ML20034G862
| ML20034G862 | |
| Person / Time | |
|---|---|
| Site: | South Texas  |
| Issue date: | 03/03/1993 |
| From: | Stetka T NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV) |
| To: | |
| Shared Package | |
| ML20034G855 | List: |
| References | |
| 50-498-92-36, 50-499-92-36, NUDOCS 9303120022 | |
| Download: ML20034G862 (35) | |
See also: IR 05000498/1992036
Text
.
i
.
.
.
.-
t
..
APPENDIX B
l
r
U.S. NUCLEAR REGULATORY COMMISSION
REGION IV
i
NRC Inspection Report:
50-498/92-36
50-499/92-36
r
!
Operating License: NPF-76
NPF-80
l
i
Licensee:
Houston Lighting & Power Company
P.O. Box 1700
Houston, Texas 77251
Facility Name:
South Texas Project Electric Generating Station,
Units 1 and 2
Inspection At: Matagorda County, Texas
Inspection Londucted: December 6, 1992, through January 16, 1993
Inspectors:
J. I. Tapia, Senior Resident Inspector
I
R. J. Evans, Resident Inspector
e
C. E. Johnson, Reactor Inspector
'
N
3/3 /93
Approv d:
,r? T. F. StetkT', Chief, Project Section D
Datel
,
,/
i
Inspection Summary
Areas Inspected: Routine, unannounced inspection of plant status, onsite
!
followup of events, operational safety verification, maintenance and
.!
surveillance observations, preparation for refueling (Unit 1), and followup on
i
a previously identified inspection followup item, three deficiencies, and six
.
observations.
,
Results:
!
Unit 2 was manually tripped when a secondary valve failed shut.
Several
additional secondary events occurred after the shutdown. Additionally,
four power maneuvers were made because of secondary equipment problems.
,
Higher levels of management oversight continue to be needed in this area
t
because of the continuing negative trend in the reliability and
,
availability of secondary components (Sections 2.1 and 3.6).
!
.
Both units were required to shut down because of the discovery of
!
incorrectly calibrated components.
The event was caused by deficient
surveillance procedures. The failure to develop and maintain safety
related surveillance procedures was a noncited violation of Technical
j
Specification (TS) requirements.
Following the Units 1 and 2 TS 3.0.3
I
!
7303120022 930305
'
ADOCK 05000498
O
-
.
.
-
=
-
.
.
-
.
.
-
I
l
-
-2-
l
,
!
required shutdowns, teams of instrumentation and controls technicians
t
were assembled to recalibrate suspect amplifiers. The work was noted to
,
be highly controlled by plant supervisors (Sections 2.2 and 5.2).
l
t
"
A surveillance test on a supplemental containment purge system valve was
not performed within the required time period specified in the TS.
This
'!
was the first example of a failure to satisfy TS requirements and was a
violation of the facility operating license (Section 3.1).
!
,
During a plant cooldown to repair a leaking seal weld on a control rod
t
I
drive mechanism housing, a steam generator power-operated relief valve
failed to operate because of a defective pressure switch (Section 3.2).
i
Emergency Diesel Generator (EDG) 11 experienced a valid failure to start-
.
during a monthly operability test as a result of excessive exhaust
,
temperature on a cylinder. The excessive temperature resulted from the
binding of a fuel lever arm which had never been lubric;ted. This
failure to lubricate resulted from a less than adequate preventive
'
maintenance (PM) procedure which did not require lubrication of the fuel
i
lever arm (Section 3.3).
'
In response to a previous commitment to review surveillance procedures
to determine their technical adequacy, a number of deficient procedures
were identified.
This was the fourth instance that deficient procedures
,
were identified during this review.
The deficient procedures were
!
considered to be noncited violations of NRC requirements.
The high
l
number of procedures being identified were a concern to the inspectors.
!
The scope of the surveillance procedure review task force should be
expanded because of the high number of deficient procedures that were
!
identified (Section 3.4).
i
l'
During the performance of a surveillance test on a component cooling
!
water (CCW) system valve, a broken terminal lug was identified.
Licensee personnel failed.to issue a station problem report (SPR) to .
j
investigate the cause of the event. After prompting by the inspector,
!
licensee personnel issued an SPR to assess the root cause of the
failure. .This was an additional example of problems in the generation
!
of SPRs and may be further addressed in NRC Operational Safety Team
Inspection Report 50-498/92-35; 50-499/92-35 (Section 3.5).
The failure to maintain at least three channels of overtemperature
i
differential . temperature (OTDT) operable was the second- example of a
'
failure to satisfy TS requirements.
The cause of the event was a
deficient procedure (Section 3.7).
-!
!
The failure to perform a daily channel calibration on a nuclear
i
instrument (NI) was the third example of a failure to satisfy TS
requirements.
A contributor to the event was the failure of a licensed
operator to record a key entry in the control room logbook
j
(Section 3.8).
}
.
.
.
.
.
. ..
>
-
i
.
o-
I
~
3
j
i
!
Problems continue to exist with one source range neutron flux monitor in
l
Unit 1.
This monitor has been intermittently inoperable since the
!
Spring of 1992 (Section 4.1).
'
i
A crack was found and repaired in the Unit 1 ECW system piping.
Although dealloying and crack problems continue to exist with the piping
of the system, the licensee's response to the problems continues to be
prompt and aggressive (Section 4.2).
'
i
'
An EDG was unintentionally tripped during a maintenance'run because of
e
inadequate venting of the lubricating oil piping (Section 4.3).
!
!
Numerous problems with the plant's toxic gas monitors were experienced
.
because of equipment malfunctions. Two examples of the failure to
adhere to TS requirements were identified.
One of the TS violations
involved the failure to maintain an out of service channel in the
tripped condition. The second violation involved the failure to perform
a channel check. The licensee's efforts to improve the reliability and
-;
availability of the toxic gas monitor systems have not been successful
i
(Section 4.4).
I
During the performance of a solid state protection system logic
functional test, problems were encountered with a test pushbutton. This
pushbutton has not worked properly since April 1992. This pushbutton
was scheduled to be replaced during the upcoming refueling outage
(Section 5.1).
f
All major work activities were successfully completed during the recent
.
Unit 1 fourth refueling outage. Management oversight of outage
activities continue to be a strength, however, additional oversight of
refueling equipment reliability issues was warranted (Section 6.2).
.!
1
Summary of Inspection findings:
Violation 498;499/9236-01 was opened (Sections 3.1, 3.7, 3.8, and 4.4).
Violation 498:499/9236-02 was opened (Section 3.3).
.j
-l
Inspection Followup Item 498:499/8868-05 was closed (Section 7.1).
]
1
Deficiency 498;499/92-201-01 was closed (Section 7.2.1).
l
Inspection Followup Item 498:499/9236-03 was opened (Section 4.2)..
j
Inspection Followup Item 498;499/9236-04 was opened (Section 7.2.1).
Deficiency 498;499/92-201-02 was closed (Section 7.2.2).
)
Violation 498:499/9236-05 was opened (Section 7.2.2).
l
1
)
i
.
,
- , .
.-
.
-
-
.
.
t
.
,
i
-4-
!
i
Deficiency 498;499/92-201-03 was closed (Section 7.2.3).
.
.
Violation 498;499/9236-06 was opened (section 7.2.3).
l
i
Observation 498;499/92-201-01 was closed (Section 7.2.4).
Observation 498;499/92-201-02 was closed (Section 7.2.4).
Observation 498;499/92-201-03 was closed (Section 7.2.4).
Observation 498:499/92-201-04 was closed (Section 7.2.4).
!
Observation 498;499/92-201-05 was closed (Section 7.2.4).
Observation 498;499/92-201-06 was closed (Section 7.2.4).
l
>
Attachments and/or Enclosures:
,
o
Attachment 1 - Persons Contacted and Exit Meeting
l
.
.
?
?
t
f
i
!
l
,
.
!
l
l
i
!
,
)
I
'
-
,
-
__
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ . ___ -
.
.
,
--
-5-
DETAILS
1 PLANT STATUS
At the beginning of this inspection period, Unit I was in Mode 5 (Cold
Shutdown) in day 77 of a scheduled 62 day refueling and maintenance outage
that began on September 18, 1992.
Unit 1 entered Mode 4 (Hot Shutdown) on
(
December 6, 1992, and entered Mode 3 (Hot Standby) 2 days later.
On December
'
10, 1992, the unit returned to Mode 5 operation to allow for the. rework of a
reactor coolant pump bearing and repair of a control rod drive mechanism leak.
Unit I reentered Mode 4 operation on December 25, 1992, and Mode 3 operation
the next day. Mode 2 (Startup) operation and criticality was achieved on
1
December 28, 1992. Mode 1 (Power Operation) was entered 2 days later. _ The
refueling outage ended on December 31, 1992, when the main generator output
breaker was closed and the unit was tied to the offsite grid. The outage
J
lasted 103 days.
Power was increased in increments to allow for startup
'
testing and the unit reached 90 percent power operation on January 11, 1993.
On January 11, 1993, Unit 1 power was reduced from 90 to 74 percent to allow
for the performance of NI calibration and troubleshooting activities. The
next day, Unit 1 entered TS 3.0.3 and a Notification of Unusual Event (NOVE)
was declared. The NOUE was declared when 9 of 12 compensated low steam line
pressure instruments and 5 of 12 steam line high pressure negative rate
instruments were determined to be inoperable because of concerns with the time
constant settings of the loops. The unit was shut down in an orderly fashion
and Unit I entered Mode 3 operation the same day. The unit remained in Mode 3
through the end of the inspection period.
-At the beginning of this inspection period, Unit 2 was operating in Mode I at
100 percent power. On December 9, 1992, Unit 2 power was reduced to
90 percent to allow for surveillance testing of the main turbine steam inlet
valves.
Power was increased to 100 percent the same day. On December 27,
1992, Unit 2 operators manually tripped the unit when a main feedwater
regulating valve unexpectedly closed and failed to open on demand from the
control room. The unit was stabilized in Mode 3 following the manual trip.
Unit 2 had remained on line for 305 days (site record) prior to the manual
trip.
Following repair of the feedwater regulating valve, the unit commenced
startup on December 29, 1992, and full power operation was achieved 2 days
later.
Several hours after full power was achieved, the unit decreased in.
power to 95 percent to allow for maintenance on a feedwater heater. The unit
returned to full power the next day.
On January 12,1993, Unit 2 also entered TS 3.0.3 and shut down to Mode 3 due
to inoperable steam pressure transmitter instrument loops. -Following repairs,
the reactor was brought critical (Mode 2) on January 14., 1993, and the unit
then entered Mode 1.
On January 15, 1993, power was reduced from 48 to
30 percent to allow for work to commence on the Steam Generator Feedwater
Pump 23 control circuits.
Power was reduced to comply with the administrative
requirements of the licensee's reactor trip prevention program.
The next day,
unit power was reduced to 25 percent when Steam Generator Feedwater pump 22
I
.
.
.
-6-
a
tripped while troubleshooting activities were in progress on Pump 23.
Following the completion of maintenance activities on the steam generator
feedwater pumps, unit power was increased on January 16, 1993, and the
inspection period ended with the unit at 50 percent power and increasing (the
unit reached 100 percent power the next day).
2 ONSITE RESPONSE TO EVENTS (93702)
!
2.1 Unit 2 Manual Trip
A manual trip of the Unit 2 reactor was required on December 27, 1993, when a
feedwater regulating valve failed shut and could not be reopened by operators.
In addition to the feedwater regulating valve failure, the unit experienced
-
several unexpected and undesired equipment and system malfunctions during the
!
reactor trip transient. The startup feedwater pump failed to start, a main
feedwater pump was secured because of no output flow, and a letdown isolation
i
occurred due to low pressurizer level resulting from an excessive secondary
steam demand. The licensee submitted Licensee Event Report (LER) 499/92-010
documenting the event.
