ML19225C263: Difference between revisions

From kanterella
Jump to navigation Jump to search
(Created page by program invented by StriderTol)
(Created page by program invented by StriderTol)
Line 21: Line 21:
: 1) the same Graver welding procedurea      (i.e., 313N and 319N) that were used on the other S&W project were also used on River Bend Unit 1, 2) the welds were performed on both projects in the same time frame, and 3) the questionable welds were confitud to those produced using the semi automatic weld procedure (313N).
: 1) the same Graver welding procedurea      (i.e., 313N and 319N) that were used on the other S&W project were also used on River Bend Unit 1, 2) the welds were performed on both projects in the same time frame, and 3) the questionable welds were confitud to those produced using the semi automatic weld procedure (313N).
In March, 1979, Gulf States determined that the potential for questionable cadsleeve welds appeared to be high and consequently notified the NRC I6E b gion IV office by telephone and followed with a " Potential Reportable Deficiency -
In March, 1979, Gulf States determined that the potential for questionable cadsleeve welds appeared to be high and consequently notified the NRC I6E b gion IV office by telephone and followed with a " Potential Reportable Deficiency -
lOCFR 50.55(e)" letter dated March 20, 1979.
10CFR 50.55(e)" letter dated March 20, 1979.
Ovei the next sixty days,          3tates had further analysis performed on the cadsleeve welds, studied the potential for safety implications, and began to review possible corrective actions in case such action became necessary. At the end of this period, Gulf States concluded that a Reportable Deficiency under 10CFR 50.55(e) did exist and not ified the NRC ISE Region IV office in a letter dated May 31, 1979.      The following is a sunmary report of the cadsleeve weld problem and the correcti.ve action to be taken.
Ovei the next sixty days,          3tates had further analysis performed on the cadsleeve welds, studied the potential for safety implications, and began to review possible corrective actions in case such action became necessary. At the end of this period, Gulf States concluded that a Reportable Deficiency under 10CFR 50.55(e) did exist and not ified the NRC ISE Region IV office in a letter dated May 31, 1979.      The following is a sunmary report of the cadsleeve weld problem and the correcti.ve action to be taken.
Description of the Affe"ted Conponents The cadweld splices are located at junctions of the drywell, reactor pedestal, weirwall, and containment vessel at the top of the reactor buildii.g mat.      They form load transfer interfaces for these structures.        The cadweld sleeve types tor 54      n1                      2002200 g g
Description of the Affe"ted Conponents The cadweld splices are located at junctions of the drywell, reactor pedestal, weirwall, and containment vessel at the top of the reactor buildii.g mat.      They form load transfer interfaces for these structures.        The cadweld sleeve types tor 54      n1                      2002200 g g

Revision as of 02:40, 6 November 2019

Deficiency Re Cadsleeve Welds Mfg by Graver Energy Sys & Produced by Semiautomatic Welding Procedure.All Defective Cadsleeves Will Be Replaced Using Manual Welding Process W/Increased Inprocess Controls
ML19225C263
Person / Time
Site: River Bend Entergy icon.png
Issue date: 06/27/1979
From:
GULF STATES UTILITIES CO.
To:
Shared Package
ML19225C260 List:
References
NUDOCS 7907260768
Download: ML19225C263 (5)


Text

.

Discoverv of the Pcoblen It ca n brought to our attention in late 1978 by our Architect-En;;ineer (Stone & !!chster) that on another of itr projects a difficulty had been encountered with the cadsleeve ' elds fabricated by Graver Energy Systems (G rave r) . Gulf States initiated an investigation 0: the potential for a simi-lar probler .n the cadsleeve welds also fabricated by Graver for the River Bend Unit 1. This investigation was initiated based on the followin;; considerations:

1) the same Graver welding procedurea (i.e., 313N and 319N) that were used on the other S&W project were also used on River Bend Unit 1, 2) the welds were performed on both projects in the same time frame, and 3) the questionable welds were confitud to those produced using the semi automatic weld procedure (313N).

In March, 1979, Gulf States determined that the potential for questionable cadsleeve welds appeared to be high and consequently notified the NRC I6E b gion IV office by telephone and followed with a " Potential Reportable Deficiency -

10CFR 50.55(e)" letter dated March 20, 1979.

Ovei the next sixty days, 3tates had further analysis performed on the cadsleeve welds, studied the potential for safety implications, and began to review possible corrective actions in case such action became necessary. At the end of this period, Gulf States concluded that a Reportable Deficiency under 10CFR 50.55(e) did exist and not ified the NRC ISE Region IV office in a letter dated May 31, 1979. The following is a sunmary report of the cadsleeve weld problem and the correcti.ve action to be taken.

