ML20035A106
| ML20035A106 | |
| Person / Time | |
|---|---|
| Site: | South Texas  |
| Issue date: | 03/19/1993 |
| From: | NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV) |
| To: | |
| Shared Package | |
| ML20035A066 | List: |
| References | |
| 50-498-93-08, 50-498-93-8, 50-499-93-08, 50-499-93-8, NUDOCS 9303240073 | |
| Download: ML20035A106 (5) | |
See also: IR 05000498/1993008
Text
.
!
i
'
I
.
-7-
,
,
issue concerned the possibility that Valve SI-31A was overtorqued during the
!
sequence of events described above. Excess torque may have been applied to
the valve actuator during the clnsing stroke (considering the potential
.'
hydraulic lock of the springpack) and/or when the valve was manually opened by
the two operators. Based on rough calculations and Limitorque vendor
information, the inspector concluded that the more likely of the two scenarios
,
to cause an overtorque condition was the manual handwheel opening stroke.
Depending on the handwheel configuration and the force appited by the two
operators, torque in excess of 200 percent of the nominal actuator rating
(850 foot-pounds) may have been applied in the effort to unseat the valve. Of
additional concern, the licensee had not addressed this consideration in the
'
i
scope of its initial corrective actions. The valve internals were not
inspected prior to returning the valve to an operable status.
The licensee
,
stated that the potential overtorquing of Valve SI-31A would be investigated
'
and all necessary corrective actions would be taken during the ongoing Unit 2
-;
refueling outage.
[
.
The inspector reviewed the licensee's response to information disseminated by
Limitorque discussing the subject of hydraulic lock of the springpack. The
licensee determined in their review process that every one of their
springpacks was of the newer design that included an internal grease relief
!
from the springpack back to the gear casing. Limitorque had stated that
l
hydraulic lock should not occur in a springpack of this design, but that if
the grease were permitted to harden, the relief function would be defeated.
,
In this case the grease appeared to be of sufficient hardness to prevent flow
!
through the relief valve.
The licensee stated that this was the first
apparent incident on springpack hydraulic lock at this site. The grease that
-
had migrated into the springpack of Valve SI-31A was approximately 5 years in
,
age.
In the licensee's MOV program, each springpack is inspected
approximately once every 5 years. The inspector discussed with the licensee
i
the possibility that the 5-year inspection frequency of actuator springpacks
may not be sufficient to detect hardening of grease in time to prevent
additional occurrences of hydraulic lock. The licensee stated that this issue
-
would be reviewed within the context of the corrective action plan developed
in response to this event.
1.2
Overtoroue of Unit 1 RHR Suction Valves
'
On February 2, 1993, while closing out the review of MOV diagnostic test data
from the recently concluded IRE 04 refueling outage, the licensee discovered
that all six Unit 1 RHR suction isolation valves were set in the closing
direction at torque levels that exceeded the manufacturer's rated torque
l
capacity of the valve actuator. These valves are 12-inch flexible wedge gate
valves and are driven by a Limitorque SB-2 actuator and a 60 foot-pound
e
Reliance motor, which together operate at an overall actuator ratio of 150:1.
The nominal torque structural rating of the actuator is 1250 foot-pounds.
!
Limitorque has authorized operation of its actuators up to 110 percent of the
nominal rating to account for inertial loads that develop after motor cutoff.
t
e
'
9303240073 930319
'
ADOCK 05000498
O
,
W -+
.
_
-
!
'
.
-
O
-8-
,
!
'
The RHR suction isolation valves are normally closed. When the unit is shut
down, these valves are opened to establish the lineup for long-term shutdown
!
cooling. The primary active safety function of these valves is to close to
!
isolate a downstream line break in the RHR system during shutdown conditions.
The primary passive safety function is to prevent the RHR system from
overpressurization during normal operations.
The licensee calculated the applied torque levels, including allowance for
diagnostic system measurement inaccuracy, to be as follows:
,
Valve Number
Closino Toroue (ft.-lbs.)
Percent of Ratino
,
AIRHMOV0060A
1772.2
142
BIRHM0V0060B
1788.6
143
CIRHM0V0060C
1262.7
101
'
BIRHM0V0061A
3204.0
256
C1RHMOV0061B
2869.1
230
-
AIRHMOV0061C
2802.1
224
i
'
In a similar manner, the licensee checked the torque levels on the
_
.
corresponding valves in Unit 2 and found that only one of the six valves had
f
been overtorqued, but that this valve was within 110 percent of nominal torque
as permitted by Limitorque. Therefore, the licensee concluded that the Unit 2
RHR suction isolation valves were acceptable.
,
The licensee initiated SPR 93-0365 to investigate the Unit 1 overtorque
situation and to determine whether the valves should be considered operable.
