ML20215J959
| ML20215J959 | |
| Person / Time | |
|---|---|
| Issue date: | 04/30/1987 |
| From: | Naidu K, Stone J Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20215J932 | List: |
| References | |
| REF-QA-99900113 NUDOCS 8705080346 | |
| Download: ML20215J959 (11) | |
Text
_.
.a ,
ORGANIZATION: ITT BART0h INSTRUMENTS C0f;PANY
,.. CITY OF INDUSTRY, CALIFORNIA REPORT INSPECTION INSPECTION N0.: 99900113/87-01 DATES: 02/?4-27/87 Ml-SITF H0llRE- ??
CORRESPONDENCE ADDRESS: ITT Barton Instruments Company ATTN: Mr. G. R. Welt, Director Quality Assurance 900 South Turnbull Canyon Road City of Industry, California 91749-1882 ORGANIZATIONAL CONTACT: J. Dwyer TELEPHONE NUMBER: 818-961-2547 NUCLEAR INDUSTRY ACTIVITY: Manufactures pressure switches, pressure transmitters, and analog trip systems.
O I
ASSIGNED INSPECTOR: L fiL 4 Us Y b y) -
K. R. Naidu, Program Development and Reactive Date Inspection Section (PDRIS)
OTHERIhSPECTOR(S):
APPROVED BY:
J/ C. Stone, Ch'ief, PDRIS, Vendor Inspection Branch a/rP/
Dat4 INSPECTION BASES AND SCOPE:
A. BASES: 10 CFR Part 21 and Appendix B to 10 CFR 50.
B.
SCOPE: Review implemeiltation of corrective action taken on noncon-formances identified in Inspection Report No. 99900113/E2-04. Review actions taken to evaluate set poir.t drift experienced at selected plants and review the design and manufacture of transmitters tc ascertain similarities and differences when compared to the product of other manufacturers which have ornarionepr set nnint drift.
3 PLANT SITE APPLICABILITY: All plants using ITT Barton equipment.
87050so346 PDR 870506 GA999 ENVBART 99900113 M PDR
e ,
ORGANIZATION: ITT BARTON INSTRUMENTS COMPANY CITY OF INDUSTRY, CALIFORNIA REFORT INSPECTION NO = 40400113/R7-01 PESULTS: PAGE 2 of 11 A. VIOLATIONS:
No violations were identified during this inspection.
B. NONCONFORMANCES:
No nonconformances were identified during this inspection.
C. UNRESOLVED ITEMS:
No unresolved items were identified during this inspection.
D. ACTION TAKEN ON PREVIOUSLY IDENTIFIED ITEMS OF l'ONCONFORMANCE:
The inspector reviewed the corrective action taken by ITT Barton (Barton) on items of nonconformance identified during an inspection conducted during November 15-18, 1982 and documented in Inspection Report No.
99900113/82-04. Barton outlined corrective actions taken to address the noncompliances in their letter dated March 4,1983 and provided additional information in their letters -dated March 31 and April 11, 1983.
- 1. (Closed) Nonconformance 99900113/82-04-1 The nonconformance identified that the following were contrary to Criterion V of Appendix B to 10 CFR 50 and paragraphs 3.5.6, 4.1, and 5.7 of Barton Engineering Instruction (EI) .No. 0764.1172.2, Revision 004, dated May 19, 1980:
- a. The noise level of the transmitter output had not been logged on the initial calibration sheet.
- b. The zero output (initial or final) had not been recorded on the initial calibration sheet,
- c. Thermal effects had not been plotted on the data sheet.
EI No. 0764.1172.2 has since been revised several times. Current EI 0764-1300.2, dated July 23, 1966, has provisions to record the above.