Description of the Affe"ted Conponents The cadweld splices are located at junctions of the drywell, reactor pedestal, weirwall, and containment vessel at the top of the reactor buildii.g mat. They form load transfer interfaces for these structures. The cadweld sleeve types tor 54 n1 2002200 g g

_2 each structe = and the number of each type is listed in Table 1.

E 1d Design Requirements The specification unJer which these components were prepared requires welding procedures and qualifications in accordance with AS'1E III subsection NE, with quality of workmanship to be "first class." Due to the partial penetration weld geometry of the component the nondestructive testing procedure specified is magnetic particle as specified in the ASME III code.

Inves_tiyation of the Welds Upon beln;; advised of the potential for internal welding inclusions, due to inadequate adherence to the vendor welding procedures using the semi-automatic welding process, S&W was directed to devise a meanc of nondestructive qualitative examination of the suspect sie l d s . A special ultrasonic testing procedure was developed to examine the weld area for identification of possible lack of fusion and slag inclusion. The proccdure developed has application only as a quali-tati"o investigative prc :edure, and is not applicable for weld acceptance due to limited weld coverage. A sample group of welds was examined for gross qualitative discont) :uities. Based on the results of this examination, which showed suffi-cient lac' of fusion and slag inclusion to demonstrate noncompliance with the specified first class weld provision of iS'1E III subsection NE, it has been decided to reject all of the cadsleeve welds Safety Concerns Due to Potentially Def ec tive l' elds The existence of gross amounts of discontiauities in the J-bevel area of the veld may lead to unacceptable st resses anc'. deformations dur trq; the operation of the plant under abnormal conditions and endanger the structural integrity of the building.

1 f* A h34 1 ) ')

L L L.

.

Corrective Action It has been decided to reject all of the cadsleeves and repince them using a progran as follows-4

..nce all defective welds were produced using the :;emi-automa tic welding procedure, the semi-automatic welding, procedure will not be used in the rework.

A ranual welding proces9 will be used and will be augmented with increased inprocess controls and hold points under the fabricator's (Graver) QA program, as well as additional surveillance by the Architect-Engineer (S&W). In addition to the specified inspection of all final weld surface by MT, an additional MT inspection will be performed midway through the weld on a random sample.

These MT inspections will be performed using the DC Prod method which will provide some depth of inspection of the weld surface both at the '.it and final weld surface. UT eyimination of the partial penetration cadsleeve weldr will not be utilized due to its lack of applicability and established procedures under the ASME III code. UT exanination of this weld design has an extremely limited ability to adequately "see" or evaluate the entire weld volume. The faulted condition in this problem was the semi-automatic weld proces, misapplication.

The use of .Inual welding and augmented Quality Assurance resolves this problem.

/l c a n 7J4 /23

.

.

TABLE 1 CAD'AELD SLEEVE TYPES FOR RIVI:R PEND STATION - UNIT 1

  1. 18 #14 #18L #18 Structure RBB-18101-JA* RBB-14101-JA* k B'.. -5 3 7-JA 1 RBW-538-J* Total
1. Reactor Pedestal 642 214 214 -

1070

2. Weirwall 645 430 215 -

1290

3. Drywell 1892 - -

172 2064

4. Containment vessel 1290 - - -

1290

5. Spar;;er Ring < 224 - - -

224

6. Weirwall Notch Area 24 4 28 Grand Total for Unit 1: 5964 NOTE:
1. The naterial of the cadweld sleeves i t, ASTM A519-1026.
2. All the above sleeves were welded to the embedment plates using a semiautomatic welding process.
3. The weld surfaces were inspected by magnetic particle examination.
  • Erico educts, Inc. catalog number 454 224

. . . .

.

[l OF SLEEVE SLEEVE

,7

'7- CFOR ADWELD "I,8 B AF(

l, J\- 5 k ,-

5" \/ \

s "" 5 s- /NN/ N,

/*

f f' 7-- SU RF ACE "B"

/N N

/ ,

POSSI BLE

' LACK OF, 7 , FUSIOkl i

^ ^

1.ACK OF PoluT A-Root FU$l0h]

\^

v R:%,,

i r

e -

e

. .- .

454 225

'

FIG.l T Y P. WE LD B ET WEEK 1 C ADWF '_D SLEEVE j PLATE