ML20006B275

LER 89-011-01:on 890528,manual Reactor Trip Initiated Following Failure of Pressurizer Safety Valve When Valve Became Unseated,Causing Rapid Depressurization of Rcs.Temp Detectors Installed on Safety valve.W/900123 Ltr

ML20006B275
Person / Time
Site:	Summer
Issue date:	06/27/1989
From:	Bradham O, Higgins W SOUTH CAROLINA ELECTRIC & GAS CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
LER-89-011, LER-89-11, NUDOCS 9002010227
Download: ML20006B275 (6)

Text

. .

10CFR50.73 .

Bo th Carolina Electr'ic & G:s C::mpany ' tile S.iredt ern .

Jenkinsville, SC 29065 Nuclear Operatior.s :

• ~

(803) 3454 040 1

SCE&G_

mn-January 23, 1990 1

l J

Document Control Desk 4

U. S. Nuclear Regulatory Commission Washington,-DC 20555:

.i

Subject:

Virgil C. Summer Nuclear. Station Docket No. 50/395 -l Operating License No. NPF-12 i LER 89-011.. Revision 1 I

Gentlemen:

Attached is Revision 1 to Licensee Event Report No.89-011 for the Virgil C.

4 Summer _ Nuclear Station. Specifically, the revision identifies the cause of.

the May 28, 1989, pressurizer safety valve'misoperation and supplements the j

" Additional Corrective Action " section. Additionally, the pressure at which the manual reactor trip was initiated has been revised to 2000 psig.- This i number is in better agreement with engineering analysis and-the Technical l Support Center event printout than the 1900 psig previously reported. This .

report is submitted pursuant to the requirements of 10CFR50.73(a)(2)(iv).- '

Should there be any questions, please call us at your convenience'.

Very truly yours,

0. S. Bradham i

EWR/0SB: led Attachment 1

c: D. A. Nauman/0. W.'Dixon, Jr./T. C. Nichols, Jr.

E. C. Roberts  !

R. V. Tanner J. C. Snelson i S. D. Ebneter R. L. Prevatte J. J. Hayes, Jr. l J. B. Knotts, Jr.  ;

General Managers INP0 Records Center i C. A. Price ANI Library  !

G. J. Taylor Marsh & McLennan J. R. Proper 4

.NSRC  !

R. B. Clary NPCF  ;

F. H. Zander RTS (ON0 0 0047)

T. L. Matlosz Files (818.05&818.07)

K. E. Nodland i, 9002010227 890627 /

PDR .ADOCK 05000395 /

?! PDC -

1

e Es4 Pere 388 U.S. NUCLEAM G.51ULMOmf C0saststous a-

• Apr'moyt0 OMS NO 3190-0104 expiats smes

- LICENSEE EVENT REPORT (LER)  !

P ACILITV Naast HI OOCKET esUesta (3) Pains (a Virail C. Summer Nuclear Station 0 Is lo lu lo lq i gi g 1l orb lg flTLt les -

Manual Roac"nr Trin nue to Press 1rizer Safety Valvp Fa il e r rn ,

SVtNT Daf t ISI Lin NU*eBER It) AGP0sti oaf t 171 OTHER P ACILIT188 INVOLVED 101

"${,* ' DoCatT Nuwstmisi MONTH DAY vlam vtAR $ MONTH DAV YEAn ' Acitivv hauts o15lojojoj i l q5 2lE 89 8l 9 -

0 l 1l 1 0l1 0l6 2l7 8 l9 oisioiotoi i i OPE A Aflese TMis alPORY es 90eadiTTED PUNSUANT TO THE Rt0Utmeesta,75 OF 10 CPR 9. (Casse sae er more e'sae do,towraP1111)

= cot = 1 m m. . . . , x .mi.H,,, 13,1, ,

POWER 30 400(eH1HH so.38(eH11 98.73teH3Het 73.711el li 0:0 O' m.m Hum mai.im umleH. ofwta is  %. A. ,

m Huo.o a n an.

wmmwHAi - *af, "" * '" ""' '*"

30 4084eH1HN) 30.734sH3 Hun go.734elGHveuHS1 30 454eH1Hel 30.734sH3Hein to.734eH3Hul LDCE8sstt CONTACT Poa THas LEA Hal NAME TELEPHONE NUuelm AREA COQt W. R. Higgins Supervisor, Regulatory Compliance p in t1 11ain I tain Iat?

