05000331/LER-1984-001, :on 840107,feedwater Recirculation Valve B, Failed Open,Decreasing Feedwater Flow to Vessel.Caused by Error in Performing Surveillance Test Procedure.Both Valves Repaired.Test Procedure Revised

:on 840107,feedwater Recirculation Valve B, Failed Open,Decreasing Feedwater Flow to Vessel.Caused by Error in Performing Surveillance Test Procedure.Both Valves Repaired.Test Procedure Revised

ML20080J598
Person / Time
Site:	Duane Arnold
Issue date:	02/06/1984
From:	Harris M, Mineck D IES UTILITIES INC., (FORMERLY IOWA ELECTRIC LIGHT
To:	NRC OFFICE OF ADMINISTRATION (ADM)
References
DAEC-84-64, LER-84-001, LER-84-1, NUDOCS 8402150127
Download: ML20080J598 (5)

LER-1984-001, on 840107,feedwater Recirculation Valve B, Failed Open,Decreasing Feedwater Flow to Vessel.Caused by Error in Performing Surveillance Test Procedure.Both Valves Repaired.Test Procedure Revised

Event date:
Report date:
3311984001R00 - NRC Website

text

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _. _ _ _. _ _. _ _. _ _ _ _ _ _ _ _ _. _

\\

NRC Form see U S. NUCLEAR 6 EGULATORY COamstN40ss 0 83' APPROYED OMS NO.3190410e LICENSEE EVENT REPORT (LER)

FACILITY NAMS til DOCKET NuesSER 12)

PAGE43 Duane Arnold Energy Center o Is l o g o g o l3 i 3l l i i lorl 0 l4 TITLE 444 Scram due to F. W. reduction and SRV actuation EVENT DATE (5)

LER NUMBER (S)

REPORT DATE 171 OTMER F ACILITIES INVOLVED (0, MONTMl DAY YEAR YEAR

,0 MONTH DAY YEAR F ACILsTV NAMES DOCKET NUMSERIS)

NONE oisioiogo, i ;

Ol1 0l 7 8 4 8 l4 0l0 l 1 0l0 0l2 0l6 8 l4 NONE o iss oioici i

~

i THIS REPORT 88 $USMITTED PURSUANT TC THE REQUIREMENTS OF 10 CFR l: (CAece one er more er the felsemes) titi OPE R AttesG j

N 20 40am n.Oswi eeni.iaH=>

n ?1m R

30.405teH1HG 30 3steH1) 90.73teH2Het 73.71(al nei 1 1 01 0 2"awHum a awim

=4y7;;y;gy;,,

anwH2H*

20 406teH1HWI to.731 alt 2Hil 90.724eH2HetelHAl 356AJ 20.e05teH1Hl.)

60.73hM2Hal go.73teH2Hvist(3) 20 40S(aH1Hvl 60.7NH2HW) 90.73(aH2Hal LICENSEE CONTACT FOR THIS LER 112)

NAME TELEPMONE NUMSER ARE A COOE Michael S. Harris, Technical Support Engineer 85l i73 O6 31,9 i i i i i COMPLETE ONE LINE FOR EACM COMPONENT FAILURE DESCRISED IN TMit aEPORT 113)

ORT

([

CAUSE

SYSTEM COMPONE NT

[

"E'n gp CAlsit SY STEM COMPONENT I

y y

g E

PR X

SlJ iTiBiG X 91 99 N

i i i i i i i i

A sib l l R IV D12l4 0 N

l l l l l l SUPPLEMENT AL REPORT EXPECTED (14 MONTH DAY YEAR Ytt 199 res, tem,rere EXPECTED SusMtSSION DA TE)

NO l

l l

A=T R ACT a,,,

,.oo s

, e.,..

. A,

,,,.w

., A n.i While in run mode at approximately 1001, power, the "B" Feedwater Recirculatica valve f ailed open, thereby decreasing feedwater flow to the vessel. Due to an unrelated error in the performance of an ongoing surveillance test procedure, a main steam relief valve was inadvertently opened 43 seconds later. Vessel level decreased to the low level setpoint (170" above top of active fuel) and initiated a scram as designed. The relief valve was reseated approximately 75 seconds later.

Both RFP's tripped on low suction pressure. The vessel level continued to fall until the HPCI/RCIC initiation setpoint was reached (119.5" above top of active fuel). Level was restored to 194" and HPCI/RCIC were secured within 4 minutes.

Prior to startup, both feedwater recirc valves were repaired or inspected and the surveillance test procedure was revised to reduce the margin for personnel error in the future.

