ML20151Y828

Responds to NRC Bulletin 88-002, Rapidly Propagating Fatigue Cracks in Steam Generator Tubes. No Mods,Preventive Tube Plugging,Or Other Preventive Actions to Preclude Such Events at Plant Required

ML20151Y828
Person / Time
Site:	Point Beach
Issue date:	03/25/1988
From:	Fay C WISCONSIN ELECTRIC POWER CO.
To:	NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III)
References
CON-NRC-88-032, CON-NRC-88-32 IEB-88-002, IEB-88-2, VPNPD-88-176, NUDOCS 8805050122
Download: ML20151Y828 (8)

Text

".

q. .

s 3

(

• h .

. 1 Wisconsin Electnc ro. sea corsparr 231 W MICHIGAN,P.O Box 2046.MfLWAUKEE Wl53201 (414)221 2345 VPNPD-8 8-17 6 ,

NRC-88-032 March 25, 1988 NRC Regional Administrator, Region III U.S. NUCLEAR REGULATORY COMMISSION 799 Roosevelt Road Glen Ellyn, Illinois 60137

Dear Sir:

1 DOCKETS 50-266 AND 50-301 NRC BULLETIN NO. 88-02, RAPIDLY PROPAGATING FATIGUE CRACKS IN STEAM GENERATOR TUBES POINT BEACH NUCLEAR PLANT UNITS 1 AND 2 This is in response to NRC Bulletin No. 88-02 regarding the potential susceptibility of steam generator tubes to rapidly propagating failure initiated by high cycle fatigue such as occurred at North Anna Unit 1 on July 15, 1987. Point Beach Nuclear Plant Units 1 and 2 are Westinghouse designed nuclear power reactors with Models 44F and 44 steam generators, respectively.

The Unit 1 steam generators are not consi.dered to be susceptible to this phenomenon. The Unit 1 steam generator lower assemblies were replaced in 1983 with a design which included tube support plates designed to minimize the potential for denting of the tubes due to corrosion in the crevice between the tube and the  ;

tube support plate. The Unit 1 steam generator tuce support plates are made of SA-240 Type 405 ferritic stainless steel and j utilize quatrefoil shaped holes. Therefore, denting at the sup-l port plates and the resulting reduction in damping and increase in mean stress required for high cycle fatigue failure are not expected in Model 44F steam generators, and NRC Bulletin No. 88-02 is not applicable to Unit 1.

i On September 22, 1987 a preliminary assessment by Westinghouse l identified the Unit 2 steam generators as being potentially l susceptible to tube failure due to high cycle fatigue. In order to more accurately assess the potential for high cycle fatigue l failure of Unit 2 steam generator tubes, we performed evalua: ions l with Westinghouse Electric Corporation during the fall 1987 l'

refueling and maintenance outage. The details of the evaluations and our conclusions w+re presented at a meeting attended by repre-8805050122 880325 0tc ' ' ~

$DR ADOCK 05000266 were NUM/c. e.,;..... . .

". .' . ... t0

NRC Regional Administrator l March 25, 1988 Page 2 sentatives of the Nuclear Regulatory Commission staff, Westing-house Electric Corporation and Wisconsin Electric on November 4, 1987 and are summarized in our November 11, 1987 letter to the Commission. Further information regarding our enhanced primary-to-secondary leakage monitoring program was provided in our November 20, 1987 letter. NRC staff concurrence with our enhanced leakage monitoring program was provided in a letter from Mr. D. H. Wagner dated November 20, 1987. Detailed reports containing our evaluations and the conclusions were submitted to the Commission in our January 11, 1988 letter which transmitted WCAP-ll666, "Point Beach Unit 2 Evaluation for Tube Vibration Induced Fatigue (Proprietary)" and WCAP-ll667, "Point Beach Unit 2 Evaluation for Tube Vibration Induced Fatigue (Non-Proprietary)."

These evaluations confirmed that Unit 2 steam generator tubes are not susceptible to fatigue failure.

Oar responses to specific actions requested in NRC Bulletin 88-02 are provided below. References are provided as appropriate to documents submitted previously and described above.

