IR 05000382/1993006
| ML20044E181 | |
| Person / Time | |
|---|---|
| Site: | Waterford  |
| Issue date: | 05/03/1993 |
| From: | Westerman T NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV) |
| To: | |
| Shared Package | |
| ML20044E177 | List: |
| References | |
| 50-382-93-06, 50-382-93-6, GL-89-10, NUDOCS 9305240045 | |
| Download: ML20044E181 (16) | |
Text
-
-. -
.-
._
.
.
-..
-.
.
.-
l
-
!
I APPENDIX
U.S. NUCLEAR REGULATORY COMMISSION
REGION IV
Inspection Report:
50-382/93-06
!
t Operating License: NPF-38
Licensee:
Entergy Operations, Inc.
P.O. Box B
'
Killona, Louisiana 70066 Facility Name: Waterford Steam Electric Station, Unit 3
l Inspection At: Killona, Louisiana
,
Inspection Conducted: April 12-16, 1993 Inspectors:
M. Runyan, Reactor Inspector, Engineering Section
'
Division of Reactor Safety R. Vickrey, Reactor Inspector, Engineering Section
Division of Reactor Safety l
!
Accompanying Personnel:
M. Holbrook, Consultant, EG&G Idaho-INEL A. Trusty, Consultant, EG&G Idaho-INEL
,
//
i Approved:
b F~f8 T.'Wes'tfrmali, Chief, Engineering Section Date
.
!
Division of Reactor Safety
Inspection Summarv i
Areas Inspected: Special, announced inspection of the implementation of the licensee's program to meet commitments to Generic 1.etter 89-10, " Safety-
Related Motor-Operated Valve Testing and Surveillance."
'
Results:
'
The licensee's motor-operated valve program was capable of
'
satisfactorily demonstrating the operability of motor-operated valves
subject to Generic Letter 89-10, and fulfilled the licensee's
!
commitments in this area (Section 1)
j Preliminary test results indicated that some of the original design
assumptions were not sufficiently conservative (Section 1.1).
i i
9305240045 93051B
!
I PDR ADOCK 05000382
,
G i
pyg
,
l
. _,
, _ - _
.
_
, _.. _., _.
_.. _.. ~. ~.. _.,.. _,..,.. _. -. ~. _ ~ _.. -.., _ _.
-
,
.
-2-An example was identified where a change to a design-basis calculation
was not reviewed for the effect on current switch settings (Section 1.1).
The licensee had discovered and was in the process of correcting an
error in the methodology of calculating target thrust windows (Section 1.2).
The licensee stated that they will have design engineering act as a
third-party reviewer for the development of or changes to motor-operated valve data record forms effectve no later than June 1,1993. The licensee further committed to have design engineering review all existing motor-operated valve data reports for consistency with the design basis calculations by January 1, 1994 (Section 1.2).
The licensee stated that they will apply a margin for Lorque measurement
uncertainty in response to the inspectors' concern that conservatism in the calculations may not always bound the measurement uncertainty (Section 1.2).
The licensee was in the process of incorporating Version 2.3 of V0TES
diagnostic software into their motor-operated valve program (Section 1.2).
For those valves selected for review, valve factors were generally
greater than 0.5, stem friction coefficients (dynamic) averaged 0.18, and load sensitive behavior averaged 10 percent (Section 1.3 and Attachment 2).
The licensee was using a straight-line extrapolation method that had not
been analytically justified (Section 1.3).
The licensee stated that data from tests conducted at between 95 and
100 percent of design basis conditions would be extrapolkted (Section 1.3).
An apparent error in the marking of a V0TES diagnostic trace was
identified, prompting the licensee to review all traces for similar problems (Section 1.3).
Two V0TES calibration curves were considered questionable because of a
lack of correlation between sensor data (Section 1.3).
The licensee had identified the actuator for Valve MS-401B as being
marginally sized and placed it on a reduced-cycle allowance; corrective actions were planned for the next refueling outage (Section 1.3).
The licensee had taken actions to address pressure locking and thermal
binding which was based in part on calculations. The staff is currently
.
-.
_
- _..
_
~
!
l
-
!
t-3-
!
'
i i
!
t evaluating calculational methods for determining the thrust required to j
overcome pressure locking and thermal binding. Additional inspection of j
this issue is pending completion of the staffs' evaluation
(Section 1.4).
[
The acceptability of the licensee's plan to use static testing for l
periodic verification will require additional justification
~
(Section 1.5).
!
The licensee had made significant progress in establishing a program to
trend motor-operated valve test and maintenance data (Section 1.6).
