ML18082A492

Forwards Schedule for Evaluation of Westinghouse Steam Generator Row One U Bends.Requests Delay of NRC Issuance of Generic Ltrs to near-term OL Plants Requiring Plugging of Row One Tubes,In Confirmation of 800415 Meeting

ML18082A492
Person / Time
Site:	Salem, North Anna, 05000450, Trojan
Issue date:	05/12/1980
From:	Anderson T WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
To:	Vollmer R Office of Nuclear Reactor Regulation
References
NS-TMA-2241, NUDOCS 8005290406
Download: ML18082A492 (12)

Text

Westinghouse Water Reactor

• Nuclear Technology Division Electric Corporation Divisions Box 355 Pittsburgh Pennsylvania 15230 NS-TMA-2241 May 12, 1980 Mr. R. H. Vollmer Director Division of Engineering U.S. Nuclear Regulatory Commission 7920 Norfolk Avenue Bethesda, Maryland 20014

Subject:

Steam Generator Row One 11 U11 Bends

Dear Mr. Vollmer:

On January 21, 1980, the Nuclear Regulatory Commission (NRC) issued a letter to Virginia Electric Power Company requiring that inspection ports be installed at the 11 U11 bend area on all North Anna Unit #2 steam generators prior to plant startup. The utility responded to the NRC by offering to plug the row one tubes in each steam generator, indicating that this action would eliminate the need for installing inspection ports. Subsequently, the NRC issued a letter to Public Service Electric and Gas requiring not only that inspection ports be installed on all Salem Unit #2 steam generators, but also that all row one tubes in each steam generator be plugged prior to plant start-up. The NRC Staff indicated that a similar action would be pursued on all other Westinghouse near-term Operating License (OL) plants.

The apparent concern of the NRC is that the 11 U11 bend area of the row one tubes may be more susceptible to sudden large leaks than other steam generator tubes.

The operating plant data indicates that the leakage experienced to date from the 11 U11 bend area of row one tubes is a very small percen-tage of the total operating plant history. (See accompanying attach-ment.) Further, lea ks in the 11 U11 bend area of row one tubes that have been manufactured by Westinghouse and that were not due to denting have been characterized by low leakage rates and stable leak size.

The large leak experienced at Surry l is attributed to denting, and the large leak at Dael 2 is attributed to excess ovality of the tubing in the 11 U11 bend region not manufactured by Westinghouse. Plugging of the row one tubes accounts for approximately 3% of the tubes in each steam generator. This is a significant portion of the tube bundle to remove from service when there is not a sound technical basis for b

s oos2 9 o.l/otP. f

Mr. R. H. Vollmer NS-TMA-2241 this decision. The history of all operating steam generators indicates that there will possibly be a need to plug additional tubes in the future. It is neither desirable nor necessary that the margin in heat transfer capability from arbitrarily plugging row l tubes should be lost at this time. If the need or the justification for plugging these tubes should arise in the future, the tubes could still be plugged. While past experience indicates that substanti~l heat transfer margins exist on older steam generators, optimization of plant design on newer units may not allow for the same level of heat transfer margin for future plants.

The nun bend leakage which has been encountered has occurred in units tubed with 7/8 11 O.D. x o.oson wall Inconel 600 tubing. There is reason to believe that newer units soon to go into operation (Model D, E and F steam generators) tubed with smaller diameter tubing represent a different population and are not expected to be susceptible to row one nun bend cracking. .

Westinghouse, in conjunction with Portland General Electric, has undertaken a program to determine the cause or causes of the row one nun bend leaks. Portland General Electric has agreed to the removal of 26 row one nun bends and three row two 11 U11 bends from one of the steam generators in their Trojan Unit as the basis for the investiga-tion. It is the objective that a detailed evaluation of these tubes, as discussed in the attached "Steam Generator Row One nun Bend Program,"

will identify these causes and lead to the development of a field inspec-tion method capable of detecting potential leaking tubes. This will then allow a preventive plugging program to be implemented on susceptible tubes.