Based on the results of the inspections of valve internals performed during
the IRE 04 refueling outage, the licensee determined that the valves could
perform their design function, but would be limited to a design duty of less
than 2000 cycles as originally designed. Based on previous test results and
records of maintenance and operations, the licensee determined that the six
i
RHR valves had been overtorqued to the levels shown above for approximately
i
50 cycles. The licensee contacted Westinghouse,-the valve supplier, and
received a calculation predicting the torque capability of the valves. The
licensee declared the valves operable with the intention of continuing
operation until the next refueling outage (Spring 1994) before taking actions
to reduce the torque applied to the actuators.
Unit I was shut down on
February 4,1993, due to unrelated problems, and the licensee decided to take
advantage of this opportunity to correct the overtorque condition.
The inspector focused on the validity of the licensee's initial determination
that the valves were operable. Of concern, was the fact that no industry
testing has been performed to uprate the torque structural limits of
Limitorque actuators.
Several qualification testing programs have increased
the qualified thrust carrying capability of the actuators, but these programs
.,
have consistently stipulated that the torque levels _ should be maintained at or
below the vendor torque ratings (including a 10 percent allowance for inertial
loads). The inspector reviewed the calculation Westinghouse supplied in
response to this event.
Based on a scaling model of the cross-sectional areas
.
'
.
d
.
-g-
of the worm gears of previously tested MOVs to the valves in question,
Westinghouse calculated that the actuator worm gears could withstand a torque
load of 1880 foot-pounds for a total of 10 cycles. The inspector questioned
how a calculation increasing the qualified actuator torque to 1880 foot-pounds
for 10 cycles could be used to justify continued operation of actuators that
had been torqued to as much as 3204 foot-pounds for 50 cycles. The licensee
stated that the Westinghouse calculation was used only on a supplementary
informational basis to lend additional support to the inspection results of
the valve internals performed during the previous outage. The valve internal
inspections, therefore, provided the actual basis for the operability
determination.
The inspector reviewed several of the work packages from the IRE 04 refueling
outage associated with the valves in question and noted that several
observations were recorded of " excessive" wear on the worm gear, a component
highly vulnerable to the application of large amounts of torque. Despite
these work package comments, none of the worm gears were replaced. After the
ensuing forced outage of February 4, the worm gears of five of the six valves
were removed and at least three of them were observed to show signs of
accelerated wear and improper alignment to the worm (the inspection of two of
the worm gears occurred after the conclusion of the inspection).
In consideration of the issues discussed above, the inspector concluded that
the licensee had apparently failed to make a proper operability determination
of February 2, 1993.
Even had the actuator inspection results from the
IRE 04 refueling outage shown no signs of unusual wear and had the Westinghouse
calculation encompassed the measured torque values, the basis for considering
the valves operable would still have been challenged due to the long-term,
consistent, and well-documented position of Limitorque, the NRC, and the
industry that Limitorque actuators should not be overtorqued.
stated in Limitorque Maintenance Update 92-01 that their actuators can be
torqued on a one-time basis up to 200 percent of the nominal rating without
damage or sacrifice to the actuator qualification.
In this same document,
limitorque recommended that if the 200 percent torque limit is exceeded or if
an overload occurred more than once, that an inspection of the actuator should
be performed to determine if any damage occurred and if the actuator is
suitable for continued service.
In this case, three of the actuators had been
torqued over 200 percent of the nominal rating for approximately 50 cycles,
and no deliberately focused evaluation of the torque-carrying components had
ever been performed (i.e., only general inspections had been performed, none
with the expressed intent of evaluating the effects of overtorque).
The operability determination of February 2 and the failure to take immediate
corrective actions to address the overtorque problem of the RHR suction
-
isolation valves is considered to be an apparent violation of 10 CFR 50,
Appendix B, Criterion XVI, " Corrective Action" (apparent violation
498/9300-04).
Based on a review of a previous enforcement issue, the inspector determined
that the apparent violation discussed above is similar to a violation of
_
.
_
_
.
-
_._ . >
_,
t
'
.
-l
!
t
e
-10-
1
i
!
,
10 CFR 50, Appendix B, Criterion XVI, that was issued in NRC Inspection
I
!
Report 50-498/92-06, 50-499/92-06. This inspection report documented the
i
first NRC inspection of the licensee's response to Generic Letter 89-10,
- Safety-Related Motor-0perated Valve Testing and Surveillance." The violation
!
was based on the licensee's unacceptable justification for continuing
l
operations with MOVs left in a condition that would result in delivered thrust
1
exceeding the actuator thrust structural rating. At the time, there were two
,
preliminary third-party test results indicating that the licensee's measured-
l
thrust levels would eventually fall within an expanded qualified thrust range.
However, the relevant test studies were still under evaluation and had not
been officially promulgated. Therefore, it was determined that the licensee's
i
operability determination was invalid at the time that it was made. The
!
overtorque issue is analogous to the overthrust issue of a year earlier, in
.
[
that both involved placing stress on the actuators in excess of the qualified
ratings. Of special note, the current issue of overtorque is considered more
'i
serious than the previous overthrust issue because there does not exist even
!
preliminary test results enveloping the observed conditions._ Further, while
l
it was recognized that the overthrust situation existed at the time the
affected actuators were visually inspected (suggesting that the inspection
'
would be appropriately focused to address that concern), the overtorque
r oblem was not recognized until after the inspections performed during the
l
last refueling outage had been accomplished, thus, rendering those inspections
}
less effective in the evaluation of operability.
