ML19256F561

Summary Rept of Feedwater Piping Repair Program & Associated Observations & Findings.

ML19256F561
Person / Time
Site:	Millstone
Issue date:	12/31/1979
From:	NORTHEAST UTILITIES
To:
Shared Package
ML19256F558	List: ML19256F557 ML19256F561
References
TAC-11793, NUDOCS 7912190414
Download: ML19256F561 (10)

Text

. .

DOCKET NO. 50-336 ATTACIDiEh"I 1 MILLSTONE NUCLEAR POWER STATION, UNIT NO. 2 A SU> DIARY REPORT OF THE FEEDWATER PIPING REPAIR PROGRAM AND ASSOCIATED OBSERVATIONS AND FINDINGS i-621 084 D% EMBER, 1979 7912190

' K3*fn5fST UTILITIES 3 , ':=l.'rm=m'2=r

. . , ~ . , me.< m "..

> CE f2o..s,'XE,m,'u.'.

,

""Z'.'.xu. . INTRODUCTION Millstone Unit No. 2 commenced plant shutdown on October 31, 1979 to reinspect the linear indications in the feedwater system piping which were detected during the IE Bulletin 79-13 inspections in August 1979. The linear indications were found adjacent to the steam generator feedwater nozzle safe-end to pipe welds (AC-G-1, BC-G-1) and the pipe to elbow welds (AC-G-2, BC-G-2). Per NRC recom-mendation, NNECO initiated the feedwater piping repair program on November 6, 1979. The primary objective of the repair program was to remove and repair all unacceptable code discontinuities detected adjacent to the steam generator feedwater nozzle safe-end to pipe and adjacent pipe to elbow welds. The feedwater piping repair program was completed on November 26, 1979.

PURPOSE The purpose of this submittal is to provide a summary report of the Millstone Unit No. 2 feedwater piping repair program and associated relevant observations.

SUMMARY

A. Repair Program 1.0 The Millstone Unit No. 2 feedwater piping repair program was based on two (2) repair methods; namely:

1.1 Option A: Removal of a portion of the steam generator shield wall; cut / replace feedwater pipe at the nozzle safe-end to pipe welds and repair indications; and, 1621 085

. .

1.2 Option B: Cut and remove first elbow outside shield wall and repair indications from the pipe ID.

On November 6, 1979 both repair methods were instituted in parallel. However, since the activities associated with repair Option B were confirmed to be feasible and proceeded with no major problems, it was decided to pursue the feedwater piping repair based strictly on repair Option B. Therefore, the removal of the steam generator shield wall was not performed, since the linear indications adjacent to the nozzle safe-end to pipe welds (AC-G-1, BC-G-1) were repaired from the pipe ID.

2.0 In order to gain access to the nozzle safe-end to pipe welds, the first elbow outside the shield wall was removed as shown in the attached Figure I. The subsequent procedural steps associated with repair Option B were as follows:

~

2.1 The first elbow outside the shield wall was removed as shown in Figure I. Welds AC-G-2 and BC-G-2 were left intact to facilitate fit-up and installation.

2.2 All welds and adjacent base material containing linear indications (i.e., AC-G-1, BC-G-1, AC-G-2, and BC-G-2) were subsequently examined by PT and/or VT. The existence of linear indications at the counterbore transition taper on both sides of the above welds was confirmed.

2.3 An internal grinding fixture was inserted inside the pipe to remove the linear indications near the nozzle safe-end to pipe welds (AC-G-1, BC-G-1). The excavated areas were 100% LP and RT examined prior to weld repair to ensure complete removal of indications.

1621 086

. .

Since the repair program was directed at the removal of the indications, no valid crack depth measurements were derived.

A qualitative assessment of the crack depths is provided later.

Furthermore, the repair Option B precluded the removal of

• metallurgical samples for evaluation.

2.4 The excavated areas at welds AC-G-1 and BC-G-1 were sub-sequently weld repaired from the pipe ID. The final weld repaired areas were LP and RT examined in accordance with the IE Bulletin 79-13 criteria. Welds AC-G-1 and BC-G-1 were post-weld heat treated to eliminate residual stresses due to repair welding.

2.5 The pipe side of welds AC-G-2 and BC-G-2 were subsequently ground out as required to remove linear indications at the counterbore transition taper. The excavated areas were LP and RT examined prior to weld repair. Weld repair was performed from the pipe ID.

2.6 Final weld repaired areas in welds AC-G-2 and BC-G-2 were LP and RT examined.

2.7 After weld repair, a cut was made through the centerline of welds AC-G-2 and BC-G-2 as required to provide the necessary bevel on the pipe side of these welds.

2.8 Limited metallurgical samples containing the linear indications on the elbow side of weld BC-G-2 were preserved for future evaluation.

2.9 New replacement elbows were used for final fit-up and installation in both feedwater piping loops. The ID surfaces adjacent to all weld repaired welds as well as new welds were ground and buffed to a smooth surf ace finish to eliminate lb2k bb

'

potential stress risers and/or crack initiation sites.

2.10 After fit-up of new replacement elbows, welding was initiated at welds AC-G-2, BC-G-2, AC-G-3 and BC-G-3. The root passes were examined by RT. The final welds were examined by LP and RT per the IE Bulletin 79-13 criteria. Welds AC-G-2, AC-G-3, BC-G-2 and BC-G-3 were PWHT.

2.11 All final acceptance radiographs were performed and evalu-ated in accordance with the IE Bulletin 70-13 criteria.

2.12 A system pressure test was performed in accordance with the applicable ASME Section XI code.

2.13 An authorized nuclear code inspector participated in the total feedwater piping repair program in accordance with the applicable ASME Section XI code.

