3F0398-14, Submits Response to GL 97-05, Steam Generator Tube Insp Techniques

Submits Response to GL 97-05, Steam Generator Tube Insp Techniques

ML20248L420
Person / Time
Site:	Crystal River
Issue date:	03/17/1998
From:	Cowan J FLORIDA POWER CORP.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
3F0398-14, 3F398-14, GL-97-05, GL-97-5, NUDOCS 9803200205
Download: ML20248L420 (12)

Text

_ - _ _ _ _ _ _ _ _ _ _ _ - _ _

l .

i Fl:rida I Power

!?"Sao,"

CLN.*.i os.m.n -

March 17,1998 )

3F0398-14 U.S. Nuclear Regulatory Commission Attn: Document Control Desk Washington, DC 205554)001

Subject:

Generic Letter 97-05, " Steam Generator Tube Inspection Techniques"

Reference:

NRC to FPC letter,3N1297-14, dated December 17,1997

Dear Sir:

The purpose of this letter is to submit the response to Generic Letter (GL) 97-05, " Steam Generator Tube Inspection Techniques." As discussed in Attachment A, there are only two degradation mechanisms for which Florida Power Corporation (FPC) utilizes a " sizing" technique for Crystal River Unit 3 (CR-3). These two degradation mechanisms are mechanical wear at tube support plates and the first span pit-like intergranular attack (IGA) in the "B" Once Through Steam Generator (OTSG). All other indications of service induced tube degradation are plugged or repaired.

If you have any questions regarding this information, please contact Ms. Sherry Bernhoft, Manager, Nuclear Licensing at (352) 563-4566.

Sincerely,

/Id. Mc  %

John Paul Cowan Vice President .- ,

Nuclear Production gil . I') I JPC/JM/LVC D Attachments ,

xc: Regional Administrator, Region II Senior Resident Inspector lllllllllllllllllllllllll*ll0lll VNRR Project Manager i

CRYSTAL RIVER ENERGY COMPLEX: 15760 W. Power Line Street

• Crystal River, Florida 344284708 + (352) 7954486 A Florida Progress Company 9803200205 980317 PDR ADOCK 05000302 P PDR _

o, U.S. Nuclear Regulatory Commission 3F0398-14 Page 2 of 2 STATE OF FLORIDA COUNTY OF CITRUS l

l John Paul Cowan states that he is the Vice President, Nuclear Production for Florida Power Corporation; that he is authorized on the part of said company to sign and file with the Nuclear Regulatory Commission the information attached hereto; and that all such statements made and matters set forth therein are true and correct to the best of his knowledge, information, and belief.

$ /2 c n John Paul Cowan Vice President Nuclear Production Sworn to and subscribed before me this Wday of //'/4fv/t ,1998, by John Paul Cowan.

0/tt// tac llNlhf Signature of Notary Public State of Florida commseONMO C000ENG

ex=s1*=a  :

! ___====*2_'*"""'h, (Print, type, or stamp Commissioned Name of Notary Public)

Personally / Produced Known -OR- Identification

U.S. Nuclear Regulatory Commission Attachment A 3F0398-14 Page 1 of 9 FLORIDA POWER CORPORATION CRISTAL RIVER UNIT 3 DOCKET NO. 50-302/ LICENSE NO. DPR-72 GENERIC LETTER 97-05 RESPONSE REGULATORY DOCUMENT INVOLVED: NRC Generic Letter (GL) 97-05, Steam Generator Tube inspection Techniques GL 97-05 REQUIRED INFORMATION:

I in GL 97-05, the NRC required licensees to reply to two items:

(1) whether it is their practice to leave steam generator tubes with indications in service based upon sizing, (2) if the response to item (1) is affirmative, those licensees should submit a written report that includes, for each type of indication, a description of the associated nondestructive examination method being used and the technical basis for the acceptability of the technique used.

FPC Response to NRC Item (I)

There are two types of indications for which CR-3 applies a sizing technique to leave tubes in service. These indications are:

1. Mechanical wear at tube support plates.

2. First span pit-like (intergranular attack) IGA in the "B" Once Through Steam Generator (OTSG).

The indications left in service are those which are sized as less than the Improved Technical Specification (ITS) criteria of 40% throughwall. The CR-3 steam generators are Babcock and Wilcox Model 177FA Once Through Steam Generators, commonly referred to as OTSGs. The tubing material in the CR-3 OTSGs is 0.625" nominal diameter, 0.037" nominal wall thickness Inconel 600 which has been sensitized due to post-fabrication stress relief of the i OTSG assemblies.

