ML20212G875
| ML20212G875 | |
| Person / Time | |
|---|---|
| Site: | Quad Cities  |
| Issue date: | 01/15/1987 |
| From: | Johnson I COMMONWEALTH EDISON CO. |
| To: | Harold Denton Office of Nuclear Reactor Regulation |
| References | |
| 3065R, NUDOCS 8701210219 | |
| Download: ML20212G875 (5) | |
Text
, -.
} One First National Plaza Chicago, Illinois Commonwealth Edison o
/
(
- 3
\\
Address Reply to: Post Office Box 767 k
Chicago,Illne 60690 0767 January 15, 1987 Mr. Harold R.
Denton U.S. Nuclear Regulatory Commission Office of Nuclear Reactor Regulation W:shington, DC.
20555
Subject:
Quad Cities Station Unit 2 Winter 1986 Refueling Outage Recirculation System Piping Inspection NRC Docket No. 50-2G5 Rsferences: 1) M.S. Turbak to H.R.
Denton dated December 24, 1986
- 2) M.S. Turbak to H.R. Denton dated December 30, 1986 D3ar Mr. Denton:
The referenced letters and attachments provided a description of flaws found, and of analytical and repair actions taken during the current Quad Cities Unit 2 refueling outage.
Particular attention was called to end cap weld 02A-S10 in Reference 1) and in Attachment 5 to Rnference 2).
The status of weld 02A-S10 in Reference 2) after weld overlay repair during this outage was that seven or eight axial flaws wsre present in the weld overlay with a remaining ligament of sound metal above the crack tips at or below the full structural weld overlay design thickness.
A two inch long circumferential flaw was also found with a remaining ligament at the weld overlay to base metal interface (0.42 inches remaining ligament).
It should be noted that the design thickness of the weld ovarlay was based on an assumed 360 degree, through wall circumferentio. flaw and provided for fatigue growth over the remaining life of the unit.
Although the preferred Commonwealth Edison option was to leave wald 02A-S10 as-is for an additional cycle, a timely NRC review of the referenced documents could not be achieved during the holiday season and CECO elected to add an additional layer of weld metal.
The intent was to increase the remaining ligament over all flaws to greater than the full structural overlay design thickness.
Low heat input and bead overlap parameters were utilized with the aim of minimizing flaw propagation due to the high strains and temperatures of the welding.
O
,\\r 8701210219 870115 PDR ADOCK 05000265 PDR
. Ultrasonic measurements after the additional layer showed a weld metal thickness of 0.48 inches on the cap side compared to the previous 0.42 inches.
Both manual and automated ultrasonic data was collected on the remaining ligaments over the flaws.
The personnel who performed the examinations and made evaluations of the data were qualified at the EpRI NDE Center for examination of weld overlays.
Manual and automated examinations were made of the entire circumference of the cap side of the weld.
The results of the GE SMART Image System generally paralleled the manual examination results with some variations in remaining ligament, length, and location.
The automated examination found one circumferential and 8 axial flaws to be present in the weld overlay metal.
A comparison of the manual and automated UT data for the flaws with the shortest ligaments of sound metal is as follows:
Least Ligament (inches)
Manual Automated Circumferential 0.24-0.32 0.38 Axial 0.3 -0.4 0.35 Axial 0.24-0.28 0.36 Axial 0.26-0.30 0.44 The manual UT data sheet noted that the weld was very noisy around the cracked areas and difficult to examine.
A letter from the GE Lovel III who conducted the automated examination is attached to this letter.
The GE letter also notes high noise levels but states these signals were non-relevant and posed no limiting interpretation problems.
The flaws with the least remaining ligaments were in locations where steam blowouts occurred during overlay welding and localized weld repairs were made.
While the most conservative basis for evaluation of weld 02A-S10 would be to use the manual data, we believe the automated data to be equally valid.
In a conversation with personnel of the EpRI NDE Center, GE, and our Technical Center, it was stated that the expected crack sizing tolerance between examinations was approximately 0.1 inches.
It was agreed that the comparisons of manual and automated data discussed were within expectations for independent examinations.
The weld overlay design thickness of 0.38 inches in Reference
- 2) was based on nominal piping wall thickness and included a i
conservatively calculated 0.030 inches of crack growth by fatigue over a l
30 year lifetime for the circumferential flaw.
A revised design thickness has been calculated based on the minimum wall thickness of 0.979 inches and elimination of the fatigue crack growth allowance.
l Basing the design on minimum wall thickness is consistent with designs previously applied to riser welds.
Elimination of the fatigue crack l
growth allowance is appropriate for an overlay intended for one cycle of l
operation.
The new design thickness for the overlay on weld 02A-S10 is l
0.33 inches.
Commonwealth Edison has considered both sets of UT data in evaluating the remaining ligament of the flaws.
