ML20056D890
| ML20056D890 | |
| Person / Time | |
|---|---|
| Issue date: | 05/27/1993 |
| From: | Jeffrey Jacobson, Potapovs U Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20056D874 | List: |
| References | |
| REF-QA-99900100 NUDOCS 9308180227 | |
| Download: ML20056D890 (16) | |
Text
O a
V wr ORGANIZATION:
Limitorque Corporation Lynchburg, Virginia REPORT NO.:
99900100/93-01 CORRESPONDENCE ADDRESS:
Dr. Ivan E. Wilkinson, P. E.
Vice President, Engineering Limitorque Corporation 5114 Woodall Road Lynchburg, Virginia 24506 ORGANIZATIONAL CONTACT:
Rory D. Segen, Quality Assurance Manager 804/528-4400 NUCLEAR INDUSTRY ACTIVITY:
Motorized valve operators, their replacement parts, and services INSPECTION CONDUCTED:
May 10-14, 1993 27f7y J.B./Ja%ctionSectionBspd', W Leader Date Tean! Irlspp Special Inspection Branch OTHER INSPECTORS:
T. Scarbrough, Hechanical Engineering Branch, Office of Nuclear Reactor Regulation (NRR)
L. Campbell, Vendor Inspection Branch, NRR D. Brewer, endogJnspectionBranch,NRR btk b
6,. 9 3 Uldis Potapovs, Chief l Date Reactive Inspection Section No. 1 Vendor Inspection Branch INSPECTION BASES:
10 CFR Part 21 and Part 50, Appendix B INSPECTION SCOPE:
Evaluate (1) Limitorque's manufacturing and engineering programs for safety-related items including design control, configuration control, and control of sub-vendors, (2) corrective actions associated with 10CFR Part 21 reports, (3)
Limitorque's comercial grade dedication activities, (4) corrective actions regarding unresolved items and concerns from previous inspection reports.
PLANT SITE APPLICABILITY:
Numerous 9308100227 930527 kij PDR GA999 EMVLMIT g
99900100 PDR w
h 9
O
-1 INSPECTION
SUMMARY
-1.1 Violations
-1.1.1 Contrary to the' requirements of Title 10 of the' Code of Federal Reculations (10 CFR) Section 21.21, Limitorque Corporation (Limitorque)-failed to complete its evaluation of the effects that;the relaxation of the actuator spring may have on the operability of motor operated valves (MOVs)
(99900100/93-01-01, see Section 3.2 of this report).
1.1.2 Contrary to the requirements of Section 21.21, Limitorque had not revised its procedures to address certain substantive revisions to 10 CFR Part 21 that became effective on October 29, 1991 (99900100/93-01-02, see Section 3.2 of this report).
1.2 Unresolved Items 1.2.1 Unresolved Item 99900100/93-01-03 (0 pen)
On July 30, 1992, Washington Public Power Supply System (WPPSS) provided preliminary notification of finding Society of Automotive Engineers (SAE)
Grade 1 or 2 screws in an application where SAE Grade 5_ screws were specified.
The screws were used-to attach the housing cap to the motor actuator housing
,on Limitorque's Model SMB-000 motor actuator. The motor actuators were-installed in a safety-related application at Washington Nuclear Plant-Unit 2.
Formal notification, "WNP-2, OPERATING LICENSE NPF-21,10 CFR PART 21 REPORT, MOTOR-OPERATOR CAP SCREW," was dated August 13,-1992.
WPPSS determined that SAE Grade 1 or 2 screws, at minimum published yield strength, could produce a service failure if the motor actuator was set at greater than 94% of the rated thrust.
Limitorque was not aware of this situation _until.two days before the inspection and had not determined the cause, extent of the condition, or preventive action. The NRC will follow the progress of Limitorque's; evaluation of this situation as Unre;olved Item 99900100/93-01-03.
1.2.2 Unresolved Item 99900100/93-01-04 (0 pen)
The NRC will follow the progress of the formal update of Limitorque's inspection plans to address its current' practice for sampling-(see Section 3.4.2 of this report).
1.2 3 Unresolved item 99900100/93-01-05 (00en)-
The 'NRC' will follow the progress of Limitorque's commitment to alert licensees to errors in the'specified value:, f-run and. stall. efficiencies for certain-SMB-3 actuators (see.Section 3.8.1 this report). -
~_:---
x,e o
- 't_
w w
2 STATUS OF PREVIOUS INSPECTION FINDINGS 2.1 Nonconformance 99900100/91-01-02 (Closed)
Limitorque's practice for accepting bar stock, tubing, and plate was reviewed during the inspection and was identified as a significant improvemqnt over its acceptance practice reviewed during the 1991 NRC inspection.
Limitorque's past practice of testing one piece of material from each supplier once a year has changed to testing one piece of certain types of material received from a supplier quarterly or monthly and in certain instances testing each heat / lot of material stock used for critical parts such as motor pinion keys.
Limitorque is developing inspection plans to formally implement this practice.