B. Relevant Observations and Findings The relevant observations and findings derived during the feed-water piping repair program are summarized below.

1.0 The nozzle ID surface around the thermal liner. was visually inspected. No evidence of significant pitting / erosion was observed on the nozzle ID in both steam generators.

2.0 Since the feedwater piping repair was directed at the complete removal and repair of the linear indications, no valid correlation between the actual crack depths and the results of the UT inspection can be validly established. This applies only to the nozzle safe-end to piping welds (AC-G-1 and BC-G-1).

Based on the inspections and measurements taken during the repair cycle of welds AC-G-1 and BC-G-1, the following qualitative assess-ment regarding the approximate size and depth of the observed linear indications can be made.

i621'088

. .

-o.

2.1 Weld AC-G-1 (a) A visual examination confirmed the existence of five (5) rather discernible linear indications running parallel to the weld joint. The linear indications were cracks running at the edge of the counterbore taper on both sides of the weld. While the longest indication was approximately four inches (4") long (12 o' clock position), the remaining indications were approxi-mately one inch (1") long. The findings of the visual examina-tion correlate in general to the RT findings.

(b) Micrometer readings of the final excavated areas were taken prior to weld repair at the north-south and east-west location on both the safe-end and pipe sides, respectively.

Since an initial diametral reading was not taken, the average amount of metal excavated was estimated based on the ID bore dimensions specified in the original welding / installation specification.

For weld AC-G-1, the average depth (d) of metal removed was as follows:

Pipe Side North-South d = 110 mils East-West d = 112 mils Safe-End Side North-South d= 88 mils East-West d= 88 mils 1621 089

.

. .

c6 2.2 Weld BC-G-2 (a) The initial PT examination of the inside surface of weld BC-G-2 revealed that the entire circumference had linear indications on both sides of the weld. Indications were more severe in the 11-2 o' clock and 5-7 o' clock locations looking into the steam generator nozzle. Although this observation correlates to the initial RT findings, it indicates that the linear indications existed also in the bottom half of the pipe.

(b) For weld BC-G-2, the average depth (d) of metal removed was as follows:

Pipe Side North-South d = 111 mils East-West d = 148 mils Safe-End Side North-South d = 204 mils East-West d = 162 mils 2.3 Welds BC-G-2 and AC-G-2 (a) Linear indications were observed on both sides of welds BC-G-2 and AC-G-2. The linear indications were confined to the counterbore transition taper and running circumferentially on an intermittent basis. Evidence of pitting in-line with the indications was noted. Metallurgical samples were preserved from the 3 o' clock and 9 o' clock positions (elbow side) of weld BC-G-2.

(b) Measurements of the crack depth in both samples reveals that the crack depths were on the order of 60-65 mils. This finding correlates to the UT measurements of 40-60 mils established during the August 1979 inspection.

090

~

-

7-

.

CONCLUSIONS

1. The Millstone Unit No. 2 feedwater piping repair program was performed in accordance with repair Option B, All repair activities were accomplished from inside the feedwater pipes.

Therefore, no shield wall removal was necessary at this time.

2. All detected linear indications were removed and weld repaired. Welds AC-G-1, 2, 3 and BC-G-1, 2, 3 were post-weld heat treated. Specific design improvements such as (1) elimination of stress risers and/or smooth transitions for unequal metal thick-nesses, (2) smooth and polished base material on both sides of the nozzle safe-end to pipe welds and the adjacent pipe-to-elbow welds, and (3) post-weld heat treatment of all six joints, were incorporated into the repair program in order to eliminate / minimize the potential for crack initiation.

3. The final acceptance radiographic examinations for welds AC-G-1, 2, 3 and BC-G-1, 2, 3 were performed in accordance with IE Bulletin 79-13.

4. An authorized nuclear code inspector participated in the total feedwater piping repair activities including the system pressure test.

5. The linear indications detected during the August 1979 inspection were confirmed to be cracks. Cracking occurred at the edge of the counterbore transition on both sides of welds AC-G-1, 2 and BC-G-1, 2. The items of interest are as follows.

5.1 The actual crack depths near welds BC-G-1 and AC-G-1 could not be validly established. However, the average depth 1621 .

091

.

.

of metal removed at weld AC-G-1 is on the order of 100 mils, while the average depth of metal removed at weld BC-G-2 ranges from 111-204 mils depending on the orientation.

5.2 The actual crack depth on the elbow side of weld BC-G-2 was confirmed to be 60-65 mils based on the evaluation of a metallurgical specimen.

1621 092

~ '

WELD NUMBERS y S/G 1 S!G 2

[ *

.* Ie*

'

@ FW AC G 21 FW BC G 18

@ FW AC G-1 FW BC G 1

@ FW AC G 2 FW DC G 2 Q

, "

-

@ FW AC G 4 FW BC G 4

-.

3

~3 l

% Ak "ip

_

,] -- . ,

,_

-

.  : .... .

' ' ' ' ~ ~ " '

T

• m

.

.

- -

. ...,

.

... .-

C A

\\ 1N <

~;l

,

i Y i asi

.

c*'

, . .

~ "

. Cur *I

-> s-*- co

-- 29}--.

..'. s.%.'e ?l.

____ _ _ _ ,

. ._

/ -

i i

- =

- -

E 5

I i i ih Cor *2 Y ,'?

'

.

6

...

,.

\ / =- WElb N lil fell {lliff rilli M AC-G-3,8C-G 3:

C: .S

_W

.

"

3 Q -

-

I Nozz FE END Pips To Eta To PIPE WELb (AC-G-i,8C-G-f) WE!.0(AC-G-2,6C 5 }

a

_

MILLSTONE II FEEDWATER

'

LINE CONFIGURATION FIGURE I 1621 .

093