On December 27, 1992, Unit 2 was in operation at full power. At 8:50 a.m.,
i
Feedwater Regulating Valve 2D unexpectedly went to the closed position.
Attempts were made by the control room operators to manually open the valve,
,'
but their attempts were unsuccessful.
The control room operators then
manually tripped the reactor because of the reduction of feedwater flow to
Steam Generator 2D with a corresponding decreasing level. The cause of the
event was subsequently determined to be the failure of a driver card in the
valve control circuitry. About 6 minutes after the manual reactor trip, the
i
unit experienced a chemical volume control system letdown isolation because of
pressurizer low level as a result of a reactor coolant system cooldown. The
'
low pressurizer level condition and system cooldown were caused by high
secondary steam demand, which was partially because the unit was providing
auxiliary steam to Unit 2.
Actions were taken to reduce the steam demand.
,
The auxiliary steam supply was isolated and the main steam isolation valves
'
were shut.
At 9:11
a.m., the control room operators attempted to start Startup Steam
Generator Feedwater Pump 24.
Pump 24 f ailed to start on low lube oil
pressure. The pump was subsequently started 3 minutes later, after the lube
,
oil duplex filters were shifted. The cause of the Feedwater Pumo 24 failure
to start was a combination of the start permissive pressure switch being out
i
of calibration and a high differential pressure of the inservice filter.
The
i
control room did not receive a high differential pressure alarm because the
l
alarm setpoint was found to be incorrectly set above the pump start permissive
setpoint. The absence of an alarm in the control room for the startup
,
feedwater pump filter differential pressure resulted in a failure to detect
the inability of the startup feedwater pump to start.
Corrective actions
planned included revising the setpoints to have a simultaneous alarm and
"
permissive start signal setpoint.
Further investigation by the licensee
indicated excessive water in the lube oil system.
!
s
1
.
--
i
.
.
.
4
-7-
A similar startup feedwater event occurred on November 21, 1992. Un that
date, Pump 24 trippei while in operation without initiating a control room
alarm. The cause of the trip was also a result of excessive water in the lube
oil system. The pump seal design allowed water intrusion into'the lube oil.
system when the pump was idle.
The licensee was investigating the adequacy of
the design and possible corrective actions. The inspectors will monitor the
licensee's efforts.
On December 30, 1992, at 1:37 a.m., the control room operators started and
rolled Steam Generator Feedwater Pump 22 to 3100 rpm, which was approximately
60 percent of the pump's rated speed. The operators tripped the pump a few
minutes later when no output flow was indicated concurrent with a high pump
casing temperature, despite the output and recirculation valves indicating
open.
Further investigations revealed that the manual isolation valve in the
recirculation line was shut. The pump was not able to discharge because the
recirculation line was improperly isolated and the discharge pressive of the
startup feedwater pump was higher than the discFarge of Feedwater Pump 22. As
a result of the pump being deadheaded in this manner, the pump _ casing
temperature increased. At 2:35 a.m., Pump 22 was placed in service, after the
manual recirculation valve was opened.
Subsequent investigations revealed
that the manual isolation valve was shut in June 1992 by a nonlicensed reactor
plant operator. The reason the valve was closed was not known at.the end of
the inspection period and the operator who shut the valve was no longer
employed at the site. Additionally, the valve was not danger or caution
tagged and.a reactor plant operator walkdown just prior to the first pump
start failed to identify the valve being shut.
The inspectors considered the
licensee's response to the mispositioned valve a weakness in that the valve
was out of position for an extendad period of time and a system walkdown was
unable to identify and correct tne problem.
2.2 Technical Specification Reauired Shutdown of Units 1 and 2
i
On January 12, 1993, both units were required to shut down to comply with TS
requirements. The forced shutdowns were necessary because the licensee
,
discovered that selected safety-related components were improperly calibrated.
The cause of the event was determined to be deficient surveillance procedures.
'
The f ailure to develop and maintain safety-related surveillance procedures was
determined to be a noncited violation of NRC requirements. The licensee
i
planned to submit LER 498/93-003 on the event.
During the recent Unit I refueling outage, the steam pressure loop calibration
surveillance procedures were performed to verify the accuracies of the
-i
compensated steam line pressure low (safety injection signal) and the steam
i
line pressure negative rate high (steam line isolation signal) instrument
.
loops. During the calibration process, selected components were identified to
l
have as-found values outside of the acceptance criteria limits.
According to
!
the requirements of Administrative Procedure OPGP03-ZM-0016, Revision 6,
l
" Installed Plant Instrumentation Calibration Verification Program,"
i
information service requests (information service requests replaced the old
'
Request for Action forms) were generated for the out-of-tolerance conditions.
,
!
i
f
- -
-
-
.
.
.
.
.
. . - - . -
-
=
. - .
-.
.
. .
-
-
--
-
I
.
-
.i'
.
I
-
!
-8-
l
.
l
During the review of the information service requests by plant engineering
. ;
department personnel, discrepancies between the TS and procedure aczeptance
!
criteria were identified.
Plant engineers noted that the tolerance band for
t
the lead and lag time constants used in the steam pressure instrument loops
1
'
appeared to violate TS required settings. TS bases state that engineered
safety features actuation system (ESFAS) setpoints were target values and
,
allow a range, but there was some uncertrinty as to whether TS allowed a
tolerance band for lead and lag time con; ants.
SPR 930104 was written on
.
January 12, 1993, at 10:30 a.m.,
about I hour after the discovery of.the
j
tolerance discrepancy. Once the licensee became aware of the discrepancies
'
between the TS limits and the tolerance errors, a. review of the most current
i
lead / lag amplifier time constant setpoints was performed.
The licensee noted
i
that a significant number of lead / lag amplifiers were incorrectly set.
Of the
,
24 amplifiers in Unit 1, 14 were incorrectly set, while in Unit 2, 19 of
l
24 amplifiers were incorrectly set.
l
.
On January 12, 1993, at 2:52 p.m., Unit 1 entered TS 3.0.3 because of the high
number of inoperable steam pressure lead / lag amplifiers. An NOUE was declared
I hour later and a unit shutdown was commenced to comply with TS requirements.
!
Unit 2 entered TS 3.0.3 at 4:03 p.m. the same day, following the completion of
i
the Unit 2 record review.
Both units were stabilized in Mode 3 operation and
i
both units exited the NOUE at 1:30 a.m. the next morning after reducing
reactor coolant system pressure to below the permissive P-ll setpoint, the
condition that permitted exiting TS 3.0.3.
A complete description of the work
'
activities performed was provided in Section 5.2 of this report.
The licensee
did not believe that a safety basis existed to justify requesting regulatory
relief from the TS requirements, therefore, a request for a temporary waiver
of compliance was not generated.
.
Immediately after the discovery of the time constant discrepancies, the
!
licensee began a review of TS and surveillance procedures to determine if
!
other time constant errors existed in their surveillance program. The
,
licensee determined the time constants used in the reactor trip system
!
instrumentation setpoints for OTDT and overpressure differential temperature
'
did not specify an allowable tolerance in the lead and lag compensator which
could account for instrument errors. However, the results of a Westinghouse
i
instrument sensitivity study indicated that sufficient margin exists to
l
accommodate a 5 percent tolerance for the lead / lag instrument characteristics.
l
The licensee's design engineering department concluded that an acceptable-
!
basis existed for the allowed tolerance.
This position was approved by the
-
Plant Operations Review Committee 'and was documented in TS Interpretation TSI-
132.
All other time constant tolerances were found to be consistent with TS
requirements.
i
The cause of deficient surveillance procedures was inadequate development and
.-
review of the original procedures.
The failure to implement and maintain
adequate surveillance procedures was a violation of TS 6.8.1 requirements.
!
However, this violation will not be subjected to enforcement action because
!
the licensee's efforts in identifying and correcting the procedures meet the
l
!
l
l
!
,- . - . _ -
,
- .-
_
.,
m--
-__m
_
..
.
-
_.
.
.
.
~
~i
!
.
_g.
i
.
criteria specified in Section VII.B.2 of Appendix C to 10 CFR 2.
The
violation was licensee identified; it was reported to the NRC operations
center within the required time interval; prompt corrective actions were
taken; and it was not a willful violation.
.
I
2.3 Conclusions
i
'
Unit 2 experienced a manual reactor trip following a secondary component
i
failure.
Problems with secondary components continued to plague the unit
.
~
R
following the unit shutdown. Higher levels of management oversight were
necessary to implement initiatives in order to redsce the number of events
that occur because of secondary component failures.
The dual unit shutdown
was the result of the discovery of inadequate surveillance procedures ~
The
.
procedures were licensee identified, which was positive. However, the high
number of surveillance procedures being discovered were a concern ti the NRC.
3 OPERATIONAL SAFETY VERIFICATION (71707)
The purpose of this inspection was to ensure that the facility was being
!
operated safely and in conformance with license and regulatory requirements.
The following paragraphs provide details of specific inspector observations
during this inspection period.
l
3.1 Missed Supplemental Purge Isolation Valve Surveillance (Unit 1)
'!
During the inspection period, the licensee failed to perform a TS required
surveillance test on a valve. The failure to perform the surveillance was a
j
failure to satisfy TS requirements. The failure to satisfy TS requirements
was a violation of the facility operating license.
Licensee personnel' failed
to perform the surveillance, in part, because of a performance coding error
l
that was input into the surveillance program documents. A contributor to the
event was that plant operations department personnel were not fully cognizant
that the surveillance requirements of a component had been revised.
.
The supplemental containment purge system was used during normal plant
l
operating conditions to reduce containment airborne radioactivity levels to
!
allow for personnel entry into the containment.
This system was designed for
a smaller flow rate than the normal purge system in order to reduce the size
!
of the containment penetration isolation valves. The supplemental containment
purge system was also utilized to maintain the normal containment pressure
i
within the required limits. The safety related portions of the system
i
consists of containment isolation valves, piping, and radiation monitors.
l
l
TS 3.6.3 requires the containment isolation valves be operable with isolation
l
times less than or equal to the required isolation times in plant operating
Modes 1-4.
TS 4.6.3.3 requires the isolation time of each power-operated or
automatic containment isolation valve be determined to be within its limits
l
when tested pursuant-to TS 4.0.5.
TS 4.0.5 provides the requirements for
l
inservice inspection and testing of American Society of Mechanical
>
Engineers (ASME) Code Class 1, 2, and 3 components.
!
i
i
l
.
- . .
.
. _ _ . .
_ . - - . .
.
.
_
.
.
-
_ . .
_
i
t
4
-
!
.
,
'
-
i
-10-
r
.
-
i
~
The reactor containment supplemental purge exhaust outside Containment
Isolation Valve Al-HC-FV-9776 was a Class 2 valve, which was required to be
tested on a quarterly basis to verify operability. A relief request to the
l
Pump and Valve Inservice Test Plan was issued to provide an alternate method
(
j
of testing the valve. According to the relief request, this valve was
j
required to be fully stroked each cold shutdown (Mode 5 and 6), not to exceed
j
once every 3 months.
The ASME code allows for relief from normal quarterly
I
testing of some valves if the test would unreasonably endanger equipment,
personnel, or plant operation.
The purge valves were included in this
category because the valves were intended to be closed as much as possible to
,
minimize the potential for a loss of primary containment integrity.
?
T
Procedure 1 PSP 03-HC-0003, Revision 4. " Reactor Containment Building Purge
System Valve Operability Test," provided instru C 's on how to test the purge
valves to comply with the requirements of TS 4.6.3.3 and the ASME Section XI
requirements. On November 27, 1992, with Unit 1 in Mode 5 operation, a
!
quarterly surveillance of Valve FV-9776 was performed in accordance with
>
Procedure 1 PSP 03-HC-0003.