.
The licensee had satisfactorily addressed weaknesses identified in l
TI 2515/109, Part 1, inspection (Section'l.7).
'
The licensee's plan to test a high percentage of motor-operated valves
!
under differential pressure conditions was considered a strength i
(Section 1.8).
l Summary of Inspection Findinas:
j Inspection Followup Item 382/9306-01 was opened (Section 1.1).
- Inspection Followup Item 382/9306-02 was opened (Section 1.2).
- Inspection Followup Item 382/9306-03 was opened (Section 1.2).
- Inspection Followup Item 382/9306-04 was opened (Section 1.2).
l
Inspection Followup Item 382/9306-05 was opened (Section 1.2).
- Inspection Followup Item 382/9306-06 was opened (Section 1.3).
- Inspection Followup Item 382/9306-07 was opened (Section 1.3).
- Inspection Followup Item 382/9306-08 was opened (Section 1.3).
- Inspection Followup Item 382/9306-09 was opened (Section 1.3).
- Inspection Followup Item 382/9306-10 was opened (Section 1.4).
!
Attachments-Attachment 1 - Persons Contacted and Exit Meeting
Attachment 2 - Waterford Gate Valve Data
.
.
i
-4-DETAILS 1 GENERIC LETTER (GL) 89-10 " SAFETY-RELATED MOTOR-OPERATED YALVE [MOV)
TESTING AND SURVEILLANCE" (2515/109)
On June 28, 1989, the NRC issued GL 89-10, which requested licensees and construction permit holders to establish a program to ensure that switch settings for safety-related MOVs were selected, set, and maintained properly.
Subsequently, four supplements to the GL have been issued and a fifth has been distributed for connent. NRC inspections of licensee actions implementing commitments to GL 89-10 and its supplements have been conducted based on guidance provided in Temporary Instruction (TI) 2515/109, " Inspection Requirements for GL 89-10, Safety-Related Motor-Operated Valve Testing and Surveill ance. "
Instruction TI 2515/109 is divided into Part 1, " Program Review," and Part 2, " Verification of Program Implementation." The TI 2515/109 Part 1 program review was conducted at Waterford 3 January 27-31, 1992, and is documented in NRC Inspection Report 50-382/92-02. The inspection documented by this report was the initial inspection at Waterford 3 under Part 2 of TI 2515/109, and thus, was focused on verification of program implementation. Nevertheless, programmatic issues were addressed during this
.
inspection in response to followup of findings in the Part 1 inspection and in the context of issues that developed in the course of the inspection.
As an overall assessment, the inspectors concluded that the licensee was making a good effort to demonstrate capabilities of the MOVs subject to GL 89-10. The program had successfully implemented the licensee's commitments to the GL.
The principle focus of the inspection was to select and review in depth several MOVs from the GL 89-10 program. The selection was based on an information matrix provided by the licensee in response to a request by the inspectors. The selection was biased toward MOVs that appeared to have less than average margin; otherwise, an attempt was made to select various sizes and differential pressure conditions.
For each MOV selected, the inspectors reviewed the design-basis calculation of the maximum expected differential pressure (MEDP), the sizing and switch setting calculation, the diagnostic test data package, and the diagnostic traces using V0TES 2.3 software. The following MOVs were selected for review:
CVC-183 Volume Control Tank Outlet Isolation Valve MS-401B Emergency Feedwater Turbine Steam Shutoff Valve SI-135B Reactor Coolant Loop Shutdown Cooling Warmup Valve SI-407A Shutdown Cooling Suction Header Isolation Valve SI-502B Reactor Coolant loop Hot Leg Injection Isolation Valve SI-219B High Pressure Safety Injection Discharge Header Orifice Bypass Valve SI-401A Reactor Coolant Loop Shutdown Cooling Isolation Valve
-
.
.
.
-5-All of the selected valves were gate valves with SMB actuators and torque closure control.
1.1 Desion-Basis Reviews The inspectors identified a concern regarding several of the licensee's design assumptions in light of test results that did not appear to substantiate the assumptions. This concern was best typified by a comparison of the assumed and observed values of valve factor and rate of loading for Valve SI-5028.
The valve factor, assumed to be 0.3 for this valve, was calculated to be 0.52 based on the differential pressure test results.
