ML20071E013
| ML20071E013 | |
| Person / Time | |
|---|---|
| Issue date: | 02/28/1983 |
| From: | Massaro S NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD) |
| To: | |
| References | |
| NUREG-BR-0051, NUREG-BR-0051-V04-N5, NUREG-BR-51, NUREG-BR-51-V4-N5, NUDOCS 8303140681 | |
| Download: ML20071E013 (23) | |
Text
NUREC/BR-0051
,.(d, POWER REACTOR EVENTS g,M.f.IF /
United States Nuclear Regulatory Commission July-August 1982/Vol. 4, No. 5 Pow r Reactor Events is a bi-monthly newsletter that compiles operating experience information about commercial nuct:::r power plants. This includes summaries of noteworthy events and listings ar.d/or abstracts of USNRC and other documents that discuss safety-related or possible generic issues. It is intended to feed back some of the lessons learned from operational experience to the various plant personnel, i.e., managers, licensed reactor operators, training coor-dinitors, and support personnel. Referenced documents are available from the USNRC Public Document Room at1717 H Street, Washington, DC 20555 for a ccpying fee. Sulscriptions and additional or back issues of Power Reactor Events miy be requested from the NRC/CPO Sales Program,(301) 492-9530, or at PHIL-016. Washington, DC 20555.
Table of Contents Page 1.0 SUMMARIES OF EVENTS 1.1 Loss of Auxiliary Power Af fects Two Units...........................................
1 1.2 Extraction Steam Line R u ptu re.........................................
4 1.3 Miscategorization of Loss of Feedwater Transient........................
7 1.4 Inoperability of Instrumentation Due to Extreme Cold Weather...................
13 1.5 Fuel Assembly Degradation in the Spent Fuel Storage Pools..........
17 1.6 Safety Relief Valves Fail to Open at Setpoint..........................................
19 1.7 Test Procedure inadequacies................................
23 1.8 Loss of 24 Volt DC Power Common to Two Units.....
24 1.9 Common Cause Feedwater Pump Failures.....................................
25 1.10 Failed Dowel Pins Disable Diesel Generator..............................................
26 1.1 1 R e ferences........................
27 2.0 ABSTRACTS OF OTHER NRC OPERATING EXPERIENCE DOCUMENTS 2.1 Abnormal Occurrence Reports (NUREG-0090)...........
29 2.2 Bulletins, Circulars, and information Notices......................................
30 2.3 Enoineering Evaluations and Case Studies....................
36 2.4 R s.y;lato ry and Tech nicaI R epo rts....................................................
40 2.5 Operating R eactor Event Memoranda...................................................
46 Editor: Sheryl A. Massaro Associate Editor:
Steven E. Trenery Office for Analysis and Evaluation of Operational Data U. S. Nuclear Regulatory Commission Published in:
February 1983 Washington, D. C. 20555 8303140681 830228 PDR NUREG BR-OO51 R PDR
. _ = -
i l
1.0 SUMMARIES OF EVENTS 1.1 Loss of Auxiliary Power Affects Two Units On June 22, 1982, the NRC was notified by Commonwealth Edison Company (the licensee) of a sequence of events at Quad Cities Nuclear Power Station
- which resulted in a total unavailability of emergency diesel generator power for Unit 1 and the loss of offsite power and one emergency diesel generator for Unit 2.
i Diesel generators (DGs) at nuclear power plants provide emergency, onsite backup ac power in the event that normal offsite sources of ac power are unavailable. Quad Cities Units 1 and 2 have a combined total of three DGs. DG-1 is dedicated to Unit 1, DG-2 is dedicated to Unit 2, and DG-1/2 is a swing diesel that can be aligned to either unit. As a result of the sequence of events described below, nomal offsite sources of ac power were available for Unit 1, but neither t
DG-1 nor DG-1/2 were available; simultaneously, all normal offsite sources of ac power were lost for approximately 40 minutes to Unit 2, and only DG-2 was available. For both units, such loss of power sources can be considered a major degradation of essential safety-related equipment. The safety signficance was increased by several other failures which occurred during the event, including loss of several instrumentation indications in the control room. Nevertheless, the actions taken by the plant staff were timely and attentive and Unit 2 was safely shut down. Unit 1 operation was not affected.
At the time of the event, Unit 2 was operating at approximately 95%
and Unit 1 at 60% power. DG-1 was out of service for maintenance; however, DG-2 and DG-1/2 were operable. While preparing to remove the Unit 2 reserve auxiliary transfomer from service for elective i
repairs, an equipment operator at 5:25 a.m. mistakenly pulled the fuses for a 4-kilovolt bus instead of pulling the transformer fuses.
(When the plant is producing electricity, the plant loads and instru-l mentation are powered by the plant's main generator via an auxiliary l
transfomer. The reserve auxiliary transformer is available to supply
. offsite power when the plant is not operating. Each unit has a separate reserve auxiliary transfomer.)
The operator error disconnected power to certain plant systems, including the 2B reactor feedwater pump. The reduced feedwater flow caused a low water level, which automatically initiated a reactor trip. The Unit 2 main generator then tripped, resulting in the loss of all nomal ac power to the Unit 2 distribution system, as the reserve auxiliary transfomer was already out of service. Both DG-2 and DG-1/2 started automatically and began to supply power to essential plant systems.
Pressure in the reactor was reduced by the automatic operation of the
- fety relief valve and subsequent manual actuation of powar-operated relief valves.
In the process, one power-operated relief valve failed to open, and reactor operators actuated a second relief valve.
O 1
Quad Cities Units 1 and 2 are 679 MWe BWRs located 20 miles northeast of Moline, Illinois, and are operated by Commonwealth Edison.
1
--,.--v-
+-,.e n---,.
w,
,, - - - - - - - - - -, -, - - - - - - ~ - -,
c
-w,,-
_ _ _ _ _ _ _ At 5:47 a.m., 22 minutes after the event began, the reactor operators started the 2A residual heat removal (RHR) service water pump to begin cooling the water in the pressure suppression pool.
(The pressure suppression pool is a doughnut-shaped tank surrounding the reactor containment. Water in the pool condenses steam released by the relief and safety valves; condensing the steam transfers the heat to the water in the suppression pool.)
DG-1/2 tripped when the RHR service water pump was being started, cutting off power being supplied to various instrumentation and safety systems. The cause of the DG-1/2 trip was the actuation of under-excitation relays which protect the DG. Power continued to be supplied from DG-?.
Loss of DG-1/2 resulted in numerous alarms and loss of several control room instrument indications.
In addition, the loss of DG-1/2 left Unit 1, which was still operating, without any backup source of power (since DG-1 was already out of service for maintenance) should it experience a loss of offsite power. With the loss of instrumen-tation, a senior operator went to the local instrument panel in the reactor building to establish communications with the control room and relay infomation on reactor pressure, water level, and containment pressure to the reactor operators. At 5:50 a.m., an unusual event was declared under the licensee's emergency plan, and appropriate notifica-tions were made.
Pressure in the Unit 2 containment increased from the nomal pressure of about 1.3 psig to about 4.3 psig. The principal causes of the pressure increase were leaking gaskets on the discharge lines of the main steam relief valves, multiple relief valve actuations to control reactor pressure, and shutdown of the drywell coolers. The latter occurred, as designed, as a result of an emergency core cooling system (ECCS) initiation signal which actuated at a drywell pressure of about 2 psig.
Also, as a result of this ECCS initiation signal, the high pressure coolant injection (HPCI) system of the ECCS started automatically and began to pump water into the reactor vessel. A core spray pump also automatically actuated, but reactor pressure remained above the pemissive pressura setpoint required for actual core spray flow. The ECCS signal also tripped the running RHR service water pump and the reactor building closed cooling water (RBCCW) pumps. A normal Group II* isolation of containment occurred and functioned properly.
Licensee personnel in the meantime were restoring the Unit 2 reserve auxiliary transformer to service. By 6:04 a.m., 39 minutes after the i
event began, the transfomer was operable and offsite power was restored for all affected plant systems.
1 1
Group II: drywell ventilation, purge and sample lines, reactor building ventilation system, transient in-core probe withdrawal l
command, shutdown cooling, and head spray modes of residual hedt removal.
l Reactor pressure continued to be controlled by manual operation of l
relief valves, and at 6:15 a.m. suppression pool cooling was established
)
using the RHR system.
Cold shutdown was achieved at about 5:00 p.m.
The plant returned to service on June 24, 1982, after appropriate maintenance and testing activities were completed.
The cause of the event can be attributed to nonconservative planning of maintenance activities, operator errors, and design error. As stated previously, the station's auxiliary electrical power system utilizes three onsite power sources (DGs). Units 1 and 2 each have one dedicated l
DG and the third DG is a swing diesel that will automatically align itself to the unit that requires it. With this arrangement, the removal of a dedicated DG from service increases the potential -for unavailability of automatic onsite emergency ac power of one unit, i
1.e.,'it is susceptible to loss by a single failure. The removal of the swing diesel generator from service causes the unavailability of onsite power to one division of the emergency electric power system of both units.
Because of this interdependence of onsite power sources between both units 6t the station, any concurrent scheduled maintenance of the offsite power system of either unit decreases the reliability of l
the overall electric power system of both units. The Unit 2 reserve auxiliary transfomer is he primary source of offsite power for the j
pl ant. Therefore, the licensee's decision to remove the Unit 2 transfomer from service for elective maintenance while the plant was in operation, i
and particularly with DG-1 already out of service for maintenance, was nonconservative (even though it was not prohibited by the plant's technical specifications).
The event was initiated by the operator error in pulling the incorrect i
fuse. The operator pulled the fuse for the bus rather than the trans-fomer. This eventually led to a Unit 2 reactor scram and Unit 2 generator trip resulting in the loss of all nomal ac power to Unit 2.
Following loss of offsite power to Unit 2, DG-2 and DG-1/2 started as desigr.ed. However, when the operator later attempted to start an l
RHR service water pump for suppression pool cooling, DG-1/2 tripped.
l Repeated attempts by the control room operator to start DG-1/2 fail ed. The cause of the trip was a design error in the DG control logic system. An underexcitation relay had been installed in 1981 on L
all three DG control logic systems as a modification recommended by the licensee. The relay is designed to protect the DG during testing when i
the DG is loaded to an energized bus, and the relay protection should be automatically blocked when an auto-start signal actuates the DG.
i This trip was not automatically blocked when the operator initiated drywell and suppression pool cooling. The characteristic of the relay is such that it can actuate erroneously when a large motor (such as the RHR service water pump) is started. Actuation of the under-4 excitation relay also trips the DG lock-out relay. With the latter relay tripped, the DG would not restart until this relay was manually reset. Resetting of the DG lock-out relay was delayed since the equip-ment operator had been sent to the switchyard to expedite restoration of offsite power at Unit 2.
The licensee has taken appropriate measures to minimize the possibility for similar operator errors, including a review of procedures and addi-tional training for operating personnel.
. The underexcitation relay that caused the DG trip has been removed, and the licensee is planning modifications to all diesel generators to prevent protective trips in an emergency situation.
The power-operated relief valve which failed to open was replaced, and the operability of all relief valves was verified prior to the unit's return to service. The licensee also replaced the leaking gaskets on the relief valve discharge lines which had contributed to the rise in containment pressure. During the next refueling, the licensee plans to modify the core spray logic so that the drywell coolers and RBCCW pumps do not trip on a core spray initiation (e.g.,
2 psig drywell pressure) if offsite power is available to the emer-gency buses.
In addition, the licensee is considering other measures which would allow a reserve auxiliary transformer to be removed from service for elective maintenance only while its reactor is shut down with power being supplied from offsite sources through the main transformer.
(Refs. 1 through 3.)
