ML20150C743
| ML20150C743 | |
| Person / Time | |
|---|---|
| Site: | Mcguire |
| Issue date: | 06/28/1988 |
| From: | Hood D Office of Nuclear Reactor Regulation |
| To: | Tucker H DUKE POWER CO. |
| Shared Package | |
| ML20150C747 | List: |
| References | |
| IEB-88-001, IEB-88-1, TAC-65955, TAC-65956, NUDOCS 8807120609 | |
| Download: ML20150C743 (5) | |
Text
_ _.
June 28, 1988 Docket Nos.: 50-369 and 50-370 Mr. H. B. Tucker, Vice President Nuclear Production Department
. Duke Power Company 422 South Church Street Charlotte, North Carolina 28242
Dear fir. Tucker:
SUBJECT:
TECHNICAL' REPORT REGARDING MCGUIRE 2 REACTOR TRIP BREAKER FAILURE AND METALL0 GRAPHIC ANALYSES OF WELDS ON THE CATAWBA REACTOR TRIP BREAKER POLE SHAFT (TACS 65955/65956)
Enclosed for your information is a final technical report by our contractor, Franklin Research Center (FRC), entitled "Investigation of Trip Breaker Failure at McGuire Unit 2."
The report explains the mechanical binding which kept the McGuire 2B reactor trip breaker (RTB) from opening July 2,1987, discusses Westinghouse's inspection recommendations of September 11, 1987 for Westinghouse Owners Group members, and-identifies issues and concerns other than those associated with pole shaft welds.
It also includes certain recommendations made prior to the December 1, 1987 revision to the Westinghouse technical bulletin and considered by the NRC prior to issuance of NRC Bulletin 88-01.
Appendix A of the enclosure presents the results of the metallurgical investigation of the Catawba RTB pole shaft that you provided the NRC.
Results of this investigation aided in our understanding of the limitations of in-situ, visual inspections of pole shaft welds to detect cracks and was instrumental in the decision reflected in the revised Westinghouse technical bulletin and NRC Bulletin 88-01 that welds not possessing the intended length and leg dimensions should receive continuing periodic inspections.
If you have questions, contact me at (301) 492-1442.
Your cooperation in this matter is appreciated.
Sincerely, Original signed by:
$$071$0[K$
0 Darl Hood, Project Manager s
Project Directorate II-3 Division of Reactor Projects I/II
Enclosure:
As stated cc:
See next page DISTRIBUTION:
,(DocketjFile--}
NRC PDR Local PDR PDII-3 Reading
^
S. Varga 14-E-4 G. Lainas 14-H-3 D. Matthews M. Rood D. Hood E. Jordan liNBB-3302 B. Grimes 9-A-2 ACRS (10) H-1016 DSR MRo&od)J-3 LA: Q PM:PDTI-3 11-3 DHood:pw D
ews 6/17/88 g/p/88 (y/%/88
(
r a.; y June 28, 1988 Docket Nos.: 50-369 and 50-370 Mr. H. B. Tucker, Vice President Nuclear Production Department
-Duke Power Company 422-South Church Street Charlotte, North Carolina 28242
Dear fir. Tucker:
SUBJECT:
TECHNICAL REPORT REGARDING MCGUIRE 2 REACTOR TRIP BREAKER FAILUP.E AND METALL0 GRAPHIC ANALYSES OF WELDS ON THE CATAWBA REACTOR TRIP BREAKER POLE SHAFT (TACS 65955/65956)
Enclosed for your information is a final technical report by our contractor, Franklin Research Center (FRC), entitled "Investigation of Trip Breaker kept the McGuire 2B reactor trip breaker (plains the mechanical binding whic Failure at McGuire Unit 2."
The report ex R18) from opening July 2,1987, discusses Westinghouse's inspection reconsnendations of September 11, 1987 for Westinghouse Owners Group members, and identifies issues and concerns other than those associated with pole shaft welds.
It also includes certain reconsnendations made prior to the December 1, 1987 revision to the Westinghouse technical bulletin and considered by the NRC prior to issuance of NRC Bulletin 88-01.
Appendix A of the enclosure presents the results of the metallurgical investigation of the Catawba RTB pole Shaft that you provided the NRC.
Results of this investigation aided in our understanding of the limitations of in-situ, visual inspections of pole shaft welds to detect cracks and was instrumental in the decision reflected in the revised Westinghouse technical bulletin and NRC Bulletin 88-01 that welds not possessing the intended length and leg dimensions should receive continuing periodic inspections.
If you have questions, contact me at (301) 492-1442. Your cooperation in this matter is appreciated.