The schedule for the completion of the various phases of the program is given in the attachment. Westirighouse is prepared to meet with .

the Staff and provide a status of the program results at the conclusion of the destructive examination of the Trojan nun bends and at the con-clusion of our effort to establish an NDE method for detecting susceptible tubes. These intermediate tasks are scheduled for 1-1/2 and 4 months, respectively, after receipt of the samples during May at Westinghouse.

By copy of this letter, and the attached program, Westinghouse is requesting the NRC to delay the issuarice of additional letters to near-term OL plants requiring the plugging of row one tubes until this investi-gative program has reached a point where a decision can be made based on sound, technical judgement. This point should be approximately four months after receipt of the Trojan 11 U" bend samples at Westinghouse.

Mr. R. H. Vollmer NS-TMA-2241 This letter confirms the discussions Westinghouse has had with the Staff in a meeting held April 15, 1980, in Bethesda, Maryland.

Should you have any questions on the Westinghouse program, please contact R. J. Sero at 412-373-4189. Westinghouse would appreciate your expedited reply to our request.

Thank you in advance.

T. M. Anderson, Manager Nuclear Safety Department RJS/TMA/rl i .

Attachment cc: V. S. Noonan - USNRC S. S. Pawlicki - USNRC

ATTACHMENT STEAM GENERATOR ROW ONE "U" BEND PROGRAM A. PURPOSE This program has been initiated as a result of several plants experi-encing leaks in their row one tubes in the "U" bend region. These leaking tubes, in conjunction with the plant operating history, are not believed to be categorized with any previously known degradation mech-anisms. These leaks are not believed to be associated with denting, phosphate thinning or any other previously identified causes of tube degradation. The program has three objectives. The first is to deter-mine the cause or causes of the tube degradation in the "U" bend. The degradation appears to be random in nature; there are steam generators in service that are apparent duplicates of the steam generators with tube "U bend 1eakage that have not had this form of tube degradation.

11 The second objective of the program is to develop an NOE method that is able to identify susceptible tubes in the field. This will allow a preventive plugging program to be implemented. The third objective of the program is to develop criteria that will be applicable to future steam generators.

B. PROGRAM OUTLINE The "U" bend program will follow the logic diagram shown in Figure 1.

The logic diagram illustrates the course of the investigation and the various options and possible results of the program.

There are several postulated causes for the tubes to leak, but any definitive answers must be determined from the metallurgical investi-gation. The metallurgical investigation should reveal th~ mode of degradation and the specific location of the degradation. It cannot be ruled out that the end result of the program may be to preventively plug the row one tubes.

ATTACHMENT C. INVESTIGATIVE PROGRAM DETAILS Task 1: Removal and Examination of Trojan 11 U11 Bends The selected tube bends will be removed through a six inch access penetration machined in the shell and wrapper at the 11 U11 bend eleva-tion. In anticipation of and in preparation for the tube removal operation, Westinghouse fabricated a full scale mock-up, designed and built the necessary tools, and successfully *demonstrated the process.

The laboratory examination of the 11 U11 bends to be removed.from Trojan Unit #1 will be divided into two distinct phases; nam~ly, non-destructive and destructive examinations. Some of the non-destructive examinations will duplicate those undertaken at the site; the purpose of this action is to assure that no marked change has taken place within the sample during the removal operation. Dimensional measurements of leg spacing will be taken before and after removal so that the amount of leg spring, if any, can be determined.

The exact number of specimens to be sectioned, etc., cannot be predicted at this time. It is expected that at least one 11 U11 bend will contain a through-wall penetration and several others will have part-wall penetration. The objective of the destructive examination is to (1) determine the cause of the penetrations, and (2) to determine why only those tubes were affected.

It is expected that, based upon the.destructive examination, a correla-tion will be found between those tubes exhibiting cracking and some characteristics of those tubes which can be detect~d non~destructively.

It would be fortunate if thi"s characteristic could be detected by the eddy current test technique, either utilizing the conventional proce-dures or some modification of the procedure which could be readily implemented. This technique could then be applied in the field with better confidence in signal interpretation.

