a CPI and a CPF of zero using the RG 1.99, Rev. 2 embrittlement correlation. This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient.

The same 50°F/hour saturation cooldown (shown in Figure 3-5) was modeled for different values of

, so as to understand the impact of shifting the RTNDT for this transient. Table 3-10 shows the results of the probabilistic analyses for the 0.25 T flaw for the 50°F/hour saturation cooldown, as a function of . All values of calculated CPI and CPF were equal to zero.

76

Table 3-10: Calculated CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a BWR Subjected to Saturation Cooldown RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 0 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 6 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 15 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 47 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 128 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 3.3.4.2 Shallow Flaw Results The stress intensity factor history for a 0.04 T ID surface flaw subjected to the P-T limit cooldown transient described in 3.1.2.4 is shown in Figure 3-28 for flaws with aspect ratios of 2, 6, 10 and infinity.

The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. Figure 3-28 shows that K increases slightly from 0 to ~20 minutes, and then decreases as both the pressure and the temperature decrease along the saturation curve. A minimum is reached between 200 and 300 minutes, with a subsequent increase in K values between the time of the minimum and 576 minutes, at which point the temperature reaches 70°F and remains constant. A second peak in K values is observed at 576 minutes that can sometimes exceed the value of the first peak early in the transient. This second peak is characteristic of shallow surface flaws that barely penetrate the base metal because these cracks are highly affected by the stresses due to the mismatch in thermal expansion coefficient between the cladding and the base metal. After 576 minutes, the temperature is held constant at 70°F, and thus the thermal stresses and associated component of K decrease rapidly, hence there is a decrease in overall K values despite the constant pressure.

77

Figure 3-28: Stress Intensity Factor for BWR 50°F/hour Saturation Cooldown for 0.04 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity A probabilistic calculation assuming a 0.04 T flaw of aspect ratio equal to 6 and the 50°F/hour saturation cooldown resulted in a CPI and a CPF of zero using the RG 1.99, Rev. 2 embrittlement correlation. This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient.

The same 50°F/hour saturation cooldown (shown in Figure 3-5) was modeled for different values of

, so as to understand the impact of shifting the RTNDT for this transient. Table 3-11 shows the results of the probabilistic analyses for the 0.25 T flaw for the 50°F/hour saturation cooldown, as a function of . All values of calculated CPI and CPF were equal to zero.

Table 3-11: Calculated CPI and CPF for a 0.04 T Flaw of Aspect Ratio 6 in a BWR Subjected to a Saturation Cooldown RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 0 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 6 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 15 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 47 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 128 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 78

3.3.5 LT40-40 Leak Test 3.3.5.1 1/4-T Flaw Results The stress intensity factor history for a 0.25 T ID surface flaw subjected to the LT40-40 leak test transient described in 3.1.2.5 is shown in Figure 3-29 for flaws with aspect ratios of 2, 6, 10 and infinity. The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. It is important to note that all probabilistic analyses were performed with a flaw of aspect ratio equal to 6. Figure 3-29 shows that K increases as the temperature and pressure increase between 0 and 146 minutes. At 146 minutes, the pressure and temperature reach their maxima of 1.1 ksi and 167°F, respectively, and remain constant until 266 minutes (for 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />).

During this time, K increases slower to eventually stabilize at its maximum value, which takes about 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />. The temperature and pressure then decrease from 266 minutes until 411 minutes, at which point they are held more or less constant until the end of the transient. This results in a steady decrease in K.

The rate of decrease in K decreases over time, and becomes near zero by 450 minutes, at which point K reaches a minimum value roughly equal to its value at the start of the transient.

Figure 3-29: Stress Intensity Factor for the BWR LT40-40 Leak Test for a 0.25 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity 79

Figure 3-30: CPI and CPF versus Number of Trials for a 0.25 T Flaw in a BWR with 0°F RTNDT Shift for the LT40-40 Leak Test A probabilistic calculation assuming a 0.25 T flaw of aspect ratio equal to 6 and the PT40-40 leak test resulted in a CPI of 2.5E-13 and a CPF of zero using the RG 1.99, Rev. 2 embrittlement correlation. This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient. Figure 3-30 shows the calculated CPI as a function of the number of RPV trials in FAVOR. Although the CPI value has not quite converged, the order of magnitude of CPI is believed to be sufficiently accurate for the purposes of this study.

The same LT40-40 leak test (shown in Figure 3-6) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. Table 3-12 and Figure 3-31 show the results of the probabilistic analyses for the 0.25 T flaw for the LT40-40 leak test, as a function of

. As expected, the CPI and CPF increase as increases. It should be noted that the analyses for 15° resulted in a CPF prediction equal to zero. Furthermore, for 47° , the CPI did not statistically converge because of the low calculated probability, but the CPI trend over the number of trials was beginning to stabilize, so the results are believed to be in the correct order of magnitude.

80

Table 3-12: Calculated CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a BWR Subjected to the LT40-40 Leak Test RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 0 2.5060E-13 NO 0.0000E+00 YES 0.0000E+00 6 1.1497E-11 NO 0.0000E+00 YES 0.0000E+00 15 1.4125E-10 NO 0.0000E+00 YES 0.0000E+00 47 8.7213E-09 NO 2.2284E-11 YES 2.2284E-14 128 1.1423E-05 YES 7.9051E-06 YES 7.9051E-09 Figure 3-31: CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a BWR Subjected to the LT40-40 Leak Test, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 81

3.3.5.2 Shallow Flaw Results The stress intensity factor history for a 0.04 T ID surface flaw subjected to the LT40-40 leak test transient described in 3.1.2.5 is shown in Figure 3-32 for flaws with aspect ratios of 2, 6, 10 and infinity. The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. Figure 3-32 shows that K initially decreases slightly and then increases as the temperature and pressure increase between 0 and 146 minutes. The initial decrease in K can be explained by the fact that the shallow flaw is subjected to compressive thermal stresses as the temperature increases, resulting in a decrease in the thermal K component. This decrease is initially greater than the increase in pressure induced membrane K, resulting in a decrease in total K until the pressure induced membrane K component increases faster than the thermal K component decreases.

At 146 minutes, the pressure and temperature reach their maxima of 1.1 ksi and 167°F, respectively, and remain constant until 266 minutes (for 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />). During this time, K increases slower to eventually stabilize at its maximum value, which takes about 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />. The temperature and pressure then decrease from 266 minutes until 411 minutes, at which point they are held more or less constant until the end of the transient. This results in a steady decrease in K. The decrease in K displays an inflection point at 411 minutes, due to the temperature being held constant after that time. At 450 minutes, K reaches a minimum value roughly equal to its value at the start of the transient.

Figure 3-32: Stress Intensity Factor for the BWR LT40-100 Leak Test for a 0.04 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity A probabilistic calculation assuming a 0.04 T flaw of aspect ratio equal to 6 and the LT40-40 leak test resulted in a CPI and a CPF of zero using the RG 1.99, Rev. 2 embrittlement correlation. This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient.

82

The same LT40-40 leak test (shown in Figure 3-6) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. Table 3-13 and Figure 3-33 show the results of the probabilistic analyses for the 0.04 T flaw for the LT40-40 leak test, as a function of

. As expected, the CPI and CPF increase as increases. For 47°, the CPI and CPF were zero.

Table 3-13: Calculated CPI and CPF for a 0.04 T Flaw of Aspect Ratio 6 in a BWR Subjected to the LT40-40 Leak Test RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 0 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 6 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 15 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 47 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 128 5.5368E-11 YES 3.8897E-12 YES 3.8897E-15 83

Figure 3-33: CPI and CPF for a 0.04 T Flaw of Aspect Ratio 6 in a BWR Subjected to the LT40-40 Leak Test, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 84

3.3.6 LT40-100 Leak Test 3.3.6.1 1/4-T Flaw Results The stress intensity factor history for a 0.25 T ID surface flaw subjected to the LT40-100 leak test transient described in 3.1.2.6 is shown in Figure 3-34 for flaws with aspect ratios of 2, 6, 10 and infinity.

The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. It is important to note that all probabilistic analyses were performed with a flaw of aspect ratio equal to 6. Figure 3-34 shows that K increases as the temperature and pressure increase between 0 and 146 minutes. At 146 minutes, the pressure and temperature reach their maxima of 1.1 ksi and 167°F, respectively, and remain constant until 266 minutes (for 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />). During this time, K increases slower to eventually stabilize at its maximum value, which takes about 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />. The temperature and pressure then decrease from 266 minutes until 324 minutes, at which point they are held more or less constant until the end of the transient. This results in a steady decrease in K. The rate of decrease in K decreases over time, and becomes near zero by 380 minutes, at which point K reaches a minimum value roughly equal to its value at the start of the transient.

Figure 3-34: Stress Intensity Factor for the BWR LT40-100 Leak Test for a 0.25 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity A probabilistic calculation assuming a 0.25 T flaw of aspect ratio equal to 6 and the LT40-100 leak test resulted in a CPI and a CPF of zero using the RG 1.99, Rev. 2 embrittlement correlation. This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient.

The same LT40-100 leak test (shown in Figure 3-7) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. Table 3-14 and Figure 3-35 show the results of the probabilistic analyses for the 0.25 T flaw for LT40-100 the leak test, as a function of

. As expected, the CPI and CPF increase as increases. It should be noted that the 85

analyses for 15° resulted in a CPF prediction equal to zero. Furthermore, for 15° , the CPI did not statistically converge because of the low calculated probability, but the CPI trend over the number of trials was beginning to stabilize, so the results are believed to be in the correct order of magnitude.

Table 3-14: Calculated CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a BWR Subjected to the LT40-100 Leak Test RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 0 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 6 1.0905E-11 NO 0.0000E+00 YES 0.0000E+00 15 1.3700E-10 NO 0.0000E+00 YES 0.0000E+00 47 8.5617E-09 YES 1.9619E-11 YES 1.9619E-14 128 1.1188E-05 YES 7.7124E-06 YES 7.7124E-09 86

Figure 3-35: CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a BWR Subjected to the LT40-100 Leak Test, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 3.3.6.2 Shallow Flaw Results The stress intensity factor history for a 0.04 T ID surface flaw subjected to the LT40-100 leak test transient described in 3.1.2.6 is shown in Figure 3-36 for flaws with aspect ratios of 2, 6, 10 and infinity.