The rate of loading, assumed to be 15 percent, was observed to be 32 percent. Other disparities of less magnitude existed between assumed and observed values for some of the other valves reviewed during the inspection. The inspectors informed the licensee that justification for all design assumptions should be firmly established by the GL 89-10 compliance date (June 1994), and that this justification should be principally based on an analysis of the licensee's test results. This issue was identified as an inspection followup item (382/9306-01). The inspectors observed that, even with the reductions in design margins as the result of testing, the licensee has been able to demonstrate valve operability due to the valve sizing margins utilized in the original design at Waterford.
During dynamic testing of Valve SI-2198, the measured differential pressure was 53 percent greater than the MEDP calculated in the design-basis reviews.
The licensee determined that the original differential pressure calculation was in error due to an incorrect assumption. However, test results indicated that Valve SI-219B had successfully stroked under the more severe differential pressure conditions observed during the test. The licensee intended to revise the design-basis calculation by incorporating the new maximum expected differential pressure.
M0V Design-Basis Review Calculation No. CVCS.001, Revision 3, for CYC-183, dated December 14, 1992, identified the maximum closing thrust allowed for this valve as 8,400 pounds-force (lbf). Two months prior to this revision of the calculation, dynamic testing was conducted that left the total thrust for CVC-183 at 13,274 lbf. The calculation in effect at the time of the testing did not include the 8,400 lbf limit.
In response to the inspectors' concern, the licensee recalculated the maximum allowed closing thrust and concluded that a higher closing thrust value of 17,000 lbf would be acceptable for i
CVC-183. This resolved the concern with the as-left thrust value.
1.2 MOV Sizino and Switch Settina i
The inspectors reviewed the Waterford 3, GL 89-10 MOV Program Plan, Revision 2, thrust calculations, MOV data records (MOVDRs), and the licensee's methodology for detarmination of thrust and torque requirements for the valves listed above.
The licensee's gate valve thrust equation incorporated valve disc friction coefficients ranging from 0.30 to 0.50.
M0V load-sensitive behavior (also
__ _ -
.
.
-6-known as " rate of loading" or "ROL") was addressed by adding a 15 percent margin. A 10 percent margin was added to account for stem lubricant degradation. Stem friction coefficient assumptions ranged from 0.15 to 0.30.
Thrust requirements were adjusted to account for diagnostic equipment inaccuracy and torque switch repeatability. When testing was conducted at
,
less than the MEDP, test results were extrapolated to design-basis conditions.
Thrust requirements were to be revised to reflect the results of dynamic test results. This feedback program was to be implemented once the test data from
>
the previous outage is fully evaluated.
During reviews, the inspectors noted a problem in computing the margin of error for the minimum thrust required to either open or close each M0V.
In these calculations, the licensee had included the V0TES diagnostic system error, torque switch repeatability, rate of loading, and stem lubrication degradation all within a square root sum of the squares (SRSS) computation.
Since the rate of loading and stem lubrication degradation are both biased effects (not +/-), it is not statistically correct to include them in a SRSS computation. The licensee was able to demonstrate that this problem had been previously recognized and that a procedure change had been made to correct the statistical methodology. However, most of the completed test data packages had not been corrected. The overall effect of recalculating the lower thrust margin will be lessened by the fact that some of the previous assumptions were being revised. The assumption for rate of loading was changed from 20 to 15 percent, and the VOTES system error was revised to the published value of 9 percent, down from the previously assumed value of 12 percent (which had been chosen earlier to add additional conservatism to the calculation and bound cases where the VOTES calibration curve did not meet the requirements for the 9 percent error). The licensee's assumed stem lubrication degradation factor of 10 percent may also be adjustable based on as-found test data.
Nevertheless, some of the thrust margins are expected to decrease.
In response to the inspectors' concer.ts, the licensee performed a quick recalculation for the MOVs subject to NRC Bulletin 85-03. This effort showed that the corrected thrust margin for every MOV subject to Bulletin 85-03, was satisfactory. The licensee stated that the thrust margins for the remainder of the MOVs in the GL 89-10 program would be corrected no later than the end of 1993. This issue was identified as an inspection followup item (382/9306-02).
Maintenance engineering used MOVDRs to adjust design-basis thrust requirements for diagnostic equipment inaccuracies, torque switch repeatability, load sensitive behavior, stem lubricant degradation, and torque correction factors associated with the V0TES diagnostic equipment. Design-Basis Calculation EC-92-027 (which addressed several valves including SI-219B)
provided thrust requirements based on both a 0.30 and 0.50 valve factor, but specified the thrust window based on the 0.50 valve factor. During the review of the MOVDR for SI-2198, the inspectors observed that the minimum required thrust based on a 0.50 valve factor was not used. Maintenance engineering had chosen to use the minimum required thrust based on a 0.30 valve factor and had cited a previous revision of the design calculation as interim justification.