Sincerely, Original signed by:
Darl Hood, Project Manager Project Directorate 11-3 Division of Reactor Projects I/II
Enclosure:
As stated CC:
See next page DISTRIBUTION:
Docket File NRC PDR Local PDR PDII-3 Reading S. Yarga 14-E-4 G. Lainas 14-H-3 D. Matthews M. Rood D. Hood E. Jordan MNBB-3302 8. Grimes 9-A-2 ACRS (10) H-1016 LA:
f-3 PM:PD I-3 P
II-3 MRoS DHood:pw D!brtthews
(,/,q/88 h g /88 6 a 7/88 2
~
p p ***%9'o UNITED STATES d
' g, NUCLEAR REGULATORY COMMISSION i
s g
E WASH MG TO N, D. C. 20555
,o June 28, 1988 l
Docket Nos.: 50-369 and 50-370 Mr. H. B. Tucker, Vice President Nuclear Production Department Duke Power Company 422 South Church Street Charlotte, North Carolina 28242
Dear Mr. Tucker:
SUBJECT:
TECHNICAL REPORT REGARDING MCGUIRE 2 REACTOR TRIP BREAKER FAILURE AND METALL0 GRAPHIC ANALYSES OF WELDS ON THE CATAWBA REACTOR TRIP BREAKER POLE SHAFT (TACS 65955/65956)
Enclosed for your information is a final technical report by our contractor, Franklin Research Center (FRC), entitled "Investigation of Trip Breaker Failure at McGuire Unit 2."
The report explains the mechanical binding which kept the McGuire 2B reactor trip breaker (RTB) from opening July 2, 1987, discusses Westinghouse's inspection recommendations of September 11, 1987 for Westinghouse Owners Group members, and identifies issues and concerns other than those associated with pole shaft welds.
It also includes certain recommendations made prior to the December 1, 1987 revision to the Westinghouse technical bulletin and considered by the NRC prior to issuance of NRC Bulletin 88-01.
Appendix A of the enclosure presents the results of the metallurgical investigation of the Catawba RTB pole shaf t that you provided the NRC.
Results of this investigation aided in our understanding of the limitations of in-situ, visual inspections of pole shaft welds to detect cracks and was instrumental in the decision reflected in the revised Westinghouse technical bulletin and NRC Bulletin 88-01 that welds not possessing the intended length and leg dimensions should receive continuing periodic inspections.
If you have questions, contact me at (301) 492-1442.
Your cooperation in this matter is appreciated.
Sincerely, l
Oh l 4
l Darl Hood, Project Manager Project Directorate Il-3 Division of Reactor Projects I/II
Enclosure:
As stated cc:
l See next page
~
.y Mr. H. B. Tucker Duke Power Company McGuire Nuclear Station cc:
Mr. A.V. Carr, Esq.
Dr. John M. Barry Duke Power Company Department of Environmental Health P. O. Box 33189 Mecklenburg County 422 South Church Street 1200 Blythe Boulevard Charlotte, North Carolina 28242 Charlotte, North Carolina 28203 County lianager of Mecklenburg County Mr. Dayne H. Brown, Chief 720 East Fourth Street Radiation Protection Branch Charlotte, North Carolina 28202 Division of Facility Services Department of Human Resources 701 Barbour Drive Mr. Robert Gill Raleigh, North Carolina 27603-2008 Duke Power Company Nuclear Production Department
- 6. McIntyre P. O. Box 33189 Westinghouse Electric Corporation Charlotte, North Carolina 28242 Nuclear Technology Division Box 355 J. Michael McGarry, III, Esq.
Pittsburgh Pennsylvania 15230 Bishop, Liberman, Cook, Purcell and Reynolds 1200 Seventeenth Street, N.W.
Washington, D. C.
20036 Senior Resident Inspector c/o U.S. Nuclear Regulatory Connission Route 4, Box 529 Hunterville, North Carolina 28078 Regional Administrator, Region II U.S. Nuclear Regulatory Connission, 101 Marietta Street, N.W., Suite 2900 Atlanta, Georgia 30323 S. S. Kilborn Area Manager, Mid-South Area ESSD Projects Westinghouse Electric Corporation MNC West Tower - Bay 239 P. O. Box 355 Pittsburgh, Pennsylvania 15230
8 l
l l
FRANKLIN RESEARCH CENTER DIVISION OF ARVIN/CALSPAN TECHNICAL REPORT i
I l
I va LEY FCRGE CORPOAATE CENTE A 2600 MONaOE SOULEvaRD NCan'STOWN PA 19403 TEL 215+666 W O c
.o F-6177-5 Investigation of Trip Breaker Failure at McGuire Unit 2 9
NRC Docket No. 50-370 FRC Project No. 6177-005 NRC Contract NRC-05-86-168, Task EL-305 FRC Report F-6177-5 FRC Group Leader:
NRC Group Leader:
H. M. Fishman D. S. Hood Submitted to U.S. Nuclear Regulatory Comission Division of Reactor Projects Project Directorate II-3 Washington, DC 20555 June 15, 1988 Prepared by:
Reviewed b -
Approved by:
fY b G JTem Princ%al[Authfr Section Hsad' epartment ead)
Gary J. Toman Howard M. Fishman Salvatore
. Carfagno Date: 6-/f"-??
Date: /f M '88 Date:
4 - / S~- F f' i
t PRANKUN RESEARCH CENTER ommon w umn/casma f
vutav sopot coarourt etNita :600 uoNaot sovi.tv Aa0 gggs00 N
~eaa,,, _,..,,.c, l
4-o F-6177-5 FORDiARD This report was prepared by Franklin Research Center (FRC), division of Arvin/Calspan, under a contract (FRC-05-86-168, Task Order EL-305), with the U.S. Nuclear Regulatory Commission (Office of Nuclear Reactor Regulation) for independent assessment and analysis.