1.2 Extraction Steam Line Rupture On June 28, 1982, Oconee Unit 2* experienced a rupture of a 24-inch steam extraction line while operating at 95% power.
Upon hearing a loud noise and observing an apparent loss of main steam turbine header pressure, the reactor operators suspected that a main steam line break had occurred. Nine seconds after the rupture, the reactor was manually tripped, initiating an automatic turbine trip.
The break was downstream of the main steam stop valves; thus, the turbine trip isolated steam supply to the extraction line. The 194 psig/380*F steam escaping through the four square foot rupture physically destroyed motor control center (MCC) 2XA-A. There were, however, no safety-related loads supplied from the MCC, nor any essential loads which precluded routine plant shutdown.
Steam impingement also destroyed several nonsafety-related instruments which were mounted on a panel board located six feet from the failure.
Two of four turbine steam header pressure transmitters were among the instruments destroyed, and caused the loss of indication of steam header pressure.
Safety-related steam generator header pressure instruments were not affected.
Other than the aforementioned equipment, no significant structural or
(
equipment damage was noted, nor was there evidence of pipe whip.
Three personnel suffered steam burns; two were hospitalized over-night.
Main steam pressure, as measured at the steam generators, peaked at 1097 psig, ultimately stabilizing at approximately 1000 psig.
Neither the pressurizer code safety valves nor the power-operated Oconee Units 1, 2, and 3 are 860 MWe PWRs located 30 miles west of i
i Greenville, South Carolina, and are operated by Duke Power.
. relief valve (PORV) were called upon to operate during the transient; primary and secondary levels remained on scale, within acceptable range; and no engineered safety features (ESF) setpoints were reached.
In an attempt to isolate the rupture, the operators isolated the "A" steam generator.
In subsequent actions, the operators sequentially 1
isolated and un-isolated both steam generators and the steam (turbine) bypass valves in their attempts to locate and isolate the rupture.
During these efforts, it was suspected that one or more of the main steam relief valves was stuck open. As a result, the steam bypass valves were isolated and the main steam relief valve (s) were used as the primary heat sink to pre-empt the possibility of an overcooling transient. Once the location of the rupture was ascertained and the main steam relief valves were verified closed, the bypass valves were un-isolated and a normal cooldown rate was established.
Subsequent evaluation of the transient and parameter recordings indicates that the main steam relief valves did not stick or hang open, but responded as would be expected for the set of circumstances described.
Seven minutes into the event, the unit experienced the loss of the process computer for a period of 3.5 minutes.
The reactor coolant subcooling margin monitors, supplied from the process computer, were rendered inoperable. The operators ascertained subcooling during the period from RCS temperature and pressure indications available in the control room. Loss of the computer posed no major impediment to the shutdown of the plant.
The computer loss was apparently the result of a computer stall -
a computer malfunction during which the computer either slows down drastically or quits as a result of excessive input information.
The computer was reinitialized with no major difficulty.
Initial indication of steam extraction line degradation at the Oconee facility was discovered in 1976, when an extraction line at Unit 3 had developed a pinhole due to steam erosion. At that time, the licensee initiated an informal, undocumented ultrasonic thickness examination program to detect similar areas of erosion degradation. They now have disclosed the following history of high pressure turbine extraction line erosion problems. Since these concern non-safety-related piping, records are not complete. The failures described below, however, were of pinhole size and do not approach the failure of the June 28, 1982 event.
Date Unit Description 10/76 3
The first pipe section between the 20-inch nozzle on the 42-inch high pressure discharge pipe on the C extraction turbine end left side developed a small hole due to erosion.
. Date Unit Description 6/77 3
The first pipe section between the 20-inch nozzle on the 42-inch high pressure discharge pipe and the first elbow on the C extraction generator end right side was ultrasonically tested and significant erosion was noted.
The spool piece was replaced.
8/78 1
The weld joining the turbine nozzle to the first pipe section on the turbine end, B extraction failed by
' erosion; attributed to a misaligned backing ring which caused eddies in the steam flow.
j i
late /79 3
The first elbow downstream of the 24-inch nozzle on the 42-inch high pressure discharge pipe C extraction turbine end developed a pinhole leak.
Subsequent to the 1979 event, the licensee had formalized their ultrasonic thickness inspection program. The procedure required that the sneasurements be performed 90* apart around the pipe, but did not specify a grid map location of the measurements.
The June 1982 rupture occurred in the outside radius of a 375-mil thick 90' elbow where a 24-inch steam extraction line branches off a 42-inch high pressure turbine exhaust line. Although ultrasonic thickness testing performed on the elbow in March 1982 had revealed significant erosion thinning, the elbow remained within minimum thickness requirements.
The thinnest area recorded was 170 mils; micrometer readings performed after the rupture revealed a thickness of 17 mils at the edge of the
/
failure.
The C extraction line failure at Oconee Unit 2 was attributed to i
steam erosion, which was accelerated by low quality steam in the
~
extraction line as a result of sustained reduced power operation.
The licensee replaced the damaged elbow, and inspected other steam extraction line piping. Equipment supplied from MCC 2XA-A was supplied temporarily from alternate sources. The licensee has revised the
-inspection program to utilize a detailed examination grid to provide a better correlation for analysis.
(Refs. 4 and 5.)
The NRC issued Inspection and Enforcement Information Notice 82-22,
" Failure in Turbine Exhaust Lines," on July 9,1982 to all nuclear power licensees to inform them of this event, and of several similar failures at other facilities since January 1,1982.
(See p. 30 for abstract of Notice.)
i I
1.3 Miscategorization of Loss of Feedwater Transient On April 24, 1982, a loss of feedwater transient occurred at 12:56 a.m. at Vermont Yankee.* The transient was initiated at 75% full power by a malfunction in the feedwater control system, and occurred during a plant load decrease to establish conditions for routine surveillance testing. Following unsuccessful operator attempts to restore feedwater flow, the reactor scrammed at 12:59 a.m. on low reactor vessel water level. The scram was followed fourteen seconds later by an emergency core cooling system (ECCS) injection on vessel low-low level. The transient was essentially teminated by 1:01 a.m.,
with vessel level restored by high pressure coolant injection (HPCI),
and reactor core isolation cooling (RCIC) system injections by a Group I** isolation which secured steaming. Plant operators reestablished nomal hot shutdown conditions by 1:25 a.m., with stable vessel level and pressure, an operating feedwater pump and steam bypass to the g
condenser.
A plant trip report was made to the NRC Duty Office at 1:21 a.m.
]
using the Emergency Notification System (ENS), but the licensee did i
not categorize the event as an " alert," as required by the Yemont Yankee Emergency Plan upon ECCS injection. At the time, the operators were apparently not aware that there was a safeguards actuation (primarily HPCI/RCIC injections). At about 2:00 a.m. operator review of control room panels and computer printouts identified i
the safeguards actuation and subsequent HPCI/RCIC injections. A second ENS notification was made to the NRC at 3:15 a.m., correcting I
earlier information. During this second discussion with the NRC, plant management decided not to declare an " alert" as prescribed by the Emergency Plan because plant conditions were then considered stable.
The plant response to a total loss of feedwater, starting from normal 100% power conditions, is analyzed in Vermont Yankee's Final Safety Analysis Report (FSAR).
In this analysis, all running feedwater planps trip, followed by a 4-to 5-second feedwater flow coastdown. An inter-lock with the recirculation flow control system reduces the recircula-l tion pumps to about 20% of rated speed after feedwater flow drops below l
20%, which reduces reactor power.
Reactor vessel indicated level drops at a maximum rate of 27 inches /second with the main turbine in operation.
The low water level scram setpoint is reached about 8 seconds into the transient. Twelve seconds into the transient, HPCI and RCIC start, the main steam isolation valves (MSIVs) close, the diesel generators start, the recirculation pumps trip, and the recirculation discharge valves close on low-low reactor vessel water level. RCIC flow is sufficient to recover vessel level with the MSIVs closed. Decay heat is expected to slowly raise the vessel pressure (assuming no operator action) until the safety-relief valve setpoint is reached, but excessive overpressurization is not expected to occur.
Vermont Yankee is a 504 MWe BWR located five miles south of Brattleboro, Yemont and is operated by Vemont Yankee Nuclear l
Power.
l l
- Group I: main steam isolation valves, steam line drains, and reactor water sample lines.
.g_
During the April 24 event, the feed pumps tripped (at 12:56 a.m.)
because of reactor vessel high water level. The main turbine remained in operation because its high water level trip setpoint was not reached.
The recirculation pumps began to run back, as required. Reactor vessel level dropped, then started to increase after two feed pumps started (one automatically and one manually). At 12:59 a.m., the feed pumps tripped again on reactor vessel high level. Level dropped at about 2 inches /
second. The recirculation pumps continued to run back. The reactor tripped on low water level about 12 seconds after the second feed pump trip.
Level continued to drop, reaching the low-low level setpoint about 36 seconds after the second feed pump trip and causing the recirculation i
pumps to trip and their discharge valves to close, the diesel generators to start, the MSIVs to close, and HPCI and RCIC to start. The main turbine automatically tripped also on low-low vessel level. Water level rose until HPCI and RCIC shut down on high water level. The MSIVs were then reopened by the operators to provide a heat sink for the reactor; consequently, no relief valve lifts occurred. The lowest reactor water level indicated during the transient was about 87 inches above the reactor core.
The transient experienced during this event was consistent with, and less severe than, that described in the plant's FSAR. No unacceptable conditions regarding sy! tem response were identified.
System Response The cause of the feedwater transient was determined to be a failure of the electro-pneumatic controller for one of the two feedwater regulating valves. The response of plant normal operating and emergency systems required to operate during the transient was reviewed and found acceptable.
The high pressure coolant injection and other safeguards systems actuated as required within the trip settings established by the technical speci-fication limiting safety system setting.
Personnel Response The personnel on shift on April 24 included the shift supervisor, the supervisory control room operator (SCRO), the control room operator (RO),
and the nuclear safety engineer (NSE). The reactor engineer was also in the control room witnessing a control rod pattern exchange.
Although the R0 had not previously experienced a plant trip, both the shift supervisor and SCR0 were experienced. The NSE, who serves at Vennont Yankee as the shift technical advisor, had about 10 months experience on shift.
(1) General Chronology Upon completion of shift turnover, shift rounds, and panel checks, preparations were made to perfonn scheduled surveillance testing.
i Plant load was decreased using recirculation flow control. At the start of the event, the SCR0 and R0 were at their normal stations (see Figure 1.) The NSE was also in the control room, but l
the shift supervisor was outside the control room discussing work assignments with the on-shift auxiliary operators.
l
I
- e P.c E
E e2 8" E $
M E *.*
k-e.
i M
i 12.j 358sEE e!E3 M
i ! E 13 E' E li # 3.,
o 9a
=
g
--a o *3e ug o-M E.:"34
- E ;; 3 w*gg 3,
o d
g I' E 5 5 ~,,
e.;.si= del $E55EE
- d..,o.'.
e 3.
i e.,.s. s x,, = x e 5,
) o..,
o
. m o = E
.. =.... W ~.. n '.
u., e
....a
. '.C
. h.
u i m W e... z w - u..m3 m
um r
u n.
- 4 w w w u 2 o.a o u. e m a. a.
waou u z
>&.oi
@WR Qb p.0;<
c n?ob Ys m,%.
p;.-.'
P..O. <
o.
$iO;a
.m o.
N id
,3-o.
D,<
o y
3.i 8a.(
.I J.d*
&.. i
.o 4o.A CRPS4 CRP 9 3 ar.'
t lE RC/RCIC PCIS/RCIS/HPCI O
2c o
w
..c...../ :. e..-
o s
- c o'R* 3
- 4 o'j.
,.. w.~ :...
'..... : Q:'....'.:.
h!
, ; c.'