The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. Figure 3-36 shows that K initially decreases slightly and then increases as the temperature and pressure increase between 0 and 146 minutes. The initial decrease in K can be explained by the fact that the shallow flaw is subjected to compressive thermal stresses as the temperature increases, resulting in a decrease in the thermal K component. This decrease is initially greater that the increase in pressure induced membrane K, resulting in a decrease in total K until the pressure induced membrane K component increases faster than the thermal K component decreases.

87

At 146 minutes, the pressure and temperature reach their maxima of 1.1 ksi and 167°F, respectively, and remain constant until 266 minutes (for 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />). During this time, K increases slower to eventually stabilize at its maximum value, which takes about 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />. The temperature and pressure then decrease from 266 minutes until 324 minutes, at which point they are held more or less constant until the end of the transient. This results in a steady decrease in K. The decrease in K displays an inflection point at 324 minutes, due to the temperature being held constant after that time. At 380 minutes, K reaches a minimum value roughly equal to its value at the start of the transient.

Figure 3-36: Stress Intensity Factor for the BWR LT40-100 Leak Test for a 0.04 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity A probabilistic calculation assuming a 0.04 T flaw of aspect ratio equal to 6 and the LT40-100 leak test resulted in a CPI and a CPF of zero using the RG 1.99, Rev. 2 embrittlement correlation. This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient.

The same LT40-100 leak test (shown in Figure 3-7) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. Table 3-15 and Figure 3-37 show the results of the probabilistic analyses for the 0.04 T flaw for the LT40-100 leak test, as a function of

. As expected, the CPI and CPF increase as increases. For 15°, the CPI was zero and for 47°, CPF was zero. Furthermore, for = 47°, the CPI did not statistically converge because of the low calculated probability, but the CPI trend over the number of trials was beginning to stabilize, so the results are believed to be in the correct order of magnitude.

88

Table 3-15: Calculated CPI and CPF for a 0.04 T Flaw of Aspect Ratio 6 in a BWR Subjected to the LT40-100 Leak Test RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 0 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 6 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 15 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 47 1.2425E-16 NO 0.0000E+00 YES 0.0000E+00 128 4.4753E-10 YES 1.4824E-11 YES 1.4824E-14 Figure 3-37: CPI and CPF for a 0.04 T Flaw of Aspect Ratio 6 in a BWR Subjected to the LT40-100 Leak Test, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 89

3.3.7 LTA Leak Test 3.3.7.1 1/4-T Flaw Results The stress intensity factor history for a 0.25 T ID surface flaw subjected to the LTA leak test transient described in 3.1.2.7 is shown in Figure 3-38 for flaws with aspect ratios of 2, 6, 10 and infinity. The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. It is important to note that all probabilistic analyses were performed with a flaw of aspect ratio equal to 6. Figure 3-38 shows that K is initially constant along while the temperature and pressure are held constant. When the temperature increases a little after 400 minutes, K become negative as the pressure is still zero and the temperature increase puts the inner surface flaw in compression. Around 825 minutes, the pressure increases rapidly up to the leak test pressure of 1.0 ksi, resulting in a corresponding increase in K to its maximum value. The pressure is momentarily reduced around 1100 minutes, and then raised back to 1.0 ksi. The K values decrease and then come back up accordingly. Around 1250 minutes, the pressure is dropped back to zero, and K decreases accordingly. Finally, as the temperature decreases (initially rapidly and then slower) beginning at 1450 minutes, a rapid rise in K is observed followed by a rapid decrease as the temperature is dropped more slowly. The K values then slowly stabilize near their values at the start of the transient.

Figure 3-38: Stress Intensity Factor for the BWR LTA Leak Test for a 0.25 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity A probabilistic calculation assuming a 0.25 T flaw of aspect ratio equal to 6 and the LTA leak test resulted in a CPI and a CPF of zero using the RG 1.99, Rev. 2 embrittlement correlation. This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient.

The same LTA leak test (shown in Figure 3-8) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. Table 3-16 and Figure 3-39 show the 90

results of the probabilistic analyses for the 0.25 T flaw for the LTA leak test, as a function of . As expected, the CPI and CPF increase as increases. It should be noted that the analyses for 47° resulted in a CPF prediction equal to zero and analyses for 15° resulted in a CPI prediction equal to zero. Furthermore, for = 47° , the CPI did not statistically converge because of the low calculated probability, but the CPI trend over the number of trials was beginning to stabilize, so the results are believed to be in the correct order of magnitude.

Table 3-16: Calculated CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a BWR Subjected to the LTA Leak Test RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 0 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 6 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 15 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 47 2.6317E-10 NO 0.0000E+00 YES 0.0000E+00 128 5.3999E-07 YES 9.5373E-08 YES 9.5373E-08 91

Figure 3-39: CPI and CPF for a 0.25 T Flaw of Aspect Ratio 6 in a BWR Subjected to the LTA Leak Test, as a Function of RTNDT Shift from RG 1.99, Rev. 2 to E90015 Embrittlement Correlations (open symbols linked by dashed lines represent results that are not statistically converged) 3.3.7.2 Shallow Flaw Results The stress intensity factor history for a 0.04 T ID surface flaw subjected to the LTA leak test transient described in 3.1.2.7 is shown in Figure 3-40 for flaws with aspect ratios of 2, 6, 10 and infinity. The arrest toughness Kia, below which no crack growth initiation can occur, as well as the median fracture toughness Kic, are also shown. Figure 3-40 shows that K is initially constant along while the temperature and pressure are held constant. When the temperature increases a little after 400 minutes, K decreases sharply as the pressure is still zero and the temperature increase puts the inner surface flaw in compression, thus offsetting some of the tensile residual stresses affecting the shallow flaw. Around 825 minutes, the pressure increases rapidly up to the leak test pressure of 1.0 ksi, resulting in a corresponding increase in K to its maximum value. The pressure is momentarily reduced around 1100 92

minutes, and then raised back to 1.0 ksi. The K values decrease and then come back up accordingly.

Around 1250 minutes, the pressure is dropped back to zero, and K decreases accordingly. Finally, as the temperature decreases (initially rapidly and then slower) beginning at 1450 minutes, a rapid rise in K is observed followed by a rapid decrease as the temperature is dropped more slowly. The K values then slowly stabilize near their values at the start of the transient.

Figure 3-40: Stress Intensity Factor for the BWR LTA Leak Test for a 0.04 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity A probabilistic calculation assuming a 0.04 T flaw of aspect ratio equal to 6 and the LTA leak test resulted in a CPI and a CPF of zero using the RG 1.99, Rev. 2 embrittlement correlation. This case corresponds to 0°F RTNDT shift (i.e. no shift) for this transient.

The same LTA leak test (shown in Figure 3-8) was modeled for different values of , so as to understand the impact of shifting the RTNDT for this transient. Table 3-17 shows the results of the probabilistic analyses for the 0.045 T flaw for the LTA leak test, as a function of . All values of calculated CPI and CPF were equal to zero.

93

Table 3-17: Calculated CPI and CPF for a 0.04 T Flaw of Aspect Ratio 6 in a BWR Subjected to the LTA Leak Test RTNDT Shift (°F) CPI CPI Converged? CPF CPF Converged? TWCF

-40 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 0 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 6 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 15 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 47 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 128 0.0000E+00 YES 0.0000E+00 YES 0.0000E+00 94

4 Summary Figure 4-1 through Figure 4-7 show the calculated CPI and CPF calculated for all the analyses performed in this study. The PT100-100, PT100-50, and PT50-50 cooldowns were simulated for both PWR and BWR. The C50P cooldown was only simulated for PWR. Finally, the LT40-40, LT40-100, and LTA leak tests, as well as the S50 saturation cooldown were only simulated for BWR. Since all the predictions were zero for the BWR S50 saturation cooldown, no summary figure is shown for this transient. As expected, in all cases resulting in non-zero predictions, the calculated CPI and CPF increase as a function of the RTNDT shift associated with going from RG 1.99, Rev. 2 to E90015 for all combinations studied.

95

Figure 4-1: CPI and CPF for PWR and BWR, 1/4 T and Shallow Flaws, as a Function of RTNDT Shift for the PT100-100 Cooldown Figure 4-1 shows that the CPI and CPF associated with the PT100-100 cooldown can be high and exceed 1x10-6 for relatively small RTNDT shift amounts (~50°F). This is particularly true for the case of the 1/4T flaw in PWR and BWR, and for the PWR shallow flaw, which produce the higher CPI and CPF for this transient. However, since the PT100-100 cooldown is estimated to have a very low frequency of occurrence of 6x10-6 events per year for PWRs and 1x10-7 events per year for BWRs per BTP 5-3 [7], the overall TWCF (though-Wall Crack Frequency) remains well below the 1x10-6 per year limit despite the high CPI and CPF calculated in some cases. It is also interesting to note that for a given set of conditions, all CPI and CPF predictions for BWR are lower than those for PWR, which was expected for this transient.

96

Figure 4-2: CPI and CPF for PWR and BWR, 1/4 T and Shallow Flaws, as a Function of RTNDT Shift for the PT100-50 Cooldown Figure 4-2 shows that the CPI and CPF associated with the PT100-50 cooldown is generally lower than for the 100°F/hour P-T limit cooldown, except for the case of PWR shallow flaws. In fact, for PWR shallow flaws, even for 0°F RTNDT shift, the calculated CPI and CPF are above 1x10-6. Furthermore, it is interesting to note that for both BWR and PWR, the CPI and CPF associated with shallow flaws exceed the CPI and CPF associated with 1/4 T flaws by several orders of magnitude for this transient. However, since the PT100-50 cooldown is estimated to have a very low frequency of occurrence of 6x10-6 events per year for PWRs and 1x10-7 events per year for BWRs per BTP 5-3 [7], the overall TWCF (though-Wall Crack Frequency) remains well below the 1x10-6 per year limit despite the high CPI and CPF calculated in some cases. Once again, all CPI and CPF predictions for BWR are lower than those for PWR, which was expected for this transient.

97

Figure 4-3: CPI and CPF for PWR and BWR, 1/4 T and Shallow Flaws, as a Function of RTNDT Shift for the PT50-50 Cooldown Figure 4-3 shows that the CPI and CPF associated with the PT50-50 cooldown can be high and exceed 1x10-6 for relatively small RTNDT shift amounts ( from 50°F to 75°F). This is particularly true for the case of the 1/4T flaw in PWR and BWR, and for the PWR shallow flaw, which produce the higher CPI and CPF for this transient. However, since the PT50-50 cooldown is estimated to have a very low frequency of occurrence of 6x10-6 events per year for PWRs and 1x10-7 events per year for BWRs per BTP 5-3 [7], the overall TWCF (though-Wall Crack Frequency) remains well below the 1x10-6 per year limit despite the high CPI and CPF calculated in some cases. Once again, all CPI and CPF predictions for BWR are lower than those for PWR, which was expected for this transient.

