ML071560510

Fourth Ten Year Inservice Inspection Interval Request for Relief No. 04-ON-007, Revision 1

ML071560510
Person / Time
Site:	Oconee
Issue date:	05/29/2007
From:	Brandi Hamilton Duke Energy Carolinas, Duke Power Co
To:	Document Control Desk, NRC/NRR/ADRO
References
04-ON-007, Rev 1
Download: ML071560510 (22)

Text

DukeBRUCE Duker H HAMILTON Vice President hEnergy, Oconee Nuclear Station Duke Energy Corporation ONO 1VP / 7800 Rochester Highway Seneca, SC 29672 864 885 3487 864 885 4208 fax bhhamilton@duke-energy.com May 29, 2007 U.S. Nuclear Regulatory Commission Document Control Desk Washington, DC 20555

Subject:

Duke Power Company LLC d/b/a Duke Energy Carolinas, LLC (Duke)

Oconee Nuclear Station, Unit 1 Docket Nos. 50-269 Fourth Ten Year Inservice Inspection Interval Request for Relief No. 04-ON-007, Revision 1 By letter dated May 17, 2004 Duke Energy Corporation (now Duke Power Company LLC d/b/a Duke Energy Carolinas) (Duke) submitted Request for Relief No. 04-ON-007 associated with the replacement of Steam Generators on Oconee Unit 1.

On July 6, 2006, Duke submitted Revision 1 to Request for Relief No. 04-ON-007 in order to address issues and questions raised by a Request for Additional Information (RAI) submitted by the Staff via e-mail.

This was followed by another RAI, also submitted by the Staff via E-mail, which is the topic of this letter. Since that RAI was received, Duke responded by E-mail and has participated in at least two conference calls with the staff reviewer and has consulted vendors and consultants in efforts to better understand the reviewer's questions and issues, and attempt to address those issues to the satisfaction of the reviewer.

The reviewer's questions, and Duke's responses to each, are attached as Enclosure A, to formally place this response on the docket. Enclosure B provides excerpts from a vendor calculation and provides information referenced in the Duke responses.

www. duke-energy, comr

U. S. Nuclear Regulatory Commission May 29, 2007 Page 2 If there are any additional questions or further information is needed you may contact R. P. Todd at (864) 885-3418.

Very truly yours, Site Vice President Enclosures xc w/att: Dr. William D. Travers Administrator, Region II U.S. Nuclear Regulatory Commission Atlanta Federal Center 61 Forsyth St., SWW, Suite 23T85 Atlanta, GA 30303 L. N. Olshan, Project Manager, Section 1 Project Directorate II Division of Licensing Project Management Office of Nuclear Reactor Regulation U. S. Nuclear Regulatory Commission Washington, DC 20555-0001 xc(w/o attch):

D. W. Rich Senior NRC Resident Inspector Oconee Nuclear Station Mr. Henry Porter Division of Radioactive Waste Management Bureau of Land and Waste Management SC Dept. of Health & Environmental Control 2600 Bull St.

Columbia, SC 29201

U. S. Nuclear Regulatory Commission May 29, 2007 Page 3 bxc w/att: R. L. Gill, Jr.

T. J. Coleman D. W. Peltola P. A. Wells R. K. Rhyne C. R. Frye V. B. Dixon B. W. Carney, Jr.

R. P. Todd L. C. Keith G. L. Brouette (ANII)

J. J. Mc Ardle III ISI Relief Request File NRIA File/ELL EC050 Document Control

Enclosure A REQUEST FOR ADDITIONAL INFORMATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION STEAM GENERATOR REPLACEMENT PROGRAM RELIEF REQUEST 04-ON-007 DUKE ENERGY CORPORATION OCONEE NUCLEAR GENERATING STATION UNIT 1 DOCKET NO. 50-269

1. The July 6, 2006, relief request indicates that the Code required calculation of peak stress intensity range was performed in accordance with Subparagraph NB-3653.2 and the cumulative usage factor was determined in accordance with Subparagraphs NB-3653.3, NB-3653.4 and NB-3653.5. The relief request also indicates that in cases where the cladding thickness was in excess of 10% of the combined thickness, the additional stresses were accounted for, as required by Subparagraph NB-3122.3.

a) Explain in detail how the additional stresses due to the cladding were calculated in those areas where the cladding exceeded 10% of the thickness.

