ML20203K299

Environ Factor Approach to Account for Reactor Water Effects in Light Water Reactor Pressure Vessel & Piping Fatigue Evaluations, Final Rept

ML20203K299
Person / Time
Site:	Calvert Cliffs
Issue date:	12/31/1995
From:	Gosselin S, Mehta H ELECTRIC POWER RESEARCH INSTITUTE, GENERAL ELECTRIC CO.
To:
Shared Package
ML20203K266	List: ML20203K263 ML20203K274 ML20203K299
References
EPRI-TR-105759, NUDOCS 9803050046
Download: ML20203K299 (85)

Text

Keywords: LPRI TR 105759 s ASME Code Work Order 332103 Fatigue (materials) Final Report, December 1995 Electric Power Research instttute Piping system Plant operating criteria Environmental effects An Environmental Factor Approach to Account for Reactor Water Effects in Light Water Reactor Pressure Vessel and Piping Fatigue Evaluations Prepared by GE Nudear Energy Sponsored by Elodne Power RaearchInstitute M ATERI AL SINGLE USER UCENSE AGi1EEMENT TWB B ALEIALLY DestIG A014 BENT DETWEEN YOU ANDM ELECTNC POWER RESEARCH INSTmlTE (EPN), PLIASE READ IT CARERILLY BtFDRE REMOVillo TrE WRAPP98G MATURAL TMS AGREEMENT C0emuuts OslTHE BACK COVm.

BY OPENING THIS SEALED REPORT YOU ARE AGREEING TO THE TtRMS OF THt$ AGREEMENT. lF YDU DO NOT AGREE TO THE TERMS Of THl5 AGREEMENTJROMPTLY RETURN THE UNDPENED REPORT TO EPRI AND THE PURCHASE PRICE WILL DE REFUNDED.

1, GRAlli 0F LICEIISE EPRI grras you the nonexclustw and nomransferable nght dunng the term of this pgreement to use ths report only for your own benefit and the benefit of your orgsruzaton. This means that the following may use the report (1) your company (at any site owned or operated by your company);(ll) ts subsidianes or other reisted entthes; and (111) a consultant to your company or related er'tfDes, if the C6nsultant has entered into a contract 30reein0 not to $sclose the report otitside of its organtzation or to use the report for its own benefit or the benem of any party other than your company.

This shnnk wrap lanse a0reemerit is suborenate to the terms of the Master Utilny Ucense A0reernent between most US. EPRI rnember utilrbes and EPRt. Any EPRI member utilny that does not have a Master Utility Lkense Agreement may get one on request 2.COPYIUGHT Trvs report. including the Informaton comnined in it,is owned by EPRI and is protected by United States and lntematonal copyngm isws. You msy not.wlthout the pnor wrttten permesson of EPRI, reproduce,tinstate or modrfy this report, ti arty form, in whole of in par', or prepare any denvatrve work based on this report.

3.RESTRIC110lIS 0

You may not rent, lease,Icense, $sciose or gNe this report to any person or organtZabon, or use tne informanon contained in this report, f or the benefit 01 any third party or for any purpose other than as specified above unless such use is with the pnor wrttien permission of EPRI. You agree to take all reasonable steps to prevent unauthon2nd @sclosure or use of ths report. Except as speertied abon,thrs 30reement does not Orant you any ngm to patents, copyngms, trade secrets, traoe names, trademarts or any other it".ehectual propeny, nghts of Icenses in respect of this report (continued an back cover) 9003050046 900209 PDR ADOCK 05000317 p PDR i

PRI in,a,- REPORT

SUMMARY

An Environmental Factor Approach to Account for Reactor Water Effects in Light Water Reactor

, Pressure Vessel and Piping Fatigue Evaluations This report provides a method that selectively applies an environmen-

' tal correction factor (K,) to the current Code fatigue evaluation proce-dures. It gives information on irnplementing screening criter:6 that allows stress ar,alysts to correct for reactor water effects only on applicable load set pairs. The report reviews current laboratory data and offers simplified procedures that account for environmental ef-facts in ASME Code-type fatigue evaluations in operating nuclear power plants. Also included are proposed changes to Section lll fatigue evaluation procedures for possible consideration by ASME Code Committees.

INTEREST CATEGORIES Plant life cycle management Piping, reactor vessel and BACKGROUND Fatigue usage in nuclear pressure vessel and piping compo-internals nonts due to stress cycles under reactor water conditions has been a major regulatory issue in the plant life extension process. Recently, the NRC ex.

panded its concem to current operating plants and the generic license approach for advanced light water reactors (LWRs). The NRC's interest stems from recent KEMWB test data showing ~ .at, when smooth test specimens were subjected to strain-controlled cyclic lo ads under reactor water conditions, fatigue failures occurred ASME Code earlier than the current ASME Section til S/N design curves would have pre-Fatigue (materials) dicted. In NUREG 5999, the NRC oroposed a set of interim S/N fatigue curves Pip!ng system that address reactor water effects in operating plants if implemented, the d " " " ' "' ' P '" *"' P " * " '"

Plant operating criteria Environmental effects E* ion a OBJECTIVES To review current laboratory data and develop simplified proce-dures for use in ASME Code type fatigue evaluations in operating nuclear power plants.

APPROACH Based on a review of past and current studies evaluating envi-ronmental fatigue effects in light water reactor applications, the Argonne statisti-cal model was used to develop an approach to calculate an environmental fatigue correction factor, F , for ASME Code Section Ill, NB 3600 and NB-3200-type fatigue analyses. When existing fatigue usage was multiplied by F ,

a new fatigue usage reflecting environmental effeMs was obtained. A threshoid

, criteria was used to eliminate load state pairs for which environmental correc-tion was not necessary. The parameters ire the mathematical expression for F, were determined using information generally available to stress analysts.

Mathematical expressions for F, and the approach for applications to NB 3600 and NB-3200 f atigue evaluations were developed. Materials considered were carbon and low alloy steels, stainless steels, and Alloy 600 materials. 't he proposed approach was applied to several example cases; boiling water reactor (BWR) feedwater piping, BWR recirculation piping, feedwater nozzle safe end, EPRI TR 105759s Electric Power Research Institute December 1995

and pressurized water reactor surge. A modest increase in calculated fatigue usage over that obtained using current Code procedures was noted; RESULTS This report summarizes past and current studies of the environmental fatigue effects in LWR applications. Current Argonne '

and Japanese research efforts are reviewed and an approach to calculate an environmental correction factor is described. A description of how the proposed approach can be imple.nonted in Section ill, NS-  ;

3600 and NB 3200 type fatigue evaluations is presented along with examples of applyiry the Approach to piping (NB 3600) and safe end fatigue evaluat6ons. These procedures were applied to several BWR and PWR rxample cases. The results of these case studies indicated that there is generally a modest increase in calculated fat 6gue usage, which is considerably less than the results obtained when the NUREG/

CR 5999 ourves are applied directly. A proposed ASME Code Section lli non mandatory appendix are proposed is included in the back of this report. Finally, Section 7 provides a summary.

EPRI PERSPECTfVE The NRC has noted in SECY 95 245 that, although no immediate licensee actions are pending for operating nuclear plants, the concems associated with Generic Safety issue (G8I) 166 ' Adequacy of Fatigue Life of Metal Components

• will be evaluated as part of the license renewal process. EPRl's work in developing the environmental fatigue procedures in this report, as well as applying flaw tolerance fatigue evaluation prcoedures (EPRI TR-104691) are expected to provide practical solutions to many of the issues in OSI 166.

PROJECT Work Order 332103 EPRI Project Manager: S.R. Gosselin Nuclear Power Group System and Component Integrity Technology Program Contractor: GE Nuclear Energy For further information on EPRI research programs, call EPRI Technical Information Specialists (415) 855 2411. f e

~ ,_ . . . _ _ _ ., __-

An Environmental Factor Approach to Account for Reactor Water Effects in

. Light Water Reactor Pressure Vessel and Piping Fatigue Evaluations TR 105759 Work Order 3321-03 Final Report, December 1995 Prepared by H.S. Mehta GE Nuclear Energy S.R. Gosselin Electric Power Research institute Prepared for ELECTRIC POWER RESEARCH INSTITUTE

, 3412 HitMew Avenue Palo Alto, Califomia 94304 EPRI Project Manager S.R. Gosselin System and Component integrity Technology Program Nacbar Power Group

DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES THis REPORT WAS PREPARED BY THE ORGAPlAT'ON(S) NAMED BELOW AS AN ACCOUh10F WORK SPONSORED OR COSPONSOYD BY THE ELICTRC POWER EstARCH INSTif tfit, NC (EN) NERHER EPRI. ANY ME MBER OF EPRI. ANY COSPONSOR, THE ORGAN 2AT10N(S) kAMED KLOW NOR ANY PERSON ACTNG ON BEHALF OF ANY OF THEM

( A) MAKES ANY W ARMNTY OR REPRESENT ATION WHATSOEVER. EXPRESS OR IMPUED,(l) WlTH RESPECT TO THE USE OF ANY NFORMATO4.

APPARATUS. METHOD. PROCESS OR $!MILAR ITEM DISCLOSED N TH$ EPORT, NCLUONG MERCHANTABILITY AND HTHESS FOR A

• PARTCULAR PURPOSE, OR (if) Tei SUCH USE DOES NOT NF RNSE ON OR WTERIRE WITH PRTVATELY DWNED RIGHTS. NCLUDING ANY PARTr$ NTELLICTUAL PROPERTY, OR (Ill) THAT TH$ REPORT IS Suri ABLI TO ANY PARTCULAR USER 1 CIRCUMST ANCE. OR (B) ASSUMES ANY RESPONSIBILITY FOR ANY DAMAGES OR OTHER UABILffY WHATSOEVER (IN*LUDNG ANY CONSEQUENTIAL DAMAGES, .

EVEN IF EPRI OR ANY EPRI REPRESENT ATlVE HAS BEIN ADVISED 0F THE POSSBILITY OF SLCH DAMAGES)RESULTW3 FROM YOUR SELICTION OR USE OF THts REPORT OR ANY WFORMATION, APPARATUS. METHOD, PROCESS OR SIMILAR ITEM D$ CLOSED IN TH:S REPORT.

ORGAN 2ATt0N(S) THAT PEPAED TH$ REPORT.

M toutent temyy ORDERING INFORMATION For information about ordonng this report, contact the EPRI Distribution Center,207 Coggins Drive, .

P.0, Box 23205, Plouant Hill, CA 94523, (510) 934-4212. There is no charge for reports Huested by EFRI member utilities.

Doctnc Power Resea*ch Insttute and EPRI are registeed service marts of Doctnc Power Resoae Instrtute, Inc.

Copynght C 1995 Doctnc Power Resoae Insttute, *nc. M nghts reserved.

EPAllkensedMeterial

. ACKNOWLEDGEMENTS The authors express thanks to Dr. A.G. Ware of Idaho National Engineering Laboratory and Dr. O.K. Chopra of Argonne National Laboratory for their review and contribu-tions to this report.

l e

111

EPRILkessed Material e

(

i 1

1

(

i

\

4 e

EPRIikennedMateria!

l l

. CONTENTS I

Section 'Page 1 in trod uctio n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2 Back g t o u nd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 2.1 A Brief Review of Eariier Studies on Environmental Effects.................... 2-2 2.2 US N RC, DOE, and ASM E Action s ..................... ............. ........................ 2-3 2.3 Objective of This Report ....... ..... .. ... .... ................ .... ... ...... . .. ...... ... .... .... . ... 2-3 2.4 R e f e re n ces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3 Review of Current Research Studies on Envi:onmental Effects................ 3-1 3.1 Review of Japanese Environmental Fatigue Research Results .............. 3-1 3.2 Review of Argonne Environmental Fatigue Research Results ................ 3-5 3.2.1 NUREG 5999. . . . . . . . . . . . . . . .. .. . . . . . . . . . . . . . . . . . . .. . . . . . . . . ... . . .. . . ... . . . . . . . . . . . . . .. . . . .. . 3-5 3.2.2 Sta tistica l Cha ra cteriza tion . . . . . . . . . . . . ... . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . .. . . .. . . . . . . 3-8 3.3 Evaluation of the Various Approaches................................................... 3-10 3.4 Environmental Effects Thresholds ......................................................... 3-16 3.5 S um ma ry of 6eview . . .. . . . .. . . . . . . . . ... .. .... . . . .. . . . ... .. . . . . . . .. . .. . . .. . . . .. . . . . . . . . . . . . . .. . . . . 3-19 3.6 R e f e ron ce s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-4 Proposed Fatigue Evaluation Prpoedure ........ .....................................4-1 4.1 ASME Section ill, NB 3600 and NB-3200 F=Ugg.a Analysis M eth odology . . . . . . . . .. . . . . . . . . . . . . . . . .. . . . . . .. .. . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . .. . . . . . . . . . 4 - 1 4.1.1 ASME Code NB-3600 Fatigue Analysis Method........................... 4-1 4.1.2 ASME Code NB-3200 Fatigue Analysis Method........................... 4 5 4.1.3 Summary of Code Fatigue Evaluation Approach.......................... 4-7 4.2 E nviron me ntal Fa cto r App roa ch . . .. . . . . . .. . .. . . . .. . . . . . . . ... . . . . . . . . . .. . . . . . . . . . . . .. . . . . . . . . . 4-7 4.2. 1 Ovetvie w . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 -7 4.2.2 Load State Pair Screening ............................................................ 4-9 4.3 N B 3600 An alysis . . . . . . . . . . . . ... . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 9 4.3.1 Determining Tempelature ........................................................... 4 10 4.3.2 Determination of Strain Rate....................................................... 4 10 4.3.3 Determination of Oxygen Concentration..................................... 4-13 4.3.4 Determining Corrected Fatigue Usage ....................................... 4-13 v

l

EPRIUcensedMaterial Section .

Page

4.4 NS 3200 Analysis . . . . . . . . . . . ... . . . . . ..... . .... . . . .... ..... . .. ........ . . . . ... . . . . . .. .. . . . . . . . . . . . 4- 1 4

• 4.4.1 Detennining Tempeinture ....................................................... . ... 4-14

4. 4.2 Determining Strsin Ra te . ... . .. . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . .. . . . .. . . . . . . . . . . . . . . . . . . 4 - 1 4 4.4.3 Determining Oxygen Concentration ............................................ 4 14 .

4.4.4 Determining Corrected Fatigue Usage ....................................... 4-14 4.5 Ref e ron ces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

, 5 Application Case Studies ...... . .... ... ........ .... ....... .. ..... ...... .... . .. . . .. . .. . ..... ..... . ...... 5-1 5.1 NB 3600 Application . . .. . . . .. . . . . . .. . .. . . . . . . . . . . . . . .. . . . ... . . .. . . . . . . . . . . . . . . . . . . ... . . . . . . . . . . . . ... 5-1 5.1.1 B WR Feedwater Piping..................................... ............................ 5-1 5.1.2 Recirculation System Piping ....................................................... 5-14 5.1. 3 PWR Surye Line .. . . . . . . . .. . .. . . . .. .. .. . . . . .. . . . .. . . . . . ... . .. . . . . . . . . . . . . . . . . . . . . . . . . . . .. 5-1 6 5.2 N B 3200 Application . . . . . . . .. .. . . . . .... . . . . . . . .. . . . . .. . . . .. . . . ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1 9 5.3 Ref e rences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

6 Recommended ASME Section Ill Changes .................................................. 6-1 6.1 NB3600..................................................................................................6-1 NB-3614 En vironmental Etfacts .... . . ... . ..... . . . . . .. .. .. . . . . . . .. . . . . . . . . . . . . . . . . . .. . . . . . . . . . 6 1 NB-3653.8 Consideration of Environmental Effects ...........:................... 6-1 6.2 NB3200..................................................................................................6-1 6.3 Non mandatory Appendix Overview ........................................................ 6-2 7 Summary.........................................................................................................71 Non-Mandatory Appendix X - Fatigue Evaluations including E n vironmental Etfocts . . .. . . . . .... .. . . . . . .. . .. . . ... . . .. . .. . . . ... .. ... . . . . . . ... . . .. . . . . . . .. . . . . .. . . . . . . . . .. . . . . . . .

X 1 000 soope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

X-1100 Environmental Fatigue Correction .......................................X 1 X-1200 Environmental Factor Definition ..........................................X-1 Carbon Steel . .. . ... ....... .. .... .. .... .. . . . ... . . . ... . . .. .. . . . . .... . . . . . .. .. . . . . ... . X-1 Low-alloy Steel .. . . ..... . . . . . . . . . . .. . . . .. . . . . . . . . . .. . . . . . . . . . . ... .. . . . . .. . . .. .. . . .. X-1 Stainless Stems Except 316NG ..........................................X-2 Type 316NG Stainless Steel ........... .................................... X 2 Alloy 600 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

X-1300 Evaluation Proced u res .................... ........... ......................... X 2 X 1400 Nomenclatu re . . ... . .. . . .. . . . . . .. . . . .. . . . . . . . . . . . . . .. . . . . . . . . .. . . . .. .. . . . . . . . . . . . . . . X A rticle X- 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Environmental Fatigue Threshold Considerations.............................. X 3 X 2000 Soope..................................................................................X-3 -

X-2100 Strain Range Threshold ...................................... X~4 X-2200 Strain Rate Threshold .... ............ ......................... X-4

• X-2300 Temperature Threshold ...................................... X-4 X-2400 Dissolved Oxygen Threshold ............................. X-4 vi

4 EPRILicensedMaterial Section . Page

• A rticle X 3000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Environmental Factor Evaluation ........................................................ X 5 X 3100 Scope..................................................................................X5

. X 3200 Evaluation Procedure for Design By Analysis.................... X 5 X 3210 Determination of Transformed Strain Rate......................... X 5 X 3220 Determination of Transformed Temperature...................... X 5 X 3230 Determination of Transformed DO for Carbon and L ow-alloy Steeis . . . . . . . ... . . .. . . . . . . . . . .. . . . . . . . . . .. . . . . . . . . . . ... . . . . . . . . . .. . . . . X-5 X 3240 Determination of Transformed Sulfur for Carbon and L ow-alloy Steels . . . . . . . .. . . . .. . . .. .. . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . .. . . . . . . . X-6 X 3250 Determina tion of F,,, .... . . ... . .......... . .... . . .. .... ... . .. ..... ... .. . . .... . .. . X~6 X-3260 Determination of F,,, Based on Damage Approach ............ X-6 X 3600 Evaluation Procedure for Piping.......................................... X-6 X 3610 General Requirements .... ........ ......... . . .. . ..... .. . . . ... ... .. ... ... .. .. . . X-6 X 3610 Determination of Transformed Strain Rate......................... X-6 X 3620 Galermination of Transformed Temperature...................... X-7 X-3630 Determination of Transformed DO for Carbon and Low-alloy Steels . . . . . . . .. . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . u . . . . . . . .. . . . .. . . X 7 X 3640 Determination of Transformed P:* fur for Carbon and Low-alloy Stee ls .. . . . . . . .. . . . . . . . . .. . . .. . .. . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . .. . . . . . . . X-7 X-36SO Determination of F,,, ...... . . .. .......... . ... .. ........... .. ... .. . .. .. ...... . .... X-7 X 3660 Determination of F,,, Based on Damage Approach ............ X-6 vii i

