05000213/LER-1985-024, Corrected LER 85-024-00:on 850910,auxiliary Feedwater auto-initiation Solenoid Operated Valve,Asco Model NP-8320a-185E Failed to Change State Upon Deenergization. Cause Not Determined.Valve Cycled After External Tapping
| ML20198A829 | |
| Person / Time | |
|---|---|
| Site: | Haddam Neck File:Connecticut Yankee Atomic Power Co icon.png |
| Issue date: | 10/10/1985 |
| From: | Evola J CONNECTICUT YANKEE ATOMIC POWER CO. |
| To: | |
| Shared Package | |
| ML20198A803 | List: |
| References | |
| LER-85-024-01, LER-85-24-1, NUDOCS 8511060193 | |
| Download: ML20198A829 (2) | |
| Event date: | |
|---|---|
| Report date: | |
| 2132085024R00 - NRC Website | |
text
,
9
- U $ NUCLEME LECULATOAY COesh88SSIO88 M.C 1 83 .F.em 336 LPPROvt0 OuS ivo 3150-41 4 LICENSEE EVENT REPORT (LER)
DOC K E T Nua.et m (21 PAGI J P&CILITT N AMI ill Haddam Neck o 15 t o t o l 0 l 211 ]3 1 l0Fl 0l4 Tit u i.,
Failure of Automatic Initiation of Auxiliary Feedwater EVENT OAYE I.I LgmasuusERt.i AEPOmf DATE (71 OYMER F ACiLITIES ihv0LVED 191 uCNTH 04, vEAm vEAR 5'gt Z
, ['[",$ wCNTw Dev vfAR 8 AC'uf v % Awes DOCA E T Nuw.E Rt 51 0 1510!010 1 1 I 0l 9 1 l0 8 5 8l 5 0l 2 l4 0l0 1 l0 1 l0 8l 5 oisto1oioi l l
'"'s atr n is sceuinED PuasuANT TO Tus naOuintuENn Or to C,a g scw. eae e,- oae neue-+es in OPE n Afineo "oo ' ' 1 n .ca.i n .o. ., so n.. nan i n ni..
n n ni.,
g _
..nina n,enii x_ = n,.n n.i
,,oi i , .
n
..ni n.4
.n,n.,
x n..ua, a n.nsua o n .n ,-
a n..nsu..n A, g=, sg..gg mA
. ..nin i o n..n n.i n,.n,n n.,
2o .o8 6.n1 Hul S. 736.n2 nml to 73 den 2nsi LICENSEE CONT ACT FOR TMit LER (12.
%AME TE LEPMONE NUMGER Amt A CODE
.Toseph H. Evola. Engineer 21 013 21 6171 -12151 516 COMPLETE ONE LINE FOR E ACH COuPONENT F AILURE DEscateEO iN Twe8 #EPORY (131 CAv5E COv'ONENT
- " ' ' ' 'L SY ST E v TO PRDS CAUSE Sv5 TEM COvPONENT "TO NPRD5 x B,A F ,S , V, A ,6,0, 9 Y , , , , , , ,
I I I I l l I I l l l l 1 l SUPPLEMENT AL REPORT E mPECTED it. uoNTH DAY
- EAR SL 8915510 %
TES U* v e come'ere EkPECTED Sumut$$ TON CA TEI NO l l l MSTR ACT IL+t to f.N seeces , e eeeeee+ere y e. hee
- s.a# e spece tsee "r'en s.eens 118 Abstract During the first perfonmnce of a preventive maintenance procedure in which each of the auxiliary feedwater auto-initiation solenoid operated valves (50Vs) is cycled, one of the SOVs failed to charge state upon deenergization. ('Ihi s procedure was developed per manufacturer's recannendations as a result of our Novmber 2,1984, event in which two feedwater bypass valves failed to open automatically. 'Ihe cause was detennined to be sticking 50Vs). 'Ihis failure would have prevented one of the four feedwater bypass valves from autanatically opening upon receipt of an Auxiliary Feedwater Systan auto-initiation signal.
It did not and would not have prevented normal valve control using the main control board controller.
After being tapped once, the stuck SOV changed state. It was then satisfactorily cycled several times. .
