ML18153C743
| ML18153C743 | |
| Person / Time | |
|---|---|
| Site: | Surry  |
| Issue date: | 08/30/1991 |
| From: | Branch M, Fredrickson P, Tingen S, York J NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II) |
| To: | |
| Shared Package | |
| ML18153C742 | List: |
| References | |
| 50-280-91-24, 50-281-91-24, NUDOCS 9109240214 | |
| Download: ML18153C743 (8) | |
See also: IR 05000280/1991024
Text
UNITED STATES
NUCLEAR REGULATORY COMMISSION
REGION II
101 MARIETTA STREET, N.W.
ATLANTA, GEORGIA 30323
Report Nos.:
50-280/91-24 and 50-281/91-24
Licensee:
Virginia Electric and Power Company
5000 Dominion Boulevard
Glen Allen, VA
23060
Docket Nos.:
50-280 and 50-281
. Facility Name:
Surry 1 and 2 *
License Nos.:
Inspection Conducted: August 2 through 26, 1991
Inspectors:
Approved by:
Scope:
/J1 I{) &_~*6'( Mt/f;dk~k~
M. W. Branch, Senlior Resident Irfspector
-},( aJ !fl~ 1; !itt/~/n~
J. W. York, Residfnt Inspector'
.
P. E; Fredrickson, Secfl-On Chief
Division of Reactor Projects
SUMMARY
olr'-l.
~
This specfal inspection was conducted on site to evaluate the failure
of the No. 3 emergency diesel generator to * reach required speed to
automatically load (if needed) on its emergency bus during an August 2, 1991,
Unit 2 safety injection with a reactor trip.
The inspection was also conducted to evaluate the inoperability of the A
charging/high head safety injection pump under certain pump configurations.
Results:
One apparent violation was identified for failure to comply with Technical
Specification 3.16.B.1 while the No. 3 emergency diesel generator was
inoperable from May 9, 1991 to August 2, 1991 (paragraph 2) .
. A second apparent violation was identified for failure to comply with Technical
Specification 3.3.B.2, while the A charging/high head safety injection pump was
inoperable for numerous durations between 1979 and the present (paragraph 3).
9109240214 910830
ADOCK 05000280
G
1.
Persons Contacteq
Licensee Employees
REPORT DETAILS
- R. Allen, Acting Superintendent of Opetations
- W. Benthall, Supetvisor, Licensing
- R. *Bilyeu, Licensing Engineer
- R. Blount, Supervisor, Procedures
D. Christian, Assistant Station Manager
J. Downs, Superintenderit of Outage and Planning
D. Erickson, Superintendent of Health Physics
- R. Gwaltney, Superintendent of Maintenance
- M. Kansler, Station Manager
T. Kendzia, Supervisor, Safety Engineering
- A. Price, Assistant Station Manager
- R. Saunders, Assistant Vice President, Nuclear Operations
- E. Smith, Site Quality Assurance Manager
- T. Sowers, Superintendent of Engineering
NRC Personnel
- M. Branch, Senior Resident Inspector
- S. Tingen, Resident Inspector*
- J. York, Resident Inspector *
- Attended exit interview on August 19, 1991
- Attended exit interview on August 26, 1991
Other licensee employees contacted included control room operators, shift
technical advisors, shift supervisors and other plant personnel.
Acronyms and initialisms used throughout this report are listed in the
last paragraph.
-
2.
Review of EOG Operallil ity
During a Unit 2 safety injection with a reactor trip on August 2, 1991,
both the No. 2 and the No. 3 EDGs received a start signal and started as
required. However, the No. 3 EDG failed to increase to the required speed*
(i.e. the EDG's output breaker would not have closed if the EDG was needed
to 1 oad on to the bus). . The EOG was running at 835 rpm.
A speed of 870
rpm is required to satisfy the output breaker closure permissive
interlock.
The detai 1 s of the August 2, 1991, event are discussed in
Inspection Report 50-280,281/91-21.
The inspector identified to the licensee that the No. 3 EOG was not
running at the required speed.