The basis for deviating from the thrust window calculated in the latest
.
-7-revision to the design calculation was not documented. However, subsequent dynamic testing indicated that a 0.35 valve factor was appropriate for SI-219B (see Attachment 2) and that the valve was operable. The inspectors were concerned that modifications to the design-basis calculations, as occurred in this case, could result in nonconservative MOV switch settings. The licensee's M0V program was structured to transfer the design authority for each MOV from design engineering to maintenance engineering prior to testing.
Maintenance engineering was responsible for taking valve information from design engineering and developing the diagnostic test thrust and torque margins in the M0VDR document. Generally, the MOVDR was used to incorporate sources of error in a calculation of diagnostic test margins. However, on occasion, maintenance engineering changed some of the basic design assumptions i
(such as valve factor) to enable the creation of a thrust margin large enough to set the torque switch. Design engineering was not necessarily consulted when these changes were made.
In the case of Valve SI-2198, this failure to communicate led to the use of a potentially nonconservative valve factor. The licensee stated that a policy change would be implemented whereby design engineering would be designated as a third party reviewer for the development of or changes to MOVDRs, effective no later than June 1,1993.
,
'
Mr. J. D. Hologa, Principal Engineer, Mechanical / Civil, committed to have design engineering review all existing MOVDRs for consistency with the design basis calculations by January 1, 1994. This issue was identified as an inspection followup item (382/9306-03).
The inspectors observed that several of the valves reviewed did not have limits associated with the maximum allowed opening and closing torque (including inertia). The MOVDR and data sheets were typically marked "N/A" for this limit.
Licensee personnel indicated that this deficiency had been previously identified and was being corrected. The inspectors reviewed newer data sheets where the limits were correctly applied and considered this issue resolved.
During the review of MOVDRs, the inspectors noted that the licensee had not accounted for diagnostic equipment inaccuracies and torque switch repeatability in their torque measurements. The licensee relied on a margin based on the use of pullout efficiency in lieu of running efficiency in the j
closing direction, and the use of a 10 percent margin in excess of the nominal torque rating allowed by Limitorque for inertial torque. The insectors did i
not consider these methods adequate to account for diagnostic inaccuracy and torque switch repeatability.
In response, the licensee committed to account for these uncertainties in their program by the end of 1993. This issue was identified as an inspection followup item (382/9306-04). At the inspectors'
request, the licensee provided an informal estimate of the torque diagnostic measurement accuracy. The inspectors used this information to perform a preliminary review of the potential effect of applying torque meawrement uncertainty to the test data. No instances were identified that would suggest an operability problem existed.
The design basis calculation for Valve SI-407A indicated that the attached 15 foot-pound (ft-lb) motor was marginally sized for the opening direction.
l
I
.
I i
!
-8-l It was necessary to assume an application factor of 1.0 and a stem friction
coefficient of 0.15 to demonstrate design capability. The licensee replaced the 15 ft-lb motor with a 40 ft-lb motor. The inspectors considered the motor replacement to be a positive indication that the licensee was inclined to make
,
'
modifications to marginal MOVs.
The inspectors noted that the licensee was incorporating Version 2.3 of the V0TES diagnostic software. This will require a review of all existing thrust calculations. The licensee stated that they will complete the incorporation
!
of the VOTES 2.3 software into their program by the end of May 1993 and will l
have completed the revision to their design-basis calculations by the end of l
July 1993. This was identified as an inspection followup item (382/9306-05).
.
1.3 Desian-Basis Capability The inspectors reviewed Design Engineering Guide MSPE-I-002, " Guideline for
'
Review of Motor-Operated Valve Test Data," Revision 0, dated September 29,
1992, static test packages, and dynamic test packages for the subject valves.
t Valve MS-401B had been tested under. static conditions only. The dynamic tests
reviewed were conducted under the following conditions.
SI-219B 89 percent of Closing MEDP
,
SI-401A 15 percent of Opening MEDP
!
CVC-183 54 percent of Closing MEDP
,
S1-502B 100 percent of Closing MEDP SI-135B 130 percent of Closing MEDP l
The inspectors reviewed the selected dynamic test data using the industry standard equation, the valve's mean seat diameter, and the dynamic test conditions. This review indicated that valve factors in excess of 0.50
existed for a majority of the selected MOVs (see Attachment 2). Stem friction i
coefficients were also observed from 0.10 to almost 0.30 for the selected MOVs (see Attachment 2).