.b..*.-
tx; ii. :
a o
k...
k *......... -
-.r
,d 0ESK
. ' '.*,/
4?.%.:.i.l.@.*.. if.......
Ag pp
.. M:l.9.;..4 8
y M
i e
.F.
3 *-
u u
i
. '/
e,,Q o
. * *. - l *.. * :: 9,.
. c... :.....
,.: :.. + r.... : m.
a.e.
c
- w..u -
.,3
.>. r.x......g, n,
8.,
.s... m...,;
....w.,..
,.r..
u m
a.w.-
e:
.:. :;-.a:.. ::c.
- s. is w
.......................:.........~....n, w
p,..,
=.
c........
. z...
u
.....m.............
c.
.. v..
.,. v....
=
,............*;.c*'.,.
o o
..s...
y,
. &. *.......*:K:l', * : :*. ll.% ;', y p
.y y.. ;; ;; y j.h th*E-Di6M.h'i'.H *@oloW XL*N g
.I*
- z g
u e
.c.
t 3.,
9e...q
.g h
9 k'N.".4*.M.MW/WfdE@b3@20F7/oWh, 5E**.H5 nim *N 4
i
~
[
1 l !
I Two key events were initially unnoticed by the operators in the sequence that followed. The first event was the second trip of the feedwater pumps at 12:59 a.m.
The second event was the HPCI/RCIC injection that accompanied the low-low level safeguards actuation. Failure to notice the first event negated shift crew attempts to mitigate the loss of feedwater (LOFW) transient.
Failure to notice the second event hindered complete assessment of all plant conditions. The following provides infonnation on why these events went unnoticed.
l The initial problem the operators became aware of and responded to i
was the first loss of feedwater pumps at 12:56 a.m.
The R0 and the SCR0 responded to the reactor and feedwater panels. The NSE took a i
position behind the computer console in front of the reactor panel. The crew's first assessment was that some problem existed in the feedwater control circuitry and that the feedwater pumps had tripped for an as yet unexplained reason. Operator attention focused on "present" reactor vessel water level and the level control station, which seemed to provide a sluggish response. Upon hearing the feedwater pumps slow down, the shift supervisor went immediately to the control room and entered it just as the A and C pumps were restarted. After checking the operating I
parameters of the pumps, he moved to the front of the reactor panel to help diagnose the response of the level control station. Vessel water I
level increased as the A and C feedwater pumps came up to speed and again reached the high level trip setpoint.
The second trip of the feedwater pumps at 12:59 a.m. started the major decrease in vessel water level, and was followed very quickly by 2
the reactor scram. Reactor scram recovery actions were initiated.
The shift supervisor anticipated that HPCI/RCIC actuations would t
occur unless the level control system, which was now in manual control, responded. Vessel water level did start to increase as a result of HPCI/RCIC injection. However, the operators attributed the l
cause of the increase to the feedwater system. Vessel water level eventually rose to about 173 inches.
(The time interval from reactor scram to vessel high water level was about 81 seconds.) At this point l
the shift supervisor noted that the feedwater pumps were off and l
concluded that a feedwater pump trip had just occurred (about 1:01 a.m.)
due to the high vessel level. This reinforced the conclusion that the feedwater system had been responsible for the level recovery.
i In the course of taking actions to recover from the scram, the operators noted that, in addition to other expected automatic scram actions, a Group I* isolation occurred, the diesels had started, and the recirculation pumps were not running. The operators associated the Group I isolation with trips which result from a combination of high steam flow or low steam header pressure conditions and the position of the reactor mode switch. The i
i reasons why the diesels started and the recirculation pumps tripped were not determined by the operators until about 2:00 a.m.
Crew activities from 1:00 a.m. to 1:30 a.m. were directed toward maintaining stable vessel level / pressure and realigning / checking i
i Group I: main steam isolation valves, steam line drains, and reactor water sample lines.
__ _ _-- systems affected by the scram. These systems included the reactor water cleanup systems, Group II through V* isolations, the main steam isolation valves, main turbine, electrical power, diesel generators, recirculation pumps, standby gas treatment system, and a
feedwater. Several of these systems (most rectably the diesel start and the recirculation pump trip) provided indications of the vessel l
low-low water level actuations. The crew did not diagnose these condi-tions, but instead concentrated on operating systems which they thought 4
were immediately needed. During this time, the shift supervisor assessed plant conditions and concluded that no loss of coolant indications were present and that the plant was stable.
Indications of HPCI/RCIC actuations included valve status lights and four annunciators that came on when the HPCI/RCIC turbines shut down on high vessel water level at 1:01 a.m.
However, these annunciators came on along with nimercus other alarms that accompanied the scram, some of which were probably cleared along with other
" acknowledged" alanns immediately following the scram. The HPCI/RCIC valve status light changes were not immediately noted.
]
From 1:20 a.m. to about 2:00 a.m., the shift supervisor made tele-i phone notifications about the event, starting with the ENS call at 1:21 a.m.
The other two licensed operators continued to work the panels and were working in particular with the reactor water l
cleanup system (due to problems in keeping it on line) and checking the status of an emergency power supply for valves in the recir-culation and residual heat removal system. The shift supervisor referred to event classification procedures at about 1:25 a.m.,
but found no match with the classification criteria, based on his j
understanding of the event at that time.
Upon completion of the notifications and during a panel check the shift supervisor noted the status of the RCIC/HPCI systems (speci-fically, the valve status lights and the operating barometric condenser pumps).
Based on this information, along with previous indications from the diesels, recirculation pumps and Group 1 isolations, he concluded that the HPCI and RCIC systems had actuated and injected. He further concluded that vessel water level had reached the low-low setpoint and HPCI/RCIC had helped the RFPs restore vessel f
l level. Subsequent operator review of computer printouts revealed the second feedwater pump trip (at 12:59 a.m.) which provided for a fuller understanding of the event sequence. The shift supervisor made a i
t l
l Group II:
drywell ventilation, purge and sample lines, reactor building ventilation system, transient in-core probe withdrawal I
command, shutdown cooling, and head spray modes of residual heat removal.
l Group III:
reactor water cleanup system.
Group IV: high pressure coolant injection system.
Group V:
reactor core isolation cooling system.
second ENS call at 3:51 a.m. to correct information given at 1:21 a.m.
regarding actuation of the safeguards system (i.e., HPCI). During this second call, it was realized that plant procedures required classification of ECCS actuation and injection as " alert" conditions.
(2) Nuclear Safety Engineer (NSE) Actions The NSE's initial efforts were to monitor the plant response and operating crew actions, and to do so without interfering with access to the control panels. This was done by referring to control panel indicators and alarm typer messages. Following the reactor scram, the NSE also verified the proper scram response actions taken. He noticed the start of the HPCI system at 1:00 a.m.
and concluded from this infomation and the startup of the diesels that a reactor vessel low-low water level actuation had occurred.
This conclusion was reached by about 1:20 a.m.
He did not communicate any of this infomation to the operators because he assumed they had reached the same conclusion.
Following plant stabilization, the NSE worked with the operators to reconstruct the event sequence from the logs and records.
(3) Findings (a)
For the most part, operator actions to restore and stabilize reactor vessel conditions, and subsequently their actions to stabilize other plant conditions, were proper. However, the operators did not, in a timely manner, diagnose available indicators of how reactor vessel water level was recovered.
(b) The shift supervisor participated at close-hand with the manipulation of feedwater controls; thus, actions to assess plant status in a timely manner were flawed because of the failure to get the "overall picture" of plant conditions.
(c) Hindrances to the operators in responding to the event included:
- lack of an annunciator for loss of feedwater pumps on high vessel level, which would have clearly identified the second feedwater pump trip before the reactor vessel low water level scram;
- failure to consult the pertinent operating procedures after plant stabilization;
- 40 minutes spent in making licensee and NRC notifications; and
- lack of input from the NSE.
(d) The purposes of NSE functions were not wholly met because of the failure to communicate pertinent plant infomation during the transient.
i l
(e) Once the safeguards system actuation was known by the shift supervisor (about 2:00 a.m.), he failed to adequately review plant procedures for classifying the event.
13 -
A Notice of Violation and a Notice of Proposed Imposition of Civil Penalty in the amount of $40,000 was issued the NRC. This action was based on the the failure of station personnel to promptly recognize changes in the status j
of safety-related equipment, and for failure to promptly classify the event t
and make the required notifications in accordance with the emergency plan.
The civil penalty was paid in full by the Vennont Yankee Nuclear Power Corporation in November 1982.
(Refs. 6 through 8.)
1.4 Inoperability of Instrumentation Due to Extreme Cold Weather On January 3,1979, an unusual event occurred at Davis Besse Unit 1.*
The event involved the freezing of the water in a portion of the high pressure coolant injection system recirculation line that is common to both high pressure coolant injection pumps.
The line was not thawed until January 5,1979, and the event was not reported to NRC until March 12, 1979.
~
This was one of several events that were discussed in IE Bulletin 79-23,
" Frozen Lines," issued September 27, 1979.
Since that time, several other failures of power and safety instrumentation have occurred due to sensing lines freezing during extreme cold weather.
(Ref. 9.)
The following information is being presented with the objective of alerting licensees, vendors, and plant personnel to the more common problem areas, and to outline preventive measures that licensees have taken associated with such problem areas.
The majority of the reported cold weather-related failures can be grouped into three categories:
those affecting (1) level instrumentation for the refueling water and borated water storage tanks, (2) main steamline pressure and flow instrumentation sensing lines, and (3) radiological effluent sampling lines. A numerical breakdown of incidents in these categories occurring during the winter months from 1972 through 1981 is shown in Table 1.
Many of these occurrences were directly related to inadequacies associated with the heat tracing provided for these sensing and sampling lines.
The most frequent causes of line freeze-up are:
- the absence of heat tracing or adequate insulation, de-energized heat trace circuits,
- improper thermostat settings, incorrect sensor location for the heat tracing, and
- space heater failures.
Several incidents for the winter of 1981/1982 are described in Table 2.
( Re f. 10. )
Davis Besse Unit is an 874 MWe PWR located 21 miles east of Toledo, Ohio, and is operated by Toledo Edison.
-w Table 1 - Numerical Breakdown of Instrumentation Failures Due to Frozen Lines Instrumentation Affected Winter Steamline Pressure Rad. Ef fluent Season Level and Flow Sampling Lines Miscellaneous 1972-73 0
0 2
0 1973-74 0
3 0
1 1974-75 3
0 2
1 1975-76 2
4 0
2 1976-77 2
3 2
1 1977-78 1
1 0
3 1978-79 6
3 2
4 1979-80 5
3 1
0 1980-81 5
2 1
5 Table 2 - Incidents Due to Frozen Lines During Winter 1981/1982 Plant Name*
LER No.
Date of Occurrence Brief Description of Event
'f Point Beach 1 31-20/03L 12/19/81 "A" steam generator pressure 81-01/03L 01/07/82 sensing lines had frozen on 01/10/82 four different occasions.
The events were caused hy inadequate freeze protection and extremely cold weather.
Zion 2 82-01/03L 01/09/82 A steam generator 2C pressure channel failed high due to frigid outside air entering the transmitter location and causing a frozen sensing line.
Robinson 2 81-34/03L 12/11/81 A steam line pressure channel was indicating higher reading than others. This was caused by freezing of the pressure transmitter sensing lines.
It was found that the freeze protection circuit for this transmitter was not in service.
- Point Beach Unit 1 is a 495 MWe PWR located 15 miles north of Manitowac, Wisconsin, and is operated hy Wisconsin Electric Power.
Zion Unit 2 is a 1040 MWe PWR located 40 miles north of Chicago, Illinois, and is operated by Canmonwealth Edison.
Robinson Unit 2 is a 665 MWe PWR located five miles northwest of Hartsville, South Carolina, and is operated by Carolina Power and Light.