The NRC will follow the progress of the formal update of Limitorque's inspection plans to address its current practice as Unresolved Item 99900100/93-01-04 (see Section 3.4.2 of this report).
2.2 Nonconformance 99900100/91-01-04 (Closed)
Limitorque's practice of dedicating motors supplied by Peerless-Winsmith was reviewed and found to adequately identify dedication activities performed by each vendor. Although this item was closed during the last inspection, the basis for activities performed by each dedication party was not clearly defined. The inspectors reviewed the basis for dedication activities performed and found them satisfactory to close this nonconformance (see Section 3.4.3 of this report).
2.3 Unresolved Item 99900100/91-01-05 (Closed)
The motor actuator characterization software system is no longer being used to obtain output torque data for actuators. Therefore, this unresolved item is closed.
2.4 Unresolved Item 999000100/91-01-06 (Closed)
Limitorque committed to notify certain NRC licensees of a possible defect concerning the required tension of Reliance motor end bolts. This notification was included in Limitorque Corporation Maintenance Update 92-2.
Therefore, this unresolved item is closed.
2.5 Unresolved Item 99900100/91-01-07 (0 pen)
Limitorque committed to notify certain NRC licensees of a possible' defect concerning improper machining of actuator limit stop housings for HBC-l' actuators. The notification was to have been made by way of Limitorque's maintenance bulletin. The inspectors found that Limitorque had failed to perform this action. This item remains open as Unresolved Item 99900100/91-01-07 until the maintenance bulletin is issued.
2.6 Unresolved Item 99900100/91-01-03 (Closed)
Limitorque failed to complete its evaluation of the effects that the relaxation of the actuator spring pack may have on the operability of MOVs and O
. ~
O incorrectly identified the evaluation as completed. This unresolved item is now being tracked as Violation 99900100/93-01-01 (see Section 3.2 of this report).
3 INSPECTION FINDINGS AND OTHER COMMENTS 3.1 Entrance and Exit Meetinas In the entrance meeting on May 10, 1993, the NRC inspectors discussed the scope of the inspection and established interfaces with Limitorque management.
During the exit meeting on May 14, 1993, the NRC inspectors discussed their findings and concerns with Limitorque's management and staff.
3.2 Implementation of 10 CFR Part 21 Procedures that Limitorque has maintained the required d
The NRC inspectors 6tr,
10 CFR Part 21 postins a procedure for implementing 10 CFR Part 21 requirements, Quality Assut once Procedure (QAP) 13.2, " Reporting of Defects for Safety Related Equipment", Revision 8, dated October 21, 1992.
Limitorque informed the NRC inspectors that they were aware of some of the changes.to 10 CFR Part 21 requirements such as the requirement to file an interim report if an evaluation has not been completed within 60 days, and had attempted to
, incorporate these changes, as they understood them, into QAP 13.2.
Limitorque, however, failed to revise QAP 13.2 to address all aspects of the revision to 10 CFR Part 21, such as defining what is considered objective evidence which demonstrates that the NRC has been adequately informed of a defect or failure to comply.
Limitorque informed the NRC inspectors that they-believed neither an evaluation nor an interim report was necessary if the NRC had been made aware of the issue during discussions at nuclear industry meetings, during undocumented casual conversations with the NRC, or by the issuance of a Limitorque technical bulletin update. The inspectors informed Limitorque that their interpretation of 10 CFR Part 21 was incorrect and that 10 CFR Part 21, Section 21.21, Paragraph (c)(2) requires written notification to the NRC. This notification should include all details as delineated in 10 CFR Part 21, Section 21.21, Paragraph (c)(4).
Limitorque's misinterpretation of 10 CFR Part 21 reporting requirements apparently resulted in the failure to complete several evaluations within 60 days or to submit interim reports to the NRC. The NRC inspectors reviewed Limitorque's 10 CFR Part 21 Log and determined that Limitorque had failed either to complete its evaluations within 60 days or to submit interim reports to the NRC for the following items (Violation 99900100/93-01-02).
Limitorque's evaluation as to reportability is in parentheses following each example.
1.
Limitorque tog No. 08, Spring Pack Relaxation (Not Reportable) 2.
Limitorque Log No. 20, AC Motor Ambient Temperature Effects i
(Reportable) 3.
Limitorque Log No. 28, Spring Pack Curve Data (Not Reportable) )
~ i
.n
w 4.
Limitorque Log No. 37, Worm Gear Failures (Reportable) 5.
Limitorque Log No. 38, HBC Over-Rated When Mounted to SMB (Not Reportable) 6.
Limitorque Log No. 39, Breather / Drain Plug (Not Reportable) 7.
Limitorque Log No. 40, Declutch Lever Seismic Test Reassurance (Not Reportable) 8.
Limitorque Log No. 41, SBD Bolt Thread Engagement (Reportable).