The principal author of this report is the former Task Leader, Mr. Gary J. Toman.
Dr. Laurence Leonard contributed the metallurgical analysis reported in Appendix A.
The report was reviewed and revised by Mr. Howard M.
Fishman, tne current Task Leader.
During the performance of the technical assistance task concerning the failure of a WestinghouJe reactor trip breaker at the McGuire Unit 2 Plant (1), FRC provided the NRC with data and recommendations that were incorporated in an NRC Information Notice No. 87-35, Supplement 1 (2) and an NRC Bulletin No. 88-01 (3].
iii l
[
.o \\
e F-6177-5 CCNTENTS
-Section
- Title, Page 1
INTRODUCTION 1
2
SUMMARY
OF FAILURE EVALUATION AT NSID 1
3 DISCUSSION OF FAILURE MODE.
4 4
DISCUSSION OF WESTINGHOUSE RECOMMENDATIONS FOR EVALUATION OF WELD DEFICIENCIES 10 5
FRC COMMENTS ON WESTINGHOUSE RECOMMENDATIONS 12 6
OTHER ISSUES AND CONCERNS 14 7
CONCLUSIONS 16 8
RECOMMENDATIONS.
17 9
REFERENCES.
APPENDIX A - ANALYSIS OF WELD JOINTS ON A REACTOR TRIP BREAKER POLE SHAFT I
I iv
{
l
- r-F-6177-5 FIGURES Figure Title Page 1
Linkages of DS-416 Breaker Mechanism Shown with CB Open and Springs Charged 16 2
Position of Mechanism with RTB Closed.
17 3
Roller Wedged Between Left Cam Segment and Right Side Frame Plate.
18 4
Binding of Roller.
19 V
= -.
4
-t'.
- 1. INTRCDUCTION On July 2, 1987, the 2B reactor trip breaker (RTB) of the McGuire Unit 2 plant failed to open on electrical command.
It had jammed mechanically. On July 3, 1987,RDuke Power Company personnel observed a second jamming of the RTB during investigation of the problem: however, the jamming of the RTB was overcome before the cause'of the jamming was determined. On July 7 through 9, 1987, further attempts were made to have the jammed condition recur in the presence of the NRC Augmented Inspection Team.
However, the circuit breaker did not jam during multiple operations even with purposeful attempts to put the RTB mechanism in an unfavorable position with regard to tripping. NRC Information Notice No. 87-3'S (1) reported this potentially significant safety problem. Thereupon, it was determined that the RTB should be shipped to the Nuclear Services Integration Division (NSID) of Westinghouse for further evaluation.
During the evaluation at the McGuire plant, it was determined that the weld between the center-pole lever and the pole shaft of the RTB had completely separated.
It was believed that this weld failure was related to the jamming of the RTB, but the exact relationship was not known.
This report describes the failure evaluation at NSID: discusses the recommendations (4) made by Westinghouse on September 11, 1987 to Westinghouse Owners' Group members with DS-type circuit breakers; critiques Westinghouse's presentation to the NRC on September 23, 1987; and provides recommendations for further actions regarding DS-type circuit breakers.
In addition, an independent analysis,by PRC of the weld joints is presented.
t
-l l
l _ _ _ _.. - - _ _
~
_ _. _. _ _,.., -..,. _ _. ~, _,. _ _
o F-6177-5 2.
SUMMARY
OF FAILURE EVALUATION AT NSID On August 24 through September 1, 1987, an evaluation of the failure of the McGuire 2B RTB (No. 24Y98508, #4) took place at Westinghouse's NSID facility in Monroeville, PA.
The RTB was tested in the as-received condition, starting with banic measurements of key components, which were compared to those of an operable circuit breaker. After initial evaluation, the RTB was cycled 42 times with various means employed for tripping the RTB (undervoltage trip, shunt trip, and manual operation): no failure to trip occurred. Then, the RTB was partially disassembled, and the roller on the main drive link was found to have been striking the right-hand (viewed from rear of RTB) side frame plate. The weld between the pole shaft and center-pole lever was observed to be completely broken, which allowed the roller end of the main drive link to be skewed to the right (viewed from rear) and the roller axis to be canted counterclockwise approximately 3 to 5 degrees.
To allow further evaluation of the operation of the RTB, the racking (levering) mechanism, charging m toe, and the undervoltage trip attachment were removed from the RTB and the remaining portions of the mechanism were returned to operating condition.
Then, the RTB was operated approximately 15 more times. Some operations were performed with shims between the roller and side of the drive link in an attempt to cancel the eifects of wear. No failure to trip occurred.
Spring tension was applied to the side of the main drive link to force it harder against the side frame plate. The circuit breaker still tripped.
An attempt to force the roller higher onto the cam (above top dead center) also failed to prevent tripping.