FPC emphasizes that there are no signal to noise (S/N) or voltage threshold based techniques currently in use at CR-3.

L s

U.S. Nuclear Regulatory Commission Attachment A

'. 3F0398-14 Page 2 of 9 FPC Response to NRC ltem (2) - General The FPC Quality Assurance Program for CR-3 implements the guidance found in the ASME Boiler and Pressure Vessel Code, Sections V and XI. At this time, CR-3 adheres to the 1983 ASME Code Edition, with Summer 1983 Addenda. Additionally, technical support for sizing degradation specific mechanisms is provided by the EPRI PWR Steam Generator Examination G_uidelines, Revision 4 (EPRI TR-106589-V1). Appendix H of the Guidelines, " Performance Demonstration for Eddy Current Examination," is used as a technical guideline for developing, qualifying and using sizing techniques.

Consistent with the industry initiative for steam generator program improvements, FPC plans to upgrade the CR-3 Steam Generator Program to be consistent with the guidance found in NEl 97-06, Steam Generator Program Guidelines, prior to the scheduled 1999 refueling outage. Adherence to these NEl guidelines will require the technical basis for our inspections to be upgraded to the current revision (Revision 5) of the EPRI PWR Stear.; Generator Examination Guidelines. -

FPC Response to NRC Item (2) - Mechanical Wear (a) Description of nondestructive examination (NDE) method used The CR-3 tube support plates are 1.5" thick carbon steel. There are 15 support plates in each OTSG. The tubing passes through broached tre-foil openings in the support plates. Because of the tre-foil design, there are three regions of contact between the support plate and the tubing. Each portion of the support plate which contacts the tubing is referred to as a " land." The uppermost support plate, the fifteenth, has drilled holes for the outer most two or three tubes around the periphery of the bundle.

The purpose of these holes is to drive the steam back toward the center of the bundle.

Mechanical wear indications are identified (detected) by the use of the bobbin coil exam at CR-3. The exam is performed with the ZETEC A-510 M/ULC mid-range bobbin probe. Wear is detected by scrolling through each support structure with a 400/200 kHz differential mix. Wear indications are initially sized using a 400/200 kHz -

differential mix phase curve established from ASME 20%,60% and 100% machined holes in the calibration standards. If the indication at the support plate is determined to be greater than or equal to 40% throughwall using this sizing technique, then the indication is considered a non-quantifiable indication (NQI) using the bobbin technique.

, At the completion of the bobbin exams, any support structure indications identified as NQI, and wear indications with a signal amplitude greater than one volt, but sized less than 40% throughwall, are inspected with a rotating coil probe. The probe used for the rotating coil inspection is the ZETEC .520 Plus Point Delta Probe with the 0.115" diameter mid-range pancake coil, the Plus Point coil, and the .080" diameter mid-range I

U.S. Nuclear Regulatory Commission Attachment A

'. 3F0398-14 Page 3 of 9 f

pancake coil. The use of the combination probe containing three diff: rent coils allows the analyst to differentiate between a crack-like indication and a mechanical wear indication. If the indication is determined to be indicative of mechanical wear, then the indication is sized using the 300/100 kHz mix on the .080" pancake coil. A calibration curve (amplitude based) is established for the .080" pancake coil from 0%, 20% and 50% nominal machined wear flaws. These machined flaws have a simulated OTSG broached tube support section installed over them to provide a representative eddy current signature. if the indication is not attributable to wear, then the indication is identified with the appropriate three letter code (for example SAI, SCI, VOL, etc.).

Indications not attributable to wear are plugged. Filters are not used when sizing wear indications. In addition, sizing of wear indications at dented support locations is not performed at CR-3. [During the 1997 inspection, only six dents were recorded at tube support plates. Dents are recorded by the analyst if the voltage amplitude is greater than or equal to 3.0 volts. Voltage normalization is set at 4.0 Volts on the 4 x 20%

flat bottom ASME holes.] Figure 1 provides a logic diagram depicting the tube support plate wear inspection process (Attachment B). .