~-

I i

S I

i e

e e

(

EPRIUenndwu48 4

. FIGURES Figure Page 3-1 Pattoms of Temperature and Strain Change Used in Reference 3-8............... 3-3 3-2 Relation between fW, and Temperature (Reforence 3-8) ........................... 3-4 3-3 Effects of Dissolved Oxygen on Fatigue Life .................................$................. 3-5

$4 Proposed EAC adjusted Design Fatigue Curves in NUREG/CR 5999 for High sulfur Carbon Steels in Oxygenated Water at 200,250, and 288'O...............................................................................................................36 3-5a Proposed EAC-adjusted Design Fatigue Curve in NUREG/CR 5999 for Carbon and Low alloy Steels in Water with s 0.1 ppm Dissolved Oxygen.............................................................................................................37 '

34b Proposed EAC-adjusted Design Fatigue Curve in NUREG/CR 5999 for Low-sulfur Carbon Steels in Water with > 0.1 ppm Dissolved Oxygen.............................................................................................................3-7 34 Proposed EAC adjusted Design Failure Curves in NUREG/CR-5999 for Austenitic Stainless Steels in Water at Temperatures between 200 and320*C.........................................................................................................3-8 3-7a Predicted F, versus Dissolved Oxygen at 289'C .......................................... 3-12 3-7b Predicted F, versus Dissolved Oxygen at 250*C .......................................... 3-12 3-7c Predicted F,,, versus Dissolved Oxygen at 200*C .......................................... 3-13 3-8a Carbon Steel Cyclic Life Test Data

- Comparison and of Nakao, Argonne Statistical Mean et al., N[it at 290*C ..................................................... 3 3-Sb Comparison of Nakao et al., N Carbon Steel Cycle Life Test Data and Argonne Statictical Mean fit at 250*C ..................................................... 3-14 ix l

CPRI Lkened Matedal Figure Page I

34c Comparlson of Nakao, et al., N Cart p >on Steel Cycle Lifo Test Data '

l and Argonne Statistical Mean Ht at 200*C ..................................................... 3 1 F, 34d Comparison of Nakao, et al., N Cart >on Steel Cycle Life Test Data ,

and Argonne Stat.stical Mean 4t at 150'C ................................... ................. 3 15 3-9 Blunt Notch CT Crack initiation Test Results................................................... f1 16 3 10 Compilation of LWR type Water Environment Test Data Satisfying Any of the Independent Criteria for Moderate Environmental Effects and Comparison to ASME Mean Curve Reduced by a Factor of 4 onLife.............................................................................................................3-19 4-1 Flow Diagram for Environmental Fatigue I ....................................................... 4-8 4-2 Effect of Strain Rate Variation during Tensile and Compressive Phases of Fatig ue Cycling (Ref erence 4 2) ...............................................................,..... 4 12 51 Mathematical Model of BWR Feedwater hping System .............. ................... 5-2 5-2 Temperature Profiles of Significant Transients ................................................. 5-7 6-3 Output of One-dimensional Heat Transfer Analysis for Transient 13 .... .......... 5-9 5-4 Mathematical Model of BWR/4 Recirculation Piping System.......................... 5-15 X

D9tlUeeneedMaterial TABLES l Table Page 41 Load State Pair Screening Threshold Criteria .................................................. 4-9 51a BWR Feedwater Piping NS 3600 Fatigue Evaluation ...................................... 6-3 51b BWR Feedwater Piping Load State Data (Sheet 1 of 2)................................... 5 4 51b BWR Feedwater Piping Load State Date (Sheet 2 of 2).......... ........................ 5-5 51c BWR Feedwater Piping Load State Pair Fatigue Usage .................................. 5 6 52 Strain Rate Calculation for Significant Load State Pairs................................. 5-10 5-3 F, Calculation for Significant Load Pairs ....................................................... 5-12 54 F,, Calculation for Loe State Pair (13 40) Using Effective Da ma ge App roach . . . . . . .. . . . . . . . . . . . . . . . . . ... .. . .. . . . . . . . . .. . . . .. . . . . . . . . .. .. . . . . . . . . . . . . . ... . . . .. . . . .. . . . 5 13 5-5 CUF Results for Example Recirculation Piping System ................................. 5-16 56 CUF R ssults for Surge Une Elbow ................................................................. 5-19 5-7 CUF Results for a Feedwater Nozzle Safe End ............................................. 5-20

EPRIflanudMaterial l

e 6

4 0

1 l

1 1

I-

EPAl LicensedMaterial 1 -

. INTRODUCTION Pressum-retaining components in the light water reactor (LWR) primary systems are designed to meet the requirements of Section E of the ASME Boiler and Pressure Vessel Code or an equivalent Code. The Class I rules of Section m requim a fatigue evaluation for the transient stresses that occur during normal / upset condition opera-tion. The fatigue design curves in Section E are based on the cyclic life observed in strain-controlled fatigue tests conducted in air environment and include either a factor of two on stress or 20 on cycles over the mean curves. The effects of a high-temperature mactor water environment was not explicitly considered, although a factor of 4 in the factor of 20 on cyclic life was attributed to atmosphem.

Although there have been relatively few corrosion fatigue failures in materials typically used in LWR applications, the laboratory data generated in various test programs (e.g.,

EPRI-sponsored testing at GE, NRC-sponsored testing at Argonne, and testing con-ducted in Japan) indicate that fatigue lives shorter than the Code design values am possible, especially under low-frequency loading conditions in oxygenated water environments at elevated temperatures.

'Ihe laboratory testing identified strain rate, temperature, strain amplitude, and oxygen content as algnificant variables affecting the fatigue-initiation life.1he laboratory test-ing has generally been with one or a combination of the significant variables held at a specified fixed value during the test. However, during the course of a typical plant transient, the temperature and the strain rate generally vary continuously; but the stmss analyses are not detalled enough to evaluate the values of these variables. Fur-thermore, it is unreasonable to burden the piping or vessel stmas analysts by requiring such detailed evaluation. Therefore, there is a need for simplified, but not overly con-servative, procedures for ASME Section E, NB-3600- and NB-3200-type analyses in which reactor water environment effects need to be accounted for. The objective of this report is to review the current laboratory data and develop simpli6ed procedures for use in ASME Code-type fatigue evaluations in operating nuclear power plants.

Section 2 of the report provides the backgrouna summary of the past and current studies conducted to evaluate the emironmer.tal fatigue effects in LWR applications.

The current Argonne and Japanese research efforts are myiewed in Section 3, and the outline of an approach is described to calculate an emironmental correction factor.

How the proposed approach can be implemented in a Section m, NB-3600 and NB-3200-type fatigue evaluations is described in Section 4. Section 5 presents examples of cpplying the approach to piping (NB-3600) and safe end (NB-3200) fatigue evaluations.

1-1

_. _.m______.__._._._____..__.____ _ _ _ _ _ . . . _ . _ . - --

EPRI L%==d M:terial Introduction Section 6 proposes changes to Section III fatigue evaluations procedures that might be considered by ASME Code Committees. A proposed ASME Section III non mandatory appendix is included in the back of this report. Finally, Section 7 provides a summary.

l' i

?

1 1

0 12

EPar Liesmead Ateerdel 7

4.

. BACKGROUND Pressure retaining components in the reactor primary systems an designed to meet the requirements of Section III of the ASME Boller and Pressure Vessel Code [2-1] or an equivalent Code. in addition to prescribing stress limits for various applied loads (e.g.,

internal pressure, dead weight thermal, and seismic), the Code also requires a fatigue evaluation for the transient stresses that occur during normal / upset condition opera-tion. Specifically, the Code provider procedums for fatigue usage calculation and re-quires that cumulative fatigue usage, based on a conservative fatigue design curve, should be less than 1.0. 'Ihe Code fatigue design cunes were obtained from the results of strain-controlled fatigue tests on small specimens of austenitic and ferritic steels by applying a factor of two on stress or 20 on cyclic life to tlie mean curve. Raference 2 2 states that the factor of 20 on life is the product of the follow!ng subfactors:

Scatter of data (minimum to mean) 2.0 Size effect 2.5 -

Surface finish, atmosphere, etc. 4.0 W.E. Cooper la Reference 2-3 states that the atmosphere in the last line was intended to reflect the effects of an industrial atmosphem in comparison with an air-conditioned lab, not the effects of a speci8c coolant.

Since the introduction of Section III fa+1gue evaluation procedures, a number of studies on carbon, low-alloy, and stainless steels have shown that the fatigue life of laboratory specimens can be affected in the presence of a high-temperature water environment typically present in the reactor pressure vessels and associated piping. The amount of laboratory data has recently increased considerably with the work done at Argonne under NRC funding and in Japan under the auspices of the Thermal and Nuclear Power Engineering Society's EFD Committee. These studies indicate that the reduction in fatigue life under environmental conditions is a function of such variables as strain rate, strain range, dissolved oxygen level, temperature, and sulfur content of the steel (in the case of carbon and low-alloy steels). Both the Japanese and Argonne research are dis-cussed in detail in the next section. Wed, other than that of the Japanese and the

- Argonne meearchers, is of some intemst in terms of its historic g.yective and is, there-fore, briefly reviewed next, Also myiewed are the activities of the USNRC, DOE, and ASME, - .

=___

21

EPRILicensed Material Backgrourul l

2.1 A Brief Review of Earlier Studies on Environmental Effects One of the earliest studies on the effects of a high-temperature water environment on low-cyde fatigue performance of materials typically used for LWR primary piping and internal structures was reported in Reference 2-4, based on the work conducted by GE for the Atomic Energy Commission under the auspices of the Reactor Primary Coolant Pipe Rupture Study. Testing was conducted in a test loop installed in Comed's Dresden '

Unit I nudear power plant. Four materials were included in the testing: Type 304 and 304L stainless steels, Inconel 600, and A 516 KC-70 carbon steel. The cyclic loading frequency was four cydes per hour. This study noted a reduction in fatigue perfor-I mmce of carbon steel material in the boiling water reactor (BWR) environment. How-ever, all data fell above the ASME Section III design curve and, thus, the material was judged fully adequate for field perfomtance.

References 2 5 and 2-6 reported on the results of a combined experimental / analytical program onducted

. under the auspices of EPRI. Two types of piping carbon steels (SA 106-Gr B and SA 333-Gr 6) were studied in air and high-temr>crature water environ-ments. The study suggested several improvements in the fatigue analysis procedures:

(1) a notch factor for local strains, (2) a mean stress factor, (3) improved fatigue strength reduction factor for butt welds, and (4) an environmental correction factor. A similar approach was presented in Reference 2-7. References 2-8 and 2-9 incorporated the results of References 2-5 and 2-6 into an approach suitable for ASME Code implementa-tion.

Fatigue life of SA 106-B carbon steel in pressurized water reactor (PWR) environments has been reported by Terrel [210). One of the condusions of the Ftudy Was that al-though PWR environment fatigue life of smooth spedmens tested in the low-cyde regime do not appear to be affected by strain ratios of 0.05 and 0.50, notched specimens tested under the same conditions suggest a degradation in fatigue life as a result of positive strain ratios.

Investigations by James and coworkers [2-11] addressed the effect of environmental contaminants on the corros!on fatigue of SA 2104r A-1 carbon steel (0.014%S) boiler tubing. Although data applied to failure of fossil-fired water-tube boilers using all-volatile-treatment (AVT) or phosphate water chemistries, the temperature studied (274*C) and the nature of the results are of interest to LWR environment. A recent review of emironmentally assisted fatigue crack initiation in low-alloy steels is given in Reference 2-12.

Some investigators have generated em>ironmental fatigue S-N curves based on the hypothesis that the cycles to initiation or failure for an initially uncracked specimen might be predicted solely from the kinetics of crack propagation. The works of O'Donnel [2-13,2-14] and Coffin [2-15] are notable in that area.

2-2

EPRILAcamandM3er'el saa rroe,ar  :

1 2.2 USNRC, DOE,and ASME Actions i i The USNRC issued a Branch To hnical Position (BTP) [2-16) outlining criteria to account l for environmental effects on cyclic fatigue. The intent of this document was to provide i ,

license extension guidelines on fatigue of nuclear power plant components, h BTP  ;

! prescribes a screening test where a penalty factor of 10 is applied to the calculated ,

fatigue usage for carbon steel components in a BWR enviisignent. Essentially, this '

meant that any carbon steel component with fatigue usage greater than 0.11s not accept-able. For low-alloy ferritic steels, the penalty factor is 3.0.

Recently, Majumdar and Shack have propc::3 interim fatigue curves that are based on l the Japanese data and other data from Argonne [2-17). These proposed curves are -

l discussed in more detail in the next section. To assess the significance of interim fatigue ,

enrves, Ware, et al. [218), perfum.ed fatigue evaluations of a sample of the components l in the reactor coolant pressure boundary.

The Department of Energy (Sandia National laboratories), in cooperation with EPRI,

, published a study [219] on evaluating conservatisms and the environmental effects in ASME Code,Section III, Class 1 fatigue analysis. The study concluded that the potential increase in predicted fatigue usage due to environmental effects should be more than

offset by decreases i: oredicted fatigue usage if reanalysis were conducted to reduce the

[ conservatiams present in eFisting Component fatigue evaluations. .

As part of its effort to reexamine ASME Code fatigue curves for environmental effects, the ASME Board of Nuclear Codes and Standards (BNCS) requested the Pressure Vessel Research Committee (PVRC) to evaluate the adequacy of the fatigue curves in Sections III and XI of the ASME Code in the light of current worldwide data on emdronmental effects. The PVRC steering committee on cyclic life and environmental effects in nuclear applications has completed an initial study and submitted a draft progress report [2-20) '

to the ASME/BNCS, More recent updates on the activities of the committee are given in References 2-21 and 2 22.

2.3 Objective of This Report The laboratory testing has generally been conducted with any or a combination of these variables held at a specified fixed value during the test. However, during the course of a typical plant transient, the temperature and the strain rate generallpre continuously varying, but the stress analyses are not detailed enough to evaluate the values of these

, variables. Furthermore, it is unreasonable to burden the piping or vessel stress analysts

, by requiring such detailed evaluation. Therefore, there is a need for simp 116ed but not 4

overly conservative procedures for both NB-3600 and NB-3200 type analyses in which i

- reactor water emdronmeni effects need to be accounted for. The main objective of this report is to develop such simplified procedures.

2-3

_ _ . . . _ _ , _ . _ _ _ _ _ _ .-- _ . u ._. _ _ .. _ . . _ . _ .___ _ _ _ _ _

Ern!Uemand Material .

B&dgic v.s 2.4 References .

[21) " Rules for Construction of Nuclear Power Plant Components."Section III, Division 1.1992 ASME Bouer and Pressure Vessel Code, American Society of 14chanical Engineers. .

[2-2) " Tentative Structural Design Basis for Reactor Pressure Vessels and Directly  !

- Associated Consponents (Pressurized, Water Cooled Systems)." PB 151987, -

U.S. Dept. of Commerce, Of8ce of Technical Services 1 Dec.1958 Revisiem. l

[2-3) W.E. Cooper. "The Initial Scope v.d Intent of the Section III Fatigue Design Procedures," Prepared for the Pa essure Vessel Research Councu (PVRC) Work-shop on Environmental Effects on Fatigue Performance, Clearwater Beach, FL January 20,1992.

[2-4) D.A. Hale, et al., " low Cycle Fatigue of Commercial Piping Steels in a BWR Primary Water Environment." Journal of Engineering Materials and Technology, Trans, of ASME Vol.103, pp.15 25, January 1981. >

[25) BWR Environmental Crocking Marginsfor Carbon Steel Piping, EPRI Report No.

NP-2406. Prepared by GE,1982.

[2-6) S. Ranganath, J.N. Kass, and J.D. Heald. " Fatigue Behavior of Carbon Steel Components in High-Temperature Water Environments." ASTM STP 770, pp.

436-459,1982.

[2-7) K. Bieniussa and H. Schulz. " Protection Against Fatigue Damage with Respect .

to the Environmental Influence of LWR Operating Conditions." Nuclear Engi-narring & Design, No. 94, pp. 317 324 (1986).

[2-8) HS. Mehta, S. Ranganath, and D. Y *F.nstein. Application of Environmental

- Fatigue Stress Rules to Carbon Steel Re for Piping. EPRI Report No. NP-4644, July 1986.

[29] H.S. Mehta and S. Ranganath. "An Environmental Fatigue Stress Rule for Carbon Steel Reactor Piping." ASME Volume PVP-241 (1992),

[2-10] J.B. Terrel. " Fatigue Life Response of ASME SA 106 B Steel in Pressurized Water Reactor Environments." International Journal of Pressure Vessels & Piping.

Vol. 39, pp. 345-374 (1989).

[2-11] B.A. James, LD Paul, and M.T. Miglin. "Iow Cycle Fatigue Crack initiation in  ;

l- SA210A1 Carbon Steel Bouer Tubing in Contaminated Bouer Water." PVP Vol. .

l 195, Fatigue Degradation and Fracture.1990, pp 13-19.

[2-12] F.P. Ford, S. Ranganath, and D. Weinstein. Environmentally Assisted Fatigue

Crack initiation in lere alloy Steels. EPRI Report No. TR-102765. August 1993.

E . _ _ _ .__ _ _ . _ _

_ - - - - = -

EPRIUamead Materdal

Background

[2 13) W.J. O'Donnel, et al. " Synthesis of S/N and da/dN Life Evaluation Technolo.

gies," PVP Conference, Pittsburgh 1988. ASME Publication 88-PVP-10.

[2 14) T.P. O'Donnel e.nd W.J. O'Donnel. " Cyclic Rate-dr#ent Fatigue Life in Reactor Water." ASME PVP-Volume 306 (1995).

[2-15] T.A. Prater and LF. Coffin. Trans. of ASME. Journal of Eng. Materials and Tech-nology 108,2,1986.

[2-16) " Fatigue Evaluation Procedures." Branch Technical Position PDLR D-1. U.S.

Nudear Regulatory Commission.

[2-17] S. Majumdar, O.K. Chopra, and W.J. Shack. " Interim Fatigue Design Curves 7 for Carbon, Low-alloy, and Austenitic Stainless Steels in LWR Environments."

NUREG/CR-5999. April 1993.

[2-18) A.G. Ware, D.K. Morton, and M.E. Nitzel, " Application of NUREG/CR 5999 Interim Fatigue Curves to Selected Nuclear Power Plant Components."

NUREG/CR-6260. March 1995.

[2-19) A.F. Deardorf and J.K. Smith. " Evaluation of Conservatisms and Environmen-tal Effects in ASME Code Section III, Class 1 Fatigue Analysis." Sandia Report No. SAND 94-0187. May 1994.

[2 20] "PVRC Progress Report on Evaluation of Fatigue Curves in Section III and XI in the Light of Current Worldwide Data." Presented to the ASME BNCS Op-erations. September 29,1992, San Diego, CA.

[221) W.A. Van Der Sluys. " Evaluation of the Available Data on the Effect of the .

Environment on the Low Cycle Fatigue Properties in Light Water Reactor Environments." Presented at the Sixth International Symposium on Environ-mental Degradation in Nuclear Power Systems - Water Reacton, TMS/NACE.

i Aug.15,1993, San Diego, CA.