._ P,lant Name*
LER No.
Date of Occurrence Brief Description of Event McGuire 1 82-07/03L 01/11/82 Exposure to cold conditions began to cause errors in instrumentation that affected automatic feedwater control.
Efforts to prevent instrument line from freezing were made all day on January 10 and January 11; but two channels of steam generator A pressure tripped due to freezing of the sensing lines. Engineered safety features was automatically actuated as a result of 2/3 low i
pressure coincidence trip of steam generator A pressure instru-mentation channels.
Inadvertent actuation of engineered safety features caused safety injection, steam line isolation and reactor and turbine trips.
Farley 2 82-04/03L 01/11/82 A main steam line pressure channel was declared inoperable due to frozen sensing lines.
Sections of the line were found to have inadequate insul ation.
Farley 2 82-02/03L 01/11/82 An instrisnentation loop associated with main feed flow transmitter was declared inoperable due to frozen transmitter sensing lines.
Farley 2 82-03/03L 01/12/82 An instrumentation loop associated with a refueling water storage tank level transmitter was declared inoperable when its indicator i
failed low. The exposed I
diaphragm on the associated level switch ruptured due to freezing, causing leakage from the sensing line.
- McGuire Unit 1 is a 1180 MWe PWR located 17 miles north of Charlotte, North Carolina, and is operated by Duke Power.
Farley Unit 2 is an 814 MWe PWR located 28 miles southeast of Dothan, Alabama, l
and is operated by Alabama Power and Light.
Plant Name*
LER No.
Date of Occurrence Brief Description of Event Pilgrim 1 82-02/03L 01/18/82 Pressure switch associated 01/19/82 with the dietel fire pump 01/23/83 was inoperable on three occasions. The sensing line freezing during cold weather was the cause of switch failures.
Sorry 1 82-04/03L 01/12/82 Unusually low temperatures caused stagnant water in the body of several fire hydrants to freeze and prevented hydrant valve operation.
Beaver Valley 1 82-02/03L 01/19/82 Steam generator drain tank recirculation line cracked due to freezing within the line as a result of an extended cold spell. The heat tracing of the line was found to be inadequately sized.
Brunswick 2 82-04/03L 01/11/82 Frozen condensate in core auxiliary cooling system vaporizer condensate return line prevented the vaporizer from adding nitrogen to the drywell; heat tracing and insulation lagging of lines was found missing.
The following are some of the actions taken, or being considered by licensees and the NRC, with the objective of lowering the frequency of these occurrences:
- Pilgrim Unit 1 is a 670 MWe BWR located four miles southeast of Plymouth, Massachusetts, and is operated by Boston Edison.
l Surry Unit 1 is a 775 MWe PWR located 17 miles northwest of Newport News, Virginia, I
and is operated by Virginia Electric and Power.
Beaver Valley Unit 1 is an 810 MWe PWR located in Pennsylvania, five miles east of East Liverpool, Ohio, and is operated by Duquesne Light.
Brunswick Unit 2 is a 790 MWe BWR located three miles north of Southport, North Carolina, and is operated by Carolina Power and Light.
_ (1) Daily surveillance during periods of extreme cold weather for lines in the three categories discussed above, and for any other susceptible lines that have safety significance. Further, procedural or technical specification changes include appropriate action statements which supplement surveillance requirements.
(2) As an alternative to daily surveillance, design changes have been considered or made, such as alarmed and/or fully redundant Class 1E heat tracing circuits including electrical power sources. These have been supplemented by less stringent technical specification surveillance requirements and action statements.
(3) As part of the NRC's operating license reviews, the adequacy of the heat tracing or nther protective measures associated with these and other lines that could be exposed to ambient temperatures have received increased emphasis.
1.5 Fuel Assembly Degradation in the Spent Fuel Storage Pools The following event involving fuel assembly degradation at Prairie Island
- may indicate a mode of degradation not previously considered, as discussed in a recent NRC study of the event. (Ref.11.) The primary safety concerns are:
(1) the mode of degradation appears to be new and occurs while fuel is in the spent fuel pool; (2) the degradation may not be readily detectable; (3) the degradation resulted in separation of the top nozzle from the remainder of the fuel assembly while moving the assembly; and (4) if the mode of degradation is generic, there is the potential for dropped fuel assemblies with resultant damage to the spent fuel and possible fission product release.
On December 16, 1981, with Unit 1 and 2 operating at power, spent fuel was being moved in the spent fuel storage pools.
A spent fuel assembly top nozzle came apart from the rest of the assembly, causing the assembly to tip toward the edge of the pool.
It became wedged at approximately a 30*
angle with the lower end of the assembly resting on top of the fuel rack, and the top end resting in a gate opening in the wall between the transfer canal and one pool.
The failure of the fuel assembly top nozzle occurred at the first bulge I
joint between the 16 stainless steel sleeves and the Zircaloy guide tubes. During fuel handling, the weight of a fuel assembly is supported from the spent fuel handling tool which is latched onto the fuel assembly l
top nozzle. The load is then transmitted down through the assembly via the 16 guide tubes which are connected to the bottom nozzle. The top nozzle is attached to the guide tubes as follows: a stainless steel sleeve, approximately 8 inches long, is welded to the top nozzle at each of the 16 guide tube locations. The Zircaloy guide tubes are then inserted into these sleeves and coupled by a fonnfit where the Zircaloy tube is bulged out into the sleeve at three elevations.
- Prairie Island Unit 1 is a 503 MWe PWR and Unit 2 is a 500 MWe PWR. They are located 28 miles southeast of Minneapolis, Minnesota, and are operated by Northern States Power.
l l
l
l Following metallurgical examinations of the fuel assembly by the fuel vendor, it was determined that the failure was caused by stress corrosion
{
cracking of the stainless steel. The cracks were intergranular and several i
cracks were present but did not propagate. Fracture surface examination revealed concentrations of iron, chromim, aluminuin, silicon, nickel, and copper. There were also impurities of chlorides, fluorides, and sulfur.
The licensee does not believe the stress corrosion cracking occurred in the reactor, because the hydrogen overpressure should result in low free oxygen.
However, the spent fuel pool has highly oxygenated water at low temperatures and could be a cause of cracking in the presence of high stresses and sensitized stainless steel. Were do not appear to be any recent changes in design or fabrication of the bulge joints that could be a root cause of the cracking.
A total of 27 assemblies at Prairie Island were examined.
Although 12 showed evidence of corrosion, this did not result in additional nozzle failures when the assemblies were moved.
The nozzle that did fail had stress corrosion cracks over approximately 90% of the rupture surface with ductile shear over the remaining 10% surface.
In addition, a review of material lots determined that assemblies with the same material have been shipped to Point Beach
- and Kewaunee.** Examinations at these plants have resulted in no evidence of stress corrosion cracking in the top nozzle.
The fuel vendor for Prairie Island determined that (1) since stress corrosion cracking takes time to develop, it must have been present at Prairie Island before December 16, 1981; (2) stress corrosion cracking is not likely to have developed during operation (when the assembly was in the reactor core),
and therefore is believed to have developed during pool storage; and (3) the pool chemistry at Prairie Island appears to have been in accordance with vendor-supplied specification limits.
Subsequent to the occurrence, fuel handling in the spent fuel pools was suspended on December 16. Recovery and storage of the assembly was accomplished on January 19 and 20, 1982. The assembly was then placed in a normal spent fuel storage location.
Fuel handling in the pools was resumed on February 19, 1982.
Of the 500 fuel assemblies at Units 1 and 2, 440 have been relocated to new high density storage racks, and should remain in place for several years.
( Re f. 12. )
The NRC has scheduled a meeting with the Prairie Island licensee and Westinghouse (the nuclear steam system supplier) to determine (1) the cause of stress cor-rosion cracking, and (2) if the event is generic in nature.
i Point Beach Units 1 and 2 are 495 MWe PWRs located 15 miles north of Manitowac, Wisconsin, and are operated by Wisconsin Electric Power.
19 -
4-1.6 Safety Relief Yalves Fail to Open at Setpoint On July 3,1982, while Hatch Unit 1* was operating at full power, a spurious trip of channel B of the reactor protection system (RPS) occurred.
_A concurrent. trip was received on channel A of the RPS, resulting in a reactor scram.. Reactor pressure was at 1005 psig and the high. pressure trip setpofnt was at,1035 psig.
The main turbine had not tripped when a Group I** isolation occurred. This included closure of the main steam isolation valves (MSIVs), and reactor pressure began increasing with the MSIVs closed.
The wide range pressure records indicated a maximum pressure of about 1180 psig. The operator elected ;
to display the GEMAC pressure indicator on the process computer digital display.
This indicator displayed a mar,foum of 1200 psig. Two other pressure indicators also displayed a maximum of 1700 psig.
During this pressure transient,11 safety relief valves (SRVs) failed to actuate at their prescribed setpoints of between 1080 and 1100 psig. Twelve minutes after the reactor trip, when reactor pressure reached 1180 psig, three SRVs lifted for approximately 30 seconds, reducing pressure to 945 psig. At that point, the MSIVs were manually opened.
Reactor water level recovered and reached the high level trip setpoint, tripping high pressure coolant injection, reactor core isolation cooling, and the reactor feed planp turbines. A normal recovery was initiated by the operators.
The SRVs were declared inoperable, and the unit was placed in cold shutdown.
Safety limits were not exceeded, and no degradation of plant equipment occurred.
However, failure of SRVs to actuate at their required setpoints is a violation of plant technical specifications.
The licensee speculated that workmen cleaning the floor near the pr~ essure sensing instrumentation may have bumped the racks, causing the, spurious trip.
The A and B reactor protection system channels are located on; separate panels, but are physically close to each other. The three SRVs that. opened ~ are all on -
the C main steam line.
All SRVs involved were two-stage Target Rock-valves.-
The licensee (Georgia Power) took the following corrective actions, which should reduce the probability of SRV malfunction:
i (1) During the subsequent cooldown, the 11 SRVs were manually actuated and all were found to operate satisfactorily.
(2) The topworks, or pH M section (see Figures 2 and 3), from all 11 SRVs were removed and : ens to Wyle Laboratories for testing. Test results are summarize' L fo' aws:
(a) each vahe was _jcied five to eight times after receipt at Kyle Laboratories; (b) six passed on first run, four on second run, and one on third run without setpoint spring adjustment; Hatch Unit 1 is a 757 MWe BWR located 11 miles north of Baxley, Georgia, and is operated by Georgia Power.
Group I: main steam isolation valves, steam line drains, and reactor water sample lines.
1 DIAPHRAGM TYPE PNEUMATIC ACTUATOR
. <p.
- ~~
HH q
"_h I-i, r
J
.I n
N 2
(_[
o)
L
/'si o
O
' O O
O O
m
' O T' T' O Q i
O
/
4#
f j/
4 s
7 DISCHARGE
',~ - _ J_ J_ _ _ Z'_-[ -____[.
~ ~ ~ ~
BONNET l
2 i l r
l
-/I
.oooood i
1l
_m.
t !
=
. -MAIN PISTON ig li
[
l l
l I I I BODY ii I I
! l li See Figure 3 For Detail i
I ii l u
A e
INLET Figure 2 Target Rock Two-Stage Pilot Actuated, Safety / Relief Valve
-- PNEUMATIC ACTUATOR
/ s A 4 crm 7'
mv
,/ 7 b
llll
,/ /
o o
o s
o o
s SUR g QO
_I O 3 N(O O
r, O x O s
" :r,:
s O
31 O
kO ggf//
'^*M'"'" q S
\\
/
k lf
~~ ~
PILOT STAGE e
DISCHARGE TUBE
/
! I t i i N
~
~
{rY_]/"N
~
' PILOT
\\
DISC
)O 0 0 0 0 01 /
,/ /
\\
MAIN
~ f l\\
\\
P!STON
,/
/
g, NBALANCING DISC l
.. _..._.5_ _ _ _ _. _ _/%t4,,., ;
r
'OOOOO l
\\ MAIN SPRING J
f' I
Figure 3 Topworks of Target Rock Two-Stage Safety / Relief Valve
i (c) one setpoint was adjusted upward on fifth run because it was approaching the lower limit; J
(d) two valves had their setpoints changed, one from 1100 psig to 1080 i
psig and the other from 1080 psig to 1100 psig. These changes were done at Georgia Power's request.
