ML20151S233

Insp Rept 99900100/88-01 on 880502-06.Nonconformance Noted. Major Areas Inspected:Operational,Qa & Environ Issues Re Supply of Motor Actuators & Parts to Nuclear Industry & Limited Review of QA Program

ML20151S233
Person / Time
Issue date:	08/01/1988
From:	Jeffrey Jacobson, Potapovs U Office of Nuclear Reactor Regulation
To:
Shared Package
ML20151S195	List: ML20151S191 ML20151S219 ML20151S233
References
REF-QA-99900100 NUDOCS 8808150112
Download: ML20151S233 (16)

Text

ORGANIZATION: LIMITORQUE CORPORATION LYNCHBURG, VIRGINIA REPORT INSPECTION INSPECTION N0.: 99900100/88-01 DATES: 05/02-06/88 ON-SITE HOURS- Pn CORRESPONDENCE ADDRESS: Mr. I. E. Wilkinson Vice President, Engineering Limitorque Corporation 5114 Woodall Road Post Office Box 11318 Lynchburg, Virginia 24506-1318 ORGANIZATIONAL CONTACT: Mr. P. McQuillan, QA Administrator TELEPHONE NUMBER: 804-528-4400 NUCLEAR INDUSTRY ACTIVITY: Manufacturer of motor operated valve actuators.

ALIGNED INSPECTOR: W h f /Pf Jeffp)!y pf Jacobson, Special Projects Inspection Date Settien(plS)

OTHER INSPECTOR (S): Ray Cilimberg, PDRIS Steve pr,SPIS APPROVED BY: j f,lb /d,w 7-l-gS Uldis Po~tapovs, Criief, TIS, Vencor Inspection Branch Date INSPECTION BASES AND SCOPE:

A. BASES: 10 CFR 50, Appendix B and 10 CFR Part 21.

B. SCOPE: This inspection was made to review operational, quality assurance, and environmental qualification issues concerning Limitorque's supply of motor actuators and parts to the nucle 7 industry. Additionally, a limited review of Limitorque's quality assurance program was conducted.

PLANT SITE APPLICABILITY: All nuclear utilities. I l

l l

8808150111 880801 l PDR OA999 EMVLMIT i 99900100 PDC

ORGANIZATION: LIMITORQUE CORPORATION LYNCHBURG, VIRGINIA REPORT INSPECTION NO.: 99900100/8_8-01 RESULTS: I? AGE 2 of 16 A. VIOLATIONS:

None.

B. NONCONFORMANCES:

1. Contrary to Criterion 111 of Appendix B to 10 CFR Part 50, Limitorque has not established measures to ensure that shaft keys and terminal blocks procured from comercial grade subvendors conform to the applicable design requirements. Specifically, measures were not established to ensure that shaft keys procured comercial grade are, in fact, mode of type 1018 steel and that terminal blocks procured commercial grade are, in fact, the same as the ones previously tested and environmentally qualified. (88-01-01)

2. Contrary to Criterion XVI of Appendix B to 10 CFR Part 50, measures have not been established to ensure that conditions adverse to quality are promptly icentified and corrected, nor have measures been estab-lished to ensure significant conditions adverse to quality are docu-mented and reported to appropriate levels of management. Specifically, failures attributed to defective shaft keys and to spring pack relaxation were not fully evaluated nor were they reported to dppropriate levels of management. No system exists at Limitorque to ensure that defects reported via field service or customer problem reports are evaluated and resolved. (88-01-02)

C. STATUS OF PREVIOUS INSPECTION FINDINGS:

1. (Closed) Violation 86-01-01: Limitorque did not pass on the require-ments of 10 CFR Part 21 to three vendors of safety-related parts and services.

The inspector reviewed several purchase orders from Limitorque to venoors of safety-related parts. All the purchase orders reviewed had properly invokeo 10 CFR Part 21.

2. (ClosM ) Violation 86-01-02: Limitorque had not properly reviewed or evaluated reports of burned internal wiring caused by energized space heaters.

The inspector verified that Limitorque procedure QCP-22 "Reporting of Defects for Safety-Related Equipment" had been revised to require de:umented evidence be maintained of Part 21 review comittee meetings. Also, the inspector verified that the wiring inspection checklist used by Limitorque inspectors had been updated to include d check for properly routed wires.