COnePLETE ONE LING FOR SACM CoasPONENT P AILunt DESCRIGto IN TMtB REPORT H31 sysitM "# # 0' L MA m A 8 2 CAU$E COMPONENT "f0 PM S CAust sysT8M COMPONENT C. l +

X ABt i i RVi Ci 7 i l:0 Y R i i i i i i i I I I I i I I i I i I I I l SUPPLIMENTAL REPORT EXPtCTED 114i MONTH Day v t A81

" ~

q vn <,1, . aon.,e amereo suewss,oN os re:

T} No A T . A C T <-,, , um . . . ,..~.,, ,.- ,,. .e ,-. , , n.i l l l_

At 0252 hours0.00292 days <br />0.07 hours <br />4.166667e-4 weeks <br />9.5886e-5 months <br />, May 28, 1989, a manual Reactor Trip was initiated following the l failure of a Pressurizer Safety Valve. The valve (XVS-8010-C) became-unseated i causing a rapid depressurization of the Reactor Coolant System. The manual Reactor i

Trip was initiated at approximately 2000 psig and the safety valve reseated prior to reaching the Safety Injection setpoint of 1850 psig. The Reactor Coolant System

! pressure recovered and was stabilized at approximately 2000 psig.

l During the transient, a condenser steam dump valve failed to close and an operator had to fail the air to the valve for closure.

Pressurizer Safety Valves (XVS-8010-B and C) were replaced and the reactor was restarted at 0049 hours5.671296e-4 days <br />0.0136 hours <br />8.101852e-5 weeks <br />1.86445e-5 months <br />, June 11, 1989. Note: Valve 8010 B was replaced due to minor leakage past the seat.

The Licensee has determined that the reason for the misoperation of XVS-8010-C is that the expected margin between normal operating pressure and the pressurizer safety valve relief setpoint pressure was reduced to zero. It was also determined that the most prevalent factor in this margin reduction was a loop seal discharge.

This discharge resulted in a reduced valve setpoint because of the steam medium imposed on the valve. As such, the Licensee is focusing the corrective action plan on the elimination of loop seal capability for the pressurizer safety valves.

NOTE: A subsequent report (LER 89-015 dated September 20,1989) documents a similar i event involving pressurizer safety valve XVS-8010-A. The Licensee has determined l that a loss of loop seal was the cause for the misoperation of XVS-8010-A.  ;

1 ih*a >'"* "*

/h NRC form 306A U S. NUCLE 17. RETUL ATORY COMMIS$40N

"' LICENSEE EVENT REPORT (LER) TEXT CONTINUATION ' urnoveo oue so mo-otm r

i l. EXPIRES 8/31/85 - ,

P ACILITY NAML.41)

• DOCKET NUMBER Q) .LER NUMSER {$) PAGE (31 .

TIA" .

• euverm <

Nuvata s Virgil- C. Summer Nuclear Station o ls jo-lo lo l3 p l5 8l 9 -

0l11 1 -

0l 1 0'l 2 oF [) l5 .

[

nx, n- .e me.s. mc m mx,nm

.b i

Plant Identification i

Westinghouse - Pressurized Water Reactor i E_auipment Identification J Pressurizer Safety Valve - EIIS - None U

.~ Identification of Event A Pressurizer Safety Valve became unseated causing a rapid decrease in Reactor Coolant  :

System (RCS) pressure. A manual Reactor Trip was initiated at approximately 2000 psig. [.

