ML20155H358

From kanterella
Jump to navigation Jump to search

Submits Clarification Re Selected Request for Addl Info for Review of Plant,Units 1 & 2,integrated Plant Assessment on Metal Fatigue
ML20155H358
Person / Time
Site: Calvert Cliffs  Constellation icon.png
Issue date: 11/02/1998
From: Dave Solorio
NRC (Affiliation Not Assigned)
To: Cruse C
BALTIMORE GAS & ELECTRIC CO.
References
REF-GTECI-166, REF-GTECI-NI, TASK-166, TASK-OR TAC-M99222, TAC-M99223, TAC-M99227, TAC-MA0601, TAC-MA0602, TAC-MA1016, TAC-MA1017, TAC-MA1108, TAC-MA1109, TAC-MA601, TAC-MA602, NUDOCS 9811100095
Download: ML20155H358 (5)


Text

- _ . _ . _ - _ _ _ _ _ . _ _ _ _ _ _ _

p

.fp-3f7hif y ea arg\

UNITED STATES

, g j e

NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 200e4-0001

    • ,,*/ November 2, 1998 Mr. Charles H. Cruse, Vice President Nuclear Energy Division Baltimore Gas and Electric Company 1650 Calvert Cliffs Parkway Lusby, MD 20657-47027

SUBJECT:

CLARIFICATION REGARDING SELECTED REQUEST FOR ADDITIONAL INFORMATION FOR THE REVIEW OF THE CALVERT CLIFFS NUCLEAR POWER PLANT, UNIT NOS 1 & 2. INTEGRATED PLANT ASSESSMENT ON METAL FATIGUE (TAC NOS. MA0601, MA0602, M99227, MA1016, MA1017, M99223, MA1108, MA1109, AND M99222)

Dear Mr. Cruse:

By letter dated April 8,1998, Baltimore Gas and Electric Company (BGE) submitted its license renewal application. The application references the industry's evaluation, in part, to address metal fatigue and Generic Safety issue (GSI) 166, " Adequacy of Fatigue Life of Metal Components." By letter dated September 2,1998, the staff issued a request for additional information (RAI) regarding metal fatigue. In a public meeting on October 1,1998, BGE requested clarification on certain RAls, including five questions in this RAI. BGE designated these questions as numbers 7.6(a) and (b),7.15,7.16,7.17, and 7.22 in the meeting handouts.

1 On October 14,1998, Mr. Don Shaw of BGE telephoned and indicated that BGE was also l requesting the same clarification on question number 7.6(c). Therefore, the subject five questions are designated by BGE as numbers 7.6,7.15,7.16,7.17, and 7.22.

l- The staff has evaluated BGE's request for clarification and has determined that it is more appropriate to address these questions to the industry. Thus, the staff is redirecting these questions and issuing them to the Nuclear Energy Institute (NEI) for an industry response.

Enclosed is a copy of the letter to the NEl transmitting the questions. Note that GSI-166 is /

closed for operating reactors; however, the staff has opened GSI-190, " Fatigue Evaluation of Metal Components for 60-Year Plant Life," to address fatigue for license renewal. The staff is /

evaluating GSI 190 to determine an appropriate resolution.

The staff evaluation of GSI-190 will follow the GSI resolution process described in NUREG- t 0933, "A Prioritization of Generic Safety lasues." Should it be determined that metal fatigue '

issues will have to be addressed during the period of extended operation, BGE will be required 4

to take appropriate actions in accordance with the existing GSI process. BGE is requested to continue their participation in industry activities related to GSI-190.

In its submittal of April 8,1998, BGE relies on the evaluations in Electric Power'Research Institute (EPRI) reports as the basis for the BGE position regarding GSI-190. BGE is requested to readdress that position for the period until GSI-190 is resolved, in the absence of the staff's endorsement of the EPRI reports. Although we expect timely resolution of GSI-190, your response should address the situation where GSI-190 is not resolved prior to the current

~

9811100095 981102' PDR ADOCK 05000317 m ._ .. eg hhh .b d b, UN

. l M Charles H. Cruss, Vico Pr:sidrnt  !

Nu ar Energy Division Balti ore Gas and Electric Company 1650 Ivert Cliffs Parkway Lusby, 20657-47027

SUBJECT:

CLARIFICATION REGARDING SELECTED REQUEST FOR ADDITIONAL NFORMATION FOR THE REVIEW OF THE CALVERT CLIFFS NUCLEAR ROWER PLANT, UNIT NOS.1 & 2, INTEGRATED PLANT ASSESSMENT ON METAL FATIGUE (TAC NOS. MA0601, MA0602, M99227, MA1016, MA1017, M9h223, MA1108, MA1109, AND M99222)

Dear Mr. Cruse:

By letter dated April 8, 98, Baltimore Gas and Electric Company (BGE) submitted its license renewal application. The pplication references the industry's evaluation, in part, to address metal fatigue and Generic afety Issue (GSI) 166, " Adequacy of Fatigue Life of Metal Components " By letter dat September 2,1998, the staff issued a request for additional information (RAI) regarding m tal fatigue. In a public meeting on October 1,1998, BGE requested clarification on certai RAls, including five questions in this RAl. BGE designated these questions as numbers 7.6(h) and (b),7.15,7.16,7.17, and 7.22 in the meeting handouts.

