ML20236E458
| ML20236E458 | |
| Person / Time | |
|---|---|
| Site: | Indian Point  |
| Issue date: | 09/23/1987 |
| From: | Cook K OAK RIDGE NATIONAL LABORATORY |
| To: | Hum M Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20236E423 | List: |
| References | |
| CON-FIN-A-9478-3 NUDOCS 8710290229 | |
| Download: ML20236E458 (10) | |
Text
{{#Wiki_filter:.
~
, -0923-26 .:-
' post omcE sox x
.CAK RIDGE NATIONAL LABORATORY '
OPERATED SV W ARTW NARIETTA ENERGY SYSTEMS. INC Septemoer 23,'1987 Mr. M.-R.. Hum Materials Engineering 8 ranch Mail Stop P-842 Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission
Dear Mr. Hum:
Travel to Madison, Pennsylvania, September 14-16,1987, to Provide Technical Assistance to NRC for Evaluation of Flaw Indication in Indian Point Unit 2 Reactor Pressure Vessel _ At your request via Work Task 87-1 (FIN A9478-3), I attended the subject meeting at the Westinghouse Waltz Mill f acility. A copy of the attendance 1ist'is. Attachment 1. As requested, John Gieske and I met with you and R. Mc8rearty the night before the demonstration meetings to be held by Consolidated Edison and Westinghouse personnel on September 16 and 16. This ratner routine p remeeting format allows NRC to provide the contractors with general guidance for the primary task. ! On Tuesday, September 15, George Wasilenko of Consolidated Edison, Rick Rishell of Westinghouse, and Frank Dodd of NES/Dynacon made p presentations. A copy of the planned agenda is Attachment 2. Wasilenko , prestnted some background information and outlined the information to be covered. The stated current goal of Consolidated Edison is to show at the planned October 1987 ibservice inspection that the indication located in
]j the August 1984 in-service inspection is essentially unchanged. This j indication was detected in the lower shell longitudinal weld (No.12) at I the 345' position. Further, Consolidated Edison would like to avoid j augmented inspection in 1990 because ti .y feel that the indication is small enough to be insignificant. .
l 1 attended five previous meetings dealing with the Indian Point Unit 2 l i ndi cation. These meetings were August 11, 1984; August 14, 1984; l October 3, 1984; May 9, 1985; and April 7, 1987. Trip reports were submitted to the NRC for each, in addition, I reviewed documents, as requested, with the most recent one being a document submitted to the NRC by M. Se han, Vice-President, Consolidated Edison of New York, Inc., dated April 21, 1987, that presented the written plans for the augmented i n-s ervice inspection in the fall of 1987. We were informed of the p rimary deviation in this written plan, which is to use Dynacon data I B710290229 871023 - PDR ALOCK 05000247-P PDR
4 M. R. Hum Page 2 September 2L le87 collection and analysis instead of their A-scan recording and enhancement system ( ARES-II) because (1) the art.i-Il system is currently much slower than the Dynacon system, and (2) the recent success of the Dynacon [ ultrasonic data recording and processing system (UDRPS)] system during the Indian Point Unit 3 in-service inspection. Rick Rishell reported that the current plans are to (1) repeat the basic inspection performed in 1984 in accordance to ASME Code requirements and (2) also repeat the special techniques performed in 1984 (i .e., the delta and shadowing techniques). In addition, Westinghouse plans to perform s tat e-o f-t he-a rt inspections (item Ill.C on the agenda) using large ultrasonic focused search units [~14 cm in diameter (5.5 in.)] purchased f rom France. Since items Ill.A and III.B will be as near identical as possible to the August 1984 in-service inspection (the same mechanical and electronic equipment to be used), nty comments will be limited to item Ill.C of the agenda. The four large focused search units purchased from Intercontrole consist o f a 2-M Hz 45' longitudinal wave unit, a 1-MHz 45* shear wave unit, a 1-MHz 60' shear wave unit, and a 1-MHz 0* longitudinal wave unit. These search units each weigh about 2.7 kg (6 lb). Tht focusing is accomplished by specially designed lenses placed on the surf ace of the piezoelectric eleinent used to generate the sound. The f requencies and design details were provided to Westinghouse by the French designers based on their experience with clad vessels inspected in Eu rope. According to Westinghouse, a focused Deam at the back surface of the approximate 22.9-cm-thick (9-in.) vessel results in a near circular diameter (somewhat
. elliptical) of about 40 mm (0.4 i n. ) . Thus, beam spread is greatly reduced and theoretically the sensitivity should be improved.
