Revision as of 16:56, 30 June 2018

Text

Official Exhibit - ENT000663-00-BD01 - Letter from A. Mcgehee & T. Hanley, EPRI, to NRC DCD, BWRVIP 2015-025, Project No. 704-Summary of Industry Position on Screening Criteria for Thermal and Irradiation Embrittlement for PWR and BWR React
	ML15334A267
Person / Time
Site:	Indian Point
Issue date:	03/09/2015
From:	; Entergy Nuclear Operations
To:	; Atomic Safety and Licensing Board Panel
	SECY RAS
References
	RAS 28136, ASLBP 07-858-03-LR-BD01, 50-247-LR, 50-286-LR
	Download: ML15334A267 (23)
	v • d • e

'e,L 2l 1I ELECTRIC POWERRESEARCH INSTITUTE 2015-025BWR Vessel & Internals Project (BWRVIP)March 09, 2015Document Control DeskU.S. Nuclear Regulatory Commission 11555 Rockville PikeRockville, MD 20852Attention: Joseph Holonich

Subject:

Project No. 704 -Summary of Industry Position on Screening Criteria forThermal and Irradiation Embrittlement for PWR and BWR Reactor Internals Fabricated of Cast Austenitic Stainless Steel

References:

1. Letter from D. Madison (BWRVIP Chairman) and A. McGehee (BWRVIP Program Manager) to J. Holonich (NRC), Project No. 704 -BWRVIPResponse to NRC Request for Additional Information on BWRVIP-234, datedMay 23, 2014. ADAMS -ML14174A841.

2. Letter from J. Holonich (NRC) to A. Mendiola (NRC), Summary of the July 15,2014, Meeting with the Electric Power Research Institute on Items Related toCast Austenitic Stainless Steel and the Materials Reliability Program-227-A,"Pressurized Water Reactor Internals Inspection And Evaluation Guidelines,"

September 9, 2014. ADAMS -ML14127A077.

3. "NRC Position on Aging Management of CASS Reactor Vessel InternalComponents,"

June 23, 2014, ADAMS -ML 14174A719.

4. Email from C. Wirtz (EPRI) to J. Holonich (NRC), October 7, 2014.5. NUREG/CR-4513, ANL-93/22, Revision 1, "Estimation of Fracture Toughness of Cast Stainless Steels During Thermal Aging in LWR Systems,"

ArgonneNational Laboratory, May 1994.The purpose of this letter is to provide the status of the Industry (PWR and BWR) activities associated with developing and proposing a generic screening criteria to address thermal andirradiation embrittlement of RPV internals fabricated of cast austenitic stainless steel (CASS).Reference I was provided to the NRC in response to a Request for Additional Information (RAI) regarding BWRVIP-234.

The RAI response contained supplemental information that provided the technical bases for an Industry screening criteria for CASS internals.

Together

...Shaping the Future of Electricity Qýý U*PALO ALTO OFFICE3420 Fillview Avenue, Palo Alto, CA 94304-1395 USA

650.855.2000

Customer Service 800.313.3774 9 www.epri.com BWRVIP 2015-025A meeting was later held on July 15, 2014 at NRC offices in Washington, DC to discuss theIndustry and NRC positions (References 2, 3). Based on that meeting, the Industry committed to investigate additional information where there was disagreement with the NRC regarding the screening criteria and provide said information to the NRC at a later date. Thatinformation was informally submitted for NRC's consideration (Reference

4) and is contained in Attachments A through D of this letter. A conference call was then held on November 20,2014 with the NRC to discuss this information.

Following this conference call the NRC indicated that they were not willing to deviate fromtheir position documented in Reference 3.Despite the NRC's current stance on the matter, the following comments regarding a genericTE and IE criteria for evaluation of reactor internals are important in forming the basis for theindustry position:

" The industry recognizes that the NRC is intending to revise its original IE position in theGrimes letter from lxl107 n/cm2 (0.00014 dpa) to a position in the region of 0.5 to 1.5dpa. Industry agrees that this position is more appropriate and has proposed aconservative value of 1 dpa.* It is not appropriate (for the staff) to penalize all non-Mo containing CASS (such as CF3and CF8) because of the lower properties associated with the Mo-containing CF8Mmaterials that are included in the ANL/NRC database, i.e., impose a lower bound.* There is no substantive data that demonstrates a synergistic effect for CASS materials that are typical of reactor internals (such as CF3 and CF8). As such the industry hasproposed criteria that do not combine TE and IE. Each mechanism is considered distinctand separate.

" The introduction of a new criteria set for the category of materials with ferrite contentbetween 15% to 20% having a lower proposed screening value for IE of 0.45 dpa isunnecessary and technically unfounded.

The introduction of this category of ferritecontent is significantly burdensome to licensees since it will require more complex andpotentially more error-prone assessments of CASS material components by virtue ofhaving more categories.

Industry maintains that the categories of materials andassociated ferrite levels contained in the Grimes letter are appropriate.

" The criteria proposed by the industry (20% Ferrite and 1 dpa) have been shown to projectsignificant margin on toughness reduction and therefore safety when compared tomeasured embrittlement behavior for CF3 and CF8 materials.