FPC emphasizes that no crack-like indications have been observed at CR-3 support plate intersections to date.

(b) The technical basisfor acceptability of the technique used CR-3 has validated the existence of mechanical wear at the tube support plate interfaces via tube pulls from the CR-3 "B" OTSG. Four tubes had sections removed in 1994. A summary of non-destructive and destructive examination results for these tubes were provided to the NRC in References 9 and 10. Additionally, EPRI TR-106483, Analysis of Steam Generator Tubing from Crystal River. Unit 3, has recently been published and is available for additional reference.

CR-3 identified two morphologies of wear during the 1994 tube pulls. The first f morphology is generally referred to as tapered wear. This morphology is characterized j by planar wear scars running axially along the tube and coincident with one of the lands of the broached support. The end of the wear scar nearest the support plate edge is typically deeper, thus giving a tapered appearance. The other morphology seen at CR-3 is rounded or D-shape wear. These wear indications are small in length and width, typically rounded, coincident with the support land, and usually occur near the edge of a support land. Both of these morphologies are routinely observed with the rotating coil examinations of wear indications at CR-3. However, the bobbin coil examinations are incapable of distinguishing between these two morphologies.

Destructive examination of the tubes pulled in 1994 indicated that the mechanical wear samples varied in depth from !! % to 35%. Several of these samples were burst tested  ;

during the destructive examination phase of the study. The wear indications tested did l l I 1

U.S. Nuclear Regulatory Commission Attachment A

. 3F0398-14 Page 4 of 9 not significantly reduce the burst pressure of the tubes when compared to a defect free sample of the tubing. The tubing sections tested with wear indications burst above 10,000 psig. The burst pressure of defect free tt'bing was 10,100 psig to 11,100 psig as recorded in EPRI TR-106483. These values are well above the Regulatory Guide '

l.121 calculated structural integrity requirement of 4050 psig for CR-3.

Mechanical wear in OTSGs is generally a slow growth phenomena.

Industry experience demonstrates that there have been no forced outages due to tube support plate mechanical wear at OTSG plants. Additionally, there are very few tubes plugged in OTSGs as a result of mechanical wear indications. The initial indications of OTSG wear were identified in the late 70s at the Oconee units.

The use of the bobbin coil probe, versus the pancake coil, for sizing OTSG wear indications is considered to provide a systematically conservative number for indicated percent throughwall. Generally, as the wear indication grows and increases in throughwall percentage, the eddy current indication signal to noise ratio will increase, giving a more representative indication of percent throughwall. The smaller eddy current indications are affected more by noise, and therefore the eddy current phase angle is more influenced towards the noise plane, resulting in higher indicated throughwall percentages. Additionally, if more than one tube support plate land is wearing the tube, the bobbin coil eddy current will provide a more conservative estimate of throughwall depth, since the bobbin coil cannot distinguish between single sided, double or triple sided wear. For these reasons, CR-3 uses the bobbin coil as the initial screening and sizing tool for wear.

From the 1994 pre-pull cddy current data, these results were obtained by using a bobbin coil 400/200 kHz differential mix:

TUBE ID TSP MAX MET DIFF MIX ROW TUBE LOCATION %TW %TW

~

68 46 7 32 41 68 46 9 13 54 68 46 9 19 20 i

72 49 7 16 24 72 49 9 11 35 109 71 3 14 NDD*

109 71 7 33 40 136 26 7 35 32

• NDD - No Detectable Degradation

U.S. Nuclear Regulatory Commission Attachment A 3F0398-14 Page 5 of 9 Using only these data points, and neglecting the 14% throughwall indication that was not detected with eddy current, the calculated Root Mean Square Error (RMSE) is about 19%. If the missed indication is included, the RMSE is about 18%. Note also that as the metallurgically determined throughwall depth increases, the indicated depth becomes more relevant to the real depth. However, the quantity of data points here is not suf0cient by itself to consider the technique qualiGed.