The licensee subsequently declared it
inoperable and entered the TS 3.16.B.1 action statement.
This TS is
2
applicable to both Unit 1 and Unit 2 because the No. 3 EDG is corranon to
both units, and is the standby power supply for each unit's J emergency
bus.
The licensee investigated why the EDG fai"led to reach its required
speed .. The initial detennination indicated that the govern.or's speed
adjustment was not at the required setting.
This was determined by an
inspection that showed the gear match marks inside the governor were not
correctly aligned.
The licensee readjusted the governor speed setting and
perfonned two consecutive fast starts of the No. 3 EDG to verify that the
engine would reach the required speed. After testing proved that the No.
3 EDG was operable, the licensee exited the TS action sta~ement.
After
the No. 3 EDG test, the licensee satisfactorily verified the fast start
capability of the other two EDGs.
Fast start testing during refueling outages is required by TS 4.6.A.l.b,
However, monthly surveillance requirements for the EDGs as specified in
TS 4.6.A.1.a only requires start of the EDGs and to manually increase the
speed to 900 rpm.
This mqnthly test_ does not verify all of the governor
automatic functions as does the refueling surveillance test.
On August 7, 1991, the inspector witnessed the fast start of the.No. 1
EOG.
Procedure 1-0P-EG-6.1, Number 1 EDG, dated May 29, 1991, was used to
accomplish this testing.
Procedure Change (PAR 91-912) was issued to
modify the procedure to allow this special test. The procedure included a
fast start of the EOG and a speed verific~tion. During EDG operation for
this fast start, the speed was adjusted to 900 rpm and the governor speed
gear and control knob were marked for future reference.
The EDG was
stopped and two additional fast starts were accomplished to verify
repeatability.
On the first fast start after reference marks for speed
control were made, the engine automatically increased speed to 900 rpm as
verified by local rpm indicators and test equipment monitoring output
frequency~
After the mechanical speed adjustment was reset to the match
mark the second fast start was initiated.
The No. 1 EDG again automati-
cally increased lo the required 900 rpm speed.
Fast start testing of the No. 2 EDG was conducted after completion of
tfle No. 1 EDG test.
During the first fast start the speed increased
to approximately 928 rpm.
This as-found condition, was outside the
upper speed limit of 918 rpm and corresponded to a generator frequency
of 61.89 HZ.
The governor speed was adjusted to 900 rpm and the EDG was
~ubsequently fast started two additional times.
Each time the speed was
within the a 11 owab le 1 imi.ts.
The as-found speed of 928 rpm was eva 1 uated
by engineering and determined to be acceptable.
After each of the above tests, the licensee had to manually adjust th~
. governor upper speed setpoint.
The governor is equipped with two mkro
switches that control power to the governor servo motor.
T_he servo motor
is designed to automatically drive the upper speed adjustment to the
required setting that corresponds to 900 rpm.
It is considered that the
out-of-band speed for the No. 2 EOG indicates that speed is not repeatable
without the manual adjustment between EDG operation.
This type of *
governor (Woodward UG-8) is* generically used to control speed in many
nuclear applications.
3
The licensee initiated a CNS review of the No. 3 EDG failure.
The
inspectors met with the group and were provided with an initial root
cause for the EOG failure to reach the required speed.
The CNS group
determined that the governor had been_ replaced in the* May 1991 timeframe *
and that adjustments had been made without a subsequent fast start test.
The inspectors questioned the licensee as to the reportability of the EOG
failure.
The licensee indicated that reportability would be evaluated
after completion of review of the above issues.
The inspectors performed an independent review of the past maintenance of
the No *. 3 EDG. governor and examined the maintenance history of the EDG
to determine if recent maintenance contributed to the failure.
On May 7
through 9, 1991, the governor was installed in accordance with
The inspectors reviewed the work package associated with
this maintenance.
The licensee did not utilize a formal procedure to
accomplish this maintenance.
The maintenance was accomplished utilizing
job steps outlined on the WO and SNSOC approved instructions for adjust-
ment of the governor.
