ML20199B056

Rev 0 GE-NE-B13-01869-122, Jet Pump Riser Weld Flaw Evaluation Handbook for Hatch Unit 1

ML20199B056
Person / Time
Site:	Hatch
Issue date:	12/31/1997
From:	Caine T, Herlekar A, Mehta H GENERAL ELECTRIC CO.
To:
Shared Package
ML20199B051	List: HL-5555, Informs That Util Reviewed Attached Rev 0 to GE-NE-B13-01869-122, Jet Pump Riser Weld Flaw Evaluation Handbook for Hatch Unit 1 & Concurs W/Analysis & Conclusions ML20199B056
References
GE-NE-B13-0186-122, GE-NE-B13-01869, GE-NE-B13-186-122, GE-NE-B13-1869, NUDOCS 9801280173
Download: ML20199B056 (30)

Text

,

GENOclear Energy GE-NE.813 01869-122 Rev. 0 JET PUMP RISER WELD FLAW EVALUATION HANDBOOK FOR HATCH UNIT 1 December 1997 Prepared by:

SlW Aroon Herlekar, Principal Engineer Structural Mechanics & Materials Verified by:

Hardayal Mehta, Technical Leader Structural Mechanics & Materials Reviewed @ h h,% C,w 1246l9l Tom A. Caine, Manager Structural Mechanics & Materials 9001280173 980123 PDR ADOCK 05000321 p

PDR 1

_- - - u

GE Nuclear Energy GE NE 213-01M9122 an. o IMPORTANT NOTICE REGARDING CONTENTS OF THIS REPORT Please Raad Carefully The only undertakings of the General Electric Company (GE) respecting information in this document are contained in the contract between Southern Nuclear Operating Company (SNC) and GE, and nothing contained in this document shall be construed as changing the contract. The use of this information by anyone other than SNC, or for any purpose other than that for which it is intended, is not authorized; and with respect to any unauthorized use, GE makes no representation or warranty, express or implied, and assumes no liability as to the completeness, accuracy, or usefulness of informatio-ontained in this document, or that its use may infringe privately owned rights, r-2

GENuclear Energy

' GS NE.Cl3 0180122 Rr

~

Executive Summary A flaw evaluation, consisting of stress and fracture mechanics analyses of the Hatch Unit 1 jet pump circumferential riser welds was conducted to develop a flaw evaluation handbook.

The procedures of BWRVIP-41, were used as a guide in determining the allowable flaw lengths. End-of-cycle allowable flaw lengths were calculated at three cirrumferential weld locations. The methodology presented in this repon can be used along with consideration of observed IGSCC, and evaluation of fatigue crack growth rates to disposition any indications detected during future inspections of the jet pumps at Hatch Unit 1.

The following table shows a summary of allowable adjusted flaw lengths for Hatch 1.

Allowabl Adjusted Circumferential Flaw Sizes Weld Flaw Length (inch)

RS-1 15.42 RL2 16.69 RS-3 19.87 Observed indications less than 5.8" in length, after consideration of IGSCC growth and NDE uncertainty, do not have to be further evaluated.

3

GENuclee Energy GE.NE.813-0180122 Ro r. 0 TABLE OF CONTENTS ggg

1. PURPOSE / OBJECTIVE -

5

2. METHODS 5

3. ASSUMPTIONS 6

i

4. DESIGN INPUTS 7

4.1. Appued Loads 7

5. LOAD COMBINATIONS AND STRESS LEVELS 12 5.1. Load Coimbinaboos 12 5.2. Calculated Stress Levels 13

6. FRACTURE MECHANICS EVALUATION 14 6.1, Llanit Load Methodology 14 6.2. Allowable Paw Length Calculation 16 6.3. Crack Growth Evaluation 16 6.4. Procedure for Evaluation of Indications using Handbook 17

7. LEAKAGE CALCULATION 18

8.

SUMMARY

AND CONCLUSIONS 19

9. REFERENCES 20 4

7 GENeeJearEanrgy GE.NE.Bl3 0CM22 Jtn.e L PURPOSE / OBJECTIVE The objective of this report is to document the results of a fracture mechanics evaluation of-the Hatch Unit 1 jet pump riser pipe circumferential welds. This evaluation results in the allowable end-of-cycle flaw lengths at the three riser pipe circumferential welds designated as welds RS-1, RS-2; and RS-3 per BWRVIP-41 (Reference 1) and are shown in Figure 1.

The results presented in the flaw evaluation handbook can be used to disposition indications found in the jet pump riser pipes at Hatch Unit 1.

t l-2.' METHOIsS This section presents the methodology and procedure used in performing the jet pump riser 1

pipe weld flaw evaluation. Following are the steps u ed in the analysis.

1. Review of the reference drawings.

2. Determine the loading and load combinations as suggested in BWRVIP-41 (Reference 1).

L In addition,' the changes to the loads due to extended power uprate (EPU) at Hatch I are also incorporated in the loads and load combinations considered.

3. Create a SAPAG07V (Reference 2) finite element model for the jet pump.

Anchor connection points are the recirculation inlet nozzle, shroud support plate and riser brace.

4.- Determine the membrane and bending stresses considering the load combinations.

5. Use the limit load methods of BWRVIP-41 as a guide to determine the allowable flaw lengths. BWRVIP-41 evaluation procedures are used as a guide, since the jet pump is not a part of the reactor pressure boundary.

