ML17292B658

Rev 0 to ME-02-98-04, Fracture Mechanics Evaluation of N1 Safe End.

ML17292B658
Person / Time
Site:	Columbia
Issue date:	05/29/1998
From:	WASHINGTON PUBLIC POWER SUPPLY SYSTEM
To:
Shared Package
ML17292B657	List: ML17292B656 ML17292B658 ML17292B659
References
ME-02-98-04-01, ME-02-98-04-R00, ME-2-98-4-1, ME-2-98-4-R, NUDOCS 9905130171
Download: ML17292B658 (30)

Text

SUPPLEMENTAL INFORMATION ANALYTICALEVALUATIONOF INSERVICE INSPECTION EXAMINATIONRESULTS Attachment A Calculation ME-02-98-04, "Fracture Mechanics Evaluation of Nl Safe End," Revision 0 9905i30i7i '990429 05000397 PDR ADOCK 8 PDR

BDC Page WA5HINGTON PtMIC FOWRR CALCULATIONCOVER SHEET k9 SUPPLY SYSTEM Equipment Piece No. Project Page Cont'd on Page WNP-2 Qoo 0 CI Discipline Calculation No.

MS-RPV-3 ME-02-98-04 MATERIALAND Quality Class WELDING 1 Remarks i"k":

TiUe/Subject FRACTURE MECHANICS EVALUATIONOF N1 SAFE END Purpose A fracture mechanics evaluation was performed to evaluate a planar indication found during in-service inspection of ISI weld number 24RRC(2)A-1. The indication is on the inside surface of the safe-end and Is located at 5:00 o'lock when looking downstream. The indication measures 3.52 inches in length and 0.29 inches deep in a pipe wall that is 2.0 inches thick. The indication exists in SA 336 Class F8 forged type 304 stainless steel safe-end. The size of the defect exceeds the ASME Code Section XI Table IWB 3514-2 allowable and thus requires an evaluation per paragraph IWB 3640 of the Code. The following calculation provides a comprehensive presentation of the fracture mechanics model, applied loads (stresses), and Code evaluations

/

Wwc- opc- <0 C IN// r a,a5 0 I Q.lcd& o~ ~

REV STATUS/ REVISION DESCRIPTION INITIATING TRANSMITTAL NO. F,P,ORS DOCUMENTS NO.

0 F Initial Issue PER 298-0600

/7ws

";;~X@5$~~~~N~PRVNN!'%'%Ã4o-".NN%-:K%%8PERFOR MAMCENERIFICATIOM!RECORD."'~%5~.';.ANNPAN%%%""':R~%%FN."N&%@5%%

REV VERIFIED BY/DATE PERFORMED BY/DATE APPROVED BY/DATE NO.

0 Tom Erwin 6-'3>-g 5 gv Study Calculations shall be used only for the purpose of evaluating alternate design options or assisting the engineer in performing assessments.

968-1 8645 R4 (6/98)

Page Cont'd on Page WA58lNGTON tURLlC tOWER CALCULATIONINDEX

43 SUPPLY SYSTEM Calculation No.

ME-02-98-04 Revision No.

0 ITEM PAGE NO. SEQUENCE Calculation Cover Sheet 1.000-Calculation Index 1.100-Verification Checklist for Calculation and CMR's 1.200-Calculation Reference List 1.300-Calculation Output Interface Document Revision Index 1.400-Calculation. Output Summary 2.000-Calculation Method 3.000-Sketches 4.000- Q,~ 0 9 Manual Calculation 5000- g 0 i~

APPENDICES:

Appendix A Pages Appendix B Pages Appendix C Pages Appendix D Pages Appendix Pages Appendix Pages Appendix Pages Appendix Pages 96S.25278 R2 (3/SS)

Page Cont'd On Page WA5HINGTON PlSLIC l'OWaa VERIFICATIONCHECKLIST FOR CALCULATIONSAND CMRs /,3QQ

.. ~3 SUPPLY SYSTEM Calculation/CMR ME-02-98-04 Revision 0 was verified using the following methods:

Checklist Below Q Alternate Calculations Checklist Item Initial Clear Statement of purpose of analysis Methodology clearly stated and sufficiently detailed and appropriate to proposed application Logical consistency of analysis

~ Completeness of documenting references

~ Completeness of documenting and updating output interface documents Completeness of input Accuracy of input data Consistency of input data with approved criteria Completeness in stating assumptions Validity of assumptions Calculation sufficiently detailed

~ Arithmetical accuracy

~ Physical units specified and correctly used Reasonableness of output conclusion Supervisor independency check (If acting as Verifier)

- Did not specify analysis approach

- Did not rule out specific analysis options

- Did not establish analysis inputs

~ If a computer program was used:

- Is the program appropriate for the proposed application?

- Have the program error notices been reviewed to determine If they pose any limitations for this application?

- Is the program name, revision number and date of run inscribed on the output?

- Is the program identified on the Calculation Method form?

If so, is it listed in chapter 10 of the Engineering Standards Manual?

Other Elements Considered

~ If a separate verifier was used for validating these functions or a portion of these functions, sign and'initial below.

Based on the foregoing, the calculation is adequate for the purpose intended.

Verifier Signature s)/Date Verifier Initials 968-2528O R1 (3I98)

PAGE CONT'D ON CALCULATIONREFERENCE LIST 00 PAGE Qg SUPFL~ SYSTEM CALCULATION NO o ME-02-98-04 REVISION NO.

