IR 05000341/1988025

Insp Rept 50-341/88-25 on 880906-1006.Apparent Violations Noted.Major Areas Inspected:Failure of Recirculation Pump B Discharge Valve to Close on 880820 & 28 & Issuance of Confirmatory Action Ltr CAL-RIII-88-24

ML20206C448
Person / Time
Site:	Fermi
Issue date:	11/08/1988
From:	Danielson D, Eick S, James Smith NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III)
To:
Shared Package
ML20206C431	List: IR 05000341/1988025 ML20206C428
References
50-341-88-25, CAL-RIII-88-24, IEB-85-003, IEB-85-3, NUDOCS 8811160279
Download: ML20206C448 (15)

Text

_ _ _ _ _ _ _ _ _

.

,.

'

, U.S. NUCt. EAR REGULATORY COMMISSION

REGION III

Report No. 50-341/88025(DRS)

Docket No. 50-341 License No. NPF-43 Licensee: The Detroit Edison Company 2000 Second Avenue Detroit, MI 48224 Facility Name: Fermt 2 Nuclear Power Station Inspection At: Fermi Site, Monroe, Michigan Inspection Conducted: September 6 through October 6, 1988 Inspectors:

Q. fa)

S. D. Eick

-

//- 8-88

.l

. . Smith //- /-As Date bM u$ ~

Approved By: D. H. Danielson, Chief #!/ M Materials and Processes Section Date Inspection Summary Inspection on September 6 through October 6. 1988 (Report No. 50-341/88025(DRS))

Areas inspected: Special inspection in response to the failure of Recircuiation Pump B Discharge Valve (B3105F031B) to close on August 20 and 28,1988, and the issuance of Confirmatory Action Letter CAL-RIII-88-02 Results: This inspection identified four apparent violations (described in Paragraph 3) of 10 CFR 50, Appendix B which contributed to the reduced reliability and failure of safety-related motor operated valve These violations were disclosed by the following problem *

MOV Torque Switches were improperly installed (preloaded) because of improperly trained contract personne *

MOV Torque Switches were improperly set as a result of inadequate guidance in defiring acceptable torque setting ranges and inadequate firmness in enforcing control of the setting GB11160279 001109 PDR ADOCK 050 4,1

---

- -------- - - ------- __ _ _ ----- _

_ _ _ - ,

.

,. :

MOV Limit Switches were improperly set as a result of improper guidance in the ' controlling procedure and conflicting objectives caused by sharing a rotor with position light *

The technical problens which reduced the reliability of the motor operated valves were previously discovered and reported on other Deviation Event Reports within the previous yea The licensee reviewed 148 of the 176 safety-related motor operated valves during the investigation effort. This sample was expanded from the original sample of 59 MOV's by including each new group of related or possibly related valves in the population whenever a deficiency was detected in a new set of NOV's. As a result of this conservative orientation, there is a high probability that the full extent of the MOV problem was encompassed by the investigatio _ - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

'

.

'

. .

.

DETAILS Persons Contacted Detroit Edison Company (Deco)

+ S. Orser, Vice President, Nuclear Operations

+ G. Catola, Vice President, Engineering and Service

+ R. Gipson, Plant Manager

'

• L. Goodman, Director, Nuclear Licensing

+P. M. Anthony, Compliance Engineer

• +i'. E. Bailey, Supervisor, Production Quality Assurance

+L. C. Fron, General Superintendent Nuclear Engineering

+C. R. Gelletly, Director, Nuclear Engineering

• +R. B. Stafford, Director, Nuclear Quality Assurance / Plant Safety USNRC Residents

+ Rogers. Senior Resident Inspector

+ Stacek, Resident Inspector

+ Denotes those personnel attending the preliminary exit meeting on September 21, 198 * Denotes those personnel attending the exit meeting on October 6, 1988.