During the conduct of the inspection, Limitorque processed a written notification to the NRC of its failure to comply with the requirements of 10 CFR Part 21 as discussed above, and also issued QAP 13.2, Revision 9, dated May 14, 1993, that incorporated current 10 CFR Part 21 reporting requirements.
The revision to QAP 13.2 identifies acceptable notification procedures and incorporates other current requirements of 10 CFR Part 21 that were missing or incorrectly stated in Revision 8 to QAP 13.2.
Additionally, the NRC inspectors verified that all outstanding Part 21 evaluations have been completed except for the issue concerning spring pack relaxation which is detailed below. These corrective actions and preventive measures that were presented during the inspection were satisfactory.
The inspectors determined that the condition associated with actuator spring pack relaxation (Log No. 08) had been incorrectly identified as closed and that the evaluation report for this condition had not been completed. Spring pack relaxation can occur on stressed spring packs, such as those on normally closed motor operated valves. The result is that for a given torque switch setting, less output torque will be delivered by the motor actuator.
Limitorque's Part 21 Log indicated that the evaluation for this condition was started on August 16, 1988. The Part 21 file for this condition contained test data which indicated that for the samples tested, spring pack relaxation of between 3 and 10 percent could occur over a two year period.
Further review of the data in the evaluation file and discussions with Limitorque revealed that Limitorque had failed to complete its evaluation of the effect that the relaxation of the actuator spring pack could have on actuator torque output. The Part 21 evaluation form had been dated as complete on June 1, 1992, but no evaluation for reportability had been documented, nor was the evaluation form signed.
Based on a review of the technical data which existed in the evaluation file, the NRC inspectors considered that sufficient data existed in June 1992 to determine that this was a reportable condition and hence constituted a violation of 10CFR Part 21 reporting requirements
.(Violation 99900100/93-01-01).
Limitorque committed to complete its evaluation of this issue and make suitable notification under 10 CFR Part 21 by June 15, 1993.
O O
3.3-Limitoroue's Actions Relative to licensee Event Reports (LERs). Part 21 Reports. and Other Reports 3.3.1 LER 275/91-021-00 LER 275/91-021-00, " Failure of Motor Pinion Keys in Limitorque SMB-3-80 Motor Operators Due to Inadequate Design of Material," dated August 28, 1992, was initiated by Pacific Gas and Electric (PG&E) Company as a result of the motor pinion key shearing in a Limitorque SMB-3-80 motor operator (3380 RPM). The i
MOV was installed on the residual heat removal heat exchanger outlet on Diablo i
Canyon Unit 1.
The date of the event was September 16, 1991. On April 29, 1992, PG&E's chemical analysis of the key material identified it as a low carbon, resulfurized, and leaded steel, such as ASTM A-29, Grade 12L13.
The Limitorque-specified key materials have been AISI 1018 for actuator Models SMB-000, SMB-00, SMB-0, SMB-1, & SMB-2, and AISI 4140 for Models SMB-3, SMB-4,
& SMB-5. Over a period of years, sheared motor pinion keys have been reported in various SMB actuators and the cause has generally been identified with the use of low carbon, resulfurized, and leaded steel or a material other than the one specified for the application.
I During an NRC inspection at Limitorque in May 1988, inspectors found, at the time of manufacture of an unspecified number of motor actuators, keys were purchased from a commercial source without certificates of conformance.
In l
. addition, no testing was performed by Limitorque to verify material requirements. On August 30, 1988, Limitorque's inspection procedures were revised to assure receipt of the correct key materials and shipment of keys made of the correct materials.
Existing stocks of keys and key materials were i
scrapped.
Limitorque's current practice for dedicating motor pinion keys is discussed in Section 3.4 of this report.
The NRC has issued three Information Notices regarding sheared motor pinion keys since such failures were first noted:
IN 81-08, " Repetitive Failures of Limitorque Operator SMB-4 Motor-to-Shaft Key;" IN 88-84, " Defective Motor Shaft Keys in Limitorque Motor Actuators;" and IN 90-37, " Sheared Pinion Gear-to-Shaft Keys in Limitorque Motor Actuators."
l Limitorque Maintenance Update 92-2,Section I., Motor Pinion Keys, states, "The material for motor pinion keys for SMB-000 through SMB-2 actuators has recently been changed from an American Iron and Steel Institute (AISI) 1018 steel to AISI 4140 steel." Requests for replacement keys will now be filled with AISI 4140 regardless of operator model. AISI 4140 steel keys and the current dedication methodology used by Limitorque are expected to minimize sheared motor pinion keys. This issue is closed. l l
._ l
w w
3.3'.2 timitoroue's Part 21 Report. Incorrect Screw tenath On October 5,1992, Limitorque reported insufficient thread engagement for screws securing the cover on the SMB to SBD-1 actuator conversion. The design requires the joint to withstand a stall thrust of 2.5 times the standard thrust rating of 45,000 lbs. Actual thread engagement produced a joint theoretically capable of withstanding 2.24 times the standard thrust rating.
