ML20024B212
| ML20024B212 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 05/14/1976 |
| From: | J. J. Barton, Toole R GENERAL PUBLIC UTILITIES CORP. |
| To: | GENERAL PUBLIC UTILITIES CORP. |
| References | |
| TASK-*, TASK-GB GPU-2485, NUDOCS 8307070346 | |
| Download: ML20024B212 (14) | |
Text
~
8 r :'
GPW2%b AEC DCCUMENT REVIEll Elant/U The attached AEC document has been reviewed for %st program and design modification requirements for the above Plant / Unit.
f 1
f QVb m
00CUMENT:
Operating Experience, dated:
l){
Current Events - Power Reactors, dated: /
'77 - 2.-l7(,
Other
, dated:
Review of the attached document has concluded.that no action is required.h lhb n
up & I atlaler Dete GA 5-l4-7(o Test'gerintencent Date
/
Review of the attached document has concluded that action is required by:
E
' T s l k m a l 0. C. S L,.A ] G e w h,o A %. a &
E C
$8 f$w Fc u[MM o.
Problem Report (s) 0,
(
has/have been issued.
b go48 Q
- 8 I
k ff,/%
oo St r
'E Test Manager
/ D&te no o<
?/
h5 j0:(
a <1 7
7/I mam Tdit Juper1Ateiident
//
~ 0ats W
07C11 DISTRIBUTICN:
R.W. Heward, Jr.
W.T. Gunn E.T. ":Ca -::t J.E. Kunkel H.A. Nelson R.J. Teole p
J.T. Faulkner
.T..,g w ed>
mw+en e+
e e
ee e
v
.= _
355eBEWFWmw=&q>.Ww m.gTEMMSWR W 5&=Tm m.m 5AM 4
2.pywnn,
-~
.mn = w.-ens =u m
.z-p A.c u
~a
,e.m.
UNITED STATES Sti n
CURRENT EVENTS NUCl. EAR REGULATORY I
POWER REACTORS comssioN SM th EVENTS SELECTED FROM REPORTS SU3MITTED TO THE UNITED STATES NUCLEAR REGULATORY COMMISSION 4
OCTOBER 1975 - FEBRUARY 1976 b i M
2x mN m
FAILURE OF REACTOR C00L1NT PUMP SEALS hg During startup activities frem a ho: shutdown on September 20, 1975 at Unit No. 1 of the Fort Calhoun Station, an increase in contain=ent sump levels indicated a' leak from the upper vapor seal region of reactor coolant pumps C and D.
Although it was within Technical Specification p
li=1ts, the plant was brought to a cold shutdown and the vapor seals of
- Kc the two reactor coolant pu=ps were replaced.
After completion of vapor seal replace =ent, the pumps were pressurized for a leak check and, at pressure, it was discovered the seal pressure breakdown for pump D was incorrect; both first and second stage seals did not indicate a pressure drop. Upon removal of the seal cartridge, it was discovered that five of the eight lower breakdown device cap screws had backed out and damaged the seal coolant recirculation impeller and the bottom supports of the lower breakdown device. All parts were recovered.
I After reassembly of the seal, reactor coolant pumps A and C did not indicate a correct pressure breakdown across the seals, so the seals for both pumps were replaced.
Cause for the initial leakage from vapor seals of pumps C and D could not be determined. However, the problem with the D pump seal caused by the backed out cap screws was the direct result of an inadequate main-tenance procedure covering the rebuilding of reactor coolant pump seals.
A lock wire had not been installed on the pump seal, and the maintenance procedure covering the rebuilding of the seals was too general.
The subsequent i= proper breakdown of seal pressure across the A and C l
pu=p seals was attributed to independent causes; the upper shaft sleeve retaining pin clearance of pump C was grea,ter than nor=al and allowed I
relative movement between the upper and lower shaft sleeves. This resulted in unloading of contact forces between the rotating and.
W 07C12
~ ~ ~ ~ -
--~~ -
~
~
~
,.,%..p i
,w w.
- $DN M6 E42R gewis t
D6 W
stationary faces of the bottom two seals. The i= proper pressure break-N down for pu=p A was the result of crud blockage in the leekoff path i
within the pressure breakdown device. After examination of the seal, it was postulated that this condition night have corrected itself if the a
system had been brought to a higher pressure before shutdown.