The right side plate and the roller were then replaced to put the RTB in a less worn condition. The RTB was operated, and the new roller was observed to strike the side plate upon closing. At the 22nd cycle, the RTB failed to l
open. The closing cam was observed to have not completely rotated by 18 degrees (The closing springs were not completely discharged.). The trip shaft was completely free of the trip latch. The edge of the trio latch had moved approximately 1/16 inch above the lower edge of the trip shaft, indicating that the trip latch had operated. The roller of the main drive link was found to be wedged between the right side frama plate and the left-hand cam segment of the closing cam. The roller was not canted about its axis as far as it had i --
7-6177-5 been in past closings (it was now riding 1 to 2 degrees counterclockwise).
The condition of the roller with respect to the cam edge was photographed via fiber optics to record the jammed condition.
After deliberation by the evaluation team, a lever was used to eclieve the pinching of the roller by the cam and side plate and thus free the roller.
The failure was replicsted twice more by manipulation of the closing cam and the rolle'r on the main drive link.
During the first replication, the con-straining link was removed to verify that the failure was not partially or fully related to binding in the trip latch bearing (evaluation of the bearing in the original side plate had revealed damage to the trip latch bearing).
During the second replication of the failure, the roller was released from the bound condition by operating the manual charging lever.
This verified that further rotation of the closing cam caused the roller to be pushed out of the jammed condition as had occurred during the failure to open in service at the McGuire plant.
In an attempt to determine if the failurs of the pole shaft weld was necessary to allow jamming to occur, two additional pole shafts were installed in the circuit breaker.
Both had center-pole levers that were out of alignment in a condition similar to the fa!1ed pole shaft lever.
The two were selected from a batch of 18 pole shafts that were available at NSID where the failure evaluation was being performed.
Multiple attempts were made to cause the RTB to fail to open; none were successful. Although the roller could be forced to rest between the cam edge and the frame, the roller would not bind, f
3
- m e
-r-6171-5 t
3.
DISC 1CSION OF FAILURE MODE Figure 1 provides a labeled view.of the RTB mechanism.
Item 2 is the i
close cam, item 15 is the main roller, item 3 is the roller constraining link, and item 19 is the side frame plate (only partially shown).
Figure 2 shows the position of the RTB mechanism with the RTB closed.
The rear of the RTB is at the right in both figures. Figut t 3 conceptually shows the roller pinched between the side frame plate and the left cam segment as viewed from the rear of the RTB. To actually observe the area where the pinching occurs, the are chutes must be removed and inspection mirrors or fiber optics must be used.
The roller becomes pinched during the closing action. As the closing cam rotates, the edge of the roller is caught between the outer laminate of the cam segment and a spacer (see Figure 4).-
Continued rotation of the cam causes the roller to straighten in a clockwise rotation about its axis. This action causes the edge of the roller (marked "W" in Figure (A) to attempt to separate the cam and the side plate. However, the cam and side plate are not free to move and therefore pinch and bind the roller. When the trip latch is released, allowing the constraining link (item 3, Figure 1) to be free, the binding or jamming of the roller prevents the roller from rolling down the cam face to allow the circuit breaker to'open. The jamming of the roller also prevents l
full discharge of the cloaing springs, leaving the closing cam 18 degrees from I
the fully rotated position.
Upon removal of the original right side frame plate, the trip latch l
bearing was found to have a broken edge. The concern that binding of the trip latch was partially causing the jamming was alleviated during the second jamming of the RTB when the constraining link was removed and the RTB remained jammed.
Binding of the crank shaft on which the close cam is mounted could i
also have partially caused the binding. However, the crank shaft was found to be free and turning properly when the RTB mechanism was disassembled. Elimi-r.ation of trip latch and crank shaf t binding confirmed that the sole cause of the failure was pinching of the roller between the edge of the close cam and the right side frame plate.
The failed attempts to jam the circuit breaker with pole shafts having unbroken center-pole-lever welds showed that both the lateral displacement of I
the roller end of the main drive link and a 3-to 5-degree canting (cocking) r f !
i
,,..,-.. - - -,,,- _ -,,,, _.,, _ -.-~ - -,
I F-6177-5 of the axis of the roller were necessary for jamming '> occur. The pole shafts with unbroken. welds did permit the lateral displacement, even allowing the roller to strike the side frame; but they did not permit sufficient axial rotation (uncocking) of the roller for it to become jammed between the cam and side plate. However, the pole shafts selected in this attempt to jam the circuit breaker were new and, as such, had not experienced the wear that could allow the necessary 3-to 5-degree canting of the roller axis.
,-,,.,_m 4
,,e n
--,,,v-
,n
~ ^7j-co:
r J-F-6177-5 4.
DISCUSSION OF WESTINGHOUSE RECOMMENDATIchS FOR EVALUATION OF WELD DEFICIDiCIES In addition to the failure of the~ center pole lever-to-pole shaft weld on the McGuire 3B RTB, three similar weld failures sre'known to have occurred on DS-type circuit breakers.
One occurred at a Duke hydroelectric power plant approximately 14 years ago and the two others occurred.at nuclear plants within the last year. In one case, the circuit breaker failed to close when both the center-pole lever and anti-bounce lever to pole-shaft welds failed.
