ML20195C773
| ML20195C773 | |
| Person / Time | |
|---|---|
| Site: | Millstone  |
| Issue date: | 10/13/1998 |
| From: | Chesworth S, Deardorff P STRUCTURAL INTEGRITY ASSOCIATES, INC. |
| To: | |
| Shared Package | |
| ML20195C762 | List: |
| References | |
| SIR-98-096, SIR-98-096-R00, SIR-98-96, SIR-98-96-R, NUDOCS 9811170216 | |
| Download: ML20195C773 (139) | |
Text
{{#Wiki_filter:1 I Repor- No.: SIR-98-096 Revision No.: O Project No.: NUSCO-22Q File No.: NUSCO-22Q-403 October 1998 Pressurizer Surge Line Leak-Before-Break Evaluation Millstone Nuclear Power Station, Unit 2 Preparedfor: Northeast Utilities System Contract 02063726 Prepared by: StructuralIntegrity Associates,Inc. San Jose, California Preparedby: f l Date: /D//5)9[
. @ariforff/P Preparedby: 0! {v V) Date: ICll5 9ff S. T. Chesworth Reviewed by: $lW 'W Date: Id/' 3kt %
W . F. W eit g f '
~
Reviewed by- & Date: /*f/3f?8 N. G. Cofie. Ph.D. Approved by: lJi / Date: A//]/ff
/ 'AMe'iircio'rff. P'.E '
~
/
U, StructuralIntegrity Associates, Inc. 9811170216 981109 PDR ADOCK 05000336s P PDR , a
REVISION CONTROL SHEET Document Number: SIR-98-096 l
Title:
Pressurizer Suree Line Leak-Before-Break Evaluation ! Millstone Nuclear Power Station, Unit 2 Client: Northeast Utilities System l l SIProject Number: NUSCO-22Q Section Pages Revision Date Comments 0 10/06/98 Draft Issue i to x 0 10/13/98 InitialIssue I l-1 to 1-7 0 2 2-1 to 2-2 0 3 3-1 to 3-3 0 4 4-I to 4-4 0 5 5-1 to 5-4 0 6 6-1 to 6-8 0 7 7-1 to 7-2 0 8 8-1 to 8-2 0 Appendix A A-0 to A-30 0 Appendix B B-0 to B-11 0 i Appendix C C-0 to C-16 0 Appendix D D-0 to D-37 0 . Total number of pages = 140 l-l
- I 6
f f StructuralIntegrityAssociates,Inc. 1 . \
l l l l l
SUMMARY
i The pressurizer surge line at Millstone Nuclear Power Station Unit 2 was evaluated for leak-before-break (LBB) behaviorin accordance with 10 CFR 50, Appendix A GDC-4 and NUREG-1061, Vol.
- 3. The pressurizer surge line begins at the hot leg and ends at the pressurizer. It is 12-inch Schedule 160 piping, fabricated from cast stainless steel. The operating pressure for the pressurizer surge line is 2235 psig. The operating temperature is 604*F at the hot leg end and 653*F at the pressurizer end 4
of the surge line, l 1 The. evale W consisted of determining critical flaw sizes and leakage flaw sizes at all weld locations u ,m n . sses at each location. The critical flaw sizes were calculated using the elastic-L plastic fracture mechanics (EPFM) J-Integral / Tearing Modulus (J/T) approach and conservative lower-bound genetic material properties. The detectable leakage at Millstone Unit 2 is 1.0 gpm. Based on the required margin of 10 on leakage in NUREG-1061, Vol. 3, the critical flaw sizes must . be greater than the crack size which produces 10 gpm leakage to assure leak-before-break. Critical j j flaw sizes were determined for normal operating + SSE loads and were shown to be at least two L times the length of 10 gpm leakage flaws. Critical flaw sizes were determined with a factor of ,[2 on normal operating loads and were shown to have a length greater than the 10 gpm leakage flaws. A fatigue crack growth analysis was also performed to determine the growth of postulated semi-elliptical, inside surface flaws with an initial size allowed by ASME Section XI acceptance standards. L The following provides a summary of the LBB evaluation and a statement of how it meets the l "llecommendations for Application of the LBB Approach"in the NUREG-1061 Vol. 3 executive summary: p i (a) The pressurizer surge line piping is constructed of very ductile stainless steel that is not susceptible to cleavage-type fracture. In addition, it has been shown to be not susceptible to the effects of corrosion and water hammer. Although fatigue is identified as a possible
- degradation mechanism,its effect will not be significant when managed by inspection.
i SIR-98-096 Rev. O iii StructuralIntegrity Associates, Inc.
(b) Loadings have been determined from piping analysis, and were due to pressure, dead weight, thermal expansion, thermal stratification and seismic effects. All weld locations in the piping and terminal ends at the nozzles were considered. (c) Material properties archived from initial plant construction were used in the evaluation. When archived data was not available, lower-bound generic industry data for the piping, safe-end, nozzle and welds were conservatively used. The effect of thermal aging was considered in determining the material properties for the cast stainless steel base metal and welds. (d) Crock gicwth analysis was conducted at all weld locations. All cyclic stresses, including thermal stratification, predicted to occur over the life of the plant were considered. The ASME Section XI .,lloivable 6:1 aspect ratio flaw for normal / upset conditions, based on the worst case stratification plus OBE with other nominal stresses, exceeds 60 percent of the wall thickness. For hypothetical flaw with an initial depth of 11.1% of the pipe wall, it takes in excess of 300 heatup/cooldown cycles for the deepest point of this flaw to reach the ASME allowable flaw sizes. For 125 heatup/cooldown cycles, the initial flaw will grow to less than 21 percent of the pipe wall thickness. For a very long hypothetical flaw at the most highly stressed location, it would take in excess of 60 heatup/cooldown cycles to grow the flaiv to the Section XI allowable size. This crack growth is not considered to be significant such that it can not be managed by inspection. (e) The detectable leakage at Millstone 2 is I gpm. Based on the required margin of 10 on leakage, the leakage flaw size has been taken as the flaw size predicted to produce 10 l gpm leakage. (f) The pressurizer surge line considered in this evaluation is approximately 60 feet in length and is not geometrically complex. All transients experienced by the surge line including thermal stratification were considered in the evaluation. Thus, no system evaluation was conducted. SIR-98-096, Rev. O iv Structural Integrity Associates, Inc.
(g) It was shown for tne controlling location that crack growth of a leakage size crack for 10 SSE seismic cycles woald grow the crack by no more than 1% of the leakage flaw size. This is not significant compared to the margin between the leakage crack size and the critical crack size. (h) For all locations, the critical size circumferential crack was determined for the l combination of nonnal plus safe shutdown eanhquake (SSE) loads. The 10 gpm leakage j size crack was shown to have a length that was at least a factor of two less than the critical crack size. All of the possible materials associated with the weld and the base material o either side were considered in this evaluation. Axial cracks were not l considered since these always exhibit higher leakage than c'ircumferentially-oriented cracks. l (i) For all locations, the critical crack size was determined for the combination of 8 times the normal plus SSE stresses. The 10 gpm leakage size crack was shown to be no greater than the critical crack size (2a) so determined. .All of the possible materials associated l with the weld and the base metal at either side were considered in this evaluation. (In no case was the limiting critical flaw size governed by this criterion.) l l (j-n) No additional testing was conducted to determine material properties beyond the properties in the cenified material test reports tCMTRs). When available, the information in the CMTRs was used. Otherwise generic lower bound piping material toughness and tensile properties were used in the evaluations. The matenal properties so determined have been shown to be applicable near the upper range of nonnal plant operation and exhibit ductile behavior at these temperatures. This data is widely accepted by industry for conducting J/T fracture mechanics analysis. I SIR-98-096, Rev. O v Structural lategrity Associates, Inc.
, ., - .. . - - -= . . . . - - . . . - _.- . _ . . - . - - _. (o)- - Limit-load analysis was not utilized in this evaluation. Thus, it is concluded that the the pressurizer surge line piping evaluated in the report qualifies for the application ofleak-before-break analysis to demonstrate that it is very unlikely that the piping could experience a large pipe break when coupled with an inspection program to address the effects of thermal fatigue.- e e SIR-98-096, Rev. 0 vi gg,,ggy,,,y,g,,,,gypggggjgggg, jng,
Table of Contents Section Pace 1.0 INTRO D U CTIO N ...... .......... ......... - .. .. .
.... ..... ......... .......... 1- 1 1.1 B ac k g rou n d . . . . . . .. . . .. . . . . . . . . . . . . . .. . .. . . . . . ...... .. . .. . .. .. .. . .. .... .... .. .. ... . . ... . . . . . . . . . . . ... .. . . . . . . . . .. . . . . . . . . . I 1.2 Leak-Before-B reak M ethodology ..... ..................................... ....... ....... ............ ..... .... I- 1 1.3 Report Organization ................. ..... ... ............. ... . ... ....... .... .... . ... .. . ...... . .. .... .. . .. . . .. . . 1 -3 2.0 CRITERIA FOR APPLICATION OF LEAK-BEFORE-BREAK APPROACH....... 2-1 2.1 Criteria for Through-Wall Flaws..... .................................... ................. ..................... 2- 1 2.2 - Criteria for Pan-Through-Wall Flaws............................ ................................ ............... 2-2 2.3 Other Mec h anis ms. ..................... .... . . .... .. ...... ................ ........... .............. ... ..... . .. . ... 2-2 3.0 CONSIDERATION OF WATER HAMMER, CORROSION AND FATIGUE........ 3-1 3.1 Wate r Ham me r . . . . . . . . . ... . . . . . . . . .. .. . . ... .. .. .. . ... .... .. .. . . ...... .. . .. . . ... .. .. . . ... . . . . . . . . . . . . . . . . . . . . . 3 3.2Conosion....................................................................................................3-2 3.3 Fatigue.....................................................................................................3-2 4.0 PIPING MATERIALS AND STRESSES .. . ..................4-1 4.1 Piping System Description ...... ........ ............. ... ....... ............................ ... . . ..... ..... . 4- 1 4.2 M ateri al Properties . .. .. . . ... . . . ... . ... . .. ... . . .... . ... ... .... . . . .... . . . . . ... . . ... . . . . . .... . . .. .. . . . . .. 4- 1 4.3 Pipin g Load s ...... .. . . ...... .... . . .. .. .. . .. .. .... ... .. .... ....... . ........................................4-2 5.0 LEAK-BEFORE-BREAK EVALUATION... .._ .............................5-1 5.1 Evaluation of Critical Flaw Sizes..... ............ .... ..... ... .......... ..........................5-1 5.2 Leakage Rate Determin ation ... .............. ..... ... . ... ......................... ..... ......... .. ...... . 5-2 5.3 LB B Evaluation Results and Discussions ...... ............. .............. ... ...... .. .... ....... ... 5-2 6.0 EVALUATION OF FATIGUE CRACK GROWTH OF SURFACE FLAWS........... 6-1 6.1 Plan t Transient Eval uation ...... ... ...... ........... . ..... . ............ . .... .. ...... .. .... . .. ..... 6-1 6.2 Thennal S tratification S tresses... ... ............... .... ... ............ ... .... .......... ..... ... .. ... .. 6-2 6.3 Weld-Related S tres ses .... .... . . ... .. . . ............ .... ........ ... ... .. ... ... .. . ... .. .. ... . . . . . .. . . .... . . .. 6-2 6.4 Loadi n g Combi natio ns ....... . ........... . .. .......... . .. . ......... ...... . .... . . . .. ... .. .. . .. ... . . . ... . . ... . . 6-2 6.5 Fracture Mechanics Model... .... .............................................................. ...6-3
- 6. 6 Crack G ro wth Laws . ... . ... ... .. ..... . .. ... ... . ....... . ........ ... . . . .. .. .. . . .. . . . . .. ... . . .. . . . . . s . . . .. . . . 6-3 6.7 Allowable Flaw Size Evaluation . .. ......... ........... ... ..... .. ....... .......... ......... ....6-4 6.8 Crack Growth for SSE during Normal Operation...... . ........ . . . .. ...... .. .... ..... . 6-4 6.9 Thermal Striping Crack Growth.. ....... ........... . . .. . . . . . . . . . . . . . ..................6-5 6.10 Crack Growth Results . .. . ........... ... . . ...... . ..... . . . . . . . . . . . . . . . . . ... .. ...... 6-5 7.0
SUMMARY
AND CONCLUSIONS. . ...... . . ... .............................. 7-1 8.0 RE FE R E N C ES ........ ................ ....... ...... ....... . .. .. ....... ....... ................... 8 - 1 1 SIR-98-096, Rev. O vii StructuralIntegrity Associates, Inc.
1 1 i l Table of Contents (continued) Section - Pages 1 l 1 l APPENDIX A M aterials Evaluation...................... . .......................... ............. ............ A0 to A3 APPENDIX B Pipi ng Loads .. ............................................................ .. ........... ............. B0 to B 1 1 I i APPENDIX C Critical Flaw Size and Leakage Evaluation.................... ........ . ............C0 to C16 APPENDIX D Fatigue Crack Growth Evaluation .... .......... ...... ......... ........................ D0 to D37 ) i i I 1 l i 1 l I 6 s S1R-98-096, Rev. 0 viii Structurallategrity Associates, Inc.
Table of Contents (continued) Section ' Pages . APPENDIX A Materials Evaluation............ ................... ......... .......... ............ ........... A0 to A3l APPENDIX B Piping Loads ................. .... .......... ............................... .. .......... ............ B0 to B 1 1 i APPENDIX C Critical Flaw Size and Leakage Evaluation.. . .............. ....... ..............C0 to C16 APPENDIX D Fatigue Crack Growth Evaluation .... ..................... ..... ............. ......... D0 to D37 1
' SIR-98-096, Rev. 0 viii StructuralIntegrity Associates, Inc.
A
List of Tables Table Eage Table 5-1 Comparison of Critical and 10 gpm leakage Flaws for Piping Welds............ . ..... 5-3 Table 5-2 Comparison of Critical and 10 gpm Leakage Flaws for Nozzle Welds.................... 5-4 Table 6-1 Flaw Sizes After 125 and 250 Heatup/Cooldown Cycles Table D-4 Cyclic Operation S tress S tates .................................................................................. [ 4-7 p w. tolli /iE l l l i 3 SIR-98-096, Rev. O ix StructuralIntegrity Associates, Inc.
List of Figures Figure Page Figure 1-1. Representation of Postulated Cracks in Pipes for Fracture Mechanics Leak-Be fore-B reak Analysis .. .. .. .... . . . . . .... . .... . ... . . . . ........ . . . ... ... .. ... ... . . . . .... ...... .. ... . .. 1 -5 Figure 1-2. Conceptual Illustration ofISI(UT)/ Leak Detection Approach to Protection A g ai n st Pipe Ruptu re .. ... ................... ............ ...... ... . ... ... .. ..... . . .. ... ... ..... ...... .... . . . .. .. 1 -6 Figure 1-3. Leak-Before-Break Approach Based on Fracture Mechanics Analysis with In-service Inspection and Leak Detection ............ .............. ...... ........................... 1 -7 Figure 4-1. Millstone l> nit 2 Pressurizer Surge Line ...................................... .......................4-4 Figure 6-1. Maximum Deepest-Point Crack Growth Time History at Nodes 3, 4, 8, 10 and 13 ........ . .. ..... .. . .. . . . . . ..... . . . . . . .. .... .. .... . . . .. ... .. ... . . .. .. . . . .. .. . .. 6-8 SIR-98-096, Rev. 0 x StructuralIntegrity Associates, Inc. l D
l 1.0 -INTRODUCTION L 1.1 ' Background l
- This report documents evaluations performed by Structural Integrity Associates (SI) to determine l
l the leak-before-break (LBB) capabilities at all weld locations on the pressurizer surge line piping at ! Millstone Nuclear Power Station Unit 2 (Millstone 2). These evaluations were undenaken to
- address high energy line break concems at these locations. The entire portion of the pressurizer j surge line at Millstone 2 extending from the nozzle at the reactor coolant loop to the nozzle at the pressurizer was considered in the evaluation. This line is fabricated from cast stainless steel.
1.2 Leak-Before-Break Methodology NRC SECY-87-213 [1] covers a final broad scope mle to modify General Design Criterion 4 L (GDC-4) of Appendix A,10 CFR Part 50. This amendment to GDC-4 allows exclusion from the design basis of all dynamic effects associated with high energy pipe mpture by application of LBB technology. l l Definition of the LBB approach and criteria for its use are provided in NUREG-1061 (2]. Volume 3 of NUREG-1061 defines LBB as " ..the application of fracture mechanics technology to demonstrate that high energy fluid piping is very unlikely to experience double-ended ruptures or their equivalent as longitudinal or diagonal splits." The panicular crack types of interest include circumferential through-wall cracks (TWC) and part-through-wall cracks (PTWC), as well as axial or longitudinal through-wall cracks (TWC), as shown in Figure 1-1. LBB is based'on a combination ofin-service inspection (ISI) and leak detection to detect cracks, coupled with fracture mechanics analysis to show that pipe rupture will not occur for cracks smaller ( than those detectable by these methods. A discussion of the criteria for application of LBB is presented in Section 2 of this report, which summarizes NUREG-1061, Vol. 3 requirements. l i SIR-98-096, Rev. O l-1 StructuralIntegrity Associates, Inc.
The approech to LBB which has gained acceptance for demonstrating protection against high energy line break (HELB) in safety-related nuclear piping systems is schematically illustrated in Figure 1-2. Essential elements of this technique include critical flaw size evaluation, crack propagation analysis, volumetric nondestructive examination (NDE) for flaw detection / sizing, leak detection, and service experience. In Figure 1-2, a limiting circumferential crack is modeled as having both a short through-wall component, and an axisymmetric pan-through-wall crack component. Hypothetical flaws of finite aspect ratios considered in the fatigue crack growth evaluation are presented in Section 6. I2ak detection establishes an upper bound for the through-wall crack component while volumetric ISIlimits the size of undetected part-through-wall defects. These detection methods complement each other, since volumenic NDE techniques are well suited to the detection of long cracks while leakage monitoring is effective in detecting shon through-wall cracks. The level of NDE required to suppon LBB involves volumetric inspection at intervals determined by fracture mechanics crack growth analysis, which would prechide the growth of detectable part-through-wall cracks to a critical size during an inspection interval. The objective of the fatigue evaluation is to ensure that an undetected flaw xceptable per ASME Section XI, will not grow significantly during service. For through-wall defects, crack opening areas and resultant leak rates are compared with leak detection limits. The net effect of complementary leak detection and ISI is illustrated by the shaded region of Figure 1-2 as the largest undetected defect that can exist in the piping at any given time. Critical flaw size evaluation, based on elastic-plastic fracture mechanics techniques, is used to determine the length and depth of defects that would be predicted to cause pipe rupture under specific design basis loading conditions, including abnormal conditions such as a seismic event and including appropriate safety margins for each loading condition. Crack propagation analysis is used to determine the time interval in which the largest undetected crack could grow to a size which would impact plant safety margins. A summary of the elements for a leak-before-break analysis is shown in Figure 1-3. Service experience, where available, is useful to confinn analytical predictions as well as to verify that such cracking tends to develop into " leak" as opposed to " break" geometries. SIR-98-096, Rev. O l-2 StructuralIntegrity Associates, Inc.
In accordance with NUREG-1061, Vol. 3 [2] and other NRC guidance on the topic, the leak-before-break technique for high energy piping systems in a nuclear power plant should include the following considerations. t
'e Elastic-plastic fracture mechanics analysis ofload carrying capacity of through-well cracked pipes under worst case normal loading, with safe-shutdown eanhquake (SSE) loads included.
Such analysis should include elastic-plastic fracture data applicable to pipe weldments and weld heat affected zones where appropriate. Determination of through-wall sizes that will produce detectable leakage. Linear elastic fracture mechanics analysis of subcritical crack propagation to determine ISI(in-service inspection) intervals for long, part-through-wall cracks. Piping stresses have a dual role in LBB evaluations. On one hand, higher maximum (design basis) stresses tend to yield lower critical flaw sizes, which result in smaller flaws for leakage and a lower leakage rate. On the other hand, higher operating stresses tend to open cracks more for a given
. crack size and create a higher leakage rate. Because of this duality, the use of a single maximum stress location for a piping system may result in a non-conservative LBB evaluation. This LBB evaluation will, therefore, be performed in such a manner that each nodal location of the piping model of the pressurizer surge line will be specifically addressed.
L 1.3 Report Organization Section 2 of this report provides the main evaluation criteria in NUREG 1061, Vol. 3 that need to be addressed in order to assure leak-before-break for the pressurizer surge line at Millstone 2. Among these criteria is the requirement to demonstrate that there are no potential failure mechanisms than can undermine the LBB evaluation. This is discussed in Section 3. The description of the system parameters for the pressurizer surge line is provided in Section 4. Summary of the piping material properties and piping loads is also provided in this section. Details of the material properties determination and the development of piping loads and stresses SIR-98-096, Rev. O l-3 f StructuralIntegrityAssociates,Inc. \ .,........,...t .. , . .
are provided in Appendices A and B respectively. In Seuion 5, a summary of critical and leakage flaw sizes is provided with details of the evaluation provided in Appendix C. Fmigue was identified as a possible degradation mechanism for the surge line and therefore . comprehensive fatigue evaluation was performed. A summary of this evaluation is provided in Section 6 with further details provided in Appendix D. Finally, summary and conclusion of the LBB evaluation are provided in Section 7 and references used in the report are provided in Section 8. SIR-98-096, Rev. 0 - 1-4 StructuralIntegrity Associates, Inc.
' ' - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ' ~ ~ ~ ^ ~
uf 2a - 2e W l f I l 1
- a. Circumlerential and Lcna'tudinalThrough-Wall Cracks oflength 2a. .
l l
\
l R
]
t a esass<o
- b. Circunferential 360 Pbre,Through. Wall Crack of Depth a.