On October 14,1998, Mr. Don Shb of BGE telephoned and indicated that BGE was also requesting the same clarification on uestion number 7.6(c). Therefore, the subject five questions are designated by BGE as umbers 7.6,7.15,7.16,7.17, and 7.22. j The staff has evaluated BGE's request f clarification and has determined that it is more appropriate to address these questions to e induttry. Thus, the staff is withdrawing these questions and issuing them to the Nuclear ergy Institute (NEI) for an industry response.

Enclosed is a copy of the letter to the NEl tran mitting the questions. Note that GSI-166 is closed for operating reactors; however, the sta has opened GSI-190, " Fatigue Evaluation of  ;

Metal Components for 60-Year Plant Life," to ad ess fatigue for license renewal. The staff is l evaluating GSI-190 to determine an appropriate r olution. ,

i The staff evaluation of GSI-190 will follow the GSI re lution process described in NUREG-0933,"A Prioritization of Generic Safety issues." Sho it be determined that metal fatigue issues will have to be addressed during the period of ex nded operation, BGE will be required to take appropriate actions in accordance with the existin GSI process. BGE is requested to continue their participation in industry activities related to G l-190.

Sincerely, David L. Solon , Project Manager License Renewa roject Directorate Division of React Program Management Office of Nuclear actor Regulation Docket Nos. 50-317 and 50-318

Enclosure:

Letter to NEl cc w/ encl: See next page DISTRIBUTION: See next page DOCUMENT NAME G \ WORKING \ LEE \FATIGUEQ BGE OFFICE PDLR PDLR PDLR:SC LA:Pp1 , EMEB:BC %LR:D NAME SLittIegy DSolord Slee 6 > k PTKuo % RWessman CGkes DATE 10 0/98 10/Z798 10/d98 10/>5f98 10/ /98 10/ /bQ

, OFFICIAL RECORD COPY i

l

l

. Charles H. Cruse November 2, 1998 license term. Specifically, the staff requests that BGE discuss how they satisfy the relevant l portion of paragraph 54.29 of the license renewal rule as discussed in the statements of consideration (60 FR 22484, May 8,1995) and as described in Subsection 6.3.5, " Aging Management for Aging issues Associated with a Generic Safety issue (GSI) or Unresolved Safety issue," of Section 2.0," Integrated Plant Assessment Methodology," to Appendix A of the l BGE application. Consistent with the SOC, it is expected that BGE will" submit a technical  !

rationale which demonstrates that the CLB will be maintained until some later point in time in l the period of extended operation, at which time one or more reasonable options (e.g.,

replacement, analytical evaluation, or a surveillance / maintenance program) would be available ,

to adequately manage the effects of aging... and briefly describe options that are technically i l feasible during the period of extended operation to manage the effects of aging.. " l l

l Sincerely, OriGinalSWedBy David L. Solorio, Project Manager l License Renewal Project Directorate Division of Reactor Program Management Office of Nuclear Reactor Regulation l Docket Nos. 50-317 and 50-318

Enclosure:

Letter to NEl cc w/ encl: See next page DISTRIBUTION: See next page See orevious concurrence DOCUMENT NAME:G:\ WORKING \ LEE \FATIGUEQ.BGE OFFICE

[M LA:PDl-1 PDLR PDLR PDLR:SC EMEJ:BQ Mkft:0 NAME SLittle DSolorio Slee PTKuo R%feWman CGrimes DATE 10/27/98* 10/27/98* 10/27/98* 10/28/98* 1030/98 10/1/98 OFFICIAL RECORD COPY ggg is ofr %,

aso_.Det M ed I e l

Distribution:

HARD COPY PDLR R/F Dochet Elles PUBLIC MEl-Zeftawy DISTRIBUTION: E-MAIL:

FMiraglia (FJM) DMartin (DAM 3)

JRoe (JWR) WMcDowell(WDM)

DMatthews (DBM) DSolorio (DLS2)

CGrimes (CIG) SStewart (JSS1)

TEssig (THE) JFair (JRF)  :

GLainas (GCL) KManoly (KAM) ,

JStrosnider (JRS2) THiltz (TGH) l GHolahan (GMH) PDLR Staff SNewberry (SFN) TMartin (TOM 2)  ;

GBagchi(GXB1) FCherny (FCC1)

RWessman (RHW) SShaukat (SKS1)

RRothman (RLR) MMayfield (MEM2) l JBrammer (HLB)

EHackett (EMH1)

CGratton (CXG1) LLund (LXL) 1 JMoore (JEM) JVora (JPV)