After Unit 1 entered a forced outage on February 4,1993, for reasons
unrelated to the RHR MOV problems, the licensee decided to undertake actions
,
to address the overtorque situation. Beginning with the "B" train valves,
,
five of the six actuators were disassembled and inspected. Valve CIRHMOV0060C
'
(RH-60C) was not included .in the scope of work because_ its calculated _ torque
of 1262.7 foot-pounds only slightly exceeded the actuator rating of
,
i
1250 foot-pounds and was well within the margin allowed for inertial loading
as previously discussed. For each of the remaining five valves, the licensee
.
replaced the wormshaft, worm, and worm gear, which comprise some of the
components most susceptible to excessive torque loads._ In some of the valves,
including at least the two B train RHR valves, the stem nuts were replaced.
The stem nuts in each of these valves exhibited.an unusual groove running down
the middle of the thread face, perhaps caused-by a foreign particle lodged
i
between the stem nut and the stem.
In Valve C1RHMOV00618 (RH-618), an
inspection of the disassembled parts revealed a misalignment of the worm and
.
the worm gear. The worm was riding high on the worm gear and causing
I
accelerated wear of the worm gear.
In Valve BIRHMOV0060B (RH-608), a
,
misalignment of the stem to the actuator was observed.
Each of the conditions
j
discussed above could have contributed to the high torque values that were
i
observed.
t
i
The calculated stem factors (torque divided by thrust) were as high as 0.045
(for Valve RH-61A, corresponding to a stem friction coefficient'of 0.28) as-
- '
compared to the assumed value of 0.0291 (corresponding to a stem friction-
'
coefficient of 0.15). As a result, the actuator had to apply additional _
i
-
torque to deliver the same amount of thrust to the valve stem. Since the
i
l
i
.[
e
.
.
-11-
t
valves were set up to close on a limit switch (which the licensee attempts to
t
set at a valve disc position that corresponds to the calculated thrust
'
requirement) and the torque switch was removed from the motor circuit, there
was no backup protection against overtorquing the actuators. Often, high stem
factors are associated with inadequate stem lubrication, but for these valves
that did not appear to be the case. After a fresh lubrication, which preceded
the replacement of parts, the valves were retested with very little change
noted in the measured torque levels.
In addition to the replacement of some of the torque-bearing components in the
i
actuators, the licensee took additional action to lower the applied torque to
the valves. This step was to reduce the maximum expected differential
pressure (MEDP) under which the valves would be expected to operate. The
l
original MEDP was 700 psid based on the setpoint of the previously-installed
l
auto closure interlock. The interlock had been removed and no longer
i
represented a basis for the MEDP. The licensee established a new MEDP of
l
450 psid based on plant administrative procedures that require the RHR suction
isolation valves to be closed any time the reactor coolant system-(RCS)
!
pressure is above 450 pounds per square inch gage (psig). By lowering the
MEDP, the required stem thrust was likewise lowered, allowing for the limit
switch to be set earlier in the valve stroke.
This in turn would result in
l
less torque being applied to the valve actuators. The inspector questioned
the 450 psid limit because both the design pressure of the RHR piping and the
i
f
setpoint for the RHR relief valve is 600 psig and the lower setpoint of the
cold overpressure mitigation system is 485 psig. The licensee stated that the
revised MEDP calculation would be revisited and confirmed in a more rigorous
'
manner. The inspector informed the licensee that use of the 450 psid MEDP
,
!
would be considered acceptable for the current fuel cycle based on the
proceduralized administrative controls in place to prevent pressuri:' g the
RHR system above that level. However, for periods of operation beyond that
l
time frame, a more fully justified basis would be considered necessary.
With the use of the lowered MEDP and the refurbishment of the actuators
(replaced and realigned components), the licensee was successful in_ lowering
the as-left torque levels of Valves RH-608, RH-61B, and RH-61C to less than
110 percent of the nominal actuator rating. Test results from Valves RH-60A
and RH-61A were not available at the conclusion of the inspection. As stated
previously, no actions were taken for Valve RH-60C. After successful
diagnostic tests, the valves were returned to an operable status.
The inspector reviewed the final test results of Valves RH-60B, RH-61B, and
RH-61C. The as-left total torque levels (including instrument inaccuracies)
for Valves RH-60B and RH-61B were less than the nominal rating of
'
1250 foot-pounds. The as-left torque for Valve RH-61C was 1275 foot-pounds or
102 percent of the rating. This was well within the Limitorque limit of 110
percent.
For consideration of immediate operability, the inspector determined
!
that the licensee's actions and test results were sufficient to establish an
acceptable basis for declaring the valves operable. However, the inspector
,
identified four concerns that need to be addressed to establish that the
valves will remain operable over time. These issues have generic significance.
.