[2-22] W.A. Van Der Sluys and S. Yukawa. " Status of PVRC Evaluation of LWR Coolant Environmental Effects on the S N Fatigue Properties of Pressure Boundary Materials." ASME PVP-Volume 306 (1995).

a

( ,

2-5

EPRI L.icensed Material e

4 e

C

EPRILicensed M:terial 3

REVIEW OF CURRENT RESEARCH STUDIES ON ENVIRONMENTAL EFFECTS The two major experimental efforts to characterize the effect of LWR environment on the fatigue-initiation lives of materials commonly used in reactor primary pressure boundary, are by the Japan EFD committee and the NRC-sponsored work at Argonne National laborato y. The researchers associated with both of these efforts have also proposed analytical approaches to incorporate the environmental effects into the ASME Code type fatigue analyses. The proposed approach in this report for Code fatigue evaluations considering environmental effects drew upon the best features of these two efforts. Each of these efforts are summari.'.ed in this section. ,

3.1 Poview of Japanese Environmental Fatigue Research Results One of the earliest Japanese studies investigating environmenial effects on. fatigue-initiation life of carbon and low-alloy steels in oxygen-containing high-temperature water was conducted by Higuchi and lida [3-1 through 3-5]. An additional study on low-alloy steel was reported by Nagata, et al. [3-6]. The materials covered in Refenmce 3-5 were: (a) carbon steel pipe of specification ST542 in JIS G 3455, equivalent to ASME SA333 Gr.6, and (b) forged low-alloy sted of specification SFV3 in JIS G 3212-1977, equivalent to ASME SA508 C1.3. The dissolved oxygen content ranged from 0.01 to 20 ppm. All of the tests were push-pull-type tests similar to those used in generating the original ASME Code fatigue curves. The initiation life in terms of cycles, N, was de-fined as the number of cycles to a 25% drop in tensile peak stress (at the maximum tensile strain in the hysteresis loop) from the maximum value in the characteristic cyclic curve of tensile stress in a test. Based on the test results, Higuchi and lida suggested the following relationship:

N ,=Nm(i 7 F (Eq. 3-1) where N, = Fatigue life (cycles) in water at temperature Nm = Fatigue life (cycles) in air at room temperatu c ir = Strain rate during the rising phase of testing (%/sec)

P = Strain rate exponent dependent upon temperature and dissolved oxygen content 3-1

EPR1 Licensed Material Revsew of Current Research Studies on Environmental Effects Equation 3-1 can be rewritten in the following form:

F,,, = m N /N,= (6 )7 -P (Eq.3-2)

The factor F,,, can then be understood as an environmental correction in terms of cycles.

Thus, a partial fatigue usage for a load state pair based on an air fatigue curve in an ASME Code fatigue enalysis could be multiplied by F,,,to obtain the fatigue usage with '

the environmental effects factored in.

' ?.guchi and lida also defined a fatigue strength correction factor K,,, as the following:

K,,, = 2 + l(t )P8 r - 2](1-C/c,,) (Eq. 3-3) where B = -0.472 for carbon steels,-0.568 for low-alloy steels P = 0.1 + MN M = Factor to account for dissolved oxygen (DO) content

= 0.0 DO s v.1 ppm

= (DO-0.1)/0.1) 0.1 < DO < 0.2 ppm

= 1.0 DO 2 0.2 ppm N = Factor to account fo: mmperature (T)

= 0.2T/100 T < 100*C (212*F)

= 0.2 100*C (212*F) s T s 200 C (392*F)

= 0.2 + 0.4(T-200)/100 T > 200*C (392'F)

C = 0.00108 for carbon steels,0.00140 for low-alloy steels e, = Applied strain amplitude The factor K,,,is a multiplier to be applied to the ca2culated value of alternating stress amplitude. Thus,if one were to use the Higuchi-lida approach to account for the envi-ronmental effects in an ASME Code fatigue evaluation, the alterr.ating stress amplitude S, would be multiplied by this factor prior to entering the fatigue cun e for determining the allowable number of cycles.

The following observations are made regarding the Higuchi-lida model:

• Environmental effects are assumed to be significant down to 100*C (212*F). Recent data from Argonne and others indicate that the emironmental effects are insignifi-cant below 150 C [302*F] and that this threshold temperature might be as high as 200'C (392*F).

3-2

EPRI Lkensed Material Retnew of Current Research Studies on Environmental Effects The fatigue damage is assumed to saturate at the dissolved oxygen level of 0.2 ppm.

l This might have been due to the fact that a large number of tests were conducted at 8 ppm oxygen level and very few tests at 0.2 ppm level.

I

• There is a threshold strain amplitude level below which there is no emironmental fatigue damage. This threshold strain amplitude level is equal to th anstant C in Equation 3-3.

Most of the laboratory data has been generated at constant temperature and strain rate values, in contrast, plant components normally undergo varying temperatures and strain rates during plant operation. Accordingly, a number of experimental studies were performed by the Japanese researchers to determine the effective values of strain rate and temperature when these parameters are changing during the test [3-7 through 3-9). The effective values are defmed using the improved rate approach first developed by Asada [3-10). An application of the effective damage parameters to LWR plant com-ponents is described in Reference 3-11.

Strain p Strain p 0.6 -

P 0.6 -

P A 6 N /

6

@  !$ \ /

50 3 0

\

f 3 E

5 /

/ I Temperature s s &E5 \

\

/ &

/ 5

-0.6 s

m -0.6

\/ gg m b

• b (1) In phase (2) Out of phase Figure 3-1 Patterns of Temperature and Strain Change Used in Reterence 3-8.

The results from one of the important set of tests are described in Reference 3-8. In these tests, the temperature and strain rates were varied in-phase and out-of-phase as shown in Figure 3-1. The test results showed that the cyclic fatigue life under the in-phase temperature change was almost equivalent to that under the out-of-phase condition.

The authors also concluded that fatigue life under chan-ing temperature conditions could be predicted by the improved-rate method with A.gue lives at constant tempera-ture. A fatigue life reduction factor Fen for changing temperature, oxygen content, and strain rate condition, was defined as: .

F,,, = 1 + j [(F,-1)/(c,- e,,,,,,)]de (Eq. 3-4) 3-3 l

l EPRI Licensed Material l Remew of Current Research Studies on Environmental Effects where F,,, is as defined in Equation 3-2, c is minimum strain and c_ is maximum strain in a strain cycle. Altemately, the cyclic life, N', under varying temperature conditions was defined as:

11/N',1 = ll11Nw)[1RT_ - TJ]dT (iq. 3-5) -

where T_ and T_ are maximum and minimum temperatures, respectively. Using -

these concepts, the authors could establish a good correlation between predictions and actual leak cycles. Figure 3 2 shows a plot of the 1/N, as a function of test tempera-ture. The data appears to support a threshold temperature of 180*C for the environmen-tal effects. A concept similar to the preceding ones (such as F',,, and N',) is developed later in this report and is used to calculate an effective value of factor F,,,,

1 x 10 "

- f 1 1 l = 4.68 x 10-5 T-7.422 x 10 3 I( NLEAK J

~

5 x 10-1 -

~

z f' 1 i

- l = 1 x 10-3

) -

kNLEAK J e' ' '

0 O 100 200 300 Temperature T (*C)

Figure 3-2 Relationbetween 1/Nu and Temperature (Reference 3-8).

Other important information generated from the latest Japanese tests is the effect of oxygen concentration on environmental fatigue life. Figure 3-3 from Reference 3-8 shows fatigue life N, as a function of dissolved oxygen content. The data indicates that the reduction in fatigue life is gradual, from 0.1 ppm to 8 ppm oxygen level. This is in contrast to the Higuchi-lida modelin which the predicted fatigue life reduction reaches a maximum at 0.2 ppm and then does not change. Further discussion appears ,

later in this section.

EPR1 Licensed Material Retnew of Current Research Studies on Environmental Effects Temperature 290*C

! l 0

- I , 250'C g 1,000 - - - - - - - --- - - - - --

h

~

I + t l- +

!  ! ,, 150*C f -

I ## I oJ O

,g

$ g  ! 100*C

"'I E

~

l hO l l i l Oi 100 ;- Carbon Steel (STS410) J . _ . _ . _ ._ _ _ . L. . _

. ca= 0.6% i T= 0.004*/o/s  !

I l

, , . . . . , , , , ,,,,e . . . . . . .

0.01 0.1 1 10 Dissolved Oxygen Content (ppm)

Figure 3-3 Effects of Dissolved Oxygen on Fatigue Life.

3.2 Review of Argonne Environmental Fatigue Research Results Under NRC funding, extensive testing has been conducted at Argonne on the fatigue-initiation life of carbon, low-alloy, austenitic stainless steels, and Alloy 600 in LWR environments [3-12 through 3-15).

3.2.1 NUREG-5999 In NUREG-5999, the Argonne team developed interim fatigue curves based on a modi-fication of the Higuchi-lida equation:

N, = N,&CT)(t)'  %. 3 6)

~

where 4(T) = 0.6e'N+2m

. T = Temperature in Celsius 3-5 j 1

1 l

EPRI Licensed MCterial Remew of Current Research Studies on Environmental Effects The values for strain exponent P for the high-sulfur (s > 0.008% by weight) case were slightly different from those given by Higuchi lida. For the oxygenated nuter case, the oxygen level was assumed to be 0.2 ppm. Figures 3-4 through 3-6 show the interim curves for various conditions and materials.

Equation 3-6 can be recast in terms of F, as the fobowing:

F, = N)N, = M+ (T) (t)'] (Eq. 3-7) 1000 . . . . 1000

• +

: Fig.i1-9. ASME Code- ,,

j j pig.l19.1 ASSE Code' g

gg j l- -0.1W s'}4  :

j .N l- -0.1% s'1

- *C.01 % s d :

- l"*"C.0 Ms .

,.g  : SatQration : _

g\, s. l

• ""SetQration  ::

.N.;s .

.  :  :  : o  : .

im .

..;.................q.........

g ,, . N N:

. .........p.........

......gx.g$  ; ,,.:

,' .. 'sg[' ' s:'. .:.

8 i l' l. s. . i  ! -

J  ! i i i i $::s '

i '

i. C. x :.

i .

'iDJ.$+.t.-[::-

2 ! N...: n i-  !-  ! .  :  :

io 3, 101 102 103 104 1@ 10s 3oi 3o2 3os 3o4 3p ios Cycles Cycles High Sulphur Stseis 0,>0.1 ppm High Sulphur Steels 0,>0.1 ppm T = 200*C T = 250*C 1M0 . . . . .

! i Fig.i19.1 ASME Code'

- l  :- -0.1W s d d  :

g  : -* 0.01%* s .

! j"*"SatQrabon i em \,s N g.:  :

6 ig g... ..k as i'. i i  :

• : .' N s..,:'s, N ;:  :  :

l

.,*\ IN s N :

. . *l..,

• 'd ,'N . 4  %

I h

• to 101 102 103 1o4 ip jos Cycles High Sulphur Steels 02>0.1 ppm T = 288'C Figure 3-4 .

Proposed EAC-adiusted Des gn Fatigue Curves in NUREG/CR-5999 for High-sulfur i

Carbon Steels in Oxygenated Water at 200,250, and 2B8 C.

3-6

EPR1 Licmesed Material Revsew of Current Research Studies on Environmental Effects 1000 ; . , , ,

. Fig.19.:1 (ASME Code. :

)N i. i Sechon)ll. Apped,$x !)

- NN :  :

m  :

'A N  :  :

5,, goo '

~

.........j.N.4. ..

.........f...................

s:.s  !.  !

-  :  : s N t  : .

: h. s

: ' s :' .

:  : T

:- l  :

101 102 103 104 106 3ce Cycles Dissolved 02>0.1 ppm Fipm >h ,

Proposed EAC-adjusted Design Fatigue Curve in NUREG/CR-5999 for Carbon and Low-alloy Steels in Water with s 0.1 ppm Dissolved Oxygen.

1000 -

s

, j  : Fig.1-91 (ASME Code. :

's s  :

i Section)ll. Apped&x I) '

m g 's y '

3,,,3ao . . . .g q . ' .s, ...!..........!..........!..........

}s \ .:* i i  !

l N  %  :  : .

N 's, 4 '

N s%%,  ; .

j l  % : ' s' ..

to

! 't 101 102 3o3 304 305 306 Cycles Low Sulphur. Low Strength Dissolved 02>0.1 ppm Figure 3-5b Proposed EAC-adjusted Design Fatigue Curve in NUREG/CR-5999 for Low-sulfur Carbon Steels in Water with > 0.1 ppm Dissolved Oxygen.

6 3-7

\

EPRILicmsed MMal i Retnew of Current Research Studies on Environmental Effects 1000 -

s  :  :  : . 4

-s  :  : .

- s i i A5ME Code: .

l s

s!

~

s i Section,lli,i Apperidix 1, -  !

'i.s s .

Fig.1 -9.2.1 !.

-  : s  :  :  :

c i s i i m  : i. .

h jgQ ................k..........s....).g............ ..............b..............

m i i  ! 's s  !  !  :

rn - i s 4. s i i N i -

N's ,E'

- i i i i i i -

i i i i i i  ! i 10 "a t =="a? "a! ">=* "a 101 103 105 Cycles Proposed EAC-Adjusted Fatigue Curve in LWR Water Component Margins 1.5 & 20 Figure 3-6 Proposed EAC-adjusted Design Failure Curves in NUREG/CR-5999 for Austenitic Stainless Steels in Water at Temperatures between 200 and 320*C.

3.2.2 StatisticalCturectorization Following the NUREG-5999 work, the Argonne researchers presented a statistical analysis of existing fatigue S-N data, both foreign and domestic, for carbon steel and low-alloy steel, austenitic stainlesa steels, and Alloy 600 [3-15). The statistical model considers the effects of various rnaterial, loading, and environmental conditions on fatigue-initiation life of these materials. The expressions are the following:

Carbon and Low-alloy Steels in(N) = (6.667-0.7661) - (0.097-0.3821)l, + 0.52F'[x]

-(1.687+0.1841,)!n (e,-0.15+0.041,+0.026P'[1 -x]) (Eq. 3-8)

-0.00133T(1-1) + 0.554S*T*O*t*

3-8

L*RI Licensed M:terial Remew of Current Research Studies on Environmental Effects where

_ N, is the fatigue life defined as the number of cycles for the peak tensile stress to drop 25% from its initial value (consistent with Higuchi-lida definition) c, is the applied strain amplitudein %

P8/x] is the inverse of the standard normal cumulative distribution function for the xth percentile of probability T is the test temperature in *C 1, is = 1 for water and = 0 for air environment I, is = 1 for carbon steel and = 0 for low-alloy steel S*, T*, O* and t* are transformed sulfur content temperature, DO, and strain rate, respectively, defined as follows:

S* = S (0<S<0.015 wt%)

S* = 0.015 (S>0.015 wt%) -

T* = 0 (T< 150%)

T* = T-150 (T> 150%)

O* = 0 (DO<0.05 ppm)

O* = DO (0.05 ppm < DO < 0.5 ppm)

O'=0.5 (DO > 0.5 ppm) t' = 0 (t > 1%/sec) t* = In(t) (0.001 s t S 1%/sec) t* = ln(0.001) {t < 0.001%/sec)

Austenitic Stainless Steels in(N,) = 6.69 + 0.52P1/x] - 1.981n(c,-0.1225+0.016P'll-x]) (Eq. 3-9)

+ 0.3821,,

3 + 1,(0.134 t * - 0.359) where 1,is33

= 1 for Type 316NG stainless steels and = 0 otherwise. All other terms are as defined in Equation 3-7.

Alloy 600 In(N,) = 6.94 + 0.42P'lx]- 1.7761n(e,-0.12+0.021P'll-x1) (Eq. 3-10)

+ 0.498I,- 0.4011, where I,is = 0 for <150*C and is = 1 for 150-320*C. All other terms are as defmed in Equation 3-7.

3-9

EPR1 Licensed M:terial i Pemew of Current Research Studies on Environmental Effects f

3.3 Evaluation of the Various Approaches The Higuchi-Ilda approach and the NUREG/CR-5999 approach presented earlier are it, the form of an environmental correction factor on the cydes, which is convenient for use in the ASME Code fatigue evaluations. Therefore, the Argonne statistical equations .

were recast in terms of an environmental damage factor on cydes F, (as defined in Equation 3-2): ,

Carbon Steel F,,, = Nm /N 3 , = ap (+0.384 - 0.001337- 0.554S*T*O'6') (Eq. 3-11)

Low-alloy Steel F,,, = Nm/Nu, = exp (+0.766 - 0.00133T - 0.554S*T*O' t') (Eq. 3-12)

Stainless Steels Except 316NG F,, = N m/N3 , = exp (+0.359 - 0.134 s*) (Eq. 3-13)

Type 316NG Stainless Steel F,,, = Nm/Nuw = e2P (-0.023 - 0.134 i') (Eq. 3-14)

Alloy 600 F, = Nm/N3 , = exp (0.401)

= 1.49 (Eq. 3-15)

Note that the ratio in the preceding equations is between the at-temperature air cydes to at-temperature water environment cydes. In the Higuchi-lida case and in the modi-fled Higuchi-lida equations used in NUREG/CR-5999, the ratio was defined between the room temperature air cydes and the at 'emperature water environment cydes.

Some observations regarding the preceding expressions:

• Both the carbon and low-alloy steel factors have a temperature dependenc, . This comes from the fact that the air fatigue curves in the Argonne database have tem-perature dependency.

• The Alloy 600 factor is a constant, implying that no matter what the environmental parameters are, there is always a reduction in fatigue cyclic life in the LWR environ-ment. .

a The probability terms cancel out when the same percentile probability levels are used for both the air and water environments. .

3-10

EPRI r "==d Material Remew of Current Research Studies on Environmental Effects o The presence of a constant term in the exponent of carbon and low-alloy steels means that even if any of the O*,S',T* or t* terms are zero (i.e., any of these param-eters satisfy a threshold criteria value), the calculated value of F,,, is still greater than 1.0, implying some residual environmental effect. A similar conclusion also applies for stainless steels.

, A krey output from the environmental fatigue testing li, generally a predicted or implied value of F,,, (or, equivalently, K,,,) as a function of a set of environmental variables such -

as the sulfur content, DO, strain rate, and temperature. The preceding review of the Japanese and the Argonne research results indicates that there are basically three dis-tinct approach to calculating the environmental fatigue correction factor, F,,,. The first one is that proposed by Higuchi-lida, as shown in Equation 3-2. The A jonne-modifica-tion of the Higuchi-lida proposal (Equation 3-7) represents the second approach, which formed the basis of the NUREG/CR-5999 interim fatigue curves. Equations 3-11 through 3-14 are derived from the statistical characterizations in NUR2G/CR-6335 and represent the third approach. It is instructive to compare the F, values predicted by the three approaches.-

Figures 3-7(a) through (c) show a comparison ohe predicted Fn values for carbon steels using the three approaches. The pred' ( values at three constant temperatures (289*C,250*C, and 200*C) are plotted - tion of dissolved oxygen content. The assumed strain rate was 0.001% v t and the sulfur content was ass'umed as 0.015% by weight. As expected, h _, approaches yield signiScant differences in the predicted values between the DO levels of 0.1 and 0.5 ppm. In the first two approaches, DO effects are accounted for between 0.1 to 0.2 ppm; after which the DO effect is as-sumed to saturate at 0.2 ppm. In contrast, the Argonne statistical fit (the third approach) incorporates the DO effect gradually from 0.05 ppm to 0.5 ppm. The differences in the predicted F,,, values are especially significant at 0.2 ppm DO, the nominal DO level for BWRs operating with normal water chemistry (NWC). For example, at 289*C (Figure 3-

+ 7a), the Higuchi-lida approach predicts a value of 92.9 and the Argonne modification of the Higuchi-Ilda approach (the second approach) predicts a lower value of 41.9. On the other hand, the Argonne statistical St predicts : value of 4.93 only. To assess whether the predicted trend of the Argonne statistical model with respect to the DO levelis realistic, its predictions of cyclic life were compared with the experimentally obtained cyclic life where the DO level was systematically varied from 0.05 ppm to 8.0 ppm. ,

Blunt notch cyclic fatigue initiation test data reyoned in Reference 2-5 was also exam-ined.