Ek Initial leakage through the vapor seal of the pu=p allowed reactor coolant u:
leakage to the containment at=osphere. However, a certain leakage is O
permissible under the Technical Specifications, and at all ti=es the
,5??i measured leakage was below the permissible limit.
The loose parts and da= age found to the D pu=p seal did not jeopardi:e system integrity because all parts were contained within the cavity and were recovered.
The maintenance procedure has been revised to include more specific infor=ation for seal rebuilding and for proper checks for upper shaft sleeve pin clearance.1 VI3 RATION CAUSES LOW FLOW FEEDWATER LINE DMfAGE While decreasing load in preparation for shutdown of Unit 2 of the Quad-Cities Nuclear Power Station on August 31, 1975, a feedwater vibration alarm indicated excessive vibration. Because of an increase in reactor vessel water level, the unit operator started closing the feedwater regulating valve.
However, before he was able to fully close the valve, the reactor scran=ed from a turbine trip.
The reactor vessel water level was controlled, after the scram, by the feedwater system until the personnel investigating the source.of vibration reported a leak in the feedwater system. The feedwater system y
was then isolated from the reactor vessel.
The leakage was caused by severance of two 3/4-inch feedvater drain lines and the 3/4-inch bypass line around the inlet valve for the high pressure heater. Also, the feedwater regulating valve was found to be in the full open position.
The three lines broke because of high vibration of the feedwater system.
Vibration also caused the loss of the feedback spring on the valve controller.
This caused the feedwater regulating valve to go to the fully open position.
The feedwater low flow drain lines were velded and braced, and the high pressure heater bypass line was replaced from the elbow to the weldolet on the heater side.
. \\
W 07013
'~
_M
-. ~. ;. ;R.
.y _
^
hh
]
1i -1 1
The piston-cylinder air actuator on the feedwater regulating valves was replaced with a diaphrags operator in atte=pt to reduce oscillations of 4
the feedwater regulating valves.
All water from the breaks of piping was processed by the radioactive waste system; there was no excessive exposure to plant personnel nor adverse effect on health and safety of the public as a result of this
,e occurrence.
There have been several cases of excessive vibration of the feedwater system at the Quad-Cities Station. In July 1975, the low flow feedwater g
line severed at a 6-to-4-inch reducer en the downstream side of the low flow regulating valve. In June 1974, the 4-inch low flow feedwater valve failed. In September 1974, the 4-inch feedwater bypass valve had a crack about three inches long in the bottom of the valve, and cracks in the welds of the bypass valve and pipe reducers.
Excessive vibration problems have been under investigation by a co= pan {
cask force, outside consultants, and engineers from Sargent and I. undy.
OTHER CRACKS IN REACTOR FIPING OUAD-CITIES 2 On October 14,1975, Unir 2 of the Quad-Cities Nuclear Power Station was in a cold shutdown condition and ultrasonic testing of the bypass piping around the recirculation pump discharge valves was being performed in accordance with an NRC Bulletin. A crack was found in the heat-affected 3
zone on the 4-inch pipe side of a pipe-to-weldolet weld used to connect to the 28-inch discharge header on the downstream side of the bypass valve.
Although a mode of failure had not been established, the apparent cause was believed to be the same as that which caused cracks in recirculation bypass lines in the past: intergranular stress assisted corrosion.
Early detection of the crack prevented a leak. No radioactivity was released to the environment, so this incident did not present a health hazard to the public or plant personnel.
The existing recirculation pump discharge valve bypass piping of both loops was to be permanently removed and capped.
There have been crack indications in the bypass lines of the Quad-Cities
]
station in the past. On September 16, 1974, there was a crack at a weld on the 3 loop of Unic 2 which was corrected by replacing the weld I
and a short section of pipe.
W 07014
._.m,
..b 5 5 Y h_ Y Y_ _$_ h_..
m.
c m
-s
_'_.m_
m
_..~_._~___.m
__2
.n E-Mi
'"" iw
-2
'b 4-p.
During a second ultrasonic inspection at Unit 2 on Dece=ber 23, 197+,
\\
two cracks were found on bypass loop A and one on bypass loop 3; the A and the 3 loop recirculation pu=p discharge valve bypass piping was replaced.
