ML20205F663

Jet Pump Riser Weld Leakage Evaluation

ML20205F663
Person / Time
Site:	Vermont Yankee
Issue date:	03/04/1999
From:	Gmyrek M, Goodwin E, Hsieh J DUKE ENGINEERING & SERVICES
To:
Shared Package
ML20205F577	List: ML20205F574 ML20205F649 ML20205F663
References
VY9123-TD01, VY9123-TD01-R00, VY9123-TD1, VY9123-TD1-R, NUDOCS 9904060352
Download: ML20205F663 (5)

Text

i . . . -

1 1

m Duk3 Engineering Ed& Services.

A Duke Ene y Compmy l

TECHNICAL DOCUMENT COVER SHEET DOCUMENT NUMBER: VY9123-TD01 REV: 0 DOCUMENT TITLE: JET PUMP RISER WELD LEAKAGE EVALUATION PROJECT NAME: Supplemental Riser Weld Leakage Documentation PROJECT NUMBER: 00241.30.9123.00.0 CLIENT: Vermont Yankee Nuclear Power Corporation QA CONDITION I I

l DESIGN VERIFICATION METHOD:

O Design Review O Altemate Calculation O Qualification Testing Preparer Verifier Approver Signature

(( y , M4M g

. GooMn U J. S. Hsieh F M. WGyek Printed Name Date 3 - M- //(( d f f g [

9904060352 990329 // '

PDR ADOCK 05000271 P PDR 00241.30.9123.00.00000 Page1of5

r i

.b >

Duke Engineering andServices yy9;y3 399;

! JET PUMP RISER WELD LEAKAGE EVALUATION 1

Executive S'ummary Leakage rates through flawedjet pump riser welds found during the 1998 outage at Vermont Yankee were calculated reflecting the end of two subsequent cycles. The leakage rates were calculated using adjusted flaw lengths accounting for measurement uncertainty and IGSCC crack growth. The leakage rates were '

calculated for normal operation and post-LOCA conditions with pressure drops of 123 and 13 psid respectively.

l Leakage Rate, gpm l Weld ID Length for Analysis, Normal Operation LOCA Conditions  !

inches 1 28.04 9.12 N2B-RS-1 5.83 28.04 9.12 N2C-RS-1 5.83 N2H-RS-1 3.85 18.52 6.02 f 28.04 9.12 l N2K-RS-1 5.83 l

l 00241.30.9123.00.00000 Pcge 2 of 5

J, u i Duke Engineering andServices VY9123-TDo1 JET PUMP RISER WELD LEAKAGE EVALUATION

Introduction:

To address the recommended strategy on Jet Pump Leakage (Reference 1), a detailed calculation of leakage rates and assessmem ofleakage acceptability for two cycles of operation was completed in Reference 2. Reference 2 conservatively examined leakage through the fourjet pump riser welds determined to be flawed following ultrasonic inspection (Reference 3).

- In Reference 2 the flaws were rssumed to have an equal length of 5.83 inches following an additional two cycles of flaw growth beyond that most recently measured. As reported in Reference 3, the flaw length of weld N2H-RS-1 was measured at a significantly smaller value then the other flawed welds. This 'i document modifies the detailed calculation documented in Reference 2 to reflect the smaller flaw length measured for that weld.

Inputs and Assumptions: .

1. Vermont Yankee riser differential pressure at nonnal operating conditions as found in the plant specific database and referenced in Reference 3 (DP = 123' psi).

2. Vermont Yankee riser differential pressure for post-LOCA conditions as described in Reference 3.

(DP = 13 psi).  ;

3. The calculated IGSCC crack growth for two eighteen months cycles is based on a growth rate of 5E-5 inch / hot hour (Reference 4).

4. Per reference 4, a " Flaw Handbook" has been developed by General Electric for Vermont Yankee to identify the critical Flaw size for each of the jet pump circumferential welds. It was determined that the RS-1 welds could contain a flaw as large as 16.41" long and still satisfy structural integrity requirements.

5. The two-cycle time frame is assumed to be 26,280 hours0.00324 days <br />0.0778 hours <br />4.62963e-4 weeks <br />1.0654e-4 months <br /> (Reference 5).

6. No additional cracks are assumed to initiate during the period between inspections.

7. As reported in Reference 3,26 of 30 welds were ultrasonic testing (UT) inspected. The remaining four welds were inspected using a modified visual, VT-1, inspection technique. The only flaws identified  !

. were those found with using the UT technique.

8. Per Reference 4, the UT weld measurement has an uncertainty of 0.19 inches per end or 0.38 inches total.

Calculation..

Flaws Contributing to Leakane As described in Reference 3, four thermal sleeves to elbow welds were identified with existing flaws. To each measured flaw length a measurement uncertainty is acuunted for by adding a 0.191-inch evaluation factor to each end on the measured length. Additionally, to account for flaw growth due to Intergranular Stress Corcosion Cracking (IGSCC) for a two-cycle 36-month time frame (26,280 hours0.00324 days <br />0.0778 hours <br />4.62963e-4 weeks <br />1.0654e-4 months <br />) an additional length is added using a flaw growth rate of SE-5 in/hr (Reference 4) that equates to a growth of 26,280 x 5E-5 = 1.314 inch growth per end.