Regardless, this letter serves to formally submit this additional information to the NRC forreview and consideration in developing a regulatory position for the screening criteria associated with TE and IE for CASS internals and to inform revisions to NUREG-4513, Rev. 1 (Reference 5). The PWR and BWR Industry position for TE and IE screening criteria is shown inAttachment D.2 BWRVIP 2015-025Irrespective of the staff's decision and future promulgation of a regulatory position on TE and IEfor CASS internals, the BWRVIP formally requests the NRC to evaluate and resolve the RAIresponses associated with BWRVIP-234 and issue a Safety Evaluation.

If you have any questions on this subject please call Ron DiSabatino (Exelon, BWRVIPAssessment Committee Technical Chairman) at 717.456.3685.

Sincerely, 24Andrew McGehee, EPRI, BWRVIP Program ManagerTim Hanley, Exelon Corp., BWRVIP Chairman3 BWRVIP 2015-025Industry Response to July 15, 2014 Meeting with NRC on CASSOn July 15, 2014, representatives from the BWRVIP/MRP Working Group on Cast Austenitic Stainless Steels (CASS) met with the NRC staff to discuss the thermal and irradiation embrittlement screening criteria for cast austenitic stainless steel. While the working group andNRC staff were in general agreement on the thermal embrittlement (TE) screening

criteria, therewas a marked difference of opinion on the irradiation embrittlement (IE) screening

criteria, particularly for materials with delta ferrite contents of 15-20% (the NRC proposed 0.45 dpa,while the industry proposed 1 dpa).Agreement was reached on the TE screening criteria in part because the data on low-molybdenum and high-molybdenum material are considered separately, resulting in separate TEscreening criteria for these materials.

This distinction is made because high-molybdenum materials show a distinctly greater loss of fracture toughness from TE and IE effects than low-molybdenum materials.

This is reviewed and discussed further in Attachment AThe separate consideration of low- and high-molybdenum materials test data was not carriedover into the IE domain. The data set used to determine the NRC's position on IE was obtainedfrom NUREG-7027, which considered irradiated high- and low-molybdenum materials

together, along with data on irradiated welds. Because these materials were considered

together, theresulting lower bound curve is judged to be overly conservative relative to the low-molybdenum material.

A reexamination of this data (see Attachment B), shows that by considering the low-molybdenum materials test data independently of the other materials it can be concluded thatthere is a substantial safety margin in the industry's proposed I dpa screening criteria.

These screening criteria are being implemented to ensure that CASS components in reactorvessel internals maintain adequate fracture toughness during the period of extended operation.

The screening level to determine whether this requirement was met was a J value of 255 KJ/m2at a crack extension of 2.5 mm. This value was originally determined for pressure boundarycomponents (i.e., large diameter piping, etc.) and was expected to be highly conservative forapplication to reactor vessel internals.

Additional calculations (see Attachment C) demonstrate the level of this conservatism and show that reactor vessel internal components can safelyoperate with fracture toughness values much lower than the 255 KJ/m2 value specified in theGrimes letter.In summary, the actions taken by the CASS industry working group and summarized inAttachments A, B and C support the proposed screening criteria contained in Attachment D.These criteria provide a significant amount of margin for screening of TE and IE. It is concluded that that the use of the proposed screening criteria will allow for continued safe and reliableoperation of the LWR fleet.4 BWRVIP 2015-025Attachment ADiscussion on Low Molybdenum vs High Molybdenum CASS GradesIntroduction A meeting was held on July 15, 2014 at the NRC Office in Washington, DC between the staff ofthe U.S. Nuclear Regulatory Commission (NRC) and the BWRVIP/MRP Working Group onCast Austenitic Stainless Steels (CASS). The topic of discussion focused on the thermal agingand neutron irradiation embrittlement screening criteria (separately) proposed by the industryand the NRC staff. The proposals included criteria for low-molybdenum CASS grades, such asCF-3 and CF-8, and high-molybdenum CASS grades, such as CF-8M. There was agreement between industry and the NRC staff regarding the differences in such measures of embrittlement as elastic-plastic crack growth resistance (J-R curve) for the low-Mo materials versus the high-Mo materials, at least where thermal aging embrittlement effects are dominant.

In fact,screening criteria proposed by the staff are identical to the industry-proposed screening criteriafor the case of thermal aging embrittlement.

However, the areas of agreement between the industry and the staff did not carry over tocombined environments where the effects of neutron irradiation embrittlement begin to approachthe same level as those from thermal aging embrittlement.

In order to address these differences incriteria, the available data for CF-3, CF-8, and CF-8M materials in the literature are reviewedand assessed, with a recommendation to the NRC staff for a change to their proposed criteria tomore closely agree with the criteria proposed by the industry.

Review and Summary of Thermal Aging Embrittlement DataThe significant differences in thermal aging embrittlement behavior of the low-molybdenum grades of cast austenitic stainless steel, such as CF-3 and CF-8, versus the thermal agingembrittlement behavior of the high-molybdenum grades, such as CF-8M, have been known for atleast two decades, with extensive data and associated interpretations of said data documented inNUREG/CR-4513, Revision 1 [1]. The lower-bound thermal aging embrittlement estimates shown in Figures 3 (statically-cast steels) and 4 (centrifugally-cast steels) from Reference 1 withvarying amounts of delta ferrite provide an excellent illustration of those significant differences, in this case with the measure of significance being the crack-growth resistance J-R curve. Inparticular, the lower-bound formulas cited in Section 3.1.1 of Reference 1 (and plotted in Figures3 and 4) can be used to compare the estimated crack growth resistance values at 2.5 mm of crackextension for the CF-3, CF-8, and CF-8M steels with > 15% delta ferrite.