6. Evaluate IGSCC and fatigue crack growth rate. Calculate IGSCC crack growth for

- eighteen months (12,000 hrs) cycle based on a growth rate of 5x10'8 inch / hot hour.

Determine if fatigue crack growth rate due to vibration is significant by calculating the 5

1

GENedearEnergy -

GE NEcBl3 0180122 Rev.0 st[ess intensity factor due to flow induced vibration and comparing with threshold stress

- int'ensity.~ If actual stress intensity is less than threshold, than given crack is acceptable.

7.12akage curves'are calculated which show leakage versus percent of allowable flaw during normal plant operation per BWRVIP-41.

3. ASSUMPTIONS

1. The jet pump geometry is as described in the reference drawings (Reference 5). ;The.

dimensional tolerances specified on the reference drawings are such that any variations within those values will have insignificant i. t.act on the calculated stress values. It was also judged that any deviations between the as-built geometrf and the geometry indicated in the reference drawings would not be significant in terms of stress analysis and the-allowable flaw calculations.

~ 2. EThe calculations are based on one flaw per riser. However, synergistic effects of multiple flaws in one riser are negligible and would not affect the results of this analysis. The reason is that even large flaws (180 degrees) do not significantly change the stiffness of :

- the riser and therefore the response to input loading does not change.

3. Fatigue 'due to thermal stresses is negligible due to minimal temperature differentials during normal or transient conditions.

4. The jet pumps are assumed to be in the as-designed configuration because the vibration

data on which the fatigue evaluatic. is based was taken from a new plant in startup.

Issues such as jet pump fouling and restrainer bracket set screw gap may affect this

. assumption in a non-conservative manner. Even though all jet pumps have not yet been inspected, the review of the results of 6 out of 10 previously inspected and 2 out of 10

' inspected during 1997 c.itage jet pump pairs inspection data demonstrates that there is no history of set sc ew gaps or degradation at Hatch 1. Ten out of 10 Hatch 2 jet pump pairs

have been previously inspected to confinn that set screw tack welds are intact. Six out of 10 Hatch 2 jet pump pairs have been previously inspected to confirm that there are no set

- screw or restrainer wedge gaps. However, in the vibration-fatigue evaluation other conservative aspects of analysis more than adequately compensate for potential nonconservatism here.

O

GENuclearEnergy GE.NEDI3 0180122 Rev. o

'4'. DRSIGN INPUTS l

The design inputs in this evaluation consisted of the geometry of the jet pump and the applied l

loads. The geometry of the jet pump was obtained from the drawings listed in Reference 5.

The jet pump riser pipe is 10-inch schedule 30 while the thermal sleeve is 10-inch schedule 40 and the material for both is Type 304 stainless steel (Reference 5). Figure I shows a schematic of the jet pump. The welds in Figure I have been designated as RS-1 through RS-3 in accordance with BWRVIP41 designations. A finite element model was developed to

-I determine the stresses from various design loads. Figure 2 shows a line plot of the finite element model. The SAP 4G07V tinite element program was used to perform the stress analysis.

4.L AppliedLoads The applied loads on the jet pump assembly consist of the following: deadweight, seismic inertia, seismic anchor di: placements, hydraulic, fluid drag, loads due to flow induced vibrations, and thermal anchor displacements. Each of these loads are briefly discussed in the following sections.

4.1.1. Deadweight (DW)

The deadweight loading consists of the weight of the jet pump and the entrapped water. The stresses for this loading were calculated by applying one 'g' vertical acceleration in the finite element model of the jet pump assembly. For flaw evaluation purposes, the stress from this loading is treated as primary.

4.1.2 Hydraulic Loads (F1, F2)

The hydraulic loads acting on the jet pump are calculated by summing the fluid momentum and pressure forces in the venical and horizontal directions. This load definition considers any pressure differences between the annulus and the jet pump. Two hydraulic force values are calculated and applied to the jet pump. The first value is the horizontal force in the riser pipe which puts the riser elbow to sleeve weld in tension. The second value is the net vertical hydraulic load. The net vertical force is predominately caused by the pressure difference between the diffuser and inlet-mixer slip joint. Because the slip joint can not transmit a 7

GE Nwlear Energy GE NE<Bl3-0180 in Rev. 0 enic'al, load, the venicalload is carried through the riser pipe causing a bending moment o

' the riser, ihe following designations are used:

Normal llorizontal and Vertical liydraulic Load:

F1 Faulted liorizontal and Vertical Hydraulic Load:

F2 For the flaw evaluation purposes, the stresses from the hydraulic loads are treated as primary, The values of the hydraulic loads are given in the table below, flydraulic Loads Desicnation liorizontal Load (Ib)

Vertical Loadfib)

F1 16,940 9.521 F2 16,940 11.276 Since the F2 load in the venical direction is slightly higher than the F1 vertical load but the dominant horitontal load is same in both conditions, F2 loads are conservatively used in all load combinations.

4.1.3. Seismic Inertia The seismic inertia loading consists of horizontal and vertical inenia forces acting on the jet pump due to seismic excitation of the RPV (Reference 6). The seismic excitations are as a result of the core shroud repair program reanalysis of the Hatch I primary structural mcdel.