Based on testing performed to date, the licensee has observed average closing valve factors of 0.50 for gate valves and average
,
stem friction coefficients of 0.15 under static conditions and 0.18 under
'
dynamic conditions.
Load-sensitive behavior has averaged 10 percent (i.e.,
thrust at torque switch trip under dynamic conditions has averaged
approximately 90 percent of the torque switch trip thrust under static
'
conditions).
!
Where testing could not be practically conducted at design-basis conditions,
!
the licensee utilized a straight-line extrapolation of the thrust to overcome
'
differential pressure using the ratio of design-basis differential pressure to the test differential pressure. However, the licensee had not performed t
multi-point testing or developed other methods to justify the long-term acceptability of this extrapolation method. Therefore, the inspectors
.
considered the licensee's extrapolations to design-basis conditions to be the
,
first stage of a two stage approach where the valves have been setup using the
'
best available data, as discussed in GL 89-10. The licensee will be expected i
.
'
to justify its method of extrapolation by the schedcle commitment date for the
,
a
-
,
, - - -.
,-n..
--
.
,, - - -,,,
.,----,,,n
,.,,,--m-..wm
.,
s-,.,-
e
,ne-
_ _ _ _ _ _ _ _ _ _ _ - _
.
.
_g.
completion of their GL 89-10 program. This issue was identified as an inspection followup item (382/9306-06).
Section 6.1.2 of Design Engineering Guide No. MSPE-I-002 considered a dynamic test conducted at 95 percent of the design-basis conditions as being equivalent to a 100 percent design-basis test. Under this provision, the licensee did not consider extrapolations of this magnitude to be necessary.
This provision was based on " current industry experience." However, the licensee did not present justification that this method would ensure design-basis capability.
In response, the licensee agreed to extrapolate data for tests conducted between 95 and 100 percent of the MEDP. The licensee stated that the application of this policy to its current test results will not result in any operability problems.
The licensee was able to test Valve SI-401A at 15 percent of the opening MEDP, but without any appreciable flow. No differential pressure or flow could be achieved in the closing direction. The original thrust requirements assumed a 0.30 valve factor. Test results indicated that the opening valve factor was
.
'
0.31.
Based on questions from the inspectors regarding the test data, licensee personnel determined that a 0.30 valve factor assumption was not reasonable for design-basis conditions.
However, the licensee reviewed the design-basis calculations and determined that the assumptions regarding the
,
worst-case scenario were overly conservative for MEDP. The MEDP revision will allow the licensee to set up Valve SI-401A using a 0.50 valve factor.
The inspectors identified an apparent error in the marking of the dynamic i
VOTES diagnostic trace of Valve SI-5028.
Point C10, flow cutoff, was marked prior to an area of large fluctuations in measured thrust output. These accentuated fluctuations are typically associated with the stress of
,
overcoming the fast, turbulent flow that develops as the valve disc reduces i
the flow area.
In fact, flow cutoff (C10) for this valve was marked at a thrust output level that was less than the running load near the end of the stroke (CS). Since the running load is still present at the time of flow cutoff, the thrust at flow cutoff would not be less than the running thrust.
Had Valve SI-502B not been tested at greater than 100 percent of the MEDP, the as-left CIO mark could have led to a nonconservative extrapolation of thrust requirements. The licensee's extrapolation methodology used the thrust difference between C10 and the average running load multiplied by the inverse
'
of the percentage of MEDP attained during the test. The licensee stated that the original C10 mark for Valve SI-502B had been changed based on advice received from a contractor, who recommended that the point be marked at a plateau region in the trace of springpack deflection, as measured by the V0TES Torque Cartridge (VTC).
The licensee agreed that the C10 mark for Valve SI-502B was questionable and committed to review other dynamic traces for similar anomalies. The licensee also stated that an individual or group of individuals from the plant staff would attend a course in advanced VOTES signature analysis by the end of July 1993. This issue was identified as an j
inspection followup item (382/9306-07).
'
'
j
- -
.
.
.
-10-
>
The inspectors questioned the validity of the VOTES force sensor (VFS)
calibration for tests conducted on Valves SI-1358 and SI-401A.
In both instances, the VFS trace did not match the calibrator trace for a portion of the calibration stroke. Despite having an acceptable statistical analysis.as computed by the V0TES software, the lack of correlation between the two traces may render the calibration invalid at least in the area on the nonconformity.
The licensee stated that this situation would be investigated and that any necessary corrective actions would be taken. This issue was identified as an inspection followup item (382/9306-08).