In addition to the general results presented by Rishell using demonstration (calibration) block IPP-2T-REV1, he also provided some background on the Pacific Northwest Laboratory (PNL) activity. Larry Reid of PNL was collecting data with a 2.25-MHz, 50% bandwidth, 7.6-cm (3-in.) water-focused search unit fran four or five notches in the calibration block. Plans were to collect the data with Dynacon's VDRPS and then transmit a copy of the data via magnetic tape to PNL later. PNL would then have to alter the format and reprogram the synthetic aperture focusing technique (SAFT) al gorithm to analyze the data. The SAFT activity will require considerable effort on PNL's part to be ready for the scheduled in-service inspection in mid-October. Frank Dodd of Dynacon presented hard copies of data taken with the French large-diameter search units as displayed by UDRPS on a Ramtek di splay module. Two types of color displays are normally used. One is the uniform block map and the other is a modified B-scan display. In addition, other displays of the data are available, including a perspective view (sometimes referred to as mountain or waterfall views of
4 M. R. Hum
. Page 3 Septemaer 23, 1987 A-scans). Dodd stated tnat the 12. e-mm-deep (0.5-i n. ) noten C could be resolved with the 0" searcn unit, but not the shallower ones. Tne 4e longitudin61' shear wave units resolve the 7.62-mm-deep (0.3-in.) noten B, but not the 4.57-mm-deep (0.18-in.) notches G and H. The 60 snear wave unit marginally resolves the 12. 5 -mm-d eep notch. Of course, tne 0*
longitudinal wave search unit is resolving notch C partially because it is wide [~3.2 mm (0.125 in.)]. The best resolution and most useful of the
'large focused units appears to be the 2-MHz 45 longitudinal wave unit.
Based on time separation analysis of the data from notch B (the upper and lower extremes of the reflector), it appears that a 51-mm-deep (0.2-in.) notch could be sized (assuming a linear response). Tne accuracy is yet to be established; however, the data have been stored and various analyses will be performed. In fact, we expect data for review prior to the in-service inspection. I suggested that a 3-MHz longitudinal wave focused unit might increase the resolution. Dodd and Adamonis stated that the clad might preclude the use of the higher frequency, as the French su spe cted, based on their extensive experience. John Gieske pointed out that if the indicdtion is less than 5.1 mm deep that it could not be sized any less than the 5.1 mm depth. Considerable discussion ensued. After lunch, Don Adamonis presented some results of the Indian ko+ Unit 3 in-service inspection perfor.ned in June 1987. Originally three indications were reported in the intermediate shell longitudinal weld at about 300 [~20.3 cm (8 in.) above the intermediate-to-lower shell circumf erential weld]. Using the UDdPS system (in much the same manner as planned for Indian Point Unit 2), the inspectors reported that the tnree indications were detected by three different path lengths from one indication located near the outside surf ace. Tnus, the one indication is located approximately 19.8 cm (7.8 in.) below the innor diameter surface of the vessel and extends approximately 5.97 mm (0.235 in.) toward tne outer surface. The length of the buriea indication was reported as approximately 11.4 cm (4.5 in.) . No indication of the statistical accuracy of these measurements was provided. I assumed the primary purpose for the presentation was to demonstrate the use of UDRPS; however, as you requested, my comments follow.
- 1. The one-indication conclusion is plausible. ;
- 2. The buried flaw conclusion is plausible. {
- 3. The UDRPS data presentation method is very good. l 4 Tne digital stored data approach is excellent. i S. The conclusion that the indication is slag seems reasonable. !
6, The method used to determine the size of the indication appeared to be 1 a reasonable approach; however, the accuracy of the measurement was not provided. Af ter tne indian Point Unit 3 presentation, we walked to the demonstration a rea. Larry Reid of PNL was collecting SAFT data from the calibration block while we reviewed the Ramtek display of digitized data collected l l l _ _ _ _ _ _ _ _ _ _ _ _ E j
1 l
- l. .
~~ ~~ " - ~ ~ ~ ~
L M. R Hum
, .Page 4 .]