The locations where the seismic excitation is imparted to the jet pump are the vessel recirculation inlet nozzle, the shroud support plate and the riser brace. During seismic inenial loading analysis, mass of jet pump, water entrapped in it, and the hydiodynamic masses were considered. The following designations are used:

Opersting Basis Eanhquake Inenia:

OBEI Safe Shutdown (Design Basis) Eanhquake Inenia:

SSEI The natural frequency of the jet pump is high (>20 Hz) such that the horizontal acceleration values from the response spectra (envelop at RPV nodes 41 and 49) at 20 Hz are used in a j

static "g" analysis. The venical ZPA (accelerations) values were taken as 2/3 of the peak gronnd accelerations of 0.08 g (OBE) and 0.15 g (DBE). The acceleration valaes are multiplied by a conservative factor of 1.5 tc, account for higher modes effects. The values used in the evaluation are shown in the following table. The net seismic forces and moments 8

i m

GE Nuticar Energy GE NEtBl3 01869122 Rev. 0 are taken as maximum of X+Y or X+Z, where X, Y, and Z are three separate static acceleration cases analyzed using the values shown below.

Seismic Accelerations llorizontal (Y and Z)

Venical (X)

OBEl 0.45 0.08 SSEI 0.60 0.15 NOTE: liatch I has a special s...nic evaluation requirem:nt called "l/2 SME" Time llistory Evaluation. This is an alternate to the DBE loading. The acceleration values for DBE shown above are envelope of DBE or 1/2 SME. Thus no separate loading condition and combinations incorporating 1/2 SME are documented funher in this handbook.

For the purposes of flaw evaluation, the stresses from the seismic inenia loads are treated as primary.

4.1.4. Seismic Anchor Displacemer.ts The seismic anchor displacement loading consists of relative horizontal and vertical displacements between the shroud support plate, recire inlet nozzle and riser brace attachment points to the RPV. Due to relatively short elevation difference between these points and RPV itself having small displacements, relative horizontal displacements (Reference 6) have insignificant effect on flaw evaluation results, and therefore, are taken as zero. Similarly, RPV being rigid in the venical direction, there are no relative venical seis.Tuc anchor movements between recire inlet nozzle and the riser brace attachment points to the RPV. The relative displacements are negligible for both OBE and DBE loading.

Operating Basis Earthquake Displacements:

OBED Safe Shutdown Earthquake Displacements:

SSED For the purposes of flaw evaluation, the stresses from the seismic anchor displacements loads are treated as secondary.

4.1.5. FieM Diag and Acoustic Loads The drag loads consist of the forces resulting from the fluid flowing in the annulus region past the jet pump. The flow in the armulus region during normal operation exerts some 9

GE Nuclew Energy GE.NE B13 0!N9122 Rev. 0

' downward drag force on the jet pump. The magnitude of the stress from this loading is less

' than 5% of drag load due to LOCA and even less 5% of that from the weight loading. This is because the flow velocity during normal condition is only 23% of that during LOCA and the drag load is proportional to velocity square term and is therefore, neglected in the Normal / Upset load combinations. A postulated recheulation line break LOCA (suction side) subjects the jet pump to a drag force load in a tangential direction relative to vessel centerline. A previous TRACG analysis (Reference 7) performed for a similar size plant calculated the worst case flow velocities past the jet pump assembly. These velocities were adjusted ups lly for llatch 1 unique geometry. The worst case velocities correspond to a suction side recirculation line break LOCA. Other breaks do not affect the jet pump nearly as severely due to the other lines proximity o size. The velocities, used in the csiculation of the drag forces, correspond to the jet pumps nearest to the suction nozzle. The horizontal drag loads on the jet pump were determined to be approximately 900 lb on the riser and 2310 lb on the diffuser. The following designation is used:

1 Drag Loads During Normal Operation:

DRG1 Drag Loads During LOCA Condition:

DRG2 The LOCA drag load consists of the acoustic component (FAC) and the flow induced drag (DRG2) component. The acoustic component (FAC) of load is a momentary shock load and I

the flow induced drag component (DRG2) of load follows acoustic load in time. Therefore, the time phasing of the two components of this load are considered riuring Faulted load combination.

For the purposes of flaw evaluation, the stresses from the circumferential drag loads are treated as primary and the radial drag (due to vonex shedding) loads are negligible due to their low frequency (jet pump fundamental frequency being more than 3 times greater than vortex shedding frequency, thus requiring no evaluation).

4.1.6. Flow induced Vibration (FIV)

The flow induced vibration (FIV) loads are caused by turbulent flow in the piping exciting the natural frequencies of the jet pump assembly. The method of calculating the vibration stress from the 1: st data is summarized below.

1. Review the startup vibration data (Reference 8) to determine the primary modes ofinterest for the jet pump 10 i

CE Nuclear Energy GE.NE.883 0X'9122 Rev. 0

' 2. Using a finho element model of the jet pump, determine the natural frequencies, mode snapes, and modal stresses of all modes ofinterest.

3. Normalize.he modal stresses such that the they are equal to the measured data observed during startup testing.

4. Select the highest normalized modal stress at the weld location on the riser pipe from all modes. This is the FIV stress; 1150 psig peak to peak with R ratio a 0.25, bounds all three weld locations.

For liatch 1. modal extraction analysis was performed. Because the same vibration stanup data for Fitzpatrick, proto :ype of BWR/4 218 plant (Reference 9), is applicable for liatch 1, the FIY ra" sses calculated for Fitzpatrick are applicable.