MZobler/RWeisman (MLZ/RMW) MMcNeil(MBM)  :

SBajwa, (SSB1) l ADromerick (AXD)

LDoerflein (LTD)

BBores (RJB)

SDroggitis (SCD)

RArchitzel (REA))

CCraig (CMC 1)

LSpessard (RLS)

RCorreia (RPC) ,

RLatta (RML1) l

Mr. Charles H. Cruse Calvert Cliffs Nuclear Power Plant l Baltimore Gas & Electric Company Unit Nos.1 and 2 cc:

President Joseph H. Walter, Chief Engineer Calvert County Board of Public Service Commission of Commissioners Maryland

- 175 Main Street Engineering Division

. Prince Frederick, MD 20678 6 St. Paul Centre Baltimore, MD 21202-6806 James P. Bennett, Esquire -

Counsel Kristen A. Burger, Esquire  !

L_ Baltimore Gas and Electric Company Maryland People's Counsel l P.O. Box 1475 6 St. Paul Centre l

Baltimore, MD 21203 Suite 2102 i Baltimore, MD 21202-1631 l Jay E. Silberg, Esquire Shaw, Pittman, Potts, and Trowbridge Patricia T. Birnie, Esquire 2300 N Street, NW Co-Director Washington, DC 20037 Maryland Safe Energy Coalition P.O. Box 3S111 Mr. Thomas N. Prichett, Director Baltimore, MD 21218 NRM Calvert Cliffs Nuclear Power Plant Mr. Loren F. Donatell 1650 Calvert Cliffs Parkway NRC Technical Training Center Lusby, MD 20657-4702 5700 Brainerd Road l Chattanooga, TN 37411-4017 Resident inspector U.S. Nuclear Regulatory Commission David Lewis P.O. Box 287 Shaw, Pittman, Potts, and Trowbridge St. Leonard, MD 20685 2300 N Street, NW Washington, DC 20037 Mr. Richard I. McLean Nuclear Programs Douglas J. Walters Power Plant Research Program Nuclear Energy Institute Maryland Dept. of Natural Resources 1776 i Street, N.W.

Tawes State Office Building, B3 Suite 400 Annapolis, MD 21401 Washington, DC 20006-3708 Regional Administrator, Region l Barth W. Doroshuk U.S. Nuclear Regulatory Commission Baltimore Gas and Electric Company 475 Allendale Road Calvert Cliffs Nuclear Power Plant King of Prussia, PA 19406 1650 Calvert Cliffs Parkway NEF ist Floor Lusby, Maryland 20657

-, --e + , - . - - - - , ,, - - - . v y., , , - . , , .v- , -

AR 7.50 t UNITED STATES g

j NUCLEAR REGULATORY COMMISSION o WASHINGTON, D.C. 20666 4 001 November 2, 1998 Mr. Douglas J. Walters Nuclear Energy Institute 1776 l Street, N.W., Suite 400 I Washington, DC 20006-3708

SUBJECT:

REQUEST FOR ADDITIONAL INFORMATION ON THE INDUSTRY'S  ;

EVALUATION OF FATIGUE EFFECTS FOR LICENSE RENEWAL j 1

Dear Mr. Walters:

The NRC staff has had ongoing work regarding the fatigue life of metal components for several years, including the consideration of fatigue life in operating reactors, research by the Argonne National Laboratory (ANL) regarding environmentally assisted cracking of light water reactors, interaction with the American Society of Mechanical Engineers (ASME), and consideration of fatigue effects for license renewal. Generic Safety issue (GSI) 166, " Adequacy of Fatigue Life '

of Metal Components," raised questions about the anvironne,ial effects on fatigue that were j not reflected in design codes and the differences of the fatigue design requirements between  ;

newer and older vintage plants. SECY-95-245, " Completion of the Fatigue Action Plan," i indicated that a backfit of operating reactors could not be justified; however the staff recommended that additional evaluation be performed considering environmental effects for license renewal. The staff recently revied NUREG-0933,"A Prioritization of Generic Safety i Issues." in NUREG-0933, Supplemen. ;2, the staffindicated that GSI-166 was resolved for i operating reactors and opened GSI-190, " Fatigue Evaluation of Metal Components for 60-year i Plant Life," to address fatigue for license renewal.

The staff is evaluating GSI 190 to determine an appropriate resolution. ANL research indicates ,

that, under some circumstances, the current ASME design curves may be nonconservative.  !

The staff is also interacting with the ASME regarding the ANL results and their relationship to the ASME Boiler and Pressure Vessel Code (ASME Code) criteria. In addition, the staff has been interacting with the Nuclear Energy Institute (NEI) on a generic resolution of GSI-190. l 1

By lettePdated February 9,1998, Electric Power Research Institute (EPRI) submitted two EPRI reports dealing with the fatigue issue. These were TR-107515, " Evaluation of Thermal Fatigue Effects on Systems Requiring Aging Management Review for License Renewal for the Calvert Cliffs Nuclear Power Plant," which contains the industry's evaluation for the Baltimore Gas and Electric Company (BGE), and TR-105759, "An Environmental Factor Approach to Account for Reactor Water Effects in Light Water Reactor Pressure Vessel and Piping Fatigue Evaluations."