> September 23, 1987 i j
'with the_large-diameter French focused units. Frank Dodd demonstrated the measurement of notch B and obtained a value .of ' 8.03 mm- ;0.316 in.). We' also observed the perspective view of the multiple' A-scan data for this
~
We observed 0 ' longitudinal d notch and for .the shallower notch H. pulse-echo data, 45' longitudinal pulse-echo data, 45' snear_ wave pulse-echo data, anti 45 snear wave transmitter 0 longitudinal receiver data ;(delta) . _ In fact, tne latter (delta focused probe data) provided a j dz value _ of 4.39 m (0.173 in.) for notch H. However, the location of the
' start-stop signals (reflector extremities) was very marginal as 'previously-reported by Dodd. .
On . Wednesday, September 16, we : returned to the Waltz Mill facility to continue our discussions. Rick Rishell presented some data and plots showing the notch length vs the measured values for Sect. XI of the ASME-Code .(501 DAC) and also the half-maximum signal amplitude. No surprises j were presented. RishcIl also presented some data based on the French successive dB drop method for length c5 well as a 6-dB drop method. When 1 the depth of the reflector approaches 5.1 mm (0.2 in.), the length measurement becomes uncertain also. FLether information is to be provided when their analysis is completed. Gieske and I recommended tnat rf digitized data also be collected from the > Indian Point Unit 2 indication 'for possible analysis. For example, such , data _might -be processed by time-domain deconvolution techniques if the -! Dynacon UDRPS and/or PNL SAFT analyses are not successful. Dodd agreed to collect the data if needed and indicated that it would take little additional effort. We returned to the demonstration area to observe data collected by the unfocused search units. Dodd displayed the delta data f rom notches A, B, C,'G, and H. A measured value could not be obtained for notch A; however, the dz values for G and H were 4.37 and 4.39 mm (0.172 and 0.173 in.), respectively (albeit the determination was marginal also).- Tne dz for notch B was measured as 5.1 m (0.201 in.). Tnis is about 30% less than the actual depth for this one measurement. In a letter from John O'Toole of Consolidated Edison to S. A. Varga of the NRC, dated September 21, 1984, ne provided information concerning probable errors for tip location of il.5 ps [or i3.8 m (0.15 in.)] (see Attachment 3). Tnerefore, the analysis of the recently collected data must lower the 3.8 m number to increase confidence in the sizing measurement. My personal opinion is that the large focused 45* longitudinal wave search unit might provide the necessary accuracy improvement. Of course, this is to be determined. A wrap-up meeting was oegun at 1:00 p.m. Wednesday , September 16. Discussions about the goals, the use of SAFT, radiation scnooling, errar bands, quality assurance, and in-service inspection schedule were held.
M. R. Hum Page 5 September 23, 1987 Gieske and . I were. told to make reservations' for' October 20-27,- 1987, in
.Elmsford, New York, for the planned in-service inspection of Indian Point Unit.2.
The demonstration meeting was very beneficial and timely. The approach selected does' address the main issues (and prior recommendations) and will provide much more retrievable third-party data than the August 1984 i n-service inspection.. In addition, the data taken from the calibration block (IP.P-2T-REV1) are stored on -digital tape for future use as needed. The primary goal of the, upcoming'in-service inspection (to demonstrate an I essentially unchanged condition for _ the indication) should be . possible (within the uncertainty already- known). The secondary goal .of establishing better measurements of the indication's length and depth (with. adequately improved accuracy) is not as certain; however, the data needed to. evaluate the secondary goal will apparently be available (even after the in-service inspection). In addition, an immediate resolution of the 'second ary goal may not be required in October 1987 :since the successful achievement of the primary goal would apparently allow operation for 3 more years under the augmented inspection conditions. Sincerely yours, fl. hf K. V. Co ok Nondestructive Testing Group
. Metals and Ceramics Division KVC:j lb Attachments cc/att: C. Y. Cheng, NRC W. H. Hazelton, NRC .
J. B. Henderson, NRC A. P. Malinauskas R. W. McClung J. G. Pruett G. M. Siaughter K. V. Cook / File
. __ i .
[: c Dy) A. /0. T [ Kii &' h?{. ' /G5i-TC77(4.