ATTACHMENT Another possibility is that there is a geometric characteristic involved which could be revealed by ID gauging techniques.

In any case, it is the purpose of this subtask to determine what avail-able non-destructive test technique would provide the best means for identifying suspect tubes; i.e., those tubes with incipient cracks or with a high potential of developing cracks, and to recommend appropriate field inspection plans and techniques.

Task 2: Examination of Archived 11 U11 Bend Samples During 1976, a number of 11 U11 bend samples were removed from the following steam generators in conjunction with denting related 11 U11 bend cracking:

Surry Unit 2, S/G A 9 Row 1 Surry Unit 1, S/G A 15 Row 1 15 Row 2 1 Row 3 Turkey Pt. Unit 4, S/G B 15 Row 1 15 Row 2 1 Row 3 These tubes were examined using non-destructive and destructive techniques.

Primary interest was given to the 11 U11 bend apex for these samples; no NOE indications were noticed and, thus, no metallography was performed at the tangent points. Relevant data may be obtained by reexamining these archived samples utilizing the results of the Trojan tube examination.

This task will make maximum use of data previously recorded for the archive 11 U11 bends, to avoid duplication of effort in the present, supple-mentary examinations.

ATTACHMENT Task 3: Characterization of Prior and Current Production 11 U11 Bends An extensive test program is .underway at. the Westinghouse Specialty Metals Division at Blairsville, PA, consisting of the following:

1. Characterization of Prior Production Bends Model 44 and 51 'tubes: 7/8 11 OD, Row 1 and Row 2

a. Non-destructive Examination

1. PT OD surface for indications

2. UT for wall thickness

3. Ovality measurement traverse

4. ID E/C examination

b. Destructive Examination

1. Wall thickness traverse

2. PT ID surface

3. Metallography as required

4. SEM and micro-probe, if required

c. Corrosion Tests

1. Polythionic acid tests for residual stresses

2. Primary water stress corrosion tests

2. *Characterization of Current Production Bends Model D and F tubes: 3/4 11 and 11/16 11 OD tubes, Row 1 and Row 2 Tests same as la, lb and le.

3. Parametric Studies Vary bending parameters to (and beyond) failure to:

a. Define margin under normal processing

b. Provide characterized defects for S/G calibration standards

ATTACHMENT The major purpose of this work is to determine if there are any tube characteristics associated with the bending procedure which may be related to the Dbserved tube leakage.* Much of the non-destructive and destructive examinations have been completed with no evidence, to date, of any obvious condition that correlates to 11 U11 bend leakage.

11 11 Task 4: Review of Eddy Current Examinations of Row l and Row 2 U Bends in Operating Plants Westinghouse has recommended to its customers that the normal in-service 11 11 inspections of steam generator tubing include a11 the Row l U bends and a sampling of the Row 2 U bends. Where such data are available, 11 11 a review will be made to determine whether or not indications exist and whether or not there is any commonality to the findings. For example, any eddy current indications will be categorized with regard to plant, steam generator model, tube location, indication location, tube supplier, plant operating conditions, etc. The absence of indica-tions, of course, will also be assessed for its implications.

The anticipated schedule for the major milestones in each task is as follows:

Task Completion Date

1. Removal and Examination of Trojan Row 1 and Row 2 Tubes

a. Determine cause of leakage by appropriate examinations:

1. non-destructive 0.5 month~ after receipt of samples

2. destructive 1.5 months after receipt of samples

ATTACHMENT

b. Relate cause of leakage 2 months after receipt of samples to some specific attribute or characteristic of the affected 11 U11 bends

c. Identify potential non- 3 months after receipt of samples destructive technique(s) to detect the tube attributes associated with the leakage

d. Status report 4 months after receipt of samples

e. Implement field inspection To be determined utilizing identified non-destructive technique(s)