On March 19,1998, the staff met with the industry in a public meeting to discuss the status of the industry's evaluation and technical aspects of these two EPRI reports (Meeting Summary, dated March 30,1998). Certain aspects of these EPRI reports have also been reviewed by ANL, pursuant to the staff's request.

I

_ Enclosure

_gy//g yggg

9 Douglas J. Walters November 2, 1998 By letter dated June 1,1998, NEl submitted three additional EPRI reports providing information on the industry's evaluation of metal fatigue effects for license renewal. These were TR-110356," Evaluation of Environmental Thermal Fatigue Effects on Selected Components in a Boiling Water Reactor Plant," TR-110043, " Evaluation of Environmental Fatigue Effects for a

, Westinghouse Nuclear Power Plant," and TR-107943, " Environmental Fatigue Evaluations of Representative BWR Components." in the June 1,1998, letter, the industry contends that these EPRI reports provide (1) a technical basis for closing GSI 166 and (2) a technical basis for license renewal applicants to address GSI 166 in the interim while formal closure is in

, process. (Since GSI-190 supersedes GSI 166, these statements are considered applicable to GSI-190).

By letter dated April 8,1998, BGE submitted its license renewal application which references the industry's evaluation, in part, to address GSI 166 (GSI-190). Specifically, one of the EPRI reports, TR 107515, addresses Calvert Cliffs fatigue issues. The staff reviewed EPRI TR-

, 107515 and by letter dated September 2,1998, the staff issued a request for additional J

information (RAI) to BGE on metal fatigue which contains several questions regarding this EPRI report. In a public meeting with BGE on October 1,1998, BGE requested clarification on certain RAls, including five questions dealing with metal fatigue. It was agreed that the subject questions involve generic information regarding EPRI reports submitted by the industry, f

Based on the foregoing, the staff has determined that these five questions on fatigue, identified by BGE at the October 1,1998, public meeting, relate to current industry and staff activities and should be more appropriately addressed by the industry. Thus, the staff is redirecting these five questions from the BGE RAls and is transmitting them to NEl in the enclosure to this letter for an industry response. The staff also expanded Question 1(b) (from the language originally provided to BGE) by adding one more sentence at the end.

Because of the staff's ongoing work with GSI-190 and the importance of this issue to license renewal, a timely resolution to these issues is desirable. If you have any questions regarding this matter, please contact Sam Lee at (301) 415-3109.

Sincerely, 1

(\ q Christopher 1. rimes, Director

' License Renewal Project Directorate Division of Reactor Program Management Office of Nuclear Reactor Regulation Project No. 690

Enclosure:

Request for Additionalinformation cc w/ encl: See next page T--y ~- v-.r-"v-,1.mr v- e

  • sh 7 * -e t2 r % 0 ti e *-u e w + . v , - - - -- -- - =6 9 ----v-e ,- 'n

I 4

. Project No. 690 cc:

Mr. Denis Harrison Mr. Robert Gill U.S. Department of Energy Duke Energy Corporation NE-42 Mail Stop EC-12R Washington, D.C. 20585 P.O. Box 1006 Charlotte, NC 28201-1006 Mr. Ricard P. Sedano, Commissioner State Liason Officer Mr. Charles R. Pierce Department of Public Service Southern Nuclear Operating Co.

112 State Street 40 Inverness Center Parkway l Drawer 20 BIN B064 i

Montipelier, Vermont 05620-2601 Birmingham, AL 35242 1

Mr. Barth Doroshuk Baltimore Gas & Electric Company 1650 Calvert Cliffs Parkway Lusby, Maryland 20657-47027 I

Mr. John J. Carey Electric Power Research Institute 3412 Hillview Avenue Post Office Box 10412 Palo Alto, CA 94303 l

l I

~

1 1

REQUEST FOR ADDITIONAL INFORMATION l ON INDUSTRY'S EVALUATION OF FATIGUE EFFECTS FOR LICENSE RENEWAL l

NUCLEAR ENERGY INSTITUTE PROJECT NO. 690

1. Section 3.2.3 of Electrical Power Research Institute (EPRI) Report TR-107515 contains an evaluation of environmental effects on the Chemical and Volume Control System (CVCS) Charging inlet Nozzle using methodology developed in EPRI Report TR-105759, "An Environmental Factor Approach to Account for Reactor Water Effects in Light Water Reactor Pressure Vessel and Piping Fatigue '

Evaluations," dated December 1995. The attached evaluation summarizes the l staff's technical concerns regarding the methodology in EPRI Report TR 105759. ,

j Based on these comments, provide the following:

t (a) Discuss the impact of the current Argonne National Laboratory (ANL) l statistical correlations of environmental test data on the Calvert Cliffs fatigue evaluation.

l (b) The technical basis for the assertion that the American Society of l Mechanical Engineers (ASME) Code stainless steel fatigue design curve contains sufficient margin to accommodate moderate environmental effects.