98

Figure 4-4: CPI and CPF for PWR, 1/4 T and Shallow Flaws, as a Function of RTNDT Shift for the C50P Cooldown and for 42 Other PWR Cooldown Transients Simulated for a Shallow Flaw at RTNDT Shift of 100°F Figure 4-4 shows the calculated CPI and CPF for a PWR subjected to the C50P cooldown described in 2.1.2.4. It should be noted that all CPF values calculated were zero for this transient and thus are not shown on this figure. The frequency of occurrence of the C50P cooldown can be assumed to be 1.0 event per year since this is a realistic PWR cooldown, based on BTP 5-3 [7]. As a result, the CPI and CPF are equal to the annual frequency of crack initiation and TWCF, respectively. This figure shows that no matter how high the RTNDT shift applied within the range studied, the CPI are below 1E-07, and the CPF is zero, thus implying TWCF=0 events per year for this transient. Also shown on Figure 4-4 are the CPI And CPF values calculated for 42 actual PWR cooldown transients at a RTNDT shift value of 100°F. The CPI values for these 42 actual cooldowns range from ~5x10-12 to ~2x10-5 and most CPF values calculated 99

were zero, with a maximum of ~1x10-10, indicating a TWCF of ~1x10-10 for the worst case actual cooldown modeled in this study.

The cooldown transient having a frequency of occurrence equal to 1.0 for BWRs is the saturation cooldown, and in that sense, the S50 cooldown for a BWR is analog to the C50P cooldown for a PWR. All simulations performed in this study with the S50 cooldown for a BWR resulted in CPI and CPF of zero, implying a TWCF of zero per year for a saturation cooldown in a BWR.

Figure 4-5: CPI and CPF for BWR, 1/4 T and Shallow Flaws, as a Function of RTNDT Shift for the LT40-40 Leak Test Figure 4-5 shows the CPI and CPF for a BWR subjected to the LT40-40 leak test transient described in 3.1.2.5. For this transient, the 1/4 T flaw CPI and CPF are orders of magnitude larger than the shallow flaw 100

CPI and CPF, and both the CPI and CPF can reach values around 1x10-5 for high values of RTNDT shift.

However, since the frequency of occurrence of the LT40-40 leak test is estimated to be 1x10-3 event per year per BTP 5-3 [7], the overall TWCF (though-Wall Crack Frequency) remains well below the 1x10-6 per year limit despite the high CPI and CPF calculated in some cases.

Figure 4-6: CPI and CPF for BWR, 1/4 T and Shallow Flaws, as a Function of RTNDT Shift for the LT40-100 Leak Test Figure 4-6 shows the CPI and CPF for a BWR subjected to the LT40-100 leak test transient described in 3.1.2.6. The results are very similar to those for the LT40-40 leak test. For this transient, the 1/4 T flaw CPI and CPF are orders of magnitude larger than the shallow flaw CPI and CPF, and both the CPI and CPF can reach values around 1x10-5 for high values of RTNDT shift. However, since the frequency of occurrence of the LT40-40 leak test is estimated to be 1x10-3 event per year per BTP 5-3 [7], the overall 101

TWCF (though-Wall Crack Frequency) remains well below the 1x10-6 per year limit despite the high CPI and CPF calculated in some cases.

Figure 4-7: CPI and CPF for BWR, 1/4 T and Shallow Flaws, as a Function of RTNDT Shift for the LTA Leak Test Figure 4-7 shows the CPI and CPF for a BWR subjected to the LTA leak test transient described in 3.1.2.7.

The frequency of occurrence of the leak test can be assumed to be 1 event per year since this is an actual test that was performed at a plant and because leak testing must be performed as part of return to power operations in BWRs. As a result, the CPI and CPF are equal to the annual frequency of crack initiation and TWCF, respectively. This figure shows that for 1/4 T flaws, a high value of RTNDT shift of about 128°F, corresponding to the 99th percentile of RTNDT shift values (see 3.2), can result in a TWCF of 1x10-7 events per year, which is still below the 1x10-6 limit on highest allowable TWCF per the regulations. Interestingly, shallow flaws resulted in all values of CPI and CPF being zero, which indicates that shallow flaws are not a complicating factor for BWR leak tests. In fact, 1/4 T flaws result in much 102

higher CPI and CPF than shallow flaws do for BWR leak tests, indicating that the BWR leak test issue is not related to the presence of shallow flaws. This is an important point because the BWR leak test issue and the shallow flaw issue have often been incorrectly linked.

In summary, with the modeling assumptions used in this study, the transients resulting in the highest TWCF values are PWR cooldowns along the P-T limit for 1/4 T flaws at high values of RTNDT shift. However, the highest calculated TWCF still remained below 1x10-6 per year.

For BWRs, the transients resulting in the highest TWCF are leak tests, with TWCF reaching close to 1x10-7 for an actual BWR leak test and TWCF reaching close to 1x10-8 for leak tests at the P-T limit. However, once again, even the highest calculated TWCF still remained below 1x10-6 per year.

Realistic cooldown transients resulted in the lowest TWCF values, with most being zero, and the highest reaching around 1x10-10 per year for one of the actual PWR cooldowns modeled.

103

5 References

[1] U.S. Nuclear Regulatory Commission, Regulatory Guide 1.99, Revision 2: Radiation Embrittlement of Reactor Vessel Materials (ML003740284), Washington, DC: U.S. NRC, May 1988.

[2] ASTM International, ASTM E900-15: Standard Guide for Predicting Radiation-Induced Transition Temperature Shift in Reactor Vessel Materials, West Conshohocken, PA: ASTM International, 2015.

[3] B. R. Bass, T. L. Dickson, P. T. Williams, H. B. Klasky and R. H. Dodds, "ORNL/TM-2015/59531/REV-01: The Effect of Shallow Inside-Surface-Breaking Flaws on the Probability of Brittle Fracture of Reactors Subjected to Postulated and Actual Operational Cool-Down Transients: A Status Report,"

Oak Ridge National Laboratory, Oak Ridge, TN, 2015.

[4] P. T. Williams, T. L. Dickson, B. R. Bass and H. B. Klasky, "ORNL/LTR-2016/309: Fracture Analysis of Vessels - Oak Ridge FAVOR, v16.1, Computer Code: Theory and Implementation of Algorithms, Methods, and Correlations," Oak Ridge National Laboratory, Oak Ridge, TN, 2016.

[5] T. L. Dickson, P. T. Williams, B. R. Bass and H. B. Klasky, "ORNL/LTR-2016/310: Fracture Analysis of Vessels - Oak Ridge - FAVOR, v16.1, Computer Code:Users Guide," Oak Ridge National Laboratory, 2016.

[6] The American Society of Mechanical Engineers, ASME BPVC.XI-2019, Nonmandatory Appendix G:

Fracture Toughness Criteria for Protection Against Failure, New York, NY: ASME, 2019.

[7] U.S. Nuclear Regulatory Commission, Closure Memorandum Supporting the Limited Revision of NUREG-0800 Branch Technical Position 5-3 "Fracture Toughness Requirements" (ML16364A285),

Washington, DC: U.S NRC, April 11, 2017.

104

6 Appendix A: List of FAVOR Analyses 6.1 PWR Analyses Reactor Type Flaw RTNDT Shift Transient Run Type Folder Case Name PWR 025T 0F P0_PT100 Deterministic PWR 025T 0F PWR_025T_0F_P0_PT100 PWR 025T -40F PT100-100 Deterministic PWR 025T -40F PWR_025T_-40F_PT100-100 PWR 025T -40F PT100-100 Probabilistic PWR 025T -40F PWR_025T_-40F_PT100-100 PWR 003T -40F PT100-100 Deterministic PWR 003T -40F PWR_003T_-40F_PT100-100 PWR 003T -40F PT100-100 Probabilistic PWR 003T -40F PWR_003T_-40F_PT100-100 PWR 025T 0F PT100-100 Deterministic PWR 025T 0F PWR_025T_0F_PT100-100 PWR 025T 0F PT100-100 Probabilistic PWR 025T 0F PWR_025T_0F_PT100-100 PWR 003T 0F PT100-100 Deterministic PWR 003T 0F PWR_003T_0F_PT100-100 PWR 003T 0F PT100-100 Probabilistic PWR 003T 0F PWR_003T_0F_PT100-100 PWR 025T +40F PT100-100 Deterministic PWR 025T +40F PWR_025T_+40F_PT100-100 PWR 025T +40F PT100-100 Probabilistic PWR 025T +40F PWR_025T_+40F_PT100-100 PWR 003T +40F PT100-100 Deterministic PWR 003T +40F PWR_003T_+40F_PT100-100 PWR 003T +40F PT100-100 Probabilistic PWR 003T +40F PWR_003T_+40F_PT100-100 PWR 025T +78F PT100-100 Deterministic PWR 025T +78F PWR_025T_+78F_PT100-100 PWR 025T +78F PT100-100 Probabilistic PWR 025T +78F PWR_025T_+78F_PT100-100 PWR 003T +78F PT100-100 Deterministic PWR 003T +78F PWR_003T_+78F_PT100-100 PWR 003T +78F PT100-100 Probabilistic PWR 003T +78F PWR_003T_+78F_PT100-100 PWR 025T +131F PT100-100 Deterministic PWR 025T +131F PWR_025T_+131F_PT100-100 PWR 025T +131F PT100-100 Probabilistic PWR 025T +131F PWR_025T_+131F_PT100-100 PWR 003T +131F PT100-100 Deterministic PWR 003T +131F PWR_003T_+131F_PT100-100 PWR 003T +131F PT100-100 Probabilistic PWR 003T +131F PWR_003T_+131F_PT100-100 PWR 025T +193F PT100-100 Deterministic PWR 025T +193F PWR_025T_+193F_PT100-100 PWR 025T +193F PT100-100 Probabilistic PWR 025T +193F PWR_025T_+193F_PT100-100 PWR 003T +193F PT100-100 Deterministic PWR 003T +193F PWR_003T_+193F_PT100-100 105