(3) The main valve sections from the SRVs were removed and examined to ascer-tain if the cause for the malfunction could be determined. No evidence of mechanical wear or binding of components was found.
(4) To define the problem and to improve the probability that the SRVs will actuate when required, Georgia Power instituted a program at Hatch whereby nine of the 11 Unit i valves will be exercised once every 60 days. Two valves will not be exercised and will be utilized for possible future testing. Unit 2 valves will be subjected to a similar program.
(5) The Plant Review Board and the Safety Review Board met to review the event and concluded restart was acceptable based on their evaluations of the transient, corrective actions taken, and the planned SRV periodic cycling and inspection, as agreed in a Georgia Power letter to the NRC on July 9,1982.
(6) Target Rock Corporation and Wyle Laboratories concurred in the periodic, manual cycling of the SRVs to minimize the upward setpoint drift experienced on this event.
(Refs.13 and 14.)
Three additional licensees--Tennessee Valley Authority, Northeast Nuclear Energy Company, and Boston Edison--had reported that two-stage Target Rock valves, tested in the as-received condition at Wyle Laboratories, failed to actuate within 1% of the setpoint. The Hatch 1 event was potentially the most signi-ficant in terms of both (1) the fraction of valves that failed to open at their setpoint, and (2) the pressure above setpoint required to open the valves.
The General Electric Company (GE) and the Target Rock Company have joined Georgia Power in attempting to determine the cause of the failure of the valves to actuate. A GE analysis suggests that the most likely cause of the high actuation pressure is some combination of friction in the labyrinth seal area and/or sticking of the pilot disk in its seat. The slow repressurization ramp and the extended period during which the valves were not actuated are also con-sidered possible contributors to the incident.
Georgia Power has arranged with GE and with cooperating licensees for screening tests to be done on additional SRVs at Wyle Laboratories. Valves which are pressurized at the 0.5 psf /second ramp to 103% of nameplate rating without actuating are to be candidates for diagnostic testing to determine the I
magnitude of forces in the disk-to-seat interface and labyrinth seal area.
Further, examination of interior surfaces will be conducted to locate any physical damage. Two such candidates were found in the recent testing of three SRVs belonging to Northeast Nuclear Energy Company's Millstone Unit 1.*
The NRC issued Infonnation Notice No. 82-41, " Failure of Safety / Relief Valves l
to Open at a BWR," on October 22, 1982, to all nuclear power reactor facilities holding an operating license or construction pennit.
( Ref. 15. )
Millstone Unit 1 is 654 MWe BWR located five miles south of New London, Connecticut, and is operated by Northeast Nuclear Energy.
_ _ _ 1.7 Test Procedure Inadequacies During the month of August 1982, several reports were received from Brunswick Units 1 and 2* concerning the inadequacy of certain periodic test procedures.
While performing technical reviews of these procedures, the licensee for Brunswick reported two major inadequacies:
(1) Newly installed plant modifications were being tested using improper procedures, and (2) Plant modifications had been installed without incorporating a system check of the modification into the station's periodic test procedures.
The results of several of the technical reviews are discussed below.
Unit No./LER No./
Event Date Event Description 1
Following installation of fire suppression water system valves, 82-83/03L fire protection personnel failed to incorporate three system 7-23-82 valves into the test procedure. These valves include an isola-tion valve to the diesel generator building deluge valve pit, a fire protection loop sectionalizing valve, and an isolation valve to the Unit I reactor building deluge valve pit.
2 In 1980, plant modifications were installed on the core auxiliary 82-91/03L cooling (CAC) system which added additional relays to each CAC 7-27-82 isolation valve. This plant modification was installed to address concerns of all CAC valves reopening when an isolation signal was reset. The modifications install an additional relay downstream from the master relay in the logic chain to each valve, so that not only must the master relay be reset for the valve to be opened, but each valve's individ-ual isolation relay must be reset. This change and other changes performed under the modification assure positive control of the reopening of the CAC valves following an isolation. Once the modifications were complete, neither engineering nor instrumentation and control personnel identified the additional relays as requiring the response testing.
2 During a review of plant testing procedures, the licensee 82-97/03L determined that the response time test procedure for the
[
7-27-82 reactor water cleanup low level No. 2 isolation was not j
including the armature travel time of the last relay in the logic chain. The test as written used contacts on these relays that were nomally open instead of the normally closed contacts which isolate the valves. As a result, the time being measured was the time for closed contacts to break, not the time for open contacts to close.
The teminals had been improperly identified when the original test procedure was written.
I
- Brunswick Units 1 and 2 are 790 MWe BWRs located three miles north of l
Southport, North Carolina, and are operated by Carolina Power and Light Company.
i
_ _ _ _ Unit No./LER No./
Event Date Event Description 2
During a review of time response procedures for core spray 82-100/03L and low pressure coolant injection (LPCI) initiation, the 8-2-82 licensee noted that the logic associated with low vessel steam dome pressure was not being tested. The initiation logic for both core spray and LPCI requires either a low-low-low level signal or a high drywell pressure signal con-current with a low vessel steam dome pressure to initiate the system. Current time response testing only checks the low-low-low level signal and the high drywell signal.
These test procedures have been upgraded to correct inadequacies, and all required testing due to such inadequacies is being performed.
In addition, the licensee is continuing technical reviews of periodic test procedures to determine if other test procedure inadequacies exist.
It should be noted that, in general, periodic review of testing procedures to assure compliance with technical specification requirements is a good management practice. (Refs.16 through 19.)
1.8 Loss of 24 Volt DC Power Common to Two Units On June 19, 1982, during a scheduled loss of power test on Peach Bottom Unit 2, an inadvertent startup of the emergency core cooling system (ECCS) was initiated on Unit 3 while at full power.* High pressure coolant injection (HPCI) and reactor core isolation cooling (RCIC) initiated. The resulting cold water injection to the reactor pressure vessel caused the Unit 3 power level to increase rapidly to approximately 114%.
After about 30 se:onds, the transient was terminated by the operator who manually tripped HPCI dnd RCIC. Since the neutron high flux scram was 118%, the plant continued to operate throughout the event.
ECCS initiation on Unit 3 was caused by a defective ELMA de power supply which feeds a 24 V de logic circuit. The logic circuit is normally powered from the 125 V de system (Unit 3) through a " TOPAZ" inverter and the ELMA power supply.
Backup or alternate power to this logic circuit comes from a common 120 V ac distribution panel through a separate ELMA power supply. The common distri-bution panel is fed from the 4 kV emergency bus (Unit 2). The ac feeds to the ELMA power supply core were added primarily to improve availability.
Although the defective ELMA power supply was not capable of supplying power, it could run at full voltage under no-load conditions. As a result, the power supply voltage monitor did not annunciate the degraded condition, and the operator was unaware that the ECCS logic circuit was powered from the backup 120 V ac distribution panel. When this distribution panel was interrupted as required for the Unit 2 loss of power test, the Unit 3 ECCS logic lost its alternate source. Output voltage from the defective de-supplied ELMA power supply decreased, causing instrtment signals for channels B and D to decrease and indicate low water level; however, there was still sufficient voltage to operate the trip units.
In addition to HPCI and RCIC injection, the low pressure coolant injection (LPCI) and core spray systems initiated automatically. The LPCI started, but isolation valves did not open due to the high pressure interlock. The Peach Bottom Unit 2 is a 1051 MWe BWR and Unit 3 is a 1035 MWe BWR. Both are located 19 miles south of Lancaster, Pennsylvania and are operated by Phila-delphia Electric.
I core spray pumps did not start because of the load sequencing time delays.
The licensee s analysis to detennine the effect of HPCI/RCIC cold water injec-tion indicates that core thermal limits were not exceeded.
As a result of this incident, the following actions have been taken to prevent future occurrences of this nature:
(1) Power supplies are being monitored on a weekly basis pending final design changes. Since degraded operation of the ELMA power supplies is charac-terized by a high ripple voltage imposed on the normal 24 V de output, ripple voltage will be checked weekly to verify operation within accept-able limits.
4~
(2) A study has been completed which addresses plant design and possible logic cross connections. A licensee review has determined th6t there are no design errors causing logic cross-connections, and that this problem was a result of the unique configuration for the loss of power test.
(3) The ELMA ECCS power supplies will be replaced as soon as possible with j
environmentally and seismically qualified components.
(4)
Initial review indicated that additional loading on the ELMA power supplies may extend their life. The possibility of better loading the system was investigated and determined not to be feasible.
(5) An investigation has been initiated to evaluate the possibility of separating
)
the ac feeds to the power supplies from the common distribution panel.
i This incident at Peach Bottom was the result of a unique test configuration which does not exist cader normal operating conditions.
It is however, an example of a multiple unit system's interaction that needs to be considered when performing special tests to avoid initiating unnecessary plant transients.
(Ref. 20.)
1.9 Common Cause Feedwater Pump Failures In two similar events occurring in December 1981 at Zion Unit 2,* motor-driven auxiliary feedwater ( AFW) pumps failed to start on steam generator low-low level signal following reactor trips. These events are being reported at this time i
to emphasize a recent NRC finding during a review of licensee event reports, that routine single-pump testing will not detect a common cause failure of l
both pumps from a multiple pump startup.
l
(
The first event, on December 6,1981, involved AFW pump 28. While turbine driven l
AFW pump 2A was out of service for maintenance, a reactor trip occurred. On I
receiving low-low level signals from two steam generators, AFW pump 2C automa-tically started successfully, while AFW pump 2B had to be manually started from the control room. Both pumps automatically started several hours later during j
reactor protection logic tests.
l The second pump failure occurred on December 11, 1981, also during a reactor trip, when AFW pumps 2B and 2C failed to automatically start on receiving low-low level signals.
Both pumps were manually started from the control room. Turbine driven pump 2A was still out of service for maintenance at the time of the event.
Zion Unit 2 is a 1040 MWe PWR located 40 miles north of Chicago, Illinois, and is operated by Commonwealth Edison.
-_ _ _ An investigation conducted on December 12, 1981, led to the discovery that the transient effects of simultaneous pump start cause sections of the suction line near the pump to sustain temporary pressure reduction, thus causing a reverse flow in the swing line of the suction pressure switch. The sensed ' suction pressure takes a momentary drop, despite an actual abundant amount of available suction head. Thus, the simultaneous start of the two motor-driven AFW pumps caused their sensed suction pressures to drop below their low suction pressure trip setpoints, tripping the pumps.
A time delay has been installed on the starting circuitry of motor-driven AFW pumps 2B and 2C. This relay will cause the low suction pressure trip to be momentarily bypassed during pump start. There will be ample time for suction flow to stabilize before insertion of the suction pressure trip, without degrading the pumps' protection from cavitation. Tests on AFW pumps 2B and 2C yielded acceptable results following the modification.
(Refs. 21 and 22.)
1.10 Failed Dowel Pins Disable Diesel Generator 4
On July 1,1982, Brunswick Unit 1* was operating at steady state power and Unit 2 was down for refueling. While running the No. 2 diesel generator for maintenance inspection, it was observed that the diesel jacket water discharge pressure was fl uctuating. Shortly thereafter, the diesel tripped due to low jacket water pressure. This trip initiated a lockout of the diesel, making it unavailable for automatic operation. At the time, the remaining three diesel generators were availab1r for automatic emergency operation.