Using the same preventive maintenance procedure, the six SOVs were satisfactorily cycled on Septmber 11,13,16, and 19 and once per week since then. Periodic cycling wilI continue unti1 either the SOVs are replaced with an upgraded model or the cause of the sticking problan on the existing SOVs is detennined and corrected. l 10g0 $
S gC,,vw -
4RC Pere 3e6A U S NUCLE AR magutaf 0RY COMMISSION
. LICENSEE EVENT REPORT (LER) TEXT CONTINUATION Ae Rovio oMe o siso-em E nPrag5 8/3196 FactuTV hAME Hi DOCRET NUMeta til LER NUMBER tes PAGE13I "aa " 2 0 : '*#,T:
85 O y4 Haddam Neck
, p,g,, 213 _
0 9 _
00 0 ;2 o, r m n - ., we,.,, = = nn Event Description On the afternoon of September 10, 1985, with the plant in Mode 1 at 100 percent power, one of the six solenoid-operated valves (EIIS Component ID "FSV") in the Auxiliary Feedwater System (AFW) (EIIS System ID "BA") automatic initiation scheme failed to operate. This failure was detected during the first performance of the preventive maintenance procedure developed to periodically cycle each of the six SOVs to prevent a sticking problem similar to that described in our LER No.85-005.
This SOV failure would have prevented the number I feedwater bypass valve from opening automatically (normal valve control from the main control board was not affected).
The six SOVs are normally energized. Upon receipt of an auto-initiation signal, they are deenergized and change state to open six air-operated valves--two to start the two auxiliary feedwater pumps (EIIS Component ID "P") and four to open the four feedwater bypass valves (EIIS Component ID "FCV").
Reportability This event is reportable in accordance with:
- 1. 10CFR50. 73(a)(2)(1) (B) . The number 1 feedwater bypass valve would not have actuated to its correct position upon receipt of an auxiliary feedwater auto-initiation signal. This condition is implicitly prohibited by Technical Specification Sections 3.8 and 4.8.
- 2. 10CFR50.73 (a)(2)(v) . Although operator action could have provided auxiliary feedwater to the number I steam generator, the system would have failed to automatically remove residual heat.
Failure Cause The failure was a stuck SOV which, upon deenergization, failed to change state.
The SOV is an ASCO (NPRDS Manufacturer Code A609) model number NP-8320A-185E.
We have not yet determined the exact cause of the malfunction.
Corrective Action Using the same preventive maintenance procedure, the six SOVs have been cycled at least once per week since September 10. Periodic cycling will continue until either the SOVs are replaced with upgraded models (not subject to a sticking problem) or the specific cause of the existing sticking problem is determined and, if possible, corrected.
i ..e ,oaM n l
, in as,
i Nec Form 3e8A U S MUCLE AR REGULATOAV COMustslose
. LICENSEE EVENT REPORT (LER) TEXT CONTINUATION ***aovto oms =o uso-4+
f a*tatl $119S F ACILITT NAME lie DOCKET NUMSER L24 g g g gygg g g gg, pagg gy; n s= "til;;'. k',n Haddam Neck 0l q 9 TEXT ## meg spece e esgiavd ses om,menos nec penn Js5A s1117 o l5 l0 lo lo l2 l1l 3 8l5 - -
0 l0 0 l3 0F 0l4 The licensee considers that periodic cycling will preclude any further sticking problem because:
- 1. In all tests subsequent to the tailure, the six SOVs have operated satisfactorily.
- 2. The SOV manufacturer recommends periodic cycling to prevent sticking.
They have not defined " periodic" or recommended a specific cycling interval.
The latter point is very important, because prior to September 10, the last time these six SOVs were cycled was on May 16, 1985. September 10 was the first performance of a new preventive maintenance (cycling) procedure. This procedure was written af ter our LER number 85-005 (April 1,1985), but was not approved until August 29, 1985, due to a review of plant technical specification requirements of the safety evaluation for the procedure.
References
- 1. LER number 85-005-00, dated April 1, 1985, which describes the similar failure of two of the six SOVs in the auxiliary feedwater auto-initiation system.