(This meeting updated September 21 inspection results, but disclosed no significant new developments.) Followup on Licensee Event Report (LER)

(Closed) LER 88-032G0, "Recirculation Pump B Discharge Valve Failure to Close," dated September 19, 1988, Background On August 20, 1988, during Startup Testing of the Reactor Recirculation System, the "B" Reactor Recirculation Pump was trippe The "B" loop discharge valve, B3105F031B, which was signaled to close from the Control Room, failed to move to the closed position with the plant at operating temperature and pressure, during test conditions with one recirculation pump running. Subsequent troubleshooting found three loose terminations on the torque switch. Because two of the three loose connections were on the

"close" torque switch, it was concluded, at the time, that the failure was due to loose torque switch wires. After tightening the wires, the valve was tested under static cold conditions during plant shutdown and operated satisfactorily. To detennine if the loose torque switch connections were an isolated case or a generic problem, a sample inspection of 14 valves in the plant was perfonned. Valves in the sample were selected on the basis of similarity to valve B3105F0318 in design, application, and history. The inspection consisted of verifying proper tightness,

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

.

^

, :

,

.

the presence of lockwashers on the torque switch connections and the tightness of all other connections at the valve operato Based on inspection of these valves, it was determined that the loose torque switch connections found on valve B3105F031B were

an isolated case and not a generic problem, On August 28, 1988, plant conditions were established to retest the B3105F0318 valve under dynamic conditions similar to those existing on August 20, 1988. The "B" loop discharge valve again failed to l close. This was the first time since the initial repairs that the

! subject valve was given a signal to close from the Control Room, r while at normal operating temperatures and pressure On August 29, 1988, the plant was shutdown and cooled down,

and further testing was conducted. Tests were performed on valve B3105F031B and its sister valve, B3105F031A, Reactor

Recirculation Pump A Discharge Valve, under both static and dynamic

'

conditions. Both valves opened and closed normally when their associated control room pushbuttons were depressed. Running motor 4 currents and stroke times were normal except that there was no current increase as valve B3105F031B was closed into its valve seat.

'

I On August 30, 1988, a visual inspection was conducted of valve B3105F0318. All wires were tight and all contacts appeared satisfactor It was found that the torque switch was improperly installed and improperly set. The NRC, Region III, issued a

,

Confirmatory Action Letter (CAL-RIII-88-024, dated August 30,1988)

which confirmed that the licensee would conduct a full investigation

,

of the failures end would obtain the Regional Administrators

]

concurrence oefore restarting.

l Motor Operator Valve Analysis and Test System (M0 VATS) was performed j on both recirculation discharge valves. Testing of valve B3105F031B

> confirmed that the generated stem thrust was lower tl:an the specified target thrust and the torque switch was improperly installe The torque switch was replaced and the torque switch setting was j increased to the value specified by Nuclear Engineering (4.75).

MOVAIS testing was re-perfonned and the results were determined j to be acceptabl .

j b. Motor Operated Valve Failure Investigation

,

Investigation of the root causes of the failures of valve B3105F031B was divided into two areas: (1) apparent cause of failure of the valve on August 20, 1988, (loose connections) and (2) confirmed contributing causes of failures of the valve on August 28, 1988, j (improperly installed and improperly set torque switch).

The licensee's investigation of 14 other valves of similar type,

application and history for loose connections was examined by the

!' NRC inspectors and was considered to be appropriate and adequate to assure that loose connections were not a generic problem with this type of valve. No further consideration of the loose connection was necessar __ _ _ _ _ _ _ _ _ _ ___ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ _ __

.

, : i

.

The investigation found the torque switch for 83105F0318 set at 2.00 vs. the manufacturer's recommended setting of 2.75 for the original motor operator. It was subsequently identified that the correct manufacturer's torque switch setpoint is 4.75. The torque switch setpoint should have changed (from 2.75 to 4.75) in 1984 when the motor operator was replaced as a result of environmental qualification concerns. The torque switch was also preloaded i.e.,

it was not properly centered resulting in a preloaded conditio Thus, for a given switch setting it took less dispiecement of the springpack to trip the switch in the closed direction than it did in the open direction. It is believed that the valves' torque switch was incorrectly installed following springpack rework during the Spring 1988 outage, The B3105F0318 valve was prevented from t fully closing under normal operating conditions because the torque switch was incorrectly set. The torque switch being installed in  ;

a preloaded condition also contributed to the proble Based on these findings, the licensee focused the investigation on the following contributing causes of failure:

Improper installation of the torque switch (preloading);

Improper setting of the torque switch (set at 2.00 instead of the prescribed 2.75); and

Improper specification of torque switch setting (old setting of 2.75 was retained when new motor operator of different design was installed. New motor operator required a torque switch setting of 4.75).