A. Steam Generator Tube Inspections Eddy current testing was conducted during the fall 1987 refueling and maintenance outage on essentially 100% of the in-service tubes in rows 8 through 13 to identify those tubes ,

which exhibit denting and to identify the antivibration bar .

(AVB) insertion depths for each column. As expected, the eddy I current test data showed that virtually all of the tubes in these rows are dented at the sixth support plate (uppermost tube support plate for a Model 44 steam generator). The eddy current test data also showed that no tubes have wall thinning indications at the AVB locations. Therefore it is unlikely -

that those tubes have been unstable. As a ecnservative measure it was assumed for subsequent evaluations that all tubes of concern are dented at the sixth support plate.

B. Future Steam Generator Tube Inspections If No Denting Is Found Not applicable to Point Beach Unit 2.

C.1. Enhanced Primary-to-Secondary Leakage Monitoring Concurrent with the Novender 1987 evaluations to determine the susceptibility of the Unit 2 steam generator tubes to high cycle fatigue failure, we reviewed our existing primary-to-secondary leak rate monitoring capabilities and proce-dures. As a result of this review, the procedures were revised to enhance the ability to detect, assess, and take appropriate action during the progression of a tube failure

e ,

e

( NRC Regional Administrator March 25, 1988 ' '

Page 3

}

with leakage characteristics similar to that which occurred at North' Anna Unit 1. Our enhanced primary-to-secondary leakage monitoring program is described in detail in our November 11 and November 20, 1987 letters and was implemented prior to return to power following the fall 1987 refueling and maintenance outage.

Since implementation of the enhanced leakage monitoring '

program, we have re-evaluated the precision and accuracy of leak rate determinations based upon radiochemical sampic analysis. These evaluations indicate an uncertainty of approximately-20% with a reproducibility of about 1 to 2 gallons per day. We have also compared leak rate deter-minations from radiochemical sample analysis to those determined from Unit 2 air ejector monitor readings. These comparisons indicate that leak rate determinations from air ejector monitor readings are consistently and conservatively higher than those from radiochemical sample analysis.

Assuming a tube fatigue failure is in progress with time dependent leakage characteristics as shown in Figure 1 of Bulletin 88-02, the time between reaching 100 gallons per day (our administrative requirement for evaluating the need to reduce power) and total tube rupture is about 15 hours1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br />.

Applying n 5 hour5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> margin to rupture and a leak rate determi-nation uncertainty of 20%, about 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> are available for further leak rate trend and stability evaluations, confirma-tion of leak rate magnitudes by other means, and evaluations of the need to reduce power or shut down.

A power reduction to 50% power could be accomplished in less than one hour, if necessary. Even if the entire 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> available were used for further evaluations, a 50% reduction in power could be accomplished prior to exceeding the existing technical specification limit of 500 gallons per day. Thus, in view of the conservatisms in our enhanced leak rate monitoring program and the evaluations of tube fatigue susceptibility discussed in items C.2.a and C.2.b below, it is not necessary to impose absolute leak rate limits for ,

commencing plant shutdown or time limitations for reducing power or reaching cold shutdown. It should be noted that when leak rate determinations are from air ejector monitor readings, even more margin to tube rupture exists because of the higher leak rate estimates using this monitor.

i

. . +

NRC Regional Administrator  ;)

March 25, 1988 Page 4 C.2.a and C.2.b Detailed Analyses to Assess the Potential For Tube Failure Evaluations to assess the potential for highJcycle fatigue f ailure were performed with Westinghouse Elect:.-ic Corporation during the fall 1987 refueling and maintenance outage and detailed reports were submitted in WCAP-ll666 and WCAP-ll667.

These evaluations specifically address the requirements of C.2.a and C.2.b of Bulletin No. 88-02.