Duke Response: The stresses were calculated assuming two freely expanding bodies of different expansion coefficient are connected together. The Hot Leg Riser weld is used as an example (as shown on Enclosure B, page 10.2).

b) Show how these stresses were used in the calculation of peak stress intensity.

Duke Response: The cladding stresses were combined by direct addition with other pressure, moment, etc. stresses. The Hot Leg Riser weld is used as an example (as shown on Enclosure B, page 13.1).

All the other RCS butt welds impacted by the taper issue used this same procedure to calculate and qualify stresses where the cladding exceeded 10% of the wall thickness.

c) Provide a comparison of the calculated peak stress intensity determined by finite element analysis with the peak stress intensity calculated using ASME NB-3650 procedures, as provided in the certified I

design report, at the location where the cladding exceeded 10% of the thickness.

Duke Response: The peak stress intensity values for the specific welds were not calculated in the Finite Element Analysis (FEA). The supporting FEA was used to confirm the conservatism of the stress indices used in the RCS piping stress analysis. For example, the indices for the Hot Leg to ROTSG Weld are shown on Enclosure B, pages 23 and 24. Pages 36 to 43 of Enclosure B contain the evaluation of the Hot Leg to ROTSG Weld Indices (B1, B2, C1 & C2), and show that the finite element calculated indices are smaller than the indices used in the RCS piping analysis for this weld geometry. The results of the FEA and the conservative method of linearly multiplying the adjacent components 'K' indices were the methods used to ensure the stress indices used in the RCS piping stress analysis for this weld were conservative.

The above responses were provided to the Reviewer by E-mail and he responded with comments seeking additional clarification.

Reviewer Comment on the Duke response: It is not clear to me that the calculation of peak stress intensity is conservative. In response to item c, Duke indicates that the finite element model did not address peak stress intensity at the location of excess cladding thickness. In addition, the calculation does not appear to address through wall thermal transient stresses. Duke states that conservative K indices were used in the stress calculation. However, no K index was used for the cladding stress fatigue usage calculation.

These comments are addressed individually as follows:

Comment 1: It is not clear to me that the calculation of peak stress intensity is conservative.

Duke Response: The tabulated values for the calculated peak stress indices for each of the RCS butt weld geometries are documented in Enclosure B (See Tables 5.1, 5.2 and 5.3, and Section 5.7 on pages labeled 7, 15, 23, and 24). The K indices were a direct multiplication of the separate "K" indices from the intersecting piping components. All Code required stress components (pressure, moment, etc.) were directly added to produce the resulting stress values. The calculated stresses met all Code allowable values.

Comment 2: In response to item c above, Duke indicates that the finite element model did not address peak stress intensity at the location of excess cladding thickness.

2

Duke Response: The FEA was used to confirm the conservatism of the B and C indices, based on the worst case localized minimum ferritic wall thicknesses. The FEA model was not used for as-built development of the "K" indices, since the worst case combinations of thickness, weld contour and cladding effects did not extend uniformly around the length of the butt welds, but were circumferentially localized. The FEA model was developed to conservatively evaluate the impact on the worst case ferritic weld thicknesses. Including the cladding would have improved the ID weld contours and added material to produce less conservative minimum thickness results.

Comment 3: In addition, the calculation does not appear to address through wall thermal transient stresses.

Duke Response: The thermal transient stresses were determined to be conservatively represented by the existing transient stresses.

These stresses were included into the peak stress analysis as the "14.522 ksi TRG" peak stresses. (The Hot Leg Riser weld is used as an example; see Enclosure B, page 13.1.)

Comment 4: However, no K index was used for the cladding stress fatigue usage calculation.

Duke Response: The cladding stresses were taken as shear stresses, and therefore no additional "K" factor is warranted. The B and C indices were calculated based on the worst case ferritic weld geometries, without cladding. The addition of the cladding would provide a better stress contour, and therefore it was considered to be conservative not to include the cladding in the index-confirming FEA model.