An inspection of the diesel jacket water engine driven pump revealed that the shaft was not turning. Further inspection of the engine driven pump determined that the two dowel pins in the flex drive coupling drive plate had sheared.
These pins provide alignment and a mechanical union between the engine drive shaft and the plate. Metallurgical examination of the dowel pins concluded that the failures resulted from metal fatigue of the pin material, probably due to the large number of engine starts (1600). The failure of the pins, in turn, caused the failure of the eight drive plate lateral fastening capscrews.
(Since these capscrews are not designed for the torque loading of the crankshaft against the plate, their failure is expected.) The bore of the drive plate and diameter of the attachment end of the engine crankshaft were hieasured and found worn in excess of the design tolerances. This is attributed to the failure of the dowel pins.
The licensee has replaced cap screws and dowel pins on diesel generators Nos. 2 and 4, where broken screws and cracked dowel pins were found after disassaably.
The disassembly of diesel generator No.1 also has disclosed some broken cap screws. Diesel generator No. 3 will be disassembled and checked in the same
)
fashion. The licensee stated that such components on all four diesel generators will be replaced before the units return to power. As a result of these findings, Nordberg, the diesel generator manufacturer, has recommended that coupling cap screws and dowel pins be replaced after 1200 starts. The only other nuclear application of Nordberg diesel generators is at the McGuire facility, where they have had only dout 150 startups on their diesels.
The McGuire licensee has been notified of the Brunswick event.
(Refs. 23 and 24.)
Brunswick Units 1 and 2 are 790 MWe BWRs located three miles north of Southport, North Carolina, and are operated by Carolina Power and Light.
I 1.11 References 1.
NRC Memorandum from M. Chiramal, AE0D, to K. Seyfrit, AE00, transmitting engineering evaluation of " Quad Cities Nuclear Power Station Loss of Auxiliary Electrical Event of Unit Two, June 22, 1982," (AE0D/E250),
November 8, 1982.
2.
Letter from N. Kalivianakis, Commonwealth Edison, to J. Keppler, NRC/R-III, transmittirg Special Report on " Loss of Auxiliary Electrical Power Event of June 22, 1982," for Docket Nos. 50-254 and 50-265, July 9,1982.
3.
Letter from R. Spessard, NRC/R-III, to C. Reed, Commonwealth Edison, transmitting Inspection Reports 50-254/82-10(SPRP) and 50-265/
82-11(SPRP), October 22, 1982.
4.
NRC, Preliminary Notifications PNO-II-82-72 (June 28,1982) and -72A (June 29, 1982).
5.
NRC/II, Inspection Report 82-26 for 50-269, 50-270, and 50-287 p.1-4, July 1982.
6.
NRC, Inspection Report 50-271/82-07, May 12, 1982.
7.
NRC, Investigation Report 50-271/82-12, July 21,1982.
8.
NRC, Preliminary Notification, Docket No. 50-271, License No. DPR-28, October 15, 1982.
9.
NRC Memorandum from C. Michelson, Director, AE00, to H. Denton, Director, NRR, and V. Stello, Director, IE, "Inoperability of Instrumentation due to Extreme Cold Weather," June 15, 1981.
- 10. NRC Memorandum from M. Chiramal, AE0D, to C. Michelson, Director, AE00, "Inoperability of Instrumentation due to Extreme Cold Weather,"
June 18, 1982.
- 11. NRC Memorandum from E. Brown, AE00 to K. Seyfrit, AE00, " Fuel Assembly Degradation While in the Spent Fuel Storage Pool" (AE0D/E242),
October 21, 1982.
- 12. Northern States Power Company, Docket Nos. 50-282 and 50-300, Licensee Event Reports 81-31 (December 30, 1981) and 81-31 Rev. 1 (May 1982).
- 13. Georgia Power Company, Docket No. 50-321, Licensee Event Report 82-60, July 15,1982.
- 14. NRC/II, Inspection Report 50-321/82-27, approved September 2,1982.
- 15. NRC, Infonnation Notice 82-41, " Failure of Safety / Relief Valves to Open at a BWR," October 22, 1982.
. 16. Carolina Power and Light Company, Docket No. 50-325, Licensee Event Report 82-83, August 20, 1982.
- 17. Carolina Power and Light Company, Docket No. 50-324, Licensee Event Reports 82-84, 82-91, and 82-97, August 23, 1982.
- 18. Carolina Power and Light Company, Docket No. 50-324, Licensee Event Report 82-100, August 24, 1982.
- 19. Carolina Power and Light Company, Docket No. 50-325, Licensee Event Report 82-90, August 24, 1982.
- 20. Letter from M. J. Cooney, Philadelphia Electric Company to R. C. Haynes, NRC/R-I, re: Docket Nos. 50-277 and 50-278, July 14,1978.
- 21. Commonwealth Edison Company, Docket No. 50-304, Licensee Event Report 81-31, January 5,1982.
- 22. Commonwealth Edison Company, Docket No. 50-304, Licensee Event Report 81-33, January 8,1982.
- 23. Carolina Power and Light Company, Docket No. 50-325, Licensee Event Report 82-78, July 30,1982.
- 24. NRC, Preliminary Notification, PNO-II-82-88, July 28,1982.
J
. 2.0 ABSTRACTS OF OTHER NRC OPERATING EXPERIENCE DOCUMENTS 2.1 Abnormal Occurrence Reports (NUREG-0090) Issued in July-August 1982 An abnormal occurrence is defined in Section 208 of the Energy Reorganization Act of 1974 as an unscheduled incident or event which the NRC determines is significant from the standpoint of public health or safety. Under the provi-sions of Section 208, the Office for Analysis and Evaluation of Operational Data reports abnormal occurrences to the public by publishing notices in the Federal Register, and issues quarterly reports of these occurrences to Congress in the NUREG-0090 series of documents. Also included in the quarterly reports are updates of previously reported abnormal occurrences, and sunmaries of cer-tain events that may be perceived by the public as significant but do not meet the Section 208 abnormal occurrence criteria.
Date Issued Report 8/82 REPORT TO CONGRESS ON ABNORMAL OCCURRENCES: JANUARY-MARCH 1982, NUREG-0090, Vol. 5, No. 1 During the report period, there were four abnormal occurrences at the nuclear power plants licensed to operate.
The four occurrences involved: (1) diesel generator engine cooling system failures, (2) pressure transients during shutdown, (3) major deficiencies in management controls, and (4) a steam generator tube rupture. The report continues to provide update information concerning the accident at Three Mile Island and other previously reported abnormal occurrences.
In addition, low concentrations of tritium detected in groundwater at the Sheffield low-level waste disposal facility is discussed as an item of interest that does not meet abnormal occurrence criteria.
r 2.2 Bulletins, Circulars, and Information Noticos Issued in July-August 1982 The Office of Inspection and Enforcement periodically issues Bulletins, Circulars, and Information Notices to licensees and holders of construction permits. During the report period,14 Infomation Notices were issued.*
Bulletins are used primarily to communicate with industry on matters of generic importance or serious safety significance; i.e., if an event at one reactor raises the possibility of a serious generic problem, an NRC Bulletin may be issued requesting licensees to take specific actions, and requiring them to submit a written report describing actions taken and other infonnation NRC may need to assess the need for further actions.
A prompt response by affected licensees is required and, failure to respond appropriately may result-in an enforcement action, such as an order for suspension or revocation of a license. When appropriate, prior to issuing a Bulletin, the NRC may seek comments on the matter from the industry
( Atomic Industrial Forum, nuclear steam system suppliers, vendors, etc.),
a technique which has proven effective in bringing faster and better responses from licensees. Bulletins generally require one-time action and reporting. They are not intended as substitutes for rey? sed license conditions or new requirements.
Circulars notify licensees of actions NRC recommends be taken. Although written responses are not required, the licensees are asked to review the information and implement the recommendations if they are applicable to their facility.
Information Notices are rapid transmittals of infonnation which may not have been completely analyzed by NRC, but which licensees should know.
They require no acknowledgment or response, but recipients are advised to consider the applicability of the information to their facility.
Information Date Notice Issued Subject 82-22 7/9/82 FAILURES IN TURBINE EXHAUST LINES 1
This notice provides a sunmary of the June 28, 1982 event at Oconee Unit 2 where a 4-sq-ft rupture occurred in the elbow of a 24-inch diameter feedwater heat extraction line whicn is supplied steam from the high-pressure turbine exhaust. The rupture was attributed to piping degradation that results from steam erosion. Since ultrasonic inspection in March 1982 had revealed substantial erosion of the elbow (although less than the licensee's criterion for rejection), the licensee believes that sustained reduced power operation and resultant lower quality steam contributed to accelerated l
erosion.
In addition, the Institute of Nuclear Power Operations (INPO) has identified similar failures of steam lines occurring in 1982 at Yermont No Bulletins or Circulars were issued in July-August 1982.
I I
l l l
Infomation Date Notice Issued Subject Yankee, Trojan, Zion Unit 1, and Browns Ferry Unit 1.
No specific action or response to the notice was required.
82-23 7/16/82 MAIN STEAM ISOLATION VALVE (MSIV)
LEAKAGE IE has completed a survey of MSIV per-fomance at BWP,s for the years 1979 through 1981.
IE found that 19 of 25 operating BWRs had MSIVs which failed to meet, during one or more surveillance tests, the limiting condition for operation (LC0) which specifies the maximum permissible leak rate. The neber of MSIV test failures exceeded 151 and occurred with MSIVs supplied by all three MSIV vendors; i.e., Atwood &
Morrill, Crane, and Rockwell. This notice was sent to all BWR licensees and construction permit holders.
82-24 7/20/82 WATER LEAKING FROM URANIUM HEXAFLUORIDE OVERPACKS This notice provides information con-cerning incidents where water was found leaking from uranim hexafluoride overpacks during shipment from Department of Energy (DOE) facilities to NRC urania hexafluoride processors. Some overpacks are being returned to DOE facilities without all the bolts installed in the overpack to prevent rain and condensa-L tion water from building up in the f
overpack during transportation. This water may leak from the overpack through loose bolts, defective seals or rusted-through areas. This notice was sent to I
all licensed enriched uranium fuel fabri-l cation plants.
82-25 7/20/82 FAILURES OF HILLER ACTUATORS UPON GRADUAL LOSS OF AIR PRESSURE Mississippi Power and Light Company has reported that a large neber of isolation valves in the instraent air system at Grand Gulf Nuclear Station failed to pass l
Information Date Notice Issued Subject test requirements. The valves were supplied by the William Powell Company and equipped with actuators supplied by Ralph A. Hiller Company. This notice was sent to all licensees and construction permit holders.
82-26 7/22/82 RCIC AND HPCI TURDINE EXHAUST CHECK
' VALVE FAILURES A number of reactor core isolation cooling (RCIC) turbine exhaust check valve failures had been reported during the 20-month period prior to issuance of this notice.
All of the failures dealt only with RCIC turbine exhaust check valves. However, the high pressure coolant injection (HPCI) exhaust system has the same type valve in a similar system configuration; thus,it is reasonable to expect similar problems with the HPCI turbine exhaust check valve.
This notice was provided to all BWR licensees and construction permit holders as an early notification of a potentially significant problem.
82-27 8/5/82 FUEL ROD DEGRADATION RESULTING FROM BAFFLE WATER-JET IMPINGEMENT On May 6,1982, Portland General Electric reported abnormal fuel clad degradation identified during a pre-planned fuel inspection to locate suspected leaking fuel assemblies. Fuel rod damage involved 17 fuel assemblies examined at the end of Cycle 4 operation.
Portions of fuel rods were found missing and loose fuel pellets were discovered and retrieved from reactor vessel internals and the refueling cavity. This notice was distributed to all licensees and construction permit holders.