Reactor Type Flaw RTNDT Shift Transient Run Type Folder Case Name PWR 003T +193F PT100-100 Probabilistic PWR 003T +193F PWR_003T_+193F_PT100-100 PWR 025T 0F P0_PT50 Deterministic PWR 025T 0F PWR_025T_0F_P0_PT50 PWR 025T -40F PT100-50 Deterministic PWR 025T -40F PWR_025T_-40F_PT100-50 PWR 025T -40F PT100-50 Probabilistic PWR 025T -40F PWR_025T_-40F_PT100-50 PWR 003T -40F PT100-50 Deterministic PWR 003T -40F PWR_003T_-40F_PT100-50 PWR 003T -40F PT100-50 Probabilistic PWR 003T -40F PWR_003T_-40F_PT100-50 PWR 025T 0F PT100-50 Deterministic PWR 025T 0F PWR_025T_0F_PT100-50 PWR 025T 0F PT100-50 Probabilistic PWR 025T 0F PWR_025T_0F_PT100-50 PWR 003T 0F PT100-50 Deterministic PWR 003T 0F PWR_003T_0F_PT100-50 PWR 003T 0F PT100-50 Probabilistic PWR 003T 0F PWR_003T_0F_PT100-50 PWR 025T +40F PT100-50 Deterministic PWR 025T +40F PWR_025T_+40F_PT100-50 PWR 025T +40F PT100-50 Probabilistic PWR 025T +40F PWR_025T_+40F_PT100-50 PWR 003T +40F PT100-50 Deterministic PWR 003T +40F PWR_003T_+40F_PT100-50 PWR 003T +40F PT100-50 Probabilistic PWR 003T +40F PWR_003T_+40F_PT100-50 PWR 025T +78F PT100-50 Deterministic PWR 025T +78F PWR_025T_+78F_PT100-50 PWR 025T +78F PT100-50 Probabilistic PWR 025T +78F PWR_025T_+78F_PT100-50 PWR 003T +78F PT100-50 Deterministic PWR 003T +78F PWR_003T_+78F_PT100-50 PWR 003T +78F PT100-50 Probabilistic PWR 003T +78F PWR_003T_+78F_PT100-50 PWR 025T +131F PT100-50 Deterministic PWR 025T +131F PWR_025T_+131F_PT100-50 PWR 025T +131F PT100-50 Probabilistic PWR 025T +131F PWR_025T_+131F_PT100-50 PWR 003T +131F PT100-50 Deterministic PWR 003T +131F PWR_003T_+131F_PT100-50 PWR 003T +131F PT100-50 Probabilistic PWR 003T +131F PWR_003T_+131F_PT100-50 PWR 025T +193F PT100-50 Deterministic PWR 025T +193F PWR_025T_+193F_PT100-50 PWR 025T +193F PT100-50 Probabilistic PWR 025T +193F PWR_025T_+193F_PT100-50 PWR 003T +193F PT100-50 Deterministic PWR 003T +193F PWR_003T_+193F_PT100-50 PWR 003T +193F PT100-50 Probabilistic PWR 003T +193F PWR_003T_+193F_PT100-50 PWR 025T -40F PT50-50 Deterministic PWR 025T -40F PWR_025T_-40F_PT50-50 PWR 025T -40F PT50-50 Probabilistic PWR 025T -40F PWR_025T_-40F_PT50-50 PWR 003T -40F PT50-50 Deterministic PWR 003T -40F PWR_003T_-40F_PT50-50 PWR 003T -40F PT50-50 Probabilistic PWR 003T -40F PWR_003T_-40F_PT50-50 106

Reactor Type Flaw RTNDT Shift Transient Run Type Folder Case Name PWR 025T 0F PT50-50 Deterministic PWR 025T 0F PWR_025T_0F_PT50-50 PWR 025T 0F PT50-50 Probabilistic PWR 025T 0F PWR_025T_0F_PT50-50 PWR 003T 0F PT50-50 Deterministic PWR 003T 0F PWR_003T_0F_PT50-50 PWR 003T 0F PT50-50 Probabilistic PWR 003T 0F PWR_003T_0F_PT50-50 PWR 025T +40F PT50-50 Deterministic PWR 025T +40F PWR_025T_+40F_PT50-50 PWR 025T +40F PT50-50 Probabilistic PWR 025T +40F PWR_025T_+40F_PT50-50 PWR 003T +40F PT50-50 Deterministic PWR 003T +40F PWR_003T_+40F_PT50-50 PWR 003T +40F PT50-50 Probabilistic PWR 003T +40F PWR_003T_+40F_PT50-50 PWR 025T +78F PT50-50 Deterministic PWR 025T +78F PWR_025T_+78F_PT50-50 PWR 025T +78F PT50-50 Probabilistic PWR 025T +78F PWR_025T_+78F_PT50-50 PWR 003T +78F PT50-50 Deterministic PWR 003T +78F PWR_003T_+78F_PT50-50 PWR 003T +78F PT50-50 Probabilistic PWR 003T +78F PWR_003T_+78F_PT50-50 PWR 025T +131F PT50-50 Deterministic PWR 025T +131F PWR_025T_+131F_PT50-50 PWR 025T +131F PT50-50 Probabilistic PWR 025T +131F PWR_025T_+131F_PT50-50 PWR 003T +131F PT50-50 Deterministic PWR 003T +131F PWR_003T_+131F_PT50-50 PWR 003T +131F PT50-50 Probabilistic PWR 003T +131F PWR_003T_+131F_PT50-50 PWR 025T +193F PT50-50 Deterministic PWR 025T +193F PWR_025T_+193F_PT50-50 PWR 025T +193F PT50-50 Probabilistic PWR 025T +193F PWR_025T_+193F_PT50-50 PWR 003T +193F PT50-50 Deterministic PWR 003T +193F PWR_003T_+193F_PT50-50 PWR 003T +193F PT50-50 Probabilistic PWR 003T +193F PWR_003T_+193F_PT50-50 PWR 025T -40F C50P Deterministic PWR 025T -40F PWR_025T_-40F_C50P PWR 025T -40F C50P Probabilistic PWR 025T -40F PWR_025T_-40F_C50P PWR 003T -40F C50P Deterministic PWR 003T -40F PWR_003T_-40F_C50P PWR 003T -40F C50P Probabilistic PWR 003T -40F PWR_003T_-40F_C50P PWR 025T 0F C50P Deterministic PWR 025T 0F PWR_025T_0F_C50P PWR 025T 0F C50P Probabilistic PWR 025T 0F PWR_025T_0F_C50P PWR 003T 0F C50P Deterministic PWR 003T 0F PWR_003T_0F_C50P PWR 003T 0F C50P Probabilistic PWR 003T 0F PWR_003T_0F_C50P PWR 025T +40F C50P Deterministic PWR 025T +40F PWR_025T_+40F_C50P PWR 025T +40F C50P Probabilistic PWR 025T +40F PWR_025T_+40F_C50P 107

Reactor Type Flaw RTNDT Shift Transient Run Type Folder Case Name PWR 003T +40F C50P Deterministic PWR 003T +40F PWR_003T_+40F_C50P PWR 003T +40F C50P Probabilistic PWR 003T +40F PWR_003T_+40F_C50P PWR 025T +78F C50P Deterministic PWR 025T +78F PWR_025T_+78F_C50P PWR 025T +78F C50P Probabilistic PWR 025T +78F PWR_025T_+78F_C50P PWR 003T +78F C50P Deterministic PWR 003T +78F PWR_003T_+78F_C50P PWR 003T +78F C50P Probabilistic PWR 003T +78F PWR_003T_+78F_C50P PWR 025T +131F C50P Deterministic PWR 025T +131F PWR_025T_+131F_C50P PWR 025T +131F C50P Probabilistic PWR 025T +131F PWR_025T_+131F_C50P PWR 003T +131F C50P Deterministic PWR 003T +131F PWR_003T_+131F_C50P PWR 003T +131F C50P Probabilistic PWR 003T +131F PWR_003T_+131F_C50P PWR 025T +193F C50P Deterministic PWR 025T +193F PWR_025T_+193F_C50P PWR 025T +193F C50P Probabilistic PWR 025T +193F PWR_025T_+193F_C50P PWR 003T +193F C50P Deterministic PWR 003T +193F PWR_003T_+193F_C50P PWR 003T +193F C50P Probabilistic PWR 003T +193F PWR_003T_+193F_C50P PWR 003T +100F P1C1 Deterministic PWR 003T +100F PWR_003T_+100F_P1C1 PWR 003T +100F P1C1 Probabilistic PWR 003T +100F PWR_003T_+100F_P1C1 PWR 003T +100F P1C2 Deterministic PWR 003T +100F PWR_003T_+100F_P1C2 PWR 003T +100F P1C2 Probabilistic PWR 003T +100F PWR_003T_+100F_P1C2 PWR 003T +100F P1C3 Deterministic PWR 003T +100F PWR_003T_+100F_P1C3 PWR 003T +100F P1C3 Probabilistic PWR 003T +100F PWR_003T_+100F_P1C3 PWR 003T +100F P1C4 Deterministic PWR 003T +100F PWR_003T_+100F_P1C4 PWR 003T +100F P1C4 Probabilistic PWR 003T +100F PWR_003T_+100F_P1C4 PWR 003T +100F P2C1 Deterministic PWR 003T +100F PWR_003T_+100F_P2C1 PWR 003T +100F P2C1 Probabilistic PWR 003T +100F PWR_003T_+100F_P2C1 PWR 003T +100F P2C2 Deterministic PWR 003T +100F PWR_003T_+100F_P2C2 PWR 003T +100F P2C2 Probabilistic PWR 003T +100F PWR_003T_+100F_P2C2 PWR 003T +100F P3C1 Deterministic PWR 003T +100F PWR_003T_+100F_P3C1 PWR 003T +100F P3C1 Probabilistic PWR 003T +100F PWR_003T_+100F_P3C1 PWR 003T +100F P3C2 Deterministic PWR 003T +100F PWR_003T_+100F_P3C2 PWR 003T +100F P3C2 Probabilistic PWR 003T +100F PWR_003T_+100F_P3C2 108