The valve was found to have exceeded its allowed
!
stroke time change. The valve stroke time was 2.07 seconds, which was
!
]
80 percent higher than the previous stroke time, although still below the
2
4
caximum allowed time of 4 seconds
1
-
ASME Section XI requires valves that exhibit abnormal changes in stroke times
!
be tested on an increased frequency until the cause was corrected. The valve
i
,
stroke time test was subsequently increased from quarterly to monthly.
Additionally, the licensee decided to not invoke the relief request testing
requirements and transferred Valve FV-9776 back to the normal requirements
established in the inservice test plan.
The relief request requirements were
j
l
discontinued because:
(1) the licensee had to repair the valve prior to entry
r
into a mode where the valve testing was exempted and the licensee thought the
j
increased stroke time measured was only a statistical variation, and (2) the
j
'
i
valve could be opened in any mode of plant operation.
!
>
I
On December 28, 1992, Valve FV-9776 was scheduled to be tested under' the new
i
testing requirements.
Although the reasons were not fully understood at the
!
!
end of the inspection period, the licensee suspected the surveillance was not
performed because of a " performance code" error.
The performance code that
i
was listed in the test documents listed the surveillance as a cold shutdown
'
!
surveillance (Modes 5 or 6 only). With the unit in Mode 2 operation on
December 28, 1992, a decision was apparently made by the performing
organization (plant operations) to not perform the surveillance because of the
'
i
performance code, which was in error. The missed surveillance was identified
i
[
by the licensee on January 7, 1993, approximately 3 days after the end of the
j
i
Performance of the surveillance was required by Section XI of the ASMt udes
- '
to demonstrate operability of Valve Al-HC-FV-9776.
In accordance with
TS 4.0.2, each surveillance requirement shall be performed within the
l
specified surveillance interval with a maximum allowable extension not to
i
exceed 25 percent of the specified surveillance interval.
TS 4.6.3.3 requires
i
I
i
1
i
4
3
,y
,r
3
,.--
y-%--
~ -- 7
y-
..,-
e.
,
c.
, . -
,m,
.--,a
-
---
. _ _ _ _ -
-__-___--
-
,.
.
.
.
si-
the isolation time of each power-operated or automatic valve be determined to
be within-its limits when tested pursuant to TS 4.0.5 (surveillance
requirements for ASME components). The failure of the licensee to
surveillance test Valve Al-HC-FV-9776 within the required time interval was a -
violation of TS 4.0.2, 4.0.5, and 4.6.3.3.
The failure to comply with TS was
considered the first example of a violation of the facility operating license
(498:499/9236-01).
Short-term corrective actions taken included stroke testing of the valve on
January 7,1993, in accordance with Procedure IPSP03-HC-0003. The valve
cycled shut in 2.16 seconds, which was a normal stroke time for the valve.
The licensee notified the NRC Operations Center of the missed surveillance on
January 13, 1992, following the completion of the reportability review. The.
licensee planned to submit LER 498/93-004 on the incident.
I
3.2 Repair of Control Rod Drive Mechanism Leakage (Unit 1)
During the Unit I startup following the refueling and maintenance outage, a
leak from a control rod drive mechanism (CRDM) was found. The repair of the-
leak forced the unit to-depressurize the reactor coolant system.
During the
cooldown process, a steam generator power operated relief valve (PORV)~ failed
to operate because of a defective pressure switch.
Both the CRDM leak and
PORV were repaired and returned to service.
On December 9, 1992, during the performance of Surveillance
Procedure OPSP15-RC-0001, Revision 0, " Reactor Coolant System Leakage Pressure
Test," leakage was noted at a rod drive housing on the reactor vessel head.
The licensee entered TS 3.4.6.2, Action a, and initiated Service Request (SR)
188088 to repair the leak. About 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> later, TS 3.4.6.2 was exited when it
was determined that the leakage was not pressure boundary leakage but leakage
from a canopy seal weld on a mechanical connection between the vessel head and
the CRDM.
Unit 1 began a plant cooldown from Mode 3 to Mode 5.
Mode 5 was
attained the next morning. To facilitate the rework of the leak, the reactor
coolant system was drained down to below the elevation of the leak. The
licensee obtained the services of a repair contractor to repair-the leak.
This contractor utilized a proprietary canopy seal clamp assembly for the
repair. The contractor has installed 474 seal clamp assemblies at 32 reactor
plants without subsequent. failures.
During the cooldown process, PORV C1-MS-PV-7431 failed to operate from the
l
control room.
Initial troubleshooting revealed that the hydraulic pump would
not run and bad electrical contacts were suspected to be-the cause of the
problem. SR MS-156258 was issued to troubleshoot and repair the.PORV
hydraulic pump. The licensee discovered that the pump high pressure cutoff
switch (pump normally cycles on hydraulic fluid high and low pressure) was
defective. The switch was replaced, recalibrated, and returned to service.
l
Postmaintenance testing was completed on December 16, 1992.
l
l
,
-
- - - _ _ - - _ - _ - - . _ _ _ - . _ _ _ - - - - - _ - _
,
..
-
-
- -
.
.
. .
-
,
.
~
-12-
,
i
.
3.3 EDG Valid Failure
-
On December 9,1992, the licensee performed a monthly surveillance run of
EDG 11 in accordance with Procedure OPSP03-DG-0001, Revision 0, " Standby
i
Diesel 11(21) Operability Test." At the. time the test was performed,.the
plant was in Mode 3, proceeding with a plant startup following the completion
of the refueling outage. During the performance of the surveillance test, the
,
BR cylinder exhaust temperature indicated 1050 F, which was approximately
'
temperature of any cylinder was 980 F.
The EDG was subsequently unloaded and
~
200 F greater than the expected reading.
The maximum allowable operating
'
declared inoperable. This diesel' failure was reported to the NRC as a valid
failure to start on demand.
.
Subsequent troubleshooting disclosed that the lever arm on the fuel control-
,
shaft which actuates the fuel injection pump rack was binding on the affected
cylinder causing an abnormally high fuel flow. The binding lever _ arm was
freed using a penetrating lubricant. The lever arms on the other cylinders
,
associated with EDG 11 were inspected and lubricated. One other-cylinder was
i
found with a binding lever arm.
EDG 11 was successfully tested and returned
!
to service on December 11, 1992.
SRs DG 174507,
-8,_-9,--10, and -11 were
-i
issued to lubricate the fuel lever arms on the other five _ engines.
.;
t
The inspector reviewed PM Work Instruction MM-DG-92000356, Revision 1.0,
l
" Quarterly Lubrication." This review disclosed that, although other
-
components of the fuel control shaft were lubricated quarterly, the
lubrication of the lever arm was not included in-the work instructions. This
lack of lubrication led to' the binding which was the cause of the valid
!
failure of EDG 11. This represented a failure to;have adequate procedures for
.
!
the performance of PM and this failure was considered a violation
(498;499/9236-02).
The high exhaust temperature was classified as a valid failure because the
engine would not have been able to perform-its safety function.
Although the
EDGs have been analyzed to perform satisfactorily with only 18 of the
!
20 cylinders operable, operator intervention to lock out the specific cylinder
may not have occurred fast enough to prevent engine' damage. There has been
!
one valid failure in the last 20 tests of EDG 11, and the number of valid
failures within tht last 100 valid tests was less- than four.
Therefore, the
~
testing frequency for EDG 11 remains on a monthly-basis in accordance with -
l
t
3.4 Missed Main Steam Isolation Response Time Testing Surveillance Checks
!
(Units 1 and 2)
_[
Onl December 15, 1992,.during a. review of selected surveillance procedures, the
,
licensee discovered that the closing times of the main steam. isolation bypass
!
valves'were not included in the main steam isolation actuation and-response
!
time test procedure. Further_ investigation revealed that the acceptance
l
criteria for closing times were included in the valve operability-test
r
surveillance procedure. However, these valves were. determined to be incorrect
-l
!
i
._
- _-
.
-. _
--
.-
. . -
-
.
.
_
-
.
-.
.
.
..
- .
t
.
'
-13-
.
,
k
-
for the main steam isolation bypass valves. The cause of the deficiencies was
inadequate procedure development and review. The failure to develop and
f
maintain adequate surveillance procedures was determined to be a noncited
violation of TS 6.8.1 requirements.
The licensee submitted LER 498/92-021 on
,
,
the incident.
l
In rt;ponse to the May 19, 1992, TS 3.0.3 entry event, the licensee committed
i
in LER 498/92-004 to perform a review of selected surveillance procedures to
ensure that they satisfy TS requirements.
The Surveillance Review Task Force
i
was developed to perform an in-depth review of the ESFAS and reactor trip
.
!
surveillance procedures for one train of one unit. As a result of this
review, two additional examples of deficient procedures were identified and
reported to the NRC in LERs 498/92-011 and 498/92-013. The scope of the task
force was expanded to review the time response testing procedures for the
i
ESFAS and reactor protection system. The licensee later discovered a set of
inadequate time response test surveillance procedures. This was reported in
i
LER 498/92-017. During this inspection period, a second example of inadequate
time response test surveillance procedures was identified, involving the main
3
j
steam isolation bypass valves.
,
i
Each of the four main steam isolation valves was provided with a 4-inch bypass
'
l
line and bypass isolation valve for main steam line warm up during secondary
plant startup. The bypass isolation valves were normally closed valves that
fail closed. The main steam and bypass isolation valves were designed to
,
automatically close on high containment pressure, low steam line pressure, and
,
high negative steam line pressure rate signals.
Following postulated
[
accidents, the isolation valves were designed to close to maintain a
controlled reactor cooldown rate, minimize the release of steam to the
,
containment (reverse flow protection), and limit the release of radioactive
!
material to the environment.
The vendar design analysis requires that the
steam line isolation must occur within 10 seconds of the initial event
3
-identified above.
{
On December 15, 1992, the Surveillance Review Task Force identified that
'
Procedure IPSP03-SP-0024, Revision 0, " Steam Line Isolation Actuation and
.
Response Time Test," did not test the main steam isolation bypass valves. -TS
!
Table 3.3-5, " Engineered Safety Features Response Times," lists the required
response times for steam line isolation as less than or equal to 8 seconds for
~
low steam line pressure and 7 seconds for high containment pressure.
There
>
was no response time limit required for steam line. pressure negative rate high
!
events.
The cause of the deficient procedures was determined to be inadequate
i
4
procedure preparation and review. The valves were not included in the
_
i
procedure, in part, because they were normally shut valves.
Additionally, the
reference documents used during procedure development were vague and did not
,
specifically include or exclude the bypass valves as requiring a response time
j
,
test.
l
During an operability review of the bypass valves, the licensee discovered the
[
valve operability test surveillance procedure had acceptance criteria for
'
valve closure times that was less conservative than TS allowed.
The
.
?
.
I
i
l
'
.
,_
_ _ - .
_,
.
._
.
_
- .i
.
.
.
4
.
-14-
acceptance criteria for Procedure OPSP03-MS-0001, Revision 0, " Main Steam
System Valve Operability Test," was 10 seconds, which was the required
isolation time for the main steam isolation valves. The main steam isolation
and bypass valves were supposed to shut 5 seconds after the receipt of a
closure signal, assuming the closure signal took 5 seconds to reach the valve
following initiation by a double ended break accident. An evaluation
confirmed that the closing times for the four Unit 2 valves were sufficient
but two of four Unit I valves did not meet the revised criteria. Adjustments
'
have since been made to the Unit I valves to obtain a closing time of less
than or equal to 5 seconds.
The licensee committed to revising the applicable
procedures by February 15, 1993, in LER 498/92-021.