1

GRGANIZATICE: LIMITORQUE CORPORATION LYNCHBURG, VIRGINIA I

REPORT IllSPECTION NO.: 99900100/88-01 RESULTS:  ? AGE 3 of 16

3. (Closed)Nonconformance 86-01-03: Limitorque failed to impose Quality Assurance program requirements on H. K. Porter for the procurement of safety-related niotors.

The inspector verified that appropriate quality assurance requirements are now imposed on the suppliers of safety-related motors to Limitorque.

4. (0 pen) Nonconformance 86-01-04: Limitorque did not have procedures I to be followed by the Parts Department for ensuring the correctness and completeness of spare parts orders or for the initiation, review, and issue of Certificates of Compliance (C of C).

The inspector reviewec Limitorque's corrective action commitments.

Procedure QCP-21, Revision 4, cated February 20, 1988, ias issued to provide instructions for processing mre part orders. Procedure QCP-27 for the initiation, review, and issue of C of C for spare part orders had been drafted but not completed. Since QCP-27 has not been completed, this item is considered open and will be reviewed during a future NRC inspection.

5. (0 pen) Nonconformance 86-01-05: Limitorque did not process AP&L Company P0 No. 51965 for terminal strips under controlled conditie 5 to assure that prerequisites affecting quality of the order were identified and accomplished. It was nu.ed that Limitorque incorrect'y aavised AP&L personnel that Buchanan 724 terminal strips were a correct replacement for qualified Buchanan 524 terminal strips, ana issued a C of C that certified t :at Buchanan 724 terminal strips were envircamentally qualified.

The inspector reviewea a memo dated September 19, 1986, from the Quality Control Manager to personnel who prepare C of C's which defines the acceptable terminal strips for nuclear applications.

Buchanan 724 was not included as an acceptable terminal strip.

Return Goods Permit No. 14985, dated July 8, 1986, was issued for return of the unacceptable Buchanan 724 terminal strips which were received from AP&L on July 10, 1986. Limitorque committed o review previous parts orders to determine if a similar instance has raccurred, but this action has not been cumpleted. This item is considerea open and will be reviewed during a future NRC inspection.

6. (0 pen) Nonconformance d-01-05: Limitorque had failed to update Revision 2 to their Quality Assurance Manual (QAM) to address the training of non-quality control (QC) personnel. In response to this

ORGANIZATION: LIMITORQUE CCRPORATION LYNCHBURG, VIRGINIA REPORT lhSPECTION NO.: 99900100/88-01 RESULTS: PAGE 4 of 16 nonconformance, Limitorque committed to devi.op a training plan for these persennel (including Engineering, Parts and Shipping, Purchasing, Order Processing, and Field Service Departments) with training to begin by September 17, 1986.

The inspector noted that an internal memorandum requested an extension of the training start date from Limitorque management to October 15, 1986. Also, the subsequent revision of the QAM had been updated to address training of non-QC personnel. However, training for these personnel had not actually comannced ur.til much later.

Training for the Purchasing and Shipping Departments had not been completed until February 28, 1988 and February 26, 1988 respectively.

Training for three checkers in the Engineering Dept. (out of 13 personnel requiring training) hat :een completed on QC Procedure No.17, "Engineering Drawing issue and Reviev.'." The rest of the Engineering Department training was scheduled to be completed during May 1988 in the evenings. No training had been completed at the time of the inspection for Order Processin Service departments (a total of 20 people)g,This Sales, Parts, training and was to be completed in May 1988 also. Limitorque stated that they had been unable to start the training as planned due to scheduling conflicts. This issue will be reviewed further in a future NRC inspection.

7. (Closed) Nonconformance 85-01-09: Limitorque did not ef fectively implement QA program requirements for reporting defects in safety-related equipn;ent.

The inspector verified the issuance and implementation of QCP-22 "Reporting of Defects for Safety-Related Equipment" by reviewing records of Part 21 Review Comittee Meetings for several potentially reportable issues. The records provided evidence that a system is now in place tu evaluate defects for 10 CFR Part 21 reportability as long as potentiolly reportable defects are identified to the Part 21 Review Comittee.