Ohhh PDR

""S i I

esRC perm 3e4A U S NUCLE A21 KEEULATORY COMMISSION LICENSEE EVENT REPORT (LER) TEXT CONTINUATION AaPaoveo ove =o sino-moa EaPiRES S/31115 f

F ACIUTV NAA88 (1D Docket m#Unselm (21 (gn gunseg A (83 PAGE 131 Duane Arnold Energy Center m:;;pp

=,3 o ls jo lo lo l3 ]3 l1 8l 4 0 j 0l 1 0l 0 0l2 0F 0 l4 rixi n -.

e uc =.anw. m At 08:08:26 (T=0), while the reactor was at 99.8% power, the "B" Feedwater Recirculation Valve (SJ-FCV-1611) f ailed open due to a broken air supply line fitting. As a result, a portion of "B" reactor feedwater flow was returned to the condenser and the vessel level imediately began to drop.

While operators were diagnosing and compensating for the decrease in vessel level, the "B" Reactor Feedwater Pump tripped on low suction pressure and 1 of 2 safety / relief valves (SB-RV-4401) utilizing 10-10 set logic cycled open at T +43 ;econds.

After the relief valve opened, a momentary upswing in vessel level was experienced, followed by an rapid decrease in level to the low level trip setpoint (170") and an autc.natic scram initiation at T +44 seconds. At T +57 seconds, the "B" feedwater pump breaker was reset. At T +66 seconds, both feedwater pump, tripped on low suction pressure (due to condansate pumps being unable to match RFP flow with the high vessel demand and return flow to the condenser).

Ouring this time, the vessel level was continuing to fall until the 10-10 level HPCI/RCIC initiation setpoint of 119.5" was reached at T +73 seconds. At this point, HPCI and RCIC initiated per design and the level began to rise. As the vessel level fell to the 10-10 level setpoint momentarily, HPCI initiated but did not inject at any time in the event. HPCI control logic requires low-low level signal to remain present fnr the duration of the injection valve initiating time (aroroximately 15 seconds) to have HPCI actually inject. At T +112 seconds, the "B" reactor feedwater pump breaker was again closed to provide additional makeup and remaired closed with no further trips. The open safety / relief valve was reseated at T +118 seconds after operator action and remained closed. As RCIC was injecting and the "B" feedwater pump was functionino, the level quickly recovered to normal operating level (Maximum level reached uss 204.4" at T +4 min. 30 sec.).

After the level had recovercd, RCIC was secured at T +3 min. 50 sec. and HPCI was secured at T +4 min.

30 sec..

At this point, the vessel level and pressure were stable at 194" and 810 psig respectively.

All systems necessary for remaining in a safe shutdown condition were operable, and no further system perturbations were experienced.

In retrospect, the following causes and determinations can be made as to the exact component failures and related events leading up to, and af ter, the scram.

The feedwater recirculation valve (SJ-FCV-1611) is an air actuated, f ail open, control valve, Contrei Components model no. PDA964-96BW.

Investigation revealed that a flexible air supply line fitting broke due to fatigue. When the air supply was lost, the valve cycled open.

Prior to startup, the broken airline on the 'B' recirc valve was repaired. As an additional precaution, the identical airline on l

the ' A' recirc valve (SJ-FCV-1569) was inspected and found satisfactory. As a long 8

term corrective action, a maintenance action request has been generated to replace the fittings subject to fatigue with higher strength fittings.

Main steam safety / relief valve (SB-RV-4401) is one of two non-ADS safety relief valves that are utilized in the 10-10 set logic.

Following an event that has opened one of the six (6) safety / relief valves, the 10-10 set logic establishes a lower opening pressure (lifting pressure: 1020/1025 psig, reseating pressure:

900/905 psig) respectively.

Investigation revealed that the lifting of the SRV was totally unrelated to the failure of the feedwater recirc valve.

Prior to the recirc valve f ailure, a 10-10 set instrument functional test had comenced. While

g.o...

i

eseC form 306A U S NUCLt131 KELULATORY COMM888&ON LICENSEE EVENT REPORT (LER) TEXT CONTINUATION APmoveo ove *o 3 iso-oio.

E RPimf 8 8/3115 FACsLITV esAMeIH DOCA t T NUedt R (2)

LER NUMGER lt' PAGE13) vtan 580WENT'ab

• 8W***

Duane Arnold Energy Center 010 01 3 0F 014 o 1510 lo lo 131311 84

-- 01011 rm,

--c, w.nm performing the routine surveillance procedure, the instrument technician erroneously placed an ohnneter across the hi pressure switch contacts too early in the procedure. As a result, the SRV actuation solenoid relay was falsely energized with a hi pressure test signal, which, in conjunction with the procedure step to push the test button, caused the valve to cycle open (approx 40 seconds af ter the recirc valve f ailure). Approximately 75 seconds after the SRV had opened, the operators reset the 10-10 SRV logic, thereby disarming the solenoid and cycling the valve closed.

