ML20024C649

Current Events - Power Reactors,Sept-Oct,1977

ML20024C649
Person / Time
Site:	Crane
Issue date:	12/31/1977
From:	NRC
To:
Shared Package
ML20024C646	List: ML20024C645 ML20024C649
References
TASK-03, TASK-3, TASK-GB GPU-2043, NUDOCS 8307120925
Download: ML20024C649 (11)

Text

-

-L i

e UNITED STATES CURRENTEVENTS NUCLEAR REGUI.ATORY POWER REACTORS munissioN

~

THIS COMPILATION 0F SELECTED EVENTS IS PREPARED TO DISSEMINATE INFORMATION ON OPERATING EXPERIENCE AT NUCLEAR POWER PLANTS IN A TIMELY MANNER AND AS OF A FIXED DATE. THESE EVENTS ARE SELECTED FROM

~ M' PUBLIC INFORMATION SOURCES. NRC HAS OR IS TAKING CONTINUOUS ACTION ON THESE ISSUES AS APPLICABLE. FROM AN INSPECTION AND ENFORCEMENT, LICENSING AND GENERIC REVIEW STANDPOINT.

• 1 1 SEWEER - 310CTOBER 1971

\\

(PUBLISHEDDECEMBER1977)

OPERATOR ERROR l

On January 11. 1977 while the Fort Calhoun Station Unft 1 was operating, water from the Refu' ling Water Storage Tank was pumped i

e l

into the containment through the containment spray header due to an operator error.

During the perfonnance of a quarterly test of the safety injection and containment spray pumps, the operator noticed an increase in the containment sump level approximately ten minutes after the low pressure safety injection pump had been started. Approximately 3300 gallons of water had been pumped to the contatnmenc. About one

..; gig minuta later the ventilation isolation actuation signal was received.

At this time the operator realized he had failed to follow the sur-veillanca procedures and had left the discharge valve of the low head safety injection pump open. He inmediately secured the pump.

The Reactor Coolant System was checked for leakage and containment entry was made approximately one hour later.

Inspection revealed.

that a discharge from the cantainment spray nozzles had occurred.

A few minutas later power reduction was started. A second containment entry was made about an hour later, after containment air samples l

confirmed that a full face mask would provide adequate respiratory protection for the levels of radioactivity in the building. A

?f!

detailed inspection revealed no serious deficiencies and no electrical p.

grounds; the power reduction was terminated at a power level of 835.

Aithough the operator had not followed the' procedure and the discharge r-valve was open, the containment spray header isolation valve (HCV-34,5) 1 W

067S4 l

8307120925 780115 PDR ADOCK 05000289 j

P HOL

(:.,

7 2

and the Jow pressure safety in,iection to containment spray header cross-connect valve (HCV-335) should have prevented the event. The electM c/ pneumatic converter on HCV-345 had failed and both red and green position indication lights were on, indicating the valve was partially open. Prior to the event the auxiliary Building Equipment Operator had taken local control of the valve in an attempt to completely close the valve. After about 1/2 inch of stem travel, the operator removed the valve pin and the valve went back to its previous position as demanded by the valve positioner. Thethirdvalve(HCV-335) in the incident had a leakage problem that had been previously identified but no corrective action had been taken.

i The pneumatic relay on valve HCV-345 was replaced and valve HCV-335 repaired. Valve HCV-344 and HCV-345 are now required to be placed in the test mode pHor to operating the low pressure safety in,iection pump or contain spray pump for testing. This mode along with verifi-cation of an annunciator will ensure that both of thege valves are in the fully closed position prior to pump operation.'

i VALVE MALFUNCTIONS i

1.

primary System Depressurization On September 24, 1977 Davis Besse Nuclear power Station Unit s

No.1 experienced a depressurization when a pressurizar power, relief valve fail'ed in the open position. The Reactor Coolant System (RCS) pressure was reduced from 2255 psig to 875 psig in approximately twenty-one (21) minutes. At the beginning of this event, steam was being bypassed to the condenser and the reactor thermal power was at 263 MW, or 9.5%. Electricity M

was not being generated. The following systmas malfunctioned during the transient:

a.

