ML20116G478
| ML20116G478 | |
| Person / Time | |
|---|---|
| Site: | FitzPatrick  |
| Issue date: | 04/22/1985 |
| From: | POWER AUTHORITY OF THE STATE OF NEW YORK (NEW YORK |
| To: | |
| Shared Package | |
| ML20116G461 | List: |
| References | |
| NUDOCS 8505010509 | |
| Download: ML20116G478 (7) | |
Text
~
L Attachment No. 1 to JPN-85-35 o
Preliminary Summary of Interarant lar Stress,Ltrrosion Inducted Crac t Inspections Durina Marcir - April 1985 Outace James A. FitzPatrick Nuclear Power Plant Docket No. 50-333 NEN YORK POWER AUTHORITY April 22, 1985 hh33 A
K PDR b
9 r
1.0 Introduction Inspections to detect and evaluate Intergranular Stress Corrosion Cracking (IGSCC) were performed at the Authority's FitzPatrick plant in accordance with NRC Generic Letter No.
84-11 (Reference 1) during the recent March / April 1985 refueling outage.
Generic Letter No. 84-11 required that the Authority re-examine unrepaired cracked welds.
Accordingly, six welds in the Reactor Recirculation System were examined to evaluate previously reported IGSCC indications.
Table No. 1 details the results of these re-inspections and of previous inspections conducted in October of 1984.
2.0 Test Personnel The same inspection agencies were used during both the October 1984 and March /A'pril 1985 outages; (Ebasco Services, Inc. and Kraftwerk Union as Universal Testing Laboratories (KWU/UTL).
Many of the same inspectors used during the October 1984 outage also performed evaluations during the March / April 1985 outage.
The same Level III personnel used during the October 1984 inspections to perform final flaw sizing were also used during these March / April 1985 inspections in the same capacity.
Most of the Level II personnel employed during these inspections were also employed during the October 1984 inspection.
This helped to minimize any potential differences which may have resulted from variations in experience, training, etc.
3.0 Testina Techniques Between the October 1984 and April 1985, significant improvements have been made in the field of ultrasonic testing (UT) techniques.
During this period, improvements in methods
'for ultrasonically sizing IGSCC led to the development of a sizing qualification and certification program through EPRI and J. A. Jones.
These improvements can be seen in the number of different UT techniques used in IGSCC evaluation during the March / April 1985 outage.
In October 1984 Ebasco technicians used only 45' and 60* shear wave techniques to detect IGSCC and estimate circumferential length.
(Depth sizing was the respon-sibility of KWU/UTL inspectors).
During the March / April 1985 outage, Ebasco used the 45' shear wave method only for detection.
60* and 52* shear wave, 30-70-70 mode conversion (or ID creeper), 50' refracted longitudinal wave, and high angle longitudinal wave (or OD creeper) were all used, as necessary, to evaluate and size IGSCC indications.
During the October 1984 outage, KWU/UTL inspectors evaluated suspect IGSCC indications using 45* shear wave and 30-70-70 mode conversion techniques.
Later, KWU/UTL inspectors used 45*
and 60* shear wave techniques for detection.
These techniques were employed in both their P-Scan and manual inspections of 100% of each weld area.
To evaluate and size IGSCC indications, the KWU/UTL technicians primarily used the:
30-70-70 mode conversion (ID creeper) technique.
The newly developed SUS-I and SLIC-40 probes were used to obtain supplemental data, f
In cddition to tho techniques doccrib d absva, both Ebssco end KWU/UTL inOp ctors ovaluated ficw circunforential length down to 10% of the highest amplitude signal or to background noise level, when using amplitude techniques.
During the October 1984 inspections, flow and points were measured at the point of 50%
reduction of the highest amplitude signal.
This combination of the lower recording level for flaw and points, and the use of a number of recently developed techniques for sizing, has yielded improved results.
A similar approach was used to evaluate other welds inspected this outage as required by Generic Letter No. 84-11.
4.0 ConDarison With Previous InsDection Results These re-inspections verified the majority of the previous evaluations.
Table 1 and the accompanying notes compare these inspection effects.
A notable difference can be seen in the evaluation-of Recirculation System Wald No.28-112.
The October 1984 inspection showed a circumferential crack O.6" long and 17%
through wall in this 28" diameter, 1.248" thick piping.
The March / April 1985 examination showed that the crack was 4.72" long, while still 17% through wall.
This indieption size was independently arrived at by both Ebasco and KWU/UTL inspectors.
This difference is attributed to the use of improved inspection methodology, not crack growth.
Assuming that the IHSI performed on this joint had gg beneficial effects, the theoretical and empirical evidence currently available indicates that circumferential crack growth of four inches in less than six months of normal operation is extremely improbable.
Growth rates of this magnitude have never been reported.
Further, circumferential growth at this rate, with no growth through the pipe wall is even more unlikely.
To confirm this hypothesis, a re-examination was performed by Ebasco inspectors of this weld using only the techniques, calibrations, and flaw end point recording levels used during the October 1984 outage.
This re-examination was witnessed by Mr. R. McBrearty, a USNRC Region I I&E inspector.
This re-examination yielded a circumferential crack length of 0.6".
Based on these results, and the absence of any evidence of crack growth behavior which could account for such rapid, circumferential growth only, this apparent increase in length is solely attributable to improvements in examination sensitivity.
The other less significant difference in crack length is also attributable to improved inspection techniques.
In instances where re-evaluation during the March / April 1985 inspection indicates that previous indications were the result of geometry or signal levels were below recording level, the Authority will conservatively assume that the previous :
inspection results remain the basis for continued operation and future inspection. e
~
5.0 _ References 1.