The governor vendor was a 1 so at the job site
to assist in the* adjustment of the new governor *. Review of the work
package indicated that the new governor was installed and that the speed
control was adjusted.
The EOG was started and governor adjustments were
made; however, the EDG would not properly 1 oad.
It was secured and
additional adjustments were made to the fuel racks. It was restarted and
loaded properly.
Another problem was encountered during installation of
the new governor when, during steady state operation of the EOG, engine
speed slowly increased.
With the EOG running, the vendor representative
for the .Woodward governor made adjustment to stabilize the speed.
The EOG*
was operated for approximately 40 more minutes with ~ngine speed remaining
stable. It was then secured and declared operable.
The licensee concluded.and the inspectors agreed that No. 3 EOG failed to
achieve the correct speed on August 2, because the vendor ma~e adjustments
on May 9, 1991, which affected the speed of the EOG during an automatic
start.
When the vendor made the adjustment, the licensee failed to
realize that engine speed would be affected during an automatic EDG*start.
The inspectors identified several other deficiencies during review of the
work package for replacement of the governor.
These deficiencies were not
directly responsible for the August 2, 1991~ failure of No. 3 EOG, but
they may have contributed to the failure.
The first deficiency was that
- the PMT follower sheet for this maintenance did not specify all the
required post-maintenance testing.
The SNSoc*approved instructions for
governor adjustment required that the EDG be fast started to verify
correct speed following governor replacement; however, the PMT follower
only required that the EOG be slow started. This issue was discussed with
the PMT coordinator who stated that this weakness had previously been
identified in the PMT program.
This weakness was the failure to specify
all applicable testing on the PMT follower when other instructions
associated with the maintenance required additional tests.
The licensee
was in the process of cor~ecting this weakness.
The s~cond deficiency was
that the EOG was not tested in accordance with the SNSOC approved
instructions.
4
The SNSOC approved instructions required that the EDG be fast started
following governor adjustment.
These instructions were signed off as
performing a. fast sp~ed start.
The inspectors* reviewed the operator
logs and PTs utilized to accomplish the EDG testing and concluded that
No. 3 EDG was *not fast speed started as required by the SNSOC approved
instructions.
The failure to fast start the EDG was attributed to poor
corrmunications between maintenance and operations. The third deficiency
was that a formal procedure was not utilized to accomplish this
maintenance.
The WO co.ntained 16 one line steps to accomplish this
maintenance and included a two page SNSOC approved document providing
instructions for governor adjustment and testing.
A lack of a formal
procedure contributed to poor communications* between operations, ma
intenance, engineering, and the vendor.
Confusion as to the requirement
for fast starting the EDG after the vendor made the governor adjustments
also resulted from the lack of a procedure.
TSs state that an EOG is required to start within 10 secon*ds and assume
load in less than 30 seconds after receiving a start signal.
The
inspectors concluded that since May 9, 1991, the No. 3 EDG was unable to
automatically operate in accordance with. TS time constraints and was
therefore inoperable.
It should be noted, however,* that operators could
have taken manual c~ntrol of the EOG, increased its speed, and aligned the
EDG to the emergency bus.
Because No. 3 EOG is a swing EOG, inoperability
of the EOG affects both Units 1 and 2 abi 1 ity to respond to an event
requiring the EOGs to operate and load automatically onto the emergency
bus.
The inspectors reviewed the work histories of the Nos. l.and 2 EOGs
from M~y 9, i991 through August 2, 1991.
The results of this review
indicated that on July 15, 1991, No. 2 EOG was inoperable from 9:28 a.m.
to 11:06 p.m. for installation of a balance weight on the drive shaft.
Unit 2 was at power on July 15, 1991, and if an event would have occurred
that required Nos. 2 and 3 EOGs to automatically start and load, both of
the EDGs would not have been available.
However, operators could have
operated both EDGs manually.
The -i nabi 1 i ty of Nos. 2 and 3 EDGs to
automatically start for 13 hours1.50463e-4 days <br />0.00361 hours <br />2.149471e-5 weeks <br />4.9465e-6 months <br /> and 38 minutes on July 15, 1991, was
beyond the scope of TSs.