D. OTHER FINDlhGS AND COMMENTS:

1. Shaft-keyway Failures in Actuators at Watts Bar As previously reported in NRC inspection report No. 99900100/87-01, on November 12, 1986, i.ne NRC received a report from the Tennessee Valley Authority concerning a failed motor pinion key and shaft in a Limitorque actuator installed in the Watts Bar Nuclear Plant. The failure was reported to be similar to the condition reported in IE Information hotice 81-08 which addressed failure of motor pinion

ORGANIZATION: LIMITORQUE CORPORATION LYNCdBURG, VIRGINIA REPORT INSPECTION N0.: 99900100/88-01 RESULTS: PAGE 5 of 16 keys in Limttorque SMB-4 actuators with motor torques in excess of 100 ft-lbs. A subsequent letter from Limitorque rccommended replacing the original mild steel keys with high strength keys on actuator sizes three and four with motors of 150 ft-lb or greater.

The failure at Watts Bar occurreo on a size three actuator with a 100 ft-lb motor not requiring key replacement.

Recent analysis by TVA has determined that other similar actuators installeo at Watts Bar have exhibited one additional cracked shaft, four deformed keys, and 10 motor shafts with tapered key slots.

TVA has also determined that replacing the 1018 (mild steel) key material with 4140 (high strength steel) key material will not resolve the problem of key deformation and slot tapering.

Limitorque is scheduled to test an actuator returneo from Watts Bar the week of May 9, 1988 in an effort to resolve this problem. TVA is planning to witness this testing and report the results to the NRC.

2. Defective Shaft Keys at South Texas Project On October 15, 1987, tne NRC received a report from Houston Lighting and Power Company concerning sheared motor shaf t-to-pinion keys found in a Limitorque actuator at their South Texas Project, Unit 2 nuclear plant. A similar key failure had also occurred in April of 1987. The sheared key from one of the valves was sent to a metallurgical laboratory where it was determined that the properties of the material were not consistant with those of AISI type 1018 steel. AISI type 1018 steel is the material that the keys are supposed to machined from as determined from Limitorque design documents.

In August 1987, three additional keys were removed from similar actuators and sent to the lab for analysis. One of the removed keys hoo been split and the other two showed signs of deformation.

Laboratory analysis of these keys again revealed that the properties of the keys were not consistent with those of AISI type 1018 steel.

The metallurgical analysis revealed that the keys had failed longitudinally along manganese sulfide inclusions and that the key material was probably 1118 or 1119 resulphurized grade steel.

Additionally, on March 16, 1983, Westinghouse submitted a 10 CFR Part 21 report to the NRC which concerned Westinghouse's discovery of sheared keys on six actuators. These keys were also analyzed and  !

determined to be of other than AISI 1018 grade steel. l I

l l

l

ORGANIZATION: LIMITORQUE CORPORATION LYNCHBURG, VIRGINIA REPORT INSPECTION N0.: 99900100/88-01 RESULTS: PAGE 6 of 16 Review of this issue with Limitorque revealed that in the time period when the South Texas actuators were manufactured, keys were purchased from a comercial vendor without certificates of conformance for the material type or properties. No checks were performed at Limitorque to verify that the proper key material was receivec. Keys purchased today are still bought commercial grade, however, Certified Material Test RepGets (CMTRs) are received from the vendor. Limitorque also now Ges a herdness and dimensional check of the keys upon receipt inspection. This method of procure-ment is still considered unacceptable due to the fact Limitorque has never audited the current supplier of these keys, Colfin Specialty Steel Corp., as necessary to verify the validity of their CMTR's.

Also, the checks done upon receipt inspection are not sufficient to verify that the keys are, in fact, type 1018 steel. Nonconformance 86-01-01 identified in Section B of this report was written as a result.

Limitorque comitted in the exit meeting to further evaluate this problem and report the results to the NRC. The fact that this item had not been adequately evaluated and that the Limitorque Part 21 Review Committee hao not been notified of this problem is cited in hunconformance 88-01-02 in Section B of this report.

3. Cracked Worm at Fort St. Vrain Cn June 5,1987, Public Service Company of Colorado (PSC) issued a Part 21 report to the NRC concerning cracks found on the worm teeth of six actuators of their Fort St. Vrain Nuclear Generating Station.

The worms and geus were part of actuators installed in the plant and hao been in operation since 1974. The actuators had been cycled approximately 100-200 times under ambient temperature conditions.

The crocked worms were submitted by PSC for laboratory analysis where it was determined that the cracking was most likely due to a grinding process during manufacture which subjected the worms to rapid heating and cooling. The grinding process resulted in untempered martensite and associated microcracks along contact surfaces of the worm threads. The microcracks become larger due to fotigue as shown by a scanning electron microscope analysis.