In a letter dated October 5, 1992, limitorque notified its custoners that if the subject actuator in a given application has the capability of developing more than 2.24 times the standard thrust, corrective action would be necessary. Corrective action would require the replacement of existing 6.5" screws with 7" screws and the placement of a specified shim under the head of the 7" screws to prevent bottoming.
Limitorque drawing No. 60-021-0050-3, " Disk Spring Housing," for the SMB to SBD-1 conversion, was originally issued December 8,1977. The condition described above has existed since then.
Preventive action was documented in Revision D to the drawing dated July 29, 1992, when the depth of the counter-bore was changed to provide for the proper thread engagement. This issue is closed.
3.3.3 timitoraue's Part 21 Report. Seismic Effect on Declutch Mechanism On December 7,1992, Limitorque notified the NRC of a potential defect in the declutch mechanism for its SMB/SB-00 and SMB/SB/SBD-00 actuators.
Limitorque reported that, when the actuator is vibrated in the vertical axis with sufficient amplitude near the natural frequency of the declutch system, oscillations of the declutch system can cause the motor to become disengaged.
This would result in the actuator stopping its movement of the valve disc during the seismic event.
When the seismic motion ceased, the actuator would return to normal operation.
Limitorque has designed a new declutch lever to eliminate the potential problem for future orders.
In its December 7, 1992, notification, Limitorque recommended that each licensee evaluate the potential for actuator malfunction that would lengthen the valve stroke time during seismic events at the applicable frequencies.
Limitorque indicated that the declutch shaft might not oscillate if the actuator was operating under sufficient load.
Limitorque stated that licensees could purchase new declutch levers to correct the potential problem.
Limitorque recommended that, in the interim, declutch levers be secured to prevent potential movement. The December 7,1992, notification contains an attachment with a list of the recipients. The inspectors considered the December 7, 1992, notification sufficient to alert licensees to this potential problem and to provide adequate corrective action.
i 3.3.4 Texas Utilities Electric Company's Part 21 Report. Motor Deficiency In letters on December 16, 1991, and May 15, 1992, Texas Utilities Electric Company (TV) discussed a deficiency involving the potential failure of Limitorque 80 ft-lb motors to meet rated capacity during maximum expected differential pressure valve operation.
During an NRC inspection at Comanche Peak in August and September 1992 (NRC Inspection Report 50-445 and,
1 O
w w
j 446/92-34), the inspectors found that the licensee could not determine the j
~
exact cause of the motor failures, but that the-licensee believed it was j
related to the testing methodology in that manual control of the test loading-and data recording phase extended the length of time that the motor was exposed to high amperages. The inspectors considered that the licensee had 1
taken appropriate corrective action. The staff considers this issue to be I
plant specific.
i 3.3.5 Limitoroue Technical Update 92-1 In Technical Update 92-1, limitorque states that the results of a study by 1
Kalsi Engineering allowed the SMB-000 to 1 actuators to undergo.140% of their-Li thrust rating for 2000 cycles.
Limitorque also sent a letter to certain licensees with access to the details of the proprietary Kalsi report allowing the actuators to undergo 162% of their thrust rating for 2000 cycles.
In the technical update, Limitorque indicated that the actuator bolts must be torqued to a specific value to allow the increased thrust levels. During the inspection, limitorque personnel stated that the bolt torque provision remained applicable.
Limitorque stated that Kalsi is evaluating the provision on bolt torque and that it will consider modifying the' technical update.when Kalsi's review is complete. Limitorque stated that it was not currently planning to modify the actuator thrust ratings because of the desire to retain margin in the thrust capability of the actuators and to provide a foundation-for exceeding 2000 cycles of actuator operation. Limitorque stated that.new ectuators. are sized with the original thrust ratings.
Limitorque stated that it did not object to establishing a limited number of cycles beyond the Kalsi-1 allowable thrust.
3 3.3.6 Limitoroue Technical Vodate 92-2 On October 9, 1992, Limitorque released Technical Update 92-2, " Recommended Spring Pack ~ Replacement Procedures for Limitorque SMB Actuators," for-the~ use of its customers. The staff had previously reviewed the technical update as Part 21 Log No.92-205. The staff closed its review of the technical update as not a true Part 21 issue (N021) because the update provided guidance for 1
licensees in ordering new spring packs when determined to be needed.
During.
the inspection at Limitorque, the staff inspectors confirmed that Technical H
Update 92-2 was not reportable under 10 CFR Part 21.
3.3.7 Limitoroue Maintenance Update 92-1 The inspectors discussed Maintenance Update 92-1 with Limitorque personnel.-
In a section with the title of " Allowable Overloads of Limitorque SMB Actuator," Limitorque recommended inspection of the actuator if it experienced l
.two and one-half times the thrust rating (or two times the torque rating) or i
if the " overload" occurred more' than once. During the inspection, Limitorque t
('
stated that it defined multiple'" overload" as greater than 140% (or 162% where the Kalsi report is being used directly) of the thrust rating of SMB-000 to 1:
i actuators, or greater than 120% of the thrust rating of other sized actuators.