For the purposes of flaw evaluation, the FIV stresses are treated as primary.

4.1.7. Thermal Anchor Displacements The three anchor points of the jet pump (the recirculation inlet nozzle, riser braces on the vessel, and the shroud support plate) grow venically and horizontally at different rates due to differences in the materials (Iow alloy steel for the vessel, versus stainless steel for the jet pump). The loads produced by the thermal achor displacements are treated as secondary for the purposes of the flaw evaluation. The following thermal displacements are considered:

Displacements during Normal Operation:

NOD Displacements during Loss of Feed Water Pump Transient:

LWFPD Displacements during LOCA thermal effects:

LOCAD The displacements are calculated at normal operating (NOD) temperature which is 552*F (including enended power uprate) for region B according to the reactor thermal cycle diagram and power uprate design specification (Reference 9). The displacement effects for LFWPD case are about the same as NOD case but are of non controlling nature since the transient temperature is 300 F in the jet pump while RPV is still at 552'F (with very low pressures) and tims this case is considered enveloped by NOD case in all load combinations.

The LOCAD case displacements are also enveloped by NOD case since the RPV temperature (approximately equal to the drywell temperature) is 290 F per UFSAR Figure 5.2 21 (sheets 3 and 5 of 5) wHr.h is lower than NOD case and thus this case is also not considered in any load combinations The RPV pressure dialation effects in the radial direction (displacements 11

GENuclearCarrgy G&N&Bl3 0lH9122 mer. o

'of RPV.due to pressures) are same at the recirculation inlet nozzle and the riser brace

' attachment locations to the RPV for any one of the above described events. RPV pressure dialation effects in the vertical direction are considered insignificant in flaw size calculations and therefore, are taken as zero. Thus their relative values are zero and thus the pressure dialation effects of RPV is neglected. Incidentally, the hydraulle loads already include the pressure differential across the riser pipe.

5. LOAD COMBINATIONS AND STRESS LEVELS This section describes the manner in which the various loads were combined for the purpose of obtaining stress levels for the flaw evaluation. The limiting stress levels at the welds are then summarized.

5.1. Load Combinations The flaw evaluation methodology to be used makes the distinction between primary and secondary stresses by specifying different safety factors. The flaw evaluation methodology also makes the distinction between the normal / upset (Level A/B) condi ion loads, for which t

i the factor of safety is 2.77, and the emergency / faulted (Level C/D) condition loads, for which the safety factor is 1.39. The load combinations are consistent with Hatch 1 UFSAR.

The following set of the controlling load combinations were considered for the evaluation of normal / upset condition:

(1) DW(P) +Fl(P) + FIV(P) + NOD (S)

(2) DW(P) +F1(P) + FIV(P) + OBEl(P) + NOD (S)

The set of the controlling load combinations used for the Emergency / Faulted conditions are the following:

(3) DW(P) + F2(P) + SSEl(P) + NOD (S)

(4) DW(P) + F2(P) + SSEl(P) + DRG2(P) + NOD (S)

(4 ALT) DW(P) + F2(P) + SSEl(P) + FAC(P)(P) + NOD (S)

I Note that the letter 'P' or 'S' in the parenthesis indicates whether a load is primary or secondary as def'med by the ASME Code. When the neoustic load component (FAC) of 12

GENacienr Energy GE NE.513 01869122 Ror. 0

' LOC 5 fond is included along with F2(P) load in load combination (4 ALT) the DRG

' load is deleted from the load combination (4 ALT) to account for the time phasing of the drag and acoustic portions of the LOCA load.

5.2. Calculated Stress Levels The forces and moments at various nodes in the model for all of the load sources were calculated using the SAP 4007V finite element code (Reference 2). These forces and moments were then combined to obtain the total forces and moments for a given load combination. Thus, for each load combination and each node, a set of forces and moments t

were obtained. Furthermore, within each set, the forces and moments from the displacement-controlled loading were tabulated separately for the calculation of expansion stress. As i

described later, the flaw evaluation metho ilogy uses the primary membrane (P ), primary r

bending (Pn) and the expansion stress (P.).

The calculated values of P., P and P stress levels at the circumferential weld locations are summarized in the following tables for the governing load combinations for Normal / Upset and Emergency / Faulted service levels. Major contribution (more than 80%) to these stresses is from hydraulic load, Calculated Stress in Governing Normal / Upset Load Combinations Weld ID P.

P.

P.

Load (Figure 1)

(psi)

(psi)

(psi)

Combination RS 1 1795 2184 413 2

RS 2 945 2424 222 2

RS-3 1025 801 102 2

CalculrW Stress in Governing Emergency / Faulted Load Combinations Weld ID P.

Pn P.

Load (Figute 1)

(psi)

(psi)

(psi) -

Combination RS-1 3221 1652 413 4 ALT RS 2 951 1877 222 4

+

RS 3 -

1029 291 102 4

13

GENuclear Energy GEME.Bl3 01"9122 Rtv. 0

'The stress levels in the preceding tables were used in the allowable flaw evaluations as described in the next section.

6. FRACT'JRE MECIIANICS EVALUATION The limit load methodology was used in calculating the allowable flaw lengths.

This methodology is first described followed by the results of allowable flaw evaluations.