The inspectors noted that there was a lack of attention to detail during post-test evaluations for Valve SI-2198. When determining the actual closing valve factor, the licensee had incorrectly used the measured running load value for the opening direction. The reverse error was made for the calculation of the actual opening valve factor.
Further, the reviewer compared the results to a valve factor that was different from the one used in the MOVDR. The inspectors noted that the licensee had identified these errors and had made the necessary corrections prior to the inspection.
During review of the static test package for Valve MS-401B, the inspectors observed that this MOV was left with thrust at control switch trip (CST)
67 lbf below the minimum required thrust. This was justified based on the conservatism inherent in the thrust calculation.
Further, the measured final torque was 10 ft-lb over the actuator rating of 250 ft-lb. This was determined to be acceptable based on Limitorque's guidance allowing peak torque up to 110 percent of the actuator's rating. However, the licensee's torque measurements were not adjusted for diagnostic equipment inaccuracy.
In response to the inspectors' questions, licensee personnel determined that an incorrect torque correction factor (TCF) had been used for the thrust
'
measurements. Applying the correct TCF resulted in a 2,622 lbf increase in
,
the measured thrust, which removed the concern regarding the available thrust
'
at CST. Torque measurements were adjusted to account for equipment inaccuracy i
and torque switch repeatability.
Final torque was determined to be 2 ft-lb (0.7 percent) above 110 percent of the actuator's torque rating (275 ft-lb).
Torque at CST was 4 ft-lb (1.6 percent) above the nominal actuator torque rating of 250 ft-lb. This was justified, for the interim, based on l
Limitorque's peak torque allowance up to 120 percent of the actuator's torque rating for a maximum of 100 cycles. The licensee was able to demonstrate that the valve had not exceeded this number of cycles in its current condition and would remain below the cycle limitation until the next outage.
In light of the revised values for thrust, the licensee intends to lower the current torque switch setting at the next available opportunity. The inspectors considered this another example where inattention to detail resulted in a degraded condition. The staff will review the licensee's corrective actions for Valve MS-401B during future inspections.
This issue was identified as an inspection followup item (382/9306-09).
The acceptance criteria for diagnostic test results did not require an engineering evaluation of trace abnormalities, if present.
Such abnormalities may indicate mechanical or electrical problems, even for cases where thrust i
.
.
-11-and torque outputs were acceptable and the valve successfully stroked. The licensee stated that trace abnormalities would be identified and evaluated as part of a final assessment of the test data, but that this effort would not necessarily take place befure the valve is returned to service. The inspectors considered this practice to be acceptable, though not optimal, since it could place the licensee at risk if abnormal conditions are discovered at a later date.
1.4 Pressure lockina and Thermal Bindino The Office for Analysis and Evaluation of Operational Data (AE00) has
,
completed a study of pressure locking and thermal binding of gate valves.
AE00 concluded in its report that licensees have not taken sufficient action to provide assurance that pressure locking or thermal binding will not prevent a gate valve from performing its safety function. The NRC regulations require that licensees design safety-related systems to provide assurance that those systems can perform their safety functions.
In GL 89-10, licensees were requested to review the design bases of their safety-related MOVs.
The inspectors reviewed the licensee's evaluations of the potential for pressure locking and thermal binding of gate valves. Where such a potential was identified, the inspectors reviewed the actions the licensee has taken to prevent pressure locking and thermal binding. The licensee had completed a review of their valves, which are subject to thermal binding and could be expected to open in the event of an emergency. All gate valves contained in the MOV program were researched to determine their susceptibility to pressure locking and thermal binding. The results were documented in an evaluation entitled, " Gate Valve Thermal Binding and Bonnet Over Pressurization Review."
Of the 29 valves studied, 8 were determined to be potentially susceptible to pressure locking, and 8 were determined to be potentially susceptible to thermal binding. The valves that were potentially susceptible to pressure locking were further evaluated in Calculation MC-M93-005. All of the valves evaluated by the licensee for pressure locking had required opening thrusts
.
that were calculated by the licensee to be less than the thrust capability at
'
reduced voltage. Therefore, these valves were not a concern for pressure locking. The valves that were susceptible to thermal binding were further
evaluated by the licensee and found to have sufficient thrust margins to ensure opening capability.
.
Calculational methods for determining the required thrust for overcoming i
pressure locking and thermal binding is currently under staff review.
Therefore, evaluation of the licensee's calculational methodology is considered to be an inspection followup item (382/9306-10).