. r"/2C612At?7 /MEE77LCW ATTACHMENT I j c7 7 c G ,5 r s / L L S/rE ;
i, 3.z Avww, s q2 (g75D L56 -6 c'36 2
- 3. %Geoye.t A%re si k k _ Cu E Psov 4 r. d v r W R3Ib i ziz) </ w - Ys M vi z') - 3 7 4 - 3 9 !.5~
5 ]A d .Sthe c0G ll H f. l. 482, 828~ 19.T8 - i r feebe)rt p. iv Srenty ' il'6Hpft v . w g c. (v.Q 337-s-5M 4: M AST. t u +4 v u Us u g.c (3o ) ) 49 2 1 b9 Y
't K V Cook . c> n n' s. (s i s) - s 7 y - 4 9'22 5 ' J oL . E G3ies k-e 6 4.A Sos) - W4 - 63y(,
1 2. Aio tuneC Mu_.G s e c L. c.Nsio
.iC5T$ D .
't iz) 379 scic
. I 412) 25p - 639 f it. Nank11ecM yaco ~ 525-S33G l2. ffw C#JMU N&sQ)YM p (45) 9/t'-. 6f/-6 2.s ?
/J &c Jie.ol r SPA 9!Y- Oi - d.t& a l
o O . 4 e
1
. ' ATTACHMENT 2 AGENDA
.INDI AN POINT UNIT ~2 REACTOR VESSEL INDICAT10N'-
INSPECTION PLAN 1
. REVIEW OF DEMONSTRAT10N PHASE RESULTS ,
SEPTEMBER 15-16, 1987 WESTINGHOUSE WALTZ MILL SITE i "F" BUILDING AUDITORIUM
- i. OPENING REMARKS-1I. OVERVIEW 0F DEMONSTRAT10N PHASE OF INSPECT 10N PROGRAM PLAN III. GENERAL RESULTS FROM DEMONSTRATION PHASE A. ASME SECTION XI TECHNIQUES !
B. DELTA / PITCH CATCH TECHNIQUES C. FOCUSED TRANSDUCERS TECHNIQUES IV. TOUR OF DEMONSTRATION PHASE FACILITY A. REVIEW OF DEMONSTRATION PHASE DATA B. REVIEW 0F INDIAN POINT UNIT 3 INSPECTION DATA , C. REVIEW 0F INSPECT 10N EQUIPMENT AND DEMONSTRATION PHASE > SET-UP i i b
r . ATTACHMENT 3 i f.' ATTACHMENT C o. l RESPONSE TO OUTSTANDING ISSUES CONTAINED IN AN
-i NRC MEMORANDUM DATED AUGUST 28, 1984 Consolidated Edison Company of New York, Inc.
Indian Point Unit No. 2 Docket N. 50-247 September 1984 -{p-I
{
. QUESTION 2 p.
The . licensee should' establish -the uncertainty bound for ' depth sizing determined by the ' delta techniaue using the inf9rmation obtained from various
. notch depths in the curved block.
RESPONSE TO QUESTION 2: . 1 1 To determine a conclusive uncertainty bound for any technique - which contains potential operator and system variables, a large number of data points and a significant period of time is required to accumulate data and analyze the
- results. With small sample sizes even the . most accurate technique can give unwarranted indications of variation in data. flowever within the short period i of time available for these mockup exams the uncertainty range of the delta ..
technique was reviewed based on test results-on the _ reflectors in a IPP-2T mockup. These tests were performed , from both sides of the reflector - (where possible) with both of the delta arrangmenets .(TR 22 to TR 20 and -TR 24 to TR 20). Several oprators performed the examinations with the UT test system used l
~
in performing the vessel ' examination. Data from transit time measurements based on total metal path were reviewed. The uncertainty range was I approximated to be on the order of + 2 micro, seconds, which translates to a
. measuring tolerance o@Mn-i_: _over a reflector depth range from 0.1 to 2.0 inches. Data from delta nearurements was also reviewed in terms of the time interval spacing between indications representing the upper and lower extremes of the reflectors. _
The uncertainty range was approximate to be on the ordtr o croseconds for these measurements which translates to a measuring tolerance of]E_'~EM Be cause of time constraints, insufficient data points were taken and a complete statistical analysis of thes e results has not been accomplished. Therefore the uncertainty ranges discussed above are considered inconclusive and have not been factored into the overall test conclusions. l ((.
.}}