Steam and Feedwater Rupture Control System (5FRCS).

b.

pressurizar pilot Actuated Relief Yalve.

c.

No. 2 Steam Generator Auxiliary Feed pump Turbine Governor.

The event was initiated at 2134 hours0.0247 days <br />0.593 hours <br />0.00353 weeks <br />8.11987e-4 months <br />, when a spurious " half-trip" occurred in the SFRCS, resulting in closure of the No. 2 Feedwater-Startup Valve and loss of flow to No. 2 Steam Generator. Approxi-e mately one minute later, low level in the No. 2 Steam Generator 3..

caused a full SFRCS trip, closing the Main Steam Isolation Valves O

W C6735 l

~,c,

, ~,

i t

J.

3 l

i (MSIV). The loss of heat sink for the reactor caused the RCS i

tamperature, pressure, and pressurizer level to rise.

The RCS pressure increased to the pilot actuated relief valve' setpoint (2255 psig) and the valve cycled open and closed nine times in ra Meanwhile, pid succession, failing to close on the tenth opening.

the reactor operator observed the pressurizer level l

increase and manually tripped the reactor about one minute after MSIV closure (two minutes into the transient). At this point the RCS pressure was approximately 2000 psig and decreasing while the pressurizer level had reached its maximum initial rise of about 310 inches. The RCS pressure continued to decrease due to the open relief valve and upon reaching 1620 psig approxi-mately three minutes into the transient, actuated Safety Features including high pressure (water) injection and containment isolation.

Approximately five minutes into the transient the rupture disc on the pressurizer quench tank, which was receiving the RCS blowdown. burst. Bursting of the rupture disc was aggravated by the actuation of containment isolation, which had isolated the quench tank cooling systen, resulting in expedited pressuri-zation of the quench tank.

z The RCS continued to blow down through the open pressurizer power relief valve and the quench tank rupture disc opening until primary coolant saturation pressure was reached, about six minutes into the transient. The formation of steam in the RCS caused an insurge of water into the pressurizer. This insnrge and the high pressure water injection then restored pressurizer level to about 310 inches after nine minutes into y

the transient.

Approximately thirteen minutes into the transient, the secondary side of the No. 2 Steam Generator went dry. About fourteen minutes into the transient.. the operators noticed the low level condition and found that the auxiliary feed pump was operating at reduced speed. Manual control of the auxiliary feed pump

. was started and water level restored to the No. 2 Steen Gener.ator.

i l

At approximately 21 minutes into the transient, the operators j

discovered that the pressurizar power relief valve was stuck i

open. Blowdown via this valve was stopped by closing the block

)

valve, thus terminating the reactor vessel depressurization. The

?

RCS pressure recovered to normal and cooldown of the system followed.

i I

i i.

i W

C6786 i

i i-i

.Y..

~

The reason for the spurious " half-trip" of the SFRCS has not yet been determined. An extensive investigation revealed several loose connections at terminal boards, but nothing conclusive.

Investigation into the failure of the pressurizer pilot actuated I

relief valve revealed that a "close" relay was missing from the control circuit. This missing relay would normally provide a

" seal-in" circuit which would hold the valve open until the pressure dropped to 2205 psig. Without the relay the power relief.

valve cycled open and closed each time the pressure of the RCS went.above or below 2255 psig. The rapid cycling of the valve caused a failure of the pilot valve stem, and this failure caused the power relief valve to remain open.

It was detemined that the auxiliary fed pump did not go to full speed because of " binding" in the turbine governor.

The transient was analyzed by the MSSS vendor and determined to be within the design parameters analyzed for a rapid depressurization.

With exception of the above noted malfunctions, the plant functioned as designed and gre was no threat to the health and safety of the general public.2 2.