Although none of the cracked worms had completely failed, it is theorized the actuator could have become inoperable had the cefect gone undetected and the worm teeth completely broken off.

Limitorque had never been officially notified of this problem by PSC and therefore had not evaluated this issue. During the inspection, Limitorque indicated they knew of no other cases of cracked worm l -

ORGANIZATION: LIMITORQUE CORPORATION LYNCHBURG, VIRGINIA REPORT INSPECTION NO.: 999C0100/88-01 RESULTS: > AGE 7 of 16 teeth but said they would contact PSC and try to obtain the effected worms for analysis. Also, limitorque indicated they now perform a dye penetrant test on all worms, a practice which was not in effect at the time the subject worms were manufactured.

4. Cracked Worm Gear at Commanche Peak On March 18, 1988, the NRC received a Part 21 report from Limitorque concerning defective worm gears installed in two H3BC actuators at the Conmenche Peak Steam Electric Station. The defective gears were cracked and were sent to Limitorque for failure analysis.

Limitorque determineo the cracks were the result of porosity introduceo during the worm gear casting process. Limitorque knows of no other similar problems and it is assumed that one casting lot or portion of a lot m6y be defective. Efforts to obtain lot traceability have been inconclusive due to the fact that the probable casting supplier discontinued operations in 1983.

It is thought that defects of this nature would result in immediate failure upon gear loading. Limitorque stated that the actuators containing the defective worm gears at Cormianche Peak had never been loaced and that such failures should be evident at time of actuator installation.

As a result of these failures, Limitorque performed a visual inspec-tion of all worm gears presently in stock. Also, a sample of eight worm gears were sectioned or x-rayed with no defects or porosity observed. Limitorque has committed to notify all nuclear utilities of this problem by May 13, 1988 and will propose appropriate corrective action. It should be noted that this potential problem is limited to only H3BC type actuators.

5. Spring Pack Relaxation As cetoiled in NRC inspection report No. 99900100/87-01, limitorque has been notified of problems with "spring pack relaxation" by a number of nuclear utilities. Spring pack relaxation occurs when the bellville washers used to make up the spring packs installed in  ;

Limitorque actuators lose some of their spring modulus. As a result, the preloaded spring pack becomes relaxed. The spring pack will now deflect the torque switch to an open position at a lower actuator torque value than a properly working spring pack, resulting in incomplete volve actuation. This is especially significant due to the fact that properly sized and adjusted actuators could gradually

f.

ORGANIZATION: LIMITORQUE CORPORATION LYNCHBURG, VIRGINIA REPORT INSPECTION NO.: 99900100/88-01 RESULTS: PAGE 8 of 16 degrade to an inoperable state without warning. Most actuators are only required to deliver rated torque during accioent conditions and might, therefore, appear to be operational ouring normal stroke time testing.

Limitorque has analyzed one spring pack returned from the Salem Nuclear Plant and has determined that significant relaxation had occurred. Limitorque is still evaluating this problem and believes it might be limited to certain specific spring packs. Limitorque committed to complete their analysis of this problem by May 31, 1988 ano notify all nuclear utilities of the result. The fact that this item has not been acequately evaluated and that the Limitorque Part 21 Review Committee has not been notified is citeo in Nonconformance 88-01-02 in Section B of this report.

6. Hydraulic Lock Testing Program As detailed in NRC inspection report no. 99900100/87-01, spring pack hydraulic lock has been reported to occur by various nuclear utilities. Hydraulic lock occurs when the spring pack is filled with grease and then forced to quickly compress during valve seating.

The trapped grease cannot be relieved quickly, tne torque switch does not open, and the actuator essentially becomes locked. This condition results in rapid motor overheating and subsequent motor burnout.

Limitorque is currently working with the EPRl/NUMARC proup on resolu-tion of this problem. A questionnaire has been sent to a number of utilities asking questions concerning hydraulic lock. Responses to the questioner have indicated the hydraulic lock phenomena is most likely due to the trapping of grease in one of three areas within the spring pack compartment. Current prototype fixes have been developed to provide relief paths for the trapped grease in two of the three specif.ic areas. Also, modifications to prevent the flow of grease into the spring pack of some actuators are being developed.

Limitorre is planning to meet with the EPRl/NUMARC group June 9, 1988 to discu:,s future testing plans.