In order to determine the extent of the problems and to provide reliable switch settings for the valves, the following short term actions were taken:

Engineering calculated and provided records for minimum and maximum torque switch setting l

A sample population of 59 of the 176 safety-related valves l was selected for document review or field inspectio Docunent review consisted of a comparison of torque switch

'

settings, recorded settings, a list of engineering torque switch setting ranges, recorded M0 VATS settings, and settings [

recorded in the response to IE Bulletin 85-03. A portion of the valves that received a document review were also field i inspected. Field inspection consisted of opening the operator I and visually confirming that the open and close torque switch i settings conformed to the recorded data ond that torque  :

switches were properly installed (not prelodded). (

,

A program was initiated to reset all safety-related torque i switches when the actual setting was t'elow minimum or above  !

maximum settings provided by engineerin ,

5 l

. - - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ .___ __ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

'

.

,.

.

The effects of the short term actions are listed below:

(1) Engineering Validation An engineering list of minimum and maximum torque switch settings for all 176 safety-related motor-cperated valves (M0V) was generated and validated. The validated list  !

resulted in changes to the previously available requirements for 76 valves, some of which required resetting the torque switch as described in short tenn action (3) belo (2) Field Inspection and Documentation Review Field inspection and documentation review was performed on the below-listed categories of MOVs. The inspection included visual examination of actual torque switch settings, proper torque switch installation, and determination of maximum limiter plate size. NOVATS testing was performed on selected valves based on the as-found conditio l

'

32 - Torque Switch documentation was reviewed on 32 MOVs for which M0 VATS testing had been I previously performed in Fall 1987 and Spring 198 t (IE Bulletin 85-03 valves), i 8- MOVs that had the motor operators replaced in 1984 ,

for environmental qualification concerns were inspecte There was a total of 14, but 6 are included in other categorie MOVs were inspected as an additional sample to obtain ,

59 valves for a statistical sample, ,

59 - TOTAL Original Sample (NOTE: 27 required field inspection and testing and 32 required verification by a documentation review.)

Based on discrepancies discovered during inspections and review l of other design changes that could have impacted torque switch  !

settings or installation, the following valves were added to i the sampl ,

10 - Initial inspections found preloaded torque switches in MOVS that had springpack work performed during *

the Spring 1988 outage. All valves in this group l were inspecte I 4- This sample of IE Bulletin 85-03 valves was added to  :

confirm that reported torque switch setting values ,

were valid for valves which were not known to be  !

adjusted, serviced or modified since switches were se l l

l

'

I f

,____-________ _ _ _ _ _ _ _ _ _ _ _

.

,.

.

29 - Review of as-found data on torque switch settings generally showed correlation with previously recorded values, but differences of up to one division were found. All valves for which the previously recorded value was less than one division above the minimum and which were not already included in the inspection or testing program were adde These additions resulted in a total of 70 valves requiring field inspectio Five motor-operated valves were found with the torque switch installed in a preloaded condition. All of these valve operators were previously disassembled to check the springpacks for grease intrusion which required removal and reinstallation of the torque switches. This work was performed by short-term, non-licensed contractor personnel hired to support the 88-01 Spring Outage. The contractor personnel received some general training but no specific training dealing with motor-operated valve maintenance. Since all five of the preloaded torque switches were found in this population, the licensee concluded that no additional inspections were required for preloaded torque switche Although not a part of the original sample, inspections were performed for preloaded torque switches, actual torque switch settings and limiter plate size (when obtainable) when other work was performed inside the M'JV compartmen (3) Adjustments Needed as a Result of 100% Review of Actual v Minimum and Maximum Settings Based on review of recorded information on torque switch t settings of safety-related valves vs. approved minimum and maximums, 42 valves required adjustment of torque switch settings. A total of 26 valves either had their torque switch settings changed to match the minimum / maximum setting :

or were readjusted based on results of MOVATS testing. The effects of actual torque switch settings on past operability are being evaluate (4) Limit Switches During the observati::r. of MOV field inspections, the NRC inspectors raised questions on proper setpoint and tolerances for the limit switches used to provide byoass function fo- the torque switch to allow unseating of the valve. On Aany valves, the same limit switch rotor provides both position indication and the torque switch bypass function. There is a conflict between the criteria of setting the limit switch as close as possible to full open (or full closed) to provide accurate position indication vs. setting the limit switch towards mid-stroke to ensure valve unseating. Approximately 100 valves at Fermi have

- _ _ _

.