82-28 7/23/82 HYDR 0 GEN EXPLOSION WHILE GRINDING IN THE VICINITY OF DRAINED AND OPEN COOLANT SYSTEM On April 10, 1982, a hydrogen explosion occurred at Unit 1 of Arkansas Nuclear One while maintenance personnel were
Information Date Notice Issued Subject grinding a recently cut high-pressure injection (HPI) pipe, approximately.18 inches from the nozzle connecting the HPI pipe to the reactor coolant system (RCS) piping. At the time of the explosion, the RCS was partially drained and the water level in the RCS piping was just below the HPI nozzle to permit radiography of the nozzle and subsequent repair. There were no physical injuries as a result of this event. This notice was distributed to all licensees and construction permit holders.
82-29 7/23/82 CONTROL R00 DRIVE (CRD) GUIDE TUBE SUPPORT PIN FAILURES AT WESTINCHOUSE PWRs Since 1978, several failures of the CRD guide tube support pins have occurred.
Prior to May 1982, at which time a guide pin tube failed at North Anna 1, these failures had occurred only at foreign reactors (Japan and France). This notice was distributed to all licensees and construction permit holders using a i
Westinghouse designed NSSS.
82-30 7/26/82 LOSS OF THERMAL SLEEVES IN REACTOR COOLANT SYSTEM PIPING AT CERTAIN WESTINGHOUSE PWR POWER PLANTS Past operating experience has identified fatigue failure problems associated with nozzle-thermal sleeve assemblies in piping systems of both BWR and PWR plant designs.
Similar fatigue failure problems have occurred more recently in certain PWR plants designed by Babcock & Wilcox.
This notice was sent to all licensees, construction permit holders and applicants for operatit] license.
82-31 7/28/82 OVEREXPOSURE OF DIVER DURING WORK IN FUEL STORAGE P00L On June 1,1982, while installing fuel rack support plates in the Indian Point l
No. 2 fuel storage pool, a contractor diver received an exposure of about i
_--_ Information Date Notice Issued Subject 8.7 rems to the head. A second diver, also working in the pool on June 1, receiving a whole body dose of about 1.6 rems. This notice was sent to all licensees and construction permit holders.
82-32 8/19/82 CONTAMINATION OF REACTOR COOLANT SYSTEM BY ORGANIC CLEANING S0LVENTS On April 24, 1982, while at full power, the Hatch Unit 1 plant experienced a rapid increase in water conductivity in the condensate, feedwater, and primary reactor systems. This was accompanied by increasing radiation readings on the main ste3m line and offgas monitors.
Power reduction began at 1430 CST. At 1730 CST with the plant at 35% power, the reactor water conductivity exceeded the technical specification, and at 1930 CST the plant was manually scrammed.
This event resulted in a plant outage of more than a month. This notice was sent to all licensees and construction permit holders.
82-33 8/20/82 CONTROL OF RADIATION LEVELS IN UNRESTRICTED AREAS ADJACENT TO BRACHYTHERAPY PATIENTS During the course of several inspections it was reported that radiation levels exceeded regulatory limits in unrestricted areas adjacent to brachytherapy patients.
This notice discusses the applicable regulatory limits and associated require-ments, and was sent to all medical institutions.
82-34 8/20/82 WELDS IN MAIN CONTROL PANELS Inspections of welds in main control panels supplied by Systems Control of Iron Mountain, Michigan; Reliance Electric of Stone Moun-tain, Georgia; and Comsip of Linden, New Jersey have disclosed nunerous welding practices not in accordance with the Americal Welding Society Standards and several quality assurance practices not in compliance with the vendors' procedures or NRC requirements. This notice was sent to all licensees and construction permit holders.
Information Date Notice Issued Subject 82-35 8/25/82 FAILURE OF THREE CHECK VALVES OF HIGH PRESSURE INJECTION LINES TO PASS FLOW At Davis-Besse Unit 1 on June 4,1982, a stop check valve (HP-57) in the normal makeup system failed to pass flow although 120 psid was applied across the valve.
Normal opening pressure is about 5 psid.
The problem was discovered while filling the reactor coolant system (RCS) using a small low-head pump following a refueling and maintenance outage. This notice was sent to all licensees and construction permit holders.
c
___ 2.3 Engineering Evaluations and Case Studies Issued in July-August 1982 The Office for Analysis and Evaluation of Operational Data ( AE0D) has as a primary responsibility the task of reviewing the operational experience reported by NRC nuclear power plant licensees. As part of fulfilling this task, it selects events of apparent interest to safety for further review as either an engineering evaluation or a case study. An engineering evaluation is usually an immediate, general consideration to assess whether or not a more detailed, protracted case study is needed. The results are generally short reports, and the effort involved usually is a few staffdays of investigative time.
Case studies are in-depth investigations of apparently significant events or situations. They involve several staffmonths of engineering effort, and result in a formal report identifying the specific safety problems (actual or potential) illustrated by the event and recommending actions to improve safety and prevent recurrence of the event. Before issuance, this report is sent for peer review and comment to at least the applicable utility and appropriate NRC offices.
These AE00 reports are made available for information purposes and do not impose any requirements on licensees.
The findings and recommendations contained in these reports are provided in support of other ongoing NRC activities concerning the operational event (s) discussed, and do not represent the position or requirements of the responsible NRC program office.
Engineering Date Evaluation Issued Subject E230 7/7/82 WATER IN Tile FUEL OIL TANK AT SURRY POWER STATION UNIT 2 Surry reported that water from the fire suppression system was inadvertently added to the above ground diesel oil storage tank.
Four thousand gallons of water entered the tank via the fire protection sparger which is located inside the tank. Since this tank supplies diesel fuel to the underground, day and wall tanks for all diesel generators at the Surry site, the water represented a i
potential common mode failure for all the diesels. Administrative controls were increased to prevent recurrence.
E231 7/19/82 LOSS OF SHUTDOWN COOLING DUE TO TRIP 0F LOW PRESSURE SAFETY INJECTION PUMP AT MILLSTONE UNIT 2 On 12/9/81, while shutdown in mode 5 with shutdown cooling monitoring reactor coolant system (RCS) temperature at 90*F, testing of 345K switchyard breakers caused actuation of a turbine generator trip scheme which
)
.. I.
Engineering Date Evaluation Issued Subject i
provided a trip signal to a running i
low pressure safety injection (LPSI) pump. The trip of the LPSI pump caused RCS temperature to increase to 208'F 4
in about 2 minutes, which exceeded technical specification limits. Design modifications and procedure changes are being made to prevent recurrence.
E232 7/19/82 P0TENTIAL DEFICIENCY IN THE SIGMA LUMI-GRAPH INDICATORS MODEL NUMBER 9270 i
At San Onofre Unit 1, aging R2 and R6 resistors caused erroneous readings in the Model 9270 lumigraph indicator.
The R2 resistor on the alarm input assembly board can overheat, change in value, and cause a display malfunc-tion; R2 and R6 resistors on the output i
card of all dual indicators can produce similar results; and the R2 resistor in the output card may also change in value.
d While an. indicator malfunction may occur, no loss of alarm will occur in alarm
(
units. Generic implications are under review in another NRC office.
E233 7/28/82 CARBON DIOXIDE SYSTEMS USED FOR FIRE PROTECTION IN OR ADJACENT TO CRITICAL l
AREAS A recent occurrence at Grand Gulf Unit 1 involved the actuation of the carbon dioxide i
system in the emergency core cooling system penetration room. The event was initiated by i
a short in the initiation circuit relay. The system operated repeatedly until the pressure in the room forced open the locked door to the I
auxiliary building. The event raised several concerns over the use of carbon dioxide in critical plant areas and the adequacy of licensing reviews.
E234 8/11/82
. FAILURE IN A SECTION OF 4kV BUS CABLE MANUFACTURED BY OK0 NITE A 4kV bus cable, manufactured by Okonite Cable Company and installed about 10 years ago in the Beaver Valley Power Station, failed with melting of the aluminum conductor. Testing indicated that the failure was likely due to external causes or a random failure, though the exact cause is unknown.
(See PRE Vol. 4. No 3, l
pp. 9-10 for details.)
1
- Engineering Date Evaluation Issued Subject E235 8/11/82 WIRING ERROR IN HANDSWIlCH FOR SOLEN 0ID CONTROL VALVES ASSOCIATED WITH HPCI SYSTEM STEAM CONDENSING MODE PRESSURE CONTROL VALVE AT DUANE ARNOLD In response to IE Bulletin 80-06, Duane Arnold found the equipment involved in the control of the steam condensing made of residual heat removal would realign to the non-accident mode if accident signals were reset. A design change to correct the problem was incorrectly implemented. The deficiency was discovered and corrected and the QA breakdown was being pursued.
E236 8/25/82 BRUNSWICK STEAM ELECTRIC PLANT (BSEP) UNIT 2 LOSS OF RESIDUAL HEAT REMOVAL SERVICE WATER ON 1/16/82 Subsequent to a reactor scram, nonnal sup-pression pool and shutdown cooling could not immediately be established because both residual heat removal service water trains were inoperable. Both trains were operable within 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />. An open circuit breaker interrupted power to the low suction pres-sure protection logic circuit, disabling the two B train pumps; a leaking oil-filled chemical seal attached to the A train pinp suction pressure switch rendered the pressure switch inoperable and caused a start incident of the two A train pumps.
Alternate cooling was available, if needed.
System availability can be improved by reducing susceptibility to single failures.
E237 8/25/82 PORY BLOCK VALVE FAILURE AT H.B. ROBINSON, UNIT 2 Prior to hydrostatically testing the RCS, operators attempted unsuccessfully to close the pressurizer-operated relief valve (PORV) block valves. Subsequently, an operator went into containment and manually closed one valve; after manipulating the control lever on the second valve; the motor energized, drove the valve closed, continued to run, and eventually broke three of the four yoke bolts to the valve body. Review found the torque switch was not wired into the circuit, the wrong yoke bolt material was used, and the wrong lubricatioa was used for the valve stem.
The valyc operator was classified as non-safety grade, and therefore not subject to QA.
f f.
l l
Case Date Study Issued Subject i
C205 8/82 ABNORMAL TRANSIENT OPERATING GUIDELINES (AT0G) AS APPLIED TO THE APRIL 1981 OVERFILL EVENT This survey report provides:
(1) a detailed analysis of the April 8,1981, steam generator overfill transient at Arkansas Nuclear One Unit 1 ( ANO-1);
(2) an analysis of the draft AT0G pre-i pared by B&W for ANO-1, and its guidance on mitigating steam generator overfill transients; (3) a simple comparison of the draft AT0G with the actual event.
The report concludes that those units which rely on manual operator action to mitigate overfill transients will likely require equipment modifications and additions to supplement emergency procedures.
C204 7/82 LOSS OF SALT WATER COOLING EVENT This survey report concerns the loss of salt water cooling at San Onofre Unit 1 on March 10, 1980. The report provides:
(1) an analysis of the overall salt water cooling system; (2) a detailed analysis of the event and why it warranted a special study; (3) a discussion of the licensee's and NRC's actions during and after the event.
i i
l l
i l
l l
l 2.4 Regulatory and Technical Reports Issued in July-August 1982
]
The abstracts listed below have been selected from the Office of Administration's quarterly publication, Regulatory and Technical Reports (NUREG-0304). This document compiles abstracts of the fomal regulatory and technical reports issued by the NRC staff and its contractors.
Bibliographic data for the reports is also included. Copies and subscriptions of NUREG-0304 are available from the NRC/GP0 Sales Program, PHIL-016, Washington, DC 20555, or on (301) 492-9530.