7. Review of 10 CFR Part 21 Evaluations and Reporting During the inspection, a review was conducted of Limitorque's program for evaluation and reporting of Part 21 deficiencies.

QCP-22 is the applicable Limitorque procedure for accomplishing the evaluation and reporting of these deficiencies. Minutes of six

ORGANIZATION: LIMITORQUE CORPORATION LYhCHBURG, VIRGINIA REPORT INSPECTION NO.: 99900100/88-01 RESULTS: 3 AGE 9 of 16 Port 21 Review Comittee n:eetings reviewed by the inspector indicated that potentially reportable items were being evaluated and resolved.

The inspector noted however, that two problems that had been reported to Limitorque by utilities had not been passed on to the Part 21 Review Conmittee. The two items in question were the spring pack relaxation problem identified by the Salem Nuclear Plant and the defective shaft keys as identified by South Texas Project.

This problem seems to be the result of the fact that no one person at Limitorque is responsible for tracking potentially reportable items. Three individuals, the technical manager, the quality control manager, end the quality assurance administrator, all seperately keep track of potentially reportable items. No log book existed where potentially reportable Part 21 defects could be tracked. It is not until one of the aforementioned people decides that the particular item is reportable that the item is brought to the Part 21 Review Committee. Nonconformance (88-01-02) was cited in Section B of this report as a result of this problem.

S. Review of Subvendor Control During the inspection, a review was conducted of Limitorque's program for subvendor control. A number of components used in Limitorque actuotors are procured by Limitorque from subvendors.

Some of these components such as motors, shaft keys, wire, and terminal blocks are crucial in assuring that the Limitorque actuators will perform as designed in both normal and accident environments.

It is therefore necessary that Limitorque maintain varying degrees of control on these parts procured from their subvendors.

The inspector reviewed purchase orders to both nuclear motor suppliers to Limitorque, Peerless and Reliance Electric. Purchase Order 24745 to Relionce Electric invoked Reliances QA program which had been reviewed and approved by Limitorque. Also, specific motor specifica-tions containing design and electrical performance characteristics were invoked. Purchase Order 24338 to Peerless was also reviewed and found to invuke similar requirements. The requirements invoked on these purchase orders seemed adequate to ensure the quality of these motors. Purchase Order 7483 to Foster Electric for Rockbestos SIS Firewall Wire and 41301 to Shaw Plastics for torque switches were also reviewed and determined to contain adequate requirements for these parts.

Purchase Order 24972 to Rev Car Fasteners for 1000 shaft keys was found to be inodequate. The purchase order did not invoke any QA o

v (

ORGANIZATION: LIMITORQUE CORPORATION LYhCHBURG, VIRGINIA REPORT INSPECTION N0.: 99900100/88-01 RESULTS: PAGE 10 of 16__

requirements but did ask for certified n.aterial test reports (CMTRs) which indicated physical properties of the key material. This vendor, however, has never been auditeo by Limitorque so the validity of these CMTR's has not been verified. Limitorque oces, however, Jo c hardness check on a sample of these keys upon receipt but this check is not sufficient to determine that the keys are of the proper material.

Purchase order (PO) 20567 to Foster Electric for EA5-304 terminal blocks was also determined to be inadequate. No QA requirements were invoked on this P0 and no specific receipt inspection checks are done on these terminal blocks at Limitorque. Also, this vendor has not been audited by Limitorque. Terminal blocks such as these are certified as being environmentolly qualified by Limitorque, however, Limitorque has no control on the material composition or configuration of these blocks.

Both the shaft keys and the terminal blocks are items procured by Limitorque from commercial grace subvendors. These devices have not had the benefit of an adequate dedication process at Limitorque as would be required for safety-related components such as these.

Nunconform(nce 88-01-01 cited in Section B of this report was written as a result.

9. Audits and Approved Vendors List A review was conducted of Limitorque's program for performing audits and for putting vendors on their Approved Venour List. Basically, the only requirement for the Approved Vencors List is that the venoor be sent a "Vendor Appraisal form," which tells the vendor that his meterial will be inspected upon receipt at Limitorque in accordance with MIL-STD-105P, Level II, AQL 1.5. Only manufacturer's of motors for nuclear service are then audited by Limitorque at a period of at least once every three years. Consequently, appearing on Limitorque's Approved Vendor's List gives no real assurance as to the quality of the part and would not serve as an acceptable basis in itself for cedication of commercial products.