,

.  :

!

torque switch bypass and valve position indication functions sharing the same rotor; therefore, this is a definite operational concern. One solution to problems with the two-rotor limit switch design, is to use or install additional rotors to allow separate settings for valve position indication and torque switch bypass. A review of Fermi's valves showed that out of 183 valves,145 already have 4 limit switch rotors and 38 have 2 rotors. Fermi has the opportunity to rewire the majority of the valves should the need arise, l

The NRC inspectors interviews with field personnel indicated that the limit switches were set to the end of travel. It was also clear that the meaning of valve stroke (mechanical v electrical) and how setting tolerances should be applied was not consistent. In response to the deficiencies revealed by the NRC inspectors, the licensee initiated a new investigation into the setting of limit switches. A new definition of mechanical stroke, as the distance between the point of unseating and the backseat was established and tolerances were provided for setting limit switches (first 2-5% of the mechanical stroke). This chsnge affected valve stroke time as measured by valve position indication light The licensee's engineering staff reviewed 183 motor operated valves, including all 176 safety-related MOVs, for valve operating mode and safety function. Forty-two valves were selected for limit switch inspection or testing. The remaining 141 valves were eliminated as a concern for the following reasons:

14 valves were eliminated because they do not nerform an active safety functio *

94 valves were eliminated because they operate using emergency control modes in which the torque switch is bypassed with a limit switch contact for 95% of the j valve stroke in one or both directions.

! *

14 valves were elin.inate.1 because they are not automatic; their function allows ample time for manual operator

'

.

control and they are not located in a potentially harsh environmen *

7 valves were eliminated based on M0 VATS testing result *

12 valves were eliminated because their safety function is to open and a torque switch is not utilized in the opening control circui The testing consisted of monitoring the valve motor current and

limit switch position on a Visicorder or MOVATS trace that provided a plot of these parameters vs. time.

!

1

,- - _ . _ _ _

- - -

,

.

, .-

(5) Anamolies Part of the process for resolving the motor operated valve concerns involved the field adjustroent of the torque switch settings. These settings were adjusted to meet dynamic open/

closure requirements and, where necessary, were validated by calculation by the manufacturer of the actuator. During testing that followed torque switch adjustments, three valves encountered stalling. MOVATS testing showed that target thrusts were substantially exceeded with the prescribed torque switch settings. Proper thrusts were achieved during M0 VATS testing. Yalve operability was confirmed with the new setting >

Limitorque indicated that the most likely cause of this anomaly was the installation of non-standard spring packs. This possibility will be investigated during a future outag When local inspections and rework were performed on a valve operator for any reason during this program, a generel inspection was also performed inside the MOV compartmen Other deviations noted during the general inspection included missing lockwashers on torque switches, loose limiter plate mounting screws, no limiter plate or altered limiter plate installed and low grease level. These isolated deviations were dispositioned through deviation report (6) Maintenance Procedures Maintenance Procedures used for torque switch installation and settings and limit switch settings were reviewed by site, Limitorque and Stone and Webster personnel. Procedures were

,

revised as necessary to reflect the proper installation and

'

adjustment of torque limit switche (7) Training Maintenance and QC personnel were trained, with Limitorque representatives' assistance, in the proper torque switch installation methods and adjustments. Further training was provided in proper verification of limit switch settings by maintenance personne (8) Industry And Site Documentation Review Applicable industry and site documentation was reviewed to identify related experiences. The documents identified are being used to ensure that the long term corrective action addresses all identified problem '

Long Tern Actions A Long Term MOV Action Plan is being established. This plan, which will include Actions (1) thru (6) below, was scheduled to be developed by Nove7ber 4, 190 _____________ __ _________ __ ___ ___ ________ - _ _ _ _ _ _ _ __-_ '

..

.'

..

.