Li Limitorque also stated that " overload" would include 120% of the torque rating of its actuators.
Later in that section, Limitorque recommended visual inspection of the gearing and worm if " excessive torque" was applied to the a
1 e
O O
actuator. During the inspection, Limitorque indicated that " excessive torque"
. was 120% of the torque rating of the actuator. The inspectors did not consider the issues addressed in the maintenance update to be reportable under 10 CFR Part 21.
3.3.8 Limitoroue Maintenance UDdate 92-2 The inspectors discussed Maintenance Update 92-2 with Limitorque personnel.
Limitorque had evaluated the issues discussed in the maintenance update for reporting under 10 CFR Part 21. The inspectors agreed with the decision by Limitorque that the issues discussed in Maintenance Update 92-2 were not reportable under 10 CFR Part 21.
3.4 Commercial Grade Dedication With the exception of certain motors and electrical wire, Limitorque commercially purchases products used for manufacturing actuators supplied to nuclear power plants.
Limitorque's Quality Management System Manual (QMSM)
Revision 2, dated January 29, 1993, requires that all product lines identified for use in safety-related applications be subject to additional inspections and tests. Cr'tical characteristics for items to be used in safety-related applications are termed attributes and are identified and verified through inspection or test as specified in an applicable inspection plan or procedure.
Satisfactory completion of these inspections and tests form the basis for the efedication of commercially procured products as besic components. At Limitorque, dedication occurs following final acceptance for an item designated for safety-related application at the time the item is certified (Certification of Compliance issuance) by Quality Control.
The NRC inspectors selected the following dedication activities for review to determine if Limitorque's commercial grade dedication activities were being effectively implemented and as a follow up to the 1991 NRC inspection report.
3.4.1 Motor Pinion Keys Limitorque presently uses only AISI 4140 material to manufacturer safety-related motor pinion keys. Limitorque's Inspection Plan (IP) No. 023, "Bar Stock, Tubing Plate," Revision 6, dated October 2, 1991, is the current IP applicable for the receipt inspection of AISI 4140 bar stock used for the motor pinion keys (keys).
IP No. 023 does not reflect Limitorque's current practice for accepting AISI 4140 bar stock. The NRC inspectors were informed that the current practice for accepting AISI 4140 bar stock used for keys will be incorporated in the near future into IP No. 023, and will require that a piece of material from each heat of AISI 4140 bar stock received be subjected to physical testing and chemical analysis. Once the AISI 4140 material heat is acceptable, it is placed in stock. The NRC inspectors reviewed recent documentation and confirmed that these tests and analyses were being performed and were acceptable, even though they had yet to be incorporated into the i
applicable procedures.
AISI 4140 material is withdrawn and sent to a machine shop where the motor pinion keys are machined in accordance with applicable Limitorque drawing
, i i
l i
i l
a
~
S.
w w
s.
e requirements.
IP No. 076, " Keys - Clutch, Motor Pinion, Intermediate,_
Revision 1, dated March 28,~1991, is the current IP applicable for the receipt-1
- inspection of the machined keys.
IP No. 076 does not reflect Limitorque's
-l current-practice for accepting the machined keys. The NRC' inspectors were informed that IP No. 015 will be issued in the near future and will require that each lot of keys machined be' subjected to chemical analysis _and hardness testing on a sample basis. Once the machined keys are acceptable they are l
placed in stock. The NRC inspectors reviewed recent documentation and confirmed that the chemical analysis and hardness tests were being performed and were acceptable.
1 When keys are withdrawn from stock and are to be supplied for safety-related i
applications, the keys are processed in accordance with Quality Control Procedure No.10.5, " Inspection of Safety Related Parts and Orders,"
l Revision 2, dated September 4, 1992.
In. addition to standard visual
'i inspections, hardness measurements are performed on each key. The NRC l
inspectors reviewed recent documentation and confirmed that visual-inspections 1
and hardness tests had been performed on the keys with results acceptable.
i Limitorque does not audit or survey the supplier of the stock material used i
for the keys. However, the tests and chemical analysis performed on each material heat and on the machined key lot, along with the additional 1
inspections and hardness tests-performed on keys designated for safety-related applications, provide reasonable assurance that the keys will perform their l
' intended safety function.
3.4.2 Samplina Bar Stock. Tubina and Plate The 1991 NRC inspection identified that Limitorque's practice for accepting bar stock, tubing and-plate based on testing one piece of material from each l
supplier once a year as an unacceptable sampling. frequency. _ Limitorque.
informed the NRC inspectors that, for _certain critical actuator parts, each material type received from a supplier. may be tested quarterly or monthly,:and
~
in certain instances, such as motor pinion keys, testing.is performed.on.each heat or lot received. On June 10, 1991, Limitorque issued-an internal.
j directive concerning sampling for destructive testing. Also, in Limitorque's.
response to Nonconformance 99900100/91-01-02'on February 25, 1992, Limitorque-t indicated that it would develop and implement a sampling plan for.use in selecting material sampling frequencies by May 29, 1992.