0 ISSUE DATEg SEQUENCE DOCUMENT NO.

AUTHOR EDZTIONi TITLE NO ~

OR REVZSZO Failure 2.23 NASCRAC Manual Analysis Associates Supply System 5-14-98 Ultrasonic Examination Data R-R13-031 Sheet EPRI 1986 Evaluation of Flaws in NP-4690-SR Austenitic Steel Piping NRC 1988 Technical Report on Material NUREG-Selection and Processing 0313,Rev.2 Guidelines for BWR Coolant Pressure Boundary Piping ASME 1990 ASME Section XZ, Nonmandatory Fig C-3210-1 Appendix C Burns 6 Roe 5/7/76 Hanford ZZ 251" BWR Vessel T9,S9,F9 Stress Report T9,S9,F9 Recirculation Outlet Nozzle ASME 1989 ASME Section XI IWB-3640 S.T. Rolfe 1977 Fracture and Fatigue Control in J.M. Barsom Structures ASME 1986 ASME Section III, Appendices Appendix I Table Z-2.2 10 Structural March The Effect of Radiation on the SZR-97 095 Integrity 1998 Fracture Toughness of Austenitic Stainless Steel Base and Weld Material J.F.Harvey 1985 Theory and Design of Pressure pg 61 Vessels 44458 t I 0/89)

Page Cont'd On Page CALCULATIONOUTPUT IN'IXRFACE WA5BINGTON ttiaLIC?OW81 00c) Q,4 oo DOCUMENT k9 SUPPLY SYSTEM REVISION INDEX Cslculstion No.

ME-02098-04 Prepared By/Date Verified by/0 te Revision No.

Tom Etwin R 0

The below listed output interface calculations and/or documents are impacted by the current revision of the subject calculation. The listed output interfaces require revision as a result of this calculation. The documents have been revised, or the revision deferred with Manager approval, as indicated below.

CHANGED BY CHANGED DEFERRED DEPT.

AFFECTED DOCUMENT NO. (e.g., BDC, SCN, CMR, Rev.) (e.g., RFTS, LETTER NO.)

None MANAGER'equired for deferred changes only.

968-25285 R1 (3ISS)

Page Cont'd On Page WA58INGTON tulLlC tOWRR CALCULATIONOUTPUT

SUMMARY

,OOV Q OQ I> SUPPLY SYSPIM Calculation No.

ME-02-98-04 Discussion of Results Revision No. REV.

0 BAR Three computer runs were used to evaluate the indication in the N1 nozzle safe-end. The first modeled th e indication using the normal operational loads of the system.

The second model used three transients that could possibly occur in one year interval. These transients were the thermal discontinuity stress, OBE and SSE. This model was used to determine the crack growth expected from the fatigue loading at different crack depths allowing determination of when the cracking would become a significant contributor to crack growth. This allowed the determination that the crack growth would only become significant at the end of the interval selected for the next inspection.

The third model used the adjusted crack length (20:1 ratio) as required by NUREG 0313 Rev. 2 for the end of the IGSCC crack growth at R 16 as input. The required fatigue cycles for OBE and SSE were than applied to this crack dimension to determine acceptability for the interval.

The results of the computer runs are as follows:

The indication will grow to a depth of 0.983" by R 16 if IGSCC is active and the fatigue cycles are experienced In comparing the results to the 1989 ASME Section XI Code Tables IWB-3641-5 and -.6. Indication is acceptable for continued operation until R 16.

The weld will be reinspected prior to R16, see PERA 298-0600 CAP 1 PTL A149503.

Conclusions Taking into account the following conservatism's:

1. The weld residual stress distribution used is for an as welded component. The stainless steel safe-end to nozzle weld had MSIP performed on it during R 9. The distribution should be compressive at the ID.

2. The stresses are conservatively high due to the use of OBE stresses for steady state thermal. Also the pressure stress used is the hoop stress not the axial pressure stress.

3. No faucet are evident during the weld examination that would indicate IGSCC is active.

It has been determined that WNP-2 may operate until R16 before reexamination of the nozzle to safe-end weld has to occur. The evaluation demonstrates under the worst imposed loading conditions the flaw meets the acceptance criteria of the ASME Section XI IWB-'3641-5 and 3641-6. The main fracture mechanism that will propagate the flaw is intergranular stress corrosion cracking. If the IGSCC phenomena. is active the indication will increase in depth to 0.983 by R16. which is less than the ASME Code allowable.

968-1 8652 R2 (3/98)

Page Cont'd Qn Page WA5HINGTON tUSL1C toWaa O'.AT,C".T JT,ATTON MFTROD oO 7.oo I 4J SUPPLY SYSTEM Calculation No.

ME-02-98-04 Prepared By/Date Verified by/D e Revision No.

T.M.Erwin 0 Analysis Method (Check appropriate boxes)

H Manual (As required, document source of equations in Reference List)

H Computer Main Frame H Personal H In-House Program Q Computer Service Bureau Program Q BCS Q CDC Q PCC Q OTHER H Verified Program: Code name/Revision NASCRAC 2.23 Q Unverified Program: Document in Appendix B Approach/Methodology Flaw Evaluation Problem During the performance of Inservice Inspection of the reactor vessel RRC A loop an indication was

. discovered in the heat affected zone of the 24 inch RRC suction nozzle (N1A) to safe-end weld 24RRC(2)A-

1. The indication is on the inside surface of the safe-end and is located at 5:00 o'lock when looking downstream. The indication measures 3.52 inches in length and 0.29 inches deep in a pipe wall that is 2.0 inches thick. The indication exists in ductile SA 336 Class FB forged type 304 stainless steel. The design minimum wall based on faulted pressure is 1.01 inches. The remaining ligament in the safe-end is 1.71 inches.