On January 10, 1975, at Unit 1, a crack was found in the recirculation wl, E
E*
pu=p discharge valve bypass piping on the A loop weldolet running along the 4-inch side of the weld. In addition, there was a crack on the 3 E
loop weldolet running 1/2-inch to 3/4-inch fro = the veld bead.
Both the A and the B loop recirculation pu=p discharge valve bypass piping were replaced.
Si=ilar cracks were found at Dresden, Millstone, Peach Botta= Unit 3, Monticello and the Edwin I. Hatch Nuclear Power Stations.
The pipe that failed in all cases was 304 stainless steel, four-inch dia=eter, with a wall thickness of 0.377 inches.3 DRESDCT-2 lj While a local leak-rate test was being conducted with Unit 2 of the Dresden Nuclear Power Station at 700 MWe on October 7, 1975, the test i
failed, and inspection revealed a throughwall crack on the 18-inch
{
drywell/ torus nitrogen purge line. The crack occurred at an 8-to 18-inch tes connection, and extended 180* around the 8-inch connection on the 18-inch line, crossing the welded intersection, and extending approxi-
=atet.y seven inches along the 8-inch line.
3
]
It was assu=ed the crack occurred during a drywell inerting process when the heating stea= boilers, which vapori:e liquid nitrogen before ad=ission j
to the drywell, failed te=porarily. Because a previous heating stea=
.l boiler alar = had not cleared, the boilers were inoperable for approximately i
15 =inutes before the proble= beca=e evident. During this interval, i
liquid nitrogen passed through the vapori:er directly, ans i=pinge=ent on the steel tee connection caused a rapid and uneven contraction, resulting in the throughwall cracking.
The throughwall crack constituted a breach of pri=ary containment.
However, no abnor=al =akeup of nitrogen was required, so it was suspected that leakage was =ini=al.
Secondary contain=ent was in effect, and the pressure suppression syste= and all e=ergency core ecoling systa=s re=ained operable.
The i==ediate corrective action was an orderly unit shutdown at the rate of 100 MWe au hour.
After extensive =agnetic particle exa=1 nation, a 20-inch section of pipe containing the tee connection was replaced.
The l
new welds were radiographed and a successful local leak-rate test was co=pleted.
W 07C15 e
.,a a.:
..,._....~.. -.
..........~.. ~... -.-..--.....
L h
3 i
l 2
A ther=occuple and strip chart recorder were installed on the vapori:er discharge for rapid isolation of the vapori:er to prevent a similar failure. A specici operating procedure for startup was written, adding precautionary measures to the existing inerting procedures."
(
FAILED FUEL ASSEMBLY During cycle 4 core loading in December, 1975 at Unit No. 1 of the Point Beach Nuclear Station, personnel noted something protruding from the side of a fuel asse=bly as it was being lowered into the fuel assembly upender. The assembly was =oved to the spent fuel pic for anMnation.
Near grid 1, rust marks were on the grid and on rods 12 and 13. The clad of these two rods was worn so that the fuel springs were visible behind slots in the grid. A fuel fragment was observed lodg'ed between rods 13 and 14.
There were gouge = arks in rods 12 and 13 adjacent to the grid tab, and a hole was visible in red 12 at the grid tab.
In the vicinity of grid 2, the cladding of red 12 was separated with no cladding or fuel behind the upper part of the grid. Rod 13 had a cut s:
mark near the upper edge of the clad and a hole on the lef t side. A fuel frag =ent was visible between rods 13 and 14.
r i
Between grids 2 and 3, there were several holes at the contact points of 7
the grid springs on red 12; no fuel was visible. There was a split in red 13 and no fuel was visible except at the top end of the split.
bp At the top of grid 3, there was an open cut in the cladding of red 12 adjacent to the vane; the vane was severely worn. Rod 13 was completely f
separated just above the grid. Another fuel fragment was visible on top f
of the grid.
Below grid 3, holes were visible in rod 11.
There was a large hole in
)
12, with a large fuel fragment sticking out.
A section of cladding from rod 13 was missing. No fuel was visible at the top edge of this section.
Many small fuel fragments were visible behind grid 4 and rods 12 and 13.
5 Between grids 4 and 5, the top 11 inches of red 13 were missing. The end of red 13 was bent to a hori: ental position about two inches long.
The 11-inch section of red 13 was found lying diagonally across another fuel assembly, and was recovered.