Reactor Type Flaw RTNDT Shift Transient Run Type Folder Case Name PWR 003T +100F P4C1 Deterministic PWR 003T +100F PWR_003T_+100F_P4C1 PWR 003T +100F P4C1 Probabilistic PWR 003T +100F PWR_003T_+100F_P4C1 PWR 003T +100F P4C2 Deterministic PWR 003T +100F PWR_003T_+100F_P4C2 PWR 003T +100F P4C2 Probabilistic PWR 003T +100F PWR_003T_+100F_P4C2 PWR 003T +100F P4C3 Deterministic PWR 003T +100F PWR_003T_+100F_P4C3 PWR 003T +100F P4C3 Probabilistic PWR 003T +100F PWR_003T_+100F_P4C3 PWR 003T +100F P5C1 Deterministic PWR 003T +100F PWR_003T_+100F_P5C1 PWR 003T +100F P5C1 Probabilistic PWR 003T +100F PWR_003T_+100F_P5C1 PWR 003T +100F P5C2 Deterministic PWR 003T +100F PWR_003T_+100F_P5C2 PWR 003T +100F P5C2 Probabilistic PWR 003T +100F PWR_003T_+100F_P5C2 PWR 003T +100F P5C3 Deterministic PWR 003T +100F PWR_003T_+100F_P5C3 PWR 003T +100F P5C3 Probabilistic PWR 003T +100F PWR_003T_+100F_P5C3 PWR 003T +100F P5C4 Deterministic PWR 003T +100F PWR_003T_+100F_P5C4 PWR 003T +100F P5C4 Probabilistic PWR 003T +100F PWR_003T_+100F_P5C4 PWR 003T +100F P5C5 Deterministic PWR 003T +100F PWR_003T_+100F_P5C5 PWR 003T +100F P5C5 Probabilistic PWR 003T +100F PWR_003T_+100F_P5C5 PWR 003T +100F P6C1 Deterministic PWR 003T +100F PWR_003T_+100F_P6C1 PWR 003T +100F P6C1 Probabilistic PWR 003T +100F PWR_003T_+100F_P6C1 PWR 003T +100F P7C1 Deterministic PWR 003T +100F PWR_003T_+100F_P7C1 PWR 003T +100F P7C1 Probabilistic PWR 003T +100F PWR_003T_+100F_P7C1 PWR 003T +100F P7C2 Deterministic PWR 003T +100F PWR_003T_+100F_P7C2 PWR 003T +100F P7C2 Probabilistic PWR 003T +100F PWR_003T_+100F_P7C2 PWR 003T +100F P7C3 Deterministic PWR 003T +100F PWR_003T_+100F_P7C3 PWR 003T +100F P7C3 Probabilistic PWR 003T +100F PWR_003T_+100F_P7C3 PWR 003T +100F P7C4 Deterministic PWR 003T +100F PWR_003T_+100F_P7C4 PWR 003T +100F P7C4 Probabilistic PWR 003T +100F PWR_003T_+100F_P7C4 PWR 003T +100F P7C5 Deterministic PWR 003T +100F PWR_003T_+100F_P7C5 PWR 003T +100F P7C5 Probabilistic PWR 003T +100F PWR_003T_+100F_P7C5 PWR 003T +100F P7C6 Deterministic PWR 003T +100F PWR_003T_+100F_P7C6 PWR 003T +100F P7C6 Probabilistic PWR 003T +100F PWR_003T_+100F_P7C6 109

Reactor Type Flaw RTNDT Shift Transient Run Type Folder Case Name PWR 003T +100F P7C7 Deterministic PWR 003T +100F PWR_003T_+100F_P7C7 PWR 003T +100F P7C7 Probabilistic PWR 003T +100F PWR_003T_+100F_P7C7 PWR 003T +100F P7C8 Deterministic PWR 003T +100F PWR_003T_+100F_P7C8 PWR 003T +100F P7C8 Probabilistic PWR 003T +100F PWR_003T_+100F_P7C8 PWR 003T +100F P8C1 Deterministic PWR 003T +100F PWR_003T_+100F_P8C1 PWR 003T +100F P8C1 Probabilistic PWR 003T +100F PWR_003T_+100F_P8C1 PWR 003T +100F P8C2 Deterministic PWR 003T +100F PWR_003T_+100F_P8C2 PWR 003T +100F P8C2 Probabilistic PWR 003T +100F PWR_003T_+100F_P8C2 PWR 003T +100F P8C3 Deterministic PWR 003T +100F PWR_003T_+100F_P8C3 PWR 003T +100F P8C3 Probabilistic PWR 003T +100F PWR_003T_+100F_P8C3 PWR 003T +100F P8C4 Deterministic PWR 003T +100F PWR_003T_+100F_P8C4 PWR 003T +100F P8C4 Probabilistic PWR 003T +100F PWR_003T_+100F_P8C4 PWR 003T +100F P8C5 Deterministic PWR 003T +100F PWR_003T_+100F_P8C5 PWR 003T +100F P8C5 Probabilistic PWR 003T +100F PWR_003T_+100F_P8C5 PWR 003T +100F P8C6 Deterministic PWR 003T +100F PWR_003T_+100F_P8C6 PWR 003T +100F P8C6 Probabilistic PWR 003T +100F PWR_003T_+100F_P8C6 PWR 003T +100F P8C7 Deterministic PWR 003T +100F PWR_003T_+100F_P8C7 PWR 003T +100F P8C7 Probabilistic PWR 003T +100F PWR_003T_+100F_P8C7 PWR 003T +100F P9C1 Deterministic PWR 003T +100F PWR_003T_+100F_P9C1 PWR 003T +100F P9C1 Probabilistic PWR 003T +100F PWR_003T_+100F_P9C1 PWR 003T +100F P9C2 Deterministic PWR 003T +100F PWR_003T_+100F_P9C2 PWR 003T +100F P9C2 Probabilistic PWR 003T +100F PWR_003T_+100F_P9C2 PWR 003T +100F P10C1 Deterministic PWR 003T +100F PWR_003T_+100F_P10C1 PWR 003T +100F P10C1 Probabilistic PWR 003T +100F PWR_003T_+100F_P10C1 PWR 003T +100F P11C1 Deterministic PWR 003T +100F PWR_003T_+100F_P11C1 PWR 003T +100F P11C1 Probabilistic PWR 003T +100F PWR_003T_+100F_P11C1 PWR 003T +100F PAC1 Deterministic PWR 003T +100F PWR_003T_+100F_PAC1 PWR 003T +100F PAC1 Probabilistic PWR 003T +100F PWR_003T_+100F_PAC1 PWR 003T +100F PAC2 Deterministic PWR 003T +100F PWR_003T_+100F_PAC2 PWR 003T +100F PAC2 Probabilistic PWR 003T +100F PWR_003T_+100F_PAC2 110

Reactor Type Flaw RTNDT Shift Transient Run Type Folder Case Name PWR 003T +100F PBC1 Deterministic PWR 003T +100F PWR_003T_+100F_PBC1 PWR 003T +100F PBC1 Probabilistic PWR 003T +100F PWR_003T_+100F_PBC1 PWR 003T +100F PBC2 Deterministic PWR 003T +100F PWR_003T_+100F_PBC2 PWR 003T +100F PBC2 Probabilistic PWR 003T +100F PWR_003T_+100F_PBC2 PWR 003T +100F PCC1 Deterministic PWR 003T +100F PWR_003T_+100F_PCC1 PWR 003T +100F PCC1 Probabilistic PWR 003T +100F PWR_003T_+100F_PCC1 PWR 003T +100F PCC2 Deterministic PWR 003T +100F PWR_003T_+100F_PCC2 PWR 003T +100F PCC2 Probabilistic PWR 003T +100F PWR_003T_+100F_PCC2 PWR 003T +100F PDC1 Deterministic PWR 003T +100F PWR_003T_+100F_PDC1 PWR 003T +100F PDC1 Probabilistic PWR 003T +100F PWR_003T_+100F_PDC1 PWR 003T +100F PDC2 Deterministic PWR 003T +100F PWR_003T_+100F_PDC2 PWR 003T +100F PDC2 Probabilistic PWR 003T +100F PWR_003T_+100F_PDC2 PWR 003T +100F PEC1 Deterministic PWR 003T +100F PWR_003T_+100F_PEC1 PWR 003T +100F PEC1 Probabilistic PWR 003T +100F PWR_003T_+100F_PEC1 PWR 003T +100F PEC2 Deterministic PWR 003T +100F PWR_003T_+100F_PEC2 PWR 003T +100F PEC2 Probabilistic PWR 003T +100F PWR_003T_+100F_PEC2 PWR 003T +100F PFC1 Deterministic PWR 003T +100F PWR_003T_+100F_PFC1 PWR 003T +100F PFC1 Probabilistic PWR 003T +100F PWR_003T_+100F_PFC1 PWR 003T +100F PFC2 Deterministic PWR 003T +100F PWR_003T_+100F_PFC2 PWR 003T +100F PFC2 Probabilistic PWR 003T +100F PWR_003T_+100F_PFC2 111

6.2 BWR Analyses Reactor Type Flaw RTNDT Shift Transient Run Type Folder Case Name BWR 025T 0F P0_PT100 Deterministic BWR 025T 0F BWR_025T_0F_P0_PT100 BWR 025T -40F PT100-100 Deterministic BWR 025T -40F BWR_025T_-40F_PT100-100 BWR 025T -40F PT100-100 Probabilistic BWR 025T -40F BWR_025T_-40F_PT100-100 BWR 004T -40F PT100-100 Deterministic BWR 004T -40F BWR_004T_-40F_PT100-100 BWR 004T -40F PT100-100 Probabilistic BWR 004T -40F BWR_004T_-40F_PT100-100 BWR 025T 0F PT100-100 Deterministic BWR 025T 0F BWR_025T_0F_PT100-100 BWR 025T 0F PT100-100 Probabilistic BWR 025T 0F BWR_025T_0F_PT100-100 BWR 004T 0F PT100-100 Deterministic BWR 004T 0F BWR_004T_0F_PT100-100 BWR 004T 0F PT100-100 Probabilistic BWR 004T 0F BWR_004T_0F_PT100-100 BWR 025T +6F PT100-100 Deterministic BWR 025T +6F BWR_025T_+6F_PT100-100 BWR 025T +6F PT100-100 Probabilistic BWR 025T +6F BWR_025T_+6F_PT100-100 BWR 004T +6F PT100-100 Deterministic BWR 004T +6F BWR_004T_+6F_PT100-100 BWR 004T +6F PT100-100 Probabilistic BWR 004T +6F BWR_004T_+6F_PT100-100 BWR 025T +15F PT100-100 Deterministic BWR 025T +15F BWR_025T_+15F_PT100-100 BWR 025T +15F PT100-100 Probabilistic BWR 025T +15F BWR_025T_+15F_PT100-100 BWR 004T +15F PT100-100 Deterministic BWR 004T +15F BWR_004T_+15F_PT100-100 BWR 004T +15F PT100-100 Probabilistic BWR 004T +15F BWR_004T_+15F_PT100-100 BWR 025T +47F PT100-100 Deterministic BWR 025T +47F BWR_025T_+47F_PT100-100 BWR 025T +47F PT100-100 Probabilistic BWR 025T +47F BWR_025T_+47F_PT100-100 BWR 004T +47F PT100-100 Deterministic BWR 004T +47F BWR_004T_+47F_PT100-100 BWR 004T +47F PT100-100 Probabilistic BWR 004T +47F BWR_004T_+47F_PT100-100 BWR 025T +128F PT100-100 Deterministic BWR 025T +128F BWR_025T_+128F_PT100-100 BWR 025T +128F PT100-100 Probabilistic BWR 025T +128F BWR_025T_+128F_PT100-100 BWR 004T +128F PT100-100 Deterministic BWR 004T +128F BWR_004T_+128F_PT100-100 BWR 004T +128F PT100-100 Probabilistic BWR 004T +128F BWR_004T_+128F_PT100-100 BWR 025T 0F P0_PT50 Deterministic BWR 025T 0F BWR_025T_0F_P0_PT50 BWR 025T -40F PT100-50 Deterministic BWR 025T -40F BWR_025T_-40F_PT100-50 112