The cause of deficient surveillance procedures was inadequate developmeat and
review of the original procedures that perform the steam line isolation
actuation and response time test and the valve operability. test. The failure
to comply with TS constituted a violation of HRC requirements. However, this
violation will not be subjected to enforcement action because the licensee's
efforts in identifying and correcting the procedures meet the criteria
specified in Section VII.B.2 of Appendix C to 10 CFR 2.
The violation was
licensee identified; it was reported to the NRC Operations Center within the
required time interval; it was identified as the result of corrective actions
for a similar problem; it will be corrected; and it was not a willful
violation.
3.5 Failure of a Component Cooling Water Valve to Open Durina Surveillance
Test (Unit 1)
On December 17, 1992, CCW Valve CC-MOV-0130 would not fully open when-required
during the performance of Surveillanco Procedure IPSP03-CC-0008, Revision 3,
" Component Cooling Water System Train IB Valve Operability Test." This valve
was the outside containment isolation valve for the IB Residual heat Removal'
Heat Exchanger CCW outlet li: 1.
Control room personnel issued SR 187973 to
troubleshoot the cause of valve not stroking. Maintenance personnel
discovered a broken lug in the valve motor operator.
The broken lug was
replaced and the motor-operated valve functioned correctly and was returned to
service on December 18, 1992.
On December 18, 1992, the inspector noted that, although an SR had been
issued, an SPR was not' initiated to investigate the root cause of the failed-
lug. After questioning by the inspector as to how a failure analysis would be
performed without an SPR, control room personnel issued SPR 921566. This was
considered to be an example of a problem which would not have received an
adequate root cause analysis wit' ,ut prompting by the inspector. This was
similar to a finding that was moie in the recent Operational Safety Team
Inspection (50-498/92-35; 50-499/92-35)-and was determined to be unresolved
pending further NRC inspection.
-
- -
-
-.
..
-
-.
-
..
..
.
.
.
-
,
.
!
,,
!
-15-
I
3.6 Secondary Plant Eouipment Concerns (Unit 2)
I
During the inspection period, the licensee continued to experience a' negative
!
performance trend in unit power maneuvers, pump trips, or pump failures to
!
start because of Unit 2 secondary plant equipment problems.
The problems -
i
encountered included four power reductions, one feedwater pump trip during
i
troubleshooting activities, two feedwater pump manual trips on. loss of flow,
and one startup feedwater pump failure to start. The licensee plans to
f
I
implement corrective actions, including replacement of the feedwater control
circuitry during the upcoming Unit 2 refueling outage. Additional management
't
oversight was needed because a negative trend has been noted in this' area of
!
plant operations.
!
On December 27, 1992, following the Unit 2 manual . reactor trip, the startup .
l
!
feedwater pump sailed to start because of low oil pressure. Three days later,
Steam Generator Feedwater Pump 22 was manually tripped when the plant
operators discovered the pump had no output flot;.
These two incidents were
,
described in detail in Section 2.1 of this repor L.
,
On December 31, 1992, Unit 2 power was decreased from 100 to 95 percent power
in preparation for removing Feedwater Heater 21B from service because of a
j
suspected tube leak. The power was reduced to provide a margin for potential
!
feedwater temperature changes.
SR lt.6593 was issued to conduct the repair of
>
the heater.
While performing maintenance on the. heater for tube leakage,
!
'
cracks were discovered in the partition joint'to the tube sheet.
Upon
i
excavation of the defective weld area, it was further discovered that the
i
affected weld w , not a full penetration weld, as indicated by the vendor
'
manual. Also during the repair of the heater, additional indications of weld
cracks were found on the interior inlet nozzle to the shell.
SPR 930215 was
!
issued to perform a root cause analysis of the unexpected findings. The
i
feedwater heater remained out of service at the end of the inspection period.
j
On January 11, 1993, Unit 2 power was reduced from 100 to 96 percent to allow
i
for the removal of Steam Generator Feedwater Pump 21 from service. An
'l
electrohydraulic control fluid leak had developed on the high pressure
s
governor valve. The startup feedwater pump was started and Pump 21 was
-!
secured. The power reduction was required to maintain steam generator levels
!
while securing Pump 21.
In accordance with SR MS-171023, the high pressure
governor actuator was replaced because of an extruded seal. On January 15,
i
1993, Unit 2 was increasing power following a TS required shutdown when
!
Feedwater Pump 21 was again placed in service. Four minutes after the pump
j
was placed in service, the plant operators noted there was no-indicated flow
q
at the pump outlet.
Pump 21 was immediately secured. The recirculation valve
i
had failed closed and Startup Pump 24 discharge pressure, which was higher
i
than feedwater Pump 21 discharge pressure, had forced Pump 21 flow to
>
essentially zero.
Pump 21 was subsequently returned to service the next day.
!
after recirculation valve rework and satisfactory completion of
l
postmaintenance testing.
-
!
$
'{
.
-
.
.
.
.
-
.-
-
,
.
i
.
-16-
l
t
i
On January 15, 1993, a power reduction from 48 to 30 percent was performed to
allow for maintenance to be performed on the feedwater Pump 23 control
circuits in accordance with SR 158153. The power reduction was initiated to
,
comply with the licensee's reactor trip prevention program policy. At that
!
time, only Startup Feedwater Pump 24 and Steam Generator Feedwater Pump 22
,
were in service. During the maintenance activity, feedwater Pump 22 tripped
!
while a technician was inserting a logic card into the Pump 23 control
[
circuit.
The licensee was aware that the feedwater pumps' control circuits
!
were electrically interconnected and a power perturbation on one circuit can
j
affect the other pumps' control circuits.
Although the pump trip was not'
expected, the licensee prepared for the possibility of a pump trip by reducing
!
reactor power to the capacity of the startup feedwater pump. The
implementation of the reactor trip prevention policy prevented a potential
reactor trip on a rapid reduction of feedwater flow to the. steam generators.
,
Following the Pump 22 trip, reactor power was manually reduced to 25 percent
i
to match steam flow to feedwater flow from Pump 24.
Feedwater Pump 22 was
4
returned to service the next day and Pump 23 was returned to service the
-
following day.
!
!
3.7 TS 3.0.3 Entry Because of Two inoperable Differential Temperature Loops
]
(Unit 1)
j
i
On January 5,1993, Unit I entered TS 3.0.3 when two loops of OTDT were
determined to be simultaneously inoperable. The cause of the event was
subsequently determined to be an inadequate startup testing procedure. The
'
failure to comply with TS 3.3.1 requirements was considered to be a violation
t
of the facility operating license.
l
On January 5, 1993, at 7:57 a.m., Excore Power Range Channel NI-44 was removed
_!
from service to conduct corrective maintenance on the rod stop bistables in
f
accordance with SR 168092. The bistables for the Loop 4 reactor coolant
i
system OTDT were placed in the tripped condition, in accordance with the
!
requirements of TS.
This action removed Loop 4 OTDT from service. While
performing a reactor coolant system loop differential termperature (Delta-T)
calibration evaluation, the licensee discovered that the Loop 3 extrapolated
100 percent trip value was outside TS 2.2.1 limits. Delta-T evaluations were
i
'
being conducted at various power levels to ensure the calculated 100-percent
Delta-T's fell within the TS limits. The licensee then declared the Loop 3
Delta-T inoperable at 10:33 a.m.
Since Loop 4 was already out of service for
corrective maintenance, Unit 1 entered TS 3.0.3 when Loop 3 was declared
j
At 10:58 a.m., corrective maintenance was completed on HI-44 and the OTDT
I
bistables were restored to their normal positions.
At 11:15 a.m., the Loop 3
l
bistables were placed in the tripped positions to comply with TS 3.3.1 and
3.3.2 requirements. At 11:28 a.m.,
Loop 4 Delta-T was declared operable and
Unit 1 exited TS 3.0.3.
Unit shutdown was unnecessary because the TS 3.0.3
entry was exited within the 1-hour time limit.
TS 3.3.1, Table 3.3-1, Jtem 8,
lists the. minimum number of operable channels for operation in Mode 1 as three
_,
of four for OTDT. With Loops 3 and 4 out of service, TS 3.0.3 entry was-
l
!
i
!
I
i
. ..
_ --
-
- _ - _ .
.
.
-
.
.
- . - .
.
!
.
!
!
'
17-
,
required. The failure to maintain the minimum number of channels operable in.
[
accordance with the requirements of TS 3.3.1.was considered the second example
of a violation of the facility operating license (498;499/9236-01).
l
,
The licensee concluded that the cause of the event was an inadequate
I
procedure. The core reload initial startup testing procedure did not provide
!
specific criteria for declaring a channel inoperable and for coordinating the
l
Delta-T evaluations with other cross-train activities.
Corrective actions
!
completed included recalibration of the Loop 3 Delta-T. The licensee
!
committed in LER 498/93-01 to:
(1) provide additional operator training on
j
the event; (2) revise the deficient procedure to provide specific guidance;
'
(3) perform an evaluation to consider the risks and benefits of setting the
Delta-Ts to a conservative value prior to unit startup following a refueling
i
outage; and (4) perform a generic implications review of the core reload
initial startup testing procedure.
i
3.8 TS 3.0.3 Entry Because of Two Inoperable Power Range Nls (Unit 1)
i
i
i
On January 9,1993. Unit 1 entered TS 3.0.3 when the licensee discovered that
two channels of Wer range NIs were simultaneously out of service.
The cause
'
of the event was that licensed operators failed to perform a daily channel
1
calibration within the TS allowed time frame. A contributor to the event was
.
licensed operator failure to record a key entry in the control room logbook.
The failure to perform a TS required surveillance was a violation of the
.i
facility operating license,
j
1
TS Table 4.3-1, " Reactor Trip System Instrumentation Surveillance
Requirements," Item 2.a, requires a power range neutron flux high setpoint
i
channel check each shift and a channel calibration each day.
According to
Note 2 of the table, the channel calibration was performed by making a
.
i
comparison of a calorimetric (secondary heat balance calculation) to excore
power indication above 15 percent of rated thermal power, then adjusting
excore channel gains consistent with calorimetric power if the absolute
'
difference was greater than 2 percent.
!
Procedure OPSP03-ZQ-0028, Revision 5, " Operator Logs," was used to document
>
the TS required channel check once every shift. The four power range neutron
flux channels (NI-41, NI-42, NI-43, and NI-44) were visually inspected.
The
!
maximum difference allowed by procedures between operating channels was
2 percent. Procedure OPSP03-NI-0001, Revision 4, " Daily Power Range NI
l
Channel Calibration," was used to perform the TS required daily channel
>
calibration. The daily channel calibration was permitted to be performed in
one of two ways.
If the PROTEUS computer was online, Computer Point U1118,
l
'
Reactor Total Thermal Power, was used to determine the calculated percent'of
rated thermal power _ (actual power level at that time).
If the PROTEUS
computer was off line, the percent of rated thermal power was calculated based
on as-found secondary parameters. Once the percent of rated thermal power was
'
determined, the Nis were compared to the percent of rated thermal power value.
The difference between the as-found indications and the percent of rated
j
thermal power was then calculated. Accordint; to TS Table 4.3-1, the
l
!
-
.
- ,
.. -
-
-
-
--
_ _
_
_
.
_
_ _ _ _
,
_.
_ __ _ _
.
,
5
!
h
-
-
.
-
.
-18-
p
t
calculated difference must be less than or equal to 2 percent, otherwise the
nuclear instrument channels were required to be adjusted.
-
Procedure OPSP03-NI-0001 acceptance criteria was more restrictive. According
to the procedure, the calculated difference cannot be greater than 1 percent.