In response to an NRC request, Westinghouse developed inspection recommen-
.dations (4) for Westinghouse Owners Group members to use on DS-type breaker pole-sha#t welds. The basic recommendation is to perform a weld inspection with the role shaft in place in the circuit breaker and the top bracket removed from the RTB..Although removal of the top bracket improves the visibility of the weld on the center-pole lever, the weld is still located between the two side frame plates that are 1 inch apart and the weld is approximately 3 inches below the top of the plates.
Thus, assessment of the weld is difficult, especially because a closed crack may be difficult to observe visually under the best of conditions.
Westinghouse recommended short-term evaluations at the next surveillance.
t The criteria for short-term acceptance of the three pole-lever welds are as follows:
1 1.
Completely separated welds:
remove the circuit breaker from service.
2.
Cracked weld:
remove the circuit breaker from main RTB service and use it only as a reactor trip bypass breaker until the weld condition is corrected, a
3.
Excluding the ends of the weld, which may show evidence of a cold i
start, the weld should have at least 3/16-in fillet for 90' contin-uously around the pole shaf t.
If the fillet is less than 3/16 in, then the weld must have at least a 1/8-in fillet for 120' contin-uously around the pole shaft.
If these dimensions are not met, then j
the RTB should only be used as bypass RTB until the weld condition is corrected.
The long-term actions, to be performed at the next refueling outage according to the Westinghouse letter, are to inspect the remainder of the I,
welds on the pole shaft excluding the stop lever we..ds, to replace the pole shaft if necessary, and to check the alignment of the breaker mechanism.
In l
l
- i i
(
F-6177-5 i
the alignment check, it must be verified that the roller on the main-drive link is riding on the two outside cam segments and, while the breaker is in the closed position, not in contact with either side-frame plate.
The Westinghouse letter further states that the short-and long-term inspection criteria for the DS-416 apply to the DSL-416 and DS-420.
The L
timing of the inspections of the DS-206 and DSL-206 circuit breakers, which Westinghouse claims are less stressed, is left to the utilities, but must be performed on or prior to the next refueling outage.
1 i
f i t
e F-6177-5 5.
FRC COMMENTS ON WESTINGHOUSE RECOMMENDATIONS The recommendations contained in the Westinghouse letter of September 11, 1987 (4) have many shortcomings.
It is clear from the letter that a circuit breaker with a completely severed weld should be removed from service. How-ever, according to Westinghouse, welds with cracks that are not completely broken or are shorter than 90' (3/16-in fillet) or 120' (1/8-in fillet) may be used as a bypass RTB. No criteria are given for just how degraded these welds must be to be considered inadequate for bypass service, and no reasoning is provided for stating that an RTB with a degraded weld is acceptable for bypass service.
Is 10' of weld acceptable? Is 90' of cracking accepttble? At present, the criteria merely state that if the weld is not completely severed, the circuit breaker is acceptable for a bypass RTB. This is not a logical assumption.
In addition, Westinghouse claimed that the 180' weld originally specified had a conservatively calculated safety factor of 3.5.
This value was never substantiated either in the letter or in the subsequent presentation on September 23, 1987.
Complete inspection criteria for the weld are not provided in the letter. Neither enough information nor a specific standard required for use in the inspection are given.
During a meeting between Westinghouse and the NRC, held on September 23, 1987 in NRC offices in Bethesda, MD, Westinghouse personnel stated that a lack of fusion would be grounds for rejection of a weld and would require corrective action. However, the Westinghouse letter does not have any such requirement.
During the meeting, the selection of 90' of continuous weld was described by Westinghouse as being acceptable for permanent use. The original design required at least 180' of weld with a 3/16-in fillet.
The 180' weld design had been proven to be adequate by a 4000-cycle qualification test. The new 90*/120' weld acceptance critoria are based on static analyses and derating factors that do not take dynamic loads and fatigue fully into account and do not account for stress concentrations at the edges of cold starts in the welds. The visual inspection criteria given in the Westinghouse letter allow cold starts even taough there is no way of telling if good fusion has occurred after the cold start. No qualification testing has been performed with welds that are 90*/120* long that have cold starts (i.e., no breaker is known to have been cycled 4000 or more times with j
such a weld).
.g_
?
F-6177-5 The calculations of working and failure torque presented at the September 23, 1987 meeting indicated that the working torque for the 180' 3/16-in weld
[
is 1230 in-lb.
The formulae used are valid for a total of 180' are length of symmetrically placed welds, not for the one-sided weld configuration specified by Westinghouse. No explicit formulae are presented in the AISC handbook (6) for partial are fillets, but the calculation methodology that should be used is similar to that described in the section for Eccentric Loads on Weld Groups. The load capacity of partial arc fillets is significantly less than the are length ratio times the capacity for an all-arouAd circular weld.
Other loading effects, such as bending in the weld due to misalignment of the roller, were neglected in the Westinghouse analysis.
Experimental validation data were presented by Westinghouse to indicate that, under dynamic conditions and without electrical load, the torque on the weld is less than the allowable. The experiment 2nvolved putting strain gauge rosettes, wired to measure shear, on the shaft on each side of the lever and closing and opening the RTB. The calibration and the data reduction techniques described were incorrect and the torque data for the weld were meaningless, but not necessarily unconservative. This was pointed out to Westinghouse at the meating.