Include a discussion of the factor required to adjust the laboratory test data for size and surface finish effects and the margin necessary to account for scatter of the test data. Also include a discussion of the effect on the margin due to potentialinconsistencies between the ASME mean curve and l

the ANL air environment data.

(c) The technicaljustification for the strain threshold values.

2. Section 4.1 of the Baltimore Gas and Electric Company (BGE) application indicates that environmental effects do not apply to the Reactor Coolant System (RCS) l components because of the low oxygen concentrations and because the RCS i

carbon steel interior surfaces are clad with stainless steel. Discuss the applicability and impact of the latest stainless steel fatigue correlation from ANL on this conclusion (see attachment).

3. Section 3.3.3 of EPRI Report TR-107515 contains an evaluation of the Surge Line l using methodology developed in EPRI Report TR-105759. Discuss the i applicability and impact of the latest stainless steel fatigue correlation from ANL on I

this evaluation (see attachment).

i .

1 Enclosure l

2-l 4. Section 3.3.3.2 of EPRI Report TR 107515 indicates that the procedure in Section 3.1.3.2 of the EPRI report was used to develop the environmental factor used in the evaluation. Indicate whether the factor was calculated based on a " standard' treatment or " weighted average

  • approach as discussed in a June 1,1998, letter from the Nuclear Energy Institute to the NRC regarding EPRI Report TR 105759. If the " weighted average" approach was used, provide the test data used to develop l the approach. include a statistical assessment of the test data scatter Compare the results of the statistical assessment with the ANL assessment contained in NUREG/CR-6335," Fatigue Strain-Life Behavior of Carbon and Low-Alloy Ferritic Steels, Austenitic Stainless Steels, and Alloy 600 in LWR Environments." On the basis of this comparison, indicate whether the use of the " weighted average
  • approach will produce an adequate margin to account for test data scatter.

l 5. Section 5.15 of the BGE application indicates that environmental effects do not l

apply to the Safety injection components because of the low oxygen concentrations and the stainless steel components materials used in fabrication of the affected piping and valve sutmmponents. Discuss the applicability and impact of the latest l stainless steel fatigue correlation from ANL on this conclusion (see attachment).

, Enclosure i

~ -

.s , - - - - - .. , . - . - . - - -

COMMENTS ON THE APPLICATION OF THE EPRI ENVIRONMENTAL FATIGUE FACTOR TO THE CALVERT CLIFFS PLANTS The environmental factor approach described in the repon is a convenient and acceptable method to incorporate the effects of LWR coolant environments on fatigue life of pressure vessel and piping steels. However, the correlations for calculating the fatigue life correction factors F n should be updated. For carbon and low-alloy steels, the dependence of F n on dissolved oxygen (DO)is not consistent with experimental data. For austenitic stainless ,

steels, the correlations do not include the effects of DO content and temperature; panicularly the effects of DO content are important because environmental effects are more pronounced in low-DO PWR environments than in high-DO water.

Another minor point, the repon makes several references to the fact that environmental factor approach gives a lower usage factor than the interim fatigue design curves of NUREG/CR-5999, implying that this difference is due to the methodology, i.e., graphical versus mathematical representation of the best-fit curve of the experimental data. The methodology l will introduce a difference if the best-fit curves used in developing the current Code design l fatigue curves are different than the best-fit curves of the present fatigue S-N data, because l the design curves not only account for the effects of environment but also small differences that might exist between the ASME mean curve and the best-fit curve of existing fatigue data.

l For carbon and low-alloy steels, because the ASME mean curves are either comparable or somewhat conservative, the two methods should yield similar results as long as the same l correlations are used in developing the design curve and the correction factors. Minor differences between the two mentioned in this report are due to the correlations used for the interim curves. For austenitic stainless steels,it is well known (Jaske & O'Donnell,1978) that the ASME mean curve is inconsistent with the existing fatigue data. Experimental fatigue lives are a factor of up to 3 lower than those predicted by the ASME mean curve.

Consequently, usage factors based on interim design curves may be significantly higher because they account for this difference. However, for austenitic stainless steels, the margin factors on life are lower than 20 and closer to 10 or 8,i.e., there is little or no safety margin to account for environmental effects. Some specific comments on the report are as follows.

SECTIONS 2.23 & 3.13: ENVIRONMENTAL EFFECTS The repon follows the methodology of EPRI TR-105759, "An Environmental Factor Approach to Account for Reactor Water Effects in Light Water Reactor Pressure Vessel and l

Piping Fatigue Evaluations," to account for the effects of reactor coolant environment on the fatigue life of components. This approach was initially proposed by Higuchi and lida (1991).

The effects of coolant environment on fatigue life are expressed in terms of a fatigue life-correction factor F n, which is the ratio of the life in air at room temperature to that in water at the service temperature. This method is also being proposed as a non-mandatory Appendix.