Event Date ,

May 28, 1989 .

Discovery Date May 28. 1989 -'

' a~

Report Date June 27, 1989 This report was initiated by Off-Normal Occurrence Report 89-47.

Condition Prior to Event .

Mode 1 - 100 percent .:

1 Description of Event .

I a At 0252 hours0.00292 days <br />0.07 hours <br />4.166667e-4 weeks <br />9.5886e-5 months <br />, May 28, 1989, a manual Reactor Trip was initiated following the failure-of a Pressurizer Safety Valve. The valve (XVS-8010-C) became unseated causing a. rapid depressurization of the Reactor Coolant System (RCS). The. manual Reactor Trip was initiated at approximately 2000 psig and the safety valve reseated prior to reaching the l Safety Injection setpoint of 1850 psig. The RCS recovered and was stabilized at-approximately 2000 psig.

1 Indication of a possible safety valve problem had bleh noted at 0024 and 0034-hours upon intermittent actuation of the " Pressurizer Safety Valve Open" annunciator on the Main Control Board and changes in the Safety Valve tailpipe temperature. Instrument'and Controls (I&C) personnel were requested to install multichannel recorders'on the Pressurizer Safety Acoustic Monitor to aid the operators in determining the extent of the-

' problem. During the installation of the recorder, two additional events occurred at 0133 and 0206 hours0.00238 days <br />0.0572 hours <br />3.406085e-4 weeks <br />7.8383e-5 months <br />. At 0252 hours0.00292 days <br />0.07 hours <br />4.166667e-4 weeks <br />9.5886e-5 months <br />, the fifth event occurred which resulted in the depressurization of the RCS and the initiation of a manual Reactor Trip. The previous- L g,oa u..

= - _- . __ - . _ _ _ . .

y l =ac w a A . us. nucten uovurny commission  :

LICENSEE EVENT REPORT (LER) TEXT CONTINUATIO'N' urnoveo ous No. sno-om -:

l .' IXPIRES' 8/31/86 f ACitlTV NAME (n ' DOCKET NUMSER (2) .ggg gygggg (g) . . pAgg (3) . .j naa  :

" =  :

=#

Virgil C. Summer Nuclear Station o l5 lo Io lo l3 l 9l 5 49 -

011l 1 -

0 l1 0l3 oF 0 'l5 1 xi u- ,m. a.-<. .-c % =r.mn l

i l

l l events had produced no corresponding decrease in RCS. pressure or increase .in the.

Pressurizer Relief Tank pressure and temperature.

During the transient, a condenser steam dump valve failed to close and an operator had to fail the air to the valve for closure.

Cause of Event The Licensee has performed a Root Cause Analysis and h'as determined that the basic reason $

for the misoperation.of XVS-8010-C is that the expected margin of 250 psi (10 percent) . ')

between the normal operating pressure and the safety valve'setpoint pressure was reduced .

to zero. The Licensee has divided the contributing factors'in this margin elimination -

into two categories:. 1) valve-specific factors, and 2) non valve-specific-factors. Of-the 10 percent margin reduction, the analysis has quantified approximately 6.6 percent, and has identified factors which-contribute-to the remaining margin but:cannot be-quantified.

The valve-specific contributing factors are those factors which have'been tested and analyzed for XVS-8010-C. These valve-specific contributing factors, which make up approximately 6 percent of the eliminated 10 percent margin, are: 1)iloop seal discharge:

and 2) cetpoint variations. While the analysis was performed specifically for XVS-8014 C, the factors identified are expected to be generic to allisafety valves .but the numerical results are unlikely to be duplicated .  !

The first valve-specific factor contributing to the margin reduction is^the loss of loop.

seal on the valve. A test was performed on XVS-8010-C to measure the pressure-where' 1

' valve leakage exceeded condensation into the loop seal. This test, which is a sicw pressure increase ramp test performed on the test stand, is important because it specifies the pressure at which loop seal discharge occurs. For XVS-8010-C. this. -

pressure was found to be 118 psig (4.7 percent)~below the setpoint pressure of 2485 psig.