Subsequent Reviewer Comment:

Duke performed detailed finite element analyses to demonstrate that the original B and C stress indices (that are used for simplified ASME Code NB-3600 analyses) were OK given that weld geometry did not meet ASME Code requirements. Duke then asserts, without justification, that the original ASME Code K stress indices (also part of the simplified ASME Code NB-3600 analysis procedure) are conservative. Duke should have performed a complete finite element (ASME Code NB-3200) analysis of these nonconforming welds.

3

Subsequent Duke Response:

Duke justified the adequacy for not intensifying the cladding stresses in the earlier response to Comment 4 above. In addition, Duke asked the analysis vendor to review its position on intensifying the cladding stress. The analysis vendor confirmed their position that additional stress intensification was not required. In addition, Duke asked a third party vendor, knowledgeable on ASME Code stress analysis if the Duke position was appropriate. The third party vendor confirmed that the Duke approach was appropriate.

This was discussed with the NRC reviewer during a telephone conference. The NRC reviewer still disagreed with the Duke position, but stated that the NRC has performed its own stress analysis for this condition and has concluded that the results are acceptable.

Thus, the disagreement over stress intensification is inconsequential in this case because, regardless of which approach is used, both parties agree that the resultant stress values are acceptable.

2. For each of the welds covered by this relief request, indicate whether the current weld configuration can be 100% inspected in accordance with the requirements of ASME Section XI.

Duke Response: The following Reactor Coolant System welds were completed during the replacement of Steam Generators A & B on Unit 1 and were included in this relief request. Each of these listed welds received 100% coverage during the Pre-Service Inspection:

1-RC-289-7V Cold Leg 1AI 1-RC-289-8V Cold Leg 1A2 1-RC-289-3V Cold Leg 1B2 1-RC-289-4V Cold Leg 1B1 1-RC-289-6V Hot Leg 1A Riser 1-RC-289-5V Hot Leg 1A RSG Nozzle 1-RC-289-2V Hot Leg 1B Riser 1-RC-289-1V Hot Leg 1B RSG Nozzle 4

Enclosure B Response to Request for Additional Information Supporting Information The following information was extracted from Areva Calculation 32-5036882-03, in order to address the request for additional information.

The excerpts include the following pages from the calculation:

10.2 12 13.1 7 and 15 (on one sheet) 23 24 36 37 38 39 40 41 42 43

'5

EXCEAPKPI1W 3Z5034k2"03 Additional Stress Due to Cladding The ASME Code (NB-3122.1) states that no strength may be attributed to the cladding for primary stresses. NB-3122.3 states that secondary and peak stresses due to cladding may be ignored if the nominal cladding thickness is less than 10% of the total thickness of the component. Review of reference 10 shows that the minimum wall thickness is on the A loop and is 3.01". The maximum cladding thickness also exists at the A weld and is 0.643" (Ref. 10). The 0.643" is not all cladding, but is really the difference in thickness between the ferritic weld IR and the smallest cladding IR. For conservatism, this value is used as the clad thickness.

Stress in pipe wall due to cladding expansion This is calculated assuming two freely expanding bodies of different expansion coefficients are. connected together to give an averaged expansion. The averaged expansion at Trip conditions (650 F) would be calculated as:

Ea*,(a.)(th.,)

= + E, t =O.0531in/f)

(thc,) + E. (th,,)

E, where:

Ecs = 26.1 E6 psi (650 F, Ref. 5) ac = 0.0508 in/ft (650 F, Ref. 5) thes = 3.01 in (Ref. 10)

Ess = 25.1E6 psi (650 F, Ref. 1, Type 309) ass = 0.0642 in/ft (650 F, Ref. 1, Type 309) thss = 0.643 in (Ref. 10)

The membrane stress in the pipe wall is calculated as the differential expansion between the free expansion (0.0508 in/ft) and the average expansion (0.0531 in/ft) as follows:

S,,e,,, = E. (a,,,, - a,,) = 26.1E6(0.053 1-0.0508) / 12000 = 5.00ksi This will be included in the primary + secondary stress analysis (Eqn. 10).

The peak stress will be considered as that stress caused at the interface between the carbon steel and the stainless steel and is caused by the shear between the two. Therefore, the shear force per unit inch is the same as the shear stress and is calculated as:

S,,= S,,,,,,,(th) =5.00ksi(3.Olin)=15.05ldpslin= 15.05ksi The stress intensity due to a shear stress is then twice the shear stress or 30.10 ksi.