The indication has existed for some time. Due to changes in the ultrasonic techniques and technology the ability to detect material variations and conditions has increased. An example of this increase in sensitivity.

is demonstrated in this examination. The same weld was examined during the R9 outage and no indication was detected at that time. However, using the new GE ultrasonic data system the same data tape was reviewed from the R9 outage and it was determined that the same indication existed at that time. The new R13 data output and the R9 data output were compared and the indication shows no change in depth or length that is not within the inaccuracies of the equipment. The indication has been in the system since at least R9 with no change in depth or length.

The indication is required to be evaluated as an IGSCC indication even though it shows no IGSCC charactreistics.

Flaw Evaluation The linear indication was evaluated using the NASCRAC computer code developed by Failure Analysis Associates. This code uses stress field influence functions as the basis for flaw propagation. The NASCRAC model selected is a shell element containing an elliptically shaped circumferential flaw. The model is identified as 703 in the NASCRAC manual. This particular model includes three crack growth degrees of freedom encompassing the respective circumferential and crack depth coordinates. The evaluation was performed using conservative linear elastic fracture mechanics principles.

Page Cont'd On Page WA58INGTON Pl/5LIC POW51 CALCULATIONMETHO 7~~l g. 80&

Calculation No.

'lY9 SUPPLY SYSHM CONTINVATIONPAGE f='- OP ?P-o Y Revision No.

All Models The maximum fracture toughness used for the stainless steel material was 150ksi ~in . The value is conservative and is approximately one half of the fracture toughness value that is achievable for this type of stainless steel product form. (BWRVIP Report SIR-97-095) (10)

Load Combinations The load combinations used in this evaluation are provided in section 5.0 of this calculation. The following provides the combinations used by each of the models.

N1IGSCC.IN The IGSCC calculation for normal operation was performed using 11 node points from the I.D. to O.D. For each point Kmin was calculated by setting the stress value equal to zero. Kmax was determined by conservatively combining the weld residual stresses, circumferential pressure stress, deadweight and the OBE stress that includes thermal (see section 5.0). The number of cycles used is 24 since the Paris equation crack growth law is in in/hour. One load block represents 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> or one day.

N1 FAT.IN The fatigue models also used 11 nodes from I.D. to O.D. The models were set up using stresses N1FAT1. IN in psi instead of ksi. The Paris equation (fatigue) was established using psi instead of ksi . The Kmin for the fatigue models was calculated using the normal operational stresses used in the IGSCC model. The cyclic stresses were made up of cycles from three transients that re-present the potential cyclic loading the nozzle could experience in one year. These transients were: One cycle of thermal discontinuity, 300 cycles OBE (contains SRV) and ten cycles SSE.

968-25291 R1 (3/98)

Page conrd on page WA5HIHGIN kUaLlc FOWSR CALCULATIONMETHOD '3, UspW I> svppav svsnm CONTINUATIONPAGE Calculation No.

ME-02-98-04 Revision No.

0 The modeling applies the requirements identified in NRC Generic Letter 88-01. The flaw was eva Iuated as an intergranular stress corrosion crack using the crack growth rate equation provided in the generic letter. The weld residual stress distribution provided in the letter was also used even though the weld in question had Mechanical Stress Improvement (MSIP) performed on it in 1994. The weld residual stresses are developed from room temperature yield for 304 material (30 ksi) as the normalization stress outlined in the generic letter. The flaw aspect ratio was reviewed and compared to the requirements. of NUREG-0313, Rev. 2. The aspect ratio was determined to be 12:1 which requires correction in length as the crack grows until an aspect ratio of 20:1 is exceeded. Therefore, the final crack growth aspect ratio was corrected manually to comply with the requirements of NUREG-0313, Rev. 2. The correction for aspect ratio was performed at each Refueling outage time period based upon the computer output for the IGSCC model. These intervals were determined as follows:

R 14 will occur in approximately 290 days, with two subsequent 290 day intervals until R 16. The fiaw length and depth from the R16 corrected value was then used as input into the fatigue model, The fatigue model used one year of expected upset and faulted conditions as required by the Code to assure that the crack will remain within the Code allowable limits and NRC requirements.

Three input files were used to perform the IGSCC and fatigue evaluations. These files were:

N1IGSCC.IN IGSCC for normal operations N1FAT.IN Fatigue including one year of thermal discontinuity ( 1cycle), OBE (300 cycles), SSE ( 10 cycles)

Nl FAT1.IN Fatigue incorporating R16 corrected crack length for NRC 20:1 ratio and the same fatigue cycles for N1FAT.IN The following assumptions and inputs were used in developing each of the models.

All Models: The flaw model used was 703 for a semi-elliptical (circumferential) surface crack in a cylinder. (1)

Flaw Dimensions N1IGSCC.IN The crack used was 3.52" long and 0.29" deep. The half crack was calculated taking 3.52" and N1FAT.IN dividing it by 2 to yield 1.76". (2)

N1FAT1.IN The crack length for this model was the results of the 20:1 aspect ratio required by the NRC for IGSCC cracks. The value used is &om the crack depth for 870 days of IGSCC growth that would occur by R16. The values used in the model were a length of 17.8ss and a depth of

.0.89". The half crack was determined by dividing the length by 2 that results in a value of 8.9".

Crack Growth Laws N1IGSCC.IN The Paris equation used for IGSCC growth was that provided in NUREG-0313 Rev. 2. The (4) equation used:

359E 8(hK) 'n ksiWin-N1FAT.IN 'he crack growth rate for fatigue in BWR water environment was determined using the following N1FAT1.IN Paris equation: (3) 6.155E-18(tK) 'n psiWin N1IGSCC.IN The lK~ value used was 10.0 or 10000 for the fatigue N1FAT.IN N1FAT1.IN

+

5 IO

,./ I/e ltKLOS Ix 0 Nli H)8 180'00P 0 Lm 8 l

INCONEI. 182 BUrrER INC 10'SAFE- IS'e30 0 2 2UIRC12)A-I 2USC12)8-I 180'OOP O'DOP l8

'7/16 3/1 O. 06 CrrP) 0 MlCIZ)i-2 180'CCP 2UIRC12)8-2 O'CCP l8 37

~ 31/32'NCONEL OVERLAY 182 NO

~IIE 0 OEIA L I

21/32 NOZZI.E NOIE5

~IIELO OED L Cll B.OC>> ELHI)tllION l/IS G UI-I)9 NOZZLE '10 S)ELL HELD G 2 Ul-101 HOZZLE 10 SlFE EHO IEELD

~ ~ ~RPV 0 3 Ul 7 SSFE-EHO 10 PIPE )EKLD 5/8 0 Ul-101 SllE ftO IIF FORCI IX 0

EX)HIHEO) r t 5 Ul-119 HOZhE INHKR RlDIUS SAFE-ENO (0.020'CrUAL 0 3/0 R IS'55 CARBON CONIENr) 0 19/32 R

l. 180 I. 383 CLAOOINC

<< It) 1 29/32

~a Nl NOZZLE REFERENCESE CSI IR)CLElR CO.

SNI <6 SHf 47 SHI 48 151 RET RET 3 RET 4 ISOMEIAICS i

205 AE 023 NI NOZZLE FORGIHC Itl Slff-fHD NOZZLE XSSEIKILT FORCIIX'l 0 I/s~aI PIPE 2 I/O 9 11/16 INCONEL 82 Roor/NOr INCONEL 182 BALA)ICE

'17$ ~ Sl f os)Et RRC RRC 101-1 102-)

CllSSA I RET Rf T JSFC Ct)OK CLSSS ~ I ft>>RA I tf)ELE IXEOWA>> ttcl OSIEA S 10 79 SEOEC E/S ~ 1 fValC O

I O,t OZZLE

)HIS Cftlt)IN) IS IH)ftf%0 fOR USK IN FFKSKRTICE ftf) INSERT)LE IHSPf C IICESS PRt)CR>ttS OtLT, VSDIIHCIEOE SN'PLY SYSIEN Otte\Ieft. SA)stol ttot POKR tet)t g) 180')OLIlf PIPIHC 5 fS'iftt tt)EE Oll tctt ItkL t4tfRIOL SPEC tllI L C)L OLOC>>

EFEP-2

~ ~

It>>tt it)l C>>K SS 1fff )CORER KLD C COEPOKHI 0 tl'tfCfit 4$ ZS SS )5$ CR )E)E CL I )If tdM5 ICKHIIFICllltttDI)CRltt Slff IOO SO ))tt CL IR llf OE)tf t" tlt tftthf SO SCe CL 7 CS )If tE)tf PCCIRC SEC))tft

/w eeeeleo loo ~'l lo oeseeeee ees I/el EIAI eeo oeseees E4loeoltoo ESDlltlw, eeftitlto Soles FEEoeeo .EA . fl gpss SO SX) M~ CL 1 CS ti1 MQhf ll O' Et 7 )I )R IMEACO fOR It)C r~f Iff DEER HO RPV-105 HKT I

~ Et EASE C )CO t))LSE ~ f COSO OOOO

5-54 5 Crack Geomet Modei Library in NASCRAC Software 5.1.26 Semi-Elliptical (Circumferential) Surface Crack in a Cylinder FORTRAN Option Model Feature Variable Featured Model Index Number KRKTYP 703 Number of Degrees of Freedom KRKDOF 3 Crack Front Shape Semi-Elliptical Finite Width Effects Yes InQuence Function Yes Variable Thickness Effects IVTHIC No J-Integral Solutions No Data Input Description FORTRAN Input Input Description Variable Format Remarks Variable Thickness IVTHIC Tabular Not Applicable Initial Crack Size a1 AINITL(l) Constant a2 AINITL(2) Constant a3 AINITL(3)

Constant Body Widths t WIDTHS(1) Constant W'g WIDTHS(2) Constant ) Terminate W3 WIDTHS(3) Constant ) Analysis Only WIDTHS(4) Constant Crack Position Xc CENTER(1) Constant Yc CENTER(2) Constant Crack Orientation 4 CRKANG Constant Stress Input ~-(~) Equational Tabular

<-(~ v) Equational Tabular

~H Limits: 1 < aq+ as/aq < 20; 0.0 < aq/t < 1.0 Accuracy: approximately 10'or 0.0 < a,/t < 0.8 and 1 < a2+ as/a> < 12 HANItNOTOR fl&lICPO'N40 6$ 5UBKY SYFHM CALC: -Ov 8 Version 2.2 PAGE: . CSPl'sou BY: DATE: S VERIFIED. DATE

5.1 Libr of K- and J-Solutions - Model Descri tions 5-55 703 Z

cr (x,y)

I I

I X a2~ /

I t r a) X I

I a I

NASCRAC User's Manual Version 2:2

towaa Pago Cont'a On wA$HINGToN tuaLIG 4$ SUPPLY SYSTIM MkbtUALCALCULATION .00 Pago 5. CI g~

Calculation No.

Ptap d 8 /Oat Vetlliad By/Data

-e -o/

Revision No The purpose of the calculation is to determine the bounding stress in the Recirculation outlet nozzle N1 at safe end to nozzle weld.

Actual loads at the nozzle due to the pipe are lower than the allowable loads provided in the reference documents listed below. Actual pipe loads are available in calculation 8.14.107.

References:

n

1. Hanford II -251 BWR Vessel Stress Report Sections T9,S9,F9 Recirculation Outlet Nozzle.

2. Drawing 732E143, Purchase part Reactor VesselMPL item No.

B13-D003

3. Drawing 761E716, Reactor Vessel Loadings Recirculation Outlet:

Maximum Allowable Nozzle Loads for Evaluation:

H (kips) M (inch kips)

Design Mech. Load 0.0 5850 Dead Wt. 58.50 1580 Seismic Pri 164 2950 Seismic RFE 164 2950 Thermal RFE 292 7020 The above moments are. applied at the end of the safe end. The weld of concern is the safe end to nozzle weld which is 9.75 inches+/- 1/16 inch from the load application point.

Nozzle Design Pressure: 1250 psi, Nozzle Faulted Pressure: 1375 psi Nozzle Loads for Recirculation Outlet Nozzle from Calculation 8.14.107 which includes power uprate and snubber optimization of the recirculation piping.

Condition Force - Ibs Moment - inch kips Primary 5552 167.408 Secondary 34431 1805.391 Primary (Faulted) 25481. 1066.453 966.16694 Rl (6/93)

Pago Cont'a On waarrlrrGTorr ttraLIc towaa MAMJALCALCULATION .ao g 4$ suptLv sysrmr Calculation No Pago 5 DQ Praparad 8 /Dat Varifiod By/Data

- eh'-ov Aevi~ ion No.

0 Safe end material is SA-336 F8 Sm pa 16.65 ksi O575 F Z:= 9.75 inch Z is the offset distance from the application point of the loads.

Pd:= 1250 psi OD:= 25.5 inch Pf pa 1375 psi ID:= 21.6876 inch I .= 9896 in" mom A no~:= 141.292 In Calculate tangential Pressure Stresses using the thickwall formula from Theory'nd Design of Pressure Vessels, J.F. Harvey, 1985 pg 61 ID OD b r pa 10.84, 11.2.. 12.75 inch a Pd b2 Op(r) pa 1+

b - a r2 Pressure Stress Variation with radius from ID to OD is shown below.

a p(r) psl 10.84 7.79 10 11.2 7.504 10 11.56 11.92 7.244 10 12.28 7.007.10 12.64 6.791 10 ~

6.594 10 Since the range variable for r did not exactly match the outside diameter the following equation adjusts to the exact outside radius.

r pa 12.75 968-'l 8694 Al i6/93)

Pd go Cont's On IS sUPPLY sys'BM MAI'AJALCALCULATION 0 Calculation No.

Paga g Vari/lad By/Data I= Rh'-o "/

Praparad By ata Aavision No.

y2C ~/~/ f~

a Pd b2 a p(r) ps 2

1+

b2 r2 3

G I 6.536 0 PSI.

p( )

1 Thus the tangential pressure stress varies from 7800 psi to 6530 psi. This stress is tensile around the circumference of the shell. Based on the orientation of the flaw the tangential stress would not be a tensile stress for a flaw in the tangential direction.

Reset the radius to vary from lD to OD and recalculate the radial pressure stress.

r ps 10.84, 11.2.. 12.75 inch a = 10.844 b2

< pr(r) ps a Pd 1-b2 2 r2 b = 12.75 a pr(r) PSI inches

-1.253 10 10.84

-967.181 11.2

-707.494 11.56

-470.979 11.92

'254.958 12.28

-57.13 12.64 Calculate nozzle bending stresses at the safe end to nozzle weld by applying the moment plus the force times the offset to give a maximum bending moment Deadweight Loads:

Mdwt pa 167.5 + 5.552 Z in kips c ps 12.75 Mdwt = 221.632 968.18694 Rl I6/93)

4 wASNINOTON tlraLtc tow!a SUPPLY SYSHM Prepared 8y/D a MAAUALCALCULATION Verified 8y/Data Page 0

Calculation No.

--oQ- Y-o Cont'a On Page g i-.

pa+

Aewaron No zs/~8 S ~/yg Mdwt c c dWt

'=

l mom P'wt = 0.286 'si Upset Loads including Thermal The GE load combination for RPV nozzles takes the maximum of eight different combinations which include thermal, obe, obe displacements, turbine stop valve closure, srv, and srv inertia.

M pbe:= 1806 + 34.5'Z 3

M pbe = 2.142'10 in kips Mpbe c

~abc '=

l

< pbe = 2.76 ksi mom Faulted Loads:

The GE faulted loads combination does not include thermal bending on the nozzle.

Since the upset load combination includes thermal, it is conservatively added to the faulted loading without removal of the dynamic upset loads.

M sse .= 1067+ 25.5 Z+ M pbe 3

sse 3.458 10 in kip M sse'

~ sse '=

cz = 4.455 ksi l

mom 968.18694 Al {6/93)

Page Cont's On P9 S'-OOC

@ SUPPLY SYSIZM MAMJALCALCULATION Or, CelcUI ation No.

--0)- tY- 0 Prepared By/D a Verified By/Date Revision No. REV.

Y +8 B/IR Determine the discontinuity stresses due to the attachment of the stainless steel safe end to the carbon steel vessel nozzle. The vessel nozzle has a 3/8 in inconel butter on the surface and then is jointed to the safe end with an inconel weld. Thus there are three different materials to be evaluated for thermal growth.

Nozzle Forging - SA-508 CL 2 (3/4NI-1/2Mo-CR-V)

Coeffiecient of Thermal Expansion - Group A Materials at 550 F 7.34X1 0"-6in/in/F Modulus of Elasticity - 27.0X10"6 psi Safe End - SA-336 F8- (18CR - 8Ni) Group G Coefficient of Thermal Expansion - Group 9.45X1 0'-6 in/in/F Modulus of Elasticity - 25.55X1 06 psi.

Inconel Weld Metal: S8-167 N06690 (58 Ni - 29Cr - 9Fe)

Coeffiecient of Thermal Expansion - 8.13X1 0"-6in/in/F Modulus of Elasticity - 28.2X1 0"6psi Check nozzle to inconel thermal discontinuity.

27.0 10 + 28.2.10 7 Eab '= E ab = 2.76.10 psl

-6 ua .= 734 10 ub .= 8.13 10 Ta .= 550 - 70 Tb ps 550 - 70

~tdis '= Eab'a Ta ub Tb 4

a tdis = 1.047.10 psl Nozzle to safe end Check the inconel weld to safe end discontinuity.

25.5 10 + 28.2.10 7 Eab ps E ab = 2.685 10 ua .= 9.45.10 u b: 8 13.10 966.18694 R1 {6/93)

4 WA5rrlrlGTON tlraLIG toWSR SUPPLY SYSI'EM MANUALCALCULATION Varifiad By/Data Pago C ooc CaicUIatlon No.

K- oa-Cont'5 On Pago ~Q )

Praparad By/ ato -0'/'aviaion No.

Ta I= 550- 70 Tb Pa 550- 70 atdis Eab'xa Ta- abTb 4

+ tdis 1.701.10 Safe end to inconel weld.

Thus the maximum discontinuity stress is between the stainless steel safe end and the inconel weld metal.

The original vessel stress report provided calculation of the stress .

concentration factors at the locations of tapered transitions in the nozzle.

There was no stress concentration listed for the joint that we are evaluating.

Since the weld joint between the safe end and the nozzle i's a flush weld between two equivalent diameter cylinders, we can use the stress indices from a flush weld in table NB-3683.2-1.The table lists C3 as 1.0 and K3 as 1.1.

Thus for determining peak stress at the material discontinuity, the C3 and K3 indices are applied.

a d Pa 1.0'1 1'o tdis' 1 1000

+ dis 18.713 ksi Summary of Safe end to nozzle stresses:

Design Pressure Stress = 7.790 ksi Deadweight Bending Stress o dwt = 0.286 ksi Upset Primary plus Sec. Bending Stress o obe = 2.76 ksi Faulted Bending Stress, includes thermal, deadweight, obe and sse.:

0 sse = 4.455 ksi 966-16694 Ai (6/93)

4 Praparad B sUPPLY SYSTtM ata MAAUALCALCULATION Vari/iad By/Data Pago

. 00 7 Calculation No.

Cont's On Pago

--0 -P& o ~/

5,~ ~ g Ravision No. REV.

.Thermal Discontinuity Stress at the Carbon.To Stainless Steel Intersection:

dis 18'713 ksi Stresses classified as bending stresses above are based on the outer fiber stress to maximize the magnitude. Bending stress on the inner wall is obtained by factoring the stress by 10.84/12.75. Stresses though the wall thickness are linear between the minimum on the inner wall to a maximum at the outer wall.

966.t6694 Rt (6/93)

Page Cont'd On Page MANUALCALCULATIO . VIZ G.oa 43 SUPPLY SYSIXM Calculation No.

ME-02-98-04 Prepared By/Date Verified by/Date Revision No.

7.M. ENvin 0 FATIGUE CRACK GROWTH RATE BWR ENVIRONMENT (3) da/dn ~C a E a S e (hK)',n

= Material constants, C =2.0 E-19,n = 3.302 S = R-ratio correction factor = (1.0 - 0.5 Rs)"

E = Environmental factor (1.0, 2.0 and 10.0 for air,PWR, and BWR environments, respectively).

hK = Kmax - Kmin, psi in Assume R-ratio = .7

, Ca E

• S = 2,0 E-19 a 10.0 1.0 f - 0.5 (.7)s]~

= 2.0 E-18 a [1.0-0.5(.49)]

= 2.0 E-18 a [.755]

= 2.0 E-18

• 3.07

= 6.155 E-18 THEREFORE da/dn = 6.155 E-18 (dK)'or psi ~in

Page Cont'd On Page WA5HINGTON tlQLIC POWKR MANUALCALCULATIO X.oa 8 C~ oe k9 SUPPLY SYSIXM CetcUIetion No.

ME-02-98-04 Prepared By/Date VenTied by/Date Revision No.

T.M. Erwin 0 Weld Residual Stress Calculation for through wall thickness based on NuReg 0313 Rev 2 methodology. (4)

Definition of terms:

S = polynomial coefficients c = percent of through wall thickness x/t R = ratio of residual stress to room temperature yield of 30 ksi for stainless steel.

x = Point measured through wall from ID to OD.

t = Thickness of 2.00 0, = The room temperature yield strength of stainless steel 30 ksi.

cr = The calculated residual stress at location x through wall cr, + R = cr .

0.0 O.l 0.2 1.0 0.3

-6.910 0.4 S:= 8.687 i =0 4 i =0 10 G = 0.5

.480 0.6

-2.027 0.7 0.8 0.9 1.0 R,:=ps,c,

.R = f y 0/

at the% thickness, ref G above and err =30 ksi.

1.0 0.395

-0.042

-0.321

-0.457

-0.47 RES:= R 30ksi RESistheweldresidualstress

-0.385

-0.232

-0.044 0.138 027

Page Cont'd On Page WA5BINGTON tUlLlc?OWSR MANUALCALCULATIO a4 5, ]o I> sUPPLy svsrim Calculation No.

ME-02-98-04 Prepared By/Date Verified by/Date Revision No.

'T.M. Eiwin 0 Using the information developed on the previous pages the following table identifies the stresses used in performing the evaluations. The last column in Table 1 identifies the stresses used in the IGSCC calculation.

Table 1 x = in. 30 R = ksi pres.=ksi DWT= ksi OBE= ksi SSE = ksi 30'R+pressure+DWT+OBE=ksi 30 7.79 0.286 2.76 4.455 '0.83 0.2 0.395 11.85 7.79 0.286 2.76 4.455 22.68 0.4 -0.042 -1.26 7.79 0.286 2.76 4.455 9.57 0.6 -0.321 -9.63 7.79 0.286 2.76 4.455 1.2 0.8 -0.457 -13.71 7.79 0.286 2.76 4.455 -2.87

-0.47 -14.1 7.79 0.286 2.76 4.455 -3.26 1.2 -0.385 -11.55 7.79 0.286 2.76 4.455 -0.71 1.4 -0.232 -6.96 7.79 0.286 2.76 4.455 3.87 1.6 -0.044 -1.32 7.79 0.286 2.76 4.455 9.51 1.8 0.138 4.14 7.79 0.286 2.76 4.455 14.97 2 0.27 8.1 7.79 0.286 2.76 4.455 18.93

• SSE was used as input to table 4 and is a one time event of safe shutdown earthquake.

The computer Code rounds these numbers up to the nearest third decimal in scientific notation.

Table 2 contains the stresses used in developing the fatigue cycle for the thermal discontinuity stress. This occurs one time as the RRC system heats up. The minimum stress values used are the same for the lGSCC crack growth calculation for normal operation. The maximum stress is developed by conservatively adding the thermal discontinuity stress equally through wall to the normal operational stresses.

Table 2 x = in. ID-OD Thermal Discontinuity Stress ksi Stress Min) Thermal dis)ksi Stress(Max) Thermal dis)ksi 18.73 40.836 59.566 0.2 18.73 22.686 41.416 0.4 18.73 9.576 28.306 0.6 18.73 1.206 19.936 0.8 18.73 -2.874 15.856 18.73 -3.264 15.466 1.2 18.73 -0.714 18.016 1.4 18.73 3.876 22.606 1.6 18.73 9.516 28.246 1.8 18.73 14.976 33.706 18.73 18.936 37.666 The number 18.73 ksi was conservatively used instead of 18.713 ksi.

Page Cont'd On Page MANUALCALCULATIO c~ 0 Zo/

k3 sUpPLY sYSIXM Calculation No.

ME-02-98-04 Prepared By/Date Verified by/Date Revision No.

=

'.M. Etwin ze 0 Table 3 contains the stresses used in the fatigue evaluation for the upset loading (OBE). The stresses for OBE were conservatively cycled on top of the normal operating stresses that also contained the OBE stresses. For full stress reversal the minimum stresses used were calculated using the normal stresses and subtracting the OBE stress (Table 1). The maximum stress was developed using the normal stresses and adding the OBE stress. The number of cycles used in the fatigue evaluation was 300/year.

Table 3 x =in 30 R+ ressure+DWT+OBE=ksi Stress Min Fatigue OBE=ksi Stress (Max) Fatigue OBE= ksi 40.836 38.076 43.596 0.2 22.686 19.926 25.446 0.4 9.576 6.816 12.336 0.6 1.206 -1.554 3.966 0.8 -2.874 -5.634 -0.114

-3.264 -6.024 -0.504 1.2 -0.714 -3.474 2.046 1.4 3.876 1.116 6.636 1.6 9.516 6.756 12.276 1.8 14.976 12.216 17.736 18.936 16.176 21.696 Table 4 contains the stresses used in the fatigue evaluation for the faulted loading (SSE). The stresses for SSE were conservatively cycled on top of the normal operating stresses that also contained the OBE stresses. For full stress reversal the minimum stresses used were calculated using the normal stresses and subtracting the SSE stress (Table 1). The maximum stress was developed using the normal stresses and adding the SSE stress. The number of cycles used in the fatigue evaluation was 10/lifetime.