2. Examination of Archived 11 U11 Bend Samples

a. Determine condition at tan- November 19ao gent points by appropriate non-destructive and destructive examinations

b. Correlate results with those December 1980 of Trojan examinations

3. Characterization of Prior and Current Production 11 U11 Bends

a. Model 44 and 51 Tubes: 7/8 11 OD, Row l and Row 2

l. Non-destructive and destructive July 1980 examination

2. Corrosion tests (a) polythionic acid tests for September 1980 residual stress (b) primary water stress November 1980 corrosion tests

b. Model D and F Tubes: 3/4 11 and 11/16 11 OD, Row l and Row 2

l. non-destructive and destructive July 1980 examination

I ATTACHMENT

2. corrosion tests (a) polythionic acid tests September 1980 for residual stress (b) primary water stress December 1980 corrosion tests

4. Review of Eddy Current Examinations of Ongoing Row 1 and Row 2 Tubes in Operating Plants

5. Final Report 7 months after receipt of samples D. BASIS FOR CONTINUED OPERATION There has been a small number of plants that have experienced row one 11 U11 bend leaks whose cause has b.een undetermined. These are:

Date of Cumulative Series Year of Initial Date of Max.Leak No. of Plant S/G Startup Leakage Shutdown Rate Leake rs Trojan 51 1975 l /20/78 3/17/78 1 GPO l*

II 10/16/79 10/16/79 150 GPO 5 Farley 1 51 1977 8/78 3/8/79 9 GPO 1 Ringhals 2 51 1974 l/79 4/79 800 GPO 1 II 1/79 1/26/79 1 GPO 1 Plug**

II N.A. 3/13/80 1000 GPO Plug**

North Anna 1 51 1978 9/21/79 9/26/79 10 GPO

• 2 Takahama 1 51 1974 2/77 2/77 N.A. 3 II 8/78 8/78 N.A. 1

• Elevation Unknown N.A. - Not Available

• • PreviDusly plugged row one tube

ATTACHMENT All of these tube leaks remained stabl~ and did not tncrease in an unstable manner. This demonstrates that tubes subjected to this form of ~egradation are expect~d to d~velop a small, stable leak, rather than a large uncontrolled leak.

There has been a total of 14 identified row one 11 U11 bends that have leaked fro_m this form of degradation. These have all occurred in Westinghouse mod~l 51 steam generators. There are currently 8622 row one tubes in model 51 steam generators in service .. The 14 leaking tubes account for .16% of the tubes in service.

In addition to the row one leakers previously identified, there have been two incidences of row one tubes having sudden large leaks. These leaks appeared to be the result of different forms of degradation. The Surry 1

'leak was attributed to the denting phenomenon and the Dael 2 leak was attributed to excess ovality.

The 14 leaking tubes, as a minimum, have taken one year to develop into a leak with many takirig three to four yearsL

FIGURE 1 PULL TROJAN "U" BENDS NON-DESTRUCTIVE EVALUATION METALLOGRAPHIC EVALUATION CAN MODE OF DEGRADATION &

LOCATION BE DETERMINED DO ARCHIVED SURRY

&TURKEY POINT "U" BENDS 1-------1 PREVENTIVE PLUG INDICATE SAME CHARACTERISTICS ROW ONE TUBES CAN NOE TECHNIQUE CAN TUBE PROPERTIES CAN FABRICATION PROC~-;-1 CAN SERVICE HISTORY BE DEFINED TO . SPECIFICS BE BE CORRELATED WITH BE DEVELOPED FOR CORRELATE WITH LEAKS CORRELATED WITH LEAKS LEAKS FIELD USE REVIEW TUBE REVIEW FABRICATION REVIEW PRESENT IMPLEMENT PROPERTIES & RECORDS PROCESS OF TUBES & PLANT OPERATING FIELD INSP. &

& IMPLEMENT PREVENTIVE IMPLEMENT PREVENTIVE PRACTICES &

PREVENTIVE PLUGGING PLUGGING REVISE IF NEEDED PLUGGING REVIEW PRESENT FABRICATION PROCEDURE &

MODIFY IF INDICATED PREVENTIVE PLUG ROW ONE TUBES