6.1, Limit Exad Methodology Consider a circumferential crack of length, ! = 2Ra and constant depth, d.In order to dett rmine the point at which lim / load is achieved, it is necessary to apply the equations of equilibrium assuming that the cracked section behaves like a hinge. For this condition, the assumed stress state at the cracked section is as shown in Figure 3 where the maximum stress is the flow stress of the material, cf. Equilibrium oflongitudinal forces and moments about the neutral axis gives the following equations:

p = [(n - ad/t) - (Pm/or) n]/2 (1)

P ' = (2cr/n) (2 sin - d/t sin a)

(2) 5

Where, t = pipe thickness a = crack half angle as shown in Figure 3 d = crack depth R = pipe mean radius

= angle that defines the location of the neutral axis Z = weld type factor P.= piping expansion stress P = primary membrane stress Pe = primary bending stress P ' = failure bending stress 3

or = 3Sm. flow stress Sm = allowable stress The safety factor (SF) is then incorporated as follows:

P ' = Z*SF (P. + P + P./SF) - P.

6 6

(3) 14

GE Nuclear Energy GE NE= Bl3-0180122 Rev. 0 The P.land P, are primary stresses. P. is a secondary stress and include; stresses from all i

displacement controlled loadings such as thermal expansion, seismic anchor motion, etc. All three quantitles are calculated from the analysis of applied loading. The safety factor value is 2.77 for normal / upset conditions and 1.39 for emergency / faulted conditions. The crack angle (2ct) is the value for which equation 2 is equal to equation 3.

Z Factor The test data considered by the AShfE Code indicated that the welds produced by a process without using a flux had fracture toughness as good or better than the base metal. However, the welds produced by a process using flux had lower toughness. To account for the reduced toughness of the flux welds (as compared to non flux welds) the Section XI procedures presenbe a penalty factor, called a 'Z' factor. The examples of flux welds are submerged arc welds (SAW) and shielded metal are welds (ShiAW). Gas metal arc welds (Gh1AW) and i

gas tungsten arc welds (GTAW) are examples of non flux, velds. Figure IWB 3641 1 may be used to define weld base metal interface. The expressions for the value of Z factor in Appendix C are given as the following:

Z 1.15 [1 + 0.013(OD-4)) for ShiAW

=

1.30 (1 + 0.010(OD-4)) for SAW

=

where OD is the nominal pipe size (NPS) in inches. The procedures of Appendix C recommend the use of OD = 24 for pipe sizes less than 24 inches. This approach is very conservative and, therefore, the use of actual NPS (OD = 10 inches) was made in calculating the 'Z' factor. This approach is considered reasonable as recent discussions in the Section XI Code Working Group on Pipe Flaw Evaluation indicate that for small diameter pipes, such as the 10-inch diameter jet pump riser pipe, the Z factor may be close to or less than 1.0. The welding process used was a combination of shielded meta? are type (ShtAW) or submerged arc welds (SAW) and gas tungsten arc weld type (GTAW). Since a non flux process (GTAW) was specified for only part of the weld, it must be assumed that the welds are flux welds (SAW). The Z-factor is thus:

Z o4,cn = 1.30 [1 + 0.013(10-4)) = 1.38 i

15

GENuenearKnusy GE NE 313.el869122 an. o t 6.2, Allowable Finw Length Calculadon

• The stresses from the table in the preceding section were utilized to determine the axeptable -

through wall flaw lengths.- The acceptable flaw size was determined by requiring a safety factor on stress. The flow stress was taken as 3S. (S. = 16.9 ksi for Type 304 stainless steel-at 550'F). As specified in Reference 3, safety factors of 2.77 for the ormal/ upset' conditions and 1.39 for the emergency / faulted conditions, respectively, were used. The calculated-values of the end-of-cycle allowable flaw lengths are tabulated in the following table. These allowable flaw lengths conservatively assume all welds are flux welds (SAW).

i Maximum Allowable Flaw Lengths Based on Outside Diameter l

Weld Flaw Length (inch) j RS 1 -

15.42 RS 2 16.69 RS 3 19.87 These allowable values are the end of cycle values and they do not consider the crack growth due to IGSCC, fatigue, or NDE uncertainty. The crack growth is discussed in Section 6.3.

- 6.3. Creek Growth Evaluation Prior crack growth analyses performed for BWR shroud and core spray line indications have used a IGSCC crack growth rate of 5x10'5 inch / hot hour. This crack growth rate translates into a crack length increase per eighteen months cycle of approximately (12,000 hrs x 5x10'8) or 0.60 inch at each end cf an indication. Thus, the projected length, I of any indication f

whose current length at the time of inspection is, l, would be (lp + 0.60 x 2) inches. A factor p

of 2 in the preceding parenthesis is to account for the growth at each end of the indication.

-In addition to IGSCC growth, fatigue growth due to flow induced vibration (FIV) is discussedi The expected fatigue growth is a strong function of the crack size and orientation and cannot be determined until an indication is characterized. With a characterized crack, the stress Intensity factor (AK) can be computed and compared to the threshold stress intensity factor (AKa). The AKa is the value at which fatigue crack growth for high cycle stress

_becomes significant for high cycle events and must be considered. At values below AKa, fatigue growth is zero: For 304 stainless steel, Reference 10 repons a AKa value of approximately 5.5 ks! Vin at R ratio of <0.5. A threshold value of 5.0 ksiVin is used here to cover vibration loading uncertainties. For peak altemating stress intensity of 1150 psi, with 16

GE Nuclear Enorgy GE.NE.Bl3 0180122 Rev. o

' R < 0.5; the 5 ksi is conservative since AK is directly proponional to stress intensity, lowe

• amplitude of R will give lower fatigue allowable crack size. The allowable crack size due to FIV was based on the stress intensity calculation method described in Reference 11. The smallest calculated EOC crack size for the three locations is 5.8 inches. This value is less than the limit load method allowables, therefore, fatigue crack growth would begin prior to reaching the limit load c ack length.