Issue Subject NUREG-0525 SAFEGUARDS
SUMMARY
EVENT LIST R05 July 1982 The Safeguards Summary Event List (SSEL) provides brief summaries of several hundred safeguards-related events involving nuclear material or facilities regulated by the U.S. Nuclear Regulatory Commission (NRC). Events are described under the categories of bomb-related, intrusion, missing and/or allegedly stolen, transportation, vandalism, arson, fireams, sabotage and miscellaneous.
The information contained in the event descriptions is derived primarily from official NRC reporting channels.
NUREG-0714 OCCUPATIONAL RADIATION EXPOSURE REPORT Yol. 1 August 1982 This report summarizes the information report for calendar year 1979 by all NRC licensees to the Commission's centralized repository of personnel occupational radiation exposure information. The bulk of the information in the report is derived from annual reports that were required to be submitted by all NRC licensees pursuant to 10CFR20.407. This is the second year that all NRC licensees were required to submit an annual exposure report. Previously only certain categories - commercial nuclear power reactors, industrial radiographers, fuel fabricators and processors and commercial distributors of byproduct material - of NRC licensees had submitted such reports. The requirement of 10CFR20.408 for the submission of termination reports continued to apply to only these four categories, and some analysis of the
(
data contained in these reports is also presented. A i
brief description of personnel overexposures reported by NRC licensees is included as well.
NUREG-0872 STUDY OF USING LICENSEE EVENT REPORTS FOR A STATISTICAL July 1982 ASSESSMENT OF THE EFFECT OF OVERTIME AND SHIFT WORK l
ON OPERATOR ERROR A study based upon the reported licensed operator errors from January 1981 to determine if a valid statistical determination could be made of the effects of shift work and overtime on operator error. The study concludes that the data reported in the Licensee Event Reports are
. Issue Subject l
l inadequate to draw conclusions on the influence of l
overtime and shift work on operator error. The analysis did show that the errors are not uniform over the hours of the day or the days of the week; the causes of non-uniformity could not be determined.
NUREG-0899 GUIDELINES FOR THE PREPARATION OF EMERGENCY OPERATING August 1982 PROCEDURES The purpose of this document is to identify the elements for utilities to prepare and implement a program of Emergency Operating Procedures (E0Ps) for use by control room personnel to assist mitigating the consequences of a broad range of accidents and equipment failures.
This document applies only to the E0Ps so designated; it does not address emergency preparedness or emergency pl anning.
It also represents the resolution of comments on NUREG-0799, " Draft Criteria for Preparation of Emergency Operating Procedures."
NUREG-0920 U.S. NUCLEAR REGULATORY COMMISSION 1981 ANNUAL REPORT July 1982 This report addresses all NRC activities, policies, and decisions made during the reporting period, complete with i
illustrations, charts, and treatment of technical material in lay language for constanption by the lay public.
NUREG/CR-2255 EXPERT ESTIMATION OF HUMAN ERROR PROBABILITIES IN NUCLEAR August 1982 POWER PLANT OPERATIONS:
REVIEW 0F PROBABILITY ASSESSMENT AND SCALING l
This report reviews probability assessment and-psychological scaling techniques that could be used to estimate human error probabilities (HEPs) in nuclear power plant operations.
The techniques rely on expert opinion and can be used to estimate HEPs where data do not exist or are inadequate.
These techniques have been used in various other contexts and have been shown to produce reasonably accurate probabilities.
Some problems do exist, and limitations are discussed.
Additional topics covered include methods for combining t
l estimates from multiple experts, the effects of training l
on probability estimates, and some ideas on structuring the relationship between performance shaping factors and HEPs. Preliminary recommendations are provided along with cautions regarding the costs of implementing the recommendations. Additional research is required before definitive recommendations can be made.
l l
[
_ Issue Subject NUREG/CR-2781 EVALUATION OF WATER HAMMER EVENTS IN LIGHT WATER July 1982 REACTOR PLANTS This document presents the results of an evaluation of water hammer events in LWR power plants. The evaluation was based upon reports of actual events, typical plant design drawings, and operation procedures.
Included in this report are design and operating recommendations for preventing or mitigating water hammer occurrence.
NUREG/CR-2796 COMPRESSED-AIR AND BACKUP NITR0 GEN SYSTEMS IN NUCLEAR July 1982 POWER PLANTS This report reviews and evaluates the performance of the compressed-air and pressurized-nitrogen gas systems in commercial nuclear power units. The information was collected from readily available operating experiences, licensee event reports, systems designs i
in safety analysis reports, and regulatory documents.
The results are collated and analyzed for significance and impacts on power plant safety performance.
Under certain circumstances, the " fail-safe" philosophy for a piece of equipment or subsystem of the compressed-air systems initiated a series of actions culminating in reactor transient or unit scram. However, based on this
^.udy of prevailing operating experiences, reclassifying the compressed-gas systems to a higher safety level will alleviate nuclear power plant problems caused by inadequate maintenance, operating procedures, and/or practices. Conversely, because most of the problems were derived from the sources listed previously, up-grading of both maintenance and operating procedures will not only result in substantial improvement in the perfomance and availability of the compressed-air (and backup nitrogen) systems, but in improved overall plant performance.
(
NUREG/CR-2798 EVALUATION OF EVENTS INVOLVING UNPLANNED BORON DILUTIONS f
July 1982 IN NUCLEAR POWER PLANTS This report reviews and evaluates events concerned with the inadvertent dilution of boron concentrations to the reactor coolant system for pressurized-water-cooled themal reactors in commercial service. The safety concern is the unplanned addition of reactivity. The information was collected from operating experiences, licensee event reports, system designs in safety analysis reports, and regulatory documents. The results are collated and analyzed for significance and impact on power plant safety performance.
i
- Issue Subject Several operating experience events were selected for analysis because they meet the criteria for safety l
signt ficance. However, no boron dilution incidents resulted la a reactivity excursion or transient that scramaed a unit, nor was a reactor protection system
[
challenged by any of the events. The most common cause for unplanned boron dilutions was haan error, l
of which one was a common-mode / common-cause failure.
For each recorded event, the operator had sufficient time to diagnose and correct the cause of the inadver-tent dilution before the shutdown safety margin was lost or seriously challenged.
NUREG/CR-2799 EVALUATION OF EVENTS INVOLVING DECAY HEAT REMOVAL July 1982 SYSTEMS IN NUCLEAR POWER PLANTS This report reviews and evaluates events placed in the NSIC file involving the removal of decay heat in U.S.
commercial boiling-and pressurized-water reactors from June 1979 through June 1981. The information was collected from operating experiences, licensee
. event reports, systems designs in safety analysis reports, and other regulatory documents. The results were collated and analyzed according to safety sig-i nificance and cause of event.
k Thirty-eight reported events in these 2.1 years meet the criteria for safety significance.
Steam bubble formation in the reactor vessel head during natural l
circulation cooldown at St. Lucie 1 was the most j
significant event; operator awareness of the possibility of this occurrence and preparedness for dealing with it was the most important recommendation. Cavitation of residual heat removal peps during decay heat removal' operation was the most common potentially significant event. Davis-Besse 1 had several instances in which an inadvertent signal to the i
safety features actuation system caused the operating l
residual heat removal pumps to align to the dry sep causing pep cavitation.
NUREG/CR-2833 CRITICAL HUMAN FACTORS ISSUES IN NUCLEAR POWER Vols. 1-3 REGULATIONS AND RECOMMENDED COMPREHENSIVE HUMAN FACTORS August 1982 LONG RANGE PLAN This comprehensive long range haan factors plan for nuclear reactor regulation was developed by a Study Group of the Human Factors Society, Inc. This Study Group was selected by the Society to provide a balanced, experienced human factors perspective to the application of human factors scientific and engineering knowledge to the nuclear power generation.
! Issue Subject The report is presented in three volumes. Volume 1 contains an Executive Summary of the 18-month effort and its conclusions. Volume 2 samarizes all known nuclear hurnan factors activity, evaluating this activity, wherever adequate information is available, and describes the recommended long-range (10-year) plan for human factors in regulation. Vol me 3 elaborates each of the human factor issues and areas of concern that have led to recommendations in the long range plan.
)
NUREG/CR-2598 NUCLEAR POWER PLANT CONTROL ROOM TASK ANALYSIS:
July 1982 PILOT STUDY FOR PRESSURIZED WATER REACTORS Report covers nuclear plant task analysis pilot study.
Five data sources were investigated to provide informa-tion for a task analysis, (1) written operating pro-4 cedures (event-based); (2) interviews with subject matter experts (the control room operators); (3) videotapes of the control room operators (senior reactor operators and reactor operators) while responding to each event in a simulator; (4) walk-/
talk-throughs conducted by control room operators for each event; and (5) simulator data from the plant monitoring system (PMS). A Westinghouse pressurized water reactor nuclear power plant simulator was utilized in this study. Four abnormal or emergency events were studied:
nuclear instrument failure; small break loss-of coolant accident (LOCA); steam generator tube leak; and inadvertent safety injection at power. Upon completion of the task analyses, computerized data reduction was perfonned. A PRIME 1-1000 computer was used to manage the task analytic data base.
}
Videotapes, although the richest single source of.
information, frequently left some gaps in the data accumulations, e.g. the exact switch manipulation, and the sequence of manipulations was somr. times difficult to discern. The PMS data sour'.e provided l
that data. This study demonstrated the usefulness of a task analysis methodology that combines tradi-f i
tional data to prcvide a complete task analytic data set. The power of a machine-readable computer data base to support various applications of task analytic data was also demonstrated.
i NUREG/CR-2670 JOB ANALYSIS CF MAINTENANCE MECHANIC POSITION FOR THE NUCLEAR POWER PLANT MAINTENANCE PERSONNEL RELIABILITY MODEL This report is one of a series planned to describe the results of a program to define, develop, validate, and disseminate a methodology for the quantitative prediction l
.- Issue Subject l
of huan reliability in the conduct of maintenance tasks in nuclear power plants (NPPs). ORNL has subcontracted portions of this effort to Applied.
l Psychological Services, Inc. An analysis was per-l formed of the job of maintenance mechanics in i
nuclear power plants in order to provide a part l
of the information required for modeling nuclear plant maintenance activities.
It is believed j
that such a model would provide substantial in-l sights into the various h aan, equipment, and environmental factors likely to affect reliability of maintenance personnel, and thereby suggest and allow evaluation of standards, design changes or other' modifications to improve reliability and minimize public risk. The task list method of job survey was selected because the approach minimizes data acquisition costs, interferes minimally with the work of job inctnbents, is comprehensive and objective, provides quantitative data, and seemed highly appropriate for achieving the goals of the work.
In collaboration with supervisory personnel at BWR nuclear power plants, a comprehensive list of 107 tasks performed by maintenance mechanics was developed. The tasks were grouped within six generic work functions:
remove and install, test and repair, inspect and perform preventive maintenance, miscellaneous, communication, and report preparation. The lists were then assembled into an appropriate questionnaire form. The results of the survey are reported in 4
this study.
1 These documents are available in the NRC Public Document Room at 1717 H Street, Washington, D.C.
20555, for inspection and/or copying for a fee.
l l
l l
i
I
. 1.
2.5 Operating Reactor Event Memoranda Issued In July-August 1982 The Director, Division of Licensing, Office of Nuclear Reactor Regulation (NRR),
dissen;f nates information to the directors of the other divisions and program offices within NRR via the operating reactor event memoranda (OREM) system.
The OREM documents a statement of the problem, background information, the safety signficance, and short and long term actions (taken and planned).
Copies of OREMs are also sent to the Offices for Analysis and Evaluation of I
Operational Data, and of Inspection and Enforcement for their infonnation.
i No OREMs were issued during July-August 1982.
120555078877 1 ANN 4TCPM9U15C US hRC ADM DIV 0F TIOC PDR AUREG COPY POLICY & PUBLICATNS MGT BR W-SCL WASHINGTON DC 20555 I
(
.