USNRC Generic Letter No. 84-11, " Inspections of BWR Stainless Steel Piping",, dated April 19, 1984.
2.
NYPA letter, C. A. McNeill, Jr. to D. B. Vassallo, dated October 21, 1984 (JAFP-84-0979) regarding Recirculation System Weld Joints.
Includes: fracture mechanics evaluation of ultrasonic indications for five welds; summary of recirculation system weld inspections performed to date; weld overlay reports; weld overlay thickness; ultrasonic test data and polar plots for riser and safe-end welds; and, simplified recirculation system isometeric drawings.
3.
NYPA letter, C. A. McNeill, Jr. to D. B. Vassallo, dated October 25, 1984 (JAFP-84-1001) regarding Recirculation System Welds - September 1984 Outage.
4.
NYPA letter, J. P. Bayne to D. B. Vassallo, dated November 7, 1984 (JPN-84-73) regarding Recirculation System Welds.
Responds to USNRC staff questions from October 31,.
1984 telephone conversation.
5.
NYPA letter, C. A. McNeill, Jr. to D. B. Vassallo, dated March 12, 1985 (JAPP-84-0283) regarding IGSCC Inspection l
Results.
i 6.
NYPA letter, J. P. Bayne to D. B. Vassallo, dated March 9, 1985 (JPN-84-16) regarding Recirculation Piping Flaw Indication.
7.
NYPA letter, C. A. McNeill, Jr. to D. B. Vassallo, dated l
January 15, 1985 (JPN-84-02) regarding Inspections for IGSCC during February 1985 Refueling Outage.
I l -
p
N Table No. 1 JAMES A. FITZPATRICK NUCLEAR POWER PLANT SUISEARY OF RE-INSPECTION OF UNREPAIRED IG8CC MARCH / APRIL - 1985 MARCH / APRIL 1985 OCTOBER 1984 INSPECTION INDICATIONS INSPECTION WELD #
LOOP WELD LOCATION LENGTH DEPTH YES/NO TYPE LENGTH DEPTH REMARKS 5
12-4 A
Pipe to Sweep-0-Let 1.18" 7.5%
Yes-Cire.
2.5" 7,5%
Note 1, 7 i
i 12-17 A
Pipe to Safe-End 2.9" 10%
Yes Cire.
Not Recordable Note 2, 8 I
28-48 A
Pipe to Safe-End 1.1" 15%
No NA Note 9 Note 3 28-53 A
Elbow to Valve 0.4" 5%
Yes Cire.
Not Recordable Note 4, 8 l@
28-112 B
Elbow to Valva 0.6" 17%
Yes Circ.
4.72" 17%
Note 5, 7 i
28-113 B
Valve to Pipe 0.5" 10%
No NA Note 9 Note 6 l
I f
i NA = Not Applicable cire. = circumferential Not Recordable = Signal observed but below recording level.
, 1
i Notes for Table No. 1 Note 1: Length reported by letter dated 10/25/84 (JAFP 84-1001, Reference 3),
was 0.4".
However, ultrasonic test data, also submitted to the NRC, indicated a length of 1.18".
The analysis performed was based on a crack length of 2.0".
The difference in measured crack length is due to a more sensitive technique for sizing crack length than that used in October 1984. Crack depth, as measured by the technique used in October 1984, has not changed. The lack of crack depth growth i
further supports the view that no real crack growth has occurred.
This crack will be analyzed using a length of 2.5" and depth of 7.5%
'through wall, to determine feasibility of continued plant operation without repair.
Note 2: Indications reported in October 1984 were observed by KWU/UTL inspectors. However, these indications were below the recording threshold and showed no measurable through-wall dimension. Continued operation will be based on the analysis of this indication with the dimensions as reported in October 1984 (References 2, 3 and 4).
-Note 3: The indications reported in March of 1984 have since been determined to be misinterpretations attributed to defraction of the UT signal by the dendritic. structure of the weld. However, the analysis performed based on the March 1984 inspection results, provides justification for continued operation without repair. The Authority will adopt the conservative approach and use the March 1984 inspection results for this weld as the basis of future inspections.
Note 4: The indica' tion discovered in October 1984, was again detected by KWU/UTL. However, the signal was 2db below recording threshold. No apparent growth of this indication has been recorded, confirming the post-IHSI crack growth analysis, which predicted no growth. The results of the more conservative as-welded stress analysis indicates that this crack will not significantly reduce the piping structural design margin within the next 18 months. Therefore, this analysis remains valid, since no increase in crack length or depth was identified.
Note 5: The change in reported crack length is attributed solely to the use of more sensitive sizing techniques. Re-examination, in the presence of USNRC I&E inspector Robert McBrearty, using a technique similar to the one used in October 1984, duplicated results obtained in the October 1984 examination. A depth of 17% of wall was reported in both instances, indicating little or no crack growth. A new analysis, based on the reported length of 4.72" and a depth of 17%,
will be performed to determine feasibility of continued operation without repair.
1 \\
4 Note 6: The indication reported in October 1984 was evaluated as an ID surface indication with no significant depth. Due to the improved sensitivity available with the new UT techn'iques, this evaluation remains accurate. However, the Authority will adopt the conservative approach and assume that the indication reported in October 1984 still exists.
(with lengths and depths as previously measured) as a basis for future inspections.
i Note 7: This year, recording levels for sizing were lowered to 10% of the highest indication amplitude or to noise level; whichever was' lowest.
(In 1984, 50% to 50% (e.g., 6 db drop method) maximum amplitude was utilized.)
Note 8: Recording levels during detection are 20% of DAC at reference level; or any indications that appear to be suspicious in nature regardless of amplitude.
Note 9: Signal analyzed as a geometric reflection.
i.
.