Because of the similarity between the rounded wear morphology and OTSG impingement morphology, CR-3 also considers the EPRI quali6 cation of OTSG impingement to provide technicaljustification for the sizing ability of the bobbin probe for wear indications. The indications used to qualify the impingement technique consisted of OTSG tube pull samples supplemented with machined Daw samples. The indications ranged from 16% to 98% throughwall for this technique qualification. The calculated RMSE, using the 400/200 kHz differential mix technique, is recorded on EPRI Performance Demonstration Database, Examination Technique Specification Sheet (EPRI ETSS) #96002 as 11.5%. This mix is referenced for CR-3 use because the process parameters for this qualification most closely resemble the process parameters used for CR-3 inspections. CR-3 uses a different voltage normalization, more current computer hardware and software, and faster probe speeds than speciGed on the EPRI ETSS. However, site specific testing has demonstrated that these differences do not change the technical basis for the technique, For wear at. broached tube support plates, the 0.080" mid-range pancake coil 300 kHz channel and 300/100 kHz mix are used to screen the data and identify indications. Due to the support plate and crud influence, the 300/100 kHz mix is used to depth sire the indications -- if they are determined to be due to wear (see Figure 1). An amplitude based calibration curve for the mix is established, using 0%, 20%, and 50%

manufactured wear marks, which are under a simulated broached tube support plate (TSP). The recorded depth of the indication is determined by locating the largest ampli:ude signal in the indication. Filters are not used during this depth sizing. EPRI ETSS #96905 contains the data and parameters used to develop this technique. This sizing procedure is based on the analysis of 26 data sample points. The data samples are manufactured samples representative of wear indications found at OTSG plants.

The manufactured samples ranged in depth from 22% to 100% throughwall. The calculated RMSE, using the 300 kHz differential signal, is recorded on EPRI ETSS

1. 96905 as 5.8%. The RMSE for the 300/100 kHz differential mix is not explicitly determined by the EPRI ETSS. However, it is reasonable to conclude that the RMSE for the 300/100 kHz mix would be approximately the same since the mix is simply l removing the support plate and crud influences to give a more distinguishable signal. l CR-3 has a limited number (1621) of drilled tube support plate locations. These

)

l locations are limited to the 6fteenth support plate, at the periphery of the bundle. The {

purpose of the drilled locations (versus the broached locations throughout the

U.S. Nuclear Regulatory Commission Attachment A 3F0398-14 Page 6 of 9 remainder of the bundle) is to direct the steam flow back toward the center of the bundle at the uppermost elevations. During the 1997 100 % bobbin coil inspection, there were no wear indications identified at these drilled support plate locations.

In summary, CR-3 relies upon the bobbin coil for detection of mechanical wear at TSPs. Characterization of wear is performed by analyzing the bobbin coil signals, as well as the use of rotating coil exams. Detection and characterization have been confirmed via tube pulls from the "B" OTSG in 1994. Sizing is performed with both the bobbin coil and pancake coil for wear indications.

FPC Respcmse to NRC Item (2) "B" OTSG First Span Pit-Like IGA (a) Description of NDE method used The CR-3 "B" OTSG has a large number of pit-like IGA indications in the first span.

The first span is defined as the region above the lower tubesheet secondary face and below the first tube support plate. These indications are attributed to secondary side resin intrusions during the early years of plant operation. Because of the large number of indications in this region of the "B" OTSG, a site-specific sizing technique was developed and qualified (in accordance with the guidelines found in Appendix H of the EPRI Guidelines) for use only on these indications. This sizing technique is referred to as the "B" OTSG first span pit-like IGA regression technique. References 1,3, and 4 provide detailed descriptions of the eddy current protocol used to identify, disposition, and size these indications in the "B" OTSG during the 1997 inspection.

A summary of the protocol that CR-3 uses for dispositioning the first span IGA is:

1) The area of interest is limited to the first span of the B OTSG.

2) If an NQI is called by the mid-range (MR) bobbin exam [all freespan bobbin indications are identified as NQIs using the MR bobbin exam], or the tube has been previously identified, the entire length of the first span of that tube is inspected using the rotating coil exam technique. The rotating coil exams consist of a mid-range Plus Point coil, 0.115" pancake coil and 0.080" mid-range pancake coil in a motorized rotating pancake coil (MRPC) prebe head with three shoes.