1.5 Periodic Verification of MOV Capability In GL 89-10, the NRC requested that licensees prepare or revise procedures to ensure that adequate MOV switch settings were determined and maintained throughout the life of the plant. The NRC recommended that the surveillance interval be based on the safety importance of the MOV as well as its i
__
___
_.
__
.
.
..
__
_ __
.
.
t-12-
}
l l
maintenance and performance history, but that the interval not exceed 5 years
,
or 3 refueling outages.
Further, the capability of the MOV will need to be j
verified if the M0V is replaced, modified, or overhuled to an extent that the existing test results would not be representative of the MOV.
I During the GL 89-10 Part 1 inspection, the licensee indicated that it would
attempt to rely on static tests of MOVs to periodically verify MOV design i
basis capability.
For this plan to be accepted, the licensee will need to
,
provide justification for the use of data from static tests to predict the
,
performance of MOVs under design basis conditions. The licensee stated in its l
program plan that periodic verification would be performed every other refueling outage which is consistent with the frequency recommended in
!
GL 89-10. Because the licensee will not have to perform periodic testing for some time, their action in this area has been limited to monitoring NUMARC and j
EPRI efforts to develop a model for static periodic testing.
j
.
1.6 M0V Failures. Corrective Actions. and Trendino j
In GL 89-10, the NRC requested that the licensees analyze or justify each MOV failure and corrective action. The documentation should include the results and history of each as-found deteriorated condition, malfunction, test, inspection, analysis, repair, or alteration. All documentation should be
,
retained and reported in accordance with plant requirements.
It was also
suggested that the material be periodically examined as part of the monitoring and feedback effort to establish trends of MOV operability. _These trends
,
could provide the basis for a licensee revision of the testing frequency
!
established to verify M0V capability on a periodic basis. The GL indicated that a well-structured and component-oriented system would be necessary to track, capture, and share equipment history data.
During the GL 89-10 Part 1 inspection, the licensee was using existing plant programs to identify and correct M0V failures. Furthermore, the licensee was
,
in the process of improving its trending program for MOVs. During this
_
inspection, the inspectors reviewed the licensee's progress in developing a l
trending system for the GL 89-10 valves. The inspectors reviewed l
Procedure UNT-005-024, "MOV Testing, Maintenance and Trending Program,"
l Revision 1, and MD-001-031, "MOV Setting, Signature Analysis and Trending l
Evaluation," Revision 1, and Design Engineering Guide MSPE-I-002, " Guideline i
for Review of Motor Operated Valve Test Data," Revision O.
These procedures
'
contained steps, which shoulr$ result in improvements, in the licensee's M0V
-
l trending program. Additionally, the licensee had established a database of
!
work history and test data. The licensee was still in the process of i
developing a trend program to interface with the historical database.
!
Although this program was not complete, the inspectors noted that the licensee
'
had made significant progress in its development and implementation.
1.7 Followuo of Weaknesses and Observations Enumerated in the Part 1 Report l
During the Part 1 inspection (NRC Inspection Report 50-382/92-02), the inspectors had noted that the licensee had initiated a comprehensive program j
l
'
'
.
_
- -.
._
'
.
"
'
-13-I for MOVs that generally met tne recommendations and intent of GL 89-10.
However, some weaknesses were identified in the licensee's program for implementing the commitments to GL 89-10.
The licensee had documented these findings in a commitment information report that assigned priorities and responsibilities for corrective actions. The followup of these findings have been addressed in the following discussion.
The Part 1 inspectors noted that the scope, as defined in the program plan, could be interpreted to be inconsistent with the licensee's commitment to the GL. The licensee indicated that the scope would be revised to be consistent with the GL and the program to be implemented. The inspectors noted that the previous inconsistencies in the program plan scope had been corrected.
The licensee's methodology for MOV sizing and switch settings previously had not included all relevant design basis parameters nor a systematic methodology to verify these parameters. The inspectors observed that the current design-basis calculations had suitably addressed considerations that had been previously missing, such as fluid flow, temperature, and seismic conditions.
Additionally, the licensee had acceptably inserted margins for diagnostic system inaccuracy and rate of loading (load sensitive behavior) and had developed a methodology to verify the margins.
The licensee had not considered the results of its MOV tests in verifying the assumptions used in their design basis calculations and did not have procedures developed for the feedback of information from these tests. The inspectors observed that the licensee had developed a system to calculate valve factor, stem friction coefficient, and rate of loading from the results of each test. The process of analyzing this information in a feedback loop was in progress and was the subject of an inspection followup item identified in this report.
The licensee did not have justification for the planned use of static testing for the purpose of periodically verifying MOV capability. This issue is discussed in Section 1.5 of this report.