Figure 1-1. Representation of Postulated Cracks in Pipes for Fracture Mechanics Leak-Before-Break Analysis SIR-98-096, Rev. O l-5 Structuralintegrity Associates, Inc.
l Thru-Wall Length l Axisym.
% ~
+ "
- c. Depth w-Q 3
5 u_ 9 x l-- ~ w 2 2 22 k Critical Flaw Size Locus NDE 9F
- Leak Detection 93370r0 THRU 'NALL FLAW LENGTH Figure 1-2. ConceptualIllustration ofISI(UT)/ Leak Detection Approach to Protection Against Pipe Rupture SIR-98-096. Rev. O l-6 Structural Integrity Associates, Inc.
Piping Stress Analysis FRACTURE MECHANICS ANALYSIS Crack Leak Detection - . _ Detection System Critical crack size Before Pipe Leak rates Ruptures ISIintemals in-Semice inspection ' - 93371r0 Figure 1-3. Leak-Before-Break Approach Based on Fracture Mechanics Analysis with In-service Inspection and Leak Detection SIR-98-096, Rev. O l-7 j (
l , 2.0 CRITERIA FOR APPLICATION OF LEAK-BEFORE BREAK APPROACH i NUREG-1061. Vol. 3 [2] identifies several criteria to be considered in determining applicability of the leak-before-break approach to piping systems. Section 5.2 of NUREG-1061, Vol. 3 provides l extensive discussions of the criteria for performing leak-before-break analyses. The details of the l discussions will not be repeated here, but a summary of various requirements as applied to evaluation of the pressurizer surge line at Millstone 2 is provided below. 1 2.1 Criteria for Through-Wall Flaws Acceptance criteria for critical stresses and critical flaws are:
- 1. The through-wall flaw length which is required to produce an " acceptable leakage rate" during normal operating conditions must be smaller than the critical flaw length associated with normal operating loading plus safe-shutdown earthquake (SSE) by a factor of 2.
- 2. The stress required to make the "acceptab!e leakage rate" flaw critical is greater than that associated with normal operating loading plus SSE by a factor of at least [2,
- 3. The net section collapse criterion (NSCC) approach may be used to compute the critical flaw size provided that the limit moment is greater than the moment due to normal operating conditions plus SSE by a factor of 3. (This approach was not used in the pressurizer surge line evaluation.)
It has been found in previous evaluations conducted by Structural Integrity Associates (SI) that in general, the first criterion bounds the second. However, in this evaluation, both criteria will be considered for completeness. The elastic-plastic fracture mechanics (EPFM) approach is generally conservative relative to the NSCC approach when applied to stainless steel piping. Therefore, only EPF.\1 principles will be applied in this evaluation. The evaluation is presented in Section 5.0 of I this report. I SIR-98-096, Rev. 0 2-1
,7 StructuralIntegrity Associates, Inc.
1 1 l 1 2.2 Criteria for Part-Through-Wall Flaws NUREG-1061, Vol. 3 (2] requires demonstration that a part-through-wall flaw permitted by the acceptance criteria of ASME Section XI will not grow. This is demonstrated in Section 6.0 of this report, where the analysis of suberitical crack growth is discussed. ) i 2.3 Other Mechanisms NUREG-1061, Vol. 3 [2] limits applicability of the leak-before-break approach to those locations where degradation or failure by mechanisms such as water hammer, erosion / corrosion, fatigt'e, and intergranular stress corrosion cracking (IGSCC) is not a significant possibility. These mechanisms were considered for the pressurizer surge line at Millstone 2, as reported in Section 3 of this report. l 1 SIR-98-096 Rev. 0 2-2 StructuralIntegrity Associates, Inc.
I 1 l l 3.0 CONSIDERATION OF WATER HAMMER, CORROSION AND FATIGUE l l NUREG-1061, Vol. 3 [2] states that LBB should not be applied to high energy lines susceptible to l failure from the effects of water hammer, corrosion or fatigue. These potential failure mechanisms l are thus discussed below with regard to the pressurizer surge line at Millstone 2, and it is concluded l that the above failure mechanisms do not invalidate the use of LBB for this piping system. 3.1 Water Hammer i Water hammer in nuclear power plant systems is associated with sudden interruption of flow, l causing pressure waves to travel along a line. The majority of water hammers results in systems where ; 1
- 1) water is entrained in a normally steam-filled line,
- 2) water and steam flow in the same line, or 1
- 3) voiding occurs in a normally water-filled line.
The pressurizer surge line is a relatively short line that connects the reactor coolant system and i the pressurizer, both relatively large free volumes. There are no sparger or valves that can act as ' flow restrictions or that can suddenly stop the flow. The line is always filled with water with a pressure above saturation. Thus, no water hammer-causing mechanism is present. A comprehensive study performed in NUREG-0927 [3] indicated that the only water hammers reported in rextor primary systems were those due to relief valve discharge into relief valve discharge piping. This piping is connected to relief valves on the top of the pressurizer, remote from the pressurizer surge line. As such, water hammer loadings are not expected to have any effect on the pressurizer surge line. SIR-98-096, Rev. 0 3-1 . StructuralIntegrity Associates, Inc.
. . - . - _ _ ~. .. .- . . . . - . ~ . - - . - - .. _..._- - . . - - -
i 3.2 Corrosion l j-Three corrosion damage mechanisms are identified that can potentially lead to piping failure in l reactor coolant piping these are intergranular stress corrosion cracking (IGSCC), primary water stress corrosion cracking (PWSCC), and flow-assisted corrosion (FAC). IGSCC has been an issue in austenitic stainless steel piping in boiling water reactors [4] resulting from a combination of tensile stresses, susceptible material and oxygenated environment. The concentration of oxygen in a , l pressurized water reactor (PWR) is so low that IGSCC is not a problem for the primary loop piping, including the pressurizer surge line. i' The form of IGSSC which has ben obsmved in PWRs is PWSCC, affecting nickel-based alloys, where is has been seen to affect wrought Alloy 600 material. It is manifested by intergranular cracking and is aggravated by high levels of applied stress, eleveated temperature and prior strain. Field experience has demonstrated that PWSCC can affect CRD penetrations and steam generator tubing. There has been little evidence that there would be cracking in the nickel-based weld metals (Alloy 82 or 182) that may be present at the surge line to nozzle welds therefore, PWSCC is not l expected in the surge line or its welds. FAC is generally observed in carbon steels with very low alloying and in piping with high velocities and/or flow path discontinuities. Thus, FAC is not anticipated for the pressurizer surge line since I flow velocities are very low and the line is fabricated from austenitic stainless steel. 3.3 Fatigue The pressurizer surge lines in PWR plants are potentially susceptible to high fatigue usage as a
. result of thermal stratification. This phenomenon was identified in NRC Bulletin 88-11 [5] and was the subject of a Combustion Engineering Owners Group evaluation from 1988 to 1991 that
- 1) - defined the loading conditions related to thermal stratification, and
- 2) concluded generically that the fatigue usage factor in pressurizer surge lines was less than 1.0 for the design life of all CE-design plants [6).
SIR-98-096, Rev. 0 3-2 Structural Integrity Associates, Inc. l _ . _ _ . __ _ __
NRC Bulletin 88-1I was issued based on thermal stratification movements at the Trojan plant. At Trojan, unexpected thermal stratification movements had caused gap closure at pipe whip restraints, potentially causing plastic deformation of the piping system. Following the issuance of Bulletin 88-11, walkdowns were performed at Millstone 2 that did not show any indications of discemable distress or structural damage as a result of potential stratification [6]. It was also . reported [6] that the surge line welds at Millstone 2 had been inspected twice since stan of commercial operation (12n5) and that there were no indications. There were also no ISI indications or structural damage at any other CE-designed plant, except for Waterford. where a dead weight support spring can was not correctly set and a support rod was bent. Thus, because of thermal stratification loadings and other thermal transients in the surge line, there is a potential for thermal fatigue. However, it will be .shown in Section 6.0 that the number of cycles to cause significant growth of postulated cracks is very high relative to the number of cycles experienced during operation. Operating experience at Millstone 2 shows that the actual number of heatup/cooldown cycles is much less then those included in the plant design. Thus, potential fatigue cracking can be managed by surge line inspections. SIR-98-096 Rev. 0 3-3 StructuralIntegrity Associates, Inc.
4.0 PIPING MATERIALS AND STRESSES 4.1 Piping System Description The surge line begins at the nozzle connection at the hot leg and ends at the nozzle connection to the pressurizer. A schematic of the surge line and the spool pieces used in fabricating the line is shown in Figure 4-1.~ The line is fabricated from 12-inch Schedule 160 cast stainless steel piping with mi outside diameter of 12.75 inches and a wall thickness at all welds of 1.148 inches [7]. The carbon steel hot leg nozzle has a weld to a cast stainless steel safe end that has the same nominal outside diameter as the piping and has an inside diameter of 10.125 inches [7,8]. The hot leg nozzle has a minimum wall thickness of 1.25 inches that is nominally clad with 7/32 inch of stainless steel, except near the weld to the safe end where it thickens to 7/16 inch. Thus, at the thick end of the weld, the outside diameter is 13.5 inches and the total wall thickness is 1.6875 inches. The low alloy steel pressurizer nozzle is welded to a 13 inch OD cast stainless steel safe-end [9]. The nozzle has a minimum wall thickness of 1.125 inches and is clad with 7/16-inch stainless steel. Thus, the thick end of the nozzle weld has an outside diameter of 13.1875 inches and a total wall thickness of 1.53125 inches. At the weld to the piping, the wall thickness at the safe-end side of the weld is 1.273 inch:s; at the piping side, the thickness is 1.148 inches. The normal operating temperature for the hot leg end of the surge line is 604 F while that for the pressurizer is 653 F. The operating pressure is 2250 psia [10]. 4.2 Material Properties The surge line hot leg nozzle material is A-105, Grade II [8). This is followed by a safe end which is A-351, Grade CF8M. The surge line piping up to the pressurizer nozzle (including the pressurizer nozzle safe-end)is fabricated from A-35.1, Grade CF8M centrifugally cast stainless steel [7]. The pressurizer nozzle material is A-508, Class 2 [9]. All piping welds were fabricated SIR-98-096, Rev. 0 4-1 Structural Integrity Associates, Inc.
1 l i using stainless steel weld metal with shielded metal arc welding (SMAW) [20]. The safe-end-to-nozzle welds were fabricated using SMAW with Inconel 182 filler material. l l The material properties of interest for fracture mechanics and leakage calculations are the Modulus of Elasticity (E), the yield stress (S y), the ultimate stress (So), the Ramberg-Osgood parameters for describing the stress strain curve (a, and n), the fracture toughness (Jic) and power l
!aw coefficients for describing the material J Resistance curve (C and N).
Actual material properties from the certified material test reports (CMTRs) were used in the evaluation for all the components of the surge line. When material properties were not available, properties associated with the least favorable material and welding processes from industry generic material sources were used to provide a conservative assessment of critical flaw sizes and leakage rates. Appendix A provides the material properties for the various materials of the surge line and describes the methodology used in deriving these properties. 4.3 P! ping Loads The piping loads considered in the pressurizer surge LBB evaluation are due to pressure (P), dead weight (WT), surge line thermal expansion (SLTE) reactor coolant system thermal expansion (RCTE), thermal stratification (TS), and safe shutdown earthquake (SSE) consistent with the guidance provided in NUREG-1061, Vol. 3. Per the guidance provided in NUREG-1061, other secondary stresses were not included in the evaluation. l All piping system loads, except seismic, were derived specifically for this project. Loading conditions were evaluated as follows: Dead Weight (DW) included the weight of the piping, water, insulation and forces applied by dead weight hangers. SIR-98-096, Rev. 0 4-2 StructuralIntegrity Associates, Inc.
{ l i-I i
' * - Surge Line Thermal Expansion (SLTE) included the thermal expansion of the surge line with a temperature equal to the average of the pressurizer and the hot leg of 628.5*F [10].
RCS Thermal Expansion Movement (RCTE) included the loads due to the thermal expansion of the reactor coolant hot leg at 604 F [10].
^* Thermal Stratification (TS) included global stratification moments and forces for a 100*F linear p top-to-bottom temperature distribution in the piping.
The normal operating conditions used in the LBB evaluation included the algebraic sum of dead
- i. weight (DW), and themal expansion (SLTE + RCTE). Any additional moment loading due to l' thermal stratification was conservatively not included. For determination of Code-allowable flaw sizes, the maximum effects of thennal stratification were included.
Seismic stresses were derived from the original reactor coolant piping analysis for Millstone Unit 2 [11]. Pressure stresses were determined for a normal operating pressure of 2235 psig [10]. For the l fatigue crack evaluation, the loads described above were appropriately factored and additional local
- j. stresses due to thermal stratification, welding and bimetallic effects (at nozzles) were included.
More details on the piping loads determination and weld-unique moments and forces are included in
- Appendix B.
l l SIR-98-096, Rev. 0 4-3 StructuralIntegrity Associates, Inc. re -r-- - wr : ,- ,--,e . - - - -
~ . - - - - - - - , ,--
i l s (5cs 15-2) tsos-02) Ic 4e2 Sal l l (sos-03) ic4eto-21 @ o 7 (50s-04-2) Icae2041 l c-4e21 11 3 (50s 13) d l C4e191l
' 5 \
(505-05) (504-14) Ic-4eis-3l HM lc-4eis-el , Leg G (' '* ') * (sosa.i) <sosa) Ic4e20-2l lb-de18-1l PZR w v (sos-08) (s05-09)
. Ic4eisal - ,
Ic4ets.sl j G- @ O Node Numbers () Spool Piece Nuniber ( iO'A',), O u a r=ikisa m r
)
l Figure 4-1. Millstone Unit 2 Pressurizer Surge Line l 1 1 l l' SIR-98-096, Rev. 0 4-4 1 StructuralIntegrity Associates, Inc.
, - _ -I
l i i I 5.0 LEAK BEFORE-BREAK EVALUATION i The LLB approach involves the determination of critical flaw sizes and leakage through cracks in the piping. The critical flaw length for a through-wall flaw is that length for which, under a given set of applied stresses, the flaw would become marginally unstable. Similarly, the critical stress is that stress at which a given flaw size becomes marginally unstable. NUREG-1061, Vol. 3 [2] defines required margins of safety on both flaw length and applied stress. Both of these criteria will l be examined in this evaluation. Circumferential flaws are more restrictive than postulated axial flaws because the critical flaw sizes for axial flaw are very long since they are affected by only pressure stress and result in large crack opening areas due to out of plane displacements. For this reason, the evaluation presented herein will be based on assumed circumferential flaws. l The leak detection system at Millstone Unit 2 is capable of detecting i gpm leakage. Based on the safety factor of 10 on leakage required by NUREG-1061, Volume 3, the leakage flaw size is taken as that through-wall flaw size which produces 10 gpm leakage at all locations. 5.1 Evaluation of Critical Flaw Sizes Critical flaw sizes were determined for each weld location of the surge line. The critical flaw sizes were determined using the principals of elastic plastic fracture mechanics (EPFM) and J-Integral Tearing Modulus evaluation. The J-Integral calculation was based on the EPRI-developed engineering approach for EPFM [12,13]. The loadings for the evaluation included those due to normal operation + SSE. Stratification loadings were not included because they are minimal at normal operating conditions. Critical flaw sizes were determined for the unfactored load combination and for d times the loading combination. Details of the evaluation are include in Appendix C. SIR-98-096. Rev. 0 5-1 StructuralIntegrity Associates, Inc.
,_.- _ .. . . . . _m.. _ . _ _ . _ _ . _ .._-_ _ _ __ _ .. _..._._ ._. _. 1 i i 5.2 Leakage Rate Determination l l l l Calculations were performed to determine the flaw sizes that would result in 10 gpm leakage. i The calculations were performed using the PICEP computer program developed by EPRI [14,15]. The leakage was evaluated for the same loading conditions of normal operation (excluding stratification and SSE) that were used in evaluating the critical flaw sizes. This j condition was chosen since there is only a very small outflow (1.5 gpm) from the pressurizer to ! l the hot leg piping and stratification effects will be minimized. Any leakage through a crack (e.g. on the order of 10 gpm) would tend to fill the line with water from the hot leg. The leakage rates were determined uniquely for each weld location, considering a conservative combination of material properties that would minimize the leakage (maximizing the 10 gpm ! leakage flaw size). At the nozzle locations, the relatively greater thickness resulted in relatively large 10 gpm leakage flaws due to decreased stresses and increased flow path flow resistance. Details en the calculations are included in Appendix C.
]
5.3 LBB Evaluation Results and Discussions ! l Table 5-1 provides a comparison between the critical flaw size and the flaw size that produces 10 gpm leakage at all weld locations. Table 5-2 shows the evaluation at the nozzle regions. It ! can be seen that in all cases, the one-half critical flaw size based normal & SSE loads is greater than the 10 gpm leakage size flaw, thus satisfying the factor of two on flaw size required by NUREG-1061, Volume 3. Flaw size margin was also evaluated using 4 times normal plus SSE loads, and greater margin on flaw size was demonstrated for all locations. l l l i r l: I SIR-98-096, Rev. 0 5-2 \ StructuralIntegrity Associates, Inc.
l Table 5-1 l Comparison of Critical and 10 gpm Leakage Flaws for Piping Welds Cast Stainless Steel SMAW Weld One-Half Critical 10 gpm Leakage One-Half Critical 10 gpm Leakage Node Flaw Site. inches Flaw Size, inches Flaw Size, inches Flaw Size, inches i 3 7.05 7.00 7.89 6.92 1 4 5.83 5.56 6.84 5.58 5 6.20 5.90 7.21 5.91 L 7 5.19 4.9) 6.24 5.04 8 4.88 4.73 5.93 4.80 10 5.65 5.18 6.70 5.19 11 5.78 5.23 6.85 5.29 16 7.14 6.29 7.97 6.24 13 5.61 538 6.65 5.41 14 (L) 5.05 4.87 6.13 5.06 14 (H) 5.57 5.04 - -
- 1. For all cases, one-half of the critical flaw size based on normal + SSE loads is provided for direct comparison with the total length of the 10 gpm leakage size flaw. This criterion governed in all cases.
- 2. For Node 14. (L) indicates results for the low yield strength side for the cast austenitic stainless steel:(H) indicates the high strength side. For the SMAW weld, there is no difference. l
. SIR-98-096, Rev. 0 5-3 f StructuralIntegrityAssociates,Inc.
l Table 5-2 Comparison of Critical and 10 gpm Leakage Flaws for Nozzle Welds l One-Half 10 gpm Critical Flaw Leakage Flaw Node No. Material Size Size 2a SA-105 9.84 9.12 2a SMAW 10.08 9.16
. Hot Leg 2b SMAW 8.72 7.73 Nozzle 2b Cast Stainless 8.04 7.83 2c Cast Stainless 7.48 7.19 2c SMAW 8.28 7.09 15a- SA-508 6.89 5.43 15a SMAW 7.41 6.01 15b SMAW 7.00 5.72 Pressurizer 15b Cast Stainless 5.65 5.09 Nozzle 15c Cast Stainless 5.09 4.94 15c SMAW 6.53 5.31 15d SMAW 5.91 4.91 15d Cast Stainless 5.32 4.86
- 1. For all cases, one-half of the critical flaw size based on normal + SSE loads is provided for direct comparison with the total length of the 10 gpm leakage size flaw. This criteria governed in all cases.
SIR-98-096, Rev. 0 5-4 StructuralIntegrity Associates, Inc.
i 6.0 EVALUATION OF FATIGUE CRACK GROWTH OF SURFACE FLAWS ! 1 In accordance with the NRC criteria [2] set forth in Section 2 of this report, the growth of postulated I surface cracks by fatigue is evaluated to demonstrate that such growth is not significant for the plant life, when initial flaw sizes in excess of those meeting ASME Code Section XI, IWB-3514 l acceptance standards [16] are postulated. The potential for crack growth has been evaluated at each of the weld locations. In this section, the plant transients and stratification loadings are described l and the stresses used in the crack growth analysis are defined. The results of the analysis are presented in detail in Appendix D show the growth of potential flaws in the surge line. The I following summarizes the evaluation. i l 6.1 Plant Transient Evaluation
)
l For the fatigue crack growth evaluation, there are two major types of events that were considered. First, the plant transients defined in the original plant design were considered. These were taken from an updated design sprecification for the reactor coolant system (10] and l included 500 heatup/cooldown events and a range of additional transients associated with I normal, upset and emergency plant operation. Secondly, the surge line stratification transients evaluated as a result of NRC Bulletin 88-1i evaluations (5,6] were included. Detailed analysis was conducted to evaluate the throughwall thermal stress distribution for all thermal transients. For each thermal transient or stratification event, two or more thermal stress states were defined. For each state, the pressurizer pressure, surge line mean temperature, reactor coolant hot leg temperature and maximum potential surge line stratification AT were ~ defined, along with cubic polynomial stress coefficients to define the throughwall stresses. All throughwall thermal stresses were conservatively based on the nominal thickness of the surge line piping and did not take advantage of the decreased thickness at the pipe welds. SIR-98-096, Rev. 0 6-1 StructuralIntegrity Associates, Inc.
I r l l - 6.2 Thermal Stratification Stresses 1 i The stresses due to thermal stratification can be due to global thermal stratification moments and forces (as described in detail in Appendix B) and due to local circumferential temperature variations that cause local thermal stresses around each pipe section. The stratification stresses were determined in accordance with same methodology that was used in the CE Owner's Group report [6] that demonstrated that surge line fatigue usage was acceptable for a 40 year life.