Limitorque's audit of Reliance Electric, dated September 29, 1987, was reviewed by the inspector. The audit appeared to be adequate in completeness and four deficiencies were found to have had been identified by Limitorque. A follow-up audit was conducted by Limitorque on March 10, 1988 where it was identifieo that three of the four identified deficiencies still had not been corrected.

ORGANIZATION: LIttlTORQUE CORPORATI0li LYNCHBURG, VIRGINIA REPORT INSPECTION NO.: 99900100/88-01 l RESULTS: PAGE 11 of 16

10. Cracked Grease Seals on Motors Sent to San Onofre As detailed in Inspection Report No. 99900100/87-01, the San Onofre Nuclear Generating Station received a number of replacement motors from Limitorque with cracked or improperly installed grease seals.

Limitorque investigated this problem with Reliance Electric, the supplier of these motors to Limitorque. Reliance then in turn investigated the cause of the lip seal cracking with their supplier of the seals C.R. Industries. No definitive reason was ever determined for the cracked seals, however, C.R. Industries stated that exposures to high levels of ozone could cause cracking such as that observed on the subject motors. It is theorized that arc welding in the vicinity of the seals could cause high ozone levels in the area.

11. Durability and Environmental Qualification (EQ) of Motor Leaos for Lim 1 torque Actuators with DC Motors In late 1984, Peerless-Winsmith Company (formerly H.K. Porter) supplier of DC motors to Limitorque, requested Limitorque's approval to change the insulating system used on the motor leads. The original system which had been qualified under Limitorque Report No. B0009 consisted of Dupont Nomex tape covered by an epoxy-impregnated braided fiberglass jacket. The pronosed leads would be insulated with Nomex covered with a Kapton tape, in a letter to Peerless, dated November 6,1984, Limitorque d;; proved the homex/Kapton (N/K) system based on the belief that the thermal and radiation resistant characteristics of the leads would not be degraded. In a subsequent letter to Peerless, dated March 4, 1987, the chief electrical engineer for Limitorque discussed the detailed technical characteristics of the new system as compared to the original and confirmed the previous approval. Accordingly, N/K-insulated motor leads were supplied with DC motors for Limitorques for nuclear safety-related applications (including those requiring EQ) manufactured from hovember 1984 through December 1985.

Subsequently, two nuclear plants experienced operational failures of Limitorques with DC motors having N/K insulated leads, for which reports were submitted to the NRC under 10 CFR 50.73 and 10 CFR Part 21. The failures were due to electrical shorting of the motor leads together ond/or to ground on the inside of the actuator case.

The shorting was attributed to damage to the relatively fragile N/K motor lead insulation as discussed in IE Infonnation Notice (IN) 87-08.

ORGAN 12AT10h: LIMITORQUE CORPORATION LYNCHBURG, VIRGINIA REPORT INSPECTION N0.: 99900100/85-01 RESULTS: PAGE 12 of 16 Limitorque discontinueo use of the N/K system in December 1985 and a third motor lead insulation design was adopted by Limitorque consisting of Nomex primary insulation covered by a braideo fiberglass jacket with somewhat less epoxy impregnation than the original.

Wyle Laboratories (of Huntsville, AL) performed EQ testing on samples of the N/K design, the new N/ glass design and N/K samples covered with "Insultite" heatshrink sleeving. Two reports were reviewed.

Wyle Nuclear Environmental Qualification Report No. 6110-06, dated February 24, 1987, used the qualification approach of partial type testing (on unaged samples) and analysis. The second report, (Wyle 6110-06-3, Rev. A, dated September 18,1987) described a qualification dpproach based on full aging and type testing. The test program and results would be consistent with 6 10 CFR 50.49 level of qualifica-tion. However, the reports will require evaluation for applicability to the particular plant application and reported performance data evaluated for meeting plant specific functional performance requirements.

12. Mixing of Incompatible Greases in Limitorque Acuators The hRC has received several reports from licensees over the past few years of instances in which small amounts of incompatible greases found mixed with the regular lubricant in the main gear Cases and limit switch gear cases of Limitorque actuators.

Mixing of incompatible greases (those with different bases or mineral soap gelling agents) can result in chemical interaction of the bases resulting in either excessive softening of the grease, hardening of the grease, or separation of the bases and oils.