3) The Plus Point coil is the primary coil relied upon for identification (detection) and characterization of the indications. The analysts must review the C-scan plot for the entire length of data acquired. Indications that are volumetric in nature are identified as VOL. Locations that have more than one VOL indication in the same circumferential plane are identified as MVI (multiple volumetric I

l l

E-----------------------_------------------------------------------------------------------------------------------_ -

U.S. Nuclear Regulatory Commission Attachment A l -

3F0398-14 Page 7 of 9 indications). If another type of indication is observed, the appropriate three letter code is used for that indication (examples: SCI, sal, etc.).

4) A high frequency bobbin coil exam is performed on the tubes that have a VOL indication. This exam is used to ensure that the same probe and parameters used during the development of the regression technique are maintained.

5) The analyst uses the VOL indications from the rotating coil exam as a "roadmap" to obtain bobbin signals off the high frequency data. If the analyst is capable of correlating a bobbin signal to the VOL location, the analyst applies the CR-3 specific regression technique (via the ZETEC Regression Tool). If the analyst is not capable of correlating a bobbin signal to the VOL indication, the analyst identifies this situation by the use of the NQI code for that location in the high frequency bobbin results.

6) Tubes with a high frequency bobbin NQI are plugged. .

7) Tubes with an MVI call are plugged [due to the inability of the bobbin coil to acceptably distinguish individual volumetric indications when multiple indications are present at the same axial location].

8) Tubes with an indicated regression throughwall percentage 240% are plugged.

9) Tubes that have first span IGA indications which are within one inch of the OlS support lower edge or within one inch of the lower tube sheet secondary face are identified and plugged.

10) Tubes that have VOL indications which are all sized as less than 40% throughwall have their historical data reviewed. These tubes have at least one previous inspection reviewed to ensure that there is historical evidence of the indication.

Twenty seven tubes with "B" OTSG first span IGA indications do not have previous high frequency bobbin data (these tubes were last inspected in 1990, when a mid-range bobbin probe was used). For these 27 tubes, the mid-range bobbin data is reviewed to verify that these indications were present in 1990, but the regression tool is not applied. The remainder have, as a minimum, the most recent high frequency bobbin inspection reviewed and the historical indications sized using the regression technique. If the analyst is not capable of clearly sizing a bobbin indication in the historical data for that VOL indication, the code INF (indication not found) is used to denote this fact. The indication is then

conservatively assigned an initial depth of 0 when determining growth rate for I

that indication. Ilundreds of indications had multiple years of data reviewed to l further validate the CR-3 position that this degradation mechanism is dormant. As '

a result of the 1997 inspections, and based upon the growth distribution curves

U.S. Nuclear Regulatory Commission Attachment A I

. 3F0398-14 Page 8 of 9 submitted in Reference 10, CR-3 determined that these indications continue to be dormant.

11) Any tube which contains an indication that demonstrates a growth of greater than 10% throughwall by regression sizing is conservatively plugged regardless of 1997 indicated percent throughwall. In 1997, this growth criteria was applied to 1992 to 1997,1994 to 1997, and 1996 to 1997 high frequency bobbin coil data comparisons.

The OTSG inspection performed in 1997 marked the initial use of the first span pit-like IGA regression sizing technique. Reference I provides a summary of results from the CR-31997 OTSG inspection.

(b) The technical basisfor acceptability of the technique used CR-3 pulled tubes from the "B" OTSG in 1992 and 1994. These tube pulls were performed with the intent of identifying and assessing the degradation mechanism which caused the indications in the first span of the "B" OTSG. The NRC has previously been provided with the eddy current and destructive examination data from these tube pull campaigns. References 5 through 9 contain information pertaining to these tube pull evolutions at CR-3.

The CR-3 "B" OTSG first span pit-like IGA regression technique was developed exclusively with data from these 1992 and 1994 tube pulls. As documented in Reference 1, the RMSE for this technique, when used solely on the CR-3 "B" OTSG first span pit-like IGA pulled tube indications, is calculated as 9.8 % throughwall.

The NRC has been provided with the basis information which described the development, implementation, and adequacy of this technique. Specifically, References 3, 4, 5, and 6 contained FPC to NRC correspondence specifically addressing inspecting and dispositioning these "B" OTSG first span pit-like IGA indications. In the Safety Evaluation for CR-3 ITS License Amendment No.158 (Reference 2), the NRC Staff concluded that with a primary-to-secondary operational leakage rate limit of 150 gallons per day, and the changes implemented to the CR-3 inservice inspection (ISI) requirements in ITS 5.6.2.10 by License Amendment No.