Note that Figure 3provides lower-bound thermal aging estimates for static-cast steels, while Figure 4 provideslower-bound thermal aging estimates for centrifugally-cast steels.5 BWRVIP 2015-025Crack Extension ha (in.)Crack Extension As (in.)0.06 0.10 0.150.005.-3ISaIif5.-3IIa2 3 4 SLower-Bound J-R CurveStatic-Camt Stes>15% Fentbe-290"CJ =2510Aa)0-2 03-J=M4a -17&OCF-4u0 1 2 3 4 00.00 0.06 0.10 0.15* I I I ji , ' , : : : , ' , :Lower-Bound J-IRStatic-Cast StaWsi10-16% FegrktJ -274"0.3CF035j- 2740sa)03 2001&a)o36 cF-4 ..-:j -.: '. I: 32CF-Ui-4000-3000 -o-2000f.1000k0..4000-3000 I10.00.00 0.06 0.10 0.15lOOOLower-Bound J-R CurveStatic-Cat Steels80o. l-15% FenmlRoom Temp.800- 1 -31(1P40400 j =3074Uo-38 CF-C400 J 130-~036-CF-SM200-CFS01 2 3 45 01 23 450.00 0.06 0.10 0.1(0.00 0.06 0.10 0.155000"4000~30002000j1000i01 2 3 4Crack Extension, Aa (mm)Figure 3. Predicted lower-bound J-R curves at RT and 290°C for static-cast SSs with ferritecontents

>15, 10-15. or<10%6 BWRVIP 2015-025Crack Extension As (in.)0.o00. O. .O 0.1sCrack Extension ha (in.)0.00 O.06 0.10 0.15j400-~2007Lower-Bound J-R CurveCentrifugally Cast Steels>15% FmistRoom Tnp. J -304SO)0.37JF3348 39C-CF-MMFLower-Bound J-R CurveCentrifugally Cast Steels>15%FaFnite M-0C J.- 3301AP,34 0"35 CF-sCF-SCF-Sm.. ......H iI 1 23 40 1 2 3 4 10.00 0.06 010 0.15~20)0-X0-i J.30sWKa04CF-3r-Bound J-R CurveC " a olyCast Steels-. 357ý,,)°0.00 0.0S 0.10 0.15Lower-Bound J-R CurveCentrifugally Cast Steels10-15% Forte02fOC J -300ma)°35 0-36 CF-SCJ -38M&).3CF-3-4000 7C!-2000 'E-10000i4SO0O I-4000.I3000-2000 I-10000.5000-4O O"I:-3000 20001.1000Cf-M C0SM0 1 2 3 S4 1 2 3 450.00 0.0 0.10 0.150O30 0.06 0.10 0.151000 Lower-Bound J-R Curve4001200 -Lower-ilound' J-'R C'urv'e'

.Centrifugaly Cast SteDes410% FerniteRoom TSm 0 7 P41CF--3CF-SMLower-Bound J-R CurveCentrifugally Cast Steels<10% F-f-n¶J -435(Aa~5 CF-8CF-Sm0 1 2 3 4 5 0 2 3 4Crack Extension.,

A (mm) Crack Extension, A (mm)aFigure 4. Predicted lower-bound J-R curves at RT and 290'C for centrifugally cast SSs withferrite contents

> 15. 10- 15. or < 10%7 BWRVIP 2015-025For CF-3, the lower bound crack growth resistance, for statically-and centrifugally-cast materials, varies between 364 and 478 kJ/m2, and for CF-8, between 343 and 451 Id/m2.On theother hand, the CF-8M values vary between 161 and 259 kJ/m2.When the data are examinedvery closely, there is no significant difference between the static-cast lower-bound estimates forCF-3 and CF-8 in Figure 3 and the centrifugally-cast lower-bound estimates for CF-3 and CF-8in Figure 4. The data also show that these insignificant differences are roughly identical to thedifferences between CF-3 and CF-8 estimates within the static-cast and centrifugally-cast populations.

Therefore, based on this examination of lower-bound estimates, the CF-8M dataexhibit substantially lower fracture toughnesses and are therefore unsuitable as a lower-bound estimate for the low-molybdenum grades. Lower bound estimates must be conservative, but theymust also remain representative of the data. Table 1, shown below, succinctly summarizes thedata comparisons extracted from Figures 3 and 4 from Reference 1.Table 1. Lower Bound Fracture Toughness Estimates CASS Steels J-values at 2.5 cm of Crack CommentsGrowth (kJ/m2)Lower bound above the threshold CF-3 Steels 364 to 478 value of 255 kJ/M2Lower bound above the threshold CF-8 Steels 343 to 451 value of 255 kj/m2CF-8M Steels 161 to 259 Lower bound might be less thanthreshold values of 255 kJ/m2Table 1 clearly shows that lower bound fracture toughness values of CF-3 and CF-8 steels aremuch higher than CF-8M steels, and that only the fracture toughness of CF-8M steels fall belowthe threshold value of 255 kJ/m2.Therefore, the screening criteria for CF-8M (highmolybdenum) steels should be considered separately from CF-3 and CF-8 (low molybdenum) steels.Thus far, the data evaluation has been based on lower-bound fracture toughness estimates derived from Charpy impact correlations.