' A conservative local thermal stratification stress definition was developed that conservatively bounded both insurge (cold water entering the bottom of the surge line filled with hot water) and outsurge (hot water entering the top of the surge line filled with relatively colder water). Only the horizontal sections of piping were exposed to local stratification stresses, whereas the global stresses affected every weld location. ;
)
6.3 ' Weld-Related Stresses A conservative weld-residual stress was applied at each of the welds. This was taken from NUREG-0313 [4] and assumed a maximum inside surface tensile stress of 30 ksi. Although this stress is not cyclic in nature,it was included because ofits adverse effect on crack growth (high R-ratio effect). At bimetallic welds, detailed finite element analysis was conducted to determine the throughwall stress distribution on each side of a bimetallic joint that is proportional to the mean temperature i rise above an assumed zero-stress room temperature reference state. This was applied at each of the nozzle-to-safe-end weld locations.
, 6.4 Loading Combinations For each event state, specific thermal conditions were evaluated as described in Section 6.1. The
" unit load" stresses were then scaled and combined to derive a total stress condition for each SIR-98-096, Rev. 0 6-2 g f f gg g
stress state. This resulted in a set of four stress coefficients describing a cubic polynomial stress distribution. The scaling was accomplished as follows: l
- Dead weight and weld residual stresses were held constant.
l Local and global stresses thermal stratification stresses were scaled proportionally to the { everit AT.
- Surge line thermal expansion stresses and nozzle bimetalhe stresses were scaled proportionally to the surge line mean temperature.
1 Hot leg thermal anchor movement stresses were scaled proportionally to the hot leg i temperature. Pressure stresses were based upon the pressurizer pressure. I
- For the moment-related stresses and the local thermal stratification stresses, the stress distribution around the circumference of the piping was considered in developing local stresses around the pipe circumfererence.
6.5 Fracture Mechanics Model The fracture mechanics analysis was conducted based upon a hypothetical 6-to-1 aspect ratio inside surface circumferencial flaw, consistent with the hypothetical flaw used in Appendix G of Section XI of the ASME Code [16]. This flaw aspect ratio isjustified since the stress distributions are maximized locally at the top and bottom of the pipe and resulted in much faster growth rates at these locations. The fracture mechanics model predicted the stress intensity factor at the deepest point of the flaw and was from the EPRI Ductile Fractre Mechanics Handbook (17]. The input to the fracture mechanics model was the cubic polynomial stress coefficient previously described. 6.6 Crack Growth Laws The crack growth laws from ASME Section XI. Appendices A and C [16] were used as the basis for the crack growth analysis for the low alloy steel (at the nozzles) and the cast stainless steel, respectively. For the stainless steel material. an additional factor of two was applied to account SIR-98 096, Rev. 0 6-3 StructuralIntegrity Associates, Inc.
l for the difference between the crack growth rates in air and PWR water environment (as I recommended in the Section XI Appendix C basis documentation (18]). , 6.7 Allowable Flaw Size Evaluation ;
)
l- To determine the significance of crack growth, Section XI flaw evaluation was conducted for the hypothetical flaw to determine the allowable depth while still maintaining Section XI-required safety margins. The analysis was conducted for the normal / upset loading condition including OBE as the seismic loading and with the maximum specified thermal stratification of 320 F. ; Using methods from Appendix C for stainless steel weldments and Appendix H for the ferritic materials at the nozzle welds, it was determined that flaw depth up to the maximum allowed by , Section XI(60% of pipe wall for the stainless steel and 75% of pipe wall for the nozzles) was acceptable at all locations. To assess the effect of much longer hypothetical flaws, critical flaw sizing was also conducted
. for 20:1 aspect ratio flaws. In this case, the critical flaw size at the most highly stressed piping location (Node 13) is 0.452 inches (a/t = 0.393); at the pressurizer nozzle-to-safe-end weld the critical size is 0.830 inches (a/t = 0.542). ;
6.8 ' Crack Growth for SSE during Normal Operation J To show that leakage size flaws will not grow for the condition of normal operation + SSE, linear elastic fracture mechanics analysis for a throughwall flaw was conducted. The model for a cylinder with a throughwall circumferential crack in tension was conservatively used [17). Flaw sizes corresponding to 10 gpm leakage were evaluated. For the controlling locations for both stainless steel and ferritic materials, the analysis showed that the crack growth resulting for the applicatiori 50 stress ranges for the condition of normal operating stresses
- SSE (two times the SSE amplitude) resulted in a crack growth of less than one percent of the 10 gpm leakage crack l size. This is not considered to be significant.
l l 1 1 l SIR-98-096, Rev. 0 6-4 . 4 StructuralIntegrity Acciates, Inc. l i
l l 6.9 Thermal Striping Crack Growth Thermal striping is rapid cyclic thermal stresses at the interface region between hot and cold fluid layers. Thermal striping is mainly observed on the inside of pipes on the side of the pipe in the vicinity where a sharp hot-to-cold boundary exists. The effect of thermal striping was evaluated in the CE Owners Group report [6], and it was concluded that thermal striping would not propagate an existing flaw because the thermal-stress-induced stress intensity factor drops below the threshold for crack growth. Thus,it can be concluded for this analysis that thermal striping will not contribute significantly to growth of cracks through the wall. 6.10 Crack Growth Results The initial flaw size was chosen as that size allowed by ASME Section XI. For the hypothetical 6:1 flaw size, this was determined to be 11.1 percent of the pipe wall thickness. Table 6-1 shows the results of the crack growth analysis for 125 and 250 heatup cooldown cycles. The 125 cycles is the design number of heatups/cooldowns in a 10-year interval. Figure 6-1 shows the crack growth time history for the five controlling locations (Nodes 3,4,8,10, and
- 13) and shows that all are approWately the same.
From this analysis, it concluded that inspection may be used to assure that fatigue cracking will not progress significantly through the walls of the piping. As shown in Table 6-1, the maximum crack depth at 125 cycles is 0.36 inches (a/t = 0.21) at the pressurizer nozzle weld or 0.22 inches (a/t = 0.19) in the piping weldments. Even after 250 cycles the flaws are significantly less than Section XI allowables of 0.689 inches (a/t = 0.60). As indicated in Figure 6-1, there are approximately 325 cycles prior to the crack reaching the Code allowable value. Thus, there is considerable number of operating cycles between the time at which a flaw would initiate and grow to a Code allowable flaw. SIR-98-096, Rev. 0 6-5 StructuralIntegrity Associates, Inc.
l l l There is much further real conservatism in the evaluation. Millstone 2 operating history showed [ that there were 49 heatup/cooldown cycles from startup to August 1993 (= 18 years) [19]. At this rate, cracks will not grow to a significant depth to the end of plant life. To assess the uncertainty associated with hypothetical flaw aspect ratio, crack growth analysis was also conducted for longer than expected 20:1 aspect ratio flaws. In this case, a flaw with the same initial depth reached a Section XI allowable flaw depth of 0A52 inches at about 60 heatup 1 and cooldown cycles for Node 13. Other locations, including those at the nozzle-to-safe-end j welds, will have slightly larger allowable flaws and generally exhibit slower crack growth. It is expected that the crack growth rate for this case would be bounding for all locations. Thus, it is concluded that fatigue crack growth can be adequately managed by inspection and that the predicted crack growth is not significant enough to invalidate the application of leak-before-break to the pressurizer surge line. l SIR-98-096, Rev. 0 6-6 Structural Integrity Associates, Inc.
-x
Table 6-1 Flaw Sizes After 125 and 250 Heatup/Cooldown Cycles l l I End of Evaluation Flaw Size, inches and a/t (1) l 125 Cycles 250 Cycles Location a a/t a a/t Node 2 i Hot Leg Nozzle 0.358 0.212 0.498 0.295 l Hot Leg Safe-End 0.187 0.163 0.297 0.259 ! Node 3 0.217 0.189 0.443 0.386 l Node 4 0.216 0.188 0.438 0.382 i Node 5 0.184 0.160 0.285 0.248 j Node 7 0.181 0.158 0.274 0.239 l Node 8 0.214 0.186 0.437 0.381 Node 10 0.215 0.187 0.433 0.377 Node 11 0.204 0.178 0.371 0.323 Node 16 l 0.206 0.179 0.385 0.335 Node 13 0.215 0.187 0.432 0.376 Node 14 0.190 0.166 0.314 0.274 I Node 15 PZR Safe End 0.187 0.163 0.321 0.280 ; PZR Nozzle 0.263 0.172 0.504 0.329 Note: (1) All flaw sizes baed on 6:1 aspect ratio flaws with an initial flaw depth / thickness ratio of 0.111. I 1 SIR-98-096, Rev. 0 6-7 StructuralIntegrity Associates, Inc.
l-I l 1 l 0.9 Je t 0.8 y 0.7 [ 0.6 [ jas / l j f l % 0.4 j U 0.3 - - , , ~
/
- \
' 0.2
- s~ l I
0.1 l 1 0 j 0 50 100 150 200 250 m a g HeatWCoddomCy:les l Figure 6-1. Predicted Deepest-Point Crack Growth Time History at Nodes 3,4,8,10 and 13 i l l l l' l
. SIR-98-096, Rev. 0 6-8 f StructuralIntegrityAssociates,Inc.
7.0
SUMMARY
AND CONCLUSIONS Leak-before-break (LBB) evaluations are performed for the pressurizer surge line at Millstone 2 in accordance with the requirements of NUREG-106 . The analysis was performed using actual material propenies when available or otherwise using conservative generic material properties for the base metals and weldments, l.ocation specific stresses consisting of pmssure, deadweight, thermal and seismic loads were used in the analysis. In the evaluations, circumferential flaws are considered since they are more limiting than axial flaws. Critical flaw sizes and leakage flaw sizes were calculated on a location specific basis for the pressurizer surge line for several material conditions. Both base metal and weld metals properties were considered in the evaluation. The effect of thermal embrittlement was considered for the cast stainless steel base metal and the SMAW weld metal. The minimum of one half the critical flaw size with a factor of one on the stresses or the full critical flaw size with a factor of J2 on the stresses were determined and were compared to the crack sizes that would produce 10 gpm leakage. Fatigue crack growth analysis was also performed to determine the extent of growth of any pre-existing flaws. Based on these evaluations, the following conclusions can be made, e The detectable leakage at Millstone Unit 2 is 1.0 gpm. Based on the required margin of 10 on leakage in NUREG-1061, Volume 3, the critical flaw sizes were compared to the crack size i which produces 10 gpm leakage. The critical flaw sizes, based on the NUREG-1061, Vol. 3 margins, were found to be greater than the 10 gpm leakage flaw sizes at all locations for all material conditions, thus assuring leak-before-break. Fatigue crack growth of hypothetical 6:1 surface flaw with initial depth of 11.1% of pipe wall does not reach the ASME Section XI allowable flaw size for the expected number of heatup and cooldown cycles over the plant lifetime. Fatigue crack growth was also performed for a hypothetical 20:1 surface flaw and it was demonstrated that there is sufficient margin to the allowable flaw size. Therefom fatigue does not invalidate the application ofleak-before-break evaluation to the pressurizer surge line. S1R-98-096, Rev. 0 7-1 StructuralIntegrity Associates, Inc.
H
\
- The effects of other degradation mechanisms which could invalidate the LBB evaluations were -
considered in the evaluation. It was determined that there is no potential for water hammer, intergranular stress corrosion cracking (IGSCC) and flaw-assisted corrosion for the pressurizer surge line. b
- SIR-98 096, Rev. 0 - 7-2' StructuralIntegrity Associates, Inc.
8,0 REFERENCES
- 1. Stello, Jr., V., " Final Broad Scope Rule to Modify General Design Criterion 4 of Appendix A,10 CFR Part 50," NRC SECY-87-213, Rulemaking Issue (Affirmation), August 21, 1987.
- 2. NUREG-1061, Volumes 1-5,"Repon of the U. S. Nuclear Regulatory Commission Piping Review Committee," prepared by the Piping Review Committee, NRC, April 1985.
- 3. NUREG-0927, " Evaluation of Water Hammer Occurrence in Nuclear Power Plants,"
Revision 1.
- 4. W. S. Hazelton and W. H. Koo, ' Technical Repon on Material Selection and Processing Guidelines for BWR Coolant Pressure Boundary Piping," NUREG-0313, Rev. 2, USNRC, January 1988.
- 5. NRC Bulletin 88-11," Pressurizer Surge Line Themaal Stratification," U.S. Nuclear Regulatory Commission, December 20,1988.
- 6. CEN 387-NP," Pressurizer Surge Line Flow Stratification Evaluation," ABB Combustion Engineering, Rev.1-NP, December 1991.
- 7. Drawing E-234-004, " Piping Details and Assembly," Rev. 3.
- 8. Drawing E-234-005, " Nozzle Details", Rev.1.
- 9. Drawing E-233-746, " Nozzle Details for 96" 1.D. Pressurizer - Millstone Unit No. 2,"
Rev. 2.
- 10. Specification 18767-31-5,"Eagincering Specification for a Reactor Coolant Pipe and i Fittings for Northeast Utilities Service Company, Millstone Point Station, Unit 2," Rev.
16, November 1997. I 1. NU Document M2-EV-0165, " Technical Evaluation for Millstone 2 Plant Specific Surge Line OBE Loads," 8/25/95 (Attachment to ERC 25203-ER-98-0115 Rev. 2).
- 12. Kumar, V., et al.,"An Engineering Approach for Elastic-Plastic Fracture Analysis," EPRI NP-1931, Electric Power Research Institute, Palo Alto, CA, July 1981.
- 13. Kumar, V., et al., " Advances in Elastic-Plastic Fracture Analysis," EPRI NP-3607, Electric Power Research Institute, Palo Alto, CA August 1984.
- 14. EPRI Report NP-3596-SR, "PICEP: Pipe Crack Evaluation Computer Program." Rev.1, July 1987.
S1R-98-096, Rev. 0 8-1 StructuralIntegrity Associates, Inc.
l I l' 15. EPRI Report NP-3395, " Calculation of I.eak Rates Through Crack in Pipes and Tubes," December 1983,
- 16. ' ASME Boiler and Pressure Vessel Code, Section XI,1989 Edition. l l- 17. EPRI Report NP-6301-D, " Ductile Fracture Handbook," June 1989.
l 18. ASME Section XI Task Group for Piping Flaw Evaluation, ASME Code, " Evaluation of
- Flaws in Austenitic Steel Piping," Journal of Pressure Vessel Technology, Vol.108,
! l August 1986, pp. 352-366.
- 19. N-MECH-ER-016, " Implementation of the Millstone Station Unit 2 Fatigue Monitoring Program," Attachment 2, December 1993.
20.' Letter No. N-PEN-98-043 from C. Gimbrone (ABB) to Roy Terry (Northeast Utilities),
" Weld Process and Filler Material Information for Millstone Point 2 Surge Line," dated July L 31,1998 (with attachments). (Attachment A to Engineering Record Correspondence, l 25203-ER-98-0115, Rev. 2). .
l .- l l l l l l'. i u Y SIR-98-096, Rev. 0 8-2 StructuralIntegrity Associates, Inc. 1
l APPENDIX A Materials Evaluation 5 l SIR-98-096, Rev. O A-0 StructuralIntegrity Associates, Inc.
l Annendix A ) Materials Evaluation A.1 Nozzle Materials The pressurizer surge line nozzle is adjacent to the pressurizer which operates at 653 F The nozzle at the other end is on the hot leg which operates at 604 F. For use in the fracture mechanics analysis, all properties are evaluated at 605 F and 650 F. The nozzle material for the surge line piping at the hot leg is carbon steel A-105 Grade II [A-1]. I The nozzle rnaterial at the pressurizer end is low alloy carbon steel SA-508, Class 2 [A-2]. Generic lower bound properties for ferritic steel piping at 550 F which was used as the basis for ASME Section XI flaw acceptance criteria are provided in Reference A-3. The basis for the lower bound values used in ASME Section XIis provided in Reference A-4. For the stress- l strain parameters, the generic properties for carbon steel piping provided in Reference A-3 are
]
used. These properties are shown in Table A-1. The values of yield and ultimate stress shown in Table A-1 are significantly lower than the properties in the CMTRs for these components [A-5]. . Three different sets of J-R resistance curve parameters are provided in Reference A-3. The basis for these curves is provided in Reference A-4. The first of these curves (Generic CS-1) is applicable to ferritic welds. The second (Generic CS-2) is applicable to base metal for which there are no available heat specific data to define Jrc. The third (Generic CS-3) is applicable to base metal and welds where heat specific data are available tojustify J ic greater than 1050 in-lb/in 2, To determine which of these curves is applicable, the CMTRs of the nozzle materials [A-5) were reviewed. It was noted that for the A-105, Grade Il material, the Charpy values at + 10 F ranged from 24 to 43 ft-lbs. These values are consistent with those obtained by Vassilaros et al. in NUREG/CR-3740 [A-6] for A-106 Grade B at + 10 F and shown in Figure A-1. There is also good consistency between the chemical composition of the A-105 nozzle material at Millstone 2 and for the A-106 Grade B piping used in Reference A-6. Hence, Figure A-1 can be used to estimate the Charpy values of the A-105 nozzle material at 650 F. Using Figure A-1. it can be concluded that at 650 F, the nozzle material will be at the upper shelf region and more
. SIR-98-096, Rev. O A-1 StructuralIntegrity Associates, Inc.
. . _ . . . . _ _ . _ . _ . - _ _ _ _ _ _ _ __ .m importantly, the Charpy value at 650*F will be at least 110 ft-lb. Using the relationship provided in Reference A-4, this translates into J mm i 2
or Jic of at least i100 in-kip /in . This value of Jic - mures the use of Generic CS-3 parameters from Reference A-3 to define the J-R resistance curve. The J-R resistance curve parameters are shown in Table A-1. ! Similarly, for the SA-508, Class 2 material, the Charpy values range from 53 to 94 ft-lbs at + 10*F [A-5]. At 650"F, it is expected that the fracture toughness will be at least equal to that of the SA-105 Grade Il materirj making the Generic CS-3 J-R curve of Reference A-3 also conservatively applicable to this material. For use at 650 F, the stress-strain parameters were adjusted by the ratio of change in these I 1 parameters from 550 F to 650 F. The values for E, S yand So shown in Table 4-1 are adjusted by the ratio of their values in ASME Code Section III Appendices [A-7]. The basis for the adjustment of the Ramberg-Osgood parameters is provided in Section A.4. The revised properties are provided in Tables A-1 A and A-lB. The plot of the stress-strain curves at 550 F and 650 F are shown in Figures A-2 and A-3. As can be seen from these curves, thee is no significant difference between the curves at these two temperatures. The area under the stress-strain curve is related to the toughness of the material. Since the stress-strain curves are ; essentially the same at these two temperatures, it can be expected that there is no significant change in :oughness from 550 F to 650 F. Hence the toughness properties at 550 F provided in Table A-1 are used for the evaluation at 650 F. The value of the maximum valid value of J for deriving a J-R curve was obtained from Reference A-4. As explained in Section A.2.4.3 of NUREG-1061, Vol. 3 [A-8], J , is the limit of valid data above which an extrapolated J/T curve must be used in performing elastic-plastic fracture mechanics analysis. A.2 Safe End and Piping Material The safe end and piping material for surge line is A-351, Grade CF-8M centrifugal cast stainless steel [A-9. A-10]. At reactor operating temperatures (550 F- 662 F), cast stainless steels used in reactors have been shown to be susceptible to thermal aging (embrittlement) [A-11]. Thermal aging of cast stainless steels at these temperatures increases hardness and tensile strength and
' decrease ductility, impact strength and fracture toughness of the material. Hence the material SIR-98-096, Rev. O A-2 StructuralIntegrity Associates, Inc.
4
)
properties are conservatively determined for the aged condition using the correlations and data provided in References A-11 and A-12.
)
i The only information required in these correlations is the chemical composition from the CMTR. A correlation for the extent of thermal embrittlement at " saturation" (the minimum impact energy that would be achieved for the material after long term aging at a given operating temperature) is 1 given in terms of the chemical composition. The extent of thermal embrittlement as a function of time and temperature of reactor service is then estimated from the extent of embrittlement at l saturation and from the correlations describing the kinetics of embrittlement, which are also ' given in terms of the chemical composition. In this evaluation, the fracture toughness associated with the minimum impact energy assuming aging at normal LWR operating temperatures (524 - 662*F) will be used. Using the methodology of Reference A-11, the chromium equivalent (Cry ) and nickel equivalent (Nieq) are determined from the chemical composition, based on Hull's equivalent factors [A-13]: i Creq = (Cr) + 1.21 (Mo) + 0.48 (Si) - 4.99 (1) Nieq = (Ni) + 0.11 (Mn)- 0.0086 (Mn)2 + 18.4 (N) + 24.5 (C) + 2.77 (2) where the chemical composition is in wt?c. Per Reference A-11, the value of N is assumed to be 0.04 if it is not available on the CMTR. The ferrite content (6 c) is then estimated from the relationship: S e = 100.3 (Cr ey /Nieq/ - 170.72 (Creq /Nieg) + 74.22 (3) For CF8M cast stainless steel, the saturation (minimum) impact energy (Cvm) considering thermal embrittlement can be determined by two methods: SIR-98-096, Rev. O A-3 Y
'l l
In the first method, the material parameter 4 is calculated from which Cv , is determined as l follows: l I l.