OZ.

The following page lists the Delta T1 and Delta T2 (Through Wall Gradient) stresses for Jts 9000 and 9010 (Reference 5).

The dimensions used in Reference 12 to develop the Through Wall Gradients for the Hot Leg Curved Pipe are as follows:

Outside Diameter = 43.25" Ferritic Thickness = 3.375" Cladding Thickness = 0.125" The dimensions evaluated in this analysis for the Hot Leg Curved Pipe are as follows:

Outside Diameter = 44.00" Ferritic Thickness = Loop A: 3.01" Cladding Thickness = Loop A: 0.643" As shown above, the ferritic thickness at the weld location used in the analysis is less than used in the Reference 12 analysis. The reduced thickness acts to reduce the thermal gradient across the thickness because the pipe is insulated on the OD and temperature changes are applied on the ID and because the OD of a thicker pipe is slower to respond than that of a thinner pipe. This is supported by the results of Reference 12, which show that the Hot Leg Straight Pipe (Thickness = 2.8125") has lower maximum temperature gradient ranges than the Hot Leg Curved Pipe, for each transient evaluated.

The ferritic thickness listed above is the minimum thickness for either of the two hot leg elbow weld locations. As stated above, additional thickness leads to increased Through Wall Gradient stresses, but additional thickness also leads to reduced pressure and moment stresses, which are significantly greater than the stresses due to Through Wall Gradients.

Therefore, the Delta T1 and Delta T2 used in this analysis are applicable.

Note that discontinuity stresses are not considered in this analysis because there is no change in nominal thickness or material across the weld joint.

32-5036882-02 Hot Leg Riser Evaluation Geometry Primary + Secondary Range ID = 37.98 in Pressure Moment Cladding Total Allowable Ratio Th = 3.01 in Sn 25.31 . 23.31 5.00 53.63 55.2 0.97 '/

Rad 60 in rm = 20.495 in OD= 44 in I= 81845.67 in^4 18.4 ksi @650 F ri Sm kA M=

6 25336 in-kips Peak Stress Pressure Pressure Moment Cladding TRG Total P= 2750 psi Sp = 33.41 46.16 30.10./ 14.522 124.19/

Indices Alternating Stress Allowable Cycles Actual Cycles Usage h= 0.430 Salt= 62.10 2,202/ 360 D.16 v 131= 0.100 B2 = 2.282 C1 = 1.259 C2 = 3.423 C3 = 1 C'3 = 0.5 KI = 1.32 K2 1.98 K3 = 1.87

resiicP For 3H4wU*- Gu 63 Table 5.1 Stress Indices for Hot Leg 1800 Elbow - with As-Welded Girth Butt Weld A: Jt 9000 B: Jt 9010 Stress Index Standard Longitudinal Girth Butt Elbow Butt Weld - Weld - Final Indices Flush As-Welded Indices (Note 2)

B1 0.10 - 0.10 B2 2.282 - 1.0 2.282 C1 1.259 1.0 1.0 1.259 C2 3.423 1.0 1.0 3.423 C3 1.0 - (Note 1) 1.0 C3' 0.5 - 0.5 K1 1.0 1.1 1.2 1.32 K2 1.0 1.1 1.8 1.98 K3 1.0 1.1 1.7 1.87 (1) The code specifies that, for "abutting products", the standard elbow indices should be multiplied by the girth butt weld indices except for B1 and C3'. Multiplying C3 for the elbow (1.0) by C3 for the girth butt weld (0.6) reduces the index to 0.6. This index is maintained at 1.0 for conservatism.

(2) See Assumption 4. P?7 Table 5.2 Stress Indices for Cold Leg 900 Elbow - with As-Welded Girth Butt Weld A: Jt 9001, 9002 B: Jt 9011, 9012 Stress Index Standard Longitudinal Girth Butt Elbow Butt Weld - Weld - Final Indices Flush As-Welded Indices (Note 2) (Note 3)

Bi 0.10 - - 0.10 B2 2.322 - 1.0 2.322 Cl 1.310 1.0 1.0 1.310 C2 3.484 1.0 1.0 3.484 C3 1.0 - (1) 1.0 C3' 0.5 - - 0.5 K1 1.0 1.1 1.2 1.32 K2 1.0 1.1 1.8 1.98 K3 1.0 1.1 1.7 1.87 (1) The code specifies that, for "abutting products", the standard elbow indices should be multiplied by the girth butt weld indices except for B1 and C3'. Multiplying C3 for the elbow (1.0) by C3 for the girth butt weld (0.6) reduces the index to 0.6. This index is maintained at 1.0 for conservatism.