Thermal expansions were also considered in evaluating fatigue. However, the fatigue crack growth due to thermal expansion stress cycling is negligible due to the limited number of cycles and low thermal stress intensities.

The loading cycles are primarily the heatup/cooldown events. Total crack growth for thermal transients is insignificant.

6.4. Procedurefor Evaluation ofIndications using Handbook Any indications at HAZ of the welds 1,2 and 3 (Figure 1) can be evaluated by use of this handbook as follows.

With a crack growth of 0.6 inch at each end of the both indications, total actual flaw size at the End-Of -Cycle including hDE uncenainty values of '2t' (Reference 12) are as shown in the Table below.

6 Flaw Sizes After one Cycle Indication Indication IGSCC NDE Adjusted Allowab5 Allowable Location Measured Growth

• uncertainty Size Flaw Size Flaw Size Weld Length (2t)

(Inches)

(Limit Load)

OUV)

RS.I 2A" 1.2" 0.61" 2A + 1.81 '

15.42" 5.8" RS 2 2B" 1.2" 0.61" 2B + 1.81 16.69" 5.8" RS 3 2C" 1.2" 0.61" 2C + 1.81 19.87" 5.8"

(*)

Since Hatch I has installed Hydrogen Water Chemistry IGSCC mitigation measures, the calculated IGSCC crack growth of 1.2 inches is very conservatively.

If the adjusted flaw lengths after one cycle are less than the allowable Gaw sizes, then the existing flaws are acceptable for continued operation without any modifications.

17

GENuclew Energy G3NE.El3-01869133 Rev.0 l7. LEIKAGE CALCULATION 1.4akage from postulated through wall flaws with length equal to the allowab!c end of cycle

_(EOC) flaw size are calculated in this section. The leakage rate through an indication was estimated assuming incompressible Bemouill flow:

Q = CA]2g,AP/ p (5) where, Q =

Leakage C=

fiow coefficient A=

area p=

mass density of fluid AP = pressure difference across the pipe / vent l

A riser te value of 173 psid is used which bounds the jet pump normal flow conditions for l

the plant. ~his is the design value for the steady state pressure difference during the jet pump l

operation for Hatch 1.

Leak rate from the through wall indications in the riser pipe can be estimated using the preceding equation with the value of flow coefficient, C, assumed as 1.0. A key input needed is the crack opening area, A.

t The approach used in this evaluation to calculate the value of A, was to calculate a conservative value of crack opening displacement, S, and assume the crack opening config;. ration to be like a rectangular slot with one side being the crack length. Sa, and the other side as the crack opening displacement. The conservative crack opening waa. :sumed as 10 mils. The crack opening area is then simply:

A = 2a (6)

(6)

Figure 4 shows leakage rates versus the percenthge of the allowable flaw size calculated in Section 8.0.

The leakage througli the allowable flaw lengths is insignificant compared to the drive flow during normal operation and compared to LPCI flow during accident scenarios, so the allowable flaw sizes are acceptable from a leakage perspective.

18

-_____a

GENutenr Energy GE NE.R13 01~'M22 Rev.0

8.

SUMMARY

AND CONCLUSIONS A flaw evaluation, consisting of stress and fracture mechanics analyses of the Hatch Unit 1 Jet pump circumferential riser welds was conducted to develop a flaw evaluation handbook.

The procedures of BWRVIP-41, were used as a guide in detemtining the allowable flaw lengths. End-of-cycle allowable flaw lengths were calculated at three circumferential weld locations. The me'.hodology presented in this repon can be used along with consideration of observed IGSCC, and evaluation of fatigue crack growth rates to disposition any indications detected during future inspections of the jet pumps at Hatch Unit 1.

The following table shows a summary of allowable adjusted flaw lengths for Hatch 1.

Allowable Adjusted Circumferential Flaw Sizes Weld Flaw Length (inch)

RS-1 15.42 RS 2 16.69 RS-3 19.87 Observed indications less than 5.8" in length, after consideration of IGSCC growth ano NDE uncenainty, do not have to be funher evaluated. However, for indications greater than 5.8" in length, after IGSCC growth and NDE uncenainty, fatigue crack growth is predicted prior to the allowable flaw length shown in the above table, 4

19

^

GENuclear Ewy GE NE Cl3 0180122 mr. o

9. REFERENCES (1)

BWR Jet Pump Assembly Inspection and Flaw Evaluation Guidelines, (BWRVIP-41), EPRI Report No. TR 108?28. October 1997.

[2]

SAP 4G07V Users Manual, NEDO 10909, Addenda 1, Revision 1,1996.

(3)

ASME Boller and Pressure Vessel Code,Section XI, Rules for In-Service Inspection of Nuclear Power Plant Components American Society of Mechanical Engineers, 1989 Edition, Paragraph IWB 3640.