$ = 6,(Ni + Si + Mn)2(C + 0.4N)/5. (4) - The saturation value of RT impact energy, Cv,,,for steels with < 10% Ni is given by LogioCv.,= 1.10 + 2.12exp(-0.041 & ). (5)
And for steels with >10% Ni by Logio Cv ,= 1.10 + 2.64exp(-0.064 @ ). , (6) 4 In the second method, Cys, is estimated directly from the chemical compositions of the steel and i is given by: Logio Cv,, = 7.28 - 0.011 ( S ,) - 0. I 85 (Cr) - 0.369 (Mo) - 0.451 (Si) (7) l
- 0.007 (Ni) -_4.71 (C + 0.4N)
The saturation impact energy is determined using both methods given in Equations 5,6 and 7 and the lower value is used for estimating the fracture toughness. The material resistance J-R curve can be estimated from Cv,, using a power law relationship: J, = C [Cv,,]"'[Aa]" (8) 2 where: Ja is the deformation J-Integral (kJ/m ) per ASTM Specification E813-85 Aa is the crack extension (mm) l C is a constant i-m, n are power law exponents l SIR-98-096, Rev. O A-4
l l The saturation fracture toughness J-R curve at room temperature for centrifugally-cast CF8M stainless' steel is given by [A-11]: ' Ja= 20(Cv,,,) 67[Aa]" (9)
.j 2
.In English units, the J-R curve (units of J in in-kips /in and Aa in inches) is given by:
Ja = 114[25.4]"[Cv,,,] 67[Aa]" (10) l 2 where Cv.,,is in Joules /cm and the value of n at room temperature is given by: ! n = 0.25 + 0.0771ogio Cv,,, (11) Corresponding equations for the J-R curve at temperatures between 290 C and 350 C (550*F and 662 F) are given by: Ja = 57[Cv,,,] A' [Aa]" (SI units) (12) Ja = 325[25.4]"[Cv,,,] d'[Aa]" (English units) (13) i l i n = 0.23 + 0.0571ogio Cv,,, (14) Equation 13 for the J-R curve can be expressed in simple terms as: l Ja = C [Aa]" (15) SIR-98-096; Rev. O A-5 Structural Integrity Associates, Inc.
The calculation of all the above parameters, including C and'n for the surge line piping at Millstone 2 are shown in Table A-2. The calculation was performed using information contained in the CMTRs for the various spools of the surge line [A-14]. A layout of the cast stainless steel spool pieces of the surge line is shown in Figure A-4. The above correlations (Equations I through 15) account for degradation of toughness due to thermal aging, but do not explicitly consider the initial fracture propenies of the original unaged material. Fracture toughness data in Reference A-ll indicate that the J-R curva for some heats of unaged cast stainless steel may be lower than those for wrought stainless steel. To take into account the possibility of a relatively low initial unaged toughness, the methodology outlined in Reference A-11 requires that the saturation J R curves be compared to the lower bound J-R curve for the unaged cast stainless steel. The lower of the two curves is then used. For static cast stainless steel, the lowet bound unaged !-R curve is given by: Jd= 400 [Aa]** (SI units) (16) Ja = 8330[Aa]** (English units) (17) The Ramberg-Osgood parameters for the stress-strain curve for the aged condition is provided in Reference A-12. The value for the parameter is at 290"C was provided as 6.6. The parameter a was determined for a strain range <5% as 1.254 assuming the reference stress to be the yield strength. The values for the yield and ultima:e strength were obtained from the CMTRs for each spool piece [A-14]. The value of the modulus of elasticity, E, for the cast stainless steel was obtained from Reference A-15. The remaining properties at 650 F were adjusted similar to the scheme described for the nozzle material and are shown in Table A-3A. Typical stress-strain curves at 550 F and 650 F are shown in Figure A-5. The J-R curve for the aged cast stainless steels were determined for application in the range of 554 F to 662 F [A-11] and hence the J-R parameters shown in Table A-2 are also used at 650 F. SIR-98-096, Rev. O A-6 StructuralIntegrity Associates, Inc.
-.. - . .- - - - - - - - = . - - .= The maximum valid value for J for use in J/I' curve formulation was obtained from data presented Reference A-11. A.3 Weld Materials All stainless steel welds are fabricated using shielded metal arc welding (SMAW) process [A-16]. The conservative stress-strain properties for the SMAW weldments at 550 F in Reference A-3 which formed the basis for the flaw acceptance criteria in ASME Section XI [A-17] were used for the evaluation. However, for the J.R curve properties, the values provided in Reference A-3 for the SMAW welds were compared with the lower bound curve provided in NUREG-6428 [A-18] for thermally aged welds. It was found that the lower bound curve in NUREG-6428 is more conservative and was therefore used in this evaluation. The properties used for the SMAW welds are shown in Table A-4. The bimetallic nozzle to safe end field weld was fabricated using the SMAW process using Inconel 182 weld metal. The lower bound SMAW weld properties for stainless steel were also assumed for this weld. The adjusted stress - strain properties based on the ratio of ASME Code values at 650 F are shown in Table A-4A. The stress-strain curve is rhown in Figure A-6. Once again, it can be seen that the change in the stress-strain curves at the two temperatures is very small and therefore the J-R curve properties at 550 F can be assumed at 650 F. The value for the maximum value of J for the J/I' curve determination was obtained from data presented in Reference A-18. A.4 Determination of Ramberg Osgood Parameters at 650 F The Ramberg-Osgood stress-strain parameters (a and n) are necessary for elastic-plastic fracture mechanics analysis. These parameters may be a function of temperature. This section provides the methodology for making adjustment for the Ramberg-Osgood stress-strain parameters at a different temperature when the parameters for another temperature are known. In this case, the Ramberg-Osgood parameters are derived for at 650 F for given values at 550*F for the materials of the surge line at Millstone 2. SIR-98-096, Rev. 0 A-7 gg,yggy,,y jaggy,;gypggggygggg, jgg,
. _ _ . . _ .__ . . . _ . . . _ . . m.__.~._ . . . . . _. - . _ . _ . _ . _ . . _ - _ _ - _ _ . _ _ _
l 4 The Ramberg-Osgood model is in the form:
- -- = 1 + a 1 (18)-
[ Eo 00 00 l Where o and e are the true stress and true strain, a oand to are the reference stress and reference strain (in general yield stress and yield strain) and a and n are the so called Ramberg-Osgood (R-0) parameters. When the stress-strain curve at the temperature of interest is available, the R-O parameters can be abtained by curve fitting over the strain range ofinterest. In the absence of the stress-strain curve of the material, a methodology for determining the R-O parameters based on ASME Code-specified mechanical properties is provided in Reference A-19. The suggested method is described by the following equations: g y 0.002 (39)_ e, i
- t ;
1 In(1 + e,), S,(1 + e,) in - - n= -a In(1 +ey ), S,(1 + ey )- ; (20) ' S,(1 + e,) In
.S,(1 + ey ),
where So and Sy represent ultimate stress and yield stress respectively. They can be obtained from the ASME Code [A-7] for a wide range of temperatures. The engineering yield strain (ey ) is determined as: S e' =El (21) l SIR-98-096, Rev. 0 A-8 StructuralIntegrity Associates, Inc.
I where E (modulus of elasticity) can also be obtained from the ASME Code. The ultimate engineering strain (eo) is not specified at all temperatures in the ASME Code, hence the room temperature elongation value specified in the ASME Code, Section II [A-7] is assumed for all temperatures. The methodology in any case is not sensitive to the choice of eo [A-19] when determining a and n by using equation (19) and (20). It is obvious that a is a function of e y, n is a function of a, e ,oe y, So, and S y, and both are the function of temperature. Therefore, an adjustment scheme can be used as follows where the material properties at 650 F are adjusted based on the ratio of predicted properties from Equations (19) and (20) using Code minimum properties: Equation (19) 3,,, coa,. m,"* (a)63e, = (3)s=.sm r* Equation (19)33r,,(22) coa . propary 1 Equation (20)65er. code a== P'"9'"Y \ (n),3e, = (n),,,35 ,1 (23) Equation (20)33y,, coa, ;
..,,oporty Hence, Equations (19), (20), (21), (22) and (23) can be used to obtain R-O parameters at 650 F from the given values at 550 F.
The inputs into the evaluation consist of the R-O parameters for the various materials provided in Tables A-1 through A-4 and ASME Code properties at 550 F and 650 F. The input and results r i 1-of the analysis which determines the R-O parameters at 650 F are provided in Tables A-5 I through A-8. I SIR-98-096, Rev. 0 A-9 f StructuralintegrityAssociates,Inc.
I Table A-1 Generic Material Properties for Carbon Steel Nozzles at 550*F Parameter Value Temp ( F) 550 3 E (ksi) 26 x 10 Sy = co(ksi) 27.1 S. (ksi) 60.0 Ramberg-Osgood Parameter a 2.51 Ramberg-Osgood Parameter n 4.20 2 Jic(in-k/in ) 1.05 2 J-R Curve Parameter Ci (in-k/in ) 5.40 J-R Curve Parameter N 0.344 J., (in-k/in 2) 3.5 [A-4) SIR-98-096, Rev. O A-10 StructuralIntegrity Associates, Inc.
? l 1 I Table A-1 A Generic Material Properties for the A-105 Carbon Steel Surge Nozzles at 650 F l Parameter Value i i Temp ( F) 650 3 E (ksi) 25.12 x 10 I Sy = co (ksi) 25.40
]
1 S (ksi) 60.0 Ramberg-Osgood Parameter a 2.588 1 Ramberg-Osgood Parameter n 3.981 2 Ji c(in-k/in ) 1.05 l 2 J-R Curve Parameter Ci(in-k/in ) 5.40 J-R Curve Parameter N 0.344 2 Jmo (in-k/in ) 3.5 I i l SIR-98-096, Rev. O A-11 1 StructuralIntegrity Associates, Inc.
l Table A-1B Generic Material Properties for the SA-508 Class 2 Low-Alloy Steel Pressurizer Surge Nozzle at 650 F Parameter Value Temp ( F) 650 E (ksi) 25.17 x 10' Sy = co (ksi) 26.70 S. (ksi) 60.0 Ramberg-Osgood Parameter cx 2.466 Ramberg-Osgood Parameter n 4.128 2 Ji c(in-k/in ) 1.05 2 J-R Curve Parameter Ci (in-k/in ) 5.40 J-R Curve Parameter N 0.344 2 Jma (in-k/in ) 3.5 1 i SIR-98-096, Rev. O A- 12 StructuralIntegrity Associates, Inc.
U2 h 6
?
O J44 Calcadation Lower Bound Fracture Taughness of Casmg for Cast Surge Lke Components wist Thermal Emtwe8ement Propertes evatusted at 2905320C O
)O C-4618-1 C4518-1 C-4619-1 C-4619-1 C-461S2 C-4619-2 C-4619-3 C-4619-3 C-46144 C-4619-4 C-46145 C-46145 C-46194 C-46194 Malenal Matenad Matenal Matenal Matenal Matenas Malenal Malenal Material Malenal Matenal Mosenal Mosenal Matenal Q
O
.< Test #1 Test 82 Tests 1 Test s2 Test 81 Test 82 Test 81 Test 82 Test s1 . Test 82 Test 81 Test s2 Test et Test s2 8 O Upstream hoes 2 2 3 3 8 8 5 5 14
,02 h 14 le 16 11 11 Downstream Node 2 2 4 4 10 10 7 7 15 15 13 13 16 16 _.
YS. ps: 45800 45800 45800 45800 51500 51500 40700 40700 51500 51500 47000 47000 45800 45800 S. D h UTS. psa 90700 90700 90700 90700 86700 66700 82400 82400 86700 - 86700 89200 89200 90700 90700 _ C. Eing. % 44 44 44 44 48 48 46 46 48 48 40 40 44 44 O
%% m o
U2 m Cr 20 2 20 8 20.2 20 8 20 6 19 7 59 3 19 35 20 6 19 7 19.1 19 8 20 2 20 8 8p t S6 0 95 1.19 0 95 1 19 1 11 1.29 1 1 17 1.11 1.29 0 94 1 06 0 95 1.19 OO Mo 24 2.3 24 23 2 42 23 2 15 2.37 2 42 2.3 26 26 24 23 g I Nb M 945 0 91 0 9 45 0 91 0 95 0 92 0 92 0 91 0 95 0 92 0 95 0 93 0 9 45 0 91 0 g- Q [go C 0 05 0 05 0 05 0 05 0 07 0 06 0 07 0 07 0 07 0 06 0 06 0 06 0 05 0 05 Mn 0 86 0 85 0 86 0 85 0 88 0 87 08 08 0 88 0 87 0 88 0 84 0 86 0 85 C ' N (est) 0 04 0 04 0 04 0 04 0 04 0 04 0 04 0 04 0 04 0 04 0 04 0 04 0 04 0 04 OU o c. 3
=> vs(pse) R T 45800 45800 45800 45800 51500 51500 40700 40700 51500 51500 47000 47000 45800 45800 !. h I
.'. YS(psi) 650F 28259 28259 28259 28259 31776 3l776 25112 25112 31776 31776 28999 28999 28259 28259 @-
W UTS(psi).R T. 90700 90700 90700 90700 86700 86700 82400 82400 86700 86700 89200 89200 90700 90700 3. ck 'E> , UTs(ps.).650F 86800 86800 86800 86800 82972 82972 78857 78857 82972 82972 85364 85364 86800 86800 c5 o U -t
. d Flow Skets (pse)
O . g c -[ RT 68250 68250 68250 68250 69100 69100 61550 61550 69100 69100 6P100 68100 68250 68250 g00 7 650-F 57529 57529 57529 57529 57374 57374 51984 51984 57374 57374 57182 57182 57529 57529 E. 3 o m? Creq 18 6 19 2 18 6 19 2 19 1 18 1 17.4 17 8 19 1 18 1 17 7 18 5 18 6 19 2 3 7 '0 neq 14 3 13 9 14 3 13 9 14 8 14 3 14 5 14 4 14 8 14 3 14 6 14 4 14 3 13 9 i g Fende 21 9 36 7 29 3 48 0 21 9 29 3 20 7 19 2 13 7 16 4 20 7 19 2 14 9 20 5 21 9 29 3 o PHs 36 7 48 0 47.0 37 6 28 6 34 5 47 0 37 6 29 0 39 1 36 7 46 0 g3 E c:: c.us um-2: e.- 3 U- m g Po rno-m> 40 8 22 4 40 8 22 4 23 8 37.3 69 9 44 6 23 8 37.3 59 5 33 9 40 8 22 4 o U p
~ gq c..e-a m.= p a pesi 37 3 24 9 37 3 24 9 25 6 35 8 572 41 2 25 6 35 8 55 5 33 6 37.3 24 9
- EL Mwumum Cwsat 37 3 22 4 37 3 22 4 23 8 35 8 57 2 41 2 23 8 35 8 55 5 33 6 37.3 n
EC3 22 4 @
- a. -
N C (AR)in livin' 4025 9 3135 3 4025 9 3135 3 3230 9 3950.1 4965 8 4231 6 3230 9 3050
- 4896 3 3827.1 4025 9 3135 3 b n (J R) 0320 0 307 0320 0 307 0 308 0319 0 330 0 322 0 308 0 319 0 329 0317 0320 0 307 m -
O 6 O U Ei
.m .
N 9 ' i
CA b b 1R Calculatuan Lower Bound Fracture Toughness of Casing for Cast Surge Une Components wei Thermat Emtradamers
?
C Properties evaluated at 290-320C C-3223-C-4620-1 C 4620-1 C-4620'2 C-4620-2 C-4620-3 C-4620-3 C-4620-4 C-46244 C-4621-1 C-4621-1 C-3223-1R 1R
-$ Material Matenal Matenal Malenal Malenal Matenal Metettal Matertaf Malena! Malenal Motenal Material N Test s1 Test #2 Test $1 Test 82 Test #1 Test 82 Test #1 Test 82 Test 81 Test 82 Test #1 Test 82 Upstream Node 13 13 4 4 7 7 2 2 10 to 15 15 ~
h Downstream Node 14 14 5 5 0 8 3 3 11 11 15 15 6 VS. ps4 42100 42100 43800 43800 45200 45200 48200 48200 39100 39100 39400 35500 ga UTS, psi 87300 87300 86000 86000 89600 89800 91200 91200 79500 79500 77200 73300 - Eing % 50 50 44 44 44 44 46 46 46 46 47 40 h E' g* RA.% og Cr 20.2 20 19.96 20.7 20.1 20.7 20 2 20 9 19 4 20,1 19.2 20 1 gn Q Sa 09 06 1 02 0 68 1.14 1.08 1 0 66 0 92 0 74 08 0.56 gy Mo 2 65 25 2 68 2 55 2 56 2 55 2 55 2 45 2 45 2.45 2.42 2 35 0 o
% 0 0 0 0 0 0 0 0 0 0 0 0 g Ne 9 92 10.03 9 9 87 95 91 92 9 ". 9.7 10.5 10 3 .g f as C 0 04 0 04 0 05 0 05 0 06 0.06 0 05 0 05 0 04 0.04 0 02 0 04 gWp q Mn 08 0.76 0.75 0 85 0 83 0 94 0 85 0 92 0 86 0 91 0.8 0 87 tyg og N (est) 0 04 0 04 0 04 0 04 0 04 0 04 0 04 0 04 0 04 0 04 0 04 0 04 Qp O 2 A- >
p vs(pso.m T. 42100 42100 43800 43800 45200 45200 48200 48200 39100 39100 39400 35500 Ei, @ t'a i VS(psi).650F 25976 25976 27025 27025 27868 27888 29739. 29739 24125 24125 24310 21904 SM^ U,g 4 UTS(pso.R T 87300 87300 86000 86000 89600 89600 91200 9I200 79500 79500 77200 73300 co UTS(ps4.650F 83546 83546 82302 82302 85747 85747 87278 87278 76082 76082 73880 70148 53S Flow St ess (psi)
##E O C G RT 64700 64700 64900 64900 67400 67400 69700 69700 59300 59300 58300 54400 h D~ O 650T 54761 54761 54663 54663 56818 56818 58509 58509 50103 50103 49095 46026 E.E mM 9o ,
03 Creq Nq 18 8 13 6 18 3 13 8 18 7 14 8 19 1 13 8 to 8 14 9 19.3 14 6 18 8 13 9 19.2 14 0 17 8 13 9 18 4 14 3 17.5 14 8 18 2 14 9 [g
;2 O
[ Fende 30 6 24 7 18 4 30 0 18 0 24.1 26 4 28 4 20 4 21 0 13 9 15 8 ED n PHI 39 3 30 8 33 8 44 0 38 4 48 7 41.9 438 28 0 30 2 14 7 24 0 3 e i a F Q Cvsas (sem2) thom Polymman 31.3 61.2 35 9 31.5 30 0 21 5 30 6 33 4 65 9 57.7 119 7 86 0 b$ - 5 O g cv mason 2nn mmen 33 4 50 0 25.3 28.1 34 6 24 4 30 3 28 5 59 1 51 7 135 5 46 4 4lt 4 Minimum Cvsat 31.3 50 0 25 3 28 1 30 0 21 5 30 3 28 5 59 1 51.7 119 7 46 4 9 k C (J R) irtIbAn2 3697.2 4648 4 3333 0 3509 1 3621 4 30750 3637.1 3530 9 5048 6 4727.3 7132 5 44840 g n (JR) 0315 0 327 0 310 0 313 0 314 0306 0314 0 313 0.331 0 328 0 348 0 325 n - D N N D. ,
' l l Table A-3 I l Material Properties for the Cast Stainless Steel Surge Line Piping Including the Effect of Thermal Aging at 550 F l Parameter Value l Temp ( F) 550 ' 3 E (ksi) 26.1 x 10 [A-15] Sy = co(ksi) (1) S, (ksi) (1) l Ramberg-Osgood Parameter a. 1.25 [A-12] i Ramberg-Osgood Parameter n 6.60 [A-12] I 2 Jic(in-k/in ) 0.65 J-R Curve Parameter Ci(in-k/in:) (3) 1 J-R Curve Parameter N (1) 2 Jnm (in-k/in ) (2) (1) Actual Properties of each spool used in the evaluation. See Table A-2 for values for individual spool pieces. (2) ~ The value of Jnux determined for each spool at crack extension ( A a) of 0.4 inch which is the maximum crack extension from the test data in Reference A-11. l l l l SIR-98-096, Rev. O A-15 J. StructuralIntegrity Associates, Inc.
1 J l Table A-3A Material Properties for the Cast Stainless Steel Surge Line Piping Including the Effect of Thermal Aging at 650 F j l Parameter Value l Temp ( F) 650 3 E (ksi) 25.59 x 10 l l l Sy = co (ksi) (1) l Su (ksi) (1) l Ramberg-Osgood Parameter a 1.282 , 1 Ramberg-Osgood Parameter n 6.408 2 Ji c(in-k/in ) 0.65 2 J-R Curve Parameter Ci (in-k/in ) (3) J-R Curve Parameter N (1) 2 Jmo (in-k/in ) (2) I (1) Actual Properties of each spool used in the evaluation. See Table A-2 for values for individual spool pieces. (2) The value of Jma determined for each spool at crack extension ( A a) of 0.4 inch which is the maximum crack extension from the test data in Reference A-11. l l l SIR-98-096, Rev. O A-16 . Structural Integrity Associates, Inc.
Table A-4 '. l Lower Bound SMAW Material Propenies at 550 F l l l Parameter Value 1 1 Temp ( F) 550 3 E (ksi) 25 x 10 Sy = co(ksi) 49.4 S (ksi) 61.4 Ramberg-Osgood Parameter ot 9.0 i Ramberg-Osgood Parameter n 9.80 2 Jic(in-kiin ) 0.288 (1) 2 J-R Curve Parameter Ci(in-k/in ) 3.816 (2) J-R Curve Parameter N 0.643 (2)' ' 2 l Jom (in-k/in ) 2.345 (3) 2 2 (1) A very conservative value of 40 kJ/m = 0.228 in-kip /in in Reference A-18 was used. (2) The J-R curve in Reference A-18 was used and is given as 2 Ja = 40 + 83.5 ( A a) "3 kJ/m
= 0.228 + 3.816 ( A a)""3 in-kip /in 2 (3) Value of Jnm was calculated at A a = 0.4 inch.