TS 3.16.A requires that two EDGs, the dedicated EOG for each unit, and the
shared backup No. 3 EOG, be operable prior to exceeding 450 psig and.350
degrees F respectively.
TS 3.16.B.1 allows either the dedicated or shared
backup EOG to be inoperable provided the operability of the other EOG is
demonstrated daily and the inoperable EOG returned to an operable status
within seven days.
If this is not met, the unit is to be brought to a
cold shutdown condition.
The inspectors reviewed the operating history
for Units 1 and 2 over the period that the No. 3 EOG was inoperable (May
9, 1991,
to August 2, 1991).
The results of this review were that Unit 1
operated the entire period without complying with TS 3.16.B.1 action
statement.
Unit 2 operated June 1 and 2, June 5 through 11, and July 2
through August 2, 1991, without complying with TS 3.16.B.1 action
statement.
The failure to comply with this TS requirement was identified
as Apparent Violation 50-280, 281/91-24-01, Failure to Comply with TS 3.16.B.1 Requirements *With No! 3 EOG Inoperable.
5
Within the areas inspected, one apparent violation was identified.
3.
Rev*;ew Of Charging P.ump Operability
On August 19, 1991, the inspectors discussed the charging/HHS! pumps
interlocks with several system engineers.
During this discussion, the
inspectors questioned how the pumps sequenced on and off their respective
emergency bus during an undervoltage condition.
As a result of this
discussion, an engineering review of these interlocks was performed.
On
August 21, 1991, the inspectors were informed by system engineering that.
under certain charging/HHS! pump configurations, the A pump would lockout
on an undervoltage condition on its emergency bus.
Based on this
discussion, the inspectors concluded-that if an undervoltage occurred on
the H bus coincident with an ECCS actuation signal, the A pump would not
have started automatically.
There are three charging/HHS! pumps in each unit.
The A pump is powered
from the H emergency bus, the B pump is powered from the J emergency bus,
and the C pump is a swing pump that can be powered from the Hor J bus *. *
- TSs require that two charging/HHS! pumps be available in a unit when
critical, one pump powered from the H bus and the other pump powered from
the J bus.
When the A and B pumps' control switches are in the run or
automatic start positions, the C pump control switch is in the
pull-to-lock position, and the C pump J bus supply breaker is racked-out,
the configuration exists where the A pump would lockout on an H bus *
undervoltage condition.
This charging/HHS! pump configuration was
frequently*utilized in both of the units.
On initiation of an ECCS signal
in this configuration, coincident with a H bus undervoltage condition, the*
A pump would have locked out and not started automatically.
Operator
action would have been required to manually start the A pump.
The above charging/HHS! pump configuration was frequently utilized since
the implementation of Design Change No. 78~5378, dated September 19, 1979.
Prior to the implementation of this design change, all three charging/HHS!
pumps were normally aligned to automatically start upon receipt of an ECCS
initiation signal.
Upon receipt of an ECCS signal without an undervoltage -
on the H bus, all three charging/HHS! pumps would automatically start. If
an undervoltage condition developed on the H bus, the A pump would lockout
and the Band C pumps would continue to operate.
One of the results of
Design Change No. 78-5378, was to change pump configuration such that only
two charging/HHS! pumps, one off each emergency bus, were aligned to
automatically started upon receipt of an ECCS initiation signal. The pump-
. configuration change was required to ensure that the LHSI pumps could
maintain adequate NPSH to the HHSI pumps during the recirculation transfer
mode of operation and to conserve RWST inventory.
Upon discovery of the inadequate charging/HHS! pump configuration on
August 21; 1991, engineering initiated a deviation report. At the time of
the discovery, Unit 1 charging/HHS! pumps were aligned such that the A and
B pump control switches were in the automatic start and run positions and
..
.
6
the C pump control swi~ch was in pull-to-lock.
The A pump was decla~ed
inoperable and a six hour requirement to be in hot shutdown was enteredin
accordance with TS 3.0.1..
The pumps were rea1igned such that the A,
control switch was in the pull-to-lock position, B pump control switch in
run, and~ pump control switch in the automatic start position and the six
hour. clock was exited.