Limitorque informed the NRC inspectors that draft revisions for applicable IPs were still being i
developed and would be finalized in the_ near future to formally implement its current sampling practices as discussed and being implemented-during the.
inspection. The NRC will follow the progress of the formal update of Limitorque's IPs to address its current sampling practices 'as. Unresolved Item 99900100/93-01-04.
3.4.3 - Peerless-Winsmith Motors.
{
~
As a result of the issuance of Nonconformance 99900100/91-01-04. identifying -
l inadequacies during the ' dedication of Peerless-Winsmith electric motors,
' Limitorque implemented IP.No Ill, " Peerless-Winsmith Critical Material-j!
Testing," Revision 0, dated July 15, 1991. During the inspection, the NRC
!q
- {
l I
f d
e rv 1
w w
inspectors determined that the inspection requirements of IP No. Ill, Revision 0, had not been followed. However, the requirements of IP No. 111, Revision Level 1, dated March 25, 1993, were being followed. Revision 0 required Limitorque to perform testing of metallic and non-metallic motor parts, whereas Revision Level 1 only required Limitorque to inspect and test I
non-metallic motor parts. The NRC inspectors questioned the basis for Limitorque not inspecting or testing metallic parts.
Limitorque informed the NRC inspectors that Peerless-Winsmith had resolved past audit findings affecting the acceptance and implementation of its quality assurance program and had provided a detailed basis description for acceptance of metallic parts used in its motors. The NRC inspectors reviewed this basis document (transmittal dated May 7,1993, from Peerless-Winsmith) and IP No. Ill, Revision Level 1, and determined if properly implemented, acceptance methods used by Limitorque and Peerless-Winsmith should provide reasonable assurance that the motors will perform their safety functions.
3.5 Hardness Testino To verify the adequacy of hardness testing procedures and equipment, two motor pinion keys were selected; part number 60-563-0268-1, for Model SMB-1, and part number 60-563-0154-1, for Models SMB-3 & 4.
These keys were AISI 4140 with a hardness requirement in the range of HB 250-345 (Brinell).
The NRC inspectors requested and received a copy of Limitorque's hardness testing procedure, 0C1-10.4, " General Instructions for Using Rockwell Hardness Testers," Revision 0, dated December 3, 1992. The procedure was determined te be in accordance with American Society for Testing and Materials (ASTM) E-18, the standard for Rockwell hardness testing.
The NRC inspectors witnessed a Limitorque QC inspector performing the hardness tests. The above procedure was followed in preparing to perform the Rockwell C hardness test. One hardness indentation was made on each of the selected motor pinion keys. Measured Rockwell C values were converted to Brinell values by the use of a standard conversion chart. The Brinell hardness values so obtained were within the specified range.
The NRC inspectors requested and received a copy of the calibration record for the hardness tester used to perform the above tests.
Examination of the record showed the machine to be within the latest calibration period.
Calibration and routine maintenance had been performed by a contractor from the Wilson Instruments Company, the manufacturer of the hardness testing equipment. The calibration record showed the tester was considered to be in good condition.
Limitorque's hardness testing procedures and equipment met the requirements of their procedures and ASTM E-18.
3.6 Desian Control The inspectors reviewed Limitorque's procedures for performing design changes as delineated in QAP 4, " Design Control Procedure," and QAP 5, " Engineering J e
(7 w
v Drawings and Standards-Issuance and Revision Procedure." The review consisted of both changes made by Limitorque which affect the Limitorque manufacturing process as well as changes made by sub-vendors to actuator components.
QAP 5 requires that a review be performed for the effect of the proposed change on the equipment's nuclear equipment qualification. The proposed change is I
documented on form L147, " Change / General Release Notice." This form contains a i
section which addresses equipment qualification and contains blocks designated as "None," "Not Applicable," and " Qualification Affected." The inspectors noted that the blocks were not filled out uniformly for similar design changes and that QAP 5 was not clear as to the difference between the "Not Applicable" and the "None" blocks.
In no case, however, did the inspectors find where a change that could have affected qualification receive an inadequate review.
{
Limitorque agreed to provide additional guidance as necessary to more l
uniformly implement the intent of QAP 5.
Limitorque's policy in regard to sub-vendor design changes is to discourage any changes which could affect the qualification of equipment. As such, very
{
few changes are made to parts contained in nuclear supplied actuators. The inspectors reviewed one change made concerning SMB-00 torque switches. This 1
change was made as a result of failures experienced with the torque switch roll pins. The change involved upgrading the roll pins to a high alloy steel and increasing the roll pin diameter from 3/32 inch to 1/8 inch. The I
inspectors found the modification process with regard to this change to be well controlled.
3.7 Bases for Motor Ratinas and Motor Data The inspectors reviewed Limitorque's basis for motor data typically supplied by Limitorque to utilities. The data could include motor speed-torque curves, motor power factors, locked rotor torque values, and locked rotor current I
values. The inspectors determined that the motor curves currently supplied by 1
l Limitorque are derived from testing done on one motor of the specific design.