Westinghouse Electric Corporation believes the initiating factor for
=
h fuel failure was water impingement when the fuel assembly was in its R
original position in the reactor core.
W 07016 tus p
_r_
=;z.,=.......
..~.... -..-
h.w -cw$
Yv$
?
5 m:x - :~m s' @ - N::mt c MfsQ:g
.4
- cv
- .
- . x :" n m.- ~ M :
6%
4 f5
'Za o
?E
\\
e
-g 19
%7 oM w
1 tith Water i=pingement at the corner or near the corner fuel rods has led to g;
vibration and fretting wear in foreign reactors. This fuel asse=bly 2.;.
occupied a corner position in cycle 2.
It is postulated a small hole e
fres fuel red vibration occurred at this ti=e.
5E The initial escalation to power at the beginning of cycle 3 was at a rate higher than the present operating guidelines, and the cycle 3 position subjected the asse=bly to a measurably higher power rating than g
its cycle 2 position. During initial escalation to power at the beginning g
of cycle 3, a sharp increase in reactor coolant activity was noted "5
between power levels of 40: to 50*.
It was presumed the fuel rods con-K taining holes in the clad beca=e water-logged in the shutdown and burst Fr#
from steam pressure during the power escalation phase.
a r
- w
] y The potential for additional fuel failures is minimized by new guide-pi lines governing the race of power escalation following a cold shutdown.
If The controlled power escalation will allow fuel rods which =ay have i
absorbed water from pinhole leaks to expel the water before a substan-
?I tial increase of stess pressure.
i a i E Acte = pts have been made to locate and remove all loose pellets from adjacent fuel assemblies and the lower core support place, and apparently no loose pellets re=ain in the reactor vessel or en fuel re=aining in the core.
E Based on a safety evaluation, operation of cycle 4 core is not con-g sidered to pose a ha:ard to the health and safety of the public.5 9g HYDRCGEN EI?LOSICNS c).
Th On November 5, 1975, the Cooper Nuclear Station was in steady state J
operation at 60: power, and station personnel were re=oving the =anhole 7
cover to the su=p below the elevated release point to investigate a N
pressurization of the su=p, a hydrogen explosion occurred when the air L
sa=pler was turned on.
An orderly shutdown of the reactor was initiated.
b$
Two persons were burned, and after cleaning the burned areas and finding b
both free of contamination, one patient was retained in the hospital and
{
treated as a burn patient, and the second worker was released.
I w
During investigation of the source of hydrogen gas, it was deter =ined F
that an isolation valve in the off-gas system was in the closed rather E=
than the nor=al open position. This caused the discharge of off-gas
=
from the steam jet air ejector to be routed through the loop seal drain
[
j line to the sump and back to the dilution fans prior to being dischargd at the elevated release point.
-E W
c7c17 1
l
~~
2 % d -2 4 M W % o.wr-Dr~~
't:Wnt'=0WM%
di.zf.S_... - Min._ K3?M. 3_E_MMi-MG. W!MWi-PW7.52"WWWf.6EX97,E_.-TE_M%.. -.. ~,
w.,.
4_
- The valve was found in the closed position although the control room valve position indicating lights and the control switch showed the valve to be open. Personnel, who had been =aking viring changes to this valve for additional off-gas treat =ent equipment, thought they had verified the proper position of the valve by noting the position of the slotted notch at the top of the stem. However, the butterfly valve gate was not aligned parallel to the slot as they had believed.
The explosion occurred when the air sampler was turned on to monitor the gaseous activity release from the su=p.
Hydrogen from the off-gas line exploded as it was drawn through the air monitor; it was ignited by the arcing of the brushes of the sampler motor. An explosive mixture meter had not been used to sample the gases from the su=p; the meter previously had been used when opening the sump, but no indication of hydrogen had been found.
r The sump at the base of the elevated release point was inspected and found to be damaged. The top of the metal lined sump had separated frem the side vall liner. Repair was made and the sump air tested.
Station operation with improper position of the valve resulted in by-passing the absolute filters in the off-gas system. The stack gas activity prior to the explosion was calculated to be approx 1mately 680 uCi/sec; after explosion, the stack gas activity was calculated to be 235 uCi/sec. Therefore, a ground level release of appror N eely 445 uCi/see occurred from the time of accident until reactor shutdown.