Subsequent to these submittals Westinghouse completed three refinements to the Unit 2 evaluations. These refinements include

1) the best estimate value of the stability ratio (without peaking factor) for the North Anna R9C51 tube using an ATHOS model of the North Anna steam generator with cartesian coordinates; 2) an up-dated value of flow peaking for North Anna R9C51 tuber and 3) the effect of operation of Unit 2 during ast-down periods at the end of fuel cycles. The overall effect <' hese refinements on the fatigue uscage, stability ratios, anc ..ress ration for the Unit 2 steam generator tubes is a reduction in the corresponding values reported in WCAP-ll666 and WCAP-ll667. This demonstrates a greater margin to the lower limit for susceptability.

The initial three dimensional evaluations for North Anna were performed using an ATHOS model in cylindrical coordinates. These evaluations were repeated using an ATHOS model in cartesian coor-dinates, including a fine resolution cell mesh, for the peripheral columns of tubes in Rows 9 to 12. The cartesian coordinate model allows better simulation of the U-tube region including the AVB's.

As a result, a more accurate solution is obtained using this model.

Modifications were made to the air test model used to determine flow peaking factors resulting from ncn-uniform AVB insertion.

These modifications eliminated air leakage paths that occurred in the original tests near slots used to adjust AVB positions in the model. All AVB configurations were retested to obtain the final test values for the peaking factors. In addition, a detailed ,

uncertainty evaluation was performed and the results were used conservatively to adjust the peaking factors used for the tube fatigue analyses. Included in the uncertainty evaluation were

1) a reassessment of the North Anna R9C51 AVL configuration using AVB extrapolation methods developed since the original North Anna evaluation; 2) test uncertainties from extrapolating the canti-lever configuration in the air test model to an actual U-bend;

3) uncertainties from extrapolating the air test to a steam / water mixture; and 4) uncertainties in AVB position, particularly for low flow peaking configurations. The overall result of the air

', NRC Regional Administrator March 25, 1988 Page 5 e

test model modification and uncertainty evaluation was an increase of 9% in the North Anna R9C51 flow peaking factor. This increase leads to a reduction in the maximum relative stability and stress ratios for Unit 2 steam generator tubes and demonstrates a greater margin to the lower limit for susceptibility.

The following table presents the results of applying the revised values for flow peaking and stability ratio for North Anna R9C51.

For the purpose of comparison, this list of tubes and associated revised relative stability ratios and stress ratios also contains the corresponding values as presented in Table 8-2 of WCAP-ll666 and WCAP-ll667. All stress ratios have decreased which demon-strates the conservatism of the original evaluation results. This is a direct result of the decrease in the relative stability ratio by including flow peaking effects. The largest relative stability ratio is now 0.82 as compared to the prior value of 0.88.

STEAM GENERATOR A Relative Stability Ratio Stress Ratio Tube Original Revised Original Revised R12C2 0.84 0.79 0 48 0.34 -

R12C91 0.84 0.79 0.48 0.34 RllC2 0.61 0.57 0.09 0.06 R11C3 0.86 0.81 0.65 0.46 R11C4 0.88 0.82 0.76 0.49 R11C91 0.61 0.57 0.09 0.06 R10C3 0.71 0.67 0.24 0.17 R10C4 0.70 0.66 0.21 0.16 R10C5 0.69 0.65 0.20 0.14 4

<,--m -_- - ~ -- ,- - , . , _ , , - - - , ,, .-,,,.._,.-r.,_ -

__--1,

NRC Regional Administrator March 25, 1988 '

Page 6 STEAM GENERATOR B i Relative .etability Ratio Stress Ratio Tube Original Revised Original Revised

.R12C2 0.84 0.79 0.48 0.34 R12C91 0.B4 0.79 0.48 0.34 R11C2 0.61 0.57 0.09 0.06 R11C91 0.61 0.57 0.09 0.06 R10C4 0.70 0.66 0.21 0.16 R10C5 0.82 0.77 0.57 0.39 R10C90 0.71 0.67 0.24 0.17 Note: Ratios are relative to North Anna Unit 1 R9C51.

Westinghouse has developed conservative acceptance criteria to determine susceptibility to high cycle fatigue failure of steam generator tubes as referenced to North Anna R9C51. Specifically, the acceptance criteria for a tube which is dented at the top tube support plate and without AVB support are 1) a relative stability ratio which it less than or equal to 0.9 and 2) a relative stress ratio less than or equal to 1.0. The results described above are well within these criteria.