(1) Validation of Setting For Non-Safety-Related MOV's Torque switch settings for non-safety-related MOVs will be reviewed by engineering. A determination will be made of which settings require validation. After validation of these settings, the information will be included as controlled information under the existing design control progra (2) Preventive / Corrective Maintenance Program Preventive / Corrective Maintenance Programs will be reviewed with respect to work activities and controls for MOV Procedures will be revised as necessary to ensure activities which can affect torque switch or limit switch settings are correct and comply with current configuration control practice (3) Post Maintenance Testing The Post Maintenance Testing program will be reviewed and revised as necessary to provide confidence of valve operability after maintenance. Expected completion was Ja'iuary 1,198 Increased emphasis was placed on performing thtse tests with dynamic conditions at normal temperature and pressure when

, possible and/or performing diagnostic testing (Visicorder or M0 VATS).

'

(4) Procedure Review All applicable procedures will be reviewed to assure that MOVs and their settings will be properly controlled from procurement to installation.

(5) Maintenance Training The ongoing training program for maintenance personnel was and will continue to be updated to incorporate applica)le

'

lessons learne (6) Contractor Training Contractors who are nonnally used during an outage and who perform safety-related activities will be task qualifirc through specific training or work under the close supervision of qualified utility personne Comittee Revi :w The Nuclear Safety Review Group reviewed and approved the root cause determination and status of corrective actions on September 2,14 and 23, 1988. The Onsite Nfety Review Comittee reviewed the root cause determination and .orrective actions and concurred with these actions prior to ra Gor startu _ _ - _ - _ _ _ . - _ _ _ _ ___________ ____ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

.

..

c. Nht Analysis of Problems identified in the Investigation (1) Improper Installation of the Torque Switch: One type of torque switch has a shaft projecting from it, with a gear on the end of the shaft. The switches operate by rotation of the shaft in either the clockwise or counterclockwise -

.

direction. Closing torque will cause the shaft to rotate  !

in one direction until it reaches the preset close torque r value and then interrupts the current to the motor. Opening  !

torque will cause rotation in the opposite direction until the preset open torque value is reached and the opening L current is interrupted, i When the torque switch is installed, the gear at the end i of the shaft engages a mating part in the operator and the i switch base is bolted to the operator. If the bolt holes l in the switch base are not aligned with the bolt holes in  !

the operator when the gear is engaged, the shaft should be (

withdrawn and reinserted. Rotation of the switch without  ;

disengaging the gear has the same effect on the torque switch shaft as applying torque to the valve stem: it  :

rotates the shaft relative to the switch set-point. In the extreme case, the switch case can be rotated to the  !

extent that one set of contacts - either open or close -  :

,

can be opened without any torque being applied. In the ,

absence of any torque switch bypass, this will prevent

'

!

operation of this valve in the affected direction.

i Improper installation for the torque switch as described  !

above, results in a condition called "preloading." t Preloaded torque switches were identified in five MOVs '

during this investigation. Two of those valves, B3105F031B and E1150-F028A, were detennined to be inoperable by the l licensee's engineering revie ! .

The licensee concluded that the preloading of the torque i switches occurred during the 88-01 Spring outage when motor operators were disassembled and then reassembled after checks were made on the springpacks for grease intrusio ;

All work done on the valves exhibiting the preloaded [

torque switch was perfonned by contractor personnel hired ,

to support the outage. The contractor personnel who worked on the MOVs and set the MOV switches were not trained by the contractor or the utility to disassemble / i

- reassemble MOVs or set MOV switches. Their qualifications l

'

to perform the MOV work were not reviewed and procurement -

documents for their services included no requirements for -

proficiency or training in MOVs. Work packages provided i'

.

guidance that might have been adequate for trained l personnel, but was not adequate for the contracted  !

personne ;

(2) Improperly Set Toraue Swwhes: Errors in torque switch settings can result in tri pTng prior to completion of a  :

i 11  ;

l

.

__ - ____-____ -_____________ -. _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ -___ _ _ ______ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ . _ _ _ _ _ _

'

.

,.  :

close cycle or in overtorquing or stalling on closure. Errors reported in settings occurred as a result of.several causes:

Torque switch setting dials were never designed to be read closely. Mr. D. S. Warsing of Limitorque indicated that any reading on a large operator should be considered to carry a i i unit tolerance. Smaller units cannot be expected to be set closer than 2 i unit. Two inspectors reading settings t would be expected to record similar differences when L

'

reading the some torque switch settings. Examination of the torque switches by the NRC inspectors confirmed the logic of this positio Graduation markings are aroad and on different levels, making parallax reading errors possibl *

Engineering's list of acceptable torque switch settings

< ranges was originally calculated by Limitorque for euch safety-relattd valve and application during the design and  !

