ML17261A220

Revised LER 79-023/01X-1:on 791207,during Inservice Insp, Liquid Penetrant Examination of Pressurizer nozzle-to-safe- End Weld Showed Linear Indications.Caused by Inadequate Delta Ferrite Content & Geometric Discontinuity

ML17261A220
Person / Time
Site:	Ginna
Issue date:	08/05/1980
From:	Curtis A ROCHESTER GAS & ELECTRIC CORP.
To:	NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
Shared Package
ML17250A501	List: ML17250A500 ML17261A220
References
LER-79-023-01X, LER-79-23-1X, NUDOCS 8008080400
Download: ML17261A220 (8)

Text

NRC F OR M 366

( ~), LER 79-023/01X-1 Update Report - U. S. NUCLEAR REGULATORY COIEIIVIISSION Previous Report Date lZ/21/79 LICENSEE EVENT REPORT (CAR 1249)

CONTROL BLOCK 1 (PLEASE PRINT OR TYPE ALL REQUIRED INFORMATION) 1 6 7

~ET 8 9 N Y R EG IQ20 LICENSEE CODE 14 15 0 0 0 0 LICENSE NUMBER 0 0 0 OQ24 25 26 I I I LICENSE TYPE 104~02 30 57 CAT 58 CON'T

[opia 7 8 ooomooE~LOEO 60 61 5 0 0 0 DOCKET NUMBER 2 4 40 69'VENT 68 1 2 0 7 DATE 7 90E

'74 75 8 8 O 5 REPORT DATE 8 O 0

EVENT DESCRIPTION AND PROBABLE CONSEQUENCES Q1P During normal inservice inspection of pressurizer relief nozzle-to-safe-end weld,

~os Itq Rid penetrant exam showed linear indications I-I/8" and 3/8" long . (7. S . 6 . 9 . 2 a(3)) UT did not show any. Followup X-ray exam showed no volumetric indications.

6 Further liquid penetrant and replication showed area of attack 6-7" by 1/8" consisting of a family of fine cracks similar to inter ranular t e attack. All ressurizer head nozzle-to-safe-end weld areas were examined w'ith UT and liquid penetrant with no

~DB other relevant indications noted 7 8 9 80 SYSTEM CAUSE CAUSE COMP. VALVE CODE CODE SUBCODE COMPONENT CODE SUBCODE SUBCODE 7 8

~>O>>

9 10 11 E O'~0>> 12 13 I I p E X X 04 18

~BOTE 19

~ZOTE 20 SEQUENTIAL OCCURRENCE REPORT REVISION 012 LERIRO REPQRT ACTION FUTURE EVENT YEAR

~79 21 22 EFFECT

~

23 SHUTDOWN

~02 24 REPORT NO.

3 26

~W 27

~1 28 ATTACHMENT CODE 29 NPRD.4 TYPE

~X 30

~

PRIME COMP.

31 NO.

~I 32 COMPONENT TAKEN ACTION ON PLANT METHOD HOURS O22 SUBMITTED FORM SUB. SUPPI.IER MANUFACTURER QE~XOTE ~A022 ~AQ>> 0 4 1 6 Y Q22 ~Y024 N QEE W I 2 0 Qto 33 34 35 36 37 40 41 42 43 44 47 CAUSE DESCRIPTION AND CORRECTIVE ACTIONS 27 p Super solidus cracking during original construction due mainly to inadequate delta ferrite content and geometric discontinuity at nozzle weld interface. Investigation included examining boat sample with scanning electron microscope to determine corro-3 s ive species, and insulation leaching chemical analysis ~ Cloth cover replaced with iberglass. Report of metallurgical analysis results was submitted 5 28 80. Pressur-4 izerha n z 7 8 9 CatiOns 80

~

~

FACILITY ~3p METHOD OF STATUS  % POWER OTHER STATUS DISCOVERY DISCOVERY DESCRIPTION Q32

~ls G Q22 ~00 0 Q22 NA ~8031 Routine ISI 7 8 9 10 12 13 44 45 46 80 ACTIVITY CONTENT RELEASED OF RELEASE AMOUNT OF ACTIVITYQ36 LOCATION OF RELEASE Q36 6 ~Z Q33 ~ZQ34 NA NA 7 8 9 10 11 44 45 80 PERSONNEL EXPOSURES NUI'4 8ER TYPE DESCRIPTION Q39

~TT ~00 0 QET ~ZQ32 NA 7 8 9 11 12 13 80 PERSONNEL INJURIES 7 8 9 11 12 80 LOSS OF OR DAMAGE TO FACILITY TYPE DESCRIPTION Q43

~TE Z 042 NA 7 8 9 10 80 PUBLICITY o

ISSUED

~Y044 DESCRIPTION Q News release was made December 11, lg7g.