It may be possible for Westinghouse to reinterpret the data and derive meaningful dynamic weld torque information.
To demonstrate that a 1/8-in fillet could be used for 120', Westinghouse performed a static load test.
The weld configuration was a ground-down 3/16-in weld.
Such a modified weld may have fewer surface irregularities than a fillet with an 1/8-in bead, and may be significantly stronger than the weld it is intended to simulate.
In addition to problems with weld acceptance criteria, the 4000-cycle qualification limit is being approached by some RTBs. The estimated number of cycles on the McGuire 2B RTB that failed is between 2500 and 3500 cycles.
Westinghouse states that some test circuit breakers have been cycled to at least 10,000 cycles: however, it must be assumed that these circuit breakers had 180* welds unless the pole shaf ts are available and can be inspected to determine weld condition and length.
_9
T e
.-6177-5 6.
OTHER ISSUES AND CONCERNS The observed damage to the latch bearing is the result of the torsional forces on the constraining link and trip latch caused by the unwanted lateral' displacement of the roller end of the main drive link.
The lateral displace-ment could be caused by the breaking of the center-pole-lever weld or by the center-pole lever not being fully perpendicular to the pole shaf t.
It is possible that another type of failure could occur due to the misalignment of main drive link and roller if the damage to the trip latch bearing progresses and binds the trip latch pivot pin.
During inspection of the RTB components, damage was noted on the surfaces-of the closing cam. The cam in the McGuire RTB is composed of four steel segments that are sandwiched together and held by three rivets. The two outer segments are heat-treated steel; the two inner segmentr are non-hardened steel. The surface of the segments is supposed to be of uniform shape except in the area of the stop roller that is fixed in a hollow in the two center segments. However, on the McGuire RTB, the two outer segments are slightly larger than the inner segments, providing the edge for the roller to catch upon.
The cam had also been mushroomed by the drive link roller in a number of areas. Of key concern was mushrooming in the area of the stop roller (item 1 in Figure 1).
The stop roller holds the mechanism in readiness for release of the closing latch. The extreme mushrooming impeded rotation of the stop roller.
It is possible that continued mushrooming could totally prevent stop-roller operation, which could prevent closure of the circuit breaker upon demand. While not of concern for an RTB, a failure-to-close condition would be of concern for DS-type circuit breakers used in safety applications requiring energitation of the connected loads.
Eighteen new pole shafts were evaluated by Westinghouse for use in the RTB to determine the importance of a broken weld to a jamming condition. The welds on the center-pole lever are supposed to extend a minimum of 180' around j
the surface of the shaft.
Because of the geometry of the adjacent anti-bounce lever, the center-pole-lever weld must be made in two segments: one approximately 120' and the other approximately 60'.
It was noted that approximately half of the pole shafts had only the 120' segment of the weld T-6177-5 and that some of these welds app 3ared to lack fusion to the base mecal of the lever and/or shaft.
It should also be noted that the weld that failed on the McGuire RTB also was only 120' long and appeared to have a lack of fusion for more than two-thirds of its length. These conditions were probably key to its failure.
From a metallurgical examination of welds on a shaft that had not failed, but which had been removed from service because of apparent defects, FRC found that the defects were more extensive than visually evident.
However, FRC fot.nd no in-servica cracking that had propagated from pre-existing defects.
This analysis is detailed in Appendix A.
1c.
F-6177-5 7.
CONCLUSIONS The McGuire 2B RTB failed to open due to pinching of the main roller between the raised edge of the closin; cam and the right-hand side plate.
Both lateral displacement of the roller #end of the main drive link and axial rotation of roller are necessary to allow jamming.
Some pole shaf ts without broken welds will allow the lateral displacement and even allow the roller to strike the side plate. However, a new pole shaft without broken welds would not allow sufficient axial rotation for jamming to occur.
It may be possible that 3000 or more cycles could cause wear that would allow the necessary axial rotation.
I A raised edge on the close cam is necessary to allow jamming.
In addition, the distance between the inner surface of the cam edge and the side frame plate must be nearly the same as the width of the roller.
Other failure modes may also be developing in the DS-416 circuit breakers.
If the roller hits the right-hand side frame plate (viewed from back of circuit breaker), the linkage exerts a lateral force on the constraining link and the trip latch that could result in trip latch bearing damage. Ultimately, such bearing damage could jam che trip latch and prevent the circuit breaker from opening.
(On September 23, 1987, Westinghouse personnel discounted this theory, but precented no formal evaluation of the problem.)
In addition, a failure-to-close condition could occur due to the mush-rooming of the cam if it causes binding of the stop roller. Such binding would prevent operation of the close release mechanism.
1 With respect to weld acceptance criteria, a detailed weld inspection standard is needed for evaluating the welds on the pole shaf t.
The standard must be either an existing industry standard or one specifically prepared for the pole shaft welds. The inspectors need to know exactly how to judge the welds and what is or is not acceptable.
It is doubtful that an adequate inspection of the weld can be made with a pole shaft still in the circuit breaker, especially an inspection of the center-pole-lever weld. The restricted space and viewing angle do not allow proper inspection.