0 3-11

EPRI Licenud Material Review of Current Research Studies on Environmental Effects I == ,__

_. a m a m m e,xan _____

eu nnSaoc, _____

e coin *

. . .. . Asgerve uman R $290 C. -----

p9 e coin sono i l'!'% ' -a I i i iii

_m -- 0404 West s

} q -

.s .

I

- X_

g .x

\

}

i.. . __ . ______ ___ _____ .__. __

w I 0010 E100 is tad (D (N) h Ouygen Cenesse (ppen)

Figure 3-Ba Comparison of Nakao, et al., N3Carbon Steel Cyclic Life Test Data and'Argonne Statistical Mean Fit at 290*C.

mo

(

a ===omoc .___.

Aegame usan n 3250c I i1 I cn cao<w e

_ oma f k ..,

\I I.

g i.. _.

s s

}

\ '

T

\ i.

m 0 410 0 100 1400 10 2 (N) Deseehod Onygen Cators (pprn)

• Figure 3-8b Comparison of Nakao et al., N3Carbon Steel Cycle Life Test Data and Argonne Statistical Mean Fit at 250 C.

3-14

EPRI Licensed M:terial Retnew of Current Research Studies mt Environmental Effects icooo e Nuno,ea330)C Arprew Mem Pe g200 C -----

I I II aM%

9 s 4 111MiWenc

^ 012%

l

===."

N "

I g $o noII * -\

h ico 0W 11100 10SX) o=*=d o yea c.== w Figure 3-8c Comparison of Nakao, et al., NuCarbon Steel Cycle Life Test Data and Argonne Statistical Mean Fit at 200 C, .

mo

% w eeimc e y==oisoc a e45%

a 4 0049Weme

~

m nsw Im ll ll 11 8

Cio 12) o==*=d ours= comm e.=>

Figure 3-8d Ccmparison of Nakao, et al., N 3Carbon Steel Cycle Life Test Data c.nd Argonne Statistical Mean Fit at 150 C.

4 3-15

EPRI Licensed Mcterial Retnew of Current Research Studies on Environmental Effects l Figure 3-9 shows the cyclic fatigue-initiation life results for blunt notch test specimens at various DO levels. The data on the extreme left, although plotted as with 0.01 ppm DO, is, in fact, in the air environment. The material was SA 333 Gr. 6 carbon steel and the test temperature was 550*F. The initiation was defined as a crack growth of 0.016 inch.

A review of Figure 3-9 clearly indicau s that there is a significant difference lietween the cyc11e life at 0.2 ppm DO and that at 8 ppm DO.

10ft10 ,

I

~~'**- . . . . ,

,, s 2 1000 -

t

~

" Z Notch Radius =0.002 in -

Temp =550 F - -

[

_ delta K=20 ksl*in^(1/2) '..

Material: SA 333 Gr.6 '-

Testing Frequency: 1.25 epm , ,

sao

$r)

Figure 3-9 Blunt Notch CT Crack Initiation Test Results.

From the preceding review of cyclic fatigue life data,it can be concluded that the pre-dicted trend in cyclic life reduction as a function of DO by the Argonne statistical model is far more realistic compared to the other two models. Therefore, Equations 3-11 through 3-14 were used in this report for application to ASME Code fatigue evaluations, as described in Section 4.

3.4 Environmental Effects Thresholds The information in this subsection is based on the work reported by Van Der Sluys and Yukawa [2-21 and 2-22] as a part of the PVRC effort.

3-16

EPRIUcensed Material Remew of Current Research Studies on Environmental Effects Although the detrimental effect on S N life can be large for the worst combinations of environmental parameters, these worst-case combinations generally are not typical of LWR operating conditions, The cardination of very low strain rates and relatively large strain ranges that result in large enviremnental effects do not seem to be typical of

- events in operating plants. In addition, the high oxygen levels at which much of the data have been obtained are above the levels typical of BWR plants. Therefore, one of

, the tasks in the PVRC activity consisted of defining a tentative set of criterion values for test and material parameters where the envig,rer. ental effects would be expected to be moderate or acceptable. This required quantifying moderate or acceptable environmental effects with respect to the air environment data used in developing the ASME Code ,

fatigue design curves Recalling that the analysis of the collected air environment test data indicated a factor of about four for temperature and data scatter effects, a factor of four on the ASME mean life was chosen as a working de6nition of moderate or accept-eble water environment effect.'

Based on the examination of the duabase, Van Der Sluys and Yukawa determined that values of independent parameters listed below should result in only a m'oderate detri-mental effect on cyna life of carbon and low-alloy steels.

Parameter Rangt Strain amplitude s 0.1 % -

Strain rate 2 0.1%/sec Oxygen content s 0.1 ppm Temperature 2150*C or 300*F Sulfur content < 0.003%

Fluid velocity > 10 ft/sec or 3 m/sec Note that independent means that only one criterion needs to be satisfied, regardless of the values of the other parameters. It has been observed that, to have e, large effect of the environment on the S-N fatigue life, a critical combination of conditions is necessary. If

. any one of the condtions is missing, the Cieet of the environment on fatigue life will be moderate, For example,if the strain rate is greater than 0.1% per second, only a moder-ate environmental effect is expected, even if the dissolved oxygen is high, the tempera-

. ture is 288'C, and the material has a high sulfur content.

- Reference 2-22 presented a plot, shown here as Figure 310, which demonstrated the validity of the values derived for each of the inaependent criterion for moderate envi-ronmental effects for carbon and low-alloy steels. From Figure 3-10, it is seen that a factor of four on the ASME mean curve encompasses a large portion of the data for tests that meet any one of the independent criterion value. Another consistency check of the criterion values can be made by using the Argonne statistical model. From Figure 3-7a, it is seen that the predicted value of F;for parameter values of 0.1 ppm DO,289'C, 0.015% sulfur, and 0.001% per second strain rate is 2.2. The corresponding value for the low-alloy steelis 3.2. These values are well within the factor of four.

3-17

EPRI Licensed Material Retneto of Current Research Studies on Environmental Effects A major task of the PVRC Working Group on S-N data analysis is the validation of each of the criterion listed above, specifically the sulfur content and the flow-velocity crite-rion. Nevertheless, for this report, it was judgd that each threshold criterion value is reasonable and can be used in evaluating the environmental fatigue life of carbon and low-alloy steels.

For the stainless steels, '.he PVRC Working Group has not yet recommended the thresh-

• old values for strain amplitude and strain rate. Nevertheless, a review of the LWR-type water envirorment test data for ennealed austenitic steels and nickel-base Alloy 600 presented in Reference 242 indicates that a threshold strain amplitude level of 0.1%

might be justified for these materials also. The strain rate threshold for the purpose of this report was assumed as 0.1% per second, as in the case of carbon and low-alloy steels.

A strain amplitude of 0.1% is equivalent to a pseudo stress amplitude of (0.1x30000/

100) or 30 ksi. Because the Code fatigue curve for carbon and low-alloy steels uses a value of 30,000 ksi for E, the same value was used in this calculation also. In the case of stainless steel and Alloy 600 for which the Code fatigue curve uses a value of 28,300 ksi, the corresponding threshold alternate stress amplitude would be 28.3 ksi.

If the alternating stress amplitude associated with any load set pair in an ASME Code fatigue evaluation is less than the preceding value, then that load state pair can be dropped from the environmental effects considerations. Similarly, if the strain rate exceeds 0.1% per second, then the environmental effects need not be considered. A seismic event meets this criteria as shown next.

If the seismic stress amplitude is less then the threshold alternating stress value dis-cussed earlier, then it is automatically excluded from the environmental effects consid-erations. Now, consider the case when the seismic strain amplitude is at slightly higher than the 0.1% level. To calculate the strain rate, we need to divide it by the rise time in seconds.The seismic event encompasses a range of frequencies from as low as few Hertz to 20 Hertz and higher. For the sake of this calculation, a frequency of 5 Hertz is assumed. Assuming a sine wave form, the rise time is expected to be 1/4 of the time duration for one cycle. Accordingly, the estimated rise time for this case would be 1/(5x4) or 0.05 seconds, giving a strain rate of 0.1/0.05 or 2.0% per second. This strain rate clearly meets the strain rate threshold criterion of 0.1% per second. Based on this,it is concluded that the load state pair consisting of a seismic event can be excluded from consideration of emironmental effects.

3-18

EPRI Licensed Material Retnew of Current Research Studies on Environmental Effects 5 ~ i i onnl i i unuj i a ininj i i unul i i unuj i i unut i i i nui 3 - ASME Mean for Carbon Steel at 288'C Reduced by Factor of 4 on Life -

2 o l

l O Carbon Steel Data

g

.0 7 6o A Low Ahoy St%I Data e  :

~

k -

~

.E E

3 -

( _

M 2 -

o' cP 8 b ob o.1 _

5 ' ' " ""I ""I ""'I ""I ""I ' ' " " "I ""I 1E+1 1E+2 1E+3 1E+4 1E+5 1E+6 1E+7 1E+8 Cycles to Failure Figure 3-10 Compilation of LWR-type Water Environment Test Data Satisfying Any of the Indepen-dent Criteria for Moderate Environmental Effects and Comparison to ASME Mean Curve Reduced by a Factor of 4 on Life.

3.5 Summary of Review The review indicated that the Argonne statistical model provides a reasonable basis for developing methods to include environmental effects in ASME Code fatigue evalu -

tions. Therefore, F, equations based on the Argonne statistical model were developed for use in ASME fatigue evaluations. The tentative threshold values suggested by the PVRC group appear to be reasonable and were used in developing the fatigue design rules as discussed in the next section. Load state pairs associated with the seismic events should be excluded from the consideration of environmental fatigue effects.

3.6 References

[3-1] M. Higuchi and H. Sakamoto. Trans. Iron and Steel Inst, of Japan. 24,1984, Bl%.

[3-2] M. Higuchi and H. Sakamoto. Trans. Iron and Steel Inst. of Japan. 71,1985,101-107.

[3-3] K. Iida, H. Kobayashi, and M. Higuchi. IIW XIH-1164,1985.

[3-4] M. Higuchi, K. Iida, J. Fukakura, et al. " Effects of LWR Water Environment on Fatigue Strength of Several Nuclear Structural Materials." Presented at ASME Code Subgroup on Fatigue Strength Meeting. New York City, September 1988.

3-19 l

EPRI Licensed Material Remew ofCurrent Research Studies on Environment 3 Effects

[3-5] M. higuchi and K. lida. " Fatigue Strength Correctior Factors for Carbon and low-Alloy Steels in Oxygen-Containing High-Temperature Water." Nuclear Engineering and Design.129,1991,293-306.

[3-6] N. Nagata, S. Sato, and Y. Katada. " Low-Cycle Fa tigue Behavior of Lo v-Alloy Steels in High-Temperature Pressurized Water." Trans. of the 10th International Conference on Structural Mechanics in Reactor Technology. Anaheim, CA 1989.

[3-7] M. Higuchi, K. Iida, and Y. Asada. " Effects of Strain Rate Change on Fatigue Life e' Carbon Steel in High-Temperature Water." ASME PVP-Vol. 306,1995.

[3-8] H. Kanasaki, M. Hayashi, K. Iida, and Y. Asada. " Effects of Temperature Change on Fatigue Life of Carbon Steel in High-Temperature Water." ASME PVP-Vol.

306,1995.

[3-9] G. Nakao, et al. " Effects of Temperature and Dissolved Oxygen Content on Fatigue Life of Carbon and Low-Alloy steels in LWR Water Environment."

ASME PVP-Volume 306,1995.

[3-10] Y. Asada. " Rate Approach for Fatigue Life Reduction Factor in LWR Water Environment." WRC Progress Report. Vol. No. XLVIII,9/10, p.148, September /

October 1993.

[3-11] K. Kishida, S. Suzuki, and Y. Asada. " Evaluation of Emironmental Fatigue Life for Light Water Reactor Comp ments." ASME PVP-Vol. 306,1995.

[3-12] NUREG/CR-4667. " Environmentally Assisted Cracking in Light Water Reac-tors." Semiannual Reports: Vol.12, August 1991; Vol.13, March 1992; Vol.16, September 1993, and Vol.17, June 1994.

[3-13] J. Keisler, O.K. Chopra, and W.J. Shack. " Statistical Analysis of Fatigue Strain-Life Data for Carbon and Low-Alloy Steels." NUREG/CR-6237. August 1994.

[3-14] O.K. Chopra and W.J. Shack. " Effects of LWR Emironments on Fatigue Life of Carbon and Low-Alloy Steels." ASME PVP-Vol. 306,1995.

[3-15] J. Keisler, O.K. Chopra, and W.J. Shack. " Fatigue Strain. Life Behavior of Carbon and Low-Alloy Steels, Austenitic Stainless Steels, and Alloy 600 in LWR Environ-ments." NUREG/CR-6335. August 1995.

4 1

3-20 l

EPRI Licensed Material 4

PROPOSED FATIGUE EVALUATION PROCEDURE t

This section describes the details of the proposed methodology to account for the envi-ronmental effects in the ASME Code fatigue evaluations. Before describing the details of this methodology, it is helpful to first summarize the current ASME Section III Code fatigue-evaluation approach.

4.1 ASME Section Ill, Nb-3600 and NS-3200 Fatigue Analysis Methodology Refemnces 2-18 and 2-19 summarize the current ASME Code fatigue procedures for piping and vessels. The current Code fatigue methodology described in this section is Gxtracted from these references. The stresses for the fatigue analysis are elastically computed. ASME Code NB-3600 methodology is used almost exclusively for piping and sometimes for branch nozzles. The ASME Code,Section III, NB-3200 (design by analysis) methodology is applicable to any component. It is generally used exclusively for vessels (sometimes augmented by NB-3300), fairly frequently for aozzles, and occa-sionally for piping.

4.1.1 ASME Code NB-3600 Fatigue Analysis Method The equations for service levels A and B are provided to ensure satisfactory cyclic (fatigue) behavior. To satisfy the range of primary-plus-secondary stresses, Equation 10 (Reference 4-1) must be satisfied. The stress range is calculated based on the effect of changes that occur in mechanical or thermal loadings that take place as the system goes from one load set (e.g., pressure, temperature, moment, and force loading) to any other load set that could also exist. The following must be satisfied for all pairs of load sets:

S,, = C, ' + C, M, + C,E,la,T, - a,T ]s SS,,, (Eq.4-1) where C,. C2 ,C, = secondary stress indices for the specific component under investigation (defined in Table NB-3681(a)-1 of Reference 4-1)

D, , t,1, S,,, = as defined in Equation 9 of Reference 4-1

, P, = range of service pressure, psi.

)

4-1

EPRI Licensed M:terial Proposed F tigue Evaluation Procedure M, = resultant range of moment which occurs when the system goes from one Wee load set to another, in-lb E, = average modulus of elasticity of two sides of a material or structural discontinuity at room temperature, psi a, , a, = coefficient of thermal expansion on side a and side b of a structural or material discontinuity,in/in *F T, , T, = range of average temperature on side a and side b of a struc-tural discontinuity, when the system goes from one service load to another, 'F The fatigue resistance of each piping component is assessed by evaluating the range of peak stress. For every pair of load sets, S, values are calculated using the following equation (Reference 4-1, Equation 11) :

D S, = K,C, D[P* + K,C,jM, + K,C,E,, la,T,-a,T, l y

(Eq.4-2) 7

+ 2(1- v)K,Eal AT,\ + 1-y Ea] AT,I where K, , K, , K, = local stress indices for the specific component under investiga-tion (defined in Table NB-36S1(a)-1 of Reference 4-1)

Ea = modulus of elasticity (E) times the mean coefficient of thermal expansion (a), both at room temperature, psi /*F AT, = range of the temperature difference for each load set pair between the temperature of the outside surface T, and the temperature of the inside surface T, of the piping product, assuming a moment generating equivalent linear temperature distribution,'F AT, = range for that portion of the nonline, r thermal gradient through the wall thickness not included in AT,, 'F v = Poisson's ratio A load set pair is defined as two loading sets or cases used to compute a stress range. ,

If Equation 4-1 cannot be sctisfied for all load set pairs, the alternative analysis de-scribed below may still permit qualifying the component. Only those lead set pairs that .

do not satisfy Equation 4-1 need to be considered.

4-2

EPRILianneed Material Proposed Fatigue Easluation Procedure where .

'M, = moment as ' defined for Ecuation 9 of Reference 4-1, in-lb, and all other terms as previously described C ', =

stress index (values defined in Table NB-3681 (a)-1 of Reference 4-1)

If these condit!ons are met, the value of S,shall be calculated by the following equation:

S, = K, b (Eq. 4-6) 2 where K,is as defined later and '

S, = alternating stress intensity, psi S, = peak stress intensity value calculated by Equation 4-2, psi The alternating stress for allload set pairs is computed as one-half of the p ak stress ranges calculated from Equation 4-2, or by the altemate approach of Equation 4 6 if Equation 4-1 is not met 'Ihe fatigue analysis is then performed using the applicable Code fatigue curve and the number of design cycles for each load case from the design specification. '

For ASME Section III Code editions prior to the Summer 1979 Addenda,~ Equation 4-1 contained an additional term. In these earlier Code editions, the AT, term of the peak stress in Equation 4-2 was also included in the primary plus secondary stress Equation 4-1:

S, = C * + b- M, + C,E,j a,T,-a,T,l -

2t 2I-Ea (Eq. 4-7)

+ AT, l 5 3S"-

2(1- v)

Adding this term frequently increased the stress S, above 3S,. When this occurred, Equations 4-3 and 4-5 had to be met, and the fatigue analysis was conducted using a relatively high K, factor, increasing the alternating stresses used in the fatigue analysis.

The ASME Section III Committee on Piping Design decided that this was overly conser-vative and modified the equation accordingly, starting with the Summer 1979 Addenda.

However, most current Section III plants were designed according to the earlier version of the SectionIII Code. ,

T ,

EPRI Licensed Mcterial Proposed Fatigue Evaluation Procedure

'Ihe following equation must be met (Reference 4-1, Equation 12):

S, = C M,

• s 3S,,, (Eq. 4-3) where S, = nominal value of expansion stress, psi M,* = same as M,in Equation 4-1, except that it includes only mo-ments due to thermal expansion and thermal anchor move-ments, in-lb When the limits of Equation 4-1 are exceeded, and before the rules of Equation 4-5 can be used, the value of the range of AT, cannot exceed that calculated per NB-3653.7, as follows:

AT, range s y'S' C (Eq 4-4) 0.7Eu .

wnere y' = 3.33,2.00,1.20, and 0.80 for x = 0.3,0.5,0.7, and 0.8, respec-tively x =

(PD,)2t) (1/S,)

P = maximum pressure for the set of conditions under consider-ation, psi C, = 1.1 for ferritic material

= 1.3 for austenitic material Ea = as defined in Reference 4-1, Equation 11, psi /*F S, = materid yield strength value, psi, taken at average fluid tem-Perature Note that the limitations on the AT, range are to ensure that thermal ratcheting due to the transient under consideration does not occur.