(2) See Assumption 4.

(3) Appendix C provides additional justification for using the B1, B2, Cl, C2, K1 and K2 indices shown in this table.

5.7. Hot Leg to ROTSG Weld Stress Indices In this section, ASME Code stress indices are calculated for the as-built hot leg connection to the ROTSG for the ONS-1 loop. The original stress analysis was done in Reference 5. This analysis only considers the as-built condition for the hot leg weld location at the ROTSG. The stress analysis is performed for the 1983 ASME Code (Ref. 1) as specified in the AIS (Ref. 2).

The hot leg weld is at the ROTSG nozzle locations (Jts 134. 1134, See Reference 5 and Appendix A). The stress indices for this location are calculated per the 1983 Code, Section NB-3683.7 and NB-3683.2 and Table NB-3681(a)-1. In the original analysis, no indices were applied beyond straight pipe. In this weld evaluation, an as-welded 3:1 transition will be considered as all OD surfaces meet a 3:1 slope.

BI = 0.5,B2 =1.0 t > 0.237,C1 1.0 Since tmax is not clear and the distance over which the 3:1 slope extends is not clear, the maximum C2 and C3 are used.

C2 = 2.1,C3 = 2.0 C3'=0.6,K1 = 1.2,K2 = 1.8,K3 = 1.7 As specified in the code (Section NB-3683.2), certain indices calculated above must be increased due to the longitudinal butt weld in the spool piece as discussed in the following excerpt from the code.

"For products with longitudinal butt welds, the K1, K2 and K3 indices shown shall be multiplied by 1.1 for flush welds and 1.3 for as-welded welds. At the intersection of a longitudinal butt weld in straight pipe with a girth butt weld or girth fillet weld, the C1, K1, C2, K2 and K3 indices shall be taken as the product of the respective indices."

Table 5.3 below tabulates the indices to be used for the hot leg to ROTSG weld location.

Table 5.3 Stress Indices for Hot Leg to ROTSG As-Welded 3:1 Transition A: Jt 134 B: Jt 1134 3:1 Transition Longitudinal Stress Index - As-Welded Butt Weld - Final Indices Flush Indices (Note 1) (Note 2)

B1 0.50 - 0.50 B2 1.0 - 1.0 Cl 1.0 1.0 1.0 C2 2.1 1.0 2.1 C3 2.0 - 2.0 C3' 0.6 - 0.6 K1 1.2 1.1 1.32 K2 1.8 1.1 1.98 K3 1.7 1.1 1.87 (1) See Assumption 4.

(2) See Appendix D and Section 4 for justification of the final indices used.

(taq

Appendix D: Evaluation of Hot Leg to ROTSG Weld Indices The purpose for Appendix D is to verify the indices for an as-welded butt weld with 3:1 tapers on both sides of the weld (depicted in Reference 3) whose specific weld geometry is not covered by the 1983 ASME Code Stress Indices in Table NB-3681(a)-1 or paragraph NB-3683.5.

Furthermore, this calculation validates the stress indices used in the piping calculation, or show the conservatism thereof. The method consists of obtaining the stress indices factors when subject to two cases including 1) internal pressure and 2) a bending moment for the specific geometry associated with the Oconee Hot Leg pipe welds (as depicted in Reference 3).

For the internal pressure case, a two dimensional axi-symmetric finite element model with only internal pressure applied was built, and the stress indices were obtained by calculating the ratios of the 'membrane' stress to the 'Code Pressure Term', i.e. ,D° and the 'membrane-plus-2t bending' stress intensity to the 'Code Pressure Term'.