Althdugh the ground level release was unplanned and unmonitored for a period of time, there were no indications that abnormal conditions existed outside the site boundary. Therefore, it was considered this occurrence presented no adverse potential consequences from the standpoint of public health and safety.6 Two months later, with the Cooper Nuclear Station at 83% power, an alarm in the control room indicated a low flow condition at the discharge of one of two off-gas dilution fans. The alar::. automatically started the alternate dilution fan. It was then noted the elevated release point (ERP) recorder had indicated a' gradual decrease in flow rate from approximately 2800 cm to 2200 cm over a period of several hours.
There was no increase in flow race after the alternate dilution fan was started. The standby gas treatment (SBGT) system fans were started, but there was not indication of increase ERP flow, and the SEGT flow was lov.
1 W
07G1.8 h
e.M w.m
.\\-
l h
NN5,k N_ e...s.9N% e ~EN$d.s 4 uz,.. a
.bEM., Ngs. kkb.phy@g M.
N N
h N
N
- n.
g g
m g
- %n
U Es
(" -@
?
p%- -
-8 kaw-
--b Inspection of the off-gas building did not.ev.*1 the sources of the problem. However, it was noted the constant air =enitor (CAM) was showing an increase in activity, and the building did not appear to be at its nor=al negative pressure.
After inspecting the ERP and observing no indication of the problem, personnel returned to the off-gas building. Upon reentry, they noted an unusual odor, and the constant air monitor was off scale high. The off-gas building was i==ediately evacuated and an explosion in the building occurred shorrly thereafter. Reactor power was i==ediately reduced and then the reactor was shutdown.
I The 32-foot by 48-foot =etal building was completely destroyed with the exception of some heavy =etal fra=evork. The dilution fan room ceiling and upper valls (constructed of reinforced concrete) were severely damaged.
A partially =elted ice plug was found at the bottom of the ERP several days after the explosion. It was postulated the ice plug had for=ed at the top of the 325-foot elevated release point pipe and reduced the discharge area from 153 sq. in. to approx 1=ately 12 sq. in.
The ERP is uninsulated pipe. The bottom 66 feet is constructed of 30-inch dia=eter pipe with a divider. The pipe then is reduced to an 18-inch dieter until the last 15 inches, where le is further reduced to a 14-inch diameter at point of discharge. All pipe has a 3/8-inch vall thickness.
The ice plug and subsequent reduction in ERP pipe discharge area resulted in back pressure that created the off-gas dilution fan low flow condition.
The starting of the standby gas creat=ent system compounded the problem by creating additional back pressure at the ERP.
The progression of ERP blockage and flow reduction was not easily dis-cernible. The ERP flov monitor, a Pitot-Venturi type sensor device, did not indicate an unusual flow reduction. This instrument had not been reliable and after the explosion and complete ERP flow loss, the recorder still indicated a flow of 2000 cis.
The hydrogen concentration apparently increased until an ignition source within the room caused the explosion. There were several electrical devices including limit switches and solenoids that were not of explosion-proof rating.
El 3
The event presented no adverse potential consequences from the stand--
point of public health and safety. Although there was an unplanned and un=onitored radioactive release with the explosion of the building, there was no indication of abnor=al conditions outside the site boundary.
W O'7019
-e==.um.so mm-
^*
-n.-.-..~.--..--..
- - - - - ~ ~ ~ - --
-~ -
- " ^ ~ '
4p...
m A new off-gas building has been erected. The upper 10 feet of the
).
elevated release point has been heat traced and insulated to preclude the formation of another ice plug. Also, another 10-foot section of the I
ERP around the ERP flow monitor location and the sensor flange was heat f
traced and insulated to i= prove flow monitoring reliability.
The dilution fans were re=oved from the direct path of the process flow j
stream and now take suction from the building room air only. Piping in the off-gas building that can potentially carry an explosive mixture is now designed to stand an explosion. The ERP flow pitot-venturi sensing device has been modified to improve flow monitoring re11 ability.7 EXPLOSION IN STACK FILTER HOUSE A hydrogen explosion occurred at Unic No. 2 of the Brunswick Steam Electric Plant while at 85% power. On January 19, 1976, during an attempted blowdown of the stack monitor sample line an increase in radiation activity of the stack monitors was noted, although off-gas loop seals had been determined to be filled. Two people were dispatched to the filter house and saw an alarm on the local area alars =enitor, and water was on the floor with a heavy'miot overhead. They spent about 30 seconds in the building and, upon exit, discovered they were contaminated.