Secondly, the foreign grease mr.y not be environmentally qualifted and/or could change the radiation and thermal properties of the mix so as to render it unqualified for its application in a harsh environment.

The grease recommended by Limitorque in its maintenance literature for use in the main gear case of actuators in safety-related, in-containment applications is EXXON Nebula EP, grade 0 or 1.

Nebula EP is a calcium (C6) based gelling agent (thickener) grease.

According to Limitorque. Exxon Nebula EP0 was ds used in actuators qualified under their EQ Reports and this is the only grease used by them in actuators supplied for in-containment use of Type SMB, sizes 000 through 4, after mid-1979. Prior to that time, Limitorque had used Sun Oil Co.'s Sun 50 EP in SMB-00s and 000s and Exxon

ORGANIZATION: LIMITORQUE CORPORATION LYNCHBURG, VIRGINIA REPORT lhSPECTION NO.: 99900100/88-01 RESULTS: PAGE 13 of 16 Nebula EP1 in SMB-Os through SMB-4s. When Limitorque converted to EPO for standardizotion, they recomended that their customers do likewise for SMB-000s and 00s, Serial Nos. 295810 and higher, and for SMB-0 through 4, Serial Nos. 302262 and higher. It should be noted that Sun-50 EP has a lithium-lead (Li-Pb) base (thickener) and is highly incompatible with the Exxon Nebula EPO or EPl.

The grease recommended by Limitorque in its maintenance literature for use in the limit switch gear case of safety-related actuators is Beacon 325. NRC IE Information Notice 79-03 recoamended that Mobil 28 be used in applications with relatively high temperature ambient service conditions (over 140 cegrees F continuous) due to the hardening of Beacon 325 over time above this temperature.

During this inspection, Limitorque stated that Beacon 325 was the limit switch gear case grease used in actuators in their EQ testing and was the only grease supplied in safety-related actuator limit switch gear cases unless another lubricant (Mobil 28) was specified by the customer. Limitorque contended that Beacon 325 is acceptable even in higher temperature applications if it is inspected and replaced within the Limitorque recommended intervals. It should be noted that Beacon 325 and Mobil 28 are also considered to be incompatible with each other and mixing the two could result in the problems discussed above, With respect to EQ, Limitorque stated that only those greases recommended and supplied by thcm as mentioned above (not including Mobil 28) were qualified by them and that plants desiring to use other lubricants (1) should ensure that the grease was suitable for use in Limitorques by evaluating it against the criteria listed in their maintnenance literature and (2) would have to establish qualificatiun separately.

13. Unqualified Torque Switches with Brown Insulating Block Material The NRC was recently made aware of unqualified torque switches found in safety-related and qualified SMB type Limitorque actuators at the D.C. Cook nuclear plant. The plant had returned the suspect switches to Limitorque who identified them as the old type of torque switch used in SMA type actuators. During this inspection, Limitorque stated that they had cetermined that some SMA type torque switches had been installed in some early SMB actuators until the inventory of SPA torque switches was expended. The inspector examined the SMA type torque switch that was sent to Limitorque by D.C. Cook onc noteo that it was very different in design from the

ORGANIZAT!ch: LIMITORQUE CORPORATION LYNCHBURG, VIRGINIA REPORT lhspECTION NO.: 99900100/88-01 RESULTS: PAGE 14 of 16 SMB type switch; although it functioned similarly. While the color of the insulators was similar to that of insulators in qualified SMB switches, the different physical appearance should be obvious to plant personnel who have seen both types. Limitorque has comitted to notify their customers that the potential exists that some early procured SMB actuators could contain SMA torque switches and other parts of questionable qualification.

14. Improperly Sized Limitoraue Motor Lead Lugs The River Bend nuclear plant had recently reported finding motor lead lugs (ring-tongue terminals) over-sized for the wire onto which they were crimped. This had resulted in some leads pulling out of the oversized ;rimp sleeve of the lug. The inspector reviewed this issue to determine whether a generic problem with lug sizing existed. Limitorque's records did not contain sufficient detail to determine what size lugs were used in a particular actuator.