1 Attachment

. . . - ,_-r- - . - - -- bI .-- -- < b - w =-w-- = =~O*--- ' '-1 --- " * - - -

l

! - To incorporate environmental effects into the ASME Code fatigue evaluation, a fatigue usage

! for a specific load pair based on the current Code design curve is multiplied by the correction

! factor. The correlations for Fen are based on the statistical models developed by ANL (NUREG/CR-6335,1995). The statistical models have since been updated. The models for carbon and low-a!!oy steels were first modified (Gavenda et. al. PVP Vol. 350,1997) because i it was determined that in the range of 0.05 to 0.5 ppm, the effect of DO was more logarithmic j than linear. Recently, these models have been further optimized with a larger data base (Chopra & Shack PVP 98; also NUREG/CR 6583,1998). The models in NUREG/CR-6335

for austenitic stainless steels were based on very limited data, and have also been updated to l incorporate the effects of DO, temperature, and strain rate on fatigue life (Chopra & Smith, PVP 98). These updated models should be used to estimate Fen in LWR environments.

In addition, a set of threshold values of strain amplitude, strain rate, temperature, dissolved oxygen (DO), and sulfur content are defined for environmental effects to occur. In l NUREG/CR-6335, these threshold values were defined on the basis of experimental l i observations and trends in the existing fatigue S-N data. With the exception of strain

amplitude, the same threshold values have been included in the non-mandatory Appendix. A i threshold strain amplitude of 0.1% is proposed for both carbon and low-alloy steels as well as 4

austenitic stainless steels in the Appendix; the basis for this value is not provided. The

threshold strain should be related to the rupture strain of the surface oxide film; there is little

! data to establish this value. Limited data suggest that for carbon and low-alloy steels, the j threshold strain is =20% higher than the fatigue limit of the steel (i.e., =0.11 and 0.15%,  ;

4 respectively, for carbon steels and low-alloy steels). A threshold strain amplitude of 0.16%  !

, has been observed for austenitic stainless steels. Unless it can be demonstrated otherwise, j' these values must be adjusted for the effects of mean stress and uncertainties due to material I and loading variability, which yields threshold strain amplitude of 0.07% (21 ksi or 145 MPa) for carbon and low-alloy steels and 0.097% (27.5 ksi or 189 MPa) for stainless steels.

4 The EPRI report TR-105759 gives a different set of threshold values that represent the strain rate, temperature, and DO level which results in " moderate" or " acceptable" effects of

.j environment, i.e., a factor of up to 4 decrease in fatigue life. For example, environmental j effects on life for 0.1 ppm DO level are considered acceptable, and Fen is considered to be 1.

l Although a factor of 3 or even 4 on life appears reasonable for carbon and low-alloy steels

(Chopra & Shack, PVP 98), the EPRI report does not provide a technical basis for selecting a factor of 4 as a working definition of acceptable effects. However, this approach results in a j discontinuity at the threshold value, e.g., Fen is 1 at 0.1 ppm DO and may jump to 10 or higher at 0.105 ppm. To avoid such discontinuities, experimental threshold values (e.g.,

NUREG/CR-6583) should be used to determine Fen, then to take advantage of the conservatism in design fatigue curves, the calculated values may be divided by 3. In other words, up to a factor of 3 decrease in life due to environment is ignored in the evaluations.

This approach is being considered by EPRI.

Please note that the above approach (factor of 3 decrease in life being acceptable) is applicable only for carbon and low-alloy steels and not for austenitic stainless steels. The reason being that the current ASME Code mean curve for low-alloy steels is consistent with the existing fatigue S-N data and that for carbon steels is somewhat conservative. Thus, a factor 3 margin on life may be used to account for acceptable effects of environment.

However, the current ASME Code mean curve for austenitic stainless steels are not consistent with the existing fatigue S-N data; a margin of only 10 on life and 1.5 on stress exists 2

- .- .. .. .. _ . .- . - - -. -- --- --=- - -- --.- -' -

l .

. l i

l between the Coc'e design curve and the mean curve (Chopra & Smith, PVP 98).

Consequently, a f actor ofless than 1.5 margin on life may be used to account for acceptable effects of environment.

l l

l l

l l

l 1

l I

1 l

i i

i-l.

I l

f 1

l l b 2..

t 4

-. . . .- -.--. - . .~ -.- ~. .- ..... .

1 EXECUTIVE sum 1ARY (PAGE 2, "RESULTS")

1 .... application of the effects of reactor water environments, .... produces worst-case l environmental multipliers that are already compensated for by two existing conservatisms in Class 1 ASME Code fatigue analysis procedures -(1) the low-cycle portion of the design fatigue curve margin factor of 20 that is appropriately ascribed to moderate environmental effects, and .... "

l Please note that the factors of 20 on life and 2 on stress should not be considered as safety margins but rather conversion factors that must be applied to the experimental data to obtain reasonable estimates of the lives of actual reactor components. Although in a benign environment some fraction of the factors, e.g., a factor of 3 on life, may be available as a safety margin.