They were subsequently decent m hated.
About 2-1/2 hours later, the filter pit area was reentered, and the loop seals were refilled with water although it could not be determined if the seals had blown their original water supply.
About 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> later, two off-gas annunicators alarmed. A security guard reported to the control room there had been an explosion in the filter house and thac the house was on fire.
The fire was extinguished with no apparent structural damage to the filter house. However, the hatch cover hinges were bent from the explosion.
Investigation revealed that removal of a filter cell concrete plug during subfreezing temperatures had permitted moisture buildup on the high efficiency particulate filter (HEPA). Additional moisture buildup also occurred because the HEPA filter desister was incroperly positioned and did not allow proper water removal from the process stream.
As a result of the moisture loading, the increased pressure differential I
across the HEPA caused an increase in system backpressure that caused water to be blown from an undetermined number of off-gas loop seals.
The blevn seals allowed both airborne activity and hydrogen gas to the filter house where the hydrogen gas was ignited, presumably by an are from a relay contact.
W 0'7C20
. ~. -.... -...
--..-.. ~.-...
... -.. ~.
h y_.
hs
,g -
,g
.ggig m a. _. 4g
.m h $
geW
- J"cD 309 e
"~
The operating g cup had no warning of the occurrence because excess
$m differential pressure across the filter was not annunciated.
Dg The source and sequence leading to the off-gas explosion has been accu-rately identified, and corrective =easures have been i=ple=ented to preclude occurrence of further off-gas explosions in a si=ilar =anner.8
- ~~
CONTROI, RCD DRIFT During a control rod drive scras ti=e periodic test at Unit 2 of the Brunswick Steam Electric Plant on Septe=ber 25, 1975, a control red
-l scra==ed from position 48.
Following the red scrs=, the operator withdrev the rod to position 06. When the control switches were released, the M
red continued to drift beyond position 06 to position 48.
The rod was 2*T inserted again to position 00 but five ti=es drifted out to 48.
Suspecting a problem with the hydraulic control unit, the insert and
-,M vithdrav riser valves were closed after positioning the red at 00.
3 However, the drive continued to drift to position 48.
The valves vera j
reopened, and the drive was exercised by notching in from position 48 to j
- 46. After two to three notching exercises, the drive successfully 1
latched at position 46.
The drive was then fully inserted and observed W
to stay full in.
Then, the drive was withdrawn to position 48 and left hg there.
While the control red was drif ting, it was noted there was no control rod drive hydrauli: fluid flow and the withdraw and insert lights were not lit. The t1=e to drift from full in to full out was esti=ated to be 90-120 seconds. During subsequent operation of the control red drive, it was observed to double notch out and to insert sluggishly.
Based on the control red perfor=ance, it was concluded that foreign material had entered the collet pisten area which prevented reseating of the collet piston and closure of the collet fingers.
~
5 The sluggish insert =otion was indicative of directional control valve failure.
As a result of proper red latching following the drift, it was concluded s
the interference =aterial at the collet piston had been eliminated l
during the sequences of red exercising. Demonstration of proper collet
- l finger operation under a scras condition was ter=inated after six 3!
successful scram insertions. However, nu=erous double notches were 3 !
encountered during withdrawal, apparently caused by directional control i f valve failure. Also, during testing, withdrav stall flow decreased from f
an original value of 1.5 GPM to 0.2 GPM. The insert speed control a ll Y h 1
I w
o?c21 a
l I
a u i
01,
-y.-. =
=g I
.(
4 i
" 11 -
. l I
'J needle valve was adjusted and the drive water supply, under-piston water to exhaust, and the over-piston water to exhaust filters were repisced; the withdraw exhaust and settle valve was replaced. Four scrs= tests were completed successfully; the control rod successfully notched from 00 to 48 with no indicated deviation from nor=al drive perfor=ance.