Feedwater Isolation Valves On two occasions in July, at the Tr'ojan nuclear plant. a hydraulic

.F feedwater isolation valve failed to close upon receipt of a close signal. All other equipment required to operate, functioned nomally.

The first failure, July 6,1g77, had been attributed to an improperly assembled solenoid in the hydraulic actuator.

Investigation of the second failure indicated that bnth everts were due to a lack of sufficient hydraulic pressure.

Failure of the valve to close was caused by the pressure regulator leaking and failing to close down to regulate the pressure. This caused the hydraulic system on the valve to be drained down to a point that the valve would not operate.

Inspection of the regulator revealed that a locking screw on the regulator adjusting knob was loose and would allow the knob to vibrate to any position.

With the regulator improperly set it would not close down to regulate pressure and would allow the hydraulic fluid to drain t

before the hydraulic operator could function. A similar problan was discovered on two other valves, although the maladjustment was not sufficient to prevent these valves from operating.

W C67F7 l

9.

t u

5 All of the regulators were reset and the adjusting knobs were locked in place so that they could not vibrate loose. The isolation va' ves were tested satisfactorily following these adjustments.6 3.

Off-Gas System Valves At the Oyster. Creek nuclear generating station on August 27, 1977, the reactor building ventilation system isolated and the standby gas treatment system (SGTS) automatically initiated.

Investigation revealed that at approximately 1850 hours0.0214 days <br />0.514 hours <br />0.00306 weeks <br />7.03925e-4 months <br /> a station

. employee perfoming housekeeping duties.in the main control room accidently caused the augmented off gas (A0G') mode switch to move from " isolate and bypass" to the " isolate" position. This resulted in the off gas valve and the off gas drain valve going closed.

t and since the A0G was not in service the gas flow was stopped. The l

l isolation of the reactor building ventilation system and initiation of the SETS occurred at 1905. The two off gas valves were opened 1

four minutes later and.the SGT3 was secured. The reactor building ventilation system was returned to normal at 2000 hours0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br />.

l

]

The off gas drain valve did not seat properly and was not leak l

tight. This condition allowed the gaseous radioactivity within the isolated off gas system piping to travel *up through the stack sump in the stack base and fill the air space in the ventilation tunnel. When the radiation level in the reactor building ventilation duct reached a level of 17 mr/hr the monitors located next to this duct initiated the SGTS.

The safety concern associated with this event is the possibility l

l of a submergence dose a person would have received from the radio-active gaseous atmosphere if they were in the tunnel area. The atmosphere in the tunnel area is processed through the radwasta ventilation system, which contains both roughing and absolute filters, prior to exhausting through to the stack which is monitored. The maximum radiation level sensed in the tunnel was 26 mr/hr.

i l

No personnel exposures or releases to the environment resulted from this event. The licensee is investigating the feasibility of installing an alarm to alert operations personnel to the closure of the off gas valve when the A0G is out-of-service.5 W.C6738 l

J

%.-we_m..-.---_-.,,.,..,,.--_.s.

.,--.,,._,-,w.,,-.,

~

SMALL PIPE BREAX ANALYSIS On June 9,1977, an orderly shutdown of the Yankee Nuclear power Station (Yankee Rowe), a pressurized water reactor, was initiated by the licensee because of an error discovered in the Emergency Core CoolingSystem(ECCS)perforsanceanalysis.

Yankee Atomic Electric Company YAEC),thelicensee,notifiedthe Nuclear inapar,RegulatoryConsrission( C) that an error had been discovered ticular small break loss of coolant accident (LOCA which pemitted reactor operation with Core XII in a manner) lessanalysis, conservative than assumed in the original analysis.

While performing a review of the analyzed small break accidents for the Core XIII reload, the YAEC Safety Analysis Group determined on June 7,1977 that an incorrect fluid flow resistance calculation was made in the safety injection line break analysis. The fluid flow characteristics study had taken credit for the 2-1/4 inch safety injection line thermal sleeve to retard spillage from the accumulator -

a tank which supplies horated water to the reactor core in the event of a reactor coolant systaa pipe break. The flow resistance of the sleeve should not have been included in the flow calculation, as a new worst case pipe break was identified in a 4-inch diameter line section.