(4)

Ranganath, S. and Mehta, H. S., " Engineering Methods for the Assessment of Ductile Fracture Margin in Nuclear Power Plant Piping," Elastic Plastic Fracture: Second Symposium, Volume II-Fracture Resistance Curves and Engineering Applications, ASTM STP 803, C.F. Shih and J. P. Gudas, Eds. American Society for Testing and Materials,1983, pp.11309 - II-330.

[5]

Hatch Unit 1 Jet Pump Drawing No. 729E647

)

Hatch Unit 1 Jet Pump Riser Drawing No. I17C2809 Hatch Unit 1 Jet Pump Elbow Drawing No. I17C3250 l

Hatch Unit 1 Reactor Vessel Drawing No. 920D841BB j

Hatch Unit 1 Riser Brace Drawing No. I17C4353 and 4354.

l Hatch Unit 1 Reactor Drawing No.197R589 l-(6)

Hatch Unit 1 Seismic Analysis, as follows; GENE Letter from S. H. Pang to C. V. Syx (SNOC) "In structure response spectra for a

DBE and OBE at different nozzle elevations affected by the Modified Core Shroud for Hatch Unit 1" dated October 5,1994.

b GENE -771-48-0894, Rev.1," Seismic Design Report of the Shroud for Hatch 1",

dated September 21,1994, GENE 771-48-0894, dated September 21,1994, " Hatch 1 Shroud Repair 1/2 SME c

Time History Seismic Analysis".

d GE letter to Brad Harkins from Michael P, Schrag, dated January 24,1995, "

Transmittal of various response spectra for Hatch 1".

-(7)

DRF # A7100014, Index C-13, TRACG Acoustic Loads Analysis and DRF # A13-402, Vol. X, index 3.2,3.3, Hatch 1 & 2 EPU changes.

.(8)

Fitzpatrick Vibration Measurements-Final report, NEDE-23673, Class III, August

1977, (9)

Hatch Unit 1 Reactor Thermal Cycles Drawing No. 729E762, Nozzle Thermal Cycles Drawing No.135B9990, and Power Uprate Specification 25A5541.

20

4 GENacient Energy GE NE 813 01M9122 Rev. 0 (10]. Banom, J. M., Rolfe, S. T., Fracture and Fatigue Controlin Structures, Second Edition, Prentice Hall,Inc.

[111 Zahoor, A., Ductile Fracture Handbook, Prepared for Novetech Corporation and-EPRI, EPRI Repon Number NP 6301 D.

[12]

BWR Vessel and Internals Project, Reactor Pressure Vessel and Internals Examination Guideline (BWRVIP-03), EPRI Report TR-105696, October 1995.

?

21

I s

GENuclee Energy GE NE.813 01849122 Ror. 0

~

N

/

i Ph 9

E

Ai/

li i

e, e

! )...p/

s~;

el

~~-

c e

-N 1

M z

t

\\

/

[

I i

i E

l l

l t -_. s g

Wt ec h ofs II EI 8

l0

$$E E

N A

,A wn h

\\[-

4 g

3 m

Y

\\~

i

- -- TM g

6 f

(

- mf, w

r

\\

j

%7~

ed b

e$

I

~

Ne<

Figure 1. Weld Locations on llatch Unit 1 Jet Pump 22

GENuiner Energy GE.NE;Bl3 0lN0122 gy, y

.l I

i a

N[n

/Nr i

o ll ll ll ll ll 11 Il ll Il ll ll Il ll I

ll II ll 11 y;

Y i,

~

m ti o

Il ll Il il 11 11 Figure 2. SAP Model of the Hatch Unit 1 Jet Pump 23

GENuclear Energy GEiNE+Bl3 0lRH 122 Rev. 0 Nominalareas in the Unomaind season of mpe creek t.angth a 2na P

p.

no,, or e,o,

• --+

1 I

d 4-~'

e g

\\>

v*L f

I s y i,/

t

\\./

g R

J.

t-+

+-

p f f

i i

/

g k

g.

k [/

I 1

1

\\N

(

s

-i w

I p,

M I

Asie

p. ~* *~

ames oiseribution in P. Applied Membrane 9 trees in Ug:M Section the Creoled Section at P, e Appiled Bendlag 9 trees in Unorecked Section the point ce conspm Figure 3. Stress Distribution in a Cracked Pipe at the Point of Collapse 24

GKNuclear Energy GE NE Bl3 0lN9122 R99. 0 1W.0

- 800 0 -

s' s'p*e j

80 0

• * *!* g. <

',s

'3.;,

..g s

• 'e' gll* ;,

up s',

!g

• r" W;;# '

• .1

,e c'p*

W,.gl'

'e 4tl gg a'

..b**

v'

,*o 8s J

-.. wcw ut

', g!' '.

',r*,s"#,,

s'

......weg up

... w u u, H0

,g, *

/

,, J e

it a

a w

e

=

=

.o i.

Pensmuye et Anoneme Ms anse Figure 4. Leakage Versus Percentage of Allowable Flaw Size v

25 i

f Evaluation of Indications in the Unit 1 Shroud Core Plate Support Ring (GENE-523-B1301869-12L1) l l

l l

~

V GENuclear Energy mic a:mm

'.' C flof' a reauf M a.,;; f.

• j.".'$

GFNE.$23 B1301869122L1 October 31,1997 Mr. Denver Atwood cc:

K Faynshtein, GE Southem Nuclear Co.