The inspector reviewed Els 0752-1040.2, 0763-1300.2, and 0763A1149.2 for 752, 763, and 763A type transmitters and determined that they also have provisions to record the above data.
l l
- r ORGANIZATION: ITT BARTON INSTRUMENTS COMPANY
, CITY OF INDUSTRY, CALIFORNIA REPORT INSPECTION Nn . coonn111/R7 n1 RESUlTS: PAGF 3 of 11
- 2. (Closed) Nonconformance 99900113/82-04-2 The nonconformance identified that contrary to Criterion V of Appendix B to 10 CFR 50, and paragraph 5.0 of procedure QU-11, Revision 02, dated June 1, 1982, the documented test results of the "ohmicity tests" identified in the Data Sheets of Test Procedure 0331.1027.2 had not been evaluated to assure that test requirements had been satisfied for strain gage beam assemblies identified in specific data sheets. The nonconformance also identified that QA had not reviewed the temperature compensation data sheets for conformance to requirements prior to final acceptance for inprocess activity related to specific assemblies. The lack of review was evidenced by no indication of QC/QA stamp or initial in the appropriate block of the data sheets. The current data sheets have provisions to record values of the ohmicity test. The circuit board test data sheet is completed when the assembly is initially subjected to temperature cycles. QC is required to sign off on this sheet.
During final temperature compensation test, the entire unit (circuit board and strain gage) is taken through the temperature cycle and final adjustments to the zero and span are made. QC is required to sign off on the final calibration data sheet after verifying that the final temperature compensation test was performed.
- 3. (Closed) Nonconformance 99900113/82-04-03
^
The nonconformance identified that contrary to Criterion V of Appendix B to 10 CFR 50, and the requirements contained in paragraphs 2.0 and 4.0 of Procedure QU-3, Revision 02, dated June 1, 1982, a review was not performed to assure that Barton drawings or specifications met the customer requirements as evidenced by the supply of incorrect pressure transmitters for use at Palo Verde Nuclear Generating Station. Corrective action taken by Barton, is to utilize a design control checklist to route incoming orders. The contracts department reviews the order to ascertain whether the instrument selected meets the customer requirements and whether there are any special requirements. If there are any special requirements, the order is reviewed by application design i engineering personnel. Subsequently the package is independantly verified by personnel in marketing and quality engineering and finally sent to the contract administrator. The inspector reviewed the data packages for orders issued by Fitzpatrick and South Texas Project nuclear power stations and determined that the design control checklist was implemented.
i l
{
l
ORGANIZATION: ITT BARTON INSTRUMENTS COMPANY
. CITY OF INDUSTRY, CALIFORNIA REPORT INSPECTION Nn = CCcDn111/n7 n1 RESULTS: PAGF 4 of 11
- 4. (Closed) Nonconformance 99900113/82-04-04 The nonconformance identified that contrary to Criterion V of Appendix B to 10 CFR 50, Procedure QU-12 and QA Instruction No. QAI 12-07, the hydrostatic pressure gage mounted on the test stand in Departnent 019 had not been calibrated by its due date of October 22, 1982. The inspector reviewed the calibration records and determined that the hydrostatic test pressure gage was calibrated. To prevent recurrence, QC generated a computer list of 4947 instruments with provisions to print out each measuring instrument and it's location (by area) which requires calibration during the following 30 calendar days. The cognizant area supervisor, in whose area the instrument is located, is vested with responsibility to get the instrument calibrated. QC periodically audits the calibration status.
- 5. (Closed) Noncompliance 99900113/82-04-5 '
The noncompliance identified that contrary to Criterion V of Appendix B to 10 CFR 50, and Procedure QU-5, the QA program did not contain documented instructions and procedures to cover the use of the notice of deviation form that was used to document test equipment malfunctions or deviations noted during "inhouse" qualification testing of model Nos. 763 and 764 pressure transmitters.
Barton developed QA Instruction 11-03, titled Notice of Deviation Qualification Testing, to document deviations during tests to qualify instruments. Barton stated in their letter dated March 4, 1983, that the QAI was 9-02. During a subsequent review, QC determined that it was appropriate for this procedure to be in -
Section 11 instead of Section 9.
i 6. (Closed) Nonconformance 99900113/82-04-06 The nonconformance identified that contrary to Criterion V of Appendix B to 10 CFR 50 and paragraph 4.2.2 of QA Instruction No. QAI 06-01, purchase orders (P0) issued for testing services were not entered into the purchase order review log. Barton dispensed with the log book. QA Instruction QAI 04-01, Revision F, dated February 26, 1986 requires all the originals of P0s for nuclear i
I power plant applications including calibration services to be stamped - Nuclear Contract "QA Engineering Review Required." A P0 is coded with a unique number to identify that the order is for nuclear application. When the respective material is received,
1 ORGANIZATION: ITT BARTON INSTRUMENTS COMPANY
, CITY OF INDUSTRY, CALIFORNIA REPORT INSPECTION Nn_- QQQnn113/R7-01 RESULTS- t PAGF 5 of 11 Receiving Inspection retrieves the P0, verifies from the inspection planner whether the referenced code relates to nuclear application.