On January 7, 1993, at 9:25 p.m., with Unit 1 at 73 percent power, the daily
j
channel calibration was completed in accordance with Procedure OPSP03-NI-0001.
t
No adjustments were determined to be necessary. At 1:22 p.m. the next day,
,
NI-43 was removed from service and a quarterly channel calibration was
j
commenced on NI-43 in accordance with Procedure OPSP05-NI-0043A, Revision 2,
"NIS Axial Flux Difference Calibration (N-0043A)." At 10:20 p.m. on
January 8, 1993, the daily channel calibration, Procedure OPSP03-N1-0001, was
,
performed on the remaining three channels (NI-41, NI-42, and NI-44). At
12:50 a.m., January 8, 1993, NI-43 was returned-to service, without having
_
received a daily channel calibration. A few minutes later, NI-41 was removed
l
from service for the performance of Procedure OPSP05-NI-0041A, Revision 2,
j
,
"NIS Axial flux Difference Calibration (N-0041A)." A channel check of NI-43
3
had not been performed prior to the channel being returned to service, and the
!
TS allowed outage period (including the 25 percent grace period in accordance
[
with TS 4.0.2) expired at 3:25 a.m. on January 9,1993.
i
On January 9,1993, at 7:15 a.m.,
the licensee discovered that the daily
channel calibration for NI-43 had been missed. NI-43 was immediately declared
,
inoperable. With N1-41 out of service for a calibration, TS 3.0.3 was entered
- '
because TS 3.3.1 lists three as the minimum number of channels required to be
operable in Mode 1,
NI-43 was declared operable at 7:38 a.m. the same day
l'
when the daily channel calibration was satisfactorily completed. TS 3.0.3 was
exited at that time. NI-41 was returned to service at 8:15 a.m. following the
,
satisfactory completion of Procedures OPSP05-N1-0041A and OPSP03-NI-0001.
Unit shutdown was not required because TS 3.0.3 was exited within I hour of
i
'
its entry.
i
!
The cause of the event was failure of the control room operators to perform
~
the daily channel calibration on NI-43 within the required-time interval. A
contributor to the event was the failure of the control room operators to
l
follow procedures.
Procedure OPOP01-ZQ-0022, Revision 1, " Plant Operations
Shift Routines," Section 6.4, Control Room Logbook, lists the events and
evolutions which shall be entered in the control room logbook.
Step 6.4.2.6
!
states that logbook entries were required when commencing or completing any
l
procedure used to satisfy TS surveillance requirements which requires the
!
shift supervisor's permission to perform.
Procedure OPSP03-NI-0001,
s
Step 6.1.2, stated that to obtain the shift supervisor's signature on the data
l
package coversheet, granting permission to start testing.
The performance of
i
the daily channel calibration in accordance with Procedure OPSP03-NI-0001 was
1
not logged in the control room logbook when it was performed at 10:20 p.m.,
January 8, 1993.
If it had been logged, or if an operability tracking log
!
number had been assigned to N1-43, the missed surveillance on NI-43 would not
have occurred.
The failure to perform the TS required channel calibration on
i
NI-43 as required by TS Table 4.3-1, item 2.a. was considered the third
'
example of a violation of the facility operating license (498;499/9236-01).
l
!
-
,
i-i
-
, . - . _ . - . -.- --
.
--
y
.-
..,.r
~= .
--%
v.m
.
!
..
l
-
t
..
!
..
.
-19-
'
i
!
!
The licensee subsequently considered the event to;not be reportable.
During
l
approximately 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />, 3:25 a.m. to 7:15 a.m., January 9, 1993, Unit I had
both NI-43 and NI-41 out of service, the plant operators occasionally verified
)
that all four channels were reading consistent values (less than 2 percent
!
difference). However, the NRC inspectors determined that the use of a channel
check was not equivalent to the TS required daily channel calibration in that
the comparison of Channel NI-43 to_ the other channels did not meet the intent'
of a channel to'a calorimetric value comparison. The comparison of one or
,
more channels to the calorimetric, followed by the comparison of one channel
-
to another, could result in a channel deviation greater than the allowed
,
2 percent. The licensee plans to subsit LER 498/93-02 on the TS 3.0.3. entry.
3.9 Conclusions
Several examples of failure to satisfy TS requirements were identified and
were considered violations of TS.
These examples included a failure to
!
perform a supplemental containment purge system valve surveillance test, the
failure to maintain the minimum number of 0 TDT channels operable, and the
i
i
failure _to perform a daily NI' channel calibration. A TS 3.0.3 entry was.
required because of a deficient procedure.
A daily channel calibration check
i
of a NI was missed, in part, because a key entry was not recorded in the.
control room logbook.
Unit I startup following the refueling outage was delayed because 'of a' CRDM
leak. During the unit cooldown to allow for the repair of the leak, a PORV
failed to operate on demand. Both components were reworked and subsequently
returned to service.
EDG experienced a valid failure as a result of an inadequate PM procedure
which failed to adequately lubricate the fuel linkage assemblies. The fuel
lever arm was .;ot included in the procedure for lubrication and, as a result,
the lever arm became bound during an operability run of the EDG. This
condition resulted in excessive fuel to one cylinder which resulted in
cylinder overheating.
This inadequate PM was a further example of less than
adequate maintenance of safety-related equipment.
During the review of selected surveillance procedures, a group of deficient
procedures was identified.
The procedures were determined to be noncited-
violations of TS 6.8.1 requirements. The number of self-identified procedures
were a concern to the inspectors. The licensee should expand the scope of the
surveillance procedure review task force to ensure all surveillance procedures -
were-technically adequate.
Following the discovery of a damaged lug in the motor operator for a CCW.
valve, the licensee failed to generate an SPR to analyze the cause of the
deficiency. The SPR was issued after prompting by the inspector.
The failure
to issue an SPR was previously addressed in the operational safety team
inspection report.
.
.
.
.
-20-
The licensee continued to experience problems with secondary equipment, which
resulted in power maneuvers.
The number of events being experienced were a
continuation of a negative trend. Additional management oversight was needed
in this area of plant operations.
4 MONTHLY MAINTENANCE OBSERVATIONS (62703).
Selected maintenance activities were observed to ascertain whether the
'
maintenance of safety-related systems and components was conducted in
accordance with approved procedures, TS, and appropriate codes and standards.
The inspector verified that the activities were conducted in accordance with
approved work instructions and procedures, that the test equipment was within
the current calibration cycles, and that housekeeping was being conducted in
~
an acceptable manner. All observations made were referred to the licensee for
appropriate action.
l
4.1
Source Range Nuclear Flux Monitor Repair (Unit 1)
5
During the inspection period, problems with a neutron flux source range
monitor continued to be identified by the licensee.
Source Range
t
Monitor NI-31 has been intermittently inoperable, for a variety of reasons,
since the Spring of 1992 (refer to NRC Inspection Reports 50-498/92-14;
50-499/92-14 and -92-21).
Between October 1992 and January 1993, one card in
the NI-31 circuitry has malfunctioned three times.
Following the third
,
failure and replacement of a second card in the control circuitry, an SPR was
issued to investigate the failures.
The inspector witnessed the replacement
of the two cards and card calibration conducted on January 14, 1993.
Root
cause analysis of the card failures were incomplete at the end of the
inspection period. The inspectors will continue to monitor the reliability
and operability of Source Range Monitor NI-31 and_the licensee's actions to
!
resolve this and other technical problems with the monitors.
Two source range N1 channels were provided in each unit. The monitors (NI-31
'
and NI-32) were designed for use in detecting reactor neutron flux levels
during shutdown and initial phases of reactor startup. During the Unit 1 core
of fload on October 3,1992, the licensee noticed NI-31 was indicating zero
counts per second. The core load supervisor was notified and core alterations
y
were secured pending resolution of the problem. NI-31 was declared inoperable
and troubleshooting commenced in accordance with SR HI-173381.
Driver
Assembly Card NM103, which was used to amplify the input signal and match to
impedance for input into log Pulse Integrator Card NM104, was found to be
defective and was replaced. Monitor NI-31 was returned to service the same
,
day and core alterations-were resumed.
On December 17, 1992, NI-31 was again declared inoperable when the count rate
l
was noted'to have failed low.
SR N1-187603 was issued to rework the monitor.
Driver Assembly Card NM103 was again replaced and the Monitor NI-31 was
returned to service the next day. On January 12, 1993, during the Unit 1 TS
required shutdown, Monitor NI-31 failed to energize.
SR NI-182174 was issued
l
to troubleshoot the failed monitor.
Preliminary investigations revealed that
i
>
_
..
.
..
.
.
.
.
--
-.
-
^
.61
,
.
i
.~
-21-
l
~
i
!
!
i
Driver Assembly Card NM103 had failed for the third time. An SPR was written
i
by an instrumentation and controls technician to comply with the requirements
1
of Procedure OPGP03-ZX-0002, Revision 0, " Corrective Action Program." The
l
repetitive failures of the driver card assembly were determined to be a
j
l
On January 14, 1993, Monitor NI-31 was reworked'in accordance with
i
SR NI-182174. Driver Assembly Card HM103 and Log Pulse Integrator Card NM104
'
were replaced. The log pulse integrator card was replaced because the
)
licensee suspected it was causing the premature failures of the driver
j
assembly cards.
The old driver assembly card was inspected and a failed
resistor was identified. The old log pulse integrator card was also inspected
and a small amount of corrosion was. identified on a capacitor that was
attached to the card. This corrosion was evaluated to be indicative of
t
capacitcr leakage. The licensee planned to analyze the two cards onsite to
l
determine why the driver assembly cards were failing.
The analysis was
1
incomplete at the end of the inspection period.
The log pulse' integrator card
was recalibrated in accordance with the applicable Surveillance
I
Procedure '0SP05-NI-0031, Revision 1, " Source Range Neutron Flux Channel I
Calibrati
Monttor NI-31 was subsequently returned to service the same
"
.
day.
The inspectors considered the failure of station personnel to generate an SPR
concerning the repeated failures of the driver assembly cards a continuing
weakness that has been addressed in previous NRC resident inspector reports
and the recent operational safety. team inspection-(50-498/92-35;
50-499/92-35).
4.2 Repair of Cracked Pipe Weld (Unit 1)
During the inspection period, a cracked weld was identified and repaired by
the licensee. The cause of the crack was not clearly identified by the end of
the inspection period.
The inspector witnessed portions of the repair,
including ultrasonic testing.
The work that was witnessed by the inspector
conformed to station requirements.
On December 18, 1992, during a monthly quality control inspection of the
essential cooling water (ECW) system, a through-wall seep leak was identified-
on a pipe that provided cooling water to EDG 13
The indication was in a shop.
weld between a cast tee and pipe spool on Line EW-1307-WT3.
An ultrasonic
test was performed the same day; however, no flaw was observed during the
test. SR EW-161574 was issued to obtain a defect sample piece by drilling a
1 1/4-inch diameter hole at the point of seepage and retaining'the plug for.
analysis. Three days after the discovery of the indication, the plug was cut
out of the pipe. The crack was then identified in the plug.
.j
!
Additional ultrasonic testing was performed, and a crack about 4 inches.long
was identified.
The crack traversed through the heat affected zone in the
casting and into the weld itself.
SR EW-161574 was revised to provide
instructions to grind out the crack and reweld the excavated area. Quality
'!
,
i
- - .
.
..
.
..
.
-22-
controls representatives inspected and accepted the penetrant testing results,
the root pass weld, and weld cleanliness. A plug was fabricated and then
installed in the hole that had been drilled.
SR EW-161574 was again revised
to require the removal of an indication at the toe of the plug weld. :The
indication was removed and minimum through-wall-thickness was maintained. The
postmaintenance test, which involved an inservice leak test for 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> at
system normal operating pressure, was satisfactorily performed. The Train C'
,
ECW system was returned to service December. 24, 1992.