Table 4 x from I 30'R+pressure+DWT+OBE=ksi Stress (Min) Fatigue SSE=ksi Stress (Max) Fatigue SSE=ksi 40.836 36.381 45.291 0.2 22.686 18.231 27.141 0.4 9.576 5.121 14.031 0.6 1.206 -3.249 5.661 0.8 -2.874 -7.329 1.581

-3.264 -7.719 1.191 1.2 -0.714 -5.169 3.741 1.4 3.876 -0.579 8.331 1.6 9.516 5.061 13.971 1.8 14.976 10.521 19.431 18.936 14.481 23.391

I'

~k

L Cont'd On Page WASHINGTON at/aLlC PGWR1 MANUALCALCULATIO 0/

49 SUPPLY SYSTEM Calculation No.

ME-02-98-04 Verified by/Dat Revision No.

Prepared By/Date T.M. Erwin 0 Table 5 contains the crack growth adjustments made to the computer calculated values as required by NUREG 0310 Rev. 2. For IGSCC crack growth the NRC requires an aspect ratio (crack length to depth) to be a minimum of 20:1. To calculate this new length the initial value as found during R13 was first multiplied by 20 to obtain the new crack length. This was repeated for subsequent outages and by reviewing the output data for the IGSCC crack growth depth for estimated operational days between outages. R 16 was the last interval prior to R 17 when the flaw length to depth ratio would exceed 33% of the circumference. This length would require the assumption that the flaw was the entire circumference of the pipe in accordance with NUREG 0313 Rev. 2.

Therefore, the maximum length and depth used to complete the fatigue evaluation was the R 16 value of 0.89 deep and 17.8 in length.

Table 5 Outage Days Depth=in New Crack Length=in.

13 0.29 5.8 14 290 0.544 10.88 15 580 0.746 14.92 16 870 0.89 17.8 The Input file for N1FAT1.IN contains the flaw length of 17.8" and depth of 0.89". This flaw depth and length was then ran for one year of fatigue cycles due to discontinuity, OBE and SSE in accordance with ASME Code 1989 Section XI Rules. The final length was determined to be 17.81" and 0.983" deep.

These values for Section XI Table IWB-3641-5 and IWB-3641-6 are:

lf =17.81" af = 0.983" To determine the Code acceptability of the flaws Tables IWB-3641-5 and -6 are used to determine aand a,.

These are the maximum flaw depths for normal and faulted loading conditions. Acceptability is based on a<

being less than these two values. The following calculations are used in conjunction with the referenced Section XI Tables to determined aand a<.

The indication falls into what is classified as weld zone per Fig. IWB-3641-1. This requires the flaw to be evaluated using Tables IWB 3641-5 and -6. The use of these Tables requires the calculation of the defined stress ratio and the flaw length to circumference ratio to determine the allowable depth to thickness ratio. This value is used to determine the maximum flaw depth.

s Page Cont'd On Page MANUALCALCVLATION S.'o kJ SUPPLY SYSI'EM Cetculetion No.

Prepared By/Date T.M. Etwin

~ Verified by/0 t ME-02-98-04 Revision No.

0 Circumference of the nozzle is equal to 24+3.14 = 75.36" (based on a nominal diameter of 24 )

Depth / Thickness ratio = 0.983" /2.0 = .492 I< / Circumference ratio = 17.81 /75.36" = .236 NORMAL OPERATING (INCLUDING UPSET AND TEST) CONDITIONS For Table IWB-3641-5 the stress ratio is determined by the following equation:

Stress Ratio = M(P+ P, + P ) I 277 I S(From the referenced Table)

Using the previous define stresses and an M value of 1.0 (for shielded metal arc welds when OD<24')the above equation for normal operating and upset conditions is equal to:

DWT+OBE+Pressure +OBE+Thermal Discontinuity 0.286+2.76+7.79+2.76+18.73 = 32.326 ksi NOTE: OBE is added twice conservatively to bound the normal operating and thermal stresses.

Stress Ratio = 32.326 /2.77/16.65 = .701 Using the Stress Ratio and the Circumferential Ratio the allowable Depth to thickness ratio from Table IWB-3641-5 is 0.6.

Therefore the maximum flaw = 2.0 .6 = 1.2" deep since 0.983" < 1.2 The flaw is acceptable per Table IWB-3641-5 EMERGENCY AND FAULTED CONDITIONS For Table IWB-3641-6 the stress ratio is determined using a similar equation as above with the exception of the SSE stress being substituted for one of the OBE and 2.77 being replace with 1.39. P)

Stress Ratio = M(P+ Ps + P ) I 139 I S(From the referenced Table)

Therefore: 34.021 /1.39/16.65 = 1.47 Using the Stress Ratio and the Circumferential Ratio the allowable Depth to thickness ratio from Table IWB-3641-5 is 0.538 Therefore tha maximum flaw = 2.0 *0.538 = 1.076" since 0.983" < 1.076'he flaw is acceptable per Table IWB-3641-6 Conclusion The flaw meets all the Code Section XI requirements and the N1 nozzle safe-end is acceptable for use without examination until R 16.