The mean of the manual UT data when averaged with the automated, provides a remaining ligament of 0.33 inches for the circumferential flaw.
The three deepest axial flaws have average remaining ligaments of 0.31 to 0.36 inches.
On this bssis, the short circumferential flaw meets the full structural (ntandard) overlay, design thickness.
The deepest axial flaw is marginally less than the design thickness for a circumferential flaw, but significantly exceeds the requirements for a leakage barrier overlay for axial flaws.
An ASI1E Section XI evaluation in accordance with IWB-3642 has been performed for the minimum ligaments of the manual sizing and ecceptance of the flaws for an operating cycle has been demonstrated.
Commonwealth Edison requests permission to operate Quad Cities Unit 2 for one operating cycle with end cap weld 02A-S10 in the as-is condition as dnscribed in this letter.
Quad Cities Unit 2 will continue to adhere to the restricted leakage detection and leakage limits contained in Generic Letter 84-11 for the upcoming operating cycle.
Plant shutdown shall be initiated for inspection and corrective action when any leakage system indicates, within any period of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, an increase in rate of unidentified leakage in excess of 2 GPM.
The sump level shall be monitored at 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> intervals or less.
Commonwealth Edison commits to providing a final report datailing the results of the entire Winter 1986 piping inspection to your staff by January 21, 1987.
The information entailed in this report is a compilation of reports which were submitted during the course of this inspection to NRR.
Please direct any questions that you may have regarding this matter to this office.
Very truly yours,
/- A60
/M87'k I.M.
Johnson-Nuclear Licensing Administrator RJT/1m 3065R cc:
Resident Inspector - Quad Cities T.S.
Rotella - NRR D.
Danielson - Region III M.C.
Parker - IDNS
GENERAL $ ELECTRIC NUCl, EAR ENERGY BU$lNf$$ OMRATION$
GENERAL ELECTR!C CCMDANY
- 2311 WEST 99ND STREET, SUtTE 900. CAK BRCCK tLUNoiS 60599.(319)573 3999 January 14, 1987 Mr. Hien Do ISI Engineer Quad Cities Nuclear Station 22710 206th Avenue North Cordova. IL 61242
SUBJECT:
Examination of Wald No. 02A-510 Weld Overlay utilizing GE's SMART UT during Quad Cities ISI. Unit 2
Dear Mr. Do,
This letter is written to clarify the manual and automatic examinations in response to the concerns expressed by yourself and Mark Horbaczewski regarding the Pipe to End-Cap weld overlay.
General Electric personnel examined the overlay 02A-S10 utilizing the GE SMART UT image system with RTD 60' Refracted Longitudinal and 70* Refracted longitudinal wave search units.
The examination resulted in the recording of 42 non-geometric indications with 4 being circumferential and 38 being axial orientated indications.
indications were first recordgd with a 60'RL, These and a complimentary examination with a 70 RL followed.
Analysis of this data shows 4 circumferential cracks, 27 axial cracks and 2 contamination cracks.
Remaining ligaments (amount of material from crack tip to QD surface) ranged from.350" to.680".
l Remaining ligament was determined using 3db drop method for the l
60'RL and maximum signal amplitude method for the 70'RL.
This l
is in accordance with the EPRI recommended guidelines for weld I
overlay examination.
In addition to the A-Scan presentation review, the color variations of the plot grid indicating depth and amplitude of the reflected signal are utilized to distinguish between relevent and non-relevent indications.
During scanning and data analysis high noise ;*vels were prevalent, however, these signals were non-rel
'ent and posed no limiting interpretation problems.
The results of the GE SMART UT image system ge ally paralled the manual examination results with some varia-ns in remaining ligament, length and location.
Variations between manual and automatic data can be attributed to differences in transducer charactaristics or the acoustic properties, which may fluctuate with the high temperature of the overlay material.
Further differences may be attributed to
~--
the controlled paramenters that are established with the GE SMART UT image system.
GENERAL $ ELECTRIC January 14, 1987 In concurrance with ALARA ideals, the GE SMART UT System enables the search unit to have an unlimited scan time.
Additional indications were recorded by the SMART UT System due to it's remote capabilities and precise data recording.
Final results of this technique correlate with the
)
extensive manual examinations.
I trust this meets your needs.
If you have any questions, feel free to contact me at Dresden Unit 2 ext. 595 or 815-942-6706.
Sir erely, gwpIV$$4wNhnnwn)
Tom R.
Brinkman NDE Level III Project Supervisor Inspection Services tk/rch cca M.
Horbaczewski (CECO)
K.
Modulan (CECO)
B.
Wilson (CECO)
B.
Dummer (GE)
S.
Flood (SE)
R.
Hooper (GE) l t
l i
l l
.