On the other hand, for the bending moment case, a three dimensional, 180' finite element model loaded with a bending moment on the top surface was built. The stress indices were obtained by MD calculating the ratios of the 'membrane' stress to the 'Code Bending Term', i.e. mw 0-, and the 21

'membrane-plus-bending' stress intensity to the 'Code Bending Term'.

These indices are applicable to the piping stresses calculated on the minimum joint cross-section, i.e. at the centerline of the weld.

MaterialProperties The structural properties of the modeled material at various temperatures are listed in Table 1 where E is the Modulus of Elasticity (x I Or) and p is Poisson's ratio (unitless).

Temp E 70 28.3 0.3 100 28.14 0.3 150 27.87 0.3 200 27.6 0.3 250 27.3 0.3 300 27 0.3 350 26.75 0.3 400 26.5 0.3 450 26.15 0.3 500 25.8 0.3 550 25.5 0.3 600 25.3 0.3 650 25.05 0.3 700 24.8 0.3 Reference 4 Assumed Table 1: Structural Properties of Modeled Material

Two DimensionalFEM The geometry of the joint is based on Reference 3 (contour of base metal without the cladding).

The materials are as designated in Reference 4. An axi-symmetric 2-D finite element model is developed using the 7.0 Version of ANSYS (Reference 2). The element type used in modeling the assembly was axi-symmetric PLANE82 (2-D 8-Node structural solid element).

As depicted in the figure 2, the boundary conditions can be summarized as follows:

1. Internal pressure was applied on the Hot Leg Nozzle, Elbow Spool, and weld as shown in figure 2. Because the item of interest is a ratio of stress, a representative pressure value of 1000 PSI is used.

I1. The nodes at the lowest boundary of the nozzle are constrained to zero displacement in the vertical direction. The length of the pipe was determined by calculating the attenuation length in order to limit the extent of the analysis. Per Reference 1, the attenuation length is equal to 4.94r with 'r being the inner radius of the pipe and 'T being its thickness.

AREAS !AN TYPE rM NOV 25 2003 13:06:17 This figure is non-essential to this docume (for legibility concerns)

Oconee Steam G Figure 1: Weld Geometry fl.-

ELEMENTS AN NOV 25 2003 13 :03 :46 NFOR RFOR PRES-NO1RN

~~-.----------~- I.

I This figure is non-essential to this document.

(for legibility concerns)

LX

-1554 -824.366 -94.62

-2284 -1554 -824.366 -94.62 635.127 1000

-1919 -1189 -459;493 270.254 1000 Oconee Oteaem Generator Hot Leg weld Joints Figure 2: 2-D FEM Boundary Conditions and Representative Pressure AN, NOV 25 2003 13:05:50 This figure is non-essential to this document.

i938 6612 6275 6948 Figure 3: Stress Intensity Contours at Weld Region (Pressure Loading) fi A

The output file for the pressure case solution run is 'pressure.out'.

Path Line calculation was performed between two nodes across the weld region. The linearized stress through a section defined by PATH1 is listed on the following page and could also be found in the output file 'pressure path.out'. Two stress indices are obtained using the path line results.

They are as follows.

The first stress index is calculated by performing the ratio of the 'longitudinal' membrane stress to Tem,ie.PD the 'Code Pressure Term', i.e. -- (analogous to B1):

2t 2772 2772 Stress index (B1) =2772

"-25 = 0.39 (1000x42.875) 7075 The second stress index is calculated by performing the ratio of the 'Membrane-plus-Bending' stress intensity to the 'Code Pressure Term' i.e. PD-- (analogous to Cl):

2t 6325 Stress index (Cl) = 6=

7075 0.89

PRINT LINEARIZED STRESS THROUGH A SECTION DEFINED BY PATH= PATH1 DSYS= 0 1

• ANSYS - ENGINEERING ANALYSIS SYSTEM RELEASE 7.0 ANSYS Mechanical U 00218182 VERSION=INTEL NT 16:00:51 NOV 25, 2003 CP= 0.969 Oconee Steam Generator Hot Leg Weld Joints

• • • • • POSTI LINEARIZED STRESS LISTING *****

INSIDE NODE = 26 OUTSIDE NODE = 87 LOAD STEP 1 SUBSTEP= 1 TIME= 1.0000 LOAD CASE= 0 THE FOLLOWING X,Y,Z STRESSES ARE IN GLOBAL COORDINATES.