The plug sample was sent offsite for further analysis. .The results were not
available at the end of the inspection period; however, the licensee suspects
the crack was caused by a root defect during the original shop installation of
the weld. Additionally, it was determined that the weld was repaired in the
shop prior to field installation. The condition of dealloying, combined with
crack propagation, may.have contributed to the defect.
A Justification for Continued Operation (JCO) had previously existed for1the
_i
ECW system to allow for continued system operation with cracks and dealloyed
metal in the piping. JC0 91-2/3, Revision 3, "Through Wall Cracks in ECW Pipe
Welds and Dealloying in Weld Repairs in Tees," was used as a technical basis
for issuance of a conditional release prior to the weld repair. Revision'of
'
JC0 91-273, or the development of a.new JCO, may be required because the crack
that was found may not conform to the scope of th? JCOs that currently exist
,
at the South Texas Project. .The decision to revise'the JC0 will be made when
,
the final laboratory analysis results were made available. This issue will
'
continued to be tracked for future resolution as Inspection Followup
Item 498/9236-03.
,
'
4.3 Postmaintenance Testing of an EnG (Unit 1)
!
t
During postmaintenance testing of an EDG, the machine tripped when a' valve was
manipulated. The event was apparently caused by the incomplete venting of
lube oil piping following maintenance to eliminate oil leaks. An SPR was
-l
issued to investigate the event. The event was not reportable to the NRC
,
because the maintenance run was classified as a "No Test" start.
i
As part of a routinely scheduled Train B maintenance outage, EDG 12 was
removed from service on January 12, 1993. One of the work activities
performed was SR 153353 to repair leaks on the turbocharger lube oil duplex
!
filter three-way isolation valve and associated piping.
The work consisted of
i
disassembling, cleaning, and inspecting the valve, ' removing the old 0-rings,
..
and reassembling the valve with three new 0-rings.
One 0-ring was found to be.
!
missing; however, the licensee concluded that the missing 0-ring had no effect
'
on EDG operability.
SPR 930116 was written to perform an investigation as to
why the 0-ring was missing. .The investigation of the incident was incomplete
. I
at the end of the inspection period.
i
!
On January 14, 1993, following the scheduled work activities, EDG 12 was
t
started in emergency.at 6:07 a.m. for postmaintenance testing. The
,
postmaintenance testing scheduled to be performed included an operational leak
i
,
-
- - - - - -_---
-
-
.
--
.~.
.
e
.
'
-23-
1
!
i
check and valve operability test. The maching was released from emergency and
~
synchronized to the grid about 8 minutes later.
Step 3.8 of the.SR 153353
provided instructions to verify that the fluid levels (in the filter) had
i
returned to normal. The performer of the step signed off the step with a
note, " oil level satisfactory with engine running," at 6:20 a.m.
At
j
6:38 a.m., the EDG 12 tripped on low turbocharger lube oil pressure while the
-i
filter three-way valve was being shifted for component postmaintenance
.
testing. The lube oil filter was again vented 'and air was found in one of the
'
two filters.
Further licensee investigation indicated that the low lube oil
pressure signal was probably the result of a pocket of air. in the system
piping. The machine was restarted about 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> later with no problems being
encountered.
The inspector witnessed the second postmaintenance test run of
t
the EDG. The EDG operability test was satisfactorily completed and the EDG
!'
was returned to service the same day.
The event was classified as a "No Test"
.
because the start was for maintenance and the machine was administratively out
j
of service.
,
f
4.4 Toxic Gas Monitor Operability Events
a
During the inspection period, numerous problems with the plant's toxic gas
!
monitors (TGM) were experienced by the licensee because of equipment
i
!
malfunctions. Two violations of TS were identified.
One violation involved
[
the failure to maintain an out-of-service channel in the tripped condition.
,
The second violation involved the failure to perform a channel check because
,
'
displayed data was not recognized as being erroneous data. The Unit 1
monitors were placed in service prior to the monitors actually being ready for
r
operation, which was the result of a weakness in design implementation. A
[
control room alarm was erroneously received; however, no formal'
i
troubleshooting was performed to identify the source of the problem.
Power
i
supply problems caused two monitors to unnecessarily remain out of service for
r
extended periods of time.
The NRC inspectors performed a postmodification
'
equipment walkdown in Unit I and witnessed postmaintenance testing in progress
~;
during the walkdown.
TGMs were installed in the control room envelope ventilation system outside
air intake duct to detect the presence of harmful chemicals in the event of an
j
onsite or offsite chemical spill accident.
The system was designed to detect
d
six hazardous chemicals to provide an adequate level of protection for control
!
room personnel. Two of the chemicals detected, vinyl acetate and anhydrous
ammonia / ammonium hydroxide, were TS related while the remaining four chemicals
.
(hydrogen chloride, acetic acid, naphtha, and acetaldehyde) were not governed
!
by the requirements of TS. During the recent Unit I refueling outage, the
l
TGMs were replaced and the actuation logic was revised. The Foxboro monitors
j
were replaced with Extrel monitors and the actuation logic was changed from
.;
one of two to two of three.
This change was made in an attempt to improve the
!
reliability of the system and to reduce the number of spurious actuations of
.
the TGM system. A similar modification was scheduled to be implemented during
-
the upcoming Unit 2 refueling outage. The licensee committed in
LER 499/92-009 to perform a failure mode and effects analysis on the new TGMs
l
3
to identify problems with the system.
One of the purposes of this analysis
i
i
f
,
-.
,,
- -
-.
-
.- ..
-
-
-
-
-
!
.
-
.
'
-24-
_
,
was to identify a validation process for transmission of data within the
system. The analysis was scheduled to be completed in March 1993 and
corrective actions will be developed, as necessary, based on the results of
-l
the analysis.
'
4.4.1
Unit 1 Toxic Gas Monitors
'
On November 23, 1992, TGM XE-9326 was removed from service because of an
electrically noisy power supply. The channel was placed in the tripped
condition on November 28, 1992, to comply with the actions required in
i
A false high toxic gas signal was programmed into the monitor
!
software, which caused the trip signal to be electronically sealed in.
During
a control board walkdown on December 9,1992, the control room operators
noticed that the expected alarm, " toxic gas alert," was not present.
Channel XE-9326 was found to be reset and a technician immediately tripped the
channel. NRC was notified of the potential TS 3.3.3.7 violation, failure to
i
maintain an inoperable channel in the tripped position, on the same day as
,
'
discovery. On December 12, 1992, Monitor XE-9326 was again found in'the reset
condition after being placed in the tripped condition 3 days earlier.
The
channel was then placed in the trip condition by determinating leads under the
i
'
lifted lead program to prevent inadvertent resetting of the channel.
The cause of the problem was determined to be a less-than-adequate TGM design.
Momentary losses of power were permitting a seal-in relay in the control
circuitry to reset itself. A contributing cause of the event was less-than -
adequate design implementation. Corrective action consists of installing
'
three-way switches to ensure positive control of the monitors' trip function.
This modification was scheduled to be implemented in Unit 1 by June 1993 and
{
in Unit 2 by the end of the next refueling outage.
The failure to comply with TS 3.3.3.7 operability requirements of TGM XE-9326
,
on December 9 and 12, 1992, was determined to be the fourth example of a
!
violation of the facility operating license (498;499/9236-01)..
The safety
i
significance was determined to be minimal because the other two channels,
XE-9325 and 9327, were operable or the control room ventilation system was in
l
the recirculation mode of operation during the time frame.
,
!
On December 18, 1992, the Unit I control room received a " control
!
room / electrical auxiliary building intake high toxic gas alert" alarm.
Monitor XE-9325 indicated a high level of vinyl acetate.
All other monitor
j
readings for this and. the redundant Monitor XE-9327 (XE-9326 was still out of
service) were normal. The control room ventilation system was already in
recirculation, therefore, the Monitor XE-9325 actuation had no effect on the
system. SR 187970 was generated to troubleshoot the cause of the alarm.
'
Several hours later, the control room received a second " control
room / electrical auxiliary building intake high toxic gas ESF actuation" alarm.
l
Monitor XE-9327 was indicating that a high level of vinyl acetate was present.
The Unit 2 TGMs did not display any abnormal conditions.
SR 187974 was
I
written to troubleshoot the cause of the second alarm.
l
,
I
.
m
.-
.
-25-
Investigations by the licensee revealed that the TGMs installed in Unit 1-have-
three basic modes of operation:
analysis, calibration, and sequence modes.
The monitors were designed to automatically sequence through a calibration'
cycle approximately once every 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />, then restore to the analysis mode.
Automatic calibration consists of analyzing a sample source of gas and then
using the results to automatically set the analyzer's background intensity,
fragmentation patterns (a method of correcting for interferences at a
particular ion mass), and component sensitivities. Air from the breathing air
system had been previously used durinq the calibration process. On
December 18, 1992, the licensee switched to manufactured air, bottled with
known quantities of components, for use during the 20-minute calibration
process. When the monitors recalibrated themselves to the higher quality
i
bottled air, the monitors became more sensitive to the gas levels in the
,
'
process air.
Small quantities of krypton in the process air was mistakenly
recognized as vinyl acetate because of similar fragmentation patterns
'
resulting in an inadvertent setpoint for vinyl acetate being reached. A plant
c h nge form wr issued to add krypton to the analyzer software.
Both monitors
!
wet
mturned to service by December 24, 1992.
On December 31, 1992, during the performance of a routine channel check,
Monitor XE-9325 failed the channel check. The concentration level of vinyl
acetate was found to be reading high on the monitor, when compared to the
redundant Monitor XE-9327. Again, the source of the problem was the use of
l
'
bottled air during the calibration process. The licensee then replaced the
bottled air with breathing air. The breathing air was the air of choice
because it most closely resembled the process air being monitored,
t
Monitor XE-9325 was returned to service on January 6, 1993.
l
!
On January 10, 1993, the Unit I control room received a momentary " control
room / electrical auxiliary building intake toxic gas monitors two in bypass"
alarm on the emergency response f acility data acquisition display
system (ERfDADS) computer. The control room operators then placed the control
!
room ventilation in recirculation to comply with TS 3.3.3.7 requirements. An
f
investigation of the two operable monitors (Monitor XE-9326 was still out of
service) showed no cause for the two monitors in bypass alarm.
The alarm
cleared a few minutes later and did not repeat itself, however, the control
.
room ventilation system remained in the recirculation mode as a precaution.
Neither an SR or SPR were written to investigate the cause of the spurious
computer-driven alarm, therefore, the cause of the alarm was not identified.
Monitor XE-9326 remained out of service at the end of the inspection period
because of noise in the power supply.
The noise was causing the monitor to
erroneously actuate.
Corrective actions being considered included the
installation of an isolation amplifier, signal conditioner, or fixed ground in
the power supply to the monitor. With Monitor XE-9326 out of service, the
l
Unit 1 monitors were left in a one of two logic schematic.
!
!
I
I
f
,
--.
!
.
l
.
!
'
!
.
-26-
j
?
!
'
4.4.2
Unit 2 Toxic Gas Monitors
On December 2,1992, TGM XE-9325 was declared inoperable because the monitor
I
experienced a loss of communications with the ERFDADS computer.
SR HE-186916
'
was issued to troubleshoot and repair the monitor. Maintenance personnel
reset the analyzer and normal communications with the computer resumed.
t
Components in the TGM were replaced using parts from a spare unit.
The
i
postmaintenance test was satisfactorily performed and the monitor was returned
i
i
to service on December 8, 1992.
The technicians suspected that the cause of
the problem was with the power supply. A detailed investigation of the 120
volt AC power supply concerns was not performed under this SR.
On December 10, 1992, the Unit 2 control room received an " electrical
'
auxiliary building control room intake high toxic gas non-ESF actuation"
alarm.