Also, fatigue tests conducted on 0.914 m (36 in.) diameter vessels with 19 mm (0.75 in.) wall l in room-temperature water at Southwest Research Institute for the Pressure Vessel Research Council (Kooistra, et al.,1964) show that =5 mm deep cracks can form in carbon and low-alloy steels very close to the values predicted by the ASME Code design curve. These results demonstrate clearly that the Code design fatigue curves do not necessarily guarantee any margin of safety.

I 4

--^ '

1 1

MEETING WITH ELECTRIC POWER RESEARCH INSTITUTE ON METAL l FATIGUE, MARCH 19,1998 The methodology and results from four studies on Environmental Fatigue Evaluations, e.g., 1 Calven Cliffs, Older Westinghouse Plants, Representative BWR Components, and Newer l l Vintage BWR Plants, were presented at the meeting. All studies essentially follow the

)

environmental factor approach described in the EPRI repon TR-105759, and used in the EPRI repon TR-107515 on evaluation of thermal fatigue effects for Calven Cliffs Nuclear Power Plant.

l The effects of coolant environment on fatigue life are expressed in terms of a fatigue life l correction factor or environmental factor Fen, which is the ratio of the life in air to that in i water. A fatigue usage for a specific load pair based on the current Code design curve is i multiplied by the conection factor. The correlations for Fen are based on the statistical l models developed by ANL (NUREG/CR-6335,1995), which also include a set of threshold values of strain amplitude, strain rate, temperature, and dissolved oxygen beyond which l

, environmental effects on fatigue life are significant. A detailed description of the EPRI methodology is given below.

Correlations Based on NUREG/CR-4335 Fen for carbon steels (css) and low-alloy steels (LASS) is expressed as css Fen = exp(0.384 - 0.00133 T- 0.554 S* T* d' O*) (1)

LASS Fen = exP(0.766 - 0.00133 T - 0.554 S* T* i* O*), (2) where the threshold and saturation values (the value beyond which the effect of environment saturates) of sulfur content S, temperature T, strain rate i, and DO content in water are defined as S* = S (0 < S 50.015 wt.%)

S * = 0.015 (S >0.015 wt.%) (3a)

T* = 0 (T <150*C)

T* = T- 150 (T 2150*C) (3b) d'=0 (i >1%/s) d' = In(i) (0.001 si 51%/s) d' = In(0.001) (i <0.001%/s) (3c)-

O* = 0 (DO <0.05 ppm)

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

O * = 0.5 (DO >0.5 ppm) (3d) 5 . .. a ..

. . . . , . . = . . . - . .

i

. F.n for Types 304 and 316 stainless steels (SSs) is expressed as l F n = exp(0.359-0.134 d*) (4) where the threshold and saturation values of strain rate i are defined as f*=0 (i >1%/s) l d'=1n(I) (0.001 si $1%/s) l 8* = In(0.001) (l <0.001%/s) (5) l l

Updated Correlations for Fatigue Life in LWR Environments The models for css and LASS were later updated (PVRC Meeting, Orlando, April 1996) because the existing fatigue S-N data indicate that in the range of 0.05-0.5 ppm, the effect of DO on life (Eq. 3d) was more logarithmic than linear. Thus, updated correlations of Fen for css and LASS are expressed as css F n = exp(0.384 - 0.00133 T- 0.1097 S* T* 4* O*) (6)

LASS F.a = exp(0.766 - 0.00133 T- 0.1097 S* T* k* O*), (7) where the threshold and saturation values of sulfur content S, temperature T, and strain rate are the same as those defined in Eqs. 3a-3c, and those of DO content are defined as O* = 0 (DO <0.05 ppm)

O* = In(DO/0.04) (0.05 <DO 50.5 ppm)

O* = In(12.5) (DO >0.5 ppm) (8d)

These correlations (Eqs. 6 and 7) have been funber optimized with a larger data base (Chopra

& Shack, PVP 1998). The differences between the optimized correlations and Eqs. 6 and 7 are minimal; the differences are essentially in estimates oflife in low-DO environments.

The NUREO/CR-6335 models for austenitic SSs (Eqs. 4 and 5) were based on very limited data. For example, n:arly all of the data in water were obtained at high temperatures (280-320*C) and high levels of DO (0.2-8 ppm). The data were inadequate to establish the dependence oflife on strain rate, temperature, or DO content, or to define the threshold and saturation values of these parameters. These models have now been updated with a larger data base (Chopra & Smith, PVP 1998). The updated correlation of F n for Types 304 and 316 SS is expressed as F.a = exp(0.935 -T* k* O*) (9) where the threshold and saturation values of temperature T, strain rate i, and DO content in water are defined as i-i T* = 0 (T <200*C)

! T* = 1 (T 2200*C) (10a)

......-,---.-.-.-...---------..;.7...-.--.-----.7-.;.y.--;----.