• • MEMBRANE **

Sx SY Sz SxY SYZ Sxz

-216.7 2772. 5.444. 19.51 0.000 0.000 S1 S2 S3 SINT SEQV 5444. 2772. -2 16.9 5661. 4905.

• • BENDING ** I=INSIDE C=CENTER O=OUTSIDE sx SY Sz SXY SYZ Sxz I -626.9 -1144. 24.26 -195.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 626.9 1144. -24.26 195.0 0.000 0.000 C Sl S2 S3 SINT SEQV 0 24.26 -561.6 -1209. 1234. 1069.

0.000 0.000 0.000 0.000 0.000 1209. 561.6 -24.26 1234. 1069.

I C ** MEMBRANE PLUS BENDING ** I=INSIDE C=CENTER O=OUTSIDE 0 Sx SY Sz SXY SYZ SxZ

-843.6 1628. 5469. -175.5 0.000 0.000

-216.7 2772. 5444. 19.51 0.000 0.000 1 410.2 3916. 5420. 214.5 0.000 0.000 C S1 S2 S3 SINT SEQV 0 5469. 1640. -856.0 6325. 5518.

5444. 2772. -216.9 5661. 4905.

5420. 3929. 397.1 5023. 4468.

1 C ** PEAK ** I=INSIDE C=CENTER O=OUTSIDE 0I Sx SY SZ SXY SYZ SxZ c -279.8 1322. 356.0 204.2 0.000 0.000 0 22.89 -248.2 -87.49 -90.16 0.000 0.000 I -201.2 1311. 364.4 468.6 0.000 0.000 C Sl S2 S3 SINT SEQV 0 1348. 356.0 -305.4 1653. 1441.

50.14 -87.49 -275.5 325.6 283.1 1445. 364.4 -334.6 1779. 1553.

I C ** TOTAL ** I=INSIDE C=CENTER O=OUTSIDE 0 Sx Sy SZ SXY SYZ SxZ

-1123. 2950. 5825. 28.75 0.000 0.000

-193.8 2524. 5357. -70.66 0.000 0.000 1 208.9 5228. 5785. 683.1 0.000 0.000 C S1 S2 S3 SINT SEQV TEMP 0 5825. 2951. -1124. 6948. 6047. 0.000 5357. 2526. -195.7 5553. 4809.

5785. 5319. 117.6 5667. 5449. 0.000 a &A 1w

Three DimensionalFEM A 3-D finite element model is developed using the 7.OVersion of ANSYS Reference 4 A 'plug' was used to stiffen the upper end of the model to appropriately transfer the applied moment into the pipe. The element type used in modeling the plug was SHELL63. The remainder of the model is comprised of PLANE92 elements (3-D 10-Node 3-D 10-Node Tetrahedral Structural Solid).

As depicted in the Figure 4, the boundary conditions can be summarized as follows:

1. The nodes at the lowest boundary of the pipe are constrained to zero displacement in all directions, while a bending moment is applied to the top. Because the item of interest is a ratio, a representative bending moment value of 10,000 in-lb is used. The length of the pipe was determined by calculating the attenuation length in order to limit the extent of the analysis. Per Reference 1, the attenuation length is requal to 4.9-Fr- with 'r' being the inner radius of the pipe and 't' being its thickness.

ELEMENTS A TYPE NUM* NOV 26 2003 10:24;47 U

NFOR RFOR RNOX PATH This figure is non-essential to this documei (for legibility concerns)

Oconee Steam Generator Hot Leg Weld Joints Figure 4: 3-D Finite Element Model (Moment Loading)

ANSYS 7.1 NOV 26 2003 I2:39:06 PLOT NO. I NODAL SOLUTION STEP=1 SUB -1 TIME=!

SY (AVC)

RSYS=O PowepCraphi ca EFACET= I AVRES=Mat D14X =.251E-04 SMN =-17.986 SHX =19.862

-13.781

-9.575

-5.37

-1. 165 3.041 7.246

11. 452 I 5.657 119.862 This figure is non-essential to this document.

~(for legibility concerns)

Oconee $team Generator Hot Leg Weld Joints Figure 5: Axial Stress Plot (Bending Moment Load)

The output file for the 'Moment Load' solution run is 'moment.out' A Path Line calculation was performed across the weld region. The linearized stress through a section defined by PATHI is listed on the following page and can also be found in the output file

'moment-path.out'. Two stress indices are obtained using the path line results. They are as follows.

The first stress index is calculated by calculating the ratio of the 'membrane' axial stress to the

'Code Bending' stress, i.e. MD- (analogous to B2):

21 5.16 5.16 Stress index (B2)= - =0.91

(

20000 x42..875' 2x75706 )

5.66 The second stress index is calculated by calculating the ratio of the axial 'membrane-plus-bending stress' intensity to the 'Code Bending' stress, i.e. MD0 (analogous to C2):

21 6.693 Stress index (C2) 5.66 = 1.19 6.66

PRINT LINEARIZED STRESS THROUGH A SECTION DEFINED BY PATH= P1 DSYS= 0

• • • • • POSTI LINEARIZED STRESS LISTING *****

INSIDE NODE = 320 OUTSIDE NODE = 4792 LOAD STEP 1 SUBSTEP= 1 TIME= 1.0000 LOAD CASE= 0 THE FOLLOWING X,Y,Z STRESSES ARE IN GLOBAL COORDINATES.

MEMBRANE E**

Sx SY sz SXY SYZ SXZ

-0.44S5 -5.160 -0.7509 0.6232E-01 -0.3364E-03 -0.4191E-03 Sl

-0.4446 s2 S3 SINT SEQV

-0.7509 -5.161 4.716 4.571 BENDING

• I=INS:IDE C=CENTER O=OUTSIDE Sx SY SZ SXY SYZ SxZ

-0.2872 -2.226 -0.6566 0.3001 -0.2167E-03 0.3948E-03 0.000 0.000 0.000 0.000 0.000 0.000 0.2872 2.226 0.6566 -0.3001 0.2167E-03 -0.3948E-03 si S2 S3 SINT SEQV

-0.2418 -0.6566 -2.271 2.030 1.857 0.000 0.000 0.000 0.000 0.000 2.271 0.6566 0.2418 2.030 1.857 MEMBRANE PLUS BEN[ DING ** I=INSIDE C=CENTER O=OUTSIDE SX SY SZ SXY SYZ SXZ

-0.7326 -7.386 -1.408 0.3625 -0.5531E-03 -0.2427E-04

-0.4455 -5.160 -0.7509 0.6232E-01 -0.3364E-03 -0.4191E-03

-0.1583 -2.934 -0.9433EE-01 -0.2378 -0.1196E-03 -0.8139E-03 Sl S2 53 SINT SEQV

-0.7129 -1.408 -7.406 6.693 6.374

-0,4446 -0.7509 -5.161 4.716 4.571

-0.9432E-01 -0.1381 -2.955 2.860 2.839

• • PEAK ** I-INSIDE C=CENTER O=OUTSIDE Sx SY Sz SXY SYZ SxZ 0.2720 -1.247 -0.2833 0.7273 -0.3047E-02 0.7459E-02

-0.8221E-01 0.4390 0.1069 -0.1977 -0.8487E-05 0.1497E-02 0.1735 -0.6341 -0.1372 0.2307 -0.1146E-02 -0.1091E-02 si s2 s3 SINT SEQV 0.5641 -0.2833 -1.539 2.104 1.833 0.5055 0.1069 -0.1487 0.6543 0.5711 0.2348 -0.1372 -0.6953 0.9301 0.8108

• • TOTAL ** I=INSIDE C=CENTER O=OUTSIDE Sx SY Sz SXY SYZ SxZ

-0.4606 -8.634 -1.691 1.090 -0.3600E-02 0.7435E-02

-0.5277 -4.721 -0.6440 -0.1354 -0.3448E-03 0.1078E-02 0:1522E-01 -3.568 -0.2316 -0.7133E-02 -0.1266E-02 -0.1905E-02 s1 S2 S3 SINT SEQV TEMP

-0.3178 -1.691 -8.776 8.459 7.862 0.000

-0.5233 -0.6441 -4.726 4.202 4.143 0.1525E-01 -0.2316 -3.568 3.584 3.467 0.000