SIR-98-096, Rev. O A-17 StructuralIntegrity Associates, Inc.
Table A-4A l Lower Bound SMAW Material Properties at 650 F Parameter Value Temp ( F) 650 3 E (ksi) 24.51 x 10 i l Sy = co(ksi) 47.2 S (ksi) 61.4 Ramberg-Osgood Parameter a. 9.227 Ramberg-Osgood Parameter n 9.515 2 Jic(in-k/in ) 0.288 (1) 2 J-R Curve Parameter Ci (in-k/in ) 3.816 (2) J-R Curve Parameter N 0.643 (2) 2 Jnm (in-k/in ) 2.345 (3) ; (1) A very conservative value of 40 kJ/m =2 0.228 in Reference A-18 was used. (2) The J-R curve in Reference A-18 was used and is given as 2 Ja = 40 + 83.5 ( A a) #3kJ/m
= 0.228 + 3.816 ( A a) #3 in-kip /in2 (3) Value of Jnm was calculated at A a = 0.4 inch.
SIR-98-096, Rev. O A-18 StructuralIntegrity Associates, Inc.
Table A-5 1 Determination of Ramberg-Osgood Parameters for A-105 at 650 F l Temperature F 550 650 Modulus of Elasticity, ksi 26800 25900 Yield Strength, ksi 27.85 26.1 Ultimate Tensile Strength, ksi 70 70 Ultimate Strain, in/in 0.3 0.3 Yield Strain, in/in 0.0010392 0.0010077 , a (Equation 19) 1.9245961 1.9846743 n (Equation 20) 4.1117073 3.8973976 a (Note 1) 2.51 2.588 n (Note 2) 4.2 3.981 Note 1: Value at 650 F is determined using value at 550 F [A-4] and Equation 22. Note 2: Value at 650 F is determined using value at 550 F [A-4] and Equation 23. I J STR-98-096, Rev. 0 A-19 gg,gggy,yg ygggy,,,,pgggg;gggg, fgg,
Table A-6 Determination of Ramberg-Osgood Parameters for SA-508 Class 2 at 650 F Temperature, F 550 650 Modulus of Elasticity, ksi 26700 25850 Yield Strength, ksi 44.15 43.5 Ultimate Tensile Strength. ksi 80 80 Ultimate Strain, in/in 0.3 0.3 Yield Strain, in/in 0.0016536 0.0016828 a (Equation 19) 1.209513 1.1885057 n (Equation 20) 5.6861564 5.5888515 a (Note 1) 2.51 2.466 n (Note 2) 4.2 4.128 Note 1: Value at 650 F is determined using value at 550 F [A-4] and Equation 22. Note 2: Value at 650 F is determined using value at 550 F [A-4] and Equation 23. SIR-98-096, Rev. O A-20 StructuralIntegrity Associates, Inc.
_ ~ __ - . . _ - . . .- l l Table A-7 Determination of Ramberg-Osgood Parameters for SA-351, l Grade CF-8M Cast Stainless Steel at 650 F l I Temperature, F 550 650 l Modulus of Elasticity, ksi 25550 25050 Yield Strength, ksi 19.35 18.5 Ultimate Tensile Strength, ksi 67 67 Ultimate Strain, in/in 0.3 0.3 Yield Strain, in/in 0.0007573 0.0007385 (x (Equation 19) 2.6408269 2.7081081 n (Equation 20) 3.2348215 3.1407678 et (Note 1) 1.25 1.282 l n (Note 2) 6.6 6.408 l Note 1: Value at 650 F is determined using value at 550 F [A-12] and Equation 22. l Note 2: Value at 650 F is determined using value at 550 F (A-12] and Equation 23. i l SIR-98-096, Rev. O A-21 Structural Integrity Associates, Inc. I
l Table A-8 ! Determination of Ramberg-Osgood Parameters for SMAW at 650 F Temperature, F 550l 650 Modulus of Elasticity, ksi 25550l 25050 Yield Strength, ksi 19.35 18.5 Ultimate Tensile Strength, ksi 67 67 Ultimate Strain, in/in 0.3 0.3 Yield Strain, in/in 0.0007573 0.0007385 l a (Equation 19) 2.6408269 2.7081081 I n (Equation 20) 3.2348215 3.1407678 l a (Note 1) 9.0 9.227 n (Note 2) 9.81 9.515 Note 1: Value at 650 F is determined using value at 550 F [A-3] and Equation 22. Note 2: Value at 650 F is determined using value at 550 F (A-3) and Equation 23. i l l l l l I SIR-98-096, Rev. 0 A-22 StructuralIntegrity Associates, Inc.
TEMP (*C)
-120 -100 -80 .e0 40 20 0 20 40 60 80 100
,3g 8 8 i i i i 4 i , ,
120 -
~~
180 110 - E a - 100 - 140 O 90 - 120 80 - O Filled Points A are Averages - 100 E 70 . E T
- 60 -
[ 80 $ o O 50 - O O 60 40 - 30 -' A - 40 20 - 10 . O - 20 0 - E b - 0 I i l I i i i i
-200' 150 -100 '50 0 50 100 150 200 TEMP (*F)
Figure A-l. Charpy "V" Notch Energy Transition Curve for A-106 Steel Pipe [A-6] SIR-98-096, Rev. O A-23 Structural Integrity Associates, Inc.
.. _ . . . ~ . . . _ _ _ _
l l l l 'd 120 l l 100 _ai i W' 80 60 e 40 i l 20[ - j il i 0 ': 0 0.1 0.2 0.3 0.4 0.5 - 0.6 0.7 Strain, in/in q l-D- 550 l 650 l j Figure A-2. Stress Strain Curves for A-105 Carbon Steel Nozzle at 550 F and 650 F i
)
l 2 SIR-98-096, Rev. O A-24 StructuralIntegrity Associates, Inc.
I 160 140 120 _M E.w W
. 100 F 80 60 y r
40 I 20 !! lE O ': 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Strain, Inlin f 550 ; 650l Figure A-3. Stress Strain Curves for SA-508 Class 2 Nozzle at 550 F and 650 F SIR-9S-096, Rev. O A-25 ' Structural Integrity Associates, Inc.
, ~ ... . ., - .- , . _ . ~ , . . . . . - - . _- . . . ..-, a l J (505 15-2) (505 02) 2431 (505-03)
@ ,, 7 (505-04-2) 1 IC-4620 4l i l C4621 1 3
(505-13) l C-46191 _1 5 (505-05) (504 54) { C-4619-3 l N 1C-461941 > L*9 (404-03) (505-04 1) - (508-04) IC 32231R] l C-4620-2l lC-4618-1] 1 (505 09) (50 M 8) lC 4619-5l [C 4619-4l j O- O Node Numbers ! () Spool Piece Number (505 15 1)
. l C-4620-11 O u teriaiio.atraer Figure A-4. Surge Line Piping Cast Stainless and Steel Materials Identification ,
4
. SIR-98-096, Rev. 0 A-26 Structural Integrity Associates, Inc.
i i l 60 50 . . __ m ____s gg&~J~~ Vr
.2 d30
$ \
m 20-! II 33 10 0 ': 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Strain,in/in G- 550 650 Figure A-5. Typical Stress Strain Curves for A-351 Grade CF-8M Cast Stainless Steel Nozzle at 550 F and 650 F l-l' i 1 L SIR-98-096. Rev. O A-27 StructuralIntegrity Associates, Inc.
l I-i 35
- l
~30 .._ ___ :: 0 !1 25 " *'W
._M 20 l E
15h
- i. 13 .
10 5 0 ": 0 0.1 0.2 0.3 . 0.4 0.5 0.6 0.7 Strain, in/in l --e-- 550 ; 6501 ! Figure A-6. Stress Strain Curves for SMAW Material at 550 F and 650 F l l l l t SIR-98-096, Rev. 0 A-18 StructuralIntegrity Associates, Inc.
I
- 4 l 1 Appendix A References l'
- l. Drawing No. E-234-005, " Nozzle Details", Rev.1.
- 2. Drawing E-233-746, (25203 - 29150, Sheet 13), " Nozzle Details for 96" I.D. Pressurizer ;
I - Millstone Unit No. 2," Rev. 2. i
- 3. EPRI Report NP-6301-D, " Ductile Fracture Handbook," June 1989.
- 4. EPRI Repon No. NP-6045 " Evaluation of Flaws in Ferritic Piping," October 1988.
4
- 5. Letter No. N-PENG-98-049 from CJ. Gimbrone (ABB) to Roy Terry (Northeast l
Utilities), " Piping and Pressurizer Surge Nozzle and Safe End Certified Material Test l Reports for Millstone Unit 2, dated August 26,1998. (Attachment C to Engineering l Record Correspondence,25203-ER-98-0115, Rev. 2). \ .
- 6. Vassilaros, M.G., Hays, R.A., Gudus, J.P. and Joyce J.A., "J-Integral Tearing Instability Analyses for 8-inch Diameter ASTM A106 Steel Pipe," NUREG/CR-3740, April 1984.
-7. ASME Boiler and Pressure Vessel Code, Sections II and III Appendices,1989 Edition. l
- 8. NUREG-1061, Volume 3 " Report of the U. S. Nuclear Regulatory Commission Piping Review Committee Evaluation of Potential for Pipe Breaks," prepared by the Piping Review '
Committee the Pipe Break Task Group, NRC, November 1984.
- 9. Drawing No. E-234-004, " Piping Details and Assembly," Rev. 3.
- 10. Specification P8E7a) " Purchase Specification For Austenitic Stainless Steel Castings, Section III ASME Code," Combustion Engineering, December 17.1968 (Attachment D to Engineering Record Correspondence 25203-ER-98-0115, Rev 2).
I1. NUREG/CR-4513, Rev.1," Estimation of Fracture Toughness of Cast Stainless Steels During Thermal Aging in LWR Systems," August 1994.
- 12. NUREG/CR-6142," Tensile-Propeny Characterization of Thermally Aged Cast Stainless Steels." February 1994.
- 13. L. S. Aubrey, P. F. Wieser, W. J. Pollard, and E. A. Schoefer, " Ferrite Measurement and Control in Cast Duplex Stainless Steel," in Stainless Steel Castings, V. G. Behal and A.
S. Melilli, editors, ASTM STP 756, pp.126164 (1982).
- 14. Letter No. N-PENG-98-047 from C. Gimbrone (ABB) to Roy Terry (Northeast Utilities),
" Surge Line Piping Certified MaterialTest Reports for Millstone Unit 2." dated August 26,1998 (with attachments), (Attachment B to Engineering Record Correspondence, 25203-ER-98-0115, Rev. 2).
S1R-98-096, Rev. O A-29 Structural Integrity Associates, Inc.
. ~ . _ - ._. __
.- . _ . - . _ . _ ~ _ . - . _ - _ - - - - . _ . . -
- 15. NUREG/CR-6275," Mechanical Properties of Thermally Aged Cast Stainless Steels from Shipping Port Reactor Components," April 1995.
- 16. Letter No. N-PEN-98-043 from C. Gimbrone (ABB) to Roy Terry (Northeast Utilities).
" Weld Process and Filler Material Information for Millstone Point 2 Surge Line," dated July 31,1998 (with attachments). (Attachment A to Engineering Record Correspondence, 25203-ER-98-0115, Rev. 2).
- 17. ASME Section XI Task Gmup for Piping Flaw Evaluation, ASME Code, "Ev ~ iation of Flaws in Austenitic Steel Piping," Joumal of Pressure Vessel Technology, V . . 108, August 1986, pp. 352-366.
l
- 18. NUREG/CR-6428, " Effects of Thermal Aging on Fracture Toughness and Charpy- - ,
Impact Strength of Stainless Steel Pipe Welds," May 1996. j 1
- 19. Cofie, N.G., Miessi, G.A., and
Deardorff,
A.F., " Stress-Strain Parameters in Elastic-Plastic l Fracture Mechanics," Smirt 10 Intemational Conference, August 14-18,1989. ! l i SIR-98-096, Rev. O A-30 Structural Integrity Associates, Inc.
APPENDIX B Piping Loads I SIR-98-096, Rev. O B-0 StructuralIntegrity Associates, Inc.
l Appendix B Pipine Loads l The piping loads considered in the LBB evaluation are due to pressure (P), dead weight (WT), surge line thermal expansion (SLTE), reactor coolant system thermal expansion (RCTE), thermal stratification (TS), and safe shutdown earthquake (SSE), consistent with the guidance provided in NUREG-1061, Vol. 3 [B-1]. Per the guidance provided in NUREG-1061, other secondary stresses were not included in the evaluation. All loads, except seismic, were derived using an ANSYS [B-2] model of the surge line. The model is shown in Figure B-1. Loading conditions were evaluated for a number of cases as follows: e Dead Weight (DW): This case considered the weight of the piping, water, insulation and forces applied by hangers. Loads are shown i.. Table B-1.
. Surge Line Thermal Expansion (SLTE): This case evaluated the thermal expansion of the surge line with a temperature equal to the average of the pressurizer temperature (653 F) and the hot leg (604*F), or 628.5*F [B-3]. Loads are shown in Table B-2. . RCS Thermal Expansion Movement (RCTE): This case evaluated the loads due to the thermal expansion of the reactor coolant hot leg [B-4). These loadings were considered in the thermal expansion category (as opposed to anchor movement) since the surge line is part of the reactor coolant piping and there is significant movement of the RCS hot leg nozzle. Results are shown in Table B-3. . Stratification (TS): Global stratification moments and forces were developed for a 100*F linear top-to-bottom temperature distribution in the piping. Results are shown in Table B-4.
, Table B-5 shows the combined loads for normal operating conditions (without stratification = DW l + SLTE + RCTE). In the load combination. the dead weight and thermal moments are added algebraically at the component level and then, the resulting moments are calculated. For Tables B-1 SIR-98-096, Rev. O B-1 ' Structural Integrity Associates, Inc. {
to B-5, forces and moments are shown in the global X, Y, Z coordinate system (see Figure B-1) and are the loads acting on the node from the element toward the hot leg. Axial stress is negative for compressive axial force. OBE Seismic moments and forces were derived from the original rextor coolant piping analysis for ' Millstone Unit 2 and were available at Nodes 2,4,11, and 15 based on the same coordinate system [B-5]. Conservative assumptions were used to determine loads for other locations, as shown in Table B-6.
.* It was observed that seismic moments were maximum at the hot leg nozzle and the pressurizer nozzle, as expected.
- For Node 3, the loads at Node 2 were used
.] = For Node 5, the loads at Node 4 were used.
e ~ For Nodes 7 and 8, the average of the loads at Nodes 4 and 11 were used. Since Nodes 7 and 8 are much nearer to Node 11 than Node 4, this is conservative.
- For Node 10, the loads at Node 11 were used.
- For Nodes 13 and 14,' the loads at the pressurizer nozzle (Node 15) were used
. For Node 16, a linear distribution in moments was assumed between the elbow below the pressurizer and Node 11. The forces for Node 11 are directly applicable.
For comparison, the nominal axial pressure stress in the 12 inch Schedule 160 surge line piping is 3.818 ksi at 2235 psig. Nominal stresses shown in the tables of this appendix are derived as follows for the surge line piping: F ,,, bai = A N SRSS b.= Z l l i i i SIR-.98-096, Rev. O B-2 StructuralIntegrity Associates, Inc. -l 5
where: Faw: = load acting parallel to axis of pipe (considered positive for tension), kips
-A = surge line nominal pipe area,in:
.Msass = SRSS moment,in-kips s Z = nominal pipe section modulus, in' SIR-98-096, Rev. O B-3 StructuralIntegrity Associates, Inc.
Table B-1 Dead Weight Loads Force,Ib Moments, in-lb Axial Forcei Nominal Stress, ksi Node Fx l Fy l Fz .Mx l My l Mz lSRSS kips . Moment: Axial Force 2 -8.4 1626.7 -0.1 -43321 -1513 -17107 l 46601 -1.627 l 0.42 l -0.039 3 -8.4 1124.4 -0.1 -18560 -1362 -17259 l 25381 0.000 1 0.23 l 0.000 4 -8.4 -48.3 -0.1 16950 -8% -17259 l 24204 0.000 l 0.22 ; 0.000 5 -8.4 -550.7 -0.1 11560 -652 -22650 ! 25438 -0.008 l 0.23 l 0.000 7 -8.4 -212.5 -0.1 11560 -643 9776l15153 -0.008 ; 0.14 0.000 8 -8.4 -714.9 -0.1 3212.3 -489 1428 l 3549 0.000 l 0.03 l 0.000 10 -8.4 859.3 -0.1 14034 -39 1428 l 14107 0.000 l 0.13 l 0.000 11 -8.4 691.7 -0.1 21019 37 -445 l 11024 0.004 l 0.19 l 0.000 16 -8.4 -10.0 -0.1 32672 32 4 -7180 l 33453 0.004 l 0.30 l 0.000 13 -8.4 576.2 -0.1 -19487 1394 22965!30151 0.004 l 0.27 { 0.000 14 -8.4 l 73.8 -0.1 l-14421 l 1524 19886. 24612 -0.074 ! 0.22 i -0.002 15 -8.4 -383.7 -0.1 l -14419 l 1524 19669 ' 24436 0.384 l 0.22 l 0.009 SIR-98-096, Rev. 0 B-4 StructuralIntegrity Associates, Inc.
Table B-2 Surge Line Thermal Expansion Loads for Normal Operation (628.5 F) Force, Ib Moments. in-lb Axial Force Nominal Stress, ksi Node Fx l Fy l Fz Mx l My l Mz lSRSS kips Moment Axial Force 2 l-5158.0 969.9 l 17794.0 174450 730960 48790 753071 -0.970 6.75 -0.023 3 l-5158.0 969.9 l 17794.0 -128380 823800 -44054 834906 17.794 7.49 -0.425 4 -5158.0 969.9 l 17794.0 -64366 1164200 -44054 1166810 -17.794 10.46 -0.425 5 -4650.4 969.9 l 17794.0 -46908 927640 -26596 929206 -4.650 8.33 -0.111 7 l -4650.4 807.9 l 17558.0 -46908 -1237500 77562 1240815 -4.650 11.13 -0.111 l 8 -4650.4 807.9 l19112.0 -32366 -1497800 92104 1500978 -19.112 13.46 -0.457 10 l-1904.9 1974.1 l 19112.0 18721 -1330000 92104 1333317 -19.I12 11.96 -0.457 I 11 -1904.9 1074.1 l 19112.0 28396 -1266700 89509 1270176 -15.599 11.39 -0.373 16 -1904.8 1074.1 19112.0 65118 -823830 68282 829216 -15.599 7.44 -0.373 13 -1124.7 1966.6 l 19373.0 234700 807270 -29734 841221 -16.215 7.54 -0.388' 14 l-l 124.7 1966.6 l19373.0 -83368 999310 -67694 1005064 -1.967 9.01 -0.047 15 l-l 124.7 1966.6 I19373.0 -582220 999310 -96655 1160579 -1.967 10.41 -0.047 l l l l J SIR-98-096 Rev. 0 B-5 Struc!UralIntegrity Associates, Inc.
Table B-3 i RCS Thermal Expansion Movement Loads for Normal Operation (604 F) l Force, Ib Moments, in-lb Axial Force Nominal Stress, ksi Node Fx l Fy l Fz Mx l My l Mz lSRSS kips Moment l AxialForce ! 2 -5025.2 977.2 2131.5 -63170 -442480 57930 450705 -0.977 4.04 -0.023 3 -5025.2 977.2 2131.5 -83947 -352030 -32523 363359 -2.132 3.26 -0.051 4 -5025.2 977.2 2131.5 -19451 -20364 -32523 43021 -2.132 0.39 -0.051 5 -3374.7 977.2 2131.5 -l861 2013 -14933 15183 -3.375 0.14 -0.081 7 -3374.7 351.1 1970.4 -1861 -245000 47001 249475 -3.375 2.24 -0.081 8 -3374.7 351.1 1443.3 4459 -210230 53320 216932 -1.443 1.95 -0.034 10 -20 0 .8 171.5 1443.3 17947 -68746 53320 88832 -1.443 0.80 -0.034 11 -2049.8 171.5 1443.3 19492 -46795 52906 73272 -0.225 0.66 -0.005 16 -2049.8 171.5 1443.3 25356 51814 49516 76023 -0.225 0.68 -0.005 13 -1837.1 147.5 1322.8 46388 409870 37360 414175 -0.227 3.71 -0.005 14 -1837.1 147.5 1322.8 24875 450410 2964 451106 -0.147 4.05 -0.004 15 -1837.1 147.5 1322.8 -9187 450410 -44340 452680 -0.147 l 4.06 -0.004 i SIR 98-096, Rev. O B-6 f StructuralIntegrityAssociates,Inc. l
Table B-4 Stratification Loads for 100 F Linear AT Force, Ib Moments, in-lb Axial Force l Nominal Stress. ksi Node Fx l Fy l Fz Mx l My l Mz lSRSS kips lMomenti AxialForce 2 9.9 l-387.0 394.3 659590 33126 180440 684628 0.387 0.009 6.14 l 3 9.9 l-387.0 394.3 615530 32948 18 % 20 671132 -0.394 6.02 l -0.009 4 9.9 -387.0 394.3 619980 32296 18 % 20 646562 -0.394 -0.009 5.80 l 5 10.6 l-387.0 394.3 613020 25008 173650 637631 0.011 5.72 l 0.000 7 10.6 l-619.9 441.8 613020 -27868 104450 622479 0.011 5.58 , 0.000 8 10.6 l-619.9 522.8 601860 -37469 93291 610199 -0.523 5.47 l -0.012 10 17.7 l 796.9 522.8 610500 -38246 93291 618770 -0.523 5.55 ! -0.012 11 17.7 l 796.9 522.8 617680 -37142 91366 625504 -0.462 5.61 -0.011 16 17.7 l 796.9 522.8 644930 -27415 75617 649926 -0.462 5.83 -0.011 13 9.4 l 330.2 614.3 729740 11133 26594 730309 -0.013
-0.537 l 6.55 14 9.4 l 330.2 l 614.3 723830 16519 23790 724409 -0.330 l 6.50 -0.008 15 9.4 l 330.2 614.3 708020 16519 24033 708620 -0.330 6.35 -0.008 SIR-98-096, Rev. O B-7 StructuralIntegrity Associates, Inc.