The pnmary-plus-secondary membrane plus bending stress intensity, excluding ther- .

mal bending and thermal expansion stresses, will be <3 S,,,. This requirement is satisfied by meeting the following equation:

C, '+C M + C' E.M T,-a,Td s 3S,,, (Eq. 4-5) 21 2I i

4-3

1 EPRI Licensed Mcterial Proposed Fatigue Etniuation Procedure

( Step 6. For each pair of load sets, the six components of stress are subtracted and the "

three principal stress ranges are computed. The peak stress-intensity range for each pair ,

is computed by subtracting the principal stresses, as described in Step 4, and choosing the largest.

Step 7. The S, for each load set pair is one-half the peak stress intensity range. To adjust for temperature and material, S, is multiplied by the ratio of the modulus of elasticity on the appropriate fatigue curve to the modulus of elasticity used in the analysis. The allowable number of cydes N, for each load set pair is read from the appropriate design fatigue curve.

Step 8. The individual fatigue usage factor u, at each location is determined by the ratio e of the number of design cydes (n,) to the allowable cydes (.Vp for each pair of load sett.

Once the individual usage factor for the load set pair with the largest S, is computed, the cydes associated with that load set pair are eliminated, and the process is repeated until the cydes associated with all the load sets have been exhausted.

Step 9. The cumulative usage factor (CUF) is the sum of the individual usage factors.

The ASME Code Section IIl limit is that the CUF at each location must not exceed 1.0.

This assumes a linear damage relationship, known as Miner's rule.

As stated in Step 4, if the primary plus secondary stress intensity range S, for a load state pair does not meet the 3S,11mit, a multiplier, the K, factor, is applied to the peak stress intensity to adjust for the effects of plasticity:

1 S, = - K, S, (Eq. 4-8) whtre S, = peak stress-intensity range K, = 1.0for S, s3S,,

= 1.0 + II - ") "

-1 for 3S,,, < S, < 3mS,,,

n(m - 1) (35,,, ,

g 49) 1

= -for S, 2 SmS.,

n where S,, = primary plus secondar stress intensity range S,,, = design stress intensity and m and n are defined as follows:

4-6

EPRILicensed Material Proposed Fetigue Evaluation Procedure 4.1.2 ASME Code NB 3200 Fatigue Analysis Method The first step in the NB-3200 fatigre evaluation methodology is to calculate the stress differences and the alternating stress intensity S, in accordance with NB-3216. The stress state changes occur as a msult of changes in the mechanical and thermalloadings as the system goes from one load set (e.g., pressure, temperature, moment, and force loading) to any other load set that could also exist. The following procedure is generally

• followed:

Step 1. The analyst must obtain a set of loadings for the component. This is generally in the form of a set of design- and service-level transients in the design specification.

These loadings define the temperature and pressure changes that the component is expected to undergo during its lifetime and the number of cycles n, for each of the i loadings, , ,. ,

Step 2. The analyst needs to determine the stress distribution at the most highly stressed locations in the component. Th'c includes the thermal and pressure stresses, and sometimes the preload stresses and therinal expansion stresses imposed on the component by the connecting piping.The determination of stress distributions gener-ally requires a finite element temperature and stress analysis.

Step 3. The three principal primary-plus-secondary stresses (S,, S, and S3),for each load set need to be determined. This sometimes involves separating the peak stress from the total stress, such as by linearizing the thermal stress distribution. Also, one needs to take into account the possibility of rotating the principal stresses at the point being considered during the stress cycle.

Step 4. From the results of Step 3, three stress intensities are calculated by subtracting the principal stresses.

Su= S -S Sn= S, - S, Su= S, - S, The maximtun primary-plus-secondary stress-intensity range is the largest difference between the Sn, Su , or Suvalues, determined by comparing the stress intensities of all the load sets. Tso values (one with the highest tensile stress intensity and the other with the highest compressive stress intensity of all the load sets) are used to form a load pair that determines the maximum primary-plus-secondary stress-intensity range. This stress-intensity range must meet the 3S, limit; otherwise, the simplified elastic-plastic method or a plastic analysis may be used.

Step 5. Using the stress values determined in Step 2, the peak stresses are calculated.

This might involve using stress indices, stress concentration factors, experimental stress analysis, etc. The six components of stress for each time and location of interest are .

determined for each load set.

i 4-5

EPRILiansed Material Proposed Fatigue Evaluation Procedure Material ' m n Low-alloy steel 2.0 0.2 Carbon steel 3.0 0.2

=

Austenitic stainless steel 1.7 0.3 Alloy 600 1.7 0.3 For the K, factor to be applicable, the primary-plus-secondary stress-intensity range cxcluding thermal bending, must meet the 3S, limit.

4.1.3 Summary of Code FnGgue Evaluation Approach From the preceding descriptions of the Code fatigue procedures, some of the common features of both the NB-3600 and the NB-3200 fatigue evaluations relevant, from an environmental effects point of view, can be summarized as:

o A number of distinct load states are defined at a given location where fatigue usage calculation is desired. In the case of NB-3200 analysis, the load states are defined in terms of the three principal stresses. In the NB-3600 analysis, the load states are defined in terms of the internal pressure, the three moment components, the average temperatures on the a and b s. ides (T, and T), and the two temperature' gradients, AT, and AT,.

o The load state pairs are formed and a peak stress-intensity range is calculated for each load state pair. An altemating stress-intensity amplitude S,is then calculated.

This value of S,is used to enter the appropriate Code fatigue curve to calculate the allowable number of cycles and then the fatigue usage associated with this load state pair.

o The partial fatigue usages from the various load state pairs are summed to obtain a cumulative, or total, fatigue usage factor.

4.2 Environmental Factor Approach 4.2.1 Overview The proposed approach includes as much information as typically available to the -

piping or vessel stress analyst, thus minir.tizing the need for additional information.

The essential steps of this approach are shown in the flow diagram in Figure 4-1 and are briefly summarized below:

o Determine the load state pairs that satisfy the threshold values listed in Subsection 3.3 A discussion on how to use the available information in a typical Class 1 stress report for comparison with the threshold values is provided later in this section. The fatigue usage of the load state pairs satisfying the threshold criteria remains un-changed by the environmental effects.

4-7

EPRI Licenud M:terial Proposed F:tigue Etuluation Procedure

• For the selected load state pairs that do not meet the threshold criteria, determine the appropriate values of the parameters such as T,7', S*, O*, and c'. Calculate the F, (Equations 3-11 through 3-14) and multiply it by the partial fatigue usage associ-ated with this pair to obtain a new partial fatigue usage.

Sum the partial fatigue usage factors calculated in Step 2 and add it to the partial fatigue usage of the load state pairs not selected in Step 1. This overall sum is the total fatigre usage including the environmental effects, Information Available from Pipina Strecs Report Sa, (Tr, dT , Ts, Ta fluid temperature for both Load Ehates in Every Load State Pair u

Load State Pair "i" Seismic Load States No g Yes  ;

, No Are the fluid temperatures for Yes _

both load states below 300*F o No Determine Effectrve Temperature, Tea, for the load set "i" u

Deermine strain rate fmm intc,iir,eticni "

on tenslie stress load state U4.nv = Us w v

i=1+1 Determine oxygen level (ppm) o

, Max. of the two load states v

Determine Nair/ N mme, Fen v

Muttiply the partial usage, U,, for this load pair by Fen U,,,, a Un x Fen 1

No i= n = Uwur = Uwat(i-1) + U,.n,

/ , ,Yes Uwutenv Figure 4-1 Flow Diagram for Environmental Fatigue I.

4-8

EPRIUesanedMaterial Proposed Fatigue Eoaluation Procedure 4.2.2 LoedSenen PairScreerning .

To identify which load state pairs are sensitive to reactor water effects, the threshold criteria in Table 4-1 (discussed in Subsection 3.3) can be applied.

Table 4-1

, Imad State Fair Screening 1hreshold Criteria S, s 30,000 psi (carbon steel) s 28,300 psi (stainless steel, Al!oy 600)

O s 0.1 ppm -

T s 300* F Sulfur < 0.003%

Flow > 10 ft/sec -

The first criterion that can be easily checked is alternating stress amplitude. 'lhe S,velue for each load state pair is listed in both the NB 3600 and NB-3200 fatigue evaluations.

As stated in Subsection 3.3, the threshold value, S,, for carbon and low-al!.oy steels is 30,000 psi, and for stainless steels and Alloy 600,28,300 psi. .

The next threshold criterion that can be checked is temperatare. If the highest tempera-ture of both load states in a load state pair is less than 300*F, then the temperature threshold criterion is satisfied and that load state pair can be excluded from environ---

mental considerations. The same also applies for oxygen content.

If the flow rate information is available for both load states in a load state pair, then the threshold criterion on flow rate can also be ud to eliminate appropriate load state pairs from consideration of environmental effects.

l 4.3 N8-3000 Analysis The load state pair screening conducted in the preceding subsection narrows down the load state pairs for which the environmental correction need to be applied. Beyond this point, the application of the environmental correction factors will be somewhat differ-ent for NB-3600 and NB-3200 fatigue evaluations. This section describes the pivcedures applicable to NB 3600 fatigue evaluations.

The next step in the NB-3600 fatigue evaluation process (Figure 4-1)is determining the -

, eappropriate values of temperature, strain rate, and oxygen concentration (for carbon

- and low-alloy steels) for each of there load state pairs. Once the appropriate values of these parameters has been determined for a load state pair, the er vironmental correc-

. tion factor F, can be determined by using one of the applicable equations (3-11 through 3-15) in Section 3. Partial fatigue usage, including environmental effects for that load state pair, is equal to the existing partial fatigue usage times the correction factor F,.

4-9

i EPRILicensed Material

> Proposed Fatigue Ennluntion Procedure i

Total fatigue usage, including the environmental effects,is obtained by summing these partial fatigue usages and adding to this sum the partial fati;;ue usages of load state pairs that were sva..ed out (see Figure 4-1). ,

1 A key parameter in calcul'ating F, is strain rate. The scphistication in calculating this -

parameter depends on the details of the available information on temperature tran- l sients. Therefore, two approaches are outlined. The fint one assumes that only the T,, ,

TyAT,, and AT,information is available. '!he second appoach is based on the availabil-4 i

ity of detailed elapsed time versus temperature information for a transient. Subsection ,

43.2 describes the details of these approaches. 3 4.3.1 Deenemining Temperature As discussed later in Subsection 43.2, it is the strain rate during the tensile phase, rather than the compressive phase, that is important from the viewpoint of environmental l fatigue damage. For example, a step-down temperature transient produces tensile

! stresses at the inside surface of a component that is typically in contact with the fluid.

'Iherefore, in determining an appropriate value of temperature for a load state pair, the load state associated with the step down transient is the one to consider.

Because the metal temperature is typically changing during a transient, the choices for temperature T are: ,

• Maximum temperature durirq the transient

• Average temperature during the transient

• Temperature at tae time when the maximur.i stress occurs

'Ihe last option might not be appropriate because the maximum or minimum stress wiually occurs at low temperatures, while a significant part of the tensile stress might have developed at higher temperatures. Reference 2-19 used the maximum temperature calculated for the times of maximum and minimum stresses,if known; otherwise, the maximum temperature for the load state pair was used.

'Ihe use of average temperature during the transient - .also somewhat questionable, because any averaging must also take into account the strain rate variation.

For the analyses in this subsection, the maximum temperature for the load pair was

. used conservatively. When an incremental approach to calculating F, was used, as desenbed Ir.er, the instantaneous metal temperatures and strain rates were used to obtain an instantaneous damage parameter F,.

4.3.2 Determinetton of Strain Rate Two approaches were studied to scount for stru. rate effects. In the first approach, .

information generally supplied in the Class 1 NB-3600 fatigue evaluation (i.e., load state, pressure, component moments, design cycles, T,, TyAT,, AT,, etc.) can be used to i

4-10 1

_ ~ . _ . _ . _ _ _ _ _ . . _ . _ _ _ _ _ _ . _ . _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ . _ _ _ _ _ . _ _ _ -

EPRILicensedM::terial Proposed Fatigue Evaluation Procedure calculate an average strain rate, in the second approach, the results of the one-dimen-sional heat transfer analysis can be used to calculate an effective damage factor Fg based on temperature and strain rate variations throughout the transient.

Both the Japanese and the Argonne test results show that it is the strain rate during the tensile phase, rather than the compressive phase, that is important from the viewpoint of environmental fatigue damage. Figure 4-2 shows a slide from Referer e 4-2 illustrat-ing this point. Generally, tensile stresses are produced on 'he inside surface of the pipe during a step-down temperature transient (i.e., the fluid temperature drops as the t*ansient progresses in time). Conversely, a step-up temperature transient produces compressive stresses. Therefore, we are interested in the fraction of S, of a load state pair that pertains to the load set with the positive strain rate.

[

Average Strain Rate The thermally induced strasses in the piping are expected to be proportional to the temperature gradient quantities such as l T,- T,I, AT, and AT,. Generally, the I T,- T,1, AT, and AT, values in defining a load set are selected at a point in a temperature tran-sient where the composite sum { l T,- T,I + AT, + AT3 ) reaches a maximum. Also,it is very likely that a load state pair with significant alternating stress amplitude would consist of a load set with a step-down temperature transient paired with a step up temperature transient load set, or vice versa. Therefore, it is reasonable to 'use this composite sum to determine the f. action of S, of a load sets pair associated with increas-ing strain.

Based on the preceding discussion, the following approach was used to estimate strain rate. Let l T,,- TuI, AT ,uand AT be v associated with a step-down temperature tran-sient and l T,,- T,,I, AT u and AT for u the other load set, presumably the step-up tem-perature transient.

< Now, define: T, = 1 T,,- uT l + l AT u l + l ATu l (Eq. 4-10)

T, = 1 T,,- T ,I + l AT u l + l ATu I (Eq. 4-11) nen, the peak stress magnitude associated with the increastng strain is:

S,, = 25,(T,/(T,+T,)] (8q. 4-12)

Now, assume that the elapsed time is t, where T, was determined (presumably, the maximum). Thea, the average strain rate for this load state pair is:

i = t, = S,,, /(Et,) (Eq. 4-13) 9 The value of E could be assumed the same as t'h at in the Code fatigue curve.

4-11

EPR1 Licenud M:terial Proposed Fatigue Evaluation Procedure Effects of Loading Waveform A iOG-Gr B,288'C, at =0.75% strain rate Fast / Fast SkmFast

{

A

\\\\ \Y\i Alr. 3,253; 3,753 B

YYl Air. 3,721; 3,424; 6,2,.

PWR: 1,525;2,230 PWR: 2,141 HiDO: 2,077;1,756;[1,765] Hi DO: 303;469;[342)

Fracnonof stain at skm ree:0406 Fracnon d straen at skw rate: 0.334 1 l i C

\\i; D

V/V/

Air. 5,139 Alr.

PWR: PWR:

HiDO: 545;[418) HiDO: 1,935;1,830;[682)

Frason of ersen at skm rate:0.547 Frachon of struen at slow rese: 0.16i.

E

\\tii F

h {O i Alr. 5,139 Air. 4,087 PWR: PWR:

Hi DO: 615; 553 HiDO: 1,649;2,080 Figure 4-2 Effect of Strain Rate Variation during Tensile and Compressive Phases of Fatigue Cycling (Reference 4-2).

Effective Damage Factor Tlie highest temperature in the load set was used in the preceding approach to calculate an environmental factor F,,, associated with a load state pair. Also, strain rate was ob-tained by averaging over the time period from the start of the transient to the time when the temperature stresses reach a peak. This approach 13 expceted to yield a conservative value of F,,,, However, when the detailed I T,,-u T I, DT 3 , and DT u information is available as a function of elapsed time from.the start of the transient,it is possible to calculate an effective value of F,,, that considers the variations in metal temperature and strain rate as the temperature transient progresses. This approach is expected to yield a less conservative value of F,,,. The following assumptions are made in calculating the instantaneous strain rate:

4-12

EPRI LicensalMaterial Proposal Fatigue Evaluation Procedure The peak stress value S,, as calculated in Equation 4-12, is associated with composite temperature T,. Stress at any other intermediate time point t during the transient is obtained as:

. S,,=T,/T, (Eq. 4-14)

T, = 1 T,,, - T.,,I + l ATt ,I + l ATt ,I (Eq. 4-15)

• Temperature T to be used in the F,,, is the instantaneous metal temperature, which at time t can be obtained as:

T,, a larger of(T,,,,Ts ,) + AT,,,/2 + ATz , (Eq. 4-16)

Theinstantaneous ain rate t, between time points t-At and t can then be calculated as:

i, = IS,,,- S,,,4,J/(ate) (Eq. 4-17)

The F,,, can then be incrementally calculated from the start of the transient until the metal temperature reaches 300*F (defined as f.):

Fa = (1/t) f*exp(+0.384-0.00133T,, + 0.554S*T,,,,*O* t, *)dv (Eq. 4-18)

The above formulation is for carbon steel piping. A similar expression for stainless steel piping can be also developed. The application of this approach is described in the next section.

4.3.3 Determination of Oxygsn Concentration

• Ihe oxygen concentration DO ir. u parameter in calculating F,,, for the carbon and low-alloy steels. The DO value can be conservatively taken as the maximum of applicable values for the load states constituting a load state pair. In view of the discussion in the preceding subsection, the use of DO level associated with the step-down transient is justified when the DO levels differ considerably between the two load states.

4.3.4 Determining Corrected Fatigue Usage Once the appropriate values of temperature, strain rate, and dissolved oxygen (for carbon and low-alloy steels) are detennined for a load state pair, the environmental correction factor F,,, for that load state pair can be calculated using one of the equations (3-11 through 3-15). If the effective damage approach is used, then the equation (4-18) can be used to calculate the effective value of F,,, for a load state pair. Partial fatigue usage associated with a load state pair from the previous Code calculations should be multiplied by F,,, to obtain the new value of partial fatigue usage. Total fatigue usage, including environmental effects,is obtained by summing these partial fatigue usages and then adding to this sum the partial fatigue usages of load state pairs that were screened out (see Figure 4-1).

4-13

EPRILicensed M:terial Proposed F:tigue Evaluation Proceduce 4.4 NS 3:t00 Analysis .

The approach for an NB-3200 detailed analysis would be similar to thai outlined in the l preceding subsection, the only major difference being the area of strain rate determina-tion for the NB-3600 approach. .

4.4.1 Determining Tempereture . .

l The approach is essentir.11y the same as that for NB-3600 analysis.

4.4.2 Determining Strain Rete .

The information generally available in an NB-3200 fatigue evaluation is a set of three peak principal stresses or six peak stress components (three dhect stresses and three shears) for each of the load sPt- ; 1.13 41000. 1.000 9.118 - 19. e 7500. ;01013 0 - 24 ~ l 30 '

00010.' $4531. 30040.1 1F338.  ; 1.08 . ' .37078. " 1.031 0.081; J 08.' C . --~ 40022. 7 0.0000 3 s 73 83140. 80230. 32448.? 20040. ' i.20 2 40331.' 1 1.130' 4.071 de.C. Osti. 8.0048 it =0 23 - 83062. 06787. 18015.1 21540.- 1.15 - 30137. 1.073 0.157 ' 70. 41007. 0.0000 11 0 31 50071. 40302. 13020, 20000.  ; 1.00 . 31353. 0.932 0.108 ' 08. _ 19243. 0.0042 Q' 12 1 8 48695 44344. 15932. 21849. 1.00 1.00 24347.