,

construction phase. As the plant moved from construction [

'

through start-up to operating the importance of the list '

seemed to diminish. This was corrected when engineer'ng issued an updated and validated list of torque switch setting ranges in September 198 *

Records were not updated to reflect the changing conditions i occurring when equipment was modified or replaced. Instruc- .

tions for setting torque switches on new equipment were j sufficiently vague so that settings for old units were t reused instead of finding proper settings for the new uni Settings for equipment on hand were corrected in the I recently issued engineering list. Instructions for setting f torque switches on future equipment will require use of this list, which is controlled to prevent use of obsolete

information, o l *

Local Leak Rate Testing (LLRT) contributed to the problem (

in that valves which nearly met LLRT could legitimately

,

have torque switch settings increased within the acceptable I l engineering range without authorization from engineerin !

On several occasions ranges were exceeded and limiter plates were removed without authorization from engineering. On one occasion, the valve manufacturer was consulted and authorized ,

the use of a higher torque setting without going through [

engineering. This was corrected in Procedure H!-E0043, l Revision 5, dated September 13, 1988, by requiring written l concurrence from engineering to change any torque switch  ;

'

settin *

Non-standard spring packs were installed in soue units and I 1 when the prescribed torque switch settings were used, the i M0V's did not perfom properly. When tested by M0 VATS, the j l torque switch values that give the desired results did not  ;

I conform to the calculated values. However, since MOVATS l testing provides actual thrust values, the torque switch  !

settings that satisfy the thrust requirements take l

.

i

I

._____ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _-____________ _

'

.

,.

'

precedence over calculated torque switch setting (3) [rfroperly Set Limit Switches: Improperly set limit switches chu affect MOV application by permitting a torque switch to trip before completion of an open or close operation or by allowing backseating during the open cycl The causes of improperly set limit switches are as follows:

,

The procedure for setting the limit switches (MI-E0043 Revision 4) called for a tolerance of i 4.9%, when limits were set at 5% from closed and 95% open (or 5% from back-seating). This permitted a torque switch bypass setting as low as 0.1% or essentially no br ass on opening. At the other extreme. 99.9% of the valve st.oke could have the torque switch bypassed. This would pemit the valve disk to coast into the backseat of the valve shortly after the motor was tripped by the limit switch. This condition was corrected in Revision 5 dated September i3, 1988, by I redefining the values shown in the "Valve Limit Switch Contact Development" legend of the schematic diagram for each MOV as follows:

Limit Switch Bypass Acceptable Range 0% 2- 5%

5% 2 - 10% ,

'

95% 90 - 98%

100% 95 - 98%

The initiation point for measuring torque switch bypass  :

was defined in Step 7.5.5.c.5 of Procedure MI E0043, f Revision 4 It was based on fully closed positior, rather [

than the position at which the disk begins to ope [

Because of clearance in the valve and operator, this can i easily amount to a difference in excess of E% of stem I travel. Based on using this starting point, the 5% torque switch bypass most frequently used on valves at Femi i (excluding IE Bulletin 85-03 valves) would be ineffective <

becaus' the disk would not clear the seat before the torque  !

switch ypass was opened. This condition was corrected in Revision 5 dated September 13, 1988, by defining initiation [

of disk or gate movement as mechanical 0%.

. Apr-"t . lolations

_

l l CFR Part 50, Appendix B, Criterion III, as implemented by Detroit  !