Table B-5 Combined Loads for Normal Operation Force, Ib Moments, in-lb l Axial ForceNominal Stress, ksi Node Fx l Fy l Fz l Mx My l Mz lSRSS l kips ; Moment l Ax' ' Tree 2 -10191.6 3573.9 19925.4 67959 286967 89613 308219 -3.574 2.764 -0.085 3 -10191.6 3071.6 19925.4 -230887 470408 -93836 532351 -19.925 4.774 -0.476 4 -10191.6 l 1898.8 19925.4 -66867 1143030 -93836 1148823 -19.925 10.302 -0.476 5 -8033.5 1396.5 19925.4 -37209 929001 -64179 931958 -8.034 8.357 -0.192 7 -8033.5 946.4 19528.3 -37209 -1483143 134339 -8.034 1489679l 13.358 -0.192 8 -8033.5 l 444.0 20555.2 -24695 -1708519 146852 1714997 l -20.555 15.378 -0.491 10 -3963.1 l 2104.9 20555.2 50702 -1398785 146852 1407386 l -20.555 12.620 -0.491 11
-3963.1 l 1937.3 20555.2 68907 1313458 141970 1322904 l -15.820 11.863 -0.378 16 3963.0 l 1235.6 20555.2 123146 -771693 110618 789247 l -15.820 7.077 -0.378 13 2970.2 l 2690.2 20695.7 261601 I218534 30591 1246673 l -16.438 11.I79 -0.393 14 -2970.2 l 2187.9 20695.7 -729I4 1451244 -44844 1453766! -2.I88 13.036 -0.052 15 -121326
-2970.2 l 1730.3 l 20695.7 l-6058261451244 1577293 l -1.730 14.144 -0.041 Loads are due to DW + SLTE + RCTE i
SIR-98 096. Rev. 0 B-S StructuralIntegrity Associates, Inc.
- _- ... - =. .. . -.. - -. .
Table B-6 OBE Seismic Loads Forces. Kips Nominal SRSS Moment Nominal Axial Moment Node F, F, F, (in-kips) Stress, ksi Stress, ksi 2 16.0 4.0 4.0 277.6 0.096 2.489 3 6.0 4.0 4.0 277.6 0.096 2.489 4 4.5 3 *3.5 204.4 0.084 1.833 5 4.5 3 3.5 204.4 0.084 1.833 7 3.3 2.0 3.0 156.7 0.%9 1.405 8 3.3 2.0 3.0 156.7 0.069 1.405 10 *2.0 1.0 2.5 108.9 0.077 0.977 11 2.0 . 1.0 2.5 108.9 0.077 0.977 16 2.0 1.0 2.5 125.7 0.081 1.127 13 4.0 4.0 6.0 192.7 0.096 1.728 14 4.0 14.0 6.0 192.7 0.096 1.728 15 4.0 4.0 6.0 192.7 0.096 1.728 SIR-98-096, Rev. 0 B-9 StructuralIntegrity Associates, Inc.
,-X +Z 8
9 w 11 w g Hot eg
+Z
- w
+Y +Z Pre rizer W i W = Weld location w@ 3w ]
l
.l Figure B-1. Mathematical Model of Surge Line Piping and Weld Locations l
1 SIR-98-096, Rev. O B-10 StructuralIntegrity Associates, Inc.
Appendix B References 1. NUREG-1061, Volumes 3. " Report of the U. S. Nuclear Regulatory Commission Pipmg
' Review Committee," prepared by the Piping Review Committee, NRC, November,1984. l
- 2. ANSYS Linear Plus/ Thermal, Revision 5.3, Second Release, ANSYS Inc., October 1996.
- 3. Specification No. 18767-31-5, Rev.16, " Engineering Specification for a Reactor Coolant Pipe and Fittings for Northeast Utilities Service Company, Millstone Point Station, Unit j 2," November 1997. '
- 4. Drawing 7604-M730-21-1, "Rextor Coolant Nozzle Locations and Movements" (25203-29152, Sheet 9-11),12/22n0.
- 5. Document M2-EV-98-0165, " Technical Evaluation for Millstone 2 Plant Specific Surge Line OBE Loads," Rev. 00, 8/25/98 (Attached to Engineering Record Correspondence 25203-ER-98-0115, Rev. 2).
i SIR-98-096. Rev. 0 B-l1 StructuralIntegrity Associates, Inc.
APPENDIX C Critical Flaw Size and Leakage Evaluation S1R-98-096 Rev. 0 C-0 ( StructuralIntegrity Associates, Inc.
Appendix C Critical Flaw Size and Leakane Evaluation C.1 Evaluation of Critical Flaw Sizes Critical flaw sizes may be determined using net section collapse criterion (NSCC) approach or J-Integral /rcaring Modulus (J/r) methodology. NSCC is particularly suited for materials with a considerable amount of ductility and toughness such as stainless steel materials. since it assumes that the cross-section of the pipe becomes fully plastic at the onset of failure. As such, for circumferential flaws, NSCC is less conservative compared to the J/r methodology which is based on elastic-plastic fracture mechani:s (EPFM) principles. In this evaluation, the critical flaw sizes will therefore be determined based on the J/r approach. A methodology for fracture mechanics analysis for determining the stability of through-wall circumferential flaws in cylindrical geometries such as pipes is presented in References C-1 and i C-2. This methodology was used for the determination of critical flaw sizes in the pressurizer surge line piping at Millstone 2, using computer program, pc CRACK [C-3] which has been verified under Srs Quality Assurance program. The expression for the J-integral for a through-wall circumferential crack under tension loading (C-1, C-2] which is applied in this analysis is:
"^
a,, R '- + aa,e,c 'a' h,
' 2 '
P a R P' J = f, - - s t,E t b; < b , n,-t , , P, , where: 2 faR'
' a*F
'b,t' f, a,, R '=
t, 4nR ,-t .- s ae = effective crack length including small scale yielding correction R = mean pipe radius SLR-98-096, Rev. 0 C-1 StructuralIntegrity Associates, Inc.
- t. . = actual pipe wall thickness at weld location F. . = elasticity factor (C-1, C-2]
P
= appliedload = o (2xRt) c = remote tension stress in the uncracked section
. ot =' Ramberg-Osgood material coefficient E = elastic modulus ao =. yield stress so = yield strain 2a = total crack length .
;2b = 2xR-c = b-a-hi - = plasticity factor (C-1, C-2]
Po = limit load corresponding to a perfectly plastic material
- n ~ = Ramberg-Osgood strain hardening exponent.
Similarly, the expression for the J-integral for a through-wall crack under bending loading [C-2] is given by : 1 J = f, a,, R ' M 2
+ otoo e c
'a' h 'a.,n,-
RY M i t tg E ,b j b t M,, L The parameters in the above equations are the same as the tension loading case except 2
- y. -= applied moment = o.,xR t c =- remote bending stress in the uncracked section I. =. ' moment ofinertia of the uncracked cylinder about the neutral axis
- Mo -= limit moment for a cracked pipe under pure bending corresponding to'n = * (elastic-perfectly plastic case)
' SIR-98-096, Rev. O ~ C-2 h M I/ % Q S% ides,IK v
<4 - .- , L _ , . _ ~ . . . . _
4 r 3 g '
= M', Cos --Sin (y) t
= 2 M ', _ limit moment of the uncracked cylinder = 4co R t !
The Tearing Modulus (T) is defined by the expression: T = dJ E2 da og Hence, J from the above expressions is determined as a function of crack size (a) and the slope of the J versus crack size (a) curve is calculated in order to determine T. (The flow stress, or, is taken as the mean of the yield and ultimate tensile strengths.) The material resistance J-R curve can also
- be transformed into J/r space in the same manner. The intersection of the applied and the material J/r curves is the point at which instability occurs and the crack size associated with this instability point is the critical crack size at J-valves. This is shown schematically in Figure C-1. The material J-T curve beyond the value of Jnm provided in the tables in Appendix A were extrapolated using the procedure provided in Section A.2.4.3 of NUREG-1061, Vol. 3. An example of the critical flaw
' size determination for Node Point 15 under tension loading is shown in Figure C-2.
Normal operating conditions were utilized in the evaluation. Stratification moments were not included (in both the critical flaw sizing and leakage evaluation), because the effects are minimal at normal operating conditions. The pressurizer surge line piping stresses consist of both tension and bending stresses. The tension stress is due to internal pressure and axial forces while the bending 4 stress is caused by deadweight, thermal and seismic moments. Because a fracture mechanics model for combined tension and bending loading is not readily available, an altemate analysis is performed to determine the critical flaw le'ngth under such loading condition using the tension and bending models separately. For the'first case, the stress combination is assumed to be all due to tension and the critical flaw length is determined using the tension model. For the second case, the stress combination is assumed to be all due to bending and the critical flaw length is determined as such.
-The half critical flaw sizes (lengths) obtained with the tension model (a,) and the bending model (ae)
' ^
. SIR 98' 096, Rev. 0 -
C-3 gg,yggy,,g ggggy,ggy pgggcjgggg, jng, y $ uFe-+w > p i N hr-- = ,s - , - - - y . - - - ,,e-, +-g- g y+e -'--w 1
are combined to determine the actual half critical flaw size (ac ) due to a combined tension and bending stress using linear interpolation, as described by the following equation: G, Os a,=a,O b Y0 t
+ ab O b N0 s where ci and o nare the piping tensile and bending stresses respectively. In determining the critical flaw sizes, the following assumptions were made:
- The as-designed pipe wall thickness was used for both the piping and nozzle welds.
- - The material properties determined in Appendix A at 650 for the various materials were used at all locations. At the nozzle locations, all materials (ferritic nozzle, SMAW weld and ;
cast stainless steel) were evaluated. For the cast stainless steel piping welds, the most conservative material properties on either side of the weld were used in the evaluation, except at Nodes 14 and 15, where evaluation was conducted at both sides of the weld.
+'
The models used are limited to critical flaw sizes of half circumference. For locations where the stresses are low resulting in critical flaw sizes in excess of half the circumference, extrapolation was used to determine the actual critical flaw sizes.
- For calculating the critical through-wall length, plane stress condition was used.
- A value of Poisson's is ratio oi0.3 was used.
The results of the critical flaw length detennination are presented in Tables C-1 through C-4. Tables C-1 and C-2 are for the piping welds. Tables C 3 and C-4 are for the hot leg and pressurizer
- nozzle welds. Because of very different material and geometric combinations at the nozzle welds.
SIR-98-096, Rev.' 0 C-4 gg,yggy,,y pyggy,jgy pggggggggg, jag,
1 several locations were investigated as shown in Figure C 3. In Tables C-1 and C-3, the critical flaw length is determined with a factor of unity on the normal + SSE stress combination. In Tables C-2 and C-4, the critical flaw length is determined with a factor of 8 on the nonnal + SSE stresses. NUREG-106), Vol. 3 requires a safety factor of 2 between the critical flaw size calculated with safety factor of unity on the stresses and the 10 gpm leakage size flaw that will discussed in Section C.2. It'also requires that the critical flaw rize calculated with a safety factor of 8 on the stresses
. be at least equal to the 10 gpm leakage size flaw.
C.2 Leak Rate Determination The determination of leak rate is performed using the EPRI program, PICEP [C-4]. The flow rate equations in PICEP are based on Henry's homogeneous nonequlibrium critical flow model [C-5).
' The program accounts for nonequlibrium " flashing" mass transfer between liquid and vapor phases, fluid friction due to surface roughness and flow paths with area variation through the wall thickness.
In the determination of leak rates using PICEP, the following assumptions were made: The as-designed pipe thickness at the welds was used. This accounts for the actual expected thickness at each weld location. The loading considered was that due to normal operating pressure and piping thermal expansion of the surge line and reactor coolant system piping. Stratification was not included since its effect is minimal at normal operating conditions. e For the cast austenitic stainless steel material at each weld, the side with the highest yield strength is used since this results in the greatest required crack length to produce a given leakage except for Nodes 14 and 15, where both sides are evaluated. The weld material at each location is also evaluated. Material properties are chosen at 650 F, consistent with the critical flaw size evaluations, except at Node 2, where properties at 604 F were used. A plastic zone conection is included. This is consistent with fracture mechanics principles for ductile materials. The crack is assumed to be elliptical in shape with the inlet area equal to the exit area.
- Crack roughness is taken as 0.000197 inches [C-6].
SIR-98-096, Rev. 0 C-5 Structural Integrity Associates, Inc.
There are no tuming losses assumed since the crack is assumed to be initiated by some mechanism other than IGSCC. e A sharp-edged entrance loss factor of 0.61 is used (PICEP default).
- The default friction factors of PICEP am utilized, e
The load combination used for leakage calculation is due to pressure, dead weight, surge line thermal expansion, and RCS thermal expansion. The leakage was calculated for an operating pressure of 2250 psia. Leakage was calculated based i on water at 628 F at all locations except at the hot leg nozzle where a temperature of 605*F was l used. For the hot leg nozzle, the lower temperature is used due to the close proximity to the hot leg, where turbulence will assure that the region is filled with the relatively cooler hot leg water.- For the other regions, a 10 gpm leak would be that due to 1.5 gpm normal flow from the pressurizer (at 653 F) and 8.5 gpm from the hot leg (at 604 F). Since leakage rate increases slightly with i 1 r subcooling of the liquid, the temperatures chosen for evaluation are conservative. Table C-5 shows I the predicted crack sizes for 10 gpm leakage at each pipe weld location in the surge line. Table C-6 shows the crack sizes i7r the relatively thicker safe-end to nozzle weld regions. [ l l I l l V . t [
~ SIR-98-096, Rev. 0 C-6 gg,yggy,,;ggggy,ggy pggggjgggg, jng,
Table C-1 Critical Flaw Sizes at Piping Welds Based on Factor of Unity on Normal Plus SSE Loads Critical Flaw Size (in.)(1) Cast Stainless Piping SMAW Welds Node No. a, as l ae a< as l ac 3 5.97 7.57 7.05 6.96 8.33 7.89 4 4.72 6.20 5.83 5.80 7.19 6.84 5 5.12 6.64 6.20 6.22 7.61 7.21 7 4.02 5.55 5.19 5.20 6.56 6.24 ^ 8 3.70 5.I8 4.88 4.83 6.21 5.93 10 4.52 6.00 5.65 5.65 7.04 6.70 11 4.65 6.17 5.78 5.81 7.20 6.85 16 6.08 7.68 7.14 7.07 8.43 7.97 13 4.52 5.97 5.61 5.60 6.99 6.65 14 3.89 5.41 5.05 5.09 6.46 6.13 14a 4.51 5.90 5.57 - - - Notes: (1) For comparison with the leakage flaw size for norrnal plus SSE loads, one-half of the actual critical flaw size (a c) is reponed to account for the required margin of 2. (2) 14a is side of weld with high yield strength material. SIR-98-096. Rev. 0 C-7 gg,gggy,,, ggggy,,,,pggggygggg, fag,
Table C-2 Critical Flaw Sizes at Piping Welds Based on Factor of d on Normal Plus SSE Loads Critical Flaw Size (in.)(1) Cast Stainless Piping SMAW Welds Node No. ai as l 2ac ai as 2ac 3 4.40 5.81 10.71 5.36 6.72 12.55 4 2.76 4.45 8.06 4.13 5.54 10.38 5 3.24 4.88 8.79 4.59 5.98 11.14 7 1.90 3.58 6.37 3.36 4.93 9.12 8 1.61 3. I8 5.72 2.83 4.57 8.43 10 2.40 4.16 7.47 3.98 5.40 10.11 11 2.61 4.39 7.87 4.15 5.56 10.39 16 4.52 5.94 10.92 5.48 6.84 12.76 13 2.44 4.13 7.43 3.92 5.34 9.98 14 1,79 3.44 6.09 3.20 4.81 8.86 14a 2.50 4.09 l 7.42 - - - Notes: (1) For comparison with the leakage flaw size for 4 (normal and SSE Loads), the full critical flaw size (2ae )is reported. (2) 14a is side of weld with high yield strength material. S1R-98-096. Rev. O C-8 gg,y,,,,,, ,,,,,,,gy ,gggg,3ggg, ,gg.
l l Table C-3 l Critical Flaw Sizes at Nozzle / Safe-End Welds Based on Factor of l Unity on Normal Ph.s SSE Loads ; Half Critical Flaw Size (in)(1) Node No. Material a, as ae 2a SA-105 8.70 10.56 9.84 2a SMAW 9.57 10.41 10.08 Hot Leg 2b SMAW 7.99 9.17 8.72 Nozzle 2b Cast Stainless 7.06 8.63 8.04 2c Cast Stainless 6.50 8.12 7.48 2c SMAW 7.47 8.81 8.28 15a l SA-508 5.61 7.24 6.89 15a l SMAW 6.42 7.69 7.41 15b l SMAW 5.99 7.28 7.00 Pressurizer 15b l Cast Stainless 4.52 5.97 5.65 Nozzle 15e i Cast Stainless 3.86 5.46 5.09 15c I SMAW 5.51 6.84 6.53 15d l SMAW 4.84 6.23 5.91 15d l Cast Stainless 4.24 5.64 5.32 Note: (1) For comparison with the leakage flaw size for normal plus SSE loads, one-half of the actual critical flaw size (a e) is reported to account for the required margin of 2. SIR-98-096, Rev. 0 C-9 StructuralIntegrity Associates, Inc.
Table C-4 Critical Flaw Sizes at Nozzle / Safe-End Welds Based on Factor of 4 on Normal Plus SSE Loads Critical Flaw Size (in)(1) Node No. Material a, as 2ac 2a SA-105 7.21 8.92 16.51 2a SMAW 8. I 1 9.30 17.68 Hot Leg 2b SMAW 6.49 7.79 14.59 Nozzle 2b Cast Stainless 5.51 7.02 12.90 2c Cast Stainless 4.92 6.40 11.62 2c SMAW 5.91 7.30 13.50 1Sa SA-508 3.93 5.40 10.17 1Sa SMAW 4.77 6.07 11.58 15b SMAW l 4.38 5.68 10.79 Pressurizer 15b Cast Stainless 2.37 4.14 7.52 Nozzle 15c Cast Stainless 1.68 3.40 6.01 1Sc SMAW , 3.76 5.23 9.78 15d SMAW 2.84 4.58 8.37 15d Cast Stainless 2.22 3.77 6.84 Note: (1) For comparison with the leakage flaw size for 8 (normal and SSE Loads), the full critical flaw size (2ae) is reported. SIR-98-096, Rev. 0 C-10 StructuralIntegrity Associates, Inc.
Table C-5 10 gpm Leakage Crack Sizes for Piping Welds Cast Stainless Piping SMAW Welds Length (2a), Crack Length (2a), Crack Node inches Area. in: Inches Area. in 3 7.00 0.03767 6.92 0.03754 4 5.56 0.03538 5.58 0.03543 5 5.90 0.03595 5.91 0.03596 7 4.91 0.03428 5.04 l 0.03451 8 4.73 0.03397 4.80 0.03410 10 5.18 0.03475 5.19 0.03478 11 5.7.3 0.03483 5.29 0.03495 16 6.29 0.03658 6.24 0.03650 13 5.38 0.03509 5.41 0.03515 14 5.04 0.03450 5.06 0.03455 14a (1) 4.87 0.03422 - - Note: (1) Node identified as 14a is for side of weld with low yield strength material. SIR-98-096. Rev. O C-11 StructuralIntegrity Associates, Inc.
l Table C-6 10 gpm Leakage Crack Sizes for Nozzle Welds Node No. Material Length (2a), Crack area, inches in 2 l 2a SA-105 9.12 0.03911 2a SMAW 9.16 0.03917 Hot leg 2b SMAW 7.73 0.03499 Nozzle 2b Cist Stainless 7.83 0.03513 2c Cast Stainless 7.19 0.03321 1 2c SMAW 7.09 0.03308 ) 15a SA-508 5.43 0.03725 15a SMAW 6.01 0.03831 15b SMAW 5.72 0.03729 Pressurizer 15b Cast Stainless 5.09 0.03528 Nozzle 15c Cast Stainless 4.94 0.03501 l 15c SMAW 5.31 0.03569 ' 15d SMAW 4.91 0.03428 15d C:st Stainless 4.86 0.03419 l l l \ SIR-98-096, Rev. O C-12 StructuralIntegrity Associates, Inc.
J J l INSTABILITY I
. . . . . . . . . . . . . . . ................................. ..j..........
l MATERIAL dJ., (SLOPE da I I J,, .... ............................................ 3....................................... I APPLIED I Aa dJ E T = da (cy,2) 93220ro Figure C-1, J.Integralfrearing Modulus Concept for Determination of Instability During
~ Ductile Tearing SIR 98-096, Rev. O C-13 StructuralIntegrity Associates, Inc.
l l l l Instability Analysis 5.0 l
] $ Crackincrement
- Matenal 4.0 3.5 l
3.0 l 3 2.5 kg ^
- a = 7.7"
.E -1
' ' a = 7.6"
, I \ a = 7.5" 2.0 1.5 E ++
1.0 0.5 x x 0.0 0 12.5 25 37.5 50 62.5 75 87.5 100 Tearing Modulus .b Figure C-2. Determination of Critical Flaw Size Under Tension Loading for Node Point 15 (SMAW Weld at PZR Nozzle) Using Jfr Approach SIR-93-096, Rev. O C-14 StructuralIntegrity Associates, Inc.