- 2. J. F. Opeka letter to J. A. Zwolinski, dated September 20, 1985, which discusses a number of issues regarding the auxiliary feedwater system.
This letter discusses this event as well as a wiring deficiency discovered on the same date. The wiring deficiency is not reportable under 10CFR50.72 or 50.73 because overall system operability was not and is not affected.
Generic Implications This sticking SOV problem could impact the operability of any system using ~~~
the same model valve in a similar application (normally energized, deenergized to " fail" to safety position). No other systems at our plant use these SOVs in such an application.
Safety Assessment Auto-Initiation Logic Auto-initiation of AFW logic is actuated when two-out-of-four steam generator lov level alarms occur (credited in the safety analysis) or both main feed pump breakers open (not credited in safety analysis). In the event that the auto-initiation logic fails to open the steam generator bypass valves, as in the November 2, 1984, and September 10, 1985, occurrences, the control room operator can take manual control of those valves and maintain steam generator level.
Plant procedures require the control room operator to commence feeding the steam generators via the bypass valves to restore levels to 25-50% on the narrow range level, hmC FORM 3564
.b n .
,. ,- ?.. .
I
=ac e a a u s aucuaa asautaToav co==Issio= l
. LICENSEE EVENT REPORT (LER) TEXT CONTINUATION ***=ovio ows No mso-oiod I i
soians amis F ACILaiv maMt its pocatY NUMeta 121 lea NUMeta (Si PaGG 433
'saa "U!,$ O *.O 8 g 5 O4 O4 O l5 l0 ]0 l0 l $ Il 3 q5 _
0 0l9 _,
g op g TEXT M mwe a,ece e *seured, ese eseenasd N#C Form JmLe's> (1M Four-Loop Operation Following a reactor trip from four-loop operation, reactor coolant pumps (RCP) in loops 1 and 3 trip by design. Given a total loss of feedwater, most of the decay heat is removed by the two active steam generators in the loops with the running RCPs (loops 2 and 4). For four-loop operation, the present design basis analysis show that injecting AW into all four steam generators produces acceptable results. The effects of having fewer than four operating steam generators with both AW pumps have not been specifically analyzed. This is consistent with the single failure assumption of loss of one AW pump made in the Facility Description Safety Analysis (FDSA)--other potential single failures were not considered. These other potential single failures are being evaluated 0 as part of the FDSA Chapter 10 reanalysis effort scheduled to be completed by June 1986.
The case of no AW flow to all four steam generators for 10 minutes has been analyzed and found to be acceptable if the operator takes action at that time to initiate AW flow. Since operator action (manually initiating AW flow) could be performed within 10 minutes, the subject event is not considered to be unanalyzed.
Three-Loop Operation For three-loop operation, the present analyses show that, with one AW pump operating, injecting AW into the three operable steam generators produces acceptable results. The effects of having fewer than three operable steam generators with both AW pumps operating have not been specifically analyzed.
This is consistent with the single failure assumption of loss of one AW pump made in the FDSA--other potential single failures were not considered. These other potential single failures are being evaluated as part of the FDSA Chapter 10 reanalysis effort scheduled to be completed by June 1986.
The most severe loss of feedwater event in three-loop operation occurs when the _
isolated loop contains an RCP that would normally be running following reactor trip (i.e., loop 2 or 4). When this is the case, there is forward flow through the loop with the operating RCP and reverse flow through loops 1 and 3. As a result, most of the decay heat load is placed on the single steam generator in the loop with the running RCP. Because of the heat load, the inventory in that steam generator boils down rapidly. The inventories in the steam generators in loops 1 and 3, (containing the idle RCPs) boil down much more slowly. Since the auto-initiation signal for AW utilized two-out-of-three logic, auto-initiation was predicted to occur later than 10 minutes. As a result, the auto-initiation of AW results in more severe consequences (PORVs were predicted to lif t) than a similar case assuming operator action to start AW at 10 minutes. In order to I
meet the licensee's acceptance criterion that the PORVs should not lift for this event, it is necessary for the operator to manually initiate AW prior to 10 minuter. Since operator action (manually initiating AW flow) could be performed within 10 minutes, the subject event is not considered to be unanalyzed.
m '
. ~-. . ___ _. __