In the tube fatigue evaluation of the Unit 2 steam generators pre-sented in WCAP-11666 and WCAP-ll667, the effect of reactor opera-tion during coast-down periods at the end of fuel cycles was not evaluated. Coast-down periods involve plant operation at reduced primary coolant temperatures and possibly reduced reactor power at the end of the fuel cycle. -

Operation at reduced primary temperatures results in lower steam pressures, lower U-bond density, higher U-bend velocities, and reduced damping as a result of higher void fractions in the U-bend. The analysis described below was performed to identify the ef fect on :lluidelastic vibration and hence on tube f atigue resulting f rom the end of fuel cycle coast-down periods which have occurred for Unit 2.

Using plant data for each c; eight coast-down periods the higher powered of the two steam generators was selected for evaluation.

A one-dimensional analysis was performed for the beginning and the end of each period and a "Ratio of Stability Ratio, End/ Start" was calculated. A value less than one indicates that the relative stability ratio became smaller and hence more favorable during the coast-down period. A,value greater than sne indicates that the j

.~ '

NRC Regional Administrator ,

. March 28, 1988 Page 7 relative stability ratio became larger during the coast-down period. The results of these analyses are summarized in the following table.

POINT BEACH UNIT 2 COAST-DOWN ANALYSIS RESULTS Puel Hatio of Fuel Ratio of Cycle Stability Ratio Cycle Stability Ratio Number End/ Start Number End/ Start 1 1.071 5 0.991 2 0.946 6 0.983 3 1.014 8 0.982 4 0.993 10 0.979 The values of Ratio of Stability Ratio End/ Start show that during six of the coast-down periods the relative stability ratio decreased, and in two cases the relative stability ratio increased slightly. In the original analysis, fatigue usage was calculated by assuming that all fuel cycles operated at the highest stress level calculated for any fuel cycle since plant startup (3.04 ksi for Cycle 13). Since this results in a conservative estimate of usage and since the increase in stress caused by the two coastdown periods with small increases in stability ratios is small and less than the conservative case, it is concluded that the effects of a coastdown analysis are bounded by the original analysis.

The steam flow rates for the starting conditions of some of the coast-down periodg exceeded the original analysis reference steam flow of 3.25 x 10 lb/hr. The differences between the 100% power conditions are believed to be the result of a combination of physical variability between steam generators and steam flow measurement uncertainties. The starting conditions for the coast-down periods listed in the above table can result in a 3.5%

increase in the relative stability ratio as compared to the reference condition. However, as described above, the evaluation for the reference condition has been revised to update the rela-tive stability ratios and stress ratios by incorporating a higher flow peaking factor for the North Anna, R9C51 tube. The general

4 *. &~

s .,

NRC Regional Administrator March 25, 1988 ,

Page 8 .

effect of that revision is a 6.3% reduction in-relative stability ratios. The effect of the higher than reference steam flow rates on relative stability ratios and stress ratios is therefore boun-ded by the original evaluation which incorporates the lower flow peaking factor for the North Anna, R9C51 tube.

Based on our evaluations, as originally presented, and-as supple-mented in accordance with the considerations discussed above, no tubes in the Unit 2 steam generators are susceptible to tube fatigue failure similar to that which occurred at North Anna Unit 1. Therefore, no modifications, preventive tube plugging, or other preventive actions to preclude such an event at Point Beach are required.

Please contact us if you have any questions regarding our response. .

Very truly yours, i

,v !

(iv C. W. Fay BEA/jg Copies to NRC Document Control Desk NRC Resident Inspector R. S. Cullen, Public Service Commission of Wisconsin Blind copies to Britt/Gorske/Finke, Krieser, Lipke, Newton, Zach, Gerald Charnoff, Frieling, Aronson, Seizert, [ [

Ron Steve i

Subscribed and sworn to before me this Af D of h b 1 , 1988 un  % LA Notary Public, State of Wisconsin My Commission expires S/27[9 0 .

-_,--m. -