It was later theorized that due to the relatively low internal impedance of the ohmmeter, the meter acted as a jumper and bypassed the contacts.

It is noteworthy that prompt identification of the cause of the SRV actuation was greatly aided by the technician volunteering that he had indeed committed an error.

This theory was confirmed prior to startup after running a special test procedure.

This procedure recreated the applicable sequence of test steps leading up to the opening of the SRV and revealed that the connection of an ohnneter across the contacts in question in conjunction with pressing the test button, would in fact energize the SRV solenoid.

To preclude recurrence of similar situations, all instrument technicians ware assembled and shown during a step by step reconstruction, the exact cause of the event. The importance of proc? dural complianre without deviating from the prescribed testing sequence was impressed upon them.

To reduce the margin for personnel error, the surveillance test procedure has been revised to utilize a volt meter in lieu of an ohmmeter.

This revision will eliminate the possibility of the meter acting as a jumper in a circuit under test, regardless of when it is inserted. As a long term corrective action, selected surveillance test procedures are continuing to be revised to simplify and clarify the testing sequence, thereby reducing the possibility of personnel error.

In addition to this, a joint management /uninq task force is being forme 6 to determine and implement methods of maintaining a heightened awareness of the importance of minimizing personnel errors.

Prior to startup, all systems initiated during the scram and subsequent evolutions were reviewed to determine whether they had functioned per design and technical specifications. At 170" reactor water level, the reactor scramed, valve groups 2, 3, 4 and 5 isolated, and both trains of the standby gas treatment system initiated per design. At 119.5" reactor water level valve group 8 isolated, LPCI logic tripped both recirc pumps, loop select logic initiated and performed its' function, and HPCI/RCIC initiated per design.

During and after the transient, the operators observed a normal SRV tail pipe temperature response. As an additional precaution, a walkdown was performed on the tailpipe of the opened SRV. The piping and restraints were inspected and found satisfactory.

geo.. m.

j

senC Pere 3e64 U S NUCLEAR KEEULATOTY COMMIShiON UCENSEE EVENT REPORT (LER) TEXT CONTINUATION

• • eaovio ous =o mo+o.

E mPimES S/3155 PactLITV NAAlte (1)

DOCKli NUMSER (21 LER GIUMSER IS PAGE131 Duane Arnold Energy Center "edl:;r

'JJ,q i.a 0l0 01 4 0F 014 010l 1 o1510l0l013l3l1 814 rarra -. - - m,ami.mn It was also observed that only one " side" of the reactor 10-10 level switches initiated HPCI/RCIC. This was later attributed to the location of the level switches' sensing lines in relation to the opened SRV.

It appears that when the SRV opened, the resultant pressure profile across the top of the vessel skewed the actual level such that the low level was only seen by the switches whose sensing lines were attached to the vessel opposite the steam line on which the relief valve lifted.

It appears that vessel level just reached the 119.5" setpoint and did not decrease appreciably below that level. To ensure that the low level indicating switches were functioning properly, a surveillance test procedure was performed on both sides (4 channels) of the HPCI/RCIC initiation level switches.

Prior to diartup, both sides were found to be in calibration and functioning per design.

With the exception of the unrelated failure of the recirc valve and safety / relief valve, all components and systems functioned per design throughout the event. At no time was the safety of the plant or public endangered. Apart from the corrective actions outlined above, no further action is warranted or planned at this time.

g 4 0..

t J

Iowa Electric Light and Ibwer Company February 6, 1984 DAEC 64 U. S. Nuclear Regulatory Commission Document Control Desk Washington, D. C.

20555 Subject: Duana Arnold Energy Center Docket No. 50-331 Op. License DPR-49 Licensee Event Report No.84-001 Gentlemen:

In accordance with 10 CFR 50.73 please find attached a copy of the subject Licensee Event Report.

Very truly yours, Daniel L. Mineck Plant Superintendent - Nuclear Duane Arnold Energy Center DLM/MSH/pv attachment cc: Mr. James G. Keppler Regional Administrator Region III U. S. Nuclear Regulatory Commission 799 Roosevelt Road Glen Ellyn, IL 60137 NRC Resident Inspector - DAEC File A-118a

.fEAA r

I;(

Uuane Arnold Energy Center P.O. Bo.x ::51 Cedarliapids, iowa 52406 319'852. 612 h