233%.

0.838 0.849 0.109 0.042 120.

45.

41983.

47998.

0.0029 0.0009 13 21 32 46617. 44924. 28133. 18756. t*

14 21 25 46200. 44508. 29303~ 17565. 1.00 23100. 0.841 0.042 65. 49339. 0.0013 15 9 24 43225. 43091 38471. 10349. 1.00 21613. 0.815 0.005 52. 68072. 0.0008 16 25 37 42917. 39291. 15604 20265. 1.00 21458. 0.743 0.089 80. 70535 0.0011 ,

17 25 23 42763. 41313. 21845. 17021.

12433.

1.00 1.00 21381.

19541.

0.781 0.666 0.025 0.091 46.

110.

71802.

108733.

0.0006 0.0010 g i 18 19 28 39081. 35213. 20993. m 9 38516. 37829. 32003. 10251. 1.00 19258. 0.715 0.021 70. 114592. 0.0006 j

i 19 20 26 17 39 38255. 37771. 21744. 19502. 1.00 19127. 0.714 0.009 110. 117439. 0.0009 {'

a

~'

21 2 9 37839. 37422. 32003 10484. 1.00 18919. 0.707 0.015 40. 122159. 0.0003 22 33 39 37343. 36859. 21147. 21107. 1.00 18671. 0.697 0.010 '. 128107. 0.0000 l 23 9 10 36833. 36657. 32003. 10504. 1.00 18417. 0.693 0.00R 108. 134608. 0.0008  !

24 10 29 35968. 34477. 30367. 9559. 1.00 17984. 0.652 0.021 162. 146637. 0.0011 ,

25 12 29 35954 34452. 30367. 9561. 1.00 17977. 0.651 0.020 3945. 146856 0.0269 I 26 12 28 35300. 33901. 28769. 9950. 1.00 17650. 0.641 0.023 3951. 156895 0.0252 j 34793. 20681. 20545. 1.00 17397. 0.658 0. 79. 165280. 0.0005 27 22 33 34793.  ;

28 18 22 34793. 34793. 20661. 20545. 1.00 17397. 0.658 0. 31. 165280. 0.0002 j 29 12 41 33975. 33924. 31845. 9400. 1.00 19988. 0.641 0.002 111. 180073. 0.0006 I 30 5 12 32267. 32216. 30367. 9378. 1.00 ;6134. 0.609 fl.001 40. 221601. 0.0002 j j 31 12 35 31441. 30042. 17574. 13350. 1.00 15721. 0.56E 0.032 80. 249459. 0.0003 i

I I i

EPRI UcensedMaterial Application Case Studies

. 1.5 minutes W 'e-Transent 13 216*C 21*C Transient 40

+ - 45.6*C/hr 10*C -

Transient 34 218*C -

10*C I.l.I*9 Transient 36

• "'- 32*C

]1mwe

- Transient 30 86*C 3 - *'-

55'C/hr 15 seconde Transient 14 216*C i

= = 21*C

--*! P--

4 minutes Figure 5-2 Temperature Profiles of Significant Transients.

5-7

1 EPRILianneedMaterial App!!ation Case Studies Transient 24 .

4t*Q 55'C/hr 49'C l I

i Transient 6 ,

289'c l I

10*C Transient 23 99'c V

21'C 245*C Transient 8 90'C Transient 31 132*C l 10'C --

Figure 5-2 (continued)

Temperature Profiles of Significant Transients.

5-8

EPRILicensed M:teri:.1 Application Case Studies 1

Now, consider the first load state pair,13-40, and determine the F, factor. Load state 13 l Ir this pair is associated with a step-down transient and, thus, was used to determine '

the strain rate for this load state pair. Figure 5 3 graphically shows the output results of a one-dimensional heat transfer computer run to determine the appropriate I T,- T,1, AT,, and AT, values for transient or load state 13. This figure also shows the plot of the calculated value of composite temperature T,, which, as decc %d earlier, is equal to (l T,- T,i + l AT,I + l AT,I ). Typical practice is to pick the T,, Ty AT,, ad AT, values where 'he composite temperature value reaches maximum. In this case, the compo=lte tempt ature value reached maximum at time equal to 1.501 minutes, and, therefore, the T,, TyAT,, and AT, values at this point were used to define load state 13.

ono-Demendonal Heat Transfer Analeyens Resulte ago WDunmT)

E + TATE

~ 4--QuenT2 IID 1~

Ib % g3=-gji Q g p !!N N__y --

sm Fluid Temperature Profile: Time (min.) Temp (F) 0.0 420.0 1.5 70.0 4.0 70.0 Thickness (a) = 1.03 in. Thickness (b) = 1.34 in Material: Carbon Steel 4

Figure 5-3 Output of One-dimensional Heat Transfer Analysis for Transient 13.

5-9

EPRI r Md Mcterial Appliantion Cane Studies

Table 5 2 ,

Strain Rate Calculation for Significant Load State Fairs i Los d AT i AT

T. T. T, S, (psi) -Su Tme Str.

1 Stat e (min) Rate l

, 13 40 1 49 19.6 16 L 20 U) 179.6 174096 95154 1.5 0.0035 1

! J 1' O 21 11 i 10d.0 144 38 40 1 1 0 4 2n i 24P.0 72  ! 113712 68043 1 0.nnan .

1 1'I) 28 1' , 1 102.0 14 l l 34 40 1 73 14 M 1i l55.1) 11 123914 53978 i 0.33 0.0001

, J 11') 28  !

02.0 14' I

! 14 30 1 47 9.8 38 > l 64.0 78 I amp an%1 0.25 0.0108 i J 4) 60 10 U) it) 91U 24 30 1 C '.3 0 472.6 47 .0 0.S 75952 75U16 0.25 0.0168_---

J 4) -50 156 it i.1) '10 6 23 1 -1'l 4 516 52 l. 0 33 80852 29251 0.25 0.nnn5 i J 31 ' 4 117 10 '.0 58

! 8 23 1 4) 8 487 50 u) 1 29 72274 49857 0.25 0.0111 J Sli 4 117 10 >.0 i8 -

8 31 1 4) 8 487 50 U) 129 62706 36274 0.25 0.0081 J s1 21- 197 19a.0 94

)

2 Table 5-2 shows the streb rate calculation for each of the eight load states pairs. To further illustrate this, a detailed discussion of the strain rate calculation for load state i pair 13 40 follows. The composite temperature of load set 13 is 179.6'F, and that of load 1  :

set 40 is 149'F. Again, these values were obtained from the T,, T,, AT,, and AT,informa-tion given in the load set definition of Tabic 5-1B. The altemating stress amplitude for L this load state pair is given as 87048 pai, see Table 5-1C. Therefore, the peak stress range I for this load state pair is:

S, = 2 S, I

= 2 87048 psi l

L S, = 174096 psi The tensile portion of peak stress range associated with load state 13 is calculated using Equation 4-12:

! S,., = 2*S, [T, AT,+T,)] (Eq. 5-1)

= [174096]*[179.6/(179.6+149)] psi S,,, = 95154 psi -

Because the composi'e temperature reaches a maximum at 1.5 minutes from the start of .

transient 13, the strain rate is then calculated using Equation 4-13:

5-10 ns . - - < . -- -- ,v~,ers. ,

EPRILicensed Mcterial Application Case Studies e = c, = S,/(Eet). (Eq. 5-2)

=

[95154/(30x10'*1.5x60)]e100

. c, = 0.0035 % per second

'Ihe strain rates for the other load state pairs were similarly calculated and are shown in Table 5-2.

The strain rate values soovm in Table 5-2 were then used to calculate the environmental fatigue correction factor F,,, esing Equation 3-11. This calculation for load state pair 13-40is: .

F,,, = erp (+0.384 - 0.001337 - 0.SS4S*T*O* t') (Eq. 5-3)

The values of various parameters for load state pair 13-40 are the following:

T = 421'F O = 0.2 ppm i = 0.0035%/see Based on the preceding values and Equation 3-8, the following values are Xtained Ior S*,T*,O*,andc':

S' = S = 0.01 T* =

[T-300]/1.8 = 67.2*C O' = 0 = 02

f. * = =

In(e) In (0.0035) = -5.6481 When these values of the parameters are substituted in the preceding equation for F,,,, a value of 1.677 is obtained. The partial fatigue usage factor for this load state pair is multiplied by the calev. lated value of F,,, to obtain the new partial fatigue usage reflect-ing the environmental effects:

U,,,a m

• U a m F,,, (Eq. 5-4)

= 0.0115 e 1.677

= 0.0195 Note that the existing partial fatigue usage for load state pair 13-40 based on Code (air)

, _ fatigue curve is listed in the last column of Table 5-1C. Similar calculations were per- -

formed for the remaining seven load state 2 pairs, as shown in Table 5-3. Cumulative fatigue usage (CUF) for these load state pairs is now corrected by totaling all partial usage factors. CUF increased from 0.0727 to 0.1204, or by a factor of 1.66.

5-11

EPRILicensed M:terial _

Appliattion Case Studies Table 5-3 .

F, Calculation for Significant Load Pairs Sulfur (%) = 0.01, S' = 0.01, E (ksi) = 30000 Load Set I .I Ss To is i T' O 0* Str. Rate epset* fn Un Uen (ksi)  %/see 13 40 87.0 421 270 421 67.2 0.2 0.2 0.0035 -5.6481 1.677 0.0115 0.0193 38 40 56.9 421 270 421 67.2 0.2 0.2 0.0038 -55780 1.669 0.0031 0.0052 34 40 62.0 421 270 421 67.2 0.2 0.2 0.0091 4.7009 1.563 0.0361 0.0564 14 30 41.7 421 270 421 67.2 0.2 0.2 0.0108 4.5292 1.543 0.0013 0.0020 24 30 36.0 473 270 473 96.1 0.2 0.2 0.0168 4.0463 1.638 0.0058 0.0095 6 23 40.3 552 210 552 140.0 0.2 0.2 0.0065 5.0359 2.1 84 0.0048 0.0105 8 23 36.1 473 210 473 96.1 0.2 0.2 0.0111 -4.5027 1.712 0.0059 0.0101 8 31 31.4 473 270 473 96.1 0.2 0.2 0.0081 4.8207 1.771 0.0042 0.0074 Total . 0.0727 0.1204 The corrected fatigue usage for all 31 load state pairs (CUFJ ran now be calculated:

CUF, a CUF, + (0.1204-0.0727) (Eq. 5-6)

CUF, = 0.1409 + 0.0477 CUF, = 0.1886 The F, value for the load state pair 13 40 was also calculated using the effective damage approach outlined in Subsection 4.2.5. Table 5-4 shows the details of this calculation.

The calculations are carried to the point where the calculated metal temperature (col-umn 10) reaches the threshold value of 300*F. The last column shows the calculated value of the cumulative average value of F,. It is seen that the value of F, using this approach is 1.38, compared to the earlier calculatixi value of 1.677. This is indicative of the conservatism that the effective damage approach can remove from the simplified calculations, such as those shown in Table 5-3.

l 5-12

EPRILkensedM:terial Application Case Studies Table 5-4 F, Calculation for Load State Pair (13-40) Using Effective Damage Approach Trne 011 1A TB DELT DEL 12 COM strs Strain Metal T* epsat* In *dt /n

. (min) AB BIN. (ksi) Rate Temp N/sec) F) 0.000 0.0 420.0 420.0 0.0 0.0 0.0 0.0 420.0 66.7

. 0.010 0.4 419.9 420.0 0.0 1.3 1.7 2.0 0.0121 418.5 65.8 4.4120 0.0153 1.53 0.019 1.3 419.7 419.9 0.1 2.4 3.9 4.3 0.0150 416.7 64.9 4.2027 0.0135 1.51 0.029 2.8 419.5 419.7 0.2 3.4 6.3 6.5 0.0134 414.9 63.9 4.3105 0.0150 1.51 0.041 4.7 419.0 419.4 0.4 4.6 9.6 9.5 0.0140 412.5 62.5 4.2244 0.0179 1.50 0.051 4.3 416.5 419.1 0.6 5.3 12.2 11.5 0.0129 410.7 61.5 4.3530 0.0150 1.50 0.060 8.1 418.1 418.8 0.7 5.9 14.7 13.7 0.0140 408.8 60.5 4.2696 0.0133 1.50 0.070 -9.9 417.5 418.5 0.9 4.5 17.3 15.8 0.0123 407.0 59.4 4.3971 0.0149 1.50 0.079 11.7 417.0 418.1 1.1 7.1 19.9 17.8 0.0134 405.1 58.4 4.314 0.0133 1.50 0.089 13.8 416.4 417.7 1.3 7.6 22.5 19.8 0.0118 403.3 5 '.4 4.4415 0.0148 1.49 0.102 16.1 415.5 417.1 1.6 8.2 25.9 22.3 0.0118 400.9 56.0 4.4413 0.0192 1.49 0.114 18.8 414.6 T0.5 1.9 4.7 29.2 24.8 0.0124 398.4 54.7 4.3901 0.0178 1.49 0.130 -21.7 413.3 415.6 2.3 9.4 33.4 27.8 0.0113 395.4 53.0 4.4874 0.0235 1.49 0.149 26.4 411.7 414.5 2.8 10.1 38.2 31.3 0.0109 391.7 51.0 -4.5164 0.0277 1.48 0.171 29.5 409.7 413.1 3.4 10.8 43.7 35.1 0.0103 417.1 65.1 4.5749 0.0328 1.48 0.180 31.2 408.8 412.5 3.7 11.0 46.0 36.7 0.0104 385.8 47.7 -4.5642 0.0135 1,48 0.190 42.9 401.8 3*.8 4.0 11.3 48.2 38.2 0.0092 384.1 46.7 4.6908 0.0145 1.48 0.199 34.5 406.9 411.2 -4.3 11.6 50.4 39.7 0.0100 382.3 45.7 4.6058 0.0129 1.48 0.221 -38.3 404.5 409.5 5.0 -12.2 55.5 43.1 0.0092 378.2 43.4 4.6871 0.0315 1.48

) 0.440 41.3 402.4 408.1 5.6 12.7 59.6 46.9 0.0087 374.7 41.5 4.7391 0.0271 1.47 0.258 44.3 400.2 406.5 6.3 13.1 63.7 48.8 0.0089 371.2 39.6 4.7257 0.0254 1.47 0.280 47.8 397.6 404.6 7.1 13.6 68.3 51.0 0.0081 367.2 37.3 4.8164 0.0310 1.48 0.299 -50.3 395.2 403.0 7.7 14.1 72.1 54.0 0.0077 363.7 35.4 4.8672 0.0268 1.48 0.321 53.3 392.4 400.9 8.6 14.5 76.4 56.7 0.00'2 359.8 33.2 4.933f, 0.0308 1.45 0.339 -55.8 389.9 399.1 9.3 14.8 79.9 58.8 0.0072 356.4 31.3 4.9313 0.0248 1.4d 0.361 58.5 386.9 397.0 -10.1 415.2 83.9 61.3 0.0068 352.5 29.2 -5.0184 0.0302 1.44 0.379 40.8 384.2 395.1 10.9 15.6 87.2 63.3 0.0068 349.1 27.3 5.0144 0.0245 1.44 0.401 43.3 381.1 392.8 11.7 15.9 91.0 65.5 0.0061 345.2 25.1 5.1042 0.0297 1,44 0.419 45.3 378.3 390.8 12.5 16.2 94,1 67.3 0.0061 341.9 23.3 5.0994 0.0241 1,43 0.440 -67.6 375.0 388.4 13.4 16.6 97.6 69.4 0.0058 338.0 21.1 5.1431 0.0279 1.43 0.459 49.5 372.1 386.3 14.2 16.9 100.5 71.1 0.0053 334.6 19.2 -5.2387 0.0251 1.42 0.480 71.6 368.7 383.8 15.1 -17.1 103.8 72.8 0.0050 330.8 17.1 5.2961 0.0275 1.42 0.501 -73.6 365.2 381.2 16.0 17.4 106.9 74.8 0.0049 327.0 15.0 5.3158 0.0273 1.41 0.549 77.7 357.1 375.2 -18.0 18.0 113.7 78.2 0.0048 31 8.3 10.2 -5.3881 0. % 15 1.40 0.600 81.7 348.2 368.5 20.2 18.5 120.4 81.8 0.0041 309.1 5.0 5.4943 0.0639 1.39

, 0.651 -85.1 339.1 361.5 22.4 -19.0 126.5 84.7 0.0034 300.0 0.0 5.6732 0.0624 1.38 0.701 -88.1 329.9 354.4 24.5 19.4 132.0 87.3 0.0032 290.9 0.0 5.7520 0.0606 1.36 0.751 90.8 320.4 347.0 26.6 19.8 137.2 89.7 0.0028 281.9 0.0 5.8639 0.0610 1.35 O.799 -93.1 311.3 339.9 28.6 20.0 141.7 91.6 0.0024 273.4 0.0 4.0406 0.0590 1.35 d

5-13

EPRI Lkensed M:tedal Application Case Studics Table 5-4 (continued)

F, Calculallon for Load State Pair (13-40) Using Effective Damage Approach 1,me DT1 TA TB DELTA DEL 72 COMB strs Strain Metal 7' epsvt* In'at /n (min) B IN. (ksi) Rate Temp .