NRC USE ONLY 8 9 10 68 69 80 o 80080 0 4TIIQOPPREPARER A. E. Curtis III PTTRE.716/546-2700, ext. 2644,

Attachment tc LER 7943/01X-1 Rochester Gas and Electric Corporation R. E. Ginna Nuclear Power Plant, Unit No. 1 Docket No. 50-244 Event Descri tion On Friday, December 7, 1979, during a normal inservice inspection program examination on the pressurizer relief nozzle-to-safe-end weld, linear indications 1-1/8" long and 3/8" long were noted by the liquid pene-trant examination method. The ultrasonic examination performed did not reveal any indications. On Saturday, December 8, 1979, the area of the indications was prepared by mechanical means with a flapper wheel and light grinding. A reexamination with liquid penetrant revealed an area of attack approximately 6-7 inches long by 1/8" wide. Due to other work being performed in the pressurizer head area supplementary examinations were not performed until Sunday night, December 9, 1979. These examina-tions included more liquid penetrant examinations and radiographic (X-ray) examination of -the area where the original liquid penetrant indication was found. The radiographs of the affected area revealed no apparent volumetric indications. Supplemental ultrasonic examinations of the area also revealed no apparent volumetric indications.

Inplace metallography was attempted without much success due to the restrictions imposed by the piping system that attaches to the relief nozzle. Replication using a plastic material revealed a family of very fine cracks. In a 2 inch long area the indications were ground out at about 1/8 inch depth. Confirmation of the finding occurred when in an additional 3 inch long area the indications were ground out after less than 1/8 of an inch of material was removed. On Wednesday', December 12, 1979, a boat sample was removed for further investigation and analysis of the cracks. The results of the investigations made in the field revealed that the cracks were in the 309 weld material that was used for the dissimilar weld between the carbon steel nozzle (SA-216-GR.WCC) and the stainless steel safe-end material (SA-182 TP, 316) and not the safe-end material as originally thought.

The repair procedure included removal of the remaining cracked mate-rial, preparing the excavated area for weld repair and welding with the Gas Tungsten Arc Welding Process utilizing ER-309 base filler metal. Post prep-aration, first weld pass and final weld liquid penetrant examinations were performed. A final weld preservice ultrasonic examination of the repaired area was also performed. The results of all examinations did not reveal any indications .

Also examined during this investigation were all the nozzle-to-safe-end weld areas on the pressurizer head utilizing the liquid penetrant and ultrasonic examination methods. The ultrasonic examinations included a normal inservice inspection technique and a special stress corrosion tech-

Attachment tc LER 79~/01X-1 nique. Welds in associated piping of the four nozzles were also examined utilizing the liquid penetrant method with no further relevant indications noted. There were no health or safety consequences to the public or to plant personnel due to this event.

Cause Descri tion Replication of the as polished weld surface revealed a very tight crack pattern, similar to an intergranular stress corrosion. The boat sample analysis revealed three cracks as follows: /

Crack 41 1. 4 mm from carbon steel fusion line

0. 4 mm in depth Crack 42 2.3 mm from carbon steel fusion line 0.7 mm in depth Crack 43 - 3.5 mm from carbon steel fusion line 1.1 mm in depth touching carbon steel bevel Further analysis utilizing the scanning electron microscope was later performed to determine if any corrosive species could be identified.

Initially only speculation could be made on the cause of this crack-ing. It was believed that it was possibly caused by original fabrication contamination or by a chloride concentration that has occurred during oper-ation. As part of this investigation a chemical analysis of leachable con-taminants from the thermal insulation that surrounded the relief nozzle was performed as prescribed in Regulatory Guide 1.36, Nonmetallic Thermal Insulation for Austenitic Stainless Steel. The results of this analysis are as follows for sample 1 - woven cloth cover and sample 2 - glass fibrous layered material:

Sample PPM CI PPM F 'I+F PPM PPM NA PPM Si03 PPM Na + Si03 98 9.4 107.4 100 207 307 30 1.6 31.6 40 99 139 Based on this analysis according to Figure I of Regulatory Guide 1.36, Sample 2 the glass insulation is acceptable, however Sample 1 the cloth cover is not within the acceptable region. Although it would be logical to assume that the insulating cloth provided the chlorides that might have caused this cracking phenomena, the cloth could have been contaminated during the handling of removal. Therefore, it was necessary that the results of the scanning electron microscope be further analyzed before confirming the insu-lation involvement.

Attachment tc LER 79+3/01X-1 As part of the corrective action the insulation'material was changed on the pressurizer nozzles to assure that potential leachable contaminants from insulating material will require no further consideration. Also each nozzle-to-safe-end area was cleaned to remove any contaminants prior to reinsulation. A full report on the final results of the metallurgical analysis was submitted when the investigations were completed. The May 29, 1980 submittal entitled "Metallurgical Investigation of Cracks in the Pressurizer Nozzle-to-Safe-End Weld of the Robert Emmett Ginna Nuclear Power Gener-ating Station," WCAP 9663, concluded that the cracks were caused by super solidus cracking or "hot cracking" during original fabrication as the molten weld metal cooled to the solid state. Inadequate delta ferrite con-tent of the stainless steel weld metal local to the weld interface and a geometric discontinuity at the nozzle carbon steel/weld metal interface were the main contributors to the cracking process. This cracking is exclu-sive to the particular local area on the weld and is not considered generic.

I These nozzle-to-safe-end weld areas were reexamined during the 1980 refueling and maintenance outage, and found to have no indications.

L