00241.30.9123.00.00000 Page 3 of 5

f . *, ;

Duke Engineering andServices VY9123-TD01 JET PUMP RISER WELD LEAKAGE EVALUATION Summarizing:

Weld Measured Length, Evaluation Factor Two Cycle Growth, Predicted Length for inches per end, inches per end, inches Analysis, inches N2B-RS-1 2.82 0.191 1.314 5.83 N2C-RS-1 2.82 0.191 1.314 5.83 N2H-RS-1 0.84 0.191 1.314 3.85 N2K-RS-1 2.82 0.191 1.314 5.83 Note that the predicted flaw length following two cycles is smaller than the limit for structural integrity as identified in Inouts and Assumotions: number 4.

Leakage Rate Calculation The leakage rate through the through-wall flaw can be estimated using assuming incompressible Bemoulli flow through the crack.

As described in Reference 2:

Q = CA 4 (2 g, AP/p)

Where: Q= Leakage (ft'/sec)

C= flow coefficient 2

A= Area (ft )

p= mass density of fluid at 520 F,Ib/ft' 2

AP = pressure difference across the pipe, Ib/ft 2

ge= 32.2 ft/sec To complete the calculation the flow area is assumed to correspond to a rectangular slot using the flaw length with a width of 10 mils as described in Reference 2. A AP value of 123 psi for normal operation was used in Reference 3 and will be used in this calculation. Per Reference 3, a AP of 13 psi is used for LOCA conditions. An assumed value of 1 is assumed for the flow coefficient (C), as was used in References 2 and 3. The following calculation is provided as an example; Q = 1.0 * (0.010 inch

• 5.83 inch)/1442in ffg2 *Y (64.4 ft/sec2

• 1442 in fp ,123 lb/ in2 / 47.9 lb/ft')=

= 6.248 e-2 ft' /sec = 28.04 gpm The resulting flow rates are:

Leakage Rate, gpm Weld Length for Analysis, inches Normal Operation LOCA Conditions 28.04 9.12 N2B-RS-1 5.83 28.04 9.12 N2C-RS-1 5.83 18.52 6.02 N2H-RS-1 3.85 28.04 9.12 N2K-RS-1 5.83 102.6 33.4 Total Flow Rate 00241.30.9123.00.00000 Page 4 of 5

r

.A .

Duke Engineering and Services VY9123-TD01 JET PUMP RISER WELD LEAKAGE EVALUATION Leak, age Rate Acceptability Reference 2 evaluated the impact ofleakage through four identical through-wall weld flaws of 5.83 inches length. The results presented above reflect three flaw lengths of 5.83 inches and one of 3.85 inches. The total leakage flow rate calculated here for normal conditions are therefore lower. Based on the information presented in Reference 2, the following evaluation is provided.

Normal Ooeration i The total leakage through the jet pump welds during normal operation is 102.6 gpm. This represents less then 0.12 % of the rated core flow of 48 Mlb/hr. This additionalleakage will require the recirculation pumps to operate at a slightly higher speed at the rated core flow to compensate for the additional leakage bypassing the core.

The flow biased scram setpoint is automatically adjusted based on a calculated core flow from the i recirculation pump flow indication. Because the leakage occurs between the recirculation flow indication .

[ and the core flow indication at thejet pumps, this leakage would increase the actual flow difference I between the recirculation flow and thejet pump driving flow. However, because the recirculation flow is calibrated to the jet pump core flow indication, the additional leakage is accommodated in the calibration process. Therefore, this leakage does not impact the accuracy of the core flow measurement input to the flow biased scram setpoint. Consequently, there is no impact on the Anticipated Operational Transients, which credit the flow-biased scram.

LOCA Conditions The total leakage through thejet pump welds during LOCA is 9 gpm or less per weld. Since injection is credited in only one loop, the flow is decreased by leakage through only two welds. Using the worst case combination results in 18 gpm leakage. The 18 gpm ofleakage can be accommodated by the assumption of 500 gpm through the minimum flow bypass for one LPCI pump injecting in one loop and is accommodated by the conservatism in the analysis for two pumps injecting in one loop.

l

References:

1 Memo J. Hoffman to D. Yasi, " Recommended Strategy for Jet Pump Riser Welds", OPVY 98/109,

April 22,1998.

2 Calculation VYC-1864 , Rev 0 " Jet Pump Riser Weld Leakage Evaluation", dated February 4,1999 3 Letter from Donald A. Reid, Vermont Yankee Nuclear Power Corporation to U. S. Nuclear Regulatory Commission, BVY 98-67, dated May 4,1998 4 Report GE-NE-B13-01935-LTR, Rev 1, " Jet Pump Riser Welds Allowable Flaw Sizes Letter Report for Vermont Yankee", dated April 1998.

5 Memorandum from C. B. Larsen to L. Schor, " Jet Pump Flaw Lengths for Input to Leakage Calculation, Dated January 25,1999, TS 8/99 l

00241.30.9123.00.00000 Page 5 of 5