(%/sec) (F) 0.851 95.1 301.1 331.8 30.8 20.3 146.2 93.4 0.0020 264.0 0.0 4.2047 0.0643 1.34 0900 96.8 291.3 324.0 32.8 20.5 150.1 94.3 0.0018 255.1 0.0 4.3018 0.0610 1.33 -

0.949 98.3 281.4 316.1 34.7 20.0 153.6 96.1 0.0014 246.4 0.0 -6.5471 0.0614 1.33 1.000 99.5 271.0 307.7 36.7 20.8 157.0 97.1 0.0012 237.2 0.0 6.7202 0.0643 1.33 1.099 101.3 250.0 291.1 40.5 20.9 162.7 98.5 0.0008 219.6 0.0 4.9078 0.1265 1.32 1.200 102.2 229.7 273.9 44.2 20.9 16T.3 99.1 0.0004 201.9 0.0 6.9078 0.1308 1.32 1.300 102.2 209.0 256.8 -47.8 20.6 170.6 98.8 0.0002 185.0 0.0 4.9078 0.1311 1.401 101.4 188.2 239.5 51.3 20.3 173.0 97.8 0.0006 168.5 0.0 6.9078 0.1341 l 1.501 99.2 168.1 222.9 -54.8 19.5 173.5 95.2 0.0015 153.8 0.0 6.9078 0.1342 1 1.601 48.5 151.1 207.7 56.6 15.2 160.4 81.9 0.0079 148.2 0.0 6.9078 0.1347 1.700 75.7 137.5 1 94.2 -56.7 12.9 145.2 69.9 0.0073 143.4 0.0 4.9078 0.1339 1.801 44.2 126.2 181.7 -553 10.9 130.6 59.2 0.0063 138.7 Os 4.9078 0.1370 1.901 -54.7 116.8 170.4 5'.6 9.3 117.8 50.5 -0.0052 133.7 0.0 4.9078 0.1362 1.999 47.0 109.4 160.5 c1.1 8.0 100.2 43.4 0.0043 128.9 0.0 -6.9078 0.1339 2.200 34.8 97.8 143.0 45.2 -5.9 85.9 32.2 0.0034 119.6 0.0 6.9078 0.2766 2.500 22.5 86.9 122.7 35.8 3.9 62.2 20.8 0.0023 107.6 0.0 6.9078 0.4165 2.751 15.8 81.3 110.0 28.7 2.7 47.3 14.7 0.0015 99.3 0.0 4.9078 0.3506 2.999 +11.4 77.7 100.4 22.7 2.0 36.0 10.5 0.0010 92.7 0.0 6.9078 0.3481 _

l 3.250 8.2 75.3 93.0 17.6 1.4 27.2 7.6 0.0007 87.5 0.0 -6.9078 0.3537 l 3.500 5.9 73.7 87.3 13.6 1.0 20.5 T5 -0.0005 83.3 0.0 6.9078 0.3534 1 3.750 4.3 72.7 83.1 10.4 -0.8 15.5 4.J 0.0004 80.2 0.0 -6.9078 0.3542 4.000 3.2 71.9 79.9 8.0 0.0 11.7 3 4 0.0003 77.7 0.0 6.3078 0.3548 l

Note: The material is carbon steel. The values ofstress indices, E, Alpha and Poisson's ratio are ,

the same as those in Figure 5-3.  !

l l

1 5.1.2 Recirculation System Piping A BWR/4 recirculation piping system was analyzed in Reference 2-18. Figure 5-4 shows the mathematical model of the piping system. The piping 2terialis Type 304 stainless steel. The location chosen for fatigue evaluation was tP residual heat removal (RHR) system return tee where the calculated fatigue usage ractor was found to be high. The fatigue usage calculation, using the anticipated cycles (Table 5-137 of Reference 2-18),

w as chosen for the application of methodology developed in Section 4. Table 5-5 lists .

l the alternating stress amplitudes and the strain rates used in that evaluation. A revised version of stainless steel interim fatigue curve was used to determine CUF. Reference 2-18 reported a CUF of 3.256.

5-14 l

EPRI Licenud M:terial l Application Case Studies a

f*

. 2

% y  %

RHR Retum Tee ,

' p f , h  %

j pp A "1l,

( m y

(Q" "y -

e d,

RHR Outlet RHR Retum Tee Tee eq ,

- A ,

[

/ .a Figure 5-4 Mathematical Model of ,VR/4 Recirculation Piping System.

To make an equal comparison, the alternating stress magnitudes and calculated values of strain rates for various load state pairs were assumed to be same as those used in the Reference 2-18 analysis.

5-15

EPRI Licenud Mcterial Application Can Studies l

4 Table 5-5 .

l CUF Results for Example Recirculation Piping System i

Current Code Proposed approach Fatque Fatigue Usage Usage _

Load Pair S,,, n N u, Stram f., U,,

(ksi) Rate

, g ec) .

Composite Loss E/0BE 182.76 10 266 0.0376 0.022 2.388 0.0898

. Composne Loss A/RHR 161.69 10 384 0.026 0.019 2.435 0.0633 B

l Turtune Ron A/RHR B 144.89 160 $35 0.299 0.017 2.472 0.739 i RHR A/0B5 133.56 40 693 0.0577 0.016 2.492 0.144 RHR AlTurbine Ron A 116.13 12 1086 0.011 0.014 2.537 0.028

, RHR A/ Comp. 107.48 10 1412 0.0071 0.013 2.562 0.018

Loss C RHR A/ Comp. 100.12 10 1796 0.0056 0.012 2.59 0.014 Loss D RHR Al Comp. 99.65 10 1825 0.0055 0.012 2.59 0.014 Loss G RHR ASurb. 94.26 88 2227 0.0395 0.011 2.62 ' O.103 Trip Scrams B Turn. Inp B/ 63.86 10 10107 0.001 0.001 3.613 0.004 Shutdown Turb. Tnp Al Nun & 62.87 10 10905 0.001 0.001 3.613 0.004 Cooldown l lurbine inp-Scrams 59.20 160 14611 0.011 0.001 3.613 0.04

/ Shutdown lurtune inp-Scrams 57.14 36 17358 0.0021 0.001 3.613 0.007

/ Cooldown Turtune Roll B/hluli 56.85 172 17793 0.01 0.001 3.613 0.036 WarmuhComposne 55.42 10 20146 0.0005 0.001 3.613 0.002 j Loss F Hydrotest Down/ 50.64 68 31786 0.002 0.001 3,613 0.007 Startup Reducten to Power / 50.56 139 32041 0.0043 0.001 3.613 0.016 Cooldown 5-16

EPRILicensedMaterial Application Case Studies Table 5-5 (continued) .

CUF Results for Example Recirculation Piping System Current Cooe Proposed Approach fatgue Fatigue Ussoe Usage 4

Load Pair S ,, n N u, Strain f,, u,,

4 (ksi) Rate

(%/sec)

Reducten to Power / 50.43 26 32460 0.0008 0.001 3.613 0.003 Warmup Warmup /$tartup 60.42 104 32494 0.003 0.001 3.613 0.011 46.76 25 47562 0.0005 0.001 3.613 0.002 42.83 10 76633 0.0001 0.001 3.613 0.0004 42.74 $8 77520 0.0007 0.001 3.613 0.002 41.82 10 87342 0.0001 0.001 3.613 0.004 Total 0.626 1.351 l Now consider the first load state pair in Table 5-5 with a S, of 182.76 ksi and strain rate

of 0.022%/second. The F,,, for this load state pair was calculated using Equation 3-13.

The only variable in that equation b t'. The value of L' was obtained using the strain rate of 0.022%/second:

l' = In (i)

= In (0.022)

( l' = -3.8167 The F, was then calculated as follows:

F, = exp (+0.359 - 0.134 t') (Eq.5-7)

= exp (+0.359 - 0.134.[-3.8167])

F, = 2.388 Given the S, value of 182.76 and the number of cycles at 10, the partial fatigue usage for this load state pair based on the 19?' Code fatigue curve was calculated as 0.0376. The corrected partial fadpe usage fc r tus load state pair was then obtained by multiplying this value with F,:

U, = U,,* F,,, (Eq.5-8)

= 0.0376 2.388

= 0.0898 5-17

EPRI LicensedMcterial Application Case Studies Similar calculations were repeated for other load state pairs in Table 5-5. The CUF based on the 1992 Code fatigue curve was determined to be 0.526. Table 5-5 shows the calcu-lated value CUF using the proposed reproach as 1.351. This represents an increase by a factor of (1.351/0.526) or 2.57 compared to an ' ease by a factor of (3.256/0.526) or 6.19 based on Reference 2-18.

Note that the fatigue usage factor at the RHR t 'is high due to conservatisms bw. tnto the NB-3600 procedures. A fatigue usage calcul don in Reference 5-1 for a similar recirculation line to RHR branch connection using both the NB-3600 4.nd NB-3200 meth-ods clearly illustrate this. It was reported that for one of the load states pair (28-9), the calculated fatigue usage based on the NB-3600 procedures was 0.43 versus 0.0002 when the NB-3200 procedures were used. Although such dramatic reductions in calculated

, fatigue usages are not always possible,it does illustrate that selective use of NB-3200 methods might help reduce the calculated values of fatigue usage factors.

5.1.3 PWR Surge Line In Section 5.2.3 of Reference 2-18, Ware, et al., considered a PWR surge line elbow for evaluating fatigue usage. Table 5-32 of that reference presented the results of their i evaluation using the anticipated number of cycles. The same case is considered here and the fatigue usage was calculated using the proposed methodology. The, surge line material is SA-376 Type 316 stainless steel. For consistency, the strain rate was assened as 0.001%/second for all of the load state pairs, as in Reference 2-18. The calculated fatigue usage factor in Reference 2-18 was 1.345. A revised version of stainless steel interim fatigue curve was used in that evaluation.

Table 5-6 shows the results of the evaluation using the proposed methodology. The calculation procedure for F,,, was essentially the same as that described in the preceding subsection. The last three load state pairs listed in Reference 2-18 were not included in Table 5-6 because their contribution to the total fatigue usage was insignificant. A re-view of Table 5-6 indicates that the calculated value of the cumulative fatigue usage based on the proposed methodology is 0.425, versus the reported value of 1.3E in Reference 2-18.

i 5-18

EPRILicenud M:terial Application Case Studies Table 5-6 .

CUF Results for Surge Line Elbow Current Code Proposed Appr5 fattg : s Fatioue Usaoe usa ,

Load Pair S,,, n N u, Strain t, l ' ',,

(ksi) Rate

(%/sec)

Stratif./ Loss of Flow 59.56 2 14187 0.0001 0.001 3.613 0.0004 with ReactorTrip Stratif./ Loss of Flow $8.29 1 15755 -

0.001 3.613 -

, with Loss of Load Stratif / Loss of Flow 57.01 37 17552 0.002 0.001 3.613 0.007 w/o Loss of Load Stratification / Loss of 40 20090 0.002 0.001 3.613 0.007

! Lad ( 55.45 Stratifcation/ " ~ ~ 54.25 70 22440 0.003 0.001 3.613 0.011

, Reactor Trip l Strattfcatioril 50.44 71 32430 0.0022 0.001 3.613 0.008 i StratWication Stratifcation/ 49.19 67 36815 0.0018 0.001 3.613 0.006 Reactor Trip Stratrfcation/ 45.47 5 55210 -

0.001 3.613 -

Low Pressure Strattfcation/ 44.89 202 59240 0.0034 0.001 3.613 ' O.012 Plant Unioadino Stratrication/ 36.99 17570 170080 0.103 0.001 3.613 0.372 l

Stratification Stratifcation/ 33.31 150 319215 0.0005 0.001 3.613 0.002 Leak Test A Stratrication/ 32.66 2 361010 -

0.001 3.613 -

Hydrotest Total 0.118 0.425 1

5.2 NS 3200 Application For the NB-3200 application, a feedwater nozzle safe end fatigue evaluation presented in Reference 2-18 was considemd. The same case is used here to calculate a new fatigue usage factor based on the proposed methodology.To make a direct comparison with the CUF determined h Reference 2-18, the alternating stresses, strain rates, and tempera-tures for the various load state pahs were assumed to be the same as those in that refer-ence. The safe end material is SA-508 carbon steel. The CUF determined in Refemnce 2-i 18 was 1.73.

5-19

EPRILicensed Mcterial Application Case Studies Table 5-7 CUF Results for a Feedwater Nozzle Safe End Current Code Fat gue Proposed Approach Fatgue Usage )

Usage Load Pair S,,, n N u, Temp Strain f,, u,.

(ksi) ('C) Rate

(%/sec) .

Turtune Rou A/TG 82.27 120 1024 0.117 200 0.028 1.515 0.177 Trip A Turtune Rol! A / Hot 72.60 90 1429 0.063 200 0.026 1.524 0.096 Standby A Hot Standby A 64.41 142 1967 0.072 200 0.026 1.524 0.110

/ Nun Shutdown A 38.98 555 9272 0.060 200 0.002 1.886 0.113

/Nul Turbtne Rou A 29.28 10 23830 0.0004 200 0.001 1.0

• 0.000

/rurbine Trip A 4 Turtune Ron B/ 20.85 120 81350 0.001 200 0.001 1.0' O.001 TG Trip B TG inp B/Nuu 19.21 98 115630 0.0008 200 0.001 1.0' O.000 8

Turtune inp B 17.66 10 159810 -

288 0.001 1.0' -

/ Nun OBE/NA 17.44 10 163810 - 288 0.001 1.0* -

Hot Standby 8 13.85 222 444850 0.0005 288 0.001 1.0

• 0.000

/Nul 5 Shutdown B 13.43 666 523970 0.001 288 0.001 1.0' O.001

/ Null Startup/ Nun 13.33 120 560450 0.0002 288 0.001 1.0' O.000 l 2

~

l Total 0.316 0.50 l

• The alternating stress S, rnects the threshold criteria.

l The results of CUF calculations using the proposed approach are shown in Table 5-7.

The CUF using the 1992 Code fatigue curve is 0.312 as shown in Table 5-7. The follow-ing illustrates the partial fatigue usage calculation using the proposed approach. Con-sider the first load state pa'r in Table 5-7. The temperature associated with this load state is 200*C and the strain rate is 0.02f ',/second. The F, for this load state pair was calculated using Equation 3-11. The par..neters were as follows:

S = 0.015 T = 200'C t = 0.028%/sec O = 0.2 ppm l

l 5-20

EPRILicensed Material Application Case Studies Using the preceding values, the parameters that go into the expression for F,,, were obtained using Equation 3-8:

S' = 0.015

~

P = M L' = In(s)

,= In (0.028)

= -3.5755 ,

O' = 0.2 The F,,, was calculated using Equation 311 for carbon steels as follows:

F,,, = exp (+0.384 - 0.00133T - 0.554S*T*O' 6*) (Eq. 5-9)

= exp (0.384 - 0.00133 50 - 0.554 0.015 50 0.2*[-3.5755])

= 1.515 The partial fatigue usage for this load state pair based on the 1992 Code fatigue curve was earlier calculated as 0.117 *Ihe coasted partial fatigue usage for this load state pair was then obtained by multiplying this value by F,,,:

U,,, = U,,eF, (Eq. 5-10)

= 0.117 1.515

= 0.177 Similar calculations were repeated for other load state pahs i.n Table 5-7. The CUF incorporating the emironmental effects is shown as 0.497 in Table 5-7. It is seen that the increase in calculated CUF using the proposed approach is considerably less than when using the interim curves.

In connection with the feedwater nozzle and safe end fatigue usage factors, it is note-worthy that the fatigue usage based on monitoring the actual plant transients is gener-ally considerably lower than that based on the design transients. For example, based en fatigue monitoring by General Electric of a Japanese BWR for two fuel cycles, the 40-year CUF for the feedwater nozzle was estimated at only 0.0074, compared to the design

. basis value of 0.387 [ Reference 5-2]. Similarly, it was noted in Reference 2-19 that for 12 startups and 11 shutdowns at a BWR, the computed feedwater nozzle CUF based on fatigue monitoring was about 1/30th of the design basis CUF.

5-21

~ ~ ~ - --

Ma%I Applicatiet Cane Studia 5J References .

[5-1] H.L. Hwang, J.R., and D.M. Bosi. " Fatigue Usage Factor Evaluation for an Inte-grally Reinforced Bran .h Connection using NB 3600 and NB-3200 Analysis Methods. ASME PVP. Volume No. 313-2.1995. .

[5-2] T. Sakai, K. Tokunaga, G.L. Stevens, and S. Ranganath. " Implementation of Automated, On line Fatigue Monitoring in a Boiling Water Reactor." ASME PVP. '

. Volume 252,1993.

6 4 6

4 e

e

-- _ _ . -_ ____ =-_- _ -

EPRI Licensed Matertul 6

RECOMMENDED ASME SECTION lll CHANGES 4

This section presents suggested changes in the appropriate artides of ASME Section III to provide stress analysts with enabling words to indude the environmental effects in the Code fatigue evaluailons conducted according to NB 3600 and NB-3200. The walu-ation procedures could be incorporated in the form of a non-mandatory appendix to the Code. A suggested format of such an appendix is provided in the Appendix to the report.

6.1 NB-3600 Paragraph NB 3610 spedfies the general requirements of piping design. A subpara-graph NB-3614 worded as follows might be added to provide enabling words.

NB-3614 EnvironmentalEffects When ihe environmental eJJects on ihefatigue analysis required by NB-3650 are considered signtpamt, such effects may be accountedfor by using the methods described in Appendix XX.

The other location that needs to be modified is another subparagraph, NB-3653.8, whidt

, might be added as follows:

i NB-3653.8 Consideration of EnvironmentalEffects l

When the environmental effects are considered signifcant, Ihe cumulativefatigu damage may c

be calculated using the procedures ofAppendix XX.

6.2 NB-3200 The procedure for fatigue analysis in NB-3200 is contained in NB-3224.4(e), " Procedure for Analysis for Cyclic Loading." Paragraph (5) of NB-3224.4 (e) stipulates six steps for calculating cumulative fatigue damage when there are two or more types of stress cydes that produce significant stresses. Add the following step to NB-3224.4(e):

Step 7: When ihe environmental efJects on thefatigue hfe are considered signspcant, such effects may be accountedfor by using the methods described in Appendix XX.

6-1

CPRI Licenud M:terial Recommended ASME Section fil Changes 6,3 Non-mandatory Appendix Overview At the end of this report is a non-mandatory Appendix that might be added to ASME Section III. The procedures in the Appdix are consistent with the emdronmental correedon factor approach outlined in Section 4 of the report. The piping or vessel stress analysts will then have a choice of either using the procedures outlinN in this non-mandatory Appendix or using equivalent procedures to correct the Code CUF for ,

l ificant.

environmental effects, where such effects are judged si m, 9

4 62 -

EPR!Lionnend M:terial j

' i 4

SUMMARY

e Pressure-retaining components in the light water reactor (LWR) primary systerr.s are j '

designed to meet the requirements of Section III of the ASME Boller and Pressure Vessel Code or an equivalent Code. The Class I rules of Section III requim a fatigue evaluation for the transient stresses that ocotr during normal / upset condition operation. W-fatigue design curves in Section III are based on the cyclic life observed in strain-con-trolled fatigue tests ccaducted in the air environment and include either e factor of two on stress or 20 on cycles over the mean curves. The effects of high-temper = cure reactor water environment were not explicitly considered, although a factor of 4 in the factor of

20 on cyclic life was attributed to atmosphere.

Although there have been relatively few corrosion fatigue failures in materials typically used in the LWR applications, the laboratory data generated in various test programs

(e.g., EPRI-sponsored testing at GE, NRC-sponsored testing at Argonne, and testing conducted in Japan) indicate that fatigue lives shorter than the Code design values are possible, especially under low-frequency loading conditions in oxygenated water envi-

ronments at elevated temperatures.

The laboratory testing by Argonne and in Japan identified strain rate, temperature, strain amplitude, and oxygen content as significant variables affecting fatigue-initiation life. The laboratory testing has generally been with or:e or a combination of significant variables held at a specified fixed value during the test. However, during the course of a j

typical plant trans!ent, generally, the temperature and the strain rate are continuously varying but the stress aaalyses are not detailed enough to evaluate the values of these

[ variables. Furthermore, it is unreasonable to burden the piping or vessel stress analysts

['

to require such a detailed evaluation. Therefore, there is a need for simplified, but not ovt rly conservative, procedures for ASME Section III, NB 3600- and NB-3200 type anal-w in which reactor water environment effects need to be accounted for. The basic cpp. h used in thineport was to take the existing fatigue usage and multiply it by an erwi .nental correction factor F,,, to obtain a new fatigue usage reflecting the em' iron-mentiu effects.

A review of the available laboratory test data and previous studies indicated that, l currently, two ap > roaches are available: the Higuchi-Ilda approach and the NUREG/

CR-5999 approach proposed by Argonne. The Argonne also proposed a statistical char.seterization in tha form of a mathematical expression for the cyclic fatigue-initiation

, life with significant variables, such as strain rate and temperature as the parameters.

'Ihe materials covered were carbon and low-alloy steels, stainless steels, and Alloy 600 .