The operators immediately placed the control room ventilation in the
toxic gas recirculation mode of operation.
The ERfDADS computer trend
,
indicated a spike for hydrogen chloride and naphtha.
The' Unit 1 monitors and
the redundant Unit 2 monitor trends were normal. The alarm clecred about
i
3 minutes later.
SR HE-171022 was issued to troubleshoot the monitor.
Later
the same day, Monitor XE-9325 again lost communications with ERFDADS.
SR HE-186920 was issued to troubleshoot the cause of communication failure.
The work instructions that were included in SR HE-171022 were added to SR
HE-186920, and SR HE-171022 was subsequently voided.
During troubleshooting activities in SR HE-186920, the licensee determined
that the cause of the microprocessor problem was the power supply, as
suggested previously in SR HE-186916. TGM XE-9325 was supplied with power
from nonsafety-related Distribution Panel DP-001. This panel was an
ungrounded power supply because it utilizes a battery bank as one source of
backup power.
One of the loads on DP-001 apparently had a ground which was
causing the power supply to Monitor XE-9325 to be grounded.
The voltage
regulator for Monitor XE-9325 was subsequently rewirea to allow the secondary
side of the regulator to operate ungrounded while the primary side remained'
grounded. A post-maintenance test was satisfactorily performed and the
monitor was returned to service on December 19, 1992.
During the period that Monitor XE-9325 was out of service because of the
grounded power supply problem, SR HE-168802 was issued to attempt to locate
the source of the ground. The ground was traced to an unidentified load on
the power supply panel. Operations attempted to locate and isolate the
grounded load by cycling the load supply breakers. As of December 17, 1992,
most of the breakers were cycled and the ground still had not been located.
The licensee could not cycle all breakers because of current plant conditions,
therefore, the remaining breakers were scheduled to be cycled during the next
refueling outage.
On December 10,1992. TGM XE-9326 was declared inoperable because of ERFDADS
communication problems. SR HE-185921 was issued to troubleshoot the cause of
the problem.
Garbled data was noted to be appearing on the ERFDADS host
monitor during the troubleshooting process. This condition was not
-
-
-
.
.
.
_ . - - - . -
_
.
. .
- .
.
,
.:
,
~
-
. nr
,
9-
1
.
-27-
j
,
.immediately ' identified because the system health screen was indicating that
. good data existed on the screen. A printed circuit board and a modem were
-
!
subsequently replaced. Monitor XE-9326 was returned to' service 2 days later.
SPR 921481 was issued on December 11, 1992, to document the discovery of the
"
host monitor problems.
,
On December 17, 1992, during the reportability review of the SPR,- the licensee
,
determined that the modem had been faulty since October 15, 1992. .In
,
accordance with.TS Surveillance Requirement 4.3.3.7, each chemical detection
,
system shall be demonstrated operable by performance of a channel check at.
-
least once every 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />. -This commitment was being fulfilled by'the.
'
performance of Procedure OPSP03-ZQ-0028, Revision 5,." Operator-Logs." The
?
channel check was performed on a shift basis.
Because of the modem failure,
t
plant operators were not able to meet the requirements of TS 4.3.3.7 because
the operators were not aware that the display output was erroneous. The
.;
safety significance was determined to be minimal since the two Unit 2. monitors
.;
remained in service or the control room ventilation system remained in the
recirculation mode of operation during the period in question.
t
The failure to comply with TS 4.3.3.7 requirements was the fifth example of a
f
violation of the plant operating license (498;499/9236-01). Corrective
actions taken included revising Procedure OPSP03-ZQ-0028 to provide
,
instructions that the channel check data be taken from the-analyzer printer ~
'
until the system modification was implemented during the upcoming Unit 2
refueling outage.
The licensee issued LER 499/92-009 on the event.
g
4.5 Conclusion _s
Source range Neutron flux Monitor-NI-31 was' discovered to be -inoperable twice.
f
during the inspection period. One card was replacedifor the third time. No
<2
SPR was issued to document these identified conditions adverse to-quality.
!
The monitor has been intermittently inoperable .for almost 1 year. .The
operability and availability of this monitor was of concern to the inspectors,
and the inspectors will continue to review' the licensee's actions to resolve
this and other problems with'the monitor.
A crack was found in the ECW piping and was repaired in a timely manner. The
- '
licensee continues to aggressively pursue the metallurgical problems
associated with the system piping.
t
!
An EDG trip signal was generated during a postmaintenance test run because of
-
inadequate venting of the lube oil support system.
The licensee continued to experience problems with the TGMs. .Two violations
!
of TS were identified, however, the violations were determined to be of
'
limited safety significance. Despite the efforts made by the licensee, the
reliability and availability of the TGMs has not improved.
,
,
i
'
-t
!
_ _ _ .
_,
._,
,
_
_.
.
..
.
,
<
..
,
'
-28-
i
.
'
5 BIMONTHLY SURVEILLANCE OBSERVATION (61726)
5.1
Logic Train Functional Test (Unit 2)
i
The Train S automatic trip and automatic actuation logic output functions of
the solid state protection system (SSPS) in Unit 2 were tested on January 5,
1993.
Problems were encountered with an automatic input function test
pushbutton during the performance of the surveillance test. This test was
successfully performed with no logic failures, however, the test pushbutton
,
was in need of repair or replacement. The licensee plans to rework the test
.!
I
equipment during the upcoming Unit 2 third refueling outage.
IS Table 4.3-1 requires that each logic train be tested at least once every
62 days, on a staggered test basis.
Therefore, the SSPS R or S train was
o
tested each month. On January 5, 1993, the SSPS Logic Train S was tested in
accordance with Surveillance Procedure OPSP03-SP-00055, Revision 2, "SSPS
Logic Train S Functional Test." To comply with the licensee's reactor trip
!
'
prevention program, the unit supervisor monitored the primary test performers'
'
actions. During the test, problems were encountered with the automatic input
'
function test pushbutton. The test performer had to depress the test
pushbutton multiple times for selected logic checks until the logic test for
each circuit was completed. The test was subsequently completed with no logic
,
test failures.
The licensee has experienced problems with the SSPS Logic Train S test
equipment since April 1992 (refer to NRC Inspection Reports 50-498/92-24;
50-499/92-24 and -92-32).
The licensee plans to initiate SSPS test equipment
,
!
repairs during the upcoming outage scheduled to begin in late February 1993.
Because of the potential for a reactor trip, the licensee decided not to
attempt repairs with the unit in operation.
Since the test was routine y
,
performed every other month, the surveillance procedure was not required to be
completed again prior to the unit outage.
5.2 Steam Pressure Lead / Lag Amplifier Calibrations
following the Unit I and 2 TS 3.0.3 entries on January 12, 1993, numerous
- '
recalibrations of steam pressure instrument loop lead / lag amplifiers were
performed to correct a discrepancy that was discovered in the calibration
procedures. Teams of plant personnel were quickly assembled to fully assess
the situation and to recalibrate the suspect amplifiers. The licensee
,
performed the recalibrations in a highly controlled manner. The failure to
develop and maintain adequate surveillance procedures was a violation
(noncited) of TS 6.8.1 (refer to Section 2.2 of this inspection report for
!
further information about the violation).
.
l
The ESFAS provides signals to various ESF equipment to take protective actions
needed to mitigate the consequences of postulated accidents, including loss of
,
coolant and main steam line breaks.
ESFAS instrumentation installed at STP to
protect against a steam line break include the compensated steam line pressure
low and the steam line pressure negative rate high channels.
.
L
. .
. .
i
n
rs
.
~
-29-
,
The steam line low pressure channels were used to generate a main-steam line
isolation (to prevent an uncontrolled steam generator blowdown) and safety
injection signal.
These protective functions were activated when two of.three
steamline pressure channels in any steam generator drop to 135 psig or below.
The control room operator has the capability to manually block the steam line
isolation function and safety injection actuation whenever permissive P-II,
set at a pressurizer pressure of 1985 psig, was enabled to allow for normal
plant cooldown and depressurization activities.
The steam line high pressure rate channels function to generate a main stecr.
line isolation signal only. The steam line high pressure rate sigt.al can be
manually enabled when pressurizer pressure was below the permissive P-II.
setpoint. The actuation of two of three channels in any steam generator,
concurrent with a pressurizer low pressure (P-ll) condition, will result in.a
main steam line isolation.
The nominal setpoint of the steam line pressure
negative rate high was 100 psig.
The high pressure rate steam line isolation
function was automatically disabled when pressurizer pressure increased above
the permissive P-ll setpoint and may have been manually disabled by the plant
operators below P-ll when the low main steam line pressure protection was not
blocked. The plant was designed such that one of the two steam line isolation
functions was enabled during all modes of plant operation.
Each unit has 12 steam pressure instrument channels (3 per main steam line).
Within each channel, there were two lead / lag amplifiers which function to
anticipate plant transient responses.
These lead / lag devices respond.^o e-t;;
of change. The lead portion amplifies the effect of the. rate of c6 enge whih>
the lag resists the change.
In proper combination, the two can .2r ensate te
some extent for process lags and provide early action for fast or sustau no
plant transients. According to TS Table 3.3-4, the time const ant s ut il t ic . in
s
the lead / lag amplifier for steam line pressure low were required te ne te' at
greater than or equal to 50 seconds (lead) and less than or <nua?
't
econda
1
!
(l ag) . Additionally, the time constant utilized. in the ratt Ng amplitse for
the steam line pressure negative rate high was greater than on equal to 50
seconds (lead function only).
Verification of these time constan' set t ings
were performed routinely during the refueling outages by the perfors..we of
selected surveillance procedures.
.
During procedure development, a tolerance was added to the time constants.
For example, the lead time constant tolerance was 50
5 percent or
,
2.5 seconds. The tolerance for the lag time constant was 5 i 5 percent, or
!
0.25 seconds.
During routine TS required surveillance calibrations, the time
constants were occasionally set below 50 seconds (lead) or'above. 5 seconds
(lag), but within the tolerance limits established in the procedure data
'
packages. Once the licensee became aware of the discrepancies between the TS
limits and the tolerance errors, a review of the most current lead / lag
!
amplifier time constant setpoints was performed. The licensee noted that a
i
significant number of lead / lag amplifiers were incorrectly set.
Out of the
24 amplifiers in Unit 1. 14 were incorrectly set.
In Unit 2, 19 of
24 amplifiers were incorrectly set.
l
l
[
f
r
I
-
..
.'
.
.
.
-30-
On January 12, 1993, both units entered TS 3.0.3 and a TS required shutdown
commenced because of the steam pressure instrument loop lead / lag amplifier
time constant' discrepancies. The minimum number of channels required to be
operable, in accordance with TS Table 3.3-3, was two of three in each main
steam line. The licensee had to restore two of three channels in each main
steam line loop, with the third in trip, prior to each unit being able to exit
A team of instrument and controls technicians was immediately-
assembled to recalibrate the suspect lead / lag amplifiers.
The. technicians
began the recalibration process in Unit 1.
The licensee utilized three teams
working in parallel. The number of teams was limited by the qualified
personnel and test equipment available to perform the calibrations.
The inspectors observed the work in progress. The work was being performed
using SRs to direct the technicians to perform applicable sections of
surveillance procedures. The surveillance procedures had not been revised to
correct the tolerance errors, however, a work supervisor was available at all
times to ensure the lead / lag amolifiers were being recalibrated to within the
TS limits.
The work supervisor, as well.as the plant operators, were closely
monitoring the channels being recalibrated to ensure that no work was being
performed on a different channel that could generate an actuation signal on
the two of three logic. All calibrations were finished in Unit 1 on
-
January 13, 1993, at 7:27 p.m.,
and in Unit 2 the next day at 4:20 a.m.