The trending of M0V information had not been fully developed. This area is discussed in Section 1.6 of this report.
The control of setpoints had been considered a weakness.
The inspectors concluded that an acceptable control of setpoints has been established.
1.8 Schedule The licensee expressed confidence that the initial validation of MOV design-basis capability would be completed by the GL 89-10 cc=pliance date of June 28, 1994. Tha inspectors observed that this goal appeared to be reasonably achievable considering the progress made to date. The licensee currently has 77 MOVs assigned to the GL 89-10 program. The licensee stated that an attempt would be made to test each valve under some differential pressure and flow condition. This goal was considered to represent a strength
.
-
-
.
.
.
-14-in the program. The licensee had developed a list of 7 MOVs that were currently considered to be impracticable to test under differential pressure conditions.
!
l l
l
!
I I
f l
--
.-._
,_
...
-
.
..-
.
-
..
-
,
!
.
i l.
'
>
(
!
l ATTAct9 TENT I i
1 PERSONS CONTACTED
!
1.1 Licensee Personnel
,
- R. Allen, Manager, Security and General Support
- R. Azzarello, Director, Design Engineering
- T. Becnel, Senior Electrical Maintenance Instructor
- T. Brennan, Design Engineering Technical Assistant
[
- 0. Bulich, Design Engineer, Mechanical
,
- R. Burski, Director, Nuclear Safety
-
- A. Cilluffa, Supervisor, Maintenance Engineering
.
- E. Fields, Design Engineer, Electrical l
- K. Fitzsinmons, Design Engineer, Mechanical
- T. Gates, Licensing Engineer
- J. Hoffpauir, Superintendent, Maintenance
- J. Hologa, Design Engineer, Mechanical
'
- A. Keller, Preventive Maintenance, Electrical
- L. Laughlin, Licensing Manager
- T. Leonard, Technical Services Manager
- A. Lockhart, Quality Assurance Manager
.
- G. Matharu, Supervisor, Design Engineering
!
- J. Meibaum, System Engineering
,
- D. Packer, General Manager, Plant Operations
- R. Peters, Superintenc'ent, Electrical Maintenance
'
- J. Sherrod, Preventive Maintenance, Electrical
- E. Simbles, Design Engineer
>
- K. Van Le, Maintenance Engineering l
1.2 NRC Personnel i
W. McNeill, Reactor Inspector
,
In addition to the personnel listed above, the inspectors contacted other l
personnel during the inspection.
'
- Denotes personnel that attended the exit meeting.
2 EXIT MEETING l
'
An exit meeting was conducted on April 16, 1993. During this meeting, the
,
l inspectors reviewed the scope and findings of the report. The licensee did
!
not identify as proprietary any information provided to, or reviewed by, the l
inspectors.
l l
l
. -.-
-
-
-,
.
.
,, -. -,
. -,. _,.. _,. -..... _..
...
. _ _...,-.,
r
.
.
ATTACHMENT 2 WATERFORD EATE VALVE DATA Diagnostks: VOTES Systan with LVDT for Torque Measurenents VALVE VALVE BZE TEST D1NAheC STEM ID Alf*
NLMBER
&
CONDfT10NS VALVE FEICTION SDGTIVE MANUFACTURER FACTOR *
CODTICIEP(T BEHAY3OR SI135B 8*
195 paid (Close)
0.61 (Close)
0.16 (Dynamic)
-19.9 %
Anchor Darling 195 paid (Open)
0.37 (Open)
0.20 (9 'k)
SI-219B 4*
708 paid (Close)
0.35 (Close)
0.293 (Dynamic)
20.4 %
Anchor Darling 628 psid (Open)
0.15 (Open)
0.214 (Static)
SI41A 14*
C paid (Cbse)
(No Data)
0.094 (Dynamic)
12.1%
Lunkenheimer 353 paid (Open)
0.31 (Open)
0.089 (Static)
SI-501 B 3*
958 psU (Close)
0.52 (Close)
0.183 (Dynunic)
32.4 %
Anchor Darling 958 paid (Open)
0.28 (Open)
0.081 (9_ tk)
CVC-183 4"
59 paid (Cbsc)
0.85 (Ckme)
0.104 (Dynamic)
-6.0 %
,
Powell 62 psid (Open)
0.21 (Opcm)
0.087 (hr6)
7he dynamic valve factors listed were calculated by the beensee using a mean seat diamcter
"*
A negative number indicates that the thrust observed at CST during the dynamic test was greater than the thrust observed at CST during the static test.