These expressions were recast in this report to produce mathematical expressions for 7-1

EPRI Liceved Mcterial Summary F,. This represented the third approach. A comparison of the F, values predicted by the three approaches as a function of dissolved oxygen in reactor water showed that the major differences are in the dissolved oxygen range of 0.2 to 0.5 ppm. A n. view of recent data from Japan and earlier EPRI sponsored testing of blunt notch CT specime is indicated that precUetions based on the A:gonne statistical characterization are more consistent with this data. Therefore, F,, factors developed in this report from the Argonne statistical characterization were used for subsequent implementation in the .

NB-3600 and ND-3200 fatigue analyses. The threshold values of significan; parameters ds eloped by the PVRC committee were also discussed and adopted.

The most important parameter was identified as the strain rate for a load state pair. For the NB 3600 fatigue evaluations, an approach was described to determine the strain rate. Also, an effective damage approach was described which removes conservatism in the calculated value of F,. The proposed approach was applied to several example cases such as feed water piping, recirculation piping, feedwater safe end and surge line. The results indicated that there is generally a modest increase in the calculated fatigue usage, which is considerably less than that resulting from using the NURG/CR-5999 interin sves.

The report also describes the proposed changes in Section III fatigue-evaluation proce-dures that might provide anMysts with enabling words to refer to a non-mandatory appendix to account for envirm nental fatigue effects. An example of the non-manda-tory appendix is included as e. appendix to this report.

7-2

EPRILicensed M:terial Non-mandatory Appendix X FATIGUE EVALUATIONS INCLUDING ENVIRONMENTAL EFFECTS X-1000 Scope This appendix provides methods for performing fatigue usage factor evaluations of reactor coolant system and primary pressure bondary components when the effects of reactor water on fatigue-initiation life are judged to be significant.

X-1103 Environmental Fatigue Correction The evaluation method uses as its input the partial fatigue usage factors U,, U,, U,,

.....U,, determined in Class I fatigue evaluations. In the Class I design-by-analysis proce-dure, the partial fatigue usage fattors are calculated for each type of stress cycle in paragraph NB-3222.4(e)(5). For Class I piping products designed using the NB-3600 procedure, Paragraph NB-3653 provides the procedure for calculating partial fatigue usage factors for each of the load state pairs The cumulative fatigue usage factor, Ug considering the environmental effects, is calculated as:

U, = U,*F, , + U,eF,, + U,*F,3 ... U,eFg ....+ U,*F, ,

where, Fgis the environmental fatigue correction factor for the ith stress cycle (NB-3200) orload state pair (NB-3600).

X-1200 Environmental Factor L'efinition the F, factors are to be calculated using the expressions below.

Carbon Steel

. F, = exp (+0.384 - 0.00133T - 0.554S*T*O* t *) (Eq.1)

. Low-alloy Steel F, = exp (+0.766 - 0.001337 r.S *T*O* i *) (Eq. 2)

X-1

EPRI Licensed M:terial Nonmandatory Appendix X- Fatigue Evaluations including Environmental Effects Sininless Steels Except 316NG.

F, = exp (+0.359 - 0.1316') (Eq. 3)

Type 316NG Stainless Steel F, = exp (-0.023 - 0.134 i') (Eq. 4) ,

Alloy 600 F, = 1.49 ,

(Eq. 5)

X-1300 Evaluation Procedures For some types of stress cycles or load state pairs, any one or more than one environ-mental parameters are below the threshold value for significant environmental fatigue diects. The value of the environmental fatigue correction factor F, for such types of stress cycles or load state pairs will be equal to 1.0. Article X-2000 provides procedures for threshold criteria evaluation.

The procedums for the evaluation of F, factors for design by analyds and for Class I piping products fatigue evaluations are provided in X-3000, X-1400 Nomenclature the symbols adopted in this appendix are defined as follows:

E = Young's Modulus, psi F, = Environmental correction factor applied to fatigue usage calculated using Code fatigue curves O = Oxygen content of fluid (ppm)

O* = Transformed oxygen content S = Sulfur content of carbon and low-alloy steels, weight %

S* = Transformed sulfur content S, = Alternating stress amplitude, psi T = Temperature (*C) -

T* = Transformed temperature ,

T, = Average temperature on side a during a temperature transient X-2

I

\

EPRIUcassedM:terial Nonmandatory Appendix X - Petigue Evaluations including Environmental Effects T, = Average temperature on side b during a temperature transient T, =

Sum of I T,-T,I,1 AT,I, and I AT,I for temperature transient producing -

compressive stresses at the component surface in contact with fluid T; = Metal temperature during a temperature trusient at surface in contact with fluid T, =

Sum of I T,-T,I,1 AT,I, and i AT,I for temperature transient producing tensile stresses at the component surface in contact with fluH AT, =

Linear temperature gradient through a component wall during a tempera-ture transient AT, =

Nonlinear temperature gradient through a component wall during a tem-perature transient t, = Elapsed time between the start of temperature transient and the time when T,is reached, seconds t, = Elapsed time between the start of temperature transient and the time when the metal surface in contact with fluid reaches 30(PF, seconds U, = Cumulative fatigue usage factorincluding the environmental effects U, =

Cumulative fatigue usage factor for load pair i obtained by using Code fatigue curves i = Strain rate, %/second i' = Transformed strain rate ARTICLE X-2000 l e

ENVIRONMENTAL FATIGUE THRESHOLD CONSIDERATIONS X-2000 Scope This article provides procedures for screening out types of stress cycles or load state pairs for which any one or mo:e .than one environmental parameters are below the threshold value for significant environmental fatigue effects. The value of the environ-mental fatigue correction factor F,,, for such types of stress cycles or load state pairs will be equal to 1.0.

X-3 1

EPRI Lkensed M:terial Nonmandatory Appendix X - Fatigue Evaluations including Environmental Effects X-2100 Strain Range Threshold (a) Calculate the strain range, c, assodated with a type of stress cyde or load state pair i by multiplying the alternating stress intensity S,,by 2 and dividing by the modulus of elastidty E. The value of E shall be obtained from the applicable -

design fatigue curves of Figures I-9.0.

(b) If the value of c, calculated in Step (a) for a type of stress cyde or load state pair'is less than or equal to 0.1, that type of stress cyde or load state pair satisfies the threshold criterion for strain range and the value of F,,is 1.0. No further evalua-tion with respect to other threshold values need be made for this type of stress cyde or load state pair.

X-2200 Strain Rate Threshold A type of stress cyde or the load state pair that involves seismic load state satisfies the strain rate threshold criterion for strain rate and the value of F,,,,is 1.0. No further evalu-ation with respect to other threshold values need ~t e raade for this type of stress cyde or load state pair.

X-2300 Temperature Threshold

, (a) Define the effective temperature T associated with a type of stress cyde or load state pair i as equal to the higher of the highest temperatures in the two tran-sients or load states constituting the type of stress cyde or load state pair.

(b) If the temperature calculated in Step X-2300(a)is less than or equal to 300*F (or 150*C), the stress cyde or load state pair satisfies the threshold criterion for temperature and the value of F,,,,is 1.0.

X-2400 Dissolved Oxygen Threshold (a) Defme the effective dissolved oxygen content DO associated with a type of stress cyde or load state pair I as egual to the higher of the highest oxygen content in the two transients or load states constituting the type of stress cyde or load state pair.

(b) If the value of DO determined in Step X-2400(a) fc. .ype of stress cyde or load state pair is less than or equal to 0.05 ppm, that type of stress cyde or load state pair satisfies the threshold criterion and the value of F,,,,is 1.0.

X-4

EPRILicensed M:terial Nonmandatory Appendix X - Fatigue Evaluations including Environmental Effects ARTICLE X-3000 ENVIRONMENTAL FACTOR EVALUATION '

X 3100 Scope This artide provides procedures for calculating the F, factors for types of stress cydes (NB-3200) or load eate pahs (NB-3600). Only the types of stress cydes or load state pairs that do not meet the threshold criteria of X-2000 need to be considered for F, calculation.

X-3200 Evaluation Procedure for Design By Analysis X-3210 Determination of Transformed Strain Rate The strain rate (%/sec) for a stress cyde is determined as:

t = S,, e100/Eet,

, is the stress difference range for cyde i as determined in NS-3224.4(e)(5) where, and the t_S,is the time in seconds when the stress difference reaches a maximum fro the start of the temperature transient. This calculation is performed only for the step-down temperature trantsent in the stress cydes constituting a pair. The transformed strain rate e'is obtained as:

l' = 9 (i > 1%/sec) t* = In(t) (0.001 s i s 1%/sec) t' = In(0.001) (t < 0.001%/sec)

X-3220 Determination of Transformed Temperaturn The temperature T associated with a type of stress cyde i as equal to the higher of the highest temperatures in the two transients constituting the type of stress cyde. The transformed temperature T* is obtained as:

T* = 0 (T< 150*C)

T* = T-150 (T> 150*C)

X-3230 Determination of TransformedDO for Carbon and Low alloy Steels The effective dissolved oxygen content DO associated with a type of stress cjde iis equal to the higher of the highest oxygen content in the two transients constituting the type of stress cyde. The transformed DO, O* is obtained as follows:

X-5

EPRILicensedM:tental

Nonmandatory Appendix X - Fatigue Enluations including Environmental Effects 1

O* = 0 (DO<0.05 ppm) l O' = DO (0.05 ppm s DO $ 0.5 ppm) i O* = 0.5 (DO > 0.5 ppm) ,

X3N0 Determination of Transformed Sulttor for Carbon and LownalloySteele l l .

The sulfur content S in terms of weight percent migh be obtained from the certified material test report or an equivalent source. If the sulfur coritent is unknown, then its value will be ass.uned as 0.015%. The transformed sulfur S* is obtained as:

S* = S (0<S<0.015 wt%)

l S* = 0.C15 (S>0.015 wt%)

)

X4360 Determination of F,,,

The environmental correction factor F,,,, for a type of stress cycle and the cumulative

[ fatigue usage factor will be calculated using equations given in X-1200.

t

. X4200 Determination of F,,, Bened on Demepe Appronoh -

! Procedure similar to that described in X-3660 may be used to remove some of the con-servatism built into the F,,,, determined in X-3250.

L X-3600 Evaluation Procedure for Piping

[

t X-3610 GeneralRequiremente I

The procedures in this article use the input hformation and the partial fatigue unge l results from the NB-3650 fatigue evaluation. The example of specific load state info ma- .

l tion needed is 'ntemal pressure and the three moment components, I T,-T,I , AT, and  !

_ AT,. When the detailed results of one-dimensional transient heat transfer analyses are I

available in the form of time history of i T,-T,l, AT,, and ATysuch results might be used to reduce conservatisms in the calculated values of environmental correction factor, l

X 3610 Determine %on of Transfiormed Strain Rate

, The strain rate (%/sec) for a stress cycle is determined as:

t = 2*S,,,, *[T,/(T, + T,)] Meet,)

i. where S,,,,is the alternating stress intensity for load state pair I calculated in NB-36533. .

This calculation is performed only for +'ae step-down temperature transient in a load state pair. The transformed strain rate i* is obtained as:

I X-6 l

EPRILicensed M:terial Nonmandatory Appendix X-F tigue Evaluations including Environmental Effects i* = 0 (t > 1%/sec) t' = In(e) (0.001 s t s 1%/sec) t* = tn(0.001) (t < 0.001%/sec)

X4620 Detenminadon of Transformed Temperature The temperature T associated with a load state pair ; as equal to the higher of the high-est tempe.catures in the two transients constituting the load state pair. The transformed temperature T* is obtained as the following: -

T*= 0 (T< 150*C)

T* = T-150 (T> 150*C)

X-wS30 Determinadon of Transformed DO for Carbon and Lowalloy Steels The effective dissolved oxygen content DO associated with a load state pair iis equal to the higher of the highest oxygen content in the two iransients constituting the load state pair. The transformed DO, O'is obtained as:

O* = 0 (DO<0.05 ppm)

O* = DO t0.05 ppm s DO s 0.5 ppm)

O*=0.5 (DO > 0.5 ppm)

X4640 Determinadon of Transformed Sulfur for Carbon and LowmnlloySteels The sulfur conten' S in terms of weight percent may be obtained from the cerdfied material test report or an equivalent source. If the sulfur content is unknown, then its value will be assumed as 0.015%. The transformed sulfur, S* is obtained as follows:

S* = S (0<S<0.015 wt%)

Sh 0.015 (S>0.015 wt%)

X4650 Determinadon of F, The environmental correction factor F, will be calculated using equatiom given in X-1200.

X-7

EPRI Ucensed M:terial Nonma~tdatory Appendix X - Fatigue Evaluations including Environmental Effects X-3660 Determination of F,,, Based on Damage Approach When the detailed results of one-dunensional transient heat transfer analyses are avail-able in the form of time history of I T,-T,I, AT, and AT,, such results may be used to reduce conservatisms in the calculated values of F,. The following expression or I

equivalent shall be u. sed:

Fa = (1AQ litJexp(+0.384-0.0013ST,, + 0.554S*T,,*O* t,*)Jd1 The preceding value of F, may be used in lieu of the F, value calculated in X-3650.

G e

X-8

ACOUT EPRI The trussion of the Electrile Power Flesearch insetufe is to discowr, clnodop, and deber high nk" inchnoooocal adenoon mou9h rwtworsdng andpartnershp unh the edsomdry industry.

Funded through annual membership dues from some 700 member utilities. EPRrs work covers a wide range of technologies related to the generation, delivery, and use of electndty, with special attention pNd to cost effectnreness and environmental concems.

At EPRl's headquarters in Palo Alto, Califomia, more than 350 scientists and engineers manage some 1800 ongoing projects throughout the world. Benefits accrue in the form of products, services, and inf ormation for direct application by the electric utility industry and its customers-M ATERI AL (continued from tront cover)

4. TERM AND TBUENATION This Icense and mis agreement are effective until terminated. You may temunate them at any time by destreymg this report EPRI has the right to terminate the license and this agreement immediately if you tallto comply wlth anyterm orcondrtion of this agreement upon anytermnabon you msy dostroythis report, but all obijgattuns of nondisclosure will remain in effect

5. DISCLAIMER OF WARRANTIES AND UMITATION OF UABluTIES NEITHER EPRI, ANY MEMBER OF EPRI, ANY COSPONSOR. NOR ANY PERSON OR ORGANIZAT10N ACTING ON BEHALF OF ANY OF THEM.

(A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED,(1) WITH RESPECT TO THE USE OF ANY INFORMATION, APPARATUS. METHOD, PROCESS OR SIMILAR fTEM OISCLOSED IN THIS REPORT, INCLUDING M ERCHAMTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. GR (II)THAT SUCH USE i 00ES NOT INFRINGE ON CR INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PARTY'S INTELLECTUAL PRO?ERTY, OR (Ill) THAT THIS REPORT IS SUITABLE TO ANY PARTICULAR USER'S CIRCUMSTANCE: OR '

(B) ASSJMES RESPONSIBILITY FOR ANY OAMAGES OR OTHER LIABILITY WHATSOEVER (INCLUDING ANY CONSEQUENTIAL DAMAGES, SVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR SELECTION OR USE OF THIS REPORT OR ANY INFORMATION, APFARATUS, METH00, PROCESS OR SiMILAR ITEM OISCLOSED IN THIS REPORT.

E. EXPORT Tha laws and regulations of the United States restnet the export and re-export of any porbon of this report, and you agree not to export or re-export this report or any related tecnncai data in any form without the appropnate United States and forstyn gowmment approvals.

7.CH0lCE OF LAW This agreement will be gavemed by the larws of tne State of Califomia as applied to transactions talang place entirely in Califomia De* ween Califomia residents,

8. INTEGRATION You have read and understand this agreement, and acknowiedge that it is the fina!, complete and ext.lusive h

Vd,7 agreement between you and EPRI conceming its sublect matter, superseding any pnor related under- Pinse1 cn standing or agreement. No warver, variation or diffe'snt terms 31 this agreement Will be enforteable against Racycled Paper EPRI urdess EPRI gives its pnor wrttten consent, signed Dy an officer of EPRI. Pnnard n USA

1 i

)

?

i e

i l ACOUT EPRI i The meston or me seeme Power aesearch insmuve is so chsoowr, mwaap, and eser high Malue sechnologost advancee thsouprs networtaing andpartnersnk wth the eiscencityindusey Funded through annual membeetp does froen some 700 member utilities, EPRI's work a

! covers a wide range of technologies related to the generation, dehvery, and use of electricity, Bumm

with special attention paid to cost effectiveness and environmental concems l At EDRI's headquarters in Palo Alto, Califomia, more than 350 scientets and engineers manage some 1600 ongoing projects throughout the world. Beneftts accrue in the form of

products, services, and inf ormation for direct application by the s',ectric utility industry and its customers. M AT E RI A L (continued from front cover) i 4. TERM AND TElWSNATION i This license and mis agreenvent are effecbw until terminated. You may terminate them at any time by

destroytng this report. EPRI has the right to terminate tne hcense and this agreement immediately if you l tallto compty with anyterm or condrbon of this ag reemenL Upon anyterminationyou msy destroythis report, but all obhgabons of nondtsclosure will remain in effect.

, 5.Ol8 CLAIMER OFWARRANTIES AND LIMITKT10N OF LIABILITES I NEITHER EPRI, ANY MEM'IER OF EPRI, ANY COSFONSOR, NOR ANY PERSON OR ORGANtZATION ACTING ON

! BEHALF OF ANY OF THEM:

l (A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER. EXPRESS OR IMPLIED,(1) WITH RESPECT l

! TO THE USE OF ANY INFORMATION. APPARATU3, METHOD, PROCESS OR SIMILAR ITEM DISCLOSED IN THIS RETORT. INCLUDING MERCHANTABILITY AND FITNESS FOR A PARTICULA.1 PURPOSL, OR (II)THAT SUCH USE 1 l 00ES NOT INFRINGE ON OR INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PAR 1YS l INTELLECTUAL PROPERTY, OR (Ill) THAT THIS REPORT IS GUITABLE TO ANY PARTICJLAR USER'S l CIRCUMSTANCE; 0R l (B) ASSUMIS RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER (,NCLUDING ANY i CONSEQUENTIAL DAMAGES. EVtN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED OF THE

POS$1BilffY OF SUCH DAMAGES) RErJLTING FROM YOUR SE' ECTION OR USE OF THIS REPORT OR ANY l l

INFORMATION APPARATUS, METHOD, PROCESS OR SIMILAR ITEM OISCLOSED IN THIS REPORT.

6. EXPORT

• The laws and reguiabons of the Ursted States restnct the export and re-export of any portion of this report,

, and you agree not to export or re-export this report or any related technical data in any form wtthout the appropnate United States and foreign govemment approvals. ,

i 7.CHolCE OF LAW 4 This agreement will be govemed by the laws of the State of Califomia as applied to transacnons talang place entuely in Cairfomia between Calttomia residents.

8. INTEGRATION You have read and understand this agreement, and acknowledQ4 that it is the final, complete and exclusnte h

M ,# l i agreement between you and EPRI conceming its subject matter, superseding any pnor related under. Pnnred on standing or agreeirW. No waiver, vanation or attterent terms of this agreement will be enforceable against Recycted Paper EPRI unless EPRI t'ves its prior written consent, signed by an oMicer of EPRI. Pnnled n USA i

- _ _ _