If the P0 is for nuclear application, Receipt Inspection verifies whether the document was reviewed by QA Engineering. If the necessary reviews were not performed, the received material is required to be quarantined in the designated " Holding Area," pending verification from QA Engineering.
- 7. (Closed) Unresolved Item 99900113/82-04-01 The unresolved item identified that eventhough Engineering Instruction No. 0764.1172.2, Revision 004, dated May 19, 1980, required data recording after 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> at various temperatures, there were no requirements to document the initiation and completion of temperature exposure.
The current procedure 0764-1253.2 dated July-23, 1986, contains provisions to record the initiation and completion of the respective temperature exposure during the temperature compensation tests. The temperature cycle is 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> duration at temperatures of 80 2, 130 2, and 80 2 degrees Farenheit. This provision to document the initiation and completion of temperature exposures have also been implemented for the model 763 and 752 transmitters.
E. INSPECTION FINDINGS AND OTHER COMMENTS:
- 1. Review of 10 CFR Part 21 Report on Model 580 Pressure Switches Barton notified the NRC on April 14, 1986 of two potential defects in their Model 550 series pressure indicating switches.
- a. Specifically, under loss-of-coolant accident conditions, switch malfunctions occurred when the test chamber temperature was raised to 340"F. The symptoms of the malfunctions were:
l
- 1. In three of the five instruments tested, one or both of the switches failed to operate (no change in switch state) when input pressure was varied.
- 2. Switch setpoint shifted in excess of the allowable 10 percent on two of the instruments.
l l
l t
l
s '
ORGANIZATION: ITT BARTON INSTRUMENTS COMPANY CITY OF INDUSTRY, CALIFORNIA REPORT INSPECTION Mn . 490nn113/87-01 RESULTS: PAGE 6 of 11 When the chamber temperature was lowered to room ambient, all of the switches changed state when the input pressure was varied. However, the switch setpoint shifts which occurred at the elevated temperature was permanent. Barton identified that the specific component used in these applications was a Honeywell snap-acting switch, part number 11SM 403. Barton considered this defect generic based on the premise of Honeywell's certificate of conformance that parts, materials and processes used to manufacture these switches were similar to those previously supplied to Barton.
NRC issued Information Notice IN 86-65 alerting owners of nuclear power plants of this problem.
- b. On June 16, 1986, Barton informed the NRC that in addition to the above mentioned problems, Barton determined that there is a deflection of the instrument case due to an increase in test chamber pressure. The deflection affected the position of the switch actuating mechanism and may result in either switch set point drift in excess of 15 percent or no operation of the switch. In this notification, Barton informed the NRC that the problems would be resolved by April 15, 1987.
The inspector discussed the progress of the resolution of the above problems with cognizant engineering personnel and determined that Barton procured a different type of microswitch and subjected the switches to stabilizing. temperature exposure prior to use.
This thermal stabilization process provided stress relief to the switch and it's internal components. The redesigned case, with components, is scheduled for qualification testing from mid June through December, 1987. Aging tests are performed in-house. Radiation, seismic and LOCA tests are done by outside testing laboratories.
- 2. Review of 10 CFR Part 21 Report on Model 580A Pressure Switches On October 11, 1984, Barton notified the NRC of a potential defect in their model 580A, 581A and 583A differential pressure switches.
Specifically, in these types of switches, the setpoint shift exceeded the specification requirements when the instrum mts were operated at temperatures above 225 F. Barton detergined
ORGANIZATION: ITT BARTON INSTRUMENTS COMPANY CITY OF INDUSTRY, CALIFORNIA REPORT INSPECTION Nn - oconn111/97-n1 RESULTS: PAGE 7 cf 11 that the setpoing drift was the result of relieving of stresses in a switch plate assembly which is a subassembly of the complete instrument. Use of stress relieved switchplate assemblies eliminated the setpoint drift. Barton notified the owners of nuclear power' plants where the switches were required to perfonn satisfactorily when the temperature exceeded 225*F that Barton would provide stress relieved switchplate assemblies with installation instructions.
- 3. Review of 10 CFR Part 21 Report on Model 763 Pressure Transmitters On June 30, 1983, Barton notified the NRC of a potential defect in suppressed zero model 763 static pressure transmitters. Specifically, the transmitters exhibited a negative shift in the output during initial exposure to operating pressure. The amount of shift was a function of process pressure and calibrated span of the instrument.
In interim reports dated August 16, 1983; October 14, 1983; November 16, 1983; January 16, 1984; February 7, 1984; and March 12,1984, Barton informed the NRC of progress in the identification of the defect, and corrective action taken to resolve the problem.
Initially, Barton suspected that the cause of setpoint drift in zero suppressed model 763 transmitters was the attachment material
' of the link wire connecting the pressure sensing Dourdon tube with the strain gage beam. Subsequently, Barton determined that the drift was caused by the creep of the link wire connecting the pressure sensor to the strain gage beam. A design change was mad.e to the method of mounting the strain gage beam to control the deflection. The location of the insulating pad, where the strain gage lead wires were terminated, was also charged. The modified design eliminated the setpoint drift. Test results of the modified sensor assemblies and completed transmitters indicate that the
, instruments perform to the published specifications. Barton provided the above information to their customers.
- 4. Review of 10 CFR Part 21 Report on Model 763 and 764 Transmitters On October 29, 1982, Barton informed the NRC of a potential defect in model 763 and 764 transmitters. The potential defect exhibited I itself in the form of nonrepeatability and resulted in performance outside Barton's published specification. Specifically, at elevated temperatures above 130*F the error was measured to be more that the guaranteed 1.3%. At 420*F, the error exceeded 10%. The defect l
affected both zero based and suppressed zero based instruments.
l l
l t
=.* -
ORGANIZATION: ITT BARTON INSTRUMENTS COMPANY
, CITY OF INDUSTRY, CALIFORNIA REPORT INSPECTION en . oqqnn111/97 01 RFSULTS! PACF F of 11 Barton determined that the thennal non-repeatability was caused by the following:
- a. There was electrical leakage through the wiper arms and the shafts of the potentiometer (zero and span) to the instrument Case.
- b. Incorrect temperature compensation calibration methodology was used in-house to calibrate the transmitters.
Barton corrected the above problems by implementing the following:
- a. Installing isolation washers between the potentiometer mounting bracket and the potentiometer to isolate the potentiometer from case ground.
- b. The temperature compensation calibration procedure was revised.
- 5. Evaluation of Problems Identified in Licensee Reports (LER)
The sequence coding and search system (SCSS) data base contains events reported by various nuclear power plants which are documented in LERs. A search of the SCSS yielded 497 LERs related to Barton pressure switches and transmitters during the period 1980 to mid 1986. The abstracts of the LERs were reviewed and the problems identified were discussed with Barton engineering personnel to ascertain whether there were any generic deficiencies with certain instrument models. Results of the discussions relative to each instrument model are documented in suceeding paragraphs.
- a. Model 368 Differential Pressure Transmitter (DPT) l This DPT model was primarily supplied to General Electric Company (GE) for installation in Boiling Water Reactors (BWR). One instance of setpoint drift was identified in 1980 in a BWR j which has been operating since the early sixties.
j Barton engineers were unaware of any major design problems with t
this transmitter model. GE routinely purchased a variety of instruments without stating the specific application. This problem appears to be isolated. The manufacture of this DPT has been discontinued and the DFT has been deleted from current l catalogues.
i 4 .
ORGANIZATION: ITT BARTON INSTRUMENTS COMPANY
/ . CITY OF INDUSTRY, CALIFORNIA REPORT INSPECTION NO .
CConn111/A7-01 RESULTS: PAGE 9 of 11
- b. Model 386 DPT This DFT model was primarily supplied to Westinghouse Electric Corporation (Westinghouse) for installation in Pressurized Water Redctors (PWR). In this model, the electronic components are designed to withstand higher temperatures. Occasional setpoint drift problems have been reported. Barton engineers were unaware of any major design problems. Westinghouse routinely purchased a variety of instruments without stating the specific application.