Reactor Type Flaw RTNDT Shift Transient Run Type Folder Case Name BWR 025T -40F PT100-50 Probabilistic BWR 025T -40F BWR_025T_-40F_PT100-50 BWR 004T -40F PT100-50 Deterministic BWR 004T -40F BWR_004T_-40F_PT100-50 BWR 004T -40F PT100-50 Probabilistic BWR 004T -40F BWR_004T_-40F_PT100-50 BWR 025T 0F PT100-50 Deterministic BWR 025T 0F BWR_025T_0F_PT100-50 BWR 025T 0F PT100-50 Probabilistic BWR 025T 0F BWR_025T_0F_PT100-50 BWR 004T 0F PT100-50 Deterministic BWR 004T 0F BWR_004T_0F_PT100-50 BWR 004T 0F PT100-50 Probabilistic BWR 004T 0F BWR_004T_0F_PT100-50 BWR 025T +6F PT100-50 Deterministic BWR 025T +6F BWR_025T_+6F_PT100-50 BWR 025T +6F PT100-50 Probabilistic BWR 025T +6F BWR_025T_+6F_PT100-50 BWR 004T +6F PT100-50 Deterministic BWR 004T +6F BWR_004T_+6F_PT100-50 BWR 004T +6F PT100-50 Probabilistic BWR 004T +6F BWR_004T_+6F_PT100-50 BWR 025T +15F PT100-50 Deterministic BWR 025T +15F BWR_025T_+15F_PT100-50 BWR 025T +15F PT100-50 Probabilistic BWR 025T +15F BWR_025T_+15F_PT100-50 BWR 004T +15F PT100-50 Deterministic BWR 004T +15F BWR_004T_+15F_PT100-50 BWR 004T +15F PT100-50 Probabilistic BWR 004T +15F BWR_004T_+15F_PT100-50 BWR 025T +47F PT100-50 Deterministic BWR 025T +47F BWR_025T_+47F_PT100-50 BWR 025T +47F PT100-50 Probabilistic BWR 025T +47F BWR_025T_+47F_PT100-50 BWR 004T +47F PT100-50 Deterministic BWR 004T +47F BWR_004T_+47F_PT100-50 BWR 004T +47F PT100-50 Probabilistic BWR 004T +47F BWR_004T_+47F_PT100-50 BWR 025T +128F PT100-50 Deterministic BWR 025T +128F BWR_025T_+128F_PT100-50 BWR 025T +128F PT100-50 Probabilistic BWR 025T +128F BWR_025T_+128F_PT100-50 BWR 004T +128F PT100-50 Deterministic BWR 004T +128F BWR_004T_+128F_PT100-50 BWR 004T +128F PT100-50 Probabilistic BWR 004T +128F BWR_004T_+128F_PT100-50 BWR 025T -40F PT50-50 Deterministic BWR 025T -40F BWR_025T_-40F_PT50-50 BWR 025T -40F PT50-50 Probabilistic BWR 025T -40F BWR_025T_-40F_PT50-50 BWR 004T -40F PT50-50 Deterministic BWR 004T -40F BWR_004T_-40F_PT50-50 BWR 004T -40F PT50-50 Probabilistic BWR 004T -40F BWR_004T_-40F_PT50-50 BWR 025T 0F PT50-50 Deterministic BWR 025T 0F BWR_025T_0F_PT50-50 BWR 025T 0F PT50-50 Probabilistic BWR 025T 0F BWR_025T_0F_PT50-50 BWR 004T 0F PT50-50 Deterministic BWR 004T 0F BWR_004T_0F_PT50-50 113

Reactor Type Flaw RTNDT Shift Transient Run Type Folder Case Name BWR 004T 0F PT50-50 Probabilistic BWR 004T 0F BWR_004T_0F_PT50-50 BWR 025T +6F PT50-50 Deterministic BWR 025T +6F BWR_025T_+6F_PT50-50 BWR 025T +6F PT50-50 Probabilistic BWR 025T +6F BWR_025T_+6F_PT50-50 BWR 004T +6F PT50-50 Deterministic BWR 004T +6F BWR_004T_+6F_PT50-50 BWR 004T +6F PT50-50 Probabilistic BWR 004T +6F BWR_004T_+6F_PT50-50 BWR 025T +15F PT50-50 Deterministic BWR 025T +15F BWR_025T_+15F_PT50-50 BWR 025T +15F PT50-50 Probabilistic BWR 025T +15F BWR_025T_+15F_PT50-50 BWR 004T +15F PT50-50 Deterministic BWR 004T +15F BWR_004T_+15F_PT50-50 BWR 004T +15F PT50-50 Probabilistic BWR 004T +15F BWR_004T_+15F_PT50-50 BWR 025T +47F PT50-50 Deterministic BWR 025T +47F BWR_025T_+47F_PT50-50 BWR 025T +47F PT50-50 Probabilistic BWR 025T +47F BWR_025T_+47F_PT50-50 BWR 004T +47F PT50-50 Deterministic BWR 004T +47F BWR_004T_+47F_PT50-50 BWR 004T +47F PT50-50 Probabilistic BWR 004T +47F BWR_004T_+47F_PT50-50 BWR 025T +128F PT50-50 Deterministic BWR 025T +128F BWR_025T_+128F_PT50-50 BWR 025T +128F PT50-50 Probabilistic BWR 025T +128F BWR_025T_+128F_PT50-50 BWR 004T +128F PT50-50 Deterministic BWR 004T +128F BWR_004T_+128F_PT50-50 BWR 004T +128F PT50-50 Probabilistic BWR 004T +128F BWR_004T_+128F_PT50-50 BWR 025T -40F S50 Deterministic BWR 025T -40F BWR_025T_-40F_S50 BWR 025T -40F S50 Probabilistic BWR 025T -40F BWR_025T_-40F_S50 BWR 004T -40F S50 Deterministic BWR 004T -40F BWR_004T_-40F_S50 BWR 004T -40F S50 Probabilistic BWR 004T -40F BWR_004T_-40F_S50 BWR 025T 0F S50 Deterministic BWR 025T 0F BWR_025T_0F_S50 BWR 025T 0F S50 Probabilistic BWR 025T 0F BWR_025T_0F_S50 BWR 004T 0F S50 Deterministic BWR 004T 0F BWR_004T_0F_S50 BWR 004T 0F S50 Probabilistic BWR 004T 0F BWR_004T_0F_S50 BWR 025T +6F S50 Deterministic BWR 025T +6F BWR_025T_+6F_S50 BWR 025T +6F S50 Probabilistic BWR 025T +6F BWR_025T_+6F_S50 BWR 004T +6F S50 Deterministic BWR 004T +6F BWR_004T_+6F_S50 BWR 004T +6F S50 Probabilistic BWR 004T +6F BWR_004T_+6F_S50 BWR 025T +15F S50 Deterministic BWR 025T +15F BWR_025T_+15F_S50 114

Reactor Type Flaw RTNDT Shift Transient Run Type Folder Case Name BWR 025T +15F S50 Probabilistic BWR 025T +15F BWR_025T_+15F_S50 BWR 004T +15F S50 Deterministic BWR 004T +15F BWR_004T_+15F_S50 BWR 004T +15F S50 Probabilistic BWR 004T +15F BWR_004T_+15F_S50 BWR 025T +47F S50 Deterministic BWR 025T +47F BWR_025T_+47F_S50 BWR 025T +47F S50 Probabilistic BWR 025T +47F BWR_025T_+47F_S50 BWR 004T +47F S50 Deterministic BWR 004T +47F BWR_004T_+47F_S50 BWR 004T +47F S50 Probabilistic BWR 004T +47F BWR_004T_+47F_S50 BWR 025T +128F S50 Deterministic BWR 025T +128F BWR_025T_+128F_S50 BWR 025T +128F S50 Probabilistic BWR 025T +128F BWR_025T_+128F_S50 BWR 004T +128F S50 Deterministic BWR 004T +128F BWR_004T_+128F_S50 BWR 004T +128F S50 Probabilistic BWR 004T +128F BWR_004T_+128F_S50 BWR 025T 0F P0_LT40-40 Deterministic BWR 025T 0F BWR_025T_0F_P0_LT40-40 BWR 025T -40F LT40-40 Deterministic BWR 025T -40F BWR_025T_-40F_LT40-40 BWR 025T -40F LT40-40 Probabilistic BWR 025T -40F BWR_025T_-40F_LT40-40 BWR 004T -40F LT40-40 Deterministic BWR 004T -40F BWR_004T_-40F_LT40-40 BWR 004T -40F LT40-40 Probabilistic BWR 004T -40F BWR_004T_-40F_LT40-40 BWR 025T 0F LT40-40 Deterministic BWR 025T 0F BWR_025T_0F_LT40-40 BWR 025T 0F LT40-40 Probabilistic BWR 025T 0F BWR_025T_0F_LT40-40 BWR 004T 0F LT40-40 Deterministic BWR 004T 0F BWR_004T_0F_LT40-40 BWR 004T 0F LT40-40 Probabilistic BWR 004T 0F BWR_004T_0F_LT40-40 BWR 025T +6F LT40-40 Deterministic BWR 025T +6F BWR_025T_+6F_LT40-40 BWR 025T +6F LT40-40 Probabilistic BWR 025T +6F BWR_025T_+6F_LT40-40 BWR 004T +6F LT40-40 Deterministic BWR 004T +6F BWR_004T_+6F_LT40-40 BWR 004T +6F LT40-40 Probabilistic BWR 004T +6F BWR_004T_+6F_LT40-40 BWR 025T +15F LT40-40 Deterministic BWR 025T +15F BWR_025T_+15F_LT40-40 BWR 025T +15F LT40-40 Probabilistic BWR 025T +15F BWR_025T_+15F_LT40-40 BWR 004T +15F LT40-40 Deterministic BWR 004T +15F BWR_004T_+15F_LT40-40 BWR 004T +15F LT40-40 Probabilistic BWR 004T +15F BWR_004T_+15F_LT40-40 BWR 025T +47F LT40-40 Deterministic BWR 025T +47F BWR_025T_+47F_LT40-40 BWR 025T +47F LT40-40 Probabilistic BWR 025T +47F BWR_025T_+47F_LT40-40 115