Actual motor performance could vary significantly.
For new motors, Limitorque I
requires the motor manufacturer to perform a locked rotor stall test on each i
motor. This test information is maintained by Limitorque but is not typically I
supplied with the motors. To try to quantify the uncertainty which exists i
l with the motor curves, Limitorque reviewed test data taken from 81 recently I
supplied 10 foot-pound motors of an identical design. The test data indicated that the stall torque for the motors ranged from 10.9 to 12.3 foot-pounds with the majority of the motors exhibiting stall torques of between 11.2 to l
11.6 foot-pounds.
Motor power factors for new motors can be calculated using the locked rotor stall test data and the following equation:
Po m PowerFactor, Vol tsxAmpsx{5 For older motors Limitorque has to retrieve specific motor power factors from i
the motor manufacturer. -
O O
3. 7 '. 2 Motor Ambient Temperature Effects The inspectors reviewed Limitorque's actions with respect to the potential effects that high ambient temperature can have on ac motor performance.
Limitorque has derated its dc motors for high temperature service hut had not evaluated potential ac motor torque reductions due to high temperature conditions. This issue had been discussed previously with Limitorque and has been identified as an open issue by the NRC in our review of utility programs in response to Generic Letter 89-10.
Limitorque had identified the discovery date of this issue as July 27, 1989.
In October of 1992, Limitorque completed testing of five ac motors which indicated that ac motor stall torque output decreased significantly with increasing ambient temperatures. The torque reduction measured ranged from 14 to 25 percent at 336*F. Consequently, Limitorque contracted with the motor manufacturer, Reliance Motor Co., to calculate the expected torque reduction for all Limitorque-supplied ac motor designs. The Reliance data indicated the expected motor torque reductions to vary widely with the worst case being a I
decrease of torque of 30.8 percent at 180*C (353*F).
During the inspection, as a result of the above testing and calculations, Limitorque filed a Part 21 report dated May 13, 1993. The Part 21 report provides the expected motor torque reduction for the numerous Reliance ac l
motor designs. The inspectors found the report to be technically accurate and
' comprehensive.
However, the inspectors expressed their concern regarding the long period of time taken by Limitorque to resolve this issue.
3.8 Actuator Sizino and Toroue Switch Settinos 3.8.1 Basis for Parameters in Actuator Sizino Ecuation Limitorque has a standard practice for sizing motor actuators described in its i
"SEL" documents. Limitorque predicts the torque output of a motor actuator as follows:
Actuator torque - MT x Eff x AF x OAR x DVF where MT = nominal motor starting torque Eff - gear efficiency of the actuator l
AF - application factor OAR = overall actuator ratio l
DVF - degraded voltage factor Motor Toroue In its sizing of motor actuators, Limitorque uses the nominal motor starting torque to deternine the appropriate size for the motor.
In a letter to Cleveland Electric Company on September 17, 1992, discussing a particular containment isolation MOV, Limitorque stated that there is a high probability that at least 110% of the start torque rating would be generally available in a Reliance ac motor as a by-product of the design requirement for high motor speed in revolutions per minute (RPM) at its starting torque.
Limitorque ---
~
T A
a w
further stated that, in a scenario in which the torque switch is bypassed allowing the motor to reach stall, it believed that using 110% for motor starting torque would model actuator capability with a reasonable degree of probability.
Gear Efficiency In its actuator selection data, Limitorque provides a table of pullout, run, and stall efficiencies for various actuators and their overall actuator ratici.
In its Technical Update 92-2, Limitorque states that pullout efficiency is representative of the actuator overcoming inertial loads at start-up.
Limitorque recommends that pullout efficiency be used in the sizing of actuators.
In establishing a value for pullout efficiency, Limitorque assumes that the motor is at half speed and estimates the efficiency of the worm / worm gear interface from guidance in the American Gear Manufacturers Association (AGMA) standard 440.04 (October 1971), "AGMA Standard Practice for Single and Double-Reduction Cylindrical-Worm and Helical-Worm Speed Reducers."
Limitorque then uses engineering judgement to reduce the worm efficiency predicted by the AGMA standard to estimate a pullout efficiency for the actuator.
When establishing run efficiency, Limitorque assumes that the motor is at full speed and determines the AGMA-predicted worm efficiency.
Limitorque then reduces the worm efficiency by engineering judgement to estimate run ef ficiency.
In its letter dated September 17, 1992, to Cleveland Electric, Limitorque stated that, for the containment isolation MOV in that case, the licensee could substitute the run efficiency for pullout efficiency because the application involved a close safety function with no potential for the actuator stopping at any point during the close stroke.
In Maintenance Update 92-1, Limitorque indicates that the stall efficiency is not a true efficiency and is to be used only in overload analysis.
Limitorque estimates an approximate stall efficiency by increasing the run efficiency based on engineering judgement to account for inertia.