The second valve-specific factor identified by the Licensee is the setpoint variation which XVS-80.10-C exhibits when the valve body temperature is increased. A setpoint .

reduction of 22 psig (1 percent) was measured during a steam soak test using the Setpoint I Verification Device (SPVD) with the loop seal in place. This setpoint change is due to valve stem motion as the valve body temperature increases. This setpoint reduction may.

not be evident when the valve setpoint is established under ambient operating conditions. ,

The two valve-specific factors identified above combine for a reduction in setpoint of. '

5.7 percent. This agrees with other testing on XVS-8010-C which shows the valve has a l lower lift pressure (in excess of 150 psig) when exposed to steam rather than water.

Once the loop seal is lost, the valve is exposed to steam. Since a steam medium on the valve results in a setpoint change of approximately 6 percent (150 psig), this value is i used in quantifying the setpoint change. The 4.7. percent from loop seal. discharge and the 1.0 percent from setpoint variations are included in the 6 percent.

The second category in which the Licensee has divided the contributing factors for the  !

valve misoperation is the "non-valve specific category." This category contains the- '

. factors which depend on components other than the pressurizer safety valves. These unc ,ono m.

L9 8U [

+ _ .- . _ ~ _ ,, - _ ,

+

y{

NRC* Poem 384A . Ut NUCLE AN KEOULATMY COMMISSION 'f LICENSEE EVENT REPORT (LER) TEXT CONTINUATION = AreRovEo oMe No. siso-oio4 - t

• EXPIRES 8?31/83 FACILITY NAMfdi)- DOCKET NUMBER Qi LER NUMBER (6) . PAGE {31

~

"aa s

$4 W mra*=

Virgil C. Summer Nuclear Station o ls lo j o lo l3 p9l 5' 8l9 -

L 0l 1 F 1 0 bl 0l4 oF 0 l5'  ;

1 EXT 1t1 more space in regawet one ed@tumet MC Form 366Ks) l17) a 1

factors, therefore, will impact all of the' pressurizer safety. valves.' In this category, 5 only one factor. instrumentation calibration, can be quantified as- to its effects on the

'setpoint reduction.

Instrumentation calibration checks of "As Found" conditions after-the event and subsequent evaluations have resulted in a best estimate that the pressure I.

control instrumentation may have been a maximum of 0.6 percent low in indication / control. a

'A 0.6 percent loss of setpoint margin.is therefore attributed'to pressure-instrumentation l calibration.

i Other non-valve specific contributing factors identified as having a potential for- J resulting in a setpoint margin loss but which could not be quantified are:. .

1. Loss of clearance between the tailpipe flange.and' mounting plate.-  :

By analyzing the deformation of the irregular flame cut surfaces of the mounting. t plate where the plate contacted the tailpipe flange, stress levels in the valve body ~

were calculated to be 4,925 psi. Although this stress level is cnly approximately: g 18 percent of allowable, some disc-to-nozzle misalignment probably.resulted.' The non-symmetrical pattern of degradation of the disc and nozzle seating surfaces-identified during valve disassembly also confirm that.some misalignment was present.

2. Elevated temperature of the valve body'due to the insulation configuration results in body expansion that can reduce the setpoint by-moving the spring compressed disc '

in a direction to reduce spring compression of the tiisc-to-nozzle.

3. A correlation of indicated tailpipe temperature' variations with pumping-the-Pressurizer Relief Tank has been made. Recent close_ monitoring indicates a direct  ;

correlation between pumping the tank and minor temperature excursions on the i tailpipe of a leaking safety valve. The exact cause of this pattern has not been:

• identified, and at this time it appears to be only a. minor contributor.