Reactor Type Flaw RTNDT Shift Transient Run Type Folder Case Name BWR 004T +47F LT40-40 Deterministic BWR 004T +47F BWR_004T_+47F_LT40-40 BWR 004T +47F LT40-40 Probabilistic BWR 004T +47F BWR_004T_+47F_LT40-40 BWR 025T +128F LT40-40 Deterministic BWR 025T +128F BWR_025T_+128F_LT40-40 BWR 025T +128F LT40-40 Probabilistic BWR 025T +128F BWR_025T_+128F_LT40-40 BWR 004T +128F LT40-40 Deterministic BWR 004T +128F BWR_004T_+128F_LT40-40 BWR 004T +128F LT40-40 Probabilistic BWR 004T +128F BWR_004T_+128F_LT40-40 BWR 025T 0F P0_LT40-100 Deterministic BWR 025T 0F BWR_025T_0F_P0_LT40-100 BWR 025T -40F LT40-100 Deterministic BWR 025T -40F BWR_025T_-40F_LT40-100 BWR 025T -40F LT40-100 Probabilistic BWR 025T -40F BWR_025T_-40F_LT40-100 BWR 004T -40F LT40-100 Deterministic BWR 004T -40F BWR_004T_-40F_LT40-100 BWR 004T -40F LT40-100 Probabilistic BWR 004T -40F BWR_004T_-40F_LT40-100 BWR 025T 0F LT40-100 Deterministic BWR 025T 0F BWR_025T_0F_LT40-100 BWR 025T 0F LT40-100 Probabilistic BWR 025T 0F BWR_025T_0F_LT40-100 BWR 004T 0F LT40-100 Deterministic BWR 004T 0F BWR_004T_0F_LT40-100 BWR 004T 0F LT40-100 Probabilistic BWR 004T 0F BWR_004T_0F_LT40-100 BWR 025T +6F LT40-100 Deterministic BWR 025T +6F BWR_025T_+6F_LT40-100 BWR 025T +6F LT40-100 Probabilistic BWR 025T +6F BWR_025T_+6F_LT40-100 BWR 004T +6F LT40-100 Deterministic BWR 004T +6F BWR_004T_+6F_LT40-100 BWR 004T +6F LT40-100 Probabilistic BWR 004T +6F BWR_004T_+6F_LT40-100 BWR 025T +15F LT40-100 Deterministic BWR 025T +15F BWR_025T_+15F_LT40-100 BWR 025T +15F LT40-100 Probabilistic BWR 025T +15F BWR_025T_+15F_LT40-100 BWR 004T +15F LT40-100 Deterministic BWR 004T +15F BWR_004T_+15F_LT40-100 BWR 004T +15F LT40-100 Probabilistic BWR 004T +15F BWR_004T_+15F_LT40-100 BWR 025T +47F LT40-100 Deterministic BWR 025T +47F BWR_025T_+47F_LT40-100 BWR 025T +47F LT40-100 Probabilistic BWR 025T +47F BWR_025T_+47F_LT40-100 BWR 004T +47F LT40-100 Deterministic BWR 004T +47F BWR_004T_+47F_LT40-100 BWR 004T +47F LT40-100 Probabilistic BWR 004T +47F BWR_004T_+47F_LT40-100 BWR 025T +128F LT40-100 Deterministic BWR 025T +128F BWR_025T_+128F_LT40-100 BWR 025T +128F LT40-100 Probabilistic BWR 025T +128F BWR_025T_+128F_LT40-100 BWR 004T +128F LT40-100 Deterministic BWR 004T +128F BWR_004T_+128F_LT40-100 116

Reactor Type Flaw RTNDT Shift Transient Run Type Folder Case Name BWR 004T +128F LT40-100 Probabilistic BWR 004T +128F BWR_004T_+128F_LT40-100 BWR 025T 0F P0_LTA Deterministic BWR 025T 0F BWR_025T_0F_P0_LTA BWR 025T -40F LTA Deterministic BWR 025T -40F BWR_025T_-40F_LTA BWR 025T -40F LTA Probabilistic BWR 025T -40F BWR_025T_-40F_LTA BWR 004T -40F LTA Deterministic BWR 004T -40F BWR_004T_-40F_LTA BWR 004T -40F LTA Probabilistic BWR 004T -40F BWR_004T_-40F_LTA BWR 025T 0F LTA Deterministic BWR 025T 0F BWR_025T_0F_LTA BWR 025T 0F LTA Probabilistic BWR 025T 0F BWR_025T_0F_LTA BWR 004T 0F LTA Deterministic BWR 004T 0F BWR_004T_0F_LTA BWR 004T 0F LTA Probabilistic BWR 004T 0F BWR_004T_0F_LTA BWR 025T +6F LTA Deterministic BWR 025T +6F BWR_025T_+6F_LTA BWR 025T +6F LTA Probabilistic BWR 025T +6F BWR_025T_+6F_LTA BWR 004T +6F LTA Deterministic BWR 004T +6F BWR_004T_+6F_LTA BWR 004T +6F LTA Probabilistic BWR 004T +6F BWR_004T_+6F_LTA BWR 025T +15F LTA Deterministic BWR 025T +15F BWR_025T_+15F_LTA BWR 025T +15F LTA Probabilistic BWR 025T +15F BWR_025T_+15F_LTA BWR 004T +15F LTA Deterministic BWR 004T +15F BWR_004T_+15F_LTA BWR 004T +15F LTA Probabilistic BWR 004T +15F BWR_004T_+15F_LTA BWR 025T +47F LTA Deterministic BWR 025T +47F BWR_025T_+47F_LTA BWR 025T +47F LTA Probabilistic BWR 025T +47F BWR_025T_+47F_LTA BWR 004T +47F LTA Deterministic BWR 004T +47F BWR_004T_+47F_LTA BWR 004T +47F LTA Probabilistic BWR 004T +47F BWR_004T_+47F_LTA BWR 025T +128F LTA Deterministic BWR 025T +128F BWR_025T_+128F_LTA BWR 025T +128F LTA Probabilistic BWR 025T +128F BWR_025T_+128F_LTA BWR 004T +128F LTA Deterministic BWR 004T +128F BWR_004T_+128F_LTA BWR 004T +128F LTA Probabilistic BWR 004T +128F BWR_004T_+128F_LTA 117

7 Appendix B: Analysis of Other PWR Transients This Appendix describes the characteristics and analysis results for 6 PWR transients that were not deemed to be normal PWR cooldowns, as opposed to the transients presented in Section 2.1.2.4. It is important to note that no information was available regarding how representative these transients are of actual PWR fleet operations. These transients were simply some of those that were made available to NRC by Westinghouse in 2013 through an informal exchange of data.

7.1 Transient P7C1: Partial Cooldown Figure 7-1 shows the pressure and temperature time history for the P7C1 transient. This transient appears to be a partial cooldown from operating pressure and temperature. It is possible that this partial cooldown and hold at lower pressure and temperature was performed for some online maintenance activities, but this is not known with certainty.

Figure 7-1: PWR P7C1 Transient Characteristics 118

Figure 7-2 shows the P7C1 transient on a P-T diagram, along with the percentiles of actual cooldowns as described in Section 2.1.2.4, the PT50-50 transient described in Section 2.1.2.3, and the saturation line.

Although the transient initially followed the path of a normal PWR cooldown, the pressure was then held constant a little above 1500 psi while the temperature was dropped from about 470°F to 330°F.

Figure 7-2: P-T Curve for Transient P7C1 Overlaid on P-T Histories Corresponding to Various Percentiles of the Population of PWR Cooldowns Considered (Along with the PT50-50 Limit and the Saturation Curve) 119

Figure 7-3 shows the K histories versus time for a 0.03 T shallow flaw of aspect ratios of 2, 6, 10, and infinity, along with the Kia arrest toughness and the 50th percentile of the Kic fracture toughness distribution. Overall, the K values remain very low and mostly below the arrest toughness. A FAVOR calculation performed for this transient for a small internal surface breaking flaw of depth 0.03 T and aspect ratio of 6 resulted in a CPI of 6.1E-14 and a CPF of 4.5E-16, both extremely low. Such a result is consistent with the very low K values and their position relative to the Kia curve.

Figure 7-3: Stress Intensity Factor for the PWR P7C1 Transient for a 0.03 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity 120

7.2 Transient P7C5: High-Pressure Cooldown Figure 7-4 shows the pressure and temperature time history for the P7C5 transient. This transient appears to be cooldown from operating pressure and temperature down to about 200°F and 400 psi.

This cooldown distinguishes itself from the cooldowns discussed in Section 2.1.2.4 because the high pressure is maintained while the temperature is reduced all the way down to about 360°F.

Figure 7-4: PWR P7C5 Transient Characteristics 121

Figure 7-5 shows the P7C5 transient on a P-T diagram, along with the percentiles of actual cooldowns as described in Section 2.1.2.4, the PT50-50 transient described in Section 2.1.2.3, and the saturation line.

This cooldown is different from other cooldowns studied here because it is at or close to the P-T limit until the temperature reaches about 300°F, which appears to be unusual, and may not be allowed in many plants due to various limitations on pumps and other parts of the primary that may be more limiting than the vessel beltline from a P-T limit perspective.

Figure 7-5: P-T Curve for Transient P7C5 Overlaid on P-T Histories Corresponding to Various Percentiles of the Population of PWR Cooldowns Considered (Along with the PT50-50 Limit and the Saturation Curve) 122

Figure 7-6 shows the K histories versus time for a 0.03 T shallow flaw of aspect ratios of 2, 6, 10, and infinity, along with the Kia arrest toughness and the 50th percentile of the Kic fracture toughness distribution. Overall, the K values are low early in the transient despite the high pressure. This is expected for shallow flaws because the K values are more driven by thermal stresses than pressure for such flaws. The K values exceed Kia later in the transient as the thermal stresses build up from the temperature reduction, and this occurs despite the pressure being low after about 8000 minutes. A FAVOR calculation performed for this transient for a small internal surface breaking flaw of depth 0.03 T and aspect ratio of 6 resulted in a relatively high CPI of 6.7E-6 but a CPF of only 4.8E-14, posing no real failure risk for the vessel.