The recomputed decreased flow resistance allowed increased accumulator flow to be calculated for the break and decreased the ECCS supply pressure to less than had been assum,ed, thus decreasing the core

~

reflood capability of the ECCS.

This corrected flow resistance assumption was used for the accident analysis of the present core, Core XII, which was operating at 79% of rated power in a coastdown program prior to the June 9,1977 shutdown. Operation of the reactor with Core XII consnenced in December 1975.

Upon discovering the error, the licensee reduced power level to 300 megawatts thermal (505 rated power), which was believed to conservatively acconsodate the analysis error. During subsequent analysis, however, the licensee was unable to assure himself that the 10 CFR 50.46 limits on peak fuel cladding temperature could be maintained for the postulated small break.

Therefore, the faciltty was shutdown pending resolution

~

of this matter and to proceed with the core XIII refueling outage which had been previously scheduled to comunence on July 2,1977.

e W

$6739 e

i-The licensee subsequently performed an approximate best estimate analysis of the postulated worst case small pipe break, which included assumptions based on actual facility equipment availability during Core XII operation. The results of this analysis indicated that the calculated peak fuel cladding temperature was well below 10 CFR 50.46 limits. The more conservative 10 CFR 50 Appendix K reanalysis of Core XII operation, however, indicated that 10 CFR 50.46 limits might have been exceeded in the event that the safety injection pipe break had actually occurred.

Prior to returning the Core XIII the licensee: plant to operation after refueling of

1) perfomed flow measurements tests 1

to datamine the actual flow resistance through the safety injection piping: 2) changed the flow resistance in the safety injection lines, by an ECCS modifications and 3) analyzed appropriate pipe break accidents in accordance with 10 CFR 50 Appendix K criteria. The i

changes and results of tests and analysis were submitted to the NRC and were approved prior to restart of the plant after the refueling.6-7 l

DIESEL GENERATOR TRIP 3

g During a loss-of-power test on August 26,1977,* the E-4 diesel of the peach Bottom Atomic Power Station Unit 2 started properly as a i

result of the undervoltage condition, but tripped imediately. This trip was caused by the overspeed mechanism. The circuitry was reset, an adjustment was made to the mechanical governor to limit the diesel speed during a start and the unit was started successfully. Because-the exact cause of the trip was not fimly established, surveillance

?

testing of the diesel was increased from once a week to once per shift.

.i t

During one of these tests, on August 27, 1977.the diesel tripped again. Another adjustment was made to the mechanical governor, the load capability was checked and several successful starts were perfomed.

Once per shift surveillance..was continued.

On August 29, 1977, the diesel again tripped en overspeed and was declared inoperable. The diesel was then operated in excess of synchronous speed in order to detemine the exact speed at which the overspeed mechanism would function. This test determined that the diesel would trip at 940 rpm instead of the desired setpoint of 990 f

1 rps. The trip mechanism was. adjusted to 985 rpm by a manufacturer's

{

representative and diesel was started twice, successfully.

y Investigation into the cause of the change in the trip setting 7

detemined that during the diesel maintenance in June 1977 a camshaft J

i

_t r

e, g

W C68CO

?

t l

J

~~

was replaced.

In order to replace this camshaft the overspeed mechanism had been removed. When the overspeed mechanism was replaced, some necessary shims were not installed. Although this was the only diesel requiring this maintenance during the annual check, the other diesels were operated up to a speed of 945 rpm to verify proper operation.

None of these diesels tripped on overspeed.

Analysis of this event revealed that a deficiency exists in the maintenance procedure associated with the diesel yearly inspection and the post-maintenance testing procedgre. These procedures will be revised to correct the deficiencies.*

ELECTRICAL FAULT On July 13, 1977 while the personnel at James A. Fitzpatrick nuclear power plant were conducting refueling operations a short in a cable caused 600 volts AC to be introduced into a 115 volt circuit. The

'600. volt AC supply for the refueling bridge and the 115 volt AC circuit for refueling interlocks are both located in the same cable.