4. Hydrogen concentrations in the pressurizer vapor space.-

2 Preliminary information indicates concentrations of H2 which may-alter the thermal characteristics of the valve. Hydrogen concentrations in the vapor space impact the condensation rate into the loop seal, the leak tightness of the valve (H2 vs. ,

water / steam), the control response of the pressurizer heater / spray-controls,,and.

other chemical degradation effects.

1 To summarize, the reason for the misoperation of XVS-8010-C is that the expected margin between normal operating pressure and the safety valve relief setpointt pressure was reduced to zero. Several contributing factors in the margin elimination for XVS-8010-C have been identified. Approximately 6 percent of the 10 percent reduction-has been-attributed to the difference in setpoint between steam and water. This difference occurred on XVS-8010-C when the loop seal was discharged. Because loop seal discharge '

and the resulting impact of the water /tteam difference is the most impacting of the quantifiable factors, the Licensee is focusing the corrective action plan on the loop seal system. The factors identified are expected to be applicable to all. safety valves. '

The numerical results are expected to be slightly different for each valve. -

go M uo

_ _ _ _ _- ~ J

--_.3, CC' Form 366A - U.S. NUCLE .2 KEIULAtoRY CoMMl881oh i

• • " 2 y LICENSEE EVENT REPORT (LER) TEXT CONTINUATION . " u,Rovfo ous Noso-osu :  !

,7y , EXPIRE 3. 8/31 C6 FACILITY NAMt,14 DOCKET NUMSER (2) LER NUMeER (6) PAGE (31 VIAR  :

NIa ' f U[veN ,

Virgil C. Summer Nuclear Station' o ls lo lo lo l3 l 9l5 819 -

0'-l1 l 1 -

0l1 0 15 oF 0 l5 1EKT ttf more space as reqwred, use addatena!NRC form 36liA'sl(1h I

Analysis of Event i

The safety significance of this event was moderate in that if the safety valve had'not' seated following the Reactor Trip, the Safety Injection System would.have been-challenged.

^

Immediate Corrective Action -

Immediate corrective action taken was the initiation of the manual: Reactor Trip. The  ;

operators monitored RCS pressure and were prepared :to initiate Safety Injection. Plant "

parameters were monitored and the plant was stabilized in Mode 3 at'a reduced; operating-pressure of approximately 2000 psig. The plant was subsequently taken to Mode 5 for. l removal and replacement of the: safety valves XVS 8010 B and C.-

Additional Corrective Action:

The Licensee has temporarily installed temperature detectors on the pressurizer' safety valves to' facilitate monitoring of the valve body inlet temperatures. The Licensee-has-

~

issued a special instruction detailing the monitoring of;these temperatures and the initiation of a plant shutdown should a valve body. inlet temperature reach 450*F. j NOTE: As the result of a subsequent pressurizer safety valve misoperation on August 25 s i 1989, noted in LER 89-015, the Licensee has revised.the shutdown temperature'to 390*F.-

L Additionally, at 350*F action will be taken to develop a plan for plant shutdown.taking into account, among other variables, the rate of change of temperature. 'The Licensee's.

l 1

management and the Resident NRC Inspector-will be notified when the safetyLvalve temperature reaches 350*F.

The Licensee intends to: 1) modify the pressurizer safety valves' internals for. steam application, and 2) eliminate loop seal capability on the pressurizer safety' valves.

.This modification is scheduled for the fifth refueling outage'(currently scheduled to I begin March 23,1990) and is contingent upon receiving the required materials in time for implementation.

Safety valves XVS-8010-B and C were replaced with spare valves. Valve'8010-B was.

replaced because it was experiencing minor leakage past the seat. Setpoints'for all three safeties were verified in Mode 3 prior to restart. <

Troubleshooting identified that the ralfunction of the condenser steam-dump valve was attributed to the current / pressure convertor being out of adjustment. Adjustments were made and the valve satisfactorily tested and declared operable.

Prior Event None -

L  :

p l

agroau ma C __ _ _ _ . .