Figure 7-6: Stress Intensity Factor for the PWR P7C5 Transient for a 0.03 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity 123

7.3 Transient P7C6: Unusual Cooldown Figure 7-7 shows the pressure and temperature time history for the P7C6 transient. This transient does represent a cooldown from operating pressure and temperature down to about 85°F and 400 psi, but it is unusual in the sense that there was a repressurization between 110 hours0.00127 days <br />0.0306 hours <br />1.818783e-4 weeks <br />4.1855e-5 months <br /> and 160 hours0.00185 days <br />0.0444 hours <br />2.645503e-4 weeks <br />6.088e-5 months <br />. This could have been a failed attempt to return to operation after a period of lower pressure and temperature for online maintenance, but this cannot be confirmed.

Figure 7-7: PWR P7C6 Transient Characteristics 124

Figure 7-8 shows the P7C6 transient on a P-T diagram, along with the percentiles of actual cooldowns as described in Section 2.1.2.4, the PT50-50 transient described in Section 2.1.2.3, and the saturation line.

This cooldown is different from other cooldowns studied here because of the sustained pressure between 500°F and about 330°F. In addition, the pressure fluctuations around 330°F and later around 80°F, as well as the temperature fluctuation around 400 psi appear to be atypical. Nonetheless, the transient falls overall well below the P-T limit, although it exceed the 99th percentile of normal cooldowns from Section 2.1.2.4 for temperature above ~330°F.

Figure 7-8: P-T Curve for Transient P7C6 Overlaid on P-T Histories Corresponding to Various Percentiles of the Population of PWR Cooldowns Considered (Along with the PT50-50 Limit and the Saturation Curve) 125

Figure 7-9 shows the K histories versus time for a 0.03 T shallow flaw of aspect ratios of 2, 6, 10, and infinity, along with the Kia arrest toughness and the 50th percentile of the Kic fracture toughness distribution. Overall, the K values are low early in the transient despite the high pressure and repressurization. This is expected for shallow flaws because the K values are more driven by thermal stresses than pressure for such flaws. The K values exceed Kia later in the transient as the thermal stresses build up from the temperature reduction, and this occurs despite the pressure being low after about 16000 minutes. A FAVOR calculation performed for this transient for a small internal surface breaking flaw of depth 0.03 T and aspect ratio of 6 resulted in a relatively high CPI of 1.4E-5 but a CPF of only 3.9E-10, posing no real failure risk for the vessel.

Figure 7-9: Stress Intensity Factor for the PWR P7C6 Transient for a 0.03 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity 126

7.4 Transient P7C8: Partial Cooldown Followed by Heatup Figure 7-10 shows the pressure and temperature time history for the P7C8 transient. This transient appears to be a partial cooldown from operating pressure and temperature, followed by the beginning of a return to operating conditions. It is possible that this partial cooldown and hold at lower pressure and temperature was performed for some online maintenance activities, but this is not known with certainty. In many ways, this transient is similar to the P7C1 transient described in 7.1.

Figure 7-10: PWR P7C8 Transient Characteristics 127

Figure 7-11 shows the P7C8 transient on a P-T diagram, along with the percentiles of actual cooldowns as described in Section 2.1.2.4, the PT50-50 transient described in Section 2.1.2.3, and the saturation line. Although the transient initially followed the path of a normal PWR cooldown, the pressure was then held constant a little above 1500 psi while the temperature was dropped from about 450°F to 325°F, and then pressure and temperature were both increased in the direction of a return to operating conditions. The first part of this transient is very similar to the P7C1 transient (see Figure 7-2)

Figure 7-11: P-T Curve for Transient P7C8 Overlaid on P-T Histories Corresponding to Various Percentiles of the Population of PWR Cooldowns Considered (Along with the PT50-50 Limit and the Saturation Curve) 128

Figure 7-12 shows the K histories versus time for a 0.03 T shallow flaw of aspect ratios of 2, 6, 10, and infinity, along with the Kia arrest toughness and the 50th percentile of the Kic fracture toughness distribution. Overall, the K values remain very low and entirely below the arrest toughness. A FAVOR calculation performed for this transient for a small internal surface breaking flaw of depth 0.03 T and aspect ratio of 6 resulted in a CPI and a CPF of zero. Such a result is consistent with the very low K values and their position relative to the Kia curve. The reason why the CPI and CPF are lower for this transient than for transient P7C1 is that P7C1 experienced a more rapid drop in temperature around 1750 minutes, resulting in a brief and small spike in K values that does not occur for transient P7C8 (see Figure 7-3).

Figure 7-12: Stress Intensity Factor for the PWR P7C8 Transient for a 0.03 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity 129

7.5 Transient P8C3: Cooldown Followed by Partial Heatup Figure 7-13 shows the pressure and temperature time history for the P8C3 transient. This transient appears to be a complete cooldown from operating pressure and temperature, followed by the beginning of a return to operating conditions. It is possible that this complete cooldown and hold at zero pressure and ambient temperature was performed for some short duration maintenance activities, or as a result of a reactor trip and subsequent return to power, but this is not known with certainty. In many ways, this transient is similar to the many of the normal cooldowns described in Section 2.1.2.4, except that it is followed by the beginning of a return to operating conditions.

Figure 7-13: PWR P8C3 Transient Characteristics 130

Figure 7-14 shows the P8C3 transient on a P-T diagram, along with the percentiles of actual cooldowns as described in Section 2.1.2.4, the PT50-50 transient described in Section 2.1.2.3, and the saturation line. This cooldown very similar to many of the normal cooldowns studied in Section 2.1.2.4, and lies between the 25th and 75th percentiles of the Section 2.1.2.4 cooldowns for most of the transient. Even the return to power seems to be following roughly the same path.

Figure 7-14: P-T Curve for Transient P8C3 Overlaid on P-T Histories Corresponding to Various Percentiles of the Population of PWR Cooldowns Considered (Along with the PT50-50 Limit and the Saturation Curve) 131

Figure 7-15 shows the K histories versus time for a 0.03 T shallow flaw of aspect ratios of 2, 6, 10, and infinity, along with the Kia arrest toughness and the 50th percentile of the Kic fracture toughness distribution. Overall, the K values remain low with a small spike around 1250 minutes, which relates to a relatively rapid drop in temperature, which drives the thermal K up for the shallow flaw. The K values do exceed the Kia line for most of the low temperature portion of the transient, despite there being no pressure applied, but this is expected as it is driven by the thermal stresses for the shallow flaw. It is interesting to note the expected decrease in K as the temperature is increased during the return to operating conditions at the end of the transient. A FAVOR calculation performed for this transient for a small internal surface breaking flaw of depth 0.03 T and aspect ratio of 6 resulted in a relatively high CPI of 1.9E-6 but a CPF of zero, posing zero failure risk for the vessel, as would be hoped for in normal operations as seem to be the case for this transient.

Figure 7-15: Stress Intensity Factor for the PWR P8C3 Transient for a 0.03 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity 132

7.6 Transient PBC1: Cooldown with Pressure Hold and Low Temperature Pressure Spike Figure 7-16 shows the pressure and temperature time history for the PBC1 transient. This transient does represent a cooldown from operating pressure and temperature, but it is unusual in the sense that there was a repressurization exceeding the RCS operating pressure at 58 hours6.712963e-4 days <br />0.0161 hours <br />9.589947e-5 weeks <br />2.2069e-5 months <br />. This repressurization exceed all RCS pressure limits as it reaches about 3000 psi and occurs over a period of roughly 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />.

Since NRC would have notified of such an even of it had occurred in the vessel or other portion of the RCS, one possibility is that the pressure sensor that recorded this data was isolated from the main RCS when this occurred and that the pressure spike was localized in a system that is open to the RCS during normal operation but was isolated at the time of the pressure spike. Of course, this scenario cannot be confirmed, so for the purpose of the analysis, it was assumed that this data truly represented vessel conditions, which is almost certainly not the case here.

Figure 7-16: PWR PBC1 Transient Characteristics 133

Figure 7-17 shows the PBC1 transient on a P-T diagram, along with the percentiles of actual cooldowns as described in Section 2.1.2.4, the PT50-50 transient described in Section 2.1.2.3, and the saturation line. This spears to be a stepwise cooldown where the pressure was reduced in large increments at almost constant temperature, followed by reductions in temperature at almost constant pressure, rather than a cooldown where pressure and temperature were reduced simultaneously. There were several other such stepwise cooldowns among the cooldowns studied in Section 2.1.2.4, and this transient would have likely been grouped with the normal cooldowns if it were not for the late pressure spike. As described above, this pressure spike is not believed to have been experienced by the vessel itself, but rather by an isolated portion of the primary. Nonetheless, the transient was analyzed as if this was the pressure in the vessel.

Figure 7-17: P-T Curve for Transient PBC1 Overlaid on P-T Histories Corresponding to Various Percentiles of the Population of PWR Cooldowns Considered (Along with the PT50-50 Limit and the Saturation Curve) 134

Figure 7-18 shows the K histories versus time for a 0.03 T shallow flaw of aspect ratios of 2, 6, 10, and infinity, along with the Kia arrest toughness and the 50th percentile of the Kic fracture toughness distribution. As is the case for all the transients presented in 2.1.2.4 and in this Appendix, the K values remain below Kia initially. There is a pike in K values around 1600 minutes associated with a rapid drop in temperature, and again at the very end of the transient also associated with a temperature drop.

These K increases are related to tensile thermal stresses experienced by the shallow flaw upon cooling down events. The biggest spike in K, however, is experienced at about 3500 minutes, and drives the CPI and CPF values calculate with FAVOR. In fact, a FAVOR calculation performed for this transient for a small internal surface breaking flaw of depth 0.03 T and aspect ratio of 6 resulted in a very high CPI of 1.40E-3 and a CPF of 1.38E-3 (almost equal to CPI), which is 3 orders of magnitude higher than normally acceptable. It is important to underline that this transient is not believed to have been experienced by the vessel, and could be related to a sensor malfunction or a primary system piping configuration.

Figure 7-18: Stress Intensity Factor for the PWR PBC1 Transient for a 0.03 T Flaw for Flaw Aspect Ratios of 2, 6, 10 and Infinity 135