Even the use of dye-penetrant inspection may not be feasible with the pole shaft still in the circuit breaker, because of inaccessibility to properly prepare the surfaces (a multi-step process).
! 1
4 T-6177 i The reduction of acceptable welds to the 90'/120* are length reconnended by Westinghouse has not been proven acceptable by either qualification testing or sound engineering analysis.
B.
F-6177-5 8.
RECOMMENDATIONS The inspection of pole-shaft welds is highly desirable. However, at o
present, insufficient information has been provided to the users of j
the circuit breakers to perform uniform inspections. A weld 1
inspection standard must be specified or a complete inspection guide provided.
There is no established basis for permanent usage of 90' and 120' o
welds. A qualification program should be performed for breakers with pole shafts with 90' and 120' welds.
If welds with cold starts are to be allowed for unrestricttd use, then the qualification specimens must contain welds of 90* and 120' having cold starts.
o For interim usage of the 90'/120' welds, Westinghouse should compute allowable torques based on proper formulae with conservative stress concentration factots.
In addition, it is suggested that the dynamic strain gauge measurements be either remade or properly reevaluated, establishing the dynamic torque history during opening and closing of an RTB for use in fatigue calculations, o
Since some circuit breakers are approaching the qualification limit of 4000 cycles, refurbishment criteria must be established or the qualification limit must be extended by ratest or verification that cycling test data exist. The qualificetion limit for 180' and 90*/120' welds must be established, o
With respect to restriction of the stop roller by the mushrooming of the edges of the close cam segments, inspection criteria should be added to maintenance procedures for all Class IE DS circuit breakers.
Remedial actions should also be specified.
/
F-6177-5 9.
REFERENCES 1.
NRC Information Notice No. 87-35, "Reactor Trip Breaker, Westinghouse Model DS-416, Tailed to Open on Manual Initiation from the Control Room,"
July 30, 1987.
2.
NRC Information Notice No. 87-35, Supplement 1, "Reactor Trip Breaker, Westinghouse Model DS-416, Failed to Open on Manual Initiation from the Control Room," December 16, 1987.
3.
NRC Bulletin No. 88-01, "Defects in Westinghouse Circuit Breakers,"
February 5, 1988.
4.
Westinghouse letter from H. C. Walls to the Westinghouse Owners' Group (WOG) that use its Models DS-416, DSL-416, DS-420, DS-206 and DSL-206 switchgear in Class 1E service, dated September 11, 1987.
5.
"Instructiens for Low-Voltage Power Circuit Breakers, Types DS and DSL,
Westinghouse Electric Corporation, Instruction Bulletin 33-790-1E, September, 1979.
6.
"Manual of Steel Construction," Eighth Edition, American Institute of Steel Construction, Inc. (AISC).
4 -
l p
i i
1
- 1. Stop Honor
- 2. Close Com
- 3. Roller Constraining Link
[
- 4. Pivot Pin j,4
- 6. Trip Leech iQ, *
~
i
- 6. Trip Shelt Leach'ng Susf ace Pg',
- 7. Trip Shef a N
A I
i e,
- 8. Pole Shef 4
[
s
- 9. Center Pole Lower s
g
)
- 10. Pole Lower Pin 5
'4)
/
e
- 11. Moving Conaect Arm
~,_
i
's
.Q l) i
- 12. Stettonary Arcing Contact
- 13. Moving Contret Pivot Pin
,[
()NA, N hh h s
- 14. Mein Dr6w t. ink
'N
.j (6~3
{#
l
- 16. Mein RcAtor
- 16. Sprin6 Release Leach 2
W@,',,
h
- 17. Insulating Link Aa4usting a
Stud and Locknut j
- 18. Insuledag Link
___g fg
- 19. Mechanism Side Frame
- 20. Herdened Leach Seseaces Figure 1.
Linkages of DS-416 Breaker Mechanism Shown with CB Open and Springs Charged (Source: Reference 5)
}
b
4 P-6177-5 G':
7-l15)'
-L m
es
,,.x s
O
. -fQ A.
7X B.
50X Figure 11.
Macrograph and micrograph showing the cross section of Figure 10 after repolishing. The crack, not visible at 7X, but indicated with an arrv.t in A, now extends from the free surface down into i
the weld. Also evident is a greater fusion depth into the lever, some porosity in the weld metal, and some scratches (straight j
lines) on the polished surface.
l l
i w
w-
.-------m- --
e---
e.~
i l
l l
I t
I l
6.5X Figure 12.
Macrograph showing a metallographically prepared cross section of the anti-bounce lever to shaft weld near the weld start shown in Figure 6.
There is shallow fusion into the lever and a line of no fusion (arrow) at the toe of the weld on the lever.
- l r
S
___.-._______________,_m._.m.
9 - y.,.
rw
s 2
i s
()
4 i
,s.A.
w
[Y ' ' #
3
~,
v e
s p..j-a. -
e
.n,,
sf' _ :f.- y. N 8.. ^
.s
., J.,,%
b,,' :,0
?
cf. = ~ '
,. ;o #..,
e.
g.
r..