However, the wiring standard that had been in use by Limitorque since 1980 clearly specified the correct lug for each size wire and terminal screw size combinatior to be found in Limitorque actuators supplied for safety-related aprlications. The inspector examined the specimen sent to Limitorque from River Bend noting that the lug'a terminal screw opening wa: correct for the installed terminal screws but that the crimp sleeve was sized for a larger range of wire sizes. The inspector looked through the stock of ring tongue termin61s in the assembly / wiring area of the shop and found the correct sizes available for use. The inspector observed production wiring in progress and interviewed technicians and quality control (QC) inspectors finding that the wiring standard was, in fact, being used and that personnel were familiar with its requirements. The wiring standard and QC checklist required verification of correct sizes of wires for lug selection. This had been done in the past by visual identification of the wire and lug size.

Limitorque had prepored a revision to the wiring standard that required the use of wire guages by assembly and QC personnel as an additional control. Limitorque stated that this requirement was to be implemented as soon as the wire guages could be obtained. Review of this issue indicated the controls previously in place should have provided edequate assurance of proper lug sizing and that a generic problem of improper lug sizing does not appear to have occured.

ORGANIZATION: LIMITORQUE CORPORATION LYNCHBURG, VIRGINIA REPORT INSPECTION NO.: 99900100/88-01 RESULTS: > AGE 15 of 16

15. Sand Blasting Beads Found in Lim 1 torque Motors The D.C. Cook nuclear plant had reported finding sand blasting beads or shot (grit) inside motors of safety-related Limitorque Actuators.

The first report of shot was in August of 1987.

The inspector determined that Limitorque's previous method of preparing the surface of actuators for painting was by manual sand blasting. The operator would mask motor openings with tape and use a compressed air gun with a highly directable stream to avoid direct impingement on the tape and thus prevent grit from getting inside.

In March of 1987, Limitorque adopted a new method using a machine that expelled fine shot centrifugally in combination with grit to prepare the motor surfaces, but continued to mask the motors as before. When customers reported finding grit in the motors, it became apparent that taping openings was no longer adequate, so motor end bells were fitted with a protective "can" during the sand blasting that was bolted onto motor flanges; using gaskets for larger size motors. Limitorque informed the Institute for Nuclear Power Operations (IhP0) of this problem and INP0 then issued a notice alerting the inoustry.

Subsequently, an additional problem was found. Shot was getting under motor name plates and coming out in packaging during shipment dnd then getting into the motors. Limitorque resolved this problem by instituting the practice of taping name plate edges in addition to surfaces to keep out the grit. Limitorque informed INP0 of this p roblem.

16. Crimp-Type Wire Joints in Dual Voltage Limitorques Some Limitorques supplied for safety-related applications (some of which require environmental qualification) are fitted with dual voltage AC motors. These motors have nine leads to allow them to be wired for 480 VAC or 240 VAC operation as cesired. Severdi plants hao recently reported finding types of wire joints in the switch compartments of these actuators that were not the type that Limitorque claims were qualified in their tests. Customers were told by Limitorque that only Thomas & Betts (T&B) type RC4 and RC6 wire joints "or equivalent" had been used in the manufacture of these actuators. During this inspection, Limitorque stated that use of the language "or equivalent" was only based on not having completed their review of the issue at the time. They further stated that upon completion of the review, they had determined to the best of their knowledge that only the T&B joints were used.

ORGANIZATION: LIMITORQUE CORPORATION LYNCHBURG, VIRGINIA REPORT INSPECTION NO.: 99900100/88-01 RESULTS: PAGE 16 of 16 The inspector reviewec the procurement records for wire joints and found only records ut purchasing the T&B joints. The inspector interviewed the person responsible for procurement of these components for the last 28 years. He stated that no other type of wire joints had been purchased or used. The wiring standard discussed above specifies use of the T&B joints exclusively for safety-related actuators at the Lynchburg Virginia facility. A type of Panduit joint is used exclusively at the Greensboro, North Carolina facility. The only safety-related actuators made there are SMC-04s, but these are not the ones that have been found in the reported cases. The inspector also examined the stock of wire joints in the shop including bench spares and interviewed shop personnel, finding no evidence that any other type joint was used.

E. EXIT MEETING:

At the completion of the inspection, an exit meeting was held with Limitorque management. The following people were in attendance:

R. Cilimberg, Metallurgical Engineer, USNRC J. Jecobson, Electrical Engineer USNRC B. Milustusky, Quality Control Manager, Limitorque P. G. McQuillan, GA Administrator Limitorque J. C. Land, Director of Manufacturing, Limitorque

1. E. Wilkinson, Vice President, Engineering, Limitorque