5.3 Conclusions
The SSPS logic Train S functional test was satisfactorily completed despite
the problems encountered with the test pushbutton. The operators performed
the surveillance test in a competent and controlled manner.
No concerns were
identified with this activity.
e
The discovery of the deficient steam pressure instrument loop calibration
!
procedures was indicative of the licensee's ability' to self-identify and
- l
correct procedure errors. The recalibration of the amplifiers was well
j
controlled by the licensee. The deficient procedures were identified as a
j
violation (noncited) of TS requirements (refer to Section 2.2 of this rep' ort).
!
!
6 PREPARATION FOR REFUELING - UNIT 1 (60705)
l
6.1 Unit 1 Fourth Refueling Outage Summary
i
The Unit I fourth refueling outage began on September 18, 1992, and was--
!
scheduled to last 62 days.
Prior to Unit I shutdown, an assessment of the
j
licensee's implementation of controls for refueling operations was performed.
1
As documented in NRC Inspection Report 50-498/92-26; 50-499/92-26, the
i
inspectors concluded that the refueling outage appeared to be well planned by
'!
the licensee.
Shutdown risk assessment activities and the planned use-of
~
shift outage managers and coordinators were considered positive initiatives.
1
However, the schedule appeared to be aggressive because of the number of
~
motor-operated valves-(MOV) scheduled to be tested.
.
1
l
U
.-
,
,..
,
-31-
The refueling outage ended December 31, 1992, when the main generator output
breaker was closed and the unit was tied to the offsite grid.
The outage
lasted 103 days. Delays were encountered, in part, because of a number of
unexpected equipment problems, including EDG rework, refueling equipment-
problems, rework of a reactor coolant pump lower motor be'aring, repair of a
leaking CRDM, and delays in the MOV testing process.
Items that were in the
baseline scope of the outage and were completed as scheduled included:
(1) 89
local leak rate tests; (2) 138 weld in-service inspections;-(3) 114 piping
support inservice inspections;-(4) 59 dynamic and 95 static MOV tests; (5) 68
modifications and engineering change notice packages; and (6) 187 piping
snubber functional tests. The total number of maintenance activities
scheduled, including SRs, PM, and surveillance tests, was 4591 activities, and
4568 of the activities were completed. A licensee self-critique was issued
following the outage.
The critique was an internal document that listed all
.
significant issues that impacted the outage, lessens learned from the outage,
.
and overall outage statistics.
The document was a proactive effort on the
i
part of the licensee to improve their outage performance.
{
!
The radiation exposure goal of 95 person-rem was not met, in part, because of
the length of the outage. The actual exposure was 135 person-rem.
l
Additionally, the goal of only 30 skin contamination events was exceeded by
.
f
8 events. From a personnel safety standpoint, there were no lost-time
injuries or restricted duty iN uries during the outage.
6.2 Conclusions
'
All major work activities were successfully completed during the outage.
The
overall performance of station personnel in the area of industrial safety was
,
excellent, in part, because of increased management and supervisory oversight
,
'
of the outage. Outage management continues to be a licensee strength as
evidenced by continuing improvement in reactor containment building
housekeeping and radiological controls.
A number of refueling equipment
related problems adversely effected the outage schedule.
Management oversight
should be increased in this area to prevent future outages from experiencing
unnecessary delays because of refueling equipment reliability problems.
j
7 FOLLOWUP (92701)
7.1
(Closed) Inspection Followup Item 498:499/8868-05:
Inadeauate Emergency
Lighting
During the emergency operating procedure (EOP) team inspection that was
performed in 1988, the inspectors were concerned with the lack of emergency
lighting that was available to support E0P required actions.
Various areas of
the plant had inadequate lighting to safely perform selected _ activities
required by the E0Ps during loss of all alternating current power events.
The
inspectors subsequently reviewed the status of the inspection followup item
(documented in NRC Inspection Report 50-498/89-15; 50-499/89-15) but left the
i
item open pending the completion of a corrective action plan to resolve the
emergency lighting discrepancies.
]
!
!
-
---
, _ _ _ . ___ _ _ ____ _
{f
...
,,#
.
.
-32-
-
A walkdown of the Units .1 and 2 E0Ps was performed by the licensee.in 1989.
Areas that required additional emergency lighting to assist plant operators'in
locating and operating equipment were identified. ' Modification
Packages89-158 (Unit 1) and 89-159 (Unit 2) were developed to install 30
8-hour battery packs in three buildings outside the reactor. containment
building to illuminate a total of 62 valves in each unit. The locations' were
picked to light the handwheels or actuators as applicable and to' allow the
operators to make positive identification of the valves during loss of power
events.
Modification Package 89-158 was installed in Unit I during the recent fourth
refueling outage.
The NRC inspector performed a walk.down of Unit I and
,
verified that the new emergency lights were installed and in service.
Modification Package 89-159 was scheduled to be implemented in. Unit 2 during-
the unit's third refueling outage, scheduled to begin in February 1993.
i
Additionally, standing orders exist that require reactor plant cperators to
- llways carry portable flashlights for use as needed.
'
This item is closed.
_,
j
7.2 STP Service Water Followup
The inspectors performed a followup inspection of deficiencies and
.t
observations identified by the service water operational performance
inspection which was conducted on site from June 22 through July 10. 1992, and
documented in NRC Inspection Report 50-498/92-201; 50-499/92-26.
l
'
Deficiencies were either the apparent failure of the licensee: _ (1) to comply;
with a requirement or (2) to. satisfy a written commitment or to conform to the
provisions of applicable codes, standards, guides. or other accepted industry
practices that have not been made legally binding iequirements. Observations-
!
were items considered appropriate to call to licensee mo.iagement att ytion
i
although they have no apparent direct regulatory basis.
7.2.1
(Closed) Deficiency (498:499/92-201-01):
Inadequate Heat Transfer
l
Testing program
.
The methodology used in Calculation MC-6219 to develop 'the acceptance criteria
for the essential chiller condensers and CCW pump supplementary coolers was
'
determined to be unacceptable.
The licensee has indicateo hat an alternate
method for demonstrating heat transfer capability for the essential chiller
)
condensers and CCW pump supplementary coolers will be evaluated.
A review of this issue indicated that the licensee has reviewed Electric Power
l
Research Institute (EPRI) NP-7552, " Heat Exchanger Performanct Monitoring
!
'
Guidelines." Based on these guidelines some changes to the current
performance monitoring of the ECW heat exchanger will be required.
Detailed
l
heat exchanger design information was not available for the essential chillers
l
or the CCW pump supplementary coolers; therefore, the only option remaining.
l
was a maintenance inspection program, which will include a combination of-
)
periodic visual inspection, cleaning, and water treatment to be used to
j
f
-
I
_
h
)
.g
s
q
...
-33-
.
l
l
monitor for fouling.
PM procedures will be developed to replace the heat
I
exchanger testing by December 31, 1992. Calculation MC-6219 will. be revised
to incorporate new acceptance criteria by February 1, 1994.
Deficiency 498;499/92-201-01 will be considered closed and reopened as
Inspector Followup Item 498;499/9236-04, pending completion of the PM
procedures and revision to Calculation MC 6219.
7.2.2
(Closed) Deficiency (498:499/92-201-02):
Valves Not Beino Included in
the Inservice Testing (IST) Program
l
'
A portion of the normal backwash flow path from the ECW self-cleaning strainer
in each train consists of buried piping, which was not constructed to seismic
Category I and ASME Code Class 3 requirements.
Failure of this section of
piping could result in failure of the self-cleaning strainer or loss of
essential cooling pond inventory. An emergency backwash line conforming to.
seismic Categorv I and ASME Code Class 3 requirements was provided. However,
the emergency Lackwash line discharges near the ECW intake structure e d was
normally isolated by an MOV to prevent the return of debris to the v..,nity of
the pump suction well.
Flow must also pass through a check valve installed in
the emergency backwash line. The inspectors determined that the MOVs (EWO277,
i
-278, and -279) and check valves (EWO403, -404, and -405) in each emergency
backwash line perform a safety function. These valves were not included in
the licensee's IST program as required by Article IWV-Il00 of ASME Section XI.
,
This issue was considered a violation of 10 CFR 50.55a requirements.
Deficiency 498;499/92-201-02 will be closed and reopened as
Violation 498;499/92-36-05.
7.2.3
(Closed) Deficiency (498:499/92-201-03):
Failure of the Licensee to
Request Relief from the ASME Code Requirements
1
The licensee completed an internal safety system functional assessment of the
ECW system on December 1, 1989. The safety system functional- assessment. team
determined that the ECW flow element installation for Units 1 and 2 CCW heat
exchangers was not in accordance with vendor recommendations, resulting in an
.
error of up to 10 percent in flow measurement readings. Subsequently, the
l
vendor determined actual instrument error to be no greater than 7 percent'.
l
This error results from an out-of-plane installation with respect to the last
I
upstream flow disturbance and an insufficient length of straight pipe upstream
of the instrument.
Annubars NIEW-FE-6853, -6863, and -6873 (Unit 1) and
N2EW-FE-6853, -6863, and -6873 (Unit 2) were used for measuring flowrate.
The inspectors determined that the licensee failed to note that the accuracy
of these instruments was insufficient to meet the requirements of
Article IWP-4110 of ASME Section XI and that a request for relief from the
provisions of Section XI was necessary.
This was considered a violation of
10 CFR 50.55a requirements.
-
_
_______ _ _ _ - _-______-_
_ _ _ - _ _ _ - _
/~
-
,
.
.
,
-34-
Deficiency 498;499/92-201-03 will be closed anJ reopened as
Violation 498;499/92-36-06.
7.2.4
Observations
Based on the inspectors review of the licensee's completed and proposed
corrective actions, the following observations are closed:
(Closed) Observation 498;499/92-201-01, " Design Basis Document"
(Closed) Observation 498;499/92-201-02, "De-alloying / Crack Acceptance
Criterion"
(Closed) Observation 498;499/92-201-03, " Procedural Weaknesses"
(Closed) Observation 498;499/92-201-04, "GL 89-13 Training"
(Closed) Observation 498;499/92-201-05, "ECW Flow Balance"
(Closed) Observation 498;499/92-201-06, " Independent Verification"
,
a
'
,
o
.
ATTACHMENT 1
1 PERSONS CONTACTED
1.1
Licensee Personnel
C. Ayala, Supervising Engineer, Licensing
R. Balcom, Manager, Nuclear Security
H. Bergendahl, Manager, Technical Services
S. Bollinger, Consulting Engineering Specialist, Corrective Action Group
M. Chakravorty, Executive Director, Nuclear Safety Review Board
K. Christian, Manager, Plant Operations
R. Dally-Piggott, Engineering Specialist, Licensing
J. Fast, Division Manager, Instrumentation and Control Maintenance
J. Gruber, Division Manager, Material Technical Services
S. Head, Consulting Engineer, Corrective Action Group
W. Humble, Plant Programs Manager
T. Jordan, General Manager, Nuclear Engineer
W. Jump, General Manager, Nuclear Licensing
M. Pacy, Manager, Design Engineering
G. Parkey, Plant Manager
R. Rehkugler, Director, Quality Assurance
S. Rosen, Vice President, Nuclear Engineering
D. Wohleber Director, Records Management System and Administration
1.2 Central Power and Light
B. Mclaughlin, Owners Representative
lhe personnel listed above attended the exit meeting.
In addition to the
personnel listed above, the inspectors contacted other personnel during this
inspection period.
2 EXIT MEETING
An exit meeting was conducted on January 19, 1993.
During this meeting, the
inspectors reviewed the scope and findings of the report.
The licensee did
not identify as proprietary any information provided to, or reviewed by, the
inspectors.