W Grimme, GE 40 Invemess Parkway L Patterson, GE Birmingham, AL 35242 DRF B13-01869-122

Subject:

Evaluation ofIndication in the Unit 1 Shroud Core Plate Support Ring

References:

(1) Indication Notification 197H1002.

Dear Denver,

This letter presents the results of our technical evaluation of an indication (1) in the Unit I core plate suppen ring segment weld at 180' (see Figure 1). In summary, the existing 0.3 inch long indication could grow assuming bounding IGSCC rates to 0.9 inches during the upcoming 18 month operating cycle. This is smaller than our conservatively estimated allowable flaw length of 3.6 inches, so the integrity of the ring is well within acceptable limits for at least one cycle of operation.

Acoroach The following steps and associated assumptions were used in the analysis:

1. The 0.3 inch indication, as shown in Figure 1, was assumed to txist radially across the entire 6.5 inch depth of the ring. This extends the cra:k severalinches beyond the H6B weld heat affected zone where a crack could be postulated to initiate, and is therefore a conservative flaw consideration.

2. The normal operation and LOCA delta P loadings on the ring have been multiplied by a conservative factor of S. This conservatism eliminates the need to consider detailed loading or load combinations and covers higher delta P values that would apply with extended power uprate.

3. The loading assumption on the ring takes no credit for the load camed by the core j

plate holddown bolts or by the H6a and H6bshroud welds.

4. The ring is modeled as a straight bar with an edge crack, under membrane stress equal to the hoop stress of the ring.

l

e 7

4

5. Limit load analysis of the 304 stainless steel (S.=16,950 psi) is done to determine the crack / ligament depth at which the ligament stress reaches the 3S. flow stress, also l

considering a safety factor consistent with ASME Section XI, Appendix C.

' 6. ' IOSCC growth is taken at the industry accepted bounding rate of 3x10.s in/hr, which is cordervative considering the moderate hydrogen water chemistry (HWC) levels t -

which are implemented on Unit 1. The low corrosion potentials associated with these HWC conditions would significantly slow, ifnot halt, IOSCC at the location of the Indication.

.i-Analvais The applied delta P loads are 34 psi for Normal Conditions (105% rated : ore flow) and

$3 psi for Faulted Conditions (105% rated core flow). With the Section XI safety factors of 2.77 and 1.39 for normal / upset and emergency / faulted conditions, respectively, the normal operation load and safety factor are bounding. - The conservative loading.

assumption yields a normal ' operating stress in the ring of 2 kai.

With a flow stress of 35./SF, the limit load ligame:.. stress is 18.4 ksi. The allowable ligament is therefore (2 ksi

• 4 inch) /18.4 kai = 0.4 inches. Therefore, the allowable flaw length is 3.6 inches across the entire 6.5 inch depth of the ring.

Allowing for the bounding IGSCC rate over an 18 month fuel cycle, the indication could grow 0.6 inches in one cycle of operation, to 0.9 ir.ches. This is considerably smaller than the 3.6 inch allowable flaw lengt'. so ring integrity is assured for at least one operating cycle.

Other Considerations

1. The shroud ring supporting the core plate is part of the load path which maintains tension in the shroud repair tie rods. Cracking which might degrade the ring stifthess is not a concern because the shroud shells above and below the ring are not offhet.

The ring material between the shells provides the same compressed material for the

-load path whether the ring is cracked or not. Therefore, ring cracking as described here does not degrade the shroud repair function.

2. Shroud ring cracking may provide a leakage path for higher pressure water inside the -

shroud. Such leakage, if any, would be insigni6 cant when compared to the flow through the core during normal operation or the injection into the core by ECCS systems. There would be no significant impact on operation or on design basis accident scenarios.

a

3.= There are six segment welds in the ring. The indication reported here was at the 180' segment weld. The O' segment weld was also inspected and had no indications. The other four welds are inaccessible for visual inspection. Cracking, if any, in the other segment welds should not be much different from that in the 180' weld. The results of

- this evaluation apply to cracking in all six segment welds.

Conclusions and Recommendations Thl analysis conservatively envelopes the results which would be obtained by a more rigorous analysis, analytically proves that the indication is well within allowable lindts, and can be used as the basis for a determination that the ring segment is acceptable as-is without any further action at this time.

It is recommended that the indication be reinspected in accordance with the guidelines of BWRVIP 07.

Supporting calculations and n.ience ofverification are contained in design record file B13-01869122. If you have any questions, please call me at the number below.

Regards, Crwt Verified:

TA Caine, Manager HS Mehts Technical Leader Structural Mechanics and Msterials Structural Mechanics and Materials (408) 925-4047 a

O' 6 RING SEGMENT VELDS EQUALLY SPACED.

b Jk B

g Ak Ak A

A 0.3' FLAV ON VERTICAL SURFAct 180' PLAN VIEV CORE PLATE VELD 6A l

SUPPORT RING h

j l

s i,

= r

\\

IN VELD 6B SECTION *B-B' DOCUMENTED RING SEGHENT VELD LOCATION (NOT A

IDENTIFIABLE IN THE CORE PLATE V'

EVT -1) o 0.3 '

,g FLAV

{4 s

n g

tf4, n

/

VIEV A-A' asmann Figure 1.

Location of Indication

_____- ---- j