During discussions with Limitorque personnel, the inspectors questioned the relative values for run and stall efficiencies as stated in the Limitorque efficiency table for certain SMB-3 actuators.
Limitorque subsequently determined that these efficiencies were reversed in the table.
Limitorque stated that a notice under 10 CFR Part 21 would be submitted to alert licensees to the errors in the table.
(Unresolved Item 99900100/93-01-05).
Limitorque has not performed testing to determine the accuracy of its estimates of gear efficiency. The inspectors noted that significant uncertainties exist in the use of worm efficiency and engineering judgement to predict overall gear efficiency.
In addition, Limitorque does not address the AGMA standard recommendation to establish a thermal power reting limit by testing when determining worm / worm gear efficiency under high ambient temperature applications. n
w w
Apolication Factor In Technical Update 92-2, Limitorque stated that the application factor takes into account variances of the motor starting torque and the pullout efficiency at varying voltage levels (down to 90% of nominal) and various actuator speeds and conditions.
In its sizing criteria, Limitorque provides an application factor of 0.9 for most cases with reduced application factors for more complex conditions. The inspectors found that the application factor is not based on testing, but on the engineering judgement of Limitorque. During the inspection, Limitorque reaffirmed the basis for the application factor and stated that it does not recommend removing the application factor.
Dearaded Voltace Factor The degraded voltage factor is set to one where the minimum voltage at the motor terminals is equal to or greater than 90% of the rated motor voltage.
If the minimum voltage is less than 90% but greater than 70% of the rated voltage, limitorque states that the degraded voltage factor is equal to the ratio of the minimum voltage to the rated voltage for dc motors and to this ratio squared for ac motors.
Limitorque does not have a specific relationship below 70% of the rated voltage. During the inspection, limitorque stated that it based the relationship down to 70% of rated voltage on specifications for MOVs and motor vendor statements, and not on specific testing. With regard to determining the minimum motor terminal voltage, Limitorque stated that a study of power factor was being conducted and should be complete by the end of 1993.
In summary, the inspectors found that the values for individual parameters assumed in the Limitorque criteria for sizing motor actuators are not determined by testing, but are based almost entirely on engineering judgement.
Limitorque has confidence in the sizing criteria as a result of actuators delivering the torque predicted by the sizing criteria during testing at Limitorque and nuclear power plants.
Limitorque stated that it continues to recommend the use of nominal motor starting torque, pullout efficiency, and the application factor in reviewing the capability of its motor actuators.
The NRC staff has confidence in the prediction of output torque by the Limitorque sizing criteria based on the success of the Limitorque criteria for many years in sizing actuators. However, the staff does not believe that confidence exists in the relative values for individual parameters in the sizing criteria.
In other words, one value assumed for a particular parameter may be low and the value for another parameter may be high, but the overall output torque predicted by the sizing criteria may be adequate. Limitorque stated that requests for a Limitorque position on actuator output and other technical issues may be obtained only from the Limitorque Nuclear Support Group and must be confirmed in writing.
3.8.2 Use of Sprina Pack Curves in Settina Actuator Toraue Switches In the past, Limitorque prepared curves that provided an estimate of actuator output torque based on torque switch settings or the displacement of_ the spring pack in the actuator.
Limitorque stated that the upper limit of the curve was typically based on 75 to 80% compression of the spring pack while.-
w l
' the-lower limit represented the pre-load torque of the spring pack.
Limitorque stated that these old spring pack curves were developed by a j
combination of test and analysis.
Limitorque did not have any information on i
the tolerance bands for the curves, but stated that they might be significart, Limitorque is developing new curves to reflect actuator torque as a function of spring pack compression with tolerance bands.
Limitorque also will attempt l
to develop curves of actuator torque versus te que switch setting although l
significant uncertainties exist regarding tor ae switch differences and I
s repeatability.
Limitorque stated that the data should be collected by mid-summer 1993.
Limitorque stated that it is determining a more accurate method to set the pre-load based on spring pack force or deflection using a spring pack tester. This method will require the spring pack tester to be extremely stiff to allow an accurate pre-load setting. Limitorque has not determined the length of time that the calibration of a particular spring pack will remain accurate.
Limitorque has not determined whether the new curves will be applicable to spring packs currently installed in actuators.
4 PERSONNEL CONTACTED
+
Ivan Wilkinson, Vice president, Engineering
+
Rory Segen, Quality Assurance Manager
+
William Miluszusky, Quality Control Manager
+
Frank Napoli, Quality Assurance Engineer i
+
John Franklin, Vice President Finance
+
Pat McQuillan, Nuclear Project Manager Gregory Pence, Chief Engineer Curtis Eshleman, Special Projects Engineer Jesse Puryear, Quality Control Inspector David Page, Quality Control Inspector Hubert Riley, Gage Laboratory Technician Attended the entrance meeting on May 10, 1993
+
Attended the exit meeting on May 14, 1993 i. _ - _ - _
-,