Edison's Up A ted Final Safety Analysis Report (UFSAR), Section 17. !

requires, in part, that, ". . . Appropriate quality standards are i specified and included in design documents, and that deviations from [

standards are controlled." And furthemore. "Design control measures shall be applied to items such as the following: . . naintenance and i repair; and delineation of acceptance criteria for inspection." i l

I 13

.. ._

c'

'

-

!' ,. .-

~

~

i Contrary to the above, the licensees design control measures failed

to provide a current, controlled list of torque switch setting ranges j and failed to provide suitable guidance to plant maintenance

personnel to maintain M0V's within the specified acceptable torque j twitch setting ranges. The licensee's investigation concluded that i valve B3105F0318 was prevented from fully closing under nonnal

] operating conductions because the torque switch was incorrectly

! set. Any safety-related valve might have been incorrectly set as a result of this deficienc The failure to have adequate design control measures for ensuring proper torque switch settin s is a violation of 10 CFR 50, Appendix B, Criterion III (341/88025-01 .

b. 10 CFR 50, Appendix B, Criterion IV, as implemented by Detroit Edison's UFSAR, Section 17.1.5 requires, in part "Measures shall be established to assure that applicable . . . requirements which are necessary to assure adequate quality are suitably included or referenced in the documents for procurement of material, equipment, and services . . . ."

Contrary to the above, the procurement document for contractor personnel engaged in performing maintenance on safety-related M0V's lacked the requirement for specific training necessary to perform MOV mamtenance properly. As a result of this deficiency five torque switches were improperly installed (preloaded) during the Spring 1988 outage and two of the five valves were rendered inoperabl The failure to specify the requirements necessary to assure adequate MOV maintenance capability in the procurement of contractor personnel is a violation of 10 CFR 50, Appendix B, Criterion IV (341/88025-02).

c. 10 CFR 50. Appendix B, criterion V as implemented by Detroit Edison's UFSAR Section 17.1.6 requires, that, "Activities affecting quality shall be prescribed by documented instructions, procedures, or drawings, of a type appropriete to the circumstances and shall be accceplished in accordance with these instructions, procedures or drawings. Instructions, procedures or drawings shall include appropriate quantitative or qualitative acceptance criteria for determining that important activities have been satisfactorily accomplishe Contrary to the above, the Procedure (MIE-0043. Revision 4), for MOV Electrical Testing prescribed limit switch settings which were incorrect in value, in tolerance, and in reference starting position (zero % open). It also provided no guidance in where to locate torque switch setting ranges, no guidance on how to properly install the torque switch, and no precautions on how to avoid preloading the torque switch. These deficiencies contributed to the preloading of valves, the inoperability of two of those five valves, the inter-mittent operation of one valve and had a potential to affect the operability of every safety-related M0 *

.

,. . . t

.

~

The failure to have appropriate and accurate procedure steps for reassembly and proper torque and limit switch settings for MOVs is a violation of 10 CFR 50, Appendix B, Criterion V (341/88025-03). CFR 50, Appendix B, Criterion XVI, as implemented by Detroit Edison's UFSAR Section 17.1.17, requires, in part, that, "Measures shall be established to assure that conditions adverse to quality, such as failure, malfunctions, deficiencies, deviations, defective i material and equipment, and nonconfonnances are promptly identified and corrected. In the case of significant conditions adverse to quality, the measures shall assure that the cause of the condition is determined and corrective action taken to preclude repetition."

Contrary to the above, the program failed to achieve prompt correction of deficiencies and to preclude repetition of the deficiencie For example:

August 28, 1987, DER No. 87-332: Described Torque Switch With limiter Plate Removed and Torque Switch Setting Exceeding Allowable Rang *

October 7, 1987, DER No. 87.397: Described Error in Torque Switch Settings on M0V's Which Have Operators Replaced

January 7, 1988, DER No. 87.0174 ond February 2, 1988 DER No. 87.0565: Established Need for a Document to Control Torque Switch Setting Ranges

'

January 26, 1988. DER No. 88-0119: Described Valve Inverability Due to Unbalanced (Preloaded) Torque Switc *

March 26, 1988. DER No. 88-0592: Identified Damage to an Operator as A Result of Failure of a Preloaded Torque Switch to Trip at the Intended Thrus *

June 16, 1988 Plant Meeting Minutes: Prescribed Proper Limit Switch "Zero" The failure to take prompt corrective action for previously identified MOV problems is a violation of 10 CFR 50 Appendix B, Criterion XVI (341/88025-04). Exit Interview The Region !!! inspectors met with the licensee (denoted in Paragraph 1)

at the conclusion of the inspection on October 6, 198 The inspectors suNmarized the scope and findings of the inspection ar.d discussed the likely informational content of the report with regard to documents and processes reviewed during the inspection. The licensee did not identify any sur.h documents or processes as proprietary, 15