However, these lower-bound thermal agingembrittlement estimates can be confirmed by examining actual crack extension resistance curvemeasurements for various heats of CF-3, CF-8, and CF-8M with delta ferrite content in the rangeof 20%. For example, Figure 14 from Reference 1 shows J-R curve measurements, along withthermal aging saturation estimates for two heats of CF-3 material (Heat 69 and Heat I) and oneheat of CF-8 material (Heat 68), while Figure 15 from Reference 1 shows similar results forthree heats of CF-8M material (Heats 74, 75, and 758).In this case, it should be noted that the actual fracture toughness data are based on thermal agingsaturation conditions and are therefore somewhat higher than the lower-bound fracture toughness 8

BWRVIP 2015-025estimates.

Also note that the thermal aging saturation fracture toughness data are based entirelyon static-cast data for materials with delta ferrite in the range of 15% to 25%.The crack extension resistance value at 2.5 mm of crack extension for the thermal agingsaturation conditions for CF-3 and CF-8 materials ranges between 681 and 820 kJ/m2 at roomtemperature, with values between 516 and 584 kJ/m2 at operating temperature, while the valuesfor CF-8M material are less than half of the low-molybdenum material values.Therefore, the findings from the examination of the data in Figures 14 and 15, and the derived Jvalues at 2.5 mm of crack extension, are completely consistent with the lower-bound thermalaging embrittlement estimates cited previously for both static-cast and centrifugally-cast material.

In other words, the development of screening criteria for the low-molybdenum gradesshould be treated separately from the development of screening criteria for the high-molybdenum grades.9 BWRVIP 2015-025Crack Extension, An (in.)0.0 0.1 0.2 0.3Crack Extension, An (in.)0.1 0.2 0.30.0I .Heat 69 ~CF-3 2WC~2000-~1000.8,-21%290"C -SOAN 32 a 4XOy40 0' 0.0 -eJ.366(AW).4 0 at320C2 4 6 a I]0.10.2 0.3I20aa-~1II200100Heat I CF-3o mw4g(Aa)V bwow 0 A A0 ýL0~ 0~2* 0.5 m~oC47Roomm Temp. A 0 0 UABWC'.10000 "-600I00-10000010o~I=0-10000-60002 4 6 ............

100.D 0.1 0.2 0.3......1 i 4 .1 1 1. 1- 1 l 1 1Heat 68'C"- 290C.0 =Sanm.7,40.4 S,, 32 J.46(Md2B0-0 000000ý000 --0 .2g0Ag0iIi I

IU2 4 6 8Crack Extension, A& (Mm)10Crack Extension, As (mm)Figure 14. Saturation fracture toughness J-R curves at RT and 290Co, estimated from thechemical composition of static-cast CF-3 and CF-8 steels, and determined experimentally 10 BWRVIP 2015-025Crack Extension, ha (in.)Crack Extension, ha (in.)0.0 0.1 0.2 0.31 , .I , I I .I .I I .I .i .0.0 0.1 0.20.3~1000Oj003000-492000-.[OoHiai is 'C", 'Heal 75 CF-SM Ro epStall-Caa SlabSc -25%.Walled5-WARoom Temp.Unapdj .42")0.42-0 OAO~N4WrC tUegod-~J -461(A&P.

37 -32 " 0aadI 20'8NtPA~'Co- .o32 *y oo 0 O00000OOOOC5 j .l44(AS)03

-., I ., .I I , ..I I ' ' '2 4 6 80 0.1 0.2 0.300 2 4' 6 S 140.0 0.1 0.2 0.3-15000 C-10000-3OO-000-15000 £:1m.1Heat 74 CF-SM Room TSARc--CaM Slab6Sc= 16% Uae040o Und J. 57" )o0420 000ht 0C =A324 Olp00. !aOC a *? * .o0,-- J -03*7emp.al 320*CHeat 74 CF-SM 9*taw-cau SWa6C16%32 efpy J 30(sa)03732 OOY N 290'Cj~at320*C C)00N~0 O~0!!!-7 KSaknilF- J .WM~)U. 3Ns00 0. 4 6 0 100.0 0.1 0.2 0.30.0 0.1 0.2 0.3I2000-1000.Hieat 7'581 6F4 R Temp.SONO-C*M EtowSC-21%Unap 32 apy 32 4P&I 290C at 30*CWA OA£ kj MA ALJ/ 2(a03Heat 758 CF-SM 290'CSUMc-CaN ElbowJ .487(aa)0.38~ 32Id290t Id 3py';'>eso4 -?>(V 3,3[ 16000-10000-5000CO*P0 2 4 6 8Crack Extension, Aa (mm)100Crack Extension, Aa (mm)10Figure 15. Saturation fracture toughness J-R curves at RT and 290*C. estimated from diechemical composition of static-cast CF-SM steels, and determined experimentally 11 BWRVIP 2015-025Review and Summary of Combined TE and IE DataThe available data for purely thermally-embrittled CASS materials is quite extensive.

However,that is not the case for a combination of thermal aging and neutron irradiation embrittlement.

Inspite of the limited data, the evaluation of available crack growth resistance data for CF-3, CF-8,and CF-8M material is not altered by including information on material subjected to both thermalaging and neutron irradiation embrittlement, as shown in Figures 23, 41, 49, and 58 from ANL-12/56 [2]. The results for Specimen B-1 (static-cast CF-3 material taken from Heat 69, withapproximately 24% delta ferrite, thermally aged for 10,000 hours0 days 0 hours 0 weeks 0 months at 400'C, and then subjected toa neutron irradiation dose of 0.08 dpa) are shown in Figure 23 -taken from Reference

2. Theextrapolated J value at 2.5 mm of crack extension is 789 kJ/m2, very similar to the saturated, thermally-aged result, as shown in Figure 14 for Heat 69 from Reference 1.6404801-1-743201600.5 1.0 1.Crack Extension (mm)2.0Figure 23. The J-R curve for specimen B-1.Similarly, the results for Specimen F-1 (static-cast CF-8 material taken from Heat 68, withapproximately 23% delta ferrite, thermally aged for 10,000 hours0 days 0 hours 0 weeks 0 months at 400'C and then subjected toa neutron irradiation dose of 0.08 dpa) are shown in Figure 41 -also taken from Reference 2.The extrapolated J value at 2.5 mm of crack extension is 657 kJ/m2, very similar to the saturated, thermally-aged result, as shown in Figure 14 from Reference 1 for Heat 68.12 BWRVIP 2015-025640 ....Specimen F-1CF-8, aged, 0.08 dpa480 Low-DO high-purity water, 3200CI AJ= 372"Aao.62E 320 1V, I160 AI max00.0 0.5 1.0 1.5 2.0Crack Extension (mm)Figure 41. The J-R curve of specimen F-i.Finally, the results for Specimens I-1 (CF-8M material taken from Heat 75, with approximately 28% delta ferrite, unaged but subjected to a neutron irradiation dose of 0.08 dpa) and J-1I (CF-8Mmaterial taken from Heat 75, with approximately 28% delta ferrite, thermally aged for 10,000hours at 4001C, and then subjected to a neutron irradiation dose of 0.08 dpa) are shown inFigures 49 and 58 -also taken from Reference

2. In this case the extrapolated J value at 2.5 mmof crack extension for Specimen J-1 is 466 k/m , which can be compared to the results shown inFigure 15 from Reference I for Heat 75 that appear to be of the order of 300 to 400 kJ/m2 atroom temperature and operating temperature.

13 BWR VIP 2015-025640480.. ......I .ISpecimen I-1CF-8M, unaged, 0.0Low-DO high-purity dna-~5~~cmIIIIwater, 320PC .J7mJ= 336-Aamax I'320160nIAamax=145 kJ/m2II0 0.5 1.0 1.5Crack Extension (mm)Figure 49. The JR curve of specimen I-1.2.0640480(N-3201602.0Crack Extension (mm)Figure 58. The JR curve of specimen J-1.14 BWRVIP 2015-025Conclusions and Recommendation This review and summary of available data on CF-3, CF-8, and CF-8M CASS materials subjected to both individual and combined effects of thermal aging and neutron irradiation embrittlement demonstrated that:Thermal aging embrittlement effects on high-Mo grades are significantly morepronounced than the effects on the low-Mo grades, by roughly a factor of two or morebased upon lower-bound crack growth resistance estimates.

For CF-3, the lower-bound crack growth resistance values at 2.5 mm of crack extension for > 15% delta ferrite varybetween 364 and 478 kJ/m2, and for CF-8, between 343 and 451 kJ/m2.On the otherhand, the CF-8M values vary between 161 and 259 kJ/m2 which are significantly lowercompared to CF-3/8. These results are consistent with the results obtained by examining actual crack growth resistance value measurements at 2.5 mm of crack extension for thethermal aging to saturation for CF-3 and CF-8 materials (low-Mo) that range between681 and 820 kJ/m2 at room temperature, with values between 516 and 584 kJ/m2 atoperating temperature, while the values for CF-8M material (high-Mo) are less than halfof the low-molybdenum material values.* There is no significant difference between the static-cast lower-bound estimates and thecentrifugally-cast lower-bound estimate for CF-3 and CF-8 materials, and those relatively small differences are roughly identical to differences between CF-3 and CF-8 lower-bound estimates.

Using lower-bound crack growth resistance values at 2.5 mm of crackextension as the comparative

measure, the CF-3 static-cast values of 364 and 410 kJ/m2compare very well to the 477 and 478 kJ/m2 values for centrifugally-cast

material, whilethe CF-8 static-cast values of 343 and 366 kJ/m2 compare favorably with the 427 and 451kJ/m2 values for centrifugally-cast material.

The differences between static-cast andcentrifugally-cast values, and between CF-3 and CF-8 values are both inconsequential incomparison to the differences between low-Mo and high-Mo materials.

Therefore thesetwo grades may be considered as a combined category for screening purposes.

The added effect of neutron irradiation embrittlement does not change the conclusions drawn with respect to low-Mo versus high-Mo fracture toughness.

Finally, since theconclusions are drawn primarily from static-cast fracture toughness data, arecommendation to the NRC staff to maintain the screening threshold for low-Mo static-cast grades at 20% is supported by the data.Therefore, this review shows that the CF-8M data provide an extremely conservative and thusunrepresentative lower bound to the data for the low-molybdenum grades. As such, the CF-8Mlower bound is separable and distinct from the low-molybdenum grade lower bounds, whichtherefore require their own screening category and fracture toughness estimates.

Thus, the deltaferrite levels stated in the BWRVIP-234 RAI response (from the Grimes letter) is furthersupported by this evaluation.

Finally, this review and the associated quantitative comparisons utilize the data sets as delineated in extensive documentation in NUREG/CR-4513, Revision 1 [1] and other relevant publically available references.

15 BWRVIP 2015-025References

[1]. 0. K. Chopra, "Estimation of Fracture Toughness of Cast Stainless Steels DuringThermal Aging in LWR Systems,"

NUREG/CR-4513 (ANL-93/22),

Revision 1,Argonne National Laboratory,

Argonne, Illinois (August 1994).[2]. Y. Chen, B. Alexandreanu, and K. Natesan, "Crack Growth Rate and Fracture Toughness Tests on Irradiated Cast Stainless Steels,"

Report No. ANL-12/56, Argonne National Laboartory,

Argonne, Illinois (November 2012).16 BWRVIP 2015-025Attachment BConclusions Regarding Screening Criteria for Irradiation Embrittlement of CASSMaterials The screening criteria for TE and IE in CASS components proposed and discussed with the NRC[1] appears to be based on materials testing data that includes a high-molybdenum

material, grade CF-8M, presented in Figure 64 in NUREG 7027 [2], with the following conclusions onpage 79:1. CF-8M materials represent a 'worst case' for thermal embrittlement

2. Thermal aging does not seem to lower the toughness below that expected for irradiation alone at the dose levels examined.

In regards to the use of this data, it is important to note that:* CF-8M materials are not present in LWR core internal components.

Moreover, the embrittlement identified in NUREG-7027 is conservative because manyof the heats examined had delta ferrite contents greater than 20% and as high as 42% [3].It is also important to emphasize that in LWR reactor internals, delta ferrite content doesnot typically exceed 20% as compared to the materials assessed in the NUREG database[2, 3].* Loss of fracture toughness due to thermal embrittlement is driven by the amount of deltaferrite in a given material, thus the use of high delta ferrite material data is alsoconservative.

ANL results comparing loss of fracture toughness measurements on unaged, thermally aged only, irradiation aged only and sequentially thermally aged and irradiation agedCF-3 and CF-8 materials show significant loss of toughness on all forms of aging.Measured toughnesses of thermally aged, irradiation aged and sequentially thermal plusirradiation aged materials are similar.

Given the variances that are inherent in estimating J values as material characteristics, the differences in loss of toughness cited inReference 4 cannot be taken to unequivocally demonstrate that irradiation and thermalaging are significantly additive effects to the loss of toughness in CF-3 and CF-8 castaustenitic stainless steels.* The data summarized in Figure 64 of NUREG-7027 includes many high ferrite contentand Mo-containing chemical compositions.

The correlations developed in analysesbased on this database are therefore expected to be very conservative with regard to theirradiation and thermal response of low ferrite content and low Mo containing chemicalcompositions.

17 BWRVIP 2015-025The NRC is concerned that interaction between thermal and irradiation embrittlement mechanisms may lower the fracture toughness of CASS material below levels expected from theeffect of either mechanism alone. Based on the potential for a synergistic effect and taking intoaccount the NUREG-7027 data, the NRC has concluded that a very low level of irradiation would reduce the fracture toughness of a CASS component below the 255 KJ/m2 criteria.

This led to the NRC proposing a neutron irradiation screening criteria of 0.45 dpa for lowmolybdenum static CASS components with 15-20% delta ferrite.

By reexamination of the datain NUREG-7027 a considerable amount of conservatism can be removed from the analysis whilestill providing reasonable assurance that CASS components in the LWR environment will retainadequate fracture toughness through the period of extended operations.

The irradiated materials data from NUREG-7027 were digitized from the plots of NUREG-7027 and reanalyzed to produce a trend curve for irradiated CF-3 and CF-8 materials only as shown inFigure 1 below. Note that the exposures of these materials not only represent irradiation effectsbut also some degree of thermal embrittlement.

These data for CF-3 and CF-8 materials aremost relevant to the behavior of CASS in LWR internals, as these are the materials that areprincipally known to be present in LWR reactor internals.

Thus the correlation and proposedscreening criteria are appropriately developed for these materials.

As an outcome of this analysis a plot of fracture toughness against neutron exposure was created,and a 'best fit' line was developed and the resulting curve was calculated to have a strongcorrelation with the data (R2=0.94). The best fit line was then modified using conservative engineering judgment and standard curve offsetting techniques to bound all of the irradiated CF-3 and CF-8 material measured toughness values. (i.e., the pre-exponential constant multiplier was reduced by approximately 25% and the exponential factor was increased by 35% in order tobound every data point above the red-line exponential curve fit.) A constant of 30 kJ/m2 wasalso included as a saturated fracture toughness value.This saturated fracture toughness value is proposed by Chopra in NUREG-7027, and the overallform of the equation is consistent with irradiation hardening models [6]. This results in a curvefit where all data points are above the curve, with resulting curve fit of:KJ e_.25xdpa J2.Smm = 30 + 520 -- x e xThis bounding line is shown as a red line in Figure 1.By using the red (lower bound) line as a gauge for how CF-3 and CF-8 materials will age in theLWR environment, it can be shown that these materials will not encroach upon the 255 KJ/m2fracture toughness limit until the fluence level experienced by these components approaches 3.3dpa. The industry proposed use of I dpa for neutron exposure screening

criteria, provides morethan a 2 dpa margin on neutron exposure between a CASS component at 1 dpa and the 25518 BWRVIP 2015-025KJ/m2 fracture toughness limit. At 1 dpa, the fracture toughness predicted by the bounding curvein Figure 1 is 435 KJ/m2, which results in a fracture toughness margin of 180 KJ/m2.Therefore, the industry's proposed criterion of 1 dpa has substantial margin in both the fluenceand fracture toughness domains.

This analysis and argument supports the use of the irradiation screening criteria of I dpa proposed by the industry for low molybdenum CASS components with ferrite content less than 20 percent.Additionally, a study designed to investigate the potential interaction between thermal andirradiation embrittlement mechanisms concluded that irradiation can reduce the extent of thermalaging effects [5], suggesting an antergistic effect between the two mechanisms rather than asynergistic one.This information indicates that the NRC's position is highly conservative, and that the industry's position on the screening criteria of CASS components, while also conservative, is justified.

References:

1. U.S. NRC, "NRC Staff Compiled Comments on Industry CASS Screening Position,"

July 18, 2014 (NRC ADAMS Accession Number ML14198A282.

2. NUREG/CR-7027, "Degradation of LWR Core Internal Materials due to NeutronIrradiation,"

December, 2010 (NRC ADAMS Accession Number ML 102790482.

3. NUREG/CR-4513, "Estimation of Fracture Toughness of Cast Stainless Steels DuringThermal Aging in LWR Systems,"

U.S. Nuclear Regulatory Commission, August 1994(NRC ADAMS Accession No. ML052360554).

4. ANL-12/56, "Crack Growth and Fracture Toughness Tests on Irradiated Cast Stainless Steels,"

Argonne National Lab., November 2012.5. K Fuji, K. Fukuya, "Effects of Radiation on Spinodal Decomposition of Ferrite in DuplexStainless Steel," Journal of Nuclear Materials, May 2012. Presented at the NuMat 2012Conference, pages 613 to 616, October 22-25, 2012, Osaka, Japan.6. Was, Gary S., "Fundamentals of Radiation Materials Science:

Metals and Alloys,"Springer-Verlag, 2007.7. Kim, C., R. Lott, S. Byrne, M. Burke, and G. Gerzen, "Embrittlement of Cast Austenitic Stainless Steel Reactor Internals Components,"

Proc. 6th Intl. Symp. On Contribution ofMaterials Investigations to Improve the Safety and Performance of LWRs, Fontevraud 6,French Nuclear Energy Society, SFEN, Fontevraud Roayal Abbey, France, September 18-22, 2006.19 BWRVIP 2015-025Figure 1 -Irradiated CF-3 and CF-8 Fracture Toughness Data from Literature (Refs. 2, 4 and 7)Fracture Toughness of Irradiated CF3 and CF8 (0.02 dpa and greater)1000900800700E 500~40030020010000 1 2 3 4 5 6 7 8 9 10Neutron Exposure (dpa)11 12 13 14 15 16 17 1820 Attachment CReactor Internals CASS Flaw Tolerance In the draft Interim Staff Guidance (ISG) on Aging Management of CASS Reactor VesselInternal Components, issued in June 2014, the NRC technical position observed that "Thefracture toughness screening.

value of 255 kJ/m2 specified in the Grimes Letter is based on ageneric flaw tolerance evaluation for piping, and may be overly conservative for RVI CASScomponents that are subject to mainly compressive stresses during operation, and are part of apopulation of redundant components where failure of individual components can be tolerated."

The technical position went on to add that "therefore, the staff applies the 255 kJ/m2 value forscreening purposes with the knowledge that there likely is additional conservatism present in thisscreening for non-pressure boundary RVI components."

In an effort to provide further evidencewith respect to the NRC staff technical

position, some industry efforts since the July 15, 2014meeting with the NRC staff have been directed toward this issue. The results are summarized inthe following paragraphs.

The industry selected three different geometries that can be related to typical BWR and PWRCASS reactor internals in order to estimate flaw tolerance capability:

(1) a large-diameter cylinder with a through-wall vertical flaw located in a longitudinal seam weld, subjected tocombined membrane and bending stress; (2) an edge-cracked beam-column subjected tocombined membrane and bending stress; and (3) a six-inch-diameter, Schedule 40 pipe with athrough-wall longitudinal flaw subjected to internal pressure.

The internal pressure for the thirdgeometry was selected such that the circumferential tensile stress in the pipe was identical to themembrane tensile stress level chosen for the first two geometries.

In all three case studies, theinitial flaw size (length for the first and third geometries, and depth for the second geometry) wasselected to be consistent with reactor vessel internals fabrication workmanship standards.

Noflaw growth criteria were applied to the initial flaw sizes, although the initial flaw sizes wereincreased to some extent in order to determine the rate at which the crack driving force increased as a function of the increase in flaw size.For the first geometry (large-diameter cylinder with through-wall vertical flaw), the diameterwas selected to be 452 inches with a wall thickness of either one or two inches, the membranetensile stress was selected to be 5 ksi, and the bending stress was selected to be 7.5 ksi. The totalinitial flaw length was selected to be 0.226 inches (about 1/4-inch in length),

but was increased inincrements up to 4.294 inches. The applied stress intensity factors were found using linear-elastic fracture mechanics (LEFM), and were then converted to J-crack driving forces. For thelargest initial flaw lengths examined, the typical LEFM stress intensity is of the order of 25ksiin, which converts to an elastic-plastic crack driving force of about 4 kJ/m2.This implies avery large margin relative to the screening value of 255 kJ/m2.For the second geometry (edge-cracked beam column),

the solid column diameter was selectedto be three inches, and the membrane and bending stresses were selected to be identical to thoseselected for the first geometry.

The smallest initial flaw depth was selected as 0.1 inches, andwas incremented by 0.1 inch up to a maximum depth of 1.7 inches. The applied stress intensity factors were found using linear-elastic fracture mechanics (LEFM), and were then converted toTogether

... Shaping the Future of Electricity PALO ALTO OFFICE3420 lillview Avenue, Palo Alto, CA 94304-1395 USA

650.855.2000

Customer Service 800.313.3774

www.epri.com BWRVIP 2015-025J-crack driving forces. For the largest initial flaw depth examined (1.7 inches),

the LEFM stressintensity was 42 ksiIin, which converts to an elastic-plastic crack driving force of about 10kJ/m2.Again, this calculation implies a very large margin relative to the screening value of 255kJ/m2.For the third geometry (six-inch diameter Schedule 40 pipe with a through-wall longitudinal flawsubjected to internal pressure),

the internal pressure was selected such that the circumferential stress was approximately 5 ksi, matching the membrane tensile stress used for the first andsecond geometries.

Because of the thin pipe wall, this case was found to be the most critical ofthe three geometries.

The initial flaw length was selected to be 0.1 inches, and was thenincremented in 0.1-inch increments up to a maximum flaw length of 2.0 inches. For this case,the applied stress intensity factors were found using linear-elastic fracture mechanics (LEFM),and were then converted to J-crack driving forces. In addition, this case was also analyzed usingelastic-plastic fracture mechanics, in order to compare derived crack-driving forces with directlycalculated elastic-plastic crack-driving forces. For an initial flaw length of 2.0 inches, the LEFMstress intensity factor was about 54 ksi/in, which converts to a crack driving force of about 96in-lb/in2 or about 16 kJ/m2.The directly-calculated elastic-plastic crack driving force is almostexactly the same, but very slightly lower. When the initial flaw length was doubled -to about4.0 inches, the applied LEFM stress intensity was found to be 120 ksi/in, which converts to acrack driving force of about 476 in-lb/in2 or about 83 kJ/m2.Even for this very severe example,the calculations show a sizable margin relative to the screening value of 255 kJ/m2.From this exercise, CASS reactor internals components subjected to nominal stress levels, evenin the presence of initial flaws that are well beyond fabrication workmanship acceptance

criteria, are extremely flaw tolerant, with margins against flaw instability of the order of a factor of fiveto 10 relative to the fracture toughness screening criterion of 255 kJ/m2 specified in the Grimesletter. When this flaw tolerance is coupled with the additional margin inherent in the separation of CF-3/CF-8 screening data from CF-8M screening data, the conservatism of the industrytechnical position is further confirmed.

22 BWRVIP 2015-025Attachment DTahlo I virarlld o/- larrto RPnonpe (Indivetru PnaifinnI 23

'e,L 2l 1I ELECTRIC POWERRESEARCH INSTITUTE 2015-025BWR Vessel & Internals Project (BWRVIP)March 09, 2015Document Control DeskU.S. Nuclear Regulatory Commission 11555 Rockville PikeRockville, MD 20852Attention: Joseph Holonich

Subject:

Project No. 704 -Summary of Industry Position on Screening Criteria forThermal and Irradiation Embrittlement for PWR and BWR Reactor Internals Fabricated of Cast Austenitic Stainless Steel

References:

1. Letter from D. Madison (BWRVIP Chairman) and A. McGehee (BWRVIP Program Manager) to J. Holonich (NRC), Project No. 704 -BWRVIPResponse to NRC Request for Additional Information on BWRVIP-234, datedMay 23, 2014. ADAMS -ML14174A841.

2. Letter from J. Holonich (NRC) to A. Mendiola (NRC), Summary of the July 15,2014, Meeting with the Electric Power Research Institute on Items Related toCast Austenitic Stainless Steel and the Materials Reliability Program-227-A,"Pressurized Water Reactor Internals Inspection And Evaluation Guidelines,"

September 9, 2014. ADAMS -ML14127A077.

3. "NRC Position on Aging Management of CASS Reactor Vessel InternalComponents,"

June 23, 2014, ADAMS -ML 14174A719.

4. Email from C. Wirtz (EPRI) to J. Holonich (NRC), October 7, 2014.5. NUREG/CR-4513, ANL-93/22, Revision 1, "Estimation of Fracture Toughness of Cast Stainless Steels During Thermal Aging in LWR Systems,"

ArgonneNational Laboratory, May 1994.The purpose of this letter is to provide the status of the Industry (PWR and BWR) activities associated with developing and proposing a generic screening criteria to address thermal andirradiation embrittlement of RPV internals fabricated of cast austenitic stainless steel (CASS).Reference I was provided to the NRC in response to a Request for Additional Information (RAI) regarding BWRVIP-234.

The RAI response contained supplemental information that provided the technical bases for an Industry screening criteria for CASS internals.

Together

...Shaping the Future of Electricity Qýý U*PALO ALTO OFFICE3420 Fillview Avenue, Palo Alto, CA 94304-1395 USA

650.855.2000

Customer Service 800.313.3774 9 www.epri.com BWRVIP 2015-025A meeting was later held on July 15, 2014 at NRC offices in Washington, DC to discuss theIndustry and NRC positions (References 2, 3). Based on that meeting, the Industry committed to investigate additional information where there was disagreement with the NRC regarding the screening criteria and provide said information to the NRC at a later date. Thatinformation was informally submitted for NRC's consideration (Reference