., ~ <,,
....~-
- k,,*l,. s[s y g
' i-x
. S. p
~*
g s
g *,,* 4..h,, fw *..
er i
=
a a,
l T
s.
,(*.
- .N 4
s
, w
'W -
r\\ q b&.
i.
f- . n' *
, w*.r &j% '.k,,
~ '
~
-n,.
,. f~q'.
y.p7+,
- 9, e v:,,c.
&. m %
,s
.5
. p.
. ~,.
,s ss.4
~.
e hh$4)
- N' W
.7 L
h T
A.
Shaft 100X B.
Anti-Bounce Lever 100X Figure 13.
Micrographs showing the weld of Fig re 12 after repolishing.
Some lack of fusin at the lever is stili present at the lever face and i
there is some porosity in the weld at the shaft. i l
i 0
1
l l
r l
l l
l R
E.ND OF A_
1 PoRO6tTT
. ~rw
" M:u s p
t EW R CEMTER,-BokC Su R F AC,".
.t"M.... g4-l w
-,w N 9X i
Figure 14.
Macrograph showing one of the mating fracture surfaces exposed by sectioning near the termination of the surface crack on the finishing end of the AB/PS weld.
The rough fracture surface, which reflects the solidification pattern of the weld, was darkened, apparently f rom high temperature exposure following j
cracking during cooling of the weld, j.
O i
l l
~,
.. ' !.sh,,. -
.h,. ' ' T,. -
l;_"., _
-- v...-
\\
-.g. 7 s,
- .,1, ~,,
s l
+
.i.-
~~s x
(.'
A' i
t
._r 1
i g
4 j
a i
t
,. ~
j e,
i l
'N,..'.,'
\\,
7-s -
~.
1 g
l I
. r
~
hi
,/
i g
A.
Anti-Bounce 300X v
l
\\,
41 m
l w
e
- w l
p j
t i
ft.-
(
5 s,
N s 3, N
s., '
4,b %f w
.g B.
Anti-Bounce 1000X Figure 15.
SEM micrographs showing a representative area on the weld fracture surface exposed by sectioning near the terTnination of the surface crack shown in Figures 4A and 8.
Essentially the entire fracture surface morphology is ductile dimpled tearing, t
4 i a
I
I F-6177-005 s
~>
p
\\
i n.
's [
9 '
\\
- j
(-'
z '..
v 6v',
e h
a' p
8, s~
s e '.
g f
3 g
- ' i
,X
' 4.:. ' L.
\\
A.
Center-Pole 300X k
k
\\, 'N,'
- 't i
i
~
L o
'K l
a
^
1
\\
s g
l
. p l
2 s.
s B.
Center-Pole 1000X Fioure 16.
SD4 micrographs showing a representative area on the fracture surface exposed 1,/ sectioning near the termination of the surface crack shown in Figures 4B and 7.
Partially masked by pronounced smearing (on the right) is ductile tearing similar to that on the anti-bounce cracked weld joint in Figure 15. -.
.. _ _ _ - - - ~. -. -
4 b
- l l
1 4
4 I
-yyg g
!L'a 3
.gb A
tt.
1 t
I 7X Figure 17.
Macrograph showing a metallographically prepared cross section of the center-pole to pole-shaf t weld af ter sectioning and grinding to a depth of about 80% of the surface length of the crack in Figure 7.
In this cross section, the crack runs between the two arrows from the weld surface to the bottom of the joint.
In this section, the etching delineated a different microstructure at the root of the weld, indicating that a root pass had preceded the bulk of the weld. The crack, shown at higher magnification in i
Figure 18, had a sharp change in direction at this structural change.,
e l
,~...e
,,w.
a r
i I
i s
i 1
l t
'., f, r
v..
i l
. 7'.:r??R. $n
- g
~.
l-e4 i
-i i
t I
i' Figure 18.
Composite micrograph showing a metallographically prepared cross section of the center-pole to pole-shaft weld after polishing the O
section in Figure 17 a little deeper. The crack, which was so tight at a
this location that ii could not be delineated at 7X magnification, 3
extended from the surface to the root of the weld, with a sharp change in direction near the root at the upper lef t.
The black spots are
[
either etching effects or porosity.
J
l l
l f
J I
[P ~, -
p 3
r j
~
. % +,
u
..-k'%
J k'[' q ',
a A
- a 4
0
(
i
~ ' Q +]tr '!-
' 4%,
3,.
\\'
._ f.?', ;. ', -
-l..? y,
- g%.%,,
r... f...
c6
~
g
-9 g
1 \\
.N d
- I
'S,A-(,['.t,f.
'f
)
t f~;a.,
s 1
a
- 4. + ' %
tY, g
,y i
3 t4[.
t'
/C l,
'f,"$
$h.
%l' Q'.
~
~
r.
4 t
s.
500X Figure 19.
Micrograph showing a metallographically prepared cross section near the root of the anti-bounce lever to pole-shaft weld at the end of the surface cracking shown in Figure 8.
At this stage of sectioning, about 1/2 inch of the circumferential length of the finish ig end of the weld had been ground away.
l
.l 1
.