= . . - ~. - --. -.. . - _ . - . - . = _ - . - - - . . - = - . - . . - . . - - - - . .
I- [, l ! HOT LEG NOZZLE , OD = 13.5" 7 SMAW Weld SMAW Weld OD = 12.76' /
- ['h.' C-4618-1 C-4620-4 SA-105 ^f@^ Cast Stainless $, Cast Stainless 2a 9 2b ID = 10.125" / N t = 1.6875"J L =t 1.3125" t = 1.148" PRESSURIZER NOZZLE l
OD = 13.1875" SMAWWeld SMAW Weld l OD = 13.0"
'M9 OD = 12.76' l ',. C-3223-1R Test 1 & Test 2 .. C-4619-4 SA-508 l Cast Stainless 15c Cast Stainless i
' 15a 15b 15d v: N t = 13125"J L =t 1.4375" t = 1.273" t = 1.148" l 98206r0 Figure C-3. Locations for Evaluation at Hot Leg and Pressurizer Nozzles l l i l l l i i SIR-98-096, Rev. O C-15 f StructuralIntegrityAssociates,Inc.
Appendix C References 1
- 1. Kumar, V., et al.,"An Engineering Approach for Elastic-Plastic Fracture Analysis," EPRI NP-1931, Electric Power Research Institute, Palo Alto, CA, July 1981.'
- 2. Kumar, V., et al., " Advances in Elastic-Plastic Fracture Analysis," EPRI NP-3607, Electric Power Research Institute, Palo Alto, CA, August 1984.
I
- 3. Stmcairal Integrity Associates,Inc.,"pc CRACK Fracture Mechanics Software," Version 3.0 - 3/27/97.
. 4. EPRI Report NP-3596-SR, "PICEP: Pipe Crack Evaluation Computer Program," Rev.1, July 1987.
- 5. P.E. Henry, "The two-Phase Critical Discharge of Initially Saturated or Subcooled Liquid,"
. Nuclear Science and Engineering, Vol. 41,1970.
i
- 6. EPRI Report NP-3395, " Calculation of Leak Rates Through Crack in Pipes and Tubes,"
)
December 1983. l SIR-98-096, Rev. O C-16 h StructuralIntegrity Associates, Inc. L
_ _ _ _ _ _ . _ __ _ ._ _ . ~ . _ _ _ _ . . _ _ . l l l l l l APPENDIX D Fatigue Crack Growth Evaluation I l l l l I i r r I SIR-98-096, Rev. O D-0 StructuralIntegrity Associates, Inc. I
. . . -. . - - . . - - - . . - - - - - - . ~ --
APPENDIX D Fatigue Crack Growth Evaluation D.1 Transient Definition The transients for Millstone 2 are shown in Tables D-1 and D-2 and are taken from an updated design specification for the reactor coolant system [D-1]. These are consistent with an evaluanon conducted by Combustion Engineering [D-2] to evaluate surge line stratification following issuance j of NRC Bulletin 88-11 [D-3].
' The transients are restated in Table D-3. In Table D-3, the identifier (ID) is used in this Appendix to identify each event. The " states" indicate how many stress states there are for each event. For example, heatup/cooldown contains a zero-load cold condition and an end-of-heatup hot condition. The number of occurrences (cycles) per heatup/cooldown event is derived from Tables D-1 and D-2 based upon 500 heatup/cooldown cycles for the life of the plant.
D.2 Transient Thermal Stress Evaluation Transient thermal analysis was conducted for the surge line, using the nominal pipe dimensions, to determine through-wall stresses as a result of surge line transients. This analysis was conducted with computer program PIPE-TS2 [D-4], that conducts one dimensional hear transfer and thermal stress analysis for a cylinder with specified inside and outside time histories of temperature and heat transfer coefficient. For each transient, snapshots of the through-wall axial stress distribution were evaluated for each stress peak or valley in each transient. The stress distributions were taken at the time of the maximum through-wall bending stress since this stress distribution is one which would tend to grow deeper through-wall flaws. Cubic polynomial stress fits were then determined: a = Co + Ci x + C 2 x 2+Cx 3 3 SIR-98-096, Rev. O D-1 Structural Integrity Associates, Inc
where: a = axial stress, ksi Ci = stress coefficient, ksi/in' x = distance from inside surface, inches The Co (surface stress) coefficient was then replaced with the peak surface stress that typically occurs slightly earlier in a transient to assure that a conservative representation of thermal stresses was used in the crack growth analysis. Table D-4 shows the resulting stresses, along with the associated pressure and reactor coolant system and surge line temperatures. For stratification events, the maximum potential fluid top-to-bottom AT is also shown. It was conservatively assumed that the through-wall transient thermal stresses in the ferritic nozzles would be identical to those for the stainless steel. Although the nozzle is slightly thicker, this is conservative because the much higher thermal conductivity for the ferritic material will reduce the through-wall stresses at both sides of the weld. D.3 Stratification Stress / Moments For 6 mal stratification, two effects must be considered. First, there is a global moment effect. Second, there are local stresses that develop. The stratification stresses were evaluated for CE plants in the CE Uwner's Group report (D-2]. In that report, it was stated that stratification thermal profiles could be best predicted if
- 1) the top node of the pipe model could be constrained to the pressurizer temperature for outsurges from the pressurizer (or the bottom node constrained to the hot leg for insurges into the pressurizer), and
- 2) heat transfer coefficients of 5 and 10 BTU /hr-ft2 "F were used for the hot and cold fluids, respectively.
!' SIR-98-096, Rev. O D-2 h StructuralIntegrityAssociates,Inc.
I
- To ' determine ' top-to-bottom stresses, analysis was conducted using computer program TOPBOT
[D-5], a specialized SI pmgram that computes the global moments and the local axial stresses in a pipe'with stratified conditions.
- Analysis was conducted for a AT = 200*F (653*F- 453*F) for both the top-constrained and bottom constrained temperature cases. The global moment associated with steady state conditions for both cases was 3091 in-kips or 15.455 inch-kips /*F.
i The theoretical moment for a stratified beam with a linear top-to-bottom gradient is: Z M = - x Ect AT 2 where: Z = section modulus, in' E =- modulus of elasticity, ksi .; a = coefficient of thermal expansion, in/in "F
)
This formula yields a moment of 14.889 in-kips / F for 12-inch Schedule 160 piping. Thus, the l global stratification stresses based on a linear AT = 100"F in Appendix B, should be corrected to a base AT = 96.34 F (100 x 14.889/15.455) for scaling to other AT conditions. This accounts for the fact that the temperature gradient is not linear and actually produces slightly higher global moments. The temperature difference in the piping also produces a local stress distribution that varies both through the wall and around the pipe. These are the local stratification stresses. Figure D-1 shows the fluid and pipe wall temperature distribution for a 200"F AT. Figure D-2 shows the _inside'and outside local axial stresses for the conditions in Figure D-1 with no net moment at the piping section.' Figure D-3 shows the similar stresses for the bottom-constrained conditions. In Figure D-4, the two curves are combined to produce a composite curve that conservatively L provides the highest magnitude of stress on the piping inside surface, bounding both the top- !
- . constrained and bottom constrained cases. The associated outside surface stress is also shown.
l SIR-98-096', Rev. O D-3 L Structural Integrily Associates, Inc.
l This gives a stress profile around the inside surface of the pipe for AT = 200 F that may be described by stress polynomials as: Co = Simiu ; Ci '= ""'* ~ '"* l t where: Smiu i = inside surface stress (function of 6) i Soouia = outside surface stress (function of 6)
, t = pipe wall thickness D.4 Weld Residual Stresses W-Id residual stresses can contribute to increasing crack growth rates. The weld residual stresses in stainless steel piping was taken from Reference D-6 as:
c/c i = oj (i J=0 where: oo = 1.0 ai = -6.910 c2 = 8.687 03 = -0.480 c4 = -2.027 q = x/t ci = stress magnitude at ( = 0 (inner surface)
= 30 ksi (Ref. D-6, App. A, Figure 3)
These coefficients were refit with a cubic polynomial curve that was essentially the same up to
- x/t = 0.8.
i. s l -. l SIR-98-096, Rev. O D-4 l StructuralIntegrity Associates, Inc. l i
' D.5 Nozzle / Safe End Weld Bimetallic Stresses There will be additional local stresses at the nozzle-to-safe-end welds due to bimetallic stress j effects. Stress analysis of a bimetallic transition was conducted to determined the bimetallic f-l stresses. The stresses were calculated for an A-182 F1 nozzle and an A-351 CF8M safe-end for isothermal conditions at 550 F relative to stress-free conditions at 70 F. The resulting axial stress distributions adjacent to the weld on either side of the joint were curve fit to create L polynomials for the bimetallic stresses as:
I i !- a = Co + Ci x + C 2 x 2+Cx 3 3 l with the coefficients shown in the following Table D-5. The curve fits are shown in Figures D-5 and D-6. ! The materials for the surge line are slightly different than those used in the analysis described l l above. The reference temperature for the stresses was adjusted, relative to the 550 F used in the analysis, assuming that the bimetallic stress is proportional to l c = 5(cri - cx 2) AT where: E = mean modulus of elasticity for two materials, ksi l- ai = mean coefficient of thermal expansion for each material, in/in *F AT = temperature rise from ambient To correct for the difference of material properties for the A-105 Gr. II and A-508 Cl.2 materials, a modified reference temperature may be derived as follows: I . Tg - 70_ Einaa in2 550 - 70 ~ 5,Aa, SIR-98-096, Rev. O D-5 Structural Integrity Associates, Inc.
d' where: . subscript 182 refers to A-182 F1 analysis case subscript "i" refers to the material of interest This results in adjusted reference temperatures of 582.6"F for the A-508 Class 2 and 547.3*F for the A105 Grade II material. D.6 Calculation of Stresses for Specific Events For the crack growth analysis, the stress coefficients must be calculated for each of the event states defined in Table D-3. This was completed as follows:
- 1. Table D-4 shows the throughwall stress curve fits for all transients. These were applied axisymmetrical at all weld locations for each transient.
- 2. Stratification moments are defined in Table B-4 of Appendix B. Moments for any other stratification conditions were determined based on the AT of Table D-4 by:
M = Ma,aa x AT/96.34
- 3. Local stratification thermal stresses were determined at each locaticn around the pipe
.using Figure D-4 and the AT of Table D-4. The specific inside and outside stresses (S) l were determined by i S = Sngom oa x .1T l For vertical sections of piping, there were no local stratification stresses.
- 4. Weld residual stress were assumed to be constant using the stress coefficients defined in
- Secticn D.4.
SIR-98-096, Rev. O D-6 StructuralIntegrity Associates, Inc.
- 5. ' For the bimetallic welds at each of the nozzles, the stress coefficients shown in Table D-5 '
were used, with the appropriate adjusted reference temperature. They can >e modified using the mean surge line temperature associated with each transient from Table D-4, based on the following scaling:
' Tse - 70 '
Ci = Ci.section o.5 x ( T, - 70, . where: C i, rabw o-5 = stress coefficients in Table D-5 Tst. = mean surge line temperature Ts = appropriate reference temperature discussed in Section D.5 The adjustment is made for both the stainless and the ferritic side of the weld.
- 6. The moments provided in Table B-3 of Appendix B for thermal anchor movements were
- ratioed based on the temperature of the hot leg (Tst.) from Table D-4 relative to the normal operating temperature used in the analysis:.
M = M rabi,8 3 x (Tnt -70) 4
- 7. The moments provided in Table B-2 of Appendix B for surge line thermal expansion were ratioed based on the temperature of the surge line from Table D-4 relative to the normal mean temperature used in the analysis.
(Tst - 70) M = M rabie B.2 x (628.5 - 70)
- 8. Axial forces for the thermal expansion, stratification, and thermal anchor movements are ratioed in the same manner as that for the moments.
' SIR-98-096, Rev. O D-7 StructuralIntegrity Associates, Inc.
E 1 i I ? l
- 9. The curve fits for thermal transients and other stresses were developed for the nominal pipe thickness of 1.312 inches. To conservatively assure that the same shape is l maintained at the actual specified weld thickness, the coefficients for thermal stresses, weld residual stresses, bimetallic stresses were adjusted as follows:
1 i 1 C,=C f l.312 i i i ( t s where: Ci = stress coefficient based on thickness = 1.312 inches C' i = modified stress coefficient i = exponent (0,1,2,or 3) t = specified thickness 1 1
- 10. The stresses for all moments were calculated accoucting for the angular location on the piping and the distance from the centroid to the location in the pipe.
C = (M, cos0+My sin 0)x b+ F, o
- Z ro A l
'M , cos 0 + My sin 0 ' 1 C-i Z r
( /o l C=0 C=0 3 l L where: M, = lateral moment My = venical moment . Z = section modulus ! ri = inside radius of pipe l ro = outside radius of pipe j 0 = angle from top (nominal Y direction) of pipe l F, = force in axial direction l A = cross-sectional area of pipe wall ( ! SIR-98-096, Rev. O D-8 StructuralIntegrity Associates, Inc.
For venical piping, the zero angle was taken as the -Z side of the pipe and local Mx, M y moments were used (including that in the section from Nodes 11 to 13),
- 11. Pressure stresses were constant through the wall, calculated as follows:
Pri C= o 2 2 ro - ri where: P = pressure (ksi) to = outside pipe radius ri = inside pipe radius
- 12. The individual stress coefficients wem added algebraically for use in the fracture mechanics model.
D.7 Fracture Mechanics Model The EPRI Ductile Fracture Handbook [D-7] provides a model for a circumferential semi-elliptical inside surface crack in a cylinder for 5 s R/t s 20,0.1 s a/t < 0.8, and 0.15 a/c s 1.0, where: R/t = pipe mean radius divided by wall thickness a/t = crack depth divided by wall thickness a/c = crack depth divided by crack half width at inside surface. The Mode I stress intensity factor for the deepest point on the semi-elliptical flaw is: 1 1 K i = (nt)"
- E o,(a / t)' G i
.i.o .
l 1 where e, are the coefficients of the stress polynomial describing the axial stress (c) variation ' through the cylinder wall and are defined below - SIR-98-096, Rev. 0 D-9 StructuralInte rityAssociates,Inc.
- - . - - - . _ - - ~ . . -
i l o=c o +c i
'z' 'z V
-0
' z
+o2 -
3 r t, <t, y t Here, z is the distance measured from the inner surface of the cylinder wall and t is the cylinder wall thickness. The G iparameters are the influence coefficients associated with the coefficients
~o f the stress polynomial ai and may be expressed by the following general form.
' 'R '
Gi = A i ai + A2Gi + A 3mi ' + A4 ai + A 5Gi ' + A a -5 ). it< , where:
- a=
'a' 'a" i
r t, gc, The values of Ai through A6 and m are given in Table D-6 for each G i . R is the mean radius of the cylinder. The lengths 2c and a are the flaw length measured at the cylinder inner surface and !- Jflaw depth at the deepest point of the flaw, respectively. The flaw aspect ratio (a/c) was assumed to rem'ain constant. For the nozzle locations, the radius to thickness ratio for the nozzle locations !' was slightly less than 5, but since the effect is very smalliand only affecting Go) the formulation ,. as shown was assumed to be applicable. l D.8 Fatigue Crack Growth Law I lJ For austenitic stainless steel, the crack growth law from the 1989 ASME Code. Section XI [D-8], was used in a modified form: b=CSEAK" dN where:
+io2s-"'
10b '*** **"'~"-6T , C = for "a" in inches and AK in ksi-in" h _ T. = temperature (assumed to be the mean surge line temperature for the highest of the stress states associated with AK), F L
~ SIR-98-096, Rev. O D-10 .
.. StructuralIntegrity Associates, Inc.
i
S = l' for R < 0, L
= I ~+ 1.8 R for 0 < R s 0.79,
= -43.35 + 57.97R for 0.79 < R < l.0 E =
environmental factor (assumed as 2.0 for PWR environment based on Section XI austenitic stainless steel flaw evaluation technical basis document [D-9]). Note that this term is not in ASME Section XI. R = K.io/Kmx AK = range of stress intensity factor, ksi-in u2 For ferritic steels in light water reactor environment, the 1989 Section XI [D-8] equation was used: b da '
= C S E AK"5 dN o where (for low AK values): ,
C =. 1.02 E-12 for a in inches and AK in ksi-in"2 S = 1 for 0 < R s 0.25
= 26.9 R -5.725 for 0.25 < R s 0.65
= 11.76 for R > 0.65 E = environmental factor = 1.0 da '
=CSEAK*
, , dN a where (for high AK values):
C = 'l.01 E-7 for a in inche:, and AK in ksi-in"2 S = 1 for R s 0.25
= 3.75 R +.06 for 0.25 < R s 0.65 4
I SIR-98-096, Rev. O D-i1 StructuralIntegrity Associates, Inc. i . --
= 2.5 for R > 0.65 E = environmental factor = 1.0 The result crack growth rate is then determined as:
Y da da '
- = mm of da
, or -
dN dN o dN ,2
. D.9 Crack Growth Evaluation The methodology described above was used to perform the following analysis for each weld location:
- 1. Crack growth was calculated at unique locations around the circumference at 10-degree increments, taking actual stress conditions at each circumferential location into account.
- 2. Stresses were calculated based on the previously described information for each transient and stress state. Stresses for specified conditions were ratioed from those defined at a reference temperature. Moment and local stratification stresses were distributed around the pipe circumference.
- 3. A crack growth time history was determined for each location.
The analysis was conservatively conducted based on the thickness specified for the surge line welds. A hypothetical 6:1 aspect ratio Gaw (similar to that used in Appendix G of Section XI) was assumed with a depth of 11.1 percent of the minimum wall thickness per ASME Section XI acceptance standards for piping. The analysis was terminated and maximum Daw size reported when the deepest point of the flaw reached 80c of the wall thickness (since that is the range of applicability of the fracture mechanics model). SIR-98-096, Rev. O D-12 ' StructuralIntegrity Associates, Inc.
Results from the analysis for each piping location are shown in Table D-7. This shows the relative tendency for crack growth at all locations. The critical positions for crack growth are generally near 0 and 180'since stratification stresses are greatest near these positions. Table D-8 shows the contributions to the crack growth for all of the transients. As noted, the most significant are the ST90 and ST32 stratification transients due to the large number of l defined cycles and the PUL (plant unloading) and PLD (plant loading) transients due to their relatively large numbers and high through wall thermal stresses. I l I In the alloy steel-end of the nozzle to safe-end-welds, the bending stresses reduce as compared to l the piping locatiors. The crack growth analysis was conducted to assess the time to grow an j initial flaw through the cladding and then to grow it in the base metal. Results are shown in Table D-9. This shows that the hot leg nozzle is relatively more flaw tolerant than the limiting locations in the piping. l 1 l The crack growth analysis was extended further to derive the relative number of heatup/cooldown cycles to reach 80 percent of wall depth (the limit of the fracture mechanics model) at all positions around the pipe circumference at Node 3, the location where crack growth 1 is fastest. As shown in Figure D-7, the high crack growth rates only occur near the pipe wall top I and bottom. The crack growth is compared to the allowable flaw size in the next section so the limit of 80% of pipe wall is inconsequential. I To characterize the crack growth time history, the ciack growth rate at the controlling locations is shown in Figure D-8. As expected, the crack growth rate increases with crack depth. However, the slope in v .matively earlier stages of life (e.g., less than 300 cycles) has an appreciable slope. Thus, any cracks could be detected prior to rapid growth. The crack growth analysis was further evaluated to determine the crack depth at 125 and 250 l heatup/cooldown cycles. The results are shown in Table D-10. This shows that after 125 heatup/cooldown cycles, the maximum erack depth (a/t) = 0.22. At 250 heatup/cooldown cycles. SIR-98-096, Rev. O D-13 Structural Integrity Associates, Inc.
D.10 Allowable Flaw Size Evaluation To assess the Gaw tolerance of the surge line, ASME Section XI allowable flaw sizing was conducted for the hypothetical Daw. The source equation methods of Section XI Appendix C [D-8] or Appendix H [D-10] were used. For the cast stainless welds and adjacent base metal, the Z-factors for SAW were used. For the ferritic nozzles, analyses were conducted for both EPFM and LEFM methods. For EPFM, material Category 1 from Table H-6310-2 with Ji c 21050 was used based upon properties used in this analysis. The allowable flaw sizes for the controlling locations are presented in Table D-11. In all cases, the Section XI allowable flaw sizes exceed the maximum depth cut off values permitted, and except for Node 3, exceed 80 percent of the pipe wall thickness. Thus, the surge line has significant tolerance to very deep flaws. D.11 Sensitivity to Flaw Aspect Ratio To assess the effect of potential flaws with extremely high aspect ratios (crack length-to-crack depth), crack growth analysis and critical flaw sizing was determined for some of the more critical location . A long flaw was chosen with aspect ratio 20:1. The analysis methodology was the same as previously described. The locations chosen were Nodes 2 and 13. since these ! locations exhibited the smallest number of cycles to 80-percent through wall and the pressurizer nozzle-to-safe end weld (alloy steel side). Table D-12 shows the Section XI allowable Haw sizes, including those at the hot leg nozzle. Due to the relatively longer flaws, the allowable depths are less. The crack growth analysis showed that the longer flaws grew more rapidly as shown in Tables D-13 and Figure D-9. Figure D-9 shows that the Section XI allowable Gaw depth is reached at about 60 heatup/ cooldown cycles. l SIR-98-096. Rev. O D-14 StructuralIntegrity Associates, Inc.