158, that the licensee would be able to disposition the "B" OTSG first span pit-like IGA in an acceptable fashion.

In summary, CR-3 relies initially upon the bobbin coil for detection of "B" OTSG first span pit-like IGA indications. Characterization of IGA is performed by the use of rotating coil exams. Detection and characterization has been confirmed via tube pulls from the "B" OTSG in 1992 and 1994 Sizing is performed with the use of a l regression technique applied to the bobbin coil data for "B" OTSG first span pit-like IGA indications.

i

l

, U.S. Nuclear Regulatory Commission Attachment A i 3F0398-14 Page 9 of 9 Co'nclusion CR 3 leaves only two types of indications in service by the use of sizing techniques. CR-3 has demonstrated through tube pulls that these degradation mechanisms are mechanical wear at broached tube support plate intersections and first span pit-like IGA in the "B" OTSG. CR-3 has tested and confirmed the structural integrity of tubes with these degradation mechanisms via laboratory tube burst testing and in-situ leakage testing. Based upon years of monitoring these indications, these two mechanisms do not exhibit rapid growth rates. CR-3 implements the ITS required 40% throughwall plugging criteria with respect to these indications. Thus, FPC concludes that the CR-3 eddy current inspection and dispositioning protocol for these indications is well defined and based on well established technical positions.

References

1. FPC to NRC letter,3F1297-22, dated December 5,1997 --

2. NRC to FPC letter, Amendment No.158 to DPR-72, dated October 28,1997

3. FPC to NRC letter,3F0897-09, dated August 20,1997

4. FPC to NRC letter,3F0397-16, dated March 27,1997

5. FPC to NRC letter,3F0496-18, dated April 15,1996

6. FPC to NRC letter,3F03%-19, dated March 21,19%

7. FPC to NRC letter,3F1295-03, dated December 5,1995

8. FPC to NRC letter,3F0595-07, dated May 31,1995

9. FPC to NRC letter,3F0595-05, dated May 31,1995

10. FPC to NRC letter,3F1194-10, dated November 30,1994 l

1

U.S. Nuclear Regulatory Commission Attachment B 3F0398-14 Page1of1 FIGURE 1 WEAR SIZING LOGIC DIAGRAM i

i Mid Range Bobbin Exam with Differential 400/200 kliz Mix

( >

u r 3 *NOT WEAR

• Indication Within or NO TEcliNIQUE NOT Influenced by the Tube  ;

APPLICABLE. IDENTIFY q Support Plate? j AS NQI AND SPIN WITli AIRPC TO DISPOSITION TES u

r 3 .-

Does the Support Plate Location Contain an IES . PLUG "

q Indicated Dent? j TUBE NO Characteristic' of Wear Signal and <40% No j ,

Throughwall on Mix? identify as NQI

( ) "

1ES , ,

_ _ , inspect Area ofInterest with u MRPC to Disposition gg Report as % Throughwall on Bobbin 400/200 kilz r

} ,

2 Mix Channel j Characteristic of

(

< Wear Signal?  ;

IES

" *NOT WEAR

• Greater an i Volt ( S Indicated Amplitude? IES Report as % TECIINIQUE NOT

' > Throughwall on .080" APPLICABLE. IDENTIFY Pancake 300/100 kHz Mix AS VOLUMETRIC, NO ( Channel j CRACK LIKE OR NO LEAVE DEGRADATION FOUSV TUBE i r ,

IN  : 2 40% Indicated SERVICE YO Throughwall? 17S v i Characteristics of urar on the bobbin exam include that the signal is discernible from the background noise, the signal provides a classic lissajous proGle, ar.d the signal rotates as expected on the raw frequency channels.

2 Characteristics of wrar on the MRPC exam include the absence of cracks as indicated by the Plus Point coil, the wear indication has a tapered or rounded OTSG wear profile. the wear profile is coincident with the location of a support plate land, and the absence of pit-like volumetric indications.

- _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ . . _ _ _ _ _ ._.