ML20203A797

From kanterella
Jump to navigation Jump to search
Insp Rept 99900824/86-01 on 860602-06.No Violations, Nonconformances or Unresolved Items Noted.Major Areas Inspected:Differential Pressure Switches,Review of QA Manual & Records Re Mfg of Switches
ML20203A797
Person / Time
Issue date: 07/10/1986
From: Merschoff E, Naidu K
NRC OFFICE OF INSPECTION & ENFORCEMENT (IE)
To:
Shared Package
ML20203A783 List:
References
REF-QA-99900824 NUDOCS 8607170318
Download: ML20203A797 (11)


Text

t -

ORGANIZATION: STATIC 0 RING OLATHE, KANSAS INSPECTION INSPECTION REPORT 99900826/86-01 DATE: 6/2-6/P6 OH-SITE HOURS: 42 NO.-

CORRESPONDENCE ADDRESS: Static 0 Ring ATTN: Mr. R. Johnson, Vice President and General Manager 11705 Blackbob Road Olathe, Kansas 66061 ORGANIZATIONAL CONTACT: Mr. H. Hartman, Manager, Quality Assurance TELEPHONE NUMBER: (913)764-2630 NUCLEAR INDUSTRY ACTIVITY: Current manufacture of products to the nuclear industry is less than 5% of the products manufactured for commercial grade items.

ASSIGNED INSPECTOR: s& abC

/ / v'

/

/w K. R. Naidu, Reactive Inspection Section, (RIS)

/

7O

/l

~0 ate CTHER INSPECTOR (S):

APPROVED BY: s

(' s /

')

9O! O

,f A-

-j

/M

//

E. W. Merschoff, Chief, RIS, Vendor Program Branch Date INSPECTION BASES AND SCOPE:

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

B. SCOPE: Obtain additional information on the differential pressure switches which exhibited a tendency to drif t from the set points at Oyster Creek, and LaSalle County Station, and review the QA manual and representative records associated with manufacture of switches supplied to Oyster Creek and LaSalle County Station.

PLANT SITE APPLICABILITY: Oyster Creek (50-219), LaSalle County Station 182 (50-373,374), Susquehanna (50-387), San Onofre (50-361), Browns Ferry (50-259, 260,269), Sequoyah (50-327, 328), WNPS (50-397), Millstone (50-245), South Texas Project (50-498, 499).

8607170318 860715 PDR GA999 EMVSORNG 99900826 PDR

I 4 ORGANIZATION: STATIC 0 RING OLATHE, KANSAS REPORT INSPECTION NO.- 99900826/86-01 RESULTS: 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. OTHER FINDINGS AND COMMENTS:

1. Background Information Static 0 Ring Incorporated (SOR), located in Olathe, Kansas, manu-factures pressure switches, tempe .Jre switches and differential pressure switches for installation in nuclear power plants and for commercial applications. SOR qualified their switches to IEEE-323-1974, IEEE-344-1975 and NUREG 0588. Switches qualified to these standards are being placed in operation to replace existing nonqua-lified switches.

The following malfunctions of differential pressure switches (DPS) manufactured by SOR were reported:

a. Malfunction Reported at LaSalle On June 1,1986, type 103 AS-B212-NX-JJTT X6 DPSs malfunc-tioned at the Lalalle County Station Unit 2. Specifically,  ;

while operating at 83 percent of full power, the reactor ex-perienced a feedwater transient that resulted in a low reactor water level condition for approximately 1.5 seconds. One of the four DPSs actuated resu'lting in a " half scram".

The reactor scram setpoint set forth in the technical specifi-cation is 11.5 inches. The level channels are normally set _to trip at 13 inches, and the operators are trained to expect reactor scram by the time that the water level reaches 12.5 inches. As the level passed through 10 inches, one of the four reactor scram level switches (the "D" switch) tripped, causing a " half scram" (which as designed did not initiate control rod motion). None of the other three level switches tripped during this transient. No reactor scram occurred during this transient, either automatically or manually.

O

t

)

ORGANIZATION: STATIC 0 RING OLATHE, KANSAS REPORT INSPECTION RESULTS: PAGE 3 of 11 NO.: 99900826/86-01 On Saturday, June 7, after calibrating the Static "0" Ring flow switch which actuates the minimum flow recirculation valve in the high pressure core spray system, the licensee performed a different test using actual system flow. The switch actuated when flow was at 530 gpm instead of 1000 gpm where it had been set. The licensee found similar performance for flow switches in the residual heat removal system. The licensee then consi-dered all Static "0" Ring differential pressure switches suspect and declared emergency core cooling systems in both units to be inoperable on Sunday night. Both units remained in cold shut-down. Subsequently, LaSalle reported the the above test results were not accurate because the test equipment was not calibrated against each other and that the actuation of the minimum flow valve was not electronically recorded.

On January 17, 1986, Oyster Creek nuclear power plant reported .

that, while operating at 98 percent of full power, three out of four reactor low level scram sensing differential pressure switches (CPS), type 103 AS-B-212-NX-JJTTX6 manufactured by SOR, were found to be out of calibration during monthly sur-veillance. Upon discovery, the set points were adjusted to be within acceptable ranges. On January 20, 1986, two of the DPSs were retested and both were found to be within the specifica-tion limits, but one failed to reset when the DPS was valved back onto service. The DPS was determined to have failed and was replaced. The cause of the occurrence has been attributed to instrument drift and the cause of the DPS failure was in-vestigated at the vendor's facility. During the investigation the vendor determined that the Torrington type needle bearing on which the cross shaft rotates exhibited rust on the spring end. The vendor polished the shaft and unsuccessfully attemp-ted to duplicate the failure. The safety significance of this event was considered minimal since the low level alarm was operable and the plant operating procedures require a manual scram if the automatic scrau does not cccur at 128" above the top of the active fuel (TAF). With the instrument drif t the automatic s: rem would have occurred at 136.6" above TAF. The plant reviewed the test and calibration procedure to prcvide a more restrictive tolerance or as-lef t set coints.

l l

l l

1

e

)

ORGANIZATION: STATIC 0 RING OLATHE, KANSAS REPORT INSPECTION NO.- 99900826/86-01 RESULTS: PAGE 4 of 11

2. Review of SOR Quality Assurance Manual.
a. The inspector reviewed the SOR quality assurance manual and verified the implementation of selected criteria including the relevant procedures. The documented QA program meets the requirements of Appendix B to 10 CFR 50. The implementa-tion of the program was selectively verified and determined to be adequate in the areas of design control (criterion III),

procurement document control (criterion IV), document control (criterion IV), identification and control of purchased ma-terial (criterion VIII), inspections (criterion X), test con-trol (criterion XI), control of measuring and test equipment (criterion XII), and corrective action (criterion XV).

b. 10 CFR Part 21 Procedure: SOR Procedure 8303-110, describes the implementation of 10 CFR Part 21. SOR has taken the neces-sary actions in investigating the malfunction of level switches at Oyster Creek and LaSalle. However, SOR has not identified the precise cause of failure.
3. Review of Purchase Orders Received by SOR The inspector reviewed the purchase orders (P0) issued by Commonwealth Edison Company (CECO) and General Public Utilities (GPU) to SOR for the supply of various switches to determine the technical and quality requirements.
a. CECO Purchase Order CECO P0 287842 dated May 16, 1986, to SOR required the supply of 214 pressure switches of various types for installation at the LaSalle County Nuclear Power Station, Units 1 and 2.

Switches were to meet the requirements of Sargent & Lundy (S&L)

(the architect engineer for CECO) specification T-3702, dated September 2,1983, and be qualified to meet NUREG 0588. The P0 required S0R to furnish a Certificate of Conformance (CoC) stating that the switches being provided were identical to those qualified by Acton Environmental Testing Compan.y (AETC) in materials and mechanical features. The switches were to be manufactured under the 50R quality assurance program. The P0  !

stated that Part 21 was applicable.  !

4 J

ORGANIZATION: STATIC 0 RING OLATHE, KANSAS REPORT INSPECTION NO.- 99900826/86-01 RESULTS: PAGE 5 of 11 Change Order B, dated January 17, 1984, to the above P0, required the supply of fif ty two differential pressure switches. All other requirements specified in the original P0 were applicable.

S&L specified the technical requirements including the instrument tag number, the service / design operating temperatures and pressures, and the set point for each switch in data sheets attached to the specification.

b. GPU Purchase Order GPU, owner of Oyster Creek nuclear power station, issued P0 PP-026929, dated July 8,1985, to SOR for the supply of fourteen type 103 AS-B212-NX-JJTTX6 DPSs, three type 103 AS-8202-NX-JJTTX6 DPSs, and several accessories. Data sheets, which specified the technical requirements including the instrument tag number, service / design operating temperatures and pressures, and the set point,for each DPS, were provided to SOR. The P0 required SOR to provide a CoC stating that the switches being provided were identical to those qualified by AETC in materials and mechanical features.

The inspector's review of the quality assurance records to de-termine whether 50R switches met the applicable technical and quality requirements is discussed in the following paragraph.

4. Review of Quality Assurance Records The inspector reviewed the quality assurance records to determine whether SOR met the technical and quality requirements stated in the above paragraph. The records are filed in a job folder referencing the purchase order. Each job folder typically contained the following records:
a. Assembly drawing of tra pressure switch listing the various components,
b. Assembly order for the pressure switch delineating the sequential steps for the assembly. At the completion of each steps a OA inspector inspected and verified the assembly and indicated the acceptability by his signature, including the date.

l l

l l

~

I

I .

J ORGANIZATION: STATIC 0 RING OLATHE, KANSAS REPORT INSPECTION NO - 99900826/86-01 RESULTS: PAGE 6 of 11

c. Hydrostatic test en the switch at 1.5 times the design pressure.
d. Heat stabilization and temperature influence test.
e. Test data sheet indicating the pressure switch was tested at the minimum and maximum range, plant specified set point, heat stabilization test, insulation resistance test on the electrical wiring including the microswitch optical travel, and temperature influence test.
f. Final test certificate providing final test and calibration data on the increasing and decreasing set points and deadband.

The NRC inspector reviewed the contents of the job folders for the pressure switches supplied to the Oyster Creek, and LaSalle Units 1

& 2, and determined that the manufacturer met the requirements of the P0s identified in paragraph 3.

5. Review of P0s Issued by SOR The inspector reviewed ten P0s issued by SOR to various vendors for the supply of various materials required for the manufacture of safety-related switches. P0s to Carpenter Technology, St. Louis, Missouri required various sizes of various types of stainless steel bar material. The applicable ASTM standards were specified. Perma-nent markings of the heat codes on the bar stock were required. A certified material test certificate referencing the heat code sup-plied was required with the shipment. Purchase Orders for the supply of "0" Rings required the results of compression tests performed to ASTM-D-395 method B to be furnished. Parker Seal Company, the manu-facturer of "0" Rings furnished a CoC referencing the batch number of the lot; the result of hardness, tensile, and elongation tests; and the allowable shelf-life.
6. Review of Audits Performed at SOR The inspector reviewed audits performed at SOR by tha following companies:
a. Gulf States Utilities April 23, 24, 1986.
b. Florida Power & Light Company March 11, 12, 1986.

J ORGANIZATION: STATIC 0 RING OLATHE, KANSAS REPORT INSPECTION NO.- 99900826/86-01 RESULTS: PAGE 7 ef 11

c. Commonwealth Edison Company August 14, 15, 1984 and subsequent reaudit on March 7-8, 1985 to verify the adequacy of the cor-rective action taken on the adverse findings identified during the previous audit.
d. General Public Utilities Corporation July 19, 20, 1984.
e. Arkansas Power & Light Company April 26, 27, 1984.

The audits identified no major problems in the SOR'QA program. S0R took corrective action on adverse findings identified during the audits.

7. Discussions with the Manufacturing Representative of the "0" Ring The NRC inspector discussed the properties of the "0" Rings with the manufacturer's representative who visited SOR during the inspection..

The "0" Rings are manufactured by Parker Seals, a division of Parker Hannafin Corporation. Three of them are assembled on the shaft of the DPS. The shaf t rotates on bearings and two "0" rings to transmit the deflection to a lever which actuates a microswitch. The inspector discussed the suitability of the "0" ring on the shaft when exposed to high temperature and radiation, and subjected to 1000 psi static pressure. Specifically, would the "0" rings stick to the shaf t and impede the shaf ts' movement when exposed to stagnant impure water for extended time period without exercise. The Parker Seal representative stated that the "0" Rings are manufactured from a compound known as E-893. While no actual tests were conducted pn Parker E-893-80 seals concerning radiation resistance to 1X10 rads integrated dose, certain inference can be drawn from information available.

Parker Seal group located at Culver City, California, telexed the following information to SOR on June 5,1986.

"Peports K10,063A and K10,063B show results of radiation resis-tance for Parker compounds E740-75 and E515-80 (among others) and drew the conclusion that, within a polymer type, radiation resistance correlates with compression set resistance.

Both Parker compounds E515-80 and E740-75 are being used exten-sively in the nuclear power industry with positive results.

Compression set data on 2-214 0-rings after 70 hours8.101852e-4 days <br />0.0194 hours <br />1.157407e-4 weeks <br />2.6635e-5 months <br /> at 150 C (302 F) show the superiority of E740-75 (15%, report DD3151) over the industrial standard E515-80 (51%, report KK1587),

while E893-80 is nearly comparable to E740-75 (21%, report KT 1714).

-m..-

1 ORGANIZATION: STATIC 0 RING OLATHE, KANSAS REPORT INSPECTION RESULTS: PAGE 8 of 11 NO.- 99900826/86-01 If the assumption concerning the correlation of compression set resistance and radiation resistance is true, then it can be inferred that the performance of E893-80 under radiation should be comparable to that of the recommended E740-75."

The inspector requested the representative to provide additional information on Parker-Super-0-lube which is used to lubricate the area where the "0"-ring is seated on the shaft relative to radiation resistance suitability. S0R stated that they would obtain similar information on Dow Corning Fluid 200 which is used to lubricate the shaft bearings.

8. Results of the Inspection At the inspector's request, a 103AS-B212 type switch was disassem-bled. S0R personnel demonstrated the operation of the switch and the adjustment of the set point. One switch was mounted on the test stand and the calibration of the switch was demonstrated. The inspector decided to investigate the following attributes to deter-mine the potential failure modes:
a. Cescription of Type 103 Differential Switches The type 103 differential pressure switch consists of a dia-phragm which moves a lever that rotates a cross shaft. These components are contained in a steel case designed to withstand system pressure. Both ends of the cross shaft extend out of the wetted volume and out of the case. "0" ring seals prevent leakage of water along the cross shaft. The condition of these surfaces and the condition of the "0" rings will determine the extent to which frictional forces cause a torque which opposes motion of the cross shaf t. When the cross shaft rotates, one '

lever moves a rod axially causing it to actuate a microswitch.

The other lever tears on a helical spring. An adjusting screw is used to change the enmpression of the spring and thus change the setpoint of the differential pressure switch.  ;

The case contains two ports on either side of the pisten or diaphragm. The lower port en ore side is connected to the system reference leg, and the lower port on the other side is connected to the instrument tap. The upper ports on both i

sides are connected to sents which are valved closed when the  !

switch is in service.

e

3 ORGANIZATION: STATIC 0 RING OLATHE, KANSAS REPORT INSPECTION NO.* 99900826/86-01 RESULTS: PAGE 9 of 11 The design of the cavity containing the diaphragm is such that motion of the diaphragm is limited to 0.015 inch. Vost of the time, the diaphragm is against one or the other of the mecha-nical steps which limit motion of the diaphragm. Thus the sum of the unbalanced hydraulic forces across the diaphragm is sup-ported by one stop or the other except when the microswitch is forced to change position. This occurs when the absolute value of the torque caused by the unbalanced hydraulic forces changes from a value less than to a value greater than the torque caused by the helical spring. This movement causes the cross shaft and the levers to rotate 1.8 degrees.

b. The set point of the switch appears to be time dependent. For instance, if a switch is not exercised and is permitted to sit for two months, the set point drif ts. There are two possible contributing factors. In the operation of the DPS, a diaphragm contacts a lever mounted on the cross-shaf t when the differential pressure reaches the set point. The success of the DPS to trip at the set point repeatedly depends on the unimpeded rotation of the cross-shaft. The cross-shaft rotates on two bearings located on either end and two "0" rings which seal the primary coolant and provide the pressure boundary. It is possible that the "0" rings bind or stick to the shaft when lef t unexercised (stagr. ant) over an extended period of time. Parker Seals manu-factured the "0" rings. SOR is investigating this matter with Parker Seals.

Vber. exercised, the DPS trips at the desired set point within the specified accuracy. The inspector observed that during the environmental qualifications, the DPS was exercised prior to calibration and therefore, cannot confirm whether or not this is a contributing factor to the set point drif t problem.

c. Calibration Problems The manufacturer and the power plant utilized different tech-niques to determine the set point. Furthermore, there appeans to be a pressure offset, which is the difference in calibration at atmospheric pressure and application to a system at a pressure of 1000 psig. There also appears to be a buildup of calibration tolerances within an instrument system due to variations in the location of the instrument tap / condensing pot of the reference leg.

1 ORGANIZATION: STATIC 0 RING OLATHE, KANSAS REPORT INSPECTION RESULTS: PAGE 10 of 11 NO.- 99900826/86-01

d. The SOR Vice President issued directives to his personnel to accomplish the following:

(1) The QA manager was to perform 100% inspection of all activ-ities instead of using sampling techniques.

(2) To conduct tests on switches with different springs to determine whether the strength of the spring affects the set point at 60 inches of differential water column.

(3) To simulate plant conditions by exposing the low and high side of the switch to water under pressure for extended periods of time and verifying whether the set point drifted.

(4) Dispatched plant personnel to assist LaSalle County Station personnel in investigating the cause of failure of the switches and developing revisions to the existing calibra-tion procedures.

9. Follow-up Inspection at LaSalle County Station
a. During June 23-26, 1984, the inspector conducted a follow-up inspection at the LaSalle County Station (LaSalle) to review the results of tests performed on SOR switches. On June 24, 1986, LaSalle personnel discussed the results of their investi-gation on the performance of SOR switches which were subjected to calibration verifications at pre-determined intervals. These verifications were performed with a test rig specifically constructed with the assistance of SOR personnel to closely sim-ulate operating conditions. The results of the tests indicated that the calibration technique utilized to determine the Reactor Pressure Vessel (RPV) level trip set point was not compatible with the RPV level set point during operation. Specifically, the calibration procedure did not censider the 1000 psi static pressure effect. Furthermo~re, the condensing pots had not been installed at the same elevation; there was a difference of about 1.5 inches which contributed to the difference in the set points between the level switches in the two trains. The tests determined that the performance of two level switches (instrument tag numbers B21 - No 24D and B21 - No 388) was erra tic. LaSalle decided to disassemble level switch B21 -

No 38B ( ADS level confirmation) on June 25, 1986.

6 .

i ORGANIZATION: STATIC 0 RING OLATHE, KANSAS REPORT INSPECTION RESULTS: PAGE 11 of 11 NO.- 99900826/86-01

b. On June 25, 1986, level switch B21 - No 38B was removed and disassembled. This was a SOR type 103-AS-B212-NX-CIA-JJ-TTX6 differential pressure switch which is intended to confirm to the Automatic Depressurization System that the RPV level is 12.5 inches. A S0R representative disassembled the switch utili-zing an approved procedure. NRC staff observed the following during the disassembly:
1) Some rust was observed in the helical spring housing.
2) Minor scratches were observed on the cross shaft.
3) Foreign material (brown oxide) was observed on the cross shaft at the position of the "0" rings.
4) Some rust was observed in the Torrington type needle bear-ing on the spring side of the switch. The cross shaft ro-tates on the bearing. The snR representative stated that the rust may have restricted the "0FF" travel in the switch due to frictional force.

The disassembled components were taken to SOR, Olathe for further exa:nination.

E. PERSONS CONTACTED:

Static-0-Ring Incorporated

  • R. Davidson, Project Engineer, Nuclear
  • S. Dunlap, Nuclear Sales
  • R. Engel, District Sales Manager L. Gianzer, Engineer, R&D K. Gann, Plant Manager
  • H. Hartman, Quality Assurance Manager
  • R. E. Johnson, Vice President & General Manager
  • J. Peternel, Nuclear Sales
  • R. L. Potter, Assistant Plant Manager
  • W. Priest, Chief Design Engineer Parker Hannifin Corporation D. Boyer, Territory Manager Parker Seals
  • Denotes those persons who attended the exit meeting on June 6,1986.

-