Failure of a single control red to position is not considered by the f
licensee to lead to a compromise of reactor safety systems. The failura of a control rod to insert under a scram condition is considered in the j
plant safety analysis as a worst-case situation. The apparent failure of the control rod was detected by perfor=ance of periodic tests which are designed to provide such detection.3 I
I LOSS OF POWER AND SUBSEQUENT REACTOR 3 LOWDOWN On September 13,1975, Unit 1 of the Pilgrim Station was being shut down for replacement of a flange gasket on a pressure reducing valve when, at 17% power while switching the turbine generator off-line, two 345-kV power switchyard breakers malfunctioned. This resulted in loss of power to the emergency busses, and to non-vital equipment including the reactor feedwater pumps. However, offsite power was available to the Core Spray Cooling System (CSCS); but it was not needed.
The malfunction of the breakers caused a reactor scram to occur; the emergency diesel generators started and restored station power to the safety related busses, as designed.
Primary and secondary conta1==ent isolation occurred immediately following the scram and a Reactor Core Injection Cooling (RCIC) system flow cf 400 gym was established in the test mode.
Approximately 10 minutes after primary contabant isolation occurred, one relief valve opened automatically at its design pressure. Another relief valve was opened manually to augmsnt relief of the pressure vessel, and coolant from the RCIC system was injected into the vessel at the full flow rate.
When the reactor pressure decreased to 800 psi, the manually operated relief valve was closed. However, the relief valve operating in the automatic mode failed to reseat with return to proper reactor pressure.
Reactor vessel inventory continued to decrease and caused initiation of
.the Core Standby Cooling System (CSCS) - High Pressure Coolant Inj ection f
(HPCI), Low Pressure Coolant Injection (LPCI), Residual Heat Removal (RER) system, and the Core Spray System - approximately 21 minutes after the loss of station power.
W 07022 p
e--
~ - - -
,,e--
+-
I yypypW-J5QA5G3.vg ;GhW@WW~rci:.w6e:i:.*rdGh-My yE54gyge]
._sgce m m M 9-f-s
- iNam&_95 :W9_M.t, _..M...G$m.92h,_. y
- u. g_., g. : gw.;, 5te - d1
)
w%
s
.f.
s 9IJ',
W
?55 wh MfriAJ e cx; uM The CSCS syste= responded and naintained coolant level until the relief
-' n valve subsequently resented at a pressure of 275 psi.
The =ini=u= reactor vessel level during the transient was the CSCS
]
initiation level (greater than 60 inches above the top of the reactor fuel). During the reactor depressurization, the temperature of the vessel exceeded the =axi=u: cooldown race of 100*F/hr. However, the actual rate of cooldown was less severe than a previously analyzed
(
transient at Pilgri=, and analysis of the vessel te=perature transient J
was not required.
g A pre 14-4 ary inspection of the torus showed no abnor=al conditions.
J!
However, a = ore detailed inspection below the nor=al water level of the e
torus showed the lower restraints of two discharge lines were each
=1ssing their top structural = ember, an 8-inch channel section. The other two relief valve discharge line restraints showed indication of movement at the point of contact between the pipe and the channel. One discharge line had damage to the upper structural supports and to the 12-inch discharge pipe.
~
The relief valve that failed to reseat was disasse'= bled and inspected.
Failure was attributed to pilot valve leakage; flow of stea had eroded the pilot valve asse=bly, and erosion per=itted an increase of pressure on the actuating side of the second stage piston and, thereby, reduced the closing forces on the second stage piston.10-12 o
Je
?.
- t
- U Point of Conenet
N~
Theodore C. Cintula Office of Manage =ent Infor=ation and Progra: Control U.S. Nuclear Regulatory Coc:=1ssion Ts L
t
~
j W
07023
..- ~ u.
. ~ -.
..~. _,
4
<$=&w,-
. - mmmw n
',$'% 7T~iMMViED2WF?iEg2hTE-]-Q3d]([kCNf;QQQ_T,h%-p-.y. {.. _.
a
.W-e,.
UNITED STATES NUCLEAR AEGULATORY COMMIS$1CN
]
F g
W ASHINGTON. O. C.
20555
,,,,,, g 3,, g m?m Q
o rrecs As. s usanass mm eawu o.
c o==.is.o=
PENA 4.TY PQat PRIV ATE UsE, $300 p.
h 4 N [C,N g
STduytyp A TE 5 =' 'ANAGE4 i
GE NE g a,
~
p,,
nox,y,,'3 LLC UTtLITis3 l
Qu]
i
'l00LEin
.y py s,,,S Y L V 4 N I 4 17m i:
W 0"/024
-w-
,... - - - -