Barton discontinued the manufacture of this DPT and deleted it from the current catalogues.
- c. Podel 384 DPT The model 384 DPT is similar in construction to the 368 and 386 models. Occassional setpoint drift problems were identified.
Barton engineers stated that it is possible that epoxy bonding of the piezo strain gage to the Kovar beam would yield under stress with elapse of time. Currently, Barton utilizes molecular bonding for attaching strain gages to the Kovar beam.
One LER stated that the license replaced the strain gage, and recalibrated the transmitter. It is Barton's position that instruments with repetitive setpoint drift are to be replaced.
The same corrective action is'also applicable for models 360 and 386. Barton discontinued the model 384 and replaced it with model 752. Westinghouse is the primary user of model 752.
- d. Liquid Level Measuring Sensors Model 352 and 353 liquid level sensors are equipped with pressure sensing devices at one end which transmit the sensed pressure through liquid filled capillary tubes to a remotely located pressure transmitter. The models 352 and 353 are qualified to IEEE-323-74 and IEEE-344-75 standards for use inside containment.
l Some drift problems were identified in PWRs in applications to measure the reactor cavity sump level. Barton engineers stated that problems will arise if the recommended instructions are not followed during the initial installation to hard vacuum fill the capillary tubes. Since these instruments were i
purchased by the NSSS manufacturer and installed under their supervision, they were not informed of the problems.
- J ,
ORGANIZATION: ITT BARTON INSTRUMENTS COMPANY
}ITYOFINDUSTRY, CALIFORNIA REPORT INSPECTION kn - ooonn111/P7-01 RESULTS: PAGE 10 of 11 e.
Differential Pressure Indicating Switches Models 278, 288, 288A, 289, 289A, and 290'A Numerous LERs from different NPPs reported various problems relative to inaccurate alarms, setpoint drift, and switch malfunctions. Barton engineers were unaware of any of the problems because the instruments were purchased by NSSS manufacturers. As such, when problems arise, the NPP personnel contact the HSSS manufacturers for service and spare parts.
The identified problems were discussed with Barton engineers and the causes evaluated. Evaluations indicate three types of reoccuring problems; namely, inaccurate alarm indication, switch failures, and setpoint drifts. Barton suspects that the contacts of the snap action switches were damaged by either high inrush currents when used to energize AC relay coils or by high inductive currents when used to energize DC relay coils. During the replacement of the snap action switches, the actuator arms have to be removed. The actuator arms are reconnected after the replacement and realigned for proper operation. Improper tools and techniques used during this process may result in recurring problems. Barton compiled a switch training course manual which discusses the probable causes, the recommended solution and illustrations of circuit diagrams which will not damage the switch contacts. ,
Barton offers a complete switch plate assembly to replace the ones used in the 228, 288A, 289 and 289A models. The new switch plate assembly permits the replacement of the snap action switch without removal of the actuator arms. In the original switch plate assemblies, the actuator arms had to be removed and realigned during the replacement of the snap action switches.
Barton provides the ITT Barton switch training course manual along with the replacement kit.
F. EXIT INTERVIEW:
The inspector met with Barton representatives mentioned in Section G i
at the conclusion of the inspection and discussed the scope of the inspection and findings.
l
1
- J ,
ORGANIZATION: ITT BARTON INSTRUMENTS COMPANY g CITY OF INDUSTRY, CALIFORNIA l
REPORT INSPECTION i Nn - 99Crn111/E7-01 RESULTS: PAGE 11 of 11 G. PERSONS CONTACTED:
- G. R. Welt, Director, Quality Assurance C. Watson, Manager, Development Engineering
- V. Lawford, Senior Mechnaical Engineer A. Preiser, Senior Mechnaical Engineer K. J. McLean, Contractor Administrator D. Bell, Supervisor, Manufacturing J. Dwyer, Manager, Quality Control D. A. Price, Director of Marketing J. P. Doyon, Manager, Sales and Service P. Elderton, Service Manager
- Denotes persons who attended the exit meeting on February 27, 1987.
s 1.
l 1
___