Flexing of the cable with bridge motion over the core

• caused the cable to short internally. The introduction of the 600 volts into the 115 volt circuit caused nineteen relays in the rod manual control systen to burn out. All of the refueling operations were halted until the interlocks were repaired. The rod worth minimizar and rod sequence i

. control systems were also checked for damage.

I r

A modification is being prepared that will remove the 115 volt AC interlock circuit from the ca This will prevent recurrence.gle carrying the 600 volt AC supply.

ii PIPE CRACK

{

The Brunswick Steam Electric Plant Unit 2 was in hot shutdown and k(

preparations were underway to startup the unit when the Shift Foreman noticed a small leak of the recirculation loop suction piping. This

'l discovery was made during the closecut inspection of the drywell.

li Investigation revealed the leak was from a crack in the socket weld

?

D on a three-quarter inch test connection 900 elbow that was nonisolable -

li and the plant was placed in the cold shutdown condition. The cracked 7

pipe was cut out of the system and the connection was capped. Similar connections on both Units 1 and 2 were dye-penetrant c:wcked with no other indications of cracks.

l W

C6801

. - -. - +

..m..-

. _. ~....,

___,.__,.._,_..__.,__.__,.__~_.__._-.,,,.,,,___,._._.,,-_m_m._.,

,y 9-Further investigation revealed that the crack was contained in the weld metal and intergranular stress corrosion in the heat affected zone of the base metal was ruled out.

the internal and external diameters of this section of pipe reve no other cracks.

weld joints showed that a proper gap was present betwee I

and the pipe end.

i Based on a stress analysis and the observed condition of permanent j

defomation of the failed area, along with the locat in the weld fillet area.

result of workmen (during constructionIt is believed that this deformation wa i

This use of the pipe for this purpose p)lus vibrational stress res using the pipe as a step.

l in the failure.

1 A visual inspection of similar piping on the other loop of Unit 2 and both loops of Unit i revealed no deformation as was observed on the failed pipe.

It was also noted that the location of the three remaining pipes is such that they are not.likely to be used as a step 3

or support because of physical interferences.

e These three pipes will be supported to protect them from experiencing excess:gIpr loading and vibration, or will be removed and capped.'

i

\\

Point of

Contact:

i Joseph I. McMillen Office of Management Information and Program Control

(

U.S. Nuclear Regulatory Comission I

S t

b s

~

W C6802 l

l

=

REFERENCES 1.

LER 77-2, Docket No. 50-285, January 31, 1977.

2.

LER 77-16 Docket No. 50-346, 0ctober 7, 1977.'

3.

Supplement to LEK 77-16," Docket No. 50-346, November 14,1977.

4.

LER 77-23 Docket No. 50-344, July 29, 1977.

5.

LER 77-21. Docket No. 50-219, September 23, 1977.

6.

LER 77-30. Docket No. 50-29, August 3, 1977.

}

7.

Summary of June 17 Meeting, NRC-YAEC, June 22,1977.

8.

LER 77-37A, Docket No. 50-277, September 9,1977.

9.

LER 77-43, Docket No. 50-333 August 11,1977.

{

10.

LER 77-7, Docket No. 50-324 February 28, 1977.

11.

Supplement to LER 77-7, Docket No. 50-324, September 30,1977.

e!

}

t 4

t 8

W ~ G6803 i

e

• ..r

, umano erATES

• iaACLEAR EEGULATORT -

p

• ' wammmmm. s. c asses ess

...u. ~%,

mty pga pesvaTg uns, same gg L

j i

l 11C57CC6805 1 N4 FNERAL FUS UTTLITIES SER CO HRGE WILE ISLAhD FO BCX 4E0 PICCLETChN PA 17057 8

e O

b O

e l

6 e

4 e

e W

C6804 e

e o

me S

-..----w,

. __. _ _. _.,