Table D-1 Plant Transient Events Lifetime Events Occurrences Normal Conditions
. Plant Heatup,100 F/hr 500 Plant Cooldown,100T/hr. 500 Plant Loading,5%/ Min. 15,000 Plant Unloading,5%IMin. 15,000 ,
Step Load Increase 10% 2,000 ' Step Load Decrease 10% 2,000 6 Normal Plant Variations 10 Upset Conditions Reactor Trip 400 Loss of Reactor Coolant Flow 40 Loss of Turbine Generator Load 40 Emergency Conditions Loss of Secondary Pressure 5 Test Conditions Hydrostatic Test 10 i Leak Test 200 l l l SIR-98-096, Rev. O D-15 . Structural Integrity Associates, Inc.
Table D-2 Stratification Events Pressure, Lifetime Condition Psia Tm, F A T. "F Occurrences Heatup 410 1 440 150 500 1 200 400 l 250 375 1 320 75 2250 l 653 150 500 l 200 400 l 250 375 i 320 75 Cooldown 2250 1 653 150 500 1 200 400 1 250 375 I 320 75 410 1 440 150 500 1 200 400
! 250 375 1 320 75 Hot Standby 2250 1 653 90 175.420 Normal 2250 l 653 32 2.000,000 Operation i For each stratification cycle, there is an assumed A T = 0 cycle SIR 98-096, Rev. O D-16 .
StructuralIntegrity Associates, Inc.
l l l Table D-3 Transient Identification for Each Heatup/Cooldown Cycle Cycles per ID Event States Heatup/Cooldown HUCD Heatup/Cooldown 2 1 L ST320 Hi-Temp Stratification AT = 320 F 2 0.3 ST250 Hi-Temp Stratification AT = 250 F 2 1.5 ST200 Hi-Temp Stratification AT = 200"F 2 1.6 l ST150 Hi-Temp Stratification AT = 150 F 2 2.0 ST90 Hi-Temp Stratification AT = 90*F 2 350.8 ST32 Hi-Temp Stratification A T = 32*F 2 4000 SL320 Lo-Temp Stratification AT = 320 F 2 0.3 SL250 Lo-Temp Stratification AT = 250*F 2 1.5 l SL200 Lo-Temp Stratification AT = 200 F 2 1.6 i SL150 Lo-Temp Stratification AT = 150*F 2 2.0 PUL Plant Unloading 4 30 l PLD Plant Loading 4 30 l PSD Step Down 107c 4 4 PSU Step Up 10% 4 4 PRT Reactor Trip 4 0.8 i PLFL Loss of Flow / Loss of Load 4 0.16 PLSP Loss of Secondary Pressure 2 0.01 i PNVAR Normal Variations 2 2000 HT Hydrostatic Test 2 0.02 l LT Leakage Test 2 0.4 l I f SIR-98-096, Rev. O D-17 StructuralIntegrity Associates, Inc. I
i . i I Table D-4 i Cyclic Operation Stress States ; (Page 1 of 2) ! Thennat Stress Coeffic;ents Pressurizer T. T. AT Transient Pressun:. Surge une Hotleg. Surge Une , C, Lsi C ksi/in C2. Esi4n' Ca. Esi/in' .p .p .p 4 [' 0 0 0 0 2250 598 542 0 1 IIUCD 0 0 0 0 0 70 70 0 , 19.98 -44.98 20 865 -1.7244 2240 543 543 0 . [
-3.893 20,343 -9.216 0.7032 2300 656 605 0 1 6 944 -15.889 7.203 -0.5516 2250 629 604 0 0 0 0 0 2250 629 604 0 I
6.554 -22.233 20.441 -6309 2280 601 601 0 r 0 0 0 0 2200 650 540 0 , PUL
- 19.05 -44.39 21.53 -2.21 230) 540 540 0
-153 35.1668 -16.0577 1.2829 2250 598 543 0 9.123 -20 691 9.679 -0.8331 2250 597 597 0 [
-3.796 8.2075 -3.621 0.2206 2335 650 606 0 i PSU i 6 027 -l1662 6 0929 0 4121 2250 629 6GI O ;
O O O O 2250 629 6N O i
-14.163 2260 629 604 0 t 7.21 -30 427 37Jm5 0 0 0 0 2190 650 597 0 1*SI) 7.4M -17.097 7.836 -0.643 2190 597 597 0 l
-4.28 9.712 -431 0.2918 2250 625 5% 0 0 0 0 0 2250 628 603 0 6 4 -I5.55 39.682 -30.665 7.5% 1760 616 542 0 C l'RT n 1043 -23.996 10.951 -0.873 2250 540 540 0 =
5 I g
-16.45 37.88 -17354 1.4077 2250 593 532 0 I4 14 -51.961 51 152 -17.915 2420 650 606 0 N
% 0 314 -0.5088 -5.007 2.718 1720 616 540 0 i l
2 PLit
.% 13.76 -31341 14.067 -I.0138 1720 540 540 0 l Q -16 37 37.764 -17387 1.4529 2250 593 532 0 34.29 -80.735 37.661 -3.558 155 355 355 0 l$ -4i.87 98.129 -8i.764 22.201 1700 625 505 0 8- i a
Ja Er o
~
SIR-98-096, Rev 0 D-18
' , ;{l l II l; )j!
v
- J T p 0 0 0 0 0 0 0 0 0 AT" 0 0 0 0 0 2 3
0 5 2 0 0 2 0 5 1 0 2 3 0 5 2 0 0 2 0 5 1 0 0 9 0 2 3 , p 6 0 0 0 0 0 0 0 0 0 0 0 0 3 3 3 3 3 3 3 3 3 4 , T. 10 9 0 0 2 2 9 9 4 4 9 9 3 0 3 6 5 4 7 4 7 2 2 3 0 5 5 0 5 6 6 0 1 1 1 1 2 2 3 3 4 4 4 4 5 4 5 5 6 , p 5 1 0 0 0 0 0 G 51 0 0 0 0 0 3 3 8 3 3 3 8 3 8 , 8 , T. . 3 6 1 6 0 4 7 0 4 7 8 2 4 3 4 4 4 9 4 9 5 2 5 5 5 7 5 0 s 2 , 4 4 3 4 2 4 4 6 5 6 5 6 5 6 6 e 6 , r ze
- s i r
u , 0 0 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 , y 0 , t e s s 5 3 5 1 2 1 5 1 0 5 5 1 1 4 1 4 1 4 4 8 1 4 1 4 1 4 1 4 5 2 5 2 5 2 5 2 5 2 5 2 5 2 5 2 5 2 2 5 2 a e 2 2 3 2 2 2 2 2 2 2 2 2 2 2 2 , t r P S s s e) 4 Sf r2 t Dn o 'ni
/ 9 eo2 i s O 0 0 0 0 0 0 O 0 O 0 0 0 0 0 0 0 0 0 0 0 0 O o 0 , 1 l i b t e k. -
a g a pr a C i D TeP O( i c l c s t
'ni y n /
i s C i e c L, O 0 0 0 0 0 0 O 0 u 0 0 0 0 0 0 0 0 0 0 0 0 O o 0 , i 2 f f e C o C s s e tr S l a i n m r
/
i s e O 0 0 0 0 0 0 O 0 o 0 0 0 0 0 0 0 0 0 0 0 0 O o 0 g h T L. C i s L, O 0 0 0 0 0 0 O 0 , t 0 0 0 0 0 0 0 0 0 0 0 0 O ,
, 0 ,
C. O v e R 0 5 2
'6 9
T L L S I m = s 3 T s 0 - 8 9-R 1 S ny g $e $ & a isr EDE" l i
Table D-5 Stress Coefficients for Bimetallic Stresses at Nozzle / Safe-end Welds Coefficient Stainless Steel Alloy Steel Co, ksi 6.061 -5.848 C i , ksi/in -37.934 38.492 z C2, ksi/in +34.953 -35.910 3 C3, ksi/in -2.1569 2.1292 i SIR-98-096, Rev. O D-20 StructuralIntegrity Associates, Inc.
Table D-6 Fracture Mechanics Model Coefficients Stress Distribution Ai A2 Ai A4 As A6 m Uniform: Go 1.8143 -1.9881 1.4382 -0.4680 0.056696 0.0067 0.50 Linear: Gi 1.0959 -0.9874 0.5399 -0.09303 0.0 0.0 0.38 Quadratic: 1.1836 -2.3347 2.9756 -1.7652 0.39483 0.0 0.30 G2 Cubic: G, 1.0029 -2.0160 2.5627 -1.4951 0.32759 0.0 0.25 l I SIR-98-096, Rev. O D-21 .
. StructuralIntegrity Associates, Inc.
t
Table D-7 Crack Growth at Pipe Welds Maximum Circumferential Crack Depth, Position, Location (Node) Cycles Inches Degrees Hot Leg Nozzle (2) Stainless steel weld at nozzle 500 0.678 160 Safe-end to pipe weld 500 0.631 160 3 359 0.923 180 4 363 0.920 180 5 500 0.529 180 7 500 0.490 180 8 372 0.923 180 10 366 0.921 180 11 440 0.919 180 16 418 0.919 180 13 366 0.926 180 14 500 0.766 180 Pressurizer Nozzle (15) Safe-end to pipe weld 500 0.795 180 Stainless steel weld at nozzle 500 0.808 180 Note: Results are shown for 500 heatup/cooldown cycles or for the number of heatup/ cooldown cycles for the flaw to reach 80% of the pipe local wall thickness. SLR-98-096. Rev. 0 D-22 gg,gggg,gy jgggy,jgy pggggjgggg, pgg,
. . . . . . - - . . . - - - - . - - - . - - - - ~ - - -. -. - .. - - - .
Table D-8 Transient Contributions to Crack Growth Crack Growth (inches) Node 2 Node 3 Node 10 Node 15 Transient ' (SE-Pipe) (SE-Pipe) at 160 at 180 at 180 at 180 HUCD - - - - ST320 0.002 0.008 0.007 0.004 ST250 0.005 0.019 0.017 0.010 ST200 0.003 0.010 0.009 0.006 i ST150 0.002 0.005 0.005 0.003 ST90 0.138 0.265 0.294 0.202 ST32 0.082 0.277 0.253 0.159 SL320 0.002 0.006 0.006 0.003 SL250 0.004 0.015 0.013 0.008 SL200 0.002 0.008 0.007 0.004 SL150 0.001 0.004 0.004 0.002 PSD 0.001 0.001 0.001 0.001 PSU 0.002 0.001 0.001 0.002 PUL 0.153 0.102 0.104 0.148 PLD 0.102 0.073 0.0710 0.111 PLSP - - - - PLFL - - - - PRT 0.002 0.001 0.001 0.002 PNVAR - - - - HT - - - - l LT - - - - Total 0.564 0.795 0.794 0.667 l l l k-SIR-98-096. Rev. 0 D-23 gg,gggg,g, pgggy,;yy pggggggggg, jpg, l i V
l Table D-9 Nozzle Weld Evaluation Summary l l l l Maximum ) l Node Crack Location Cycles Crack Depth, Location, i Inches Degrees l ! Through Cladding 164 0.438 340 I
- 336 In Alloy Steel 1. 340 l (500 Total) l- 246 0.407 0 )
Through Claddm.g ' 233 0.407 180 l 15 206 1.226 0 ! In Alloy Steel 226 1.227 180 (452 Total) 1.226 0 i Note: Results are shown for 500 heatup/cooldown cycles or for the number of heatup/ l cooldown cycles for the flaw to reach 80% of the pipe local wall thickness. I l l I l l l l I i SIR-98-096, Rev. 0 D-24 Structural Integrity Associates, Inc.
Table D-10 Crack Size at 125 and 250 Heatup/Cooldown Cycles Flaw Size, inches and a/t 125 Cycles 250 Cycles Location a a/t a a/t Node 2 Hot Leg Nozzle 0.358 0.212 0.498 0.295 Node 2 Hot Leg Safe-End 0.187 0.163 0.297 0.259 Node 3 0.217 0.189 0.443 0.386 Node 4 0.216 0.188 0.453 0.382 Node 5 0.184 0.160 0.285 0.248 Node 7 0.181 0.158 0.274 0.239 Node 8 0.214 0.186 0.437 0.381 Node 10 0.215 0.187 0.433 0.377 Node 11 0.204 0.178 0.371 0.323 Node 16 1 0.206 0.179 0.385 0.335 Node 13 1 0.215 0.187 0.432 0.376 Node 14 1 0.190 0.166 0.314 0.274 Node 15 PZR Safe End l 0.187 0.163 0.321 0.280 Node 15 PZR Nozzle 1 0.263 0.172 0.504 0.329 SIR-98-096, Rev. O D-25 Structural Integrity Associates, Inc.
Table D-11 Nozzle Allowable Flaw Size Summary l 1 1 Location a, inches a/t i Hot Leg Nozzle (Node 2) l (A105 Grade II) 1 LEFM 1.405 0.833 } EPFM 1.533 0.909 Section XI Cutoff 1.266 0.75 Hot Leg SE/ Pipe (Node 2) EPFM 1.032 0.899 Section XI Cutoff 0.688 0.60 Elbow Below PZR (Node 13) EPFM 0.895 0.780 Section XI Cutoff 0.688 0.60 Pressurizer Nozzle (Node 15) (A508 Class 2) LEFM 1.324 0.865 EPFM 1.456 0.951 Section XI Cutoff 1.148 0.75 I I S1R-98-096, Rev. 0 D-26 Structural Integrity Associates, Inc. i
Table D-12 Nozzle Allowable Flaw Size Summary for 20:1 Aspect Ratio Flaws Location a, inches a/t Hot Leg Nozzle (Node 2) (A105 Grade II) LEFM 1.190 0.705 EPFM (controlling) 0.863 0.511 Hot Leg SE/ Pipe (Node 2) EPFM 0.513 0.451 Elbow Ibtow PZR (Node 13) gy 0.452 0.393 Press. aer Nozzle (Node 15) (A508 Class 2) LEFM 1.120 0.731 EPFM (controlling) 0.830 0.542 SIR-98-096, Rev. O D-27 StructuralIntegrity Associates, Inc.
Table D-13 Crack Growth with 20:1 Aspect Ratio Flaw j l 1 Cycles to Exceed l Location 80 Percent of Wall Thickness { Node 3 88 Node 13 90 0 180 Node 15 Cladding 63 60 Alloy steel 46 48 Total 109 108 1 Note: All reach limit at 180* position ' I SIR-98-096, Rev. 0 D-18 StructuralIntegrity Associates, Inc.
700 6501 OOenseeeIK=eeeeeeeeeeeeene 600 - i E % ' t Vg I p - ---___ l gg. 00000GDD006G900G6DG99900D90D00000994 4X , 0 20 40 60 80 100 120 140 160 180 1 Angle from Top (degrees) I 1+ Flued + 10 -+e- 00l Figure D-1. Pipe and Fluid Temperatures for 200 F AT (Temperature Constrained at Top) I l l i SIR-98-096, Rev. 0 D-29 Structural Integrity Associates, Inc.
6
- s. ........ ........
......... \
2- - 0 ' ' 2 .
)
. ine
- : : : : l l
--+-Outsade 2: .. ..
-4........
4....... i.... . .*
...- l..
. - . i.. .... . ....-
4 ....... ........;.........:.......;........:.........(........ ..... 3 . ......:........ 6- ..-....- i....- ...I. i.... . l..
. .g . .
0 20 40 60 80 100 120 140 160 180 200 Ange %m Top, degrees Figure D-2. Local Axial Stresses for 200*F Fluid AT (Temperature Constrained at Top) ! SIR-98-096. Rev. O D-30 StructuralIntegrity Associates, Inc.
8 6 ........ ................... ........ 4 ........ . ......
. ......-l.........;........l........i... .....
3......... '. .... .
- 3. 2-
_p m I g, l l l l l l l l
.-+- Outside 2- ; . . . . . . . . .: .
. i.......... ........
4 .......: ........i.......
.........i......... ........i.... ............. -6 0 20 40 60 80 100 120 140 160 180 200 Ange from Top, degrees l
i Figure D-3. Local Axial Stresses for 200 F Fluid AT(Temperatures Constrained at Bottom) ! i i 1 i 1 i SIR-98-096, Rev. O D-31 StructuralIntegrity Associates, Inc.
8 6 ........s.......... .. ..... ..... .... . ...... . j
.r......
... .... 1 i
1 4- .........'........).........l.........l....... .... . ....... 4 .... ........ 2 .........,. ,
........,........ 4 3.. .
......t.......
3.. E. . . . . 30
- e.-Inside
--+- Outside 2 ........ ...... ......... .
i . 4 ...... .................. .... 6 .
-8 ' '
0 20 40 60 80 100 120 140 160 180 200 Ave from Top degrees Figure D-4. Combined Free-Free Stresses to Bound Top / Bottom Cases SIR-98-096, Rev. O D-32 StructuralIntegrity Associates, Inc. I
. _. . ~ . - - . . . - . . . - - . _ - _ _
S 3 . . . . . . . . . . . . . , . .. .. .......... 1 ......... .> ..........<... .......;........ ... ..... ..<............;.... ...... 1 ..... ..:...........<............;...........:.............
........)...........
3-
.......:............t.............t........................{...
.....t............
.5 - ..........*...........*.............
i............*........... ....... . .............. e-Stm W l l l l --+--Curve Fit g- . . 7 . . . . . . . . . . . ........
.g . ......... . ....... ....
11 ............ 13 15 O 0.2 0,4 0.6 0.8 1 1.2 1.4 Distance Through Wad. inches Figure D-5. Bimetallic Stresses in Ferritic Steel at 550*F I I
'1 i
J SIR-98-096, Rev. 0 D-33 gg,yggy,,; pgggy,;fy pggggjgggg, jng, . . .p.
I 14 12 - . . . . . . . . . . . .............w............
. . . . . 1
. . . . . . 1 10 -
............>............'...........4.............i............>............l..
\ . . .
- a. ............:.............;............{............)............;.............; .. .......
t, . . . I t . . . . , 6 .=.=.
?.......... .'... . . .
t*...........?.... .... ........... i i
*...t..
1 \ ,49..
........ye...........*...e........g.. .e........s........ .
...y..... . ............
-e-Stress Analyse '
l
- +-. Curve Fit 1
- g. ..... ........... ........... . . . .....
J 2- .-.......... . ...... ....-.. ...
.4 ............>........... .
-6 0 0.2 0.4 0.6 0.8 1 1.2 1.4 Distarce Through WeH. inches Figure D-6. Bimetallic Stresses in Austenitic Steel at 550*F i
\ i l l b. I I S1R-98-096, Rev. 0 D-34 StructuralIntegrity Associates, Inc.
5000 4500- . .... ....... . ,. . . . . . . . .. . ........ . r .. . - . . , . ...........
. .e . . .
4000 . . . . . . . . . . . , . ..........,........ ............. ,.......... ,.........., ..........,...........
.' g t 3500 - .' - -
- - . .. ; - 'r
-4.. . d. ;. - -
.l
' 3 coq -,..-..... .{....... ..
.l...
.......l.
.le
.-}e ....
.........-l .......
2500 ~ ........ ;...... - -- :
+
2000 - .
-{- - -
-l -
-l - -
.)- -
. . i.- --
1500 ,..........a.......
...s.
......................r. ....... .s
...a...........s............
1000 - >- *- 500 - --
- 0
- O 50 100 150 200 250 300 350 4(
Angle from Top of Pipe Figure D-7. Cycles to Grow Crack 80 Percent Through-wall
. SIR 98-096. Rev. 0 D-35 StructuralIntegrity Associates, Inc.
1.4
. 1.2
/
g1 1
/
0.6 --
~o 0.4 T '
0, _
- s I
O i 0 20 40 00 80 100 120 Nmterd Heat #CoddomCy:les j l Figure D-8. Maximum Deepest-Point Crack Growth Time History at Nodes 3,4,8,10 and 13 ' 1 l i b SIR-98-096, Rev. 0 D-36 Structural Integrity Ass 0Clates, Inc.
Appendix D References -
- 1. ABBCE Specification No. 18767-31-5, Rev.16, " Engineering Specification for a Reactor Coolant Pipe and Fittings for Northeast Utilities Service Company, Millsto ie Point Station, Unit 2," November 1997, SI File NUSCO-22Q-251p.
- 2. CEN 387-NP," Pressurizer Surge Line Flow Stratification Evaluation," ABB Combustion Engineering, Rev.1-NP, December 1991,
- 3. NRC Bulletin 88-11, " Pressurizer Surge Line Thermal Stratification," U.S. Nuclear Regulatory Commission, December 20,1988.
- 4. PIPE-TS2, Version 1.01, Structural Integrity Associates, Inc. !
l
- 5. TOPBOT, Version 3.0, Structural Integrity ' Associates, Inc.
l
- 6. _W. S. Hazelton and W. H. Koo, " Technical Report on Material Selection and Processing i Guidelines for BWR Coolant pressure Boundary Piping," NUREG-0313, Rev. 2, USNRC, I January 1988.
1 1
- 7. EPRI Report NP-6301-D, " Ductile Fracture Handbook," June 1989, 1
- 8. ASME Boiler and Pressure Vessel Code, Section XI,1989 Edition.
l
- 9. 'ASME Section XI Task Group for Piping Flaw Evaluation, ASME Code," Evaluation of Flaws in Austenitic Steel Piping," Journal of Pressure Vessel Technology, Vol.108, August 1986, pp. 352-366.
i
- 10. ASME Boiler and Pressure Vessel Code, Section XI,1992 Edition.
1 l l
- SIR-98-096 Rev. 0 D-37 f StructuralIntegrity Associates, Inc.
o - . _ . . -
. -}}