ML17138B265

Final Deficiency Rept Re Cutler-Hammer E-30 Pushbutton Switches.Sticking of Pushbutton Occurs on two-button Modules.Dislodgement of Operating Module Apparently Caused by Mechanical Locking Ring Not Being Tightly Locked

ML17138B265
Person / Time
Site:	Susquehanna
Issue date:	04/18/1980
From:	PENNSYLVANIA POWER & LIGHT CO.
To:
Shared Package
ML17138B266	List: ML17138B265 ML18017A145
References
NUDOCS 8004300383
Download: ML17138B265 (5)

Text

048 Ou9&9 Cutler Eunmer E-30 Pushbuttons Descri tion ae Cutlerdhnxner pushbutton switches installed in PGCC/ACR benchboards, are stickLng; of 189 switches inspected, ten switches do not retract when manual depression is relieved. these switches bear GE P/N 851E341P2ZKDOZ and are purchased frcm Cutler-Hamner under P/N CH E-30AC. Mesc nonconfor-mances are documented on NCR 4313.

b. A second condition, as described on Sheet 20 af NCR 3403, involves switch

{HV-1112D) CH E-30AB, which became dislodged due to faQ.ure of the mechanical locking device between the pushbutton operating module and contact block.

MhLl.e this particular switch and the second switch of the same type which have failed were installed in non-safety related equipnent, the concern is for similar switches used in safety related equipnent.

Cause

a. She sticking of the pushbuttons occurs on 2-button modules. When one pushbutton is depressed, a latch device is actuated that prevents depres-sing the other pushbutton. During the manufacturing process the lower edge of.the plastic pushbutton cap will have a sharp edge formed. this can beccme wedged sgainst the latch preventing the button frcm resetting when it is released.

b. %he dislodgement of the pushbutton operating module fran the contact block is apparently locked.

caused by the mechanical locking r~ not being tight+

Anal sis of Safe Imnlications

a. CH E-30 pushbuttons are used in Class 1E safety related circuits. Ghe sticking of the pushbuttons could result in the inoperation, misoperation, or improper lir~p of electro-mechanical or solid state logic when push-button operation is attempted.

Prospect Enquiring has deterndzed that the reported CH E-30 pushbutton sticking is a reportable deficiency under 1CCRP50-55 (e).

At.Mchment 1 Page 1 of 3

b. The dislodgement af the CH E-30 loddzg device permitted the contact blocks to disengage frcm the pushbutton operator sufficiently to prevent the con-tacts fran changing state upon depression of the pushbutton. We identi-fied deficiency, had it gone, uncorrected, could have prevented a safety system fran performing its attended function.

Pro)ect EhgineerLag has determined that the reported CH E-30 d1slodged mechanical locking device is a reportable deficiency under 10CPR50.55(e).

Corrective Actions ao GE requested that CutleMhnmer investigate the pushbutton stiddzg problan for the PGCC/ACR at the site. After the investigation Cutler-Hsaam.r reported.,

that the problem was on pushbutton switches manufactured before Novenber 1978.

In a report attached to a letter dated February 29, 1980 (DCN 116060), Cutler-Hamner stated Chat the sticking problem was first noted in June 1976. Ghey did not consider on the first it to be a critical problem because attanpt to depress, it if the button stuck could always be depressed with further attempts. Manufacturing changes were made in September 1977 and November 1978. Cutler-ihmner feels that, with these doges, it is reasonable to ass+ac that devices made after November 1978 wed. not exhibit this problem.

9he date stamp is located under the button on top af the button plate and will show the month and year the device was manufactured.

In the letter Cutlerdizaner states that a stickLng pushbutton can be corrected by scraping or buffing a slight radius to the sharp lower edge of the button where it contacts the latching device, or by operating the pushbutton several t&es. 'Ihey state that the operation of the pushbuttons should ~rove with use due to anoothing of the sharp edge af the button against the interlock surface.

For the PGCC/ACR benchboards General ELectric has agreed to replace all Cutler Hammer E-30AC operator pushbutton aadules manufactured priar Co November 1978.

The contact modules N~ not be replaced. NCR 4313 shall be revised to re-flect the corrective action taken.

b She construction of the Cutler-khmmer type E-30 pushbutton switches is such that the contact blocks are held to the aperating module by a lockfzg ring mounted on the operating module and actuated by a quarter-turn looking screw.

We dislodgement of the contact modules fran the operating modules was in-vestigated and no manufactur~ defect was apparent-Evince these seme devices, successfully passed an integrated system factory test at the manufacturers facility, 1t is assumed the locking devices worked loose during transit. We lacing devices for CH E-30 switches will be retightened to assure proper installation for both Units 1 and 2. QC inspection will verify that the lacing devices are ~t.

Page 2 of 3

c. Pro)cot Zhgineering is investigating the usage of Cutler-Hammer type E-30 pushbuttons M essential applications in the renainder of the plant. At the conclusion of the investigation any deficiencies will be documented on NCR's and the field advised of any future corrective actions to be taken+

Conclusions

a. The replacement of pushbutton operation nudules that are, or are suspected of not operating properly should resolve the problem of the E-30 AC pushbuttons on the PGCC/ACR benchboards. NCR 4313 will track the'eplacement of the pushbutton modules on the PGCC/ACR bend!boards.

Pro)cot Z~mering instructions and field verification should insure no

'urther dislodgenent of the contact modules from the operating module.

c. Pusher investigations to date have fai1ed to locate any Cutler-Ham!ner type E-30 pushbutton switches being used in the remainder of the plant.

Page 3 of 3

SOB31XT Amphenol fanale connectors The arxxaalies surrrnarized belaw have been noted during routine Quality Contxol surveillance and/or inspections of PGCC/ACR Panels.

Amphenol female connector pins are not seating properly in the dielectric..

The relative position of the pins often vary after mating and unmating of the connectors. It has been observed that scrne pins partially retract as the re-sult of turning the connector. This is especiaU.y the case after pins have been ramved arel reinserted into the connector dielectric.

Plhile performing pull tests on the individual conductors to check the holding force of &w pin crimps to the wires, the connector fana1e pins were extracted fxcrn the dielectric, in many instances, befare the required puU. test farce was achieved, One of the methods for performing the pull testing of conductor pin crimps was to use a henostat pull testing device. The device was utilized for pull testing to avoid ccraplete rival of all pins frcrn connectors. The hamstat pull test methcd tended to amplify or further aggravate the Amphenol connector problem.

General Electric conducted a quality study on the mndition identified with these Ampherol conmatars. The results were attached to their letter GB-79-283 dated Decanber 10, 1979 and foxwaxded to the NBC in our PZA&38. GE concluded that the probable cause of pin seating nonconfoxmances was not due to connector design but rather to damaged mating connectors, errors during assembly or rework at. the plant site.

ANALYSIS CF SAE'EKY 3ZPLICATICNS The identification of pins which had backed out. fran the dielectric raised a con:em relative to the connectors ability to maintain continuity. However, regardless of how well the pin seats in the dielectric, as long as the male and female pins are mated together; the circuit continuity can be maintained thxough it tie pins. In no instance was reported that the pins failed to mate together.

Project Engineering has determined that the identified Amphenol connector pins, that did not seat, properly in the dielectric, do red constitute a reportable deficiency under 10CPR50.55(e) .

Attachnent 2