, _ _ - - - - ~ ' - - ~ - -- _ _ _ _ _ _ - _ _ _ _ - - _ _ - - - - - - - - - - - - - - - - - - - - -

l'=0 (i >0.4%/s) i' = In(t/0.4)

(0.0004 5 i 50.4 %/s) l' = In(0.0004/0.4)

(i <0.0004%/s) (10b)

O* = 0.260 (DO <0.05 ppm)

O* = 0.172 (DO 20.05 ppm) (10c)

Please note that Fen is greater in low-DO PWR than in high-DO environments.

The EPRI Environmental Factor Aporoach 1)

Because the current fatigue design curves are based on data obtained in room-temperature air, an environmental correction factor should be determined with resp s

room-temperature air,i.e., Fen hould be defined as ratio of the life in air at room tempera:ure to that in water at the service temperature. It will retain the margins of 20 on life and 2 on stress that are used to develop design fatigue curves from the best-fit experimental curves. In the EPRI approach, Fen is defined as ratio of the life in air to that in water both at the service temperature. The premise being that the effect of environment alone needs to be incorporated in Fen; margins of 20 and 2 in the current design curves are adequate to account for the uncertainties that arise due to other fa 2)

The correlations for Fen are based on the statistical models of NUREG/

(Eqs.1,2, and 4). As discussed sabove, Fen hould be determined from the updated correlations (Eqs 6, 7, and 9).

3)

In EPRI report TR-105759, a different set of threshold (i.e., they result in up to a factor of 3 decrease in fatigue life). For example, when all other threshold conditions are satisfied, a DO level of 0.1 ppm may result in a factor decrease in life. Therefore, a threshold value of 0.1 ppm DO is used in the eva i.e., Fen is 1 for all load pairs with 50.1 ppm DO. Although a factor of 3 on life reasonable for defining moderate or acceptable effects of environment on life of css and LASS,it can not be used for austenitic SSs. The existing fatigue S-N data for auste SSs indicate that the difference between the ASME Code design curve and best-fit experimental curve is closer to margins of 10 on life and 1.5 on stress than the 20 and 2 originally intended. Also, care should be taken to avoid taking credit for this factor factor of up to 3 increase in CUF may be considered as " ac environment.

4)

The existing fatigue S-N data can notjustify a threshold value of 0.1% for strain amplitude, particularly for css and LASS.

7

e v

l' = 0 (t >0.4%/s) l' = In(i/0.4) (0.0004 5 i 50.4 %/s)

, l' = In(0.0004/0.4) (i <0.0004%/s) (10b) i O* = 0.260 (DO <0.05 ppm) l O* = 0.172 (DO 20.05 ppm) (10c) l Please note that Fen is greater in low-DO PWR than in high-DO environments.

The EPRI Environrnental Factor Aporoach 1)- Because the current fatigue design curves are based on data obtained in room-temperature air, an environmental correction factor should be determined with respect to room-temperature air, i.e., Fenshould be defined as ratio of the life in air at room temperature to that in water at the service temperature. It will retain the margins of 20 on life and 2 on stress that are used to develop design fatigue curves from the best-fit experimental curves. In the EPRI approach, Fen is defined as ratio of the life in air to that in water both at the service temperature. The premise being that the effect of environment alone needs to le incorporated in Fent margins of 20 and 2 in the current design curves are adequate to account for the uncertainties that arise due to other factors.

2) The correlations forFen are based on the statistical models of NUREG/CR-6335 (Eqs.1,2, and 4). As discussed above, F.a hould s be determined from the updated correlations (Eqs. 6, 7, and 9).
3) In EPRI report TR-105759, a different set of threshold values (other than Eqs. 3,8, and
10) are defined such that they result in " moderate" or " acceptable" effect of environment (i.e., they result in up to a factor of 3 decrease in fatigue life). For example, when all other threshold conditions are satisfied, a DO level of 0.1 ppm may result in a factor of 3 decrease in life. Therefore, a threshold value of 0.1 ppm DO is used in the evaluations, i.e., Fen is 1 for allload pairs with 50.1 ppm DO. Although a factor of 3 on life appears reasonable for defining moderate or acceptable effects of environment on life of css and LASS, it can not be used for austenitic SSs. The existing fatigue S-N data for austenitic SSs indicate that the difference between the ASME Code design curve and best-fit experimental curve is closer to margins of 10 on life and 1.5 on stress than the 20 and 2 originally intended. Also, care should be taken to avoid taking credit for this factor twice, e.g., after eliminating all load pairs that do not satisfy the modified thresholds, a factor of up to 3 increase in CUF may be considered as " acceptable" effect of environment.
4) The existing fatigue S-N data can notjustify a threshold value of 0.1% for strain amplitude, particularly for css and LASS.

7 l

i I

w .

. ,- .-_-- _ - _ _ ,_ ,--. m m _ _

_. _ _ _ _ _