The Unit 2 charging/HHS! pumps control switches
were positioned such that the A pump was in pull-to-lock, B pump in
automatic start, and C pump in run and therefore did not require
realignment.* In addition, on August 22, 1991, at approximately 4:18 p.m.
the licensee made a one-hour non-emergency report in accordance with 10
CFR 50.72. Since automatic HHSI initiation is assumed in the design basis
accident in coincidence with a single active failure, the units would be
outside the design bases in certain pu~p configurations if credit for
manual operator could not be taken.
The deviation report stated that during a design basis accident which
includes an undervoltage on the H bus, the inadequate charging/HH~I pump
configuration concurrent with a single failure of the B pump would result
in the unit not having a HHSl pump available for automatic start.
As
previu~sly discussed in paragraph 2 of this report~ the No. 3 EOG was
inoperable from May 7, 1991 to August 2, 1991.
Review of operator logs
. indic*ated that the Unit 1 charging/HHS! pumps were in an inadequate
configuration from April 26, through May 26, and June 20 through August 2,
1991.
With the No. 3 EOG inoperable and the charging/HHS! pumps
inadequately configured, none of the Unit 1 HHSI pumps were available for
automatic start if a design basis accident would have occurred.
TS 3.3.B.2 states that two of the three charging pumps in a unit may be
out of service, provided inrnediate attention is directed to making repairs
and one of the inoperable pumps be restored to an operable status within
24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. If one of the inoperable pumps is not restored within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />,
then the reactor shall be placed in the shutdown condition. If one of the
inoperable pumps is not restored within an additional 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br />, the
reactor shall be placed in a cold shutdown condition.
The failure to
configure Units 1 and 2 charging/HHS! pumps since approximately 1979 such
that ~he A pump would automatically operate during a design basis accident
was identified as Apparent Violation 50-280,281/91-24-02, Failure to
Comply With TS 3.3~8.2 Due To Inadequate Charging/HHS! Pump Configuration.
The inspectors were informed by the licensee that operators were t_rained
on the A charging/HHS! pump lockout and were aware that the pump would
have to be manually started if a lockout did occur.
Also, a similar
finding was identified *at the North Anna Power Station in 1983.
LER
83-058/03L-O, dated October 5, 1983, reported a condition where the A
charging/HHS! pump would have locked out and failed to start automaticilly
in response to an ECCS initiation signal during an ~ndervoltage on:the H
bus.
The corrective actions implemented at the North Anna Power Station
in response to this LER were not implemented at Surry.
Within the areas in~pected, one apparent violation was Jdentified.
7
4.
Exit Interview
The inspection scope and results were summarize-d on August 19, -1991 and
August 26, 1991
with those individuals identified in paragraph 1.
The
following summary of inspection activity was discussed by the inspectors
during this exit.
Item Number
Apparent VIO 50-280,281/91-24-0l
Apparent VIO 50-280,281/91-24-02
Description and Reference
Failure to comply with the
requirements of TS 3.16.B.1
with the No. 3 EOG inoperable.
Failure to comply with the
requirements of TS 3.3;8.2 _
with the A chargin~/HHSI pum~
- *
The licensee acknowledged the inspection conclusions with no dissenting
- comments.
The licensee did not identify as proprietary any of the*
materials provided to *or reviewed by the inspectors during this
inspection.
5.
Index of Acronyms and Initialisms
EOG
F
GPM
LER
-NPSH
NRC
SNSOC
TS
CORPORATE NUCLEAR SAFETY
FAHRENHEIT
GALLONS PER MINUTE.
HIGH HEAT SAFETY INJECTION
LICENSEE EVENT REPORT
NET POSITIVE SUCTION HEAD
NUCLEAR REGULATORY COMMISSION
POST MAINTENANCE TESTING
PERIODIC TESTS
-REFUELING WATER STORAGE TANK
STATION NUCLEAR AND SAFETY OPERATING COMMITTEE
TECHNICAL SPECIFICATIONS
WORK ORDER