ML20011E246

LER 89-036-01:on 891007,discovered That Various motor- Operated Valves Energized,Contrary to Assumptions Contained in Plant Fire Hazards Analysis.Cause Not Stated.Operations Initiated Required Fire Watches on valves.W/900131 Ltr

ML20011E246
Person / Time
Site:	River Bend
Issue date:	01/31/1990
From:	Booker J, England L GULF STATES UTILITIES CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
LER-89-036, LER-89-36, RBG-32236, NUDOCS 9002120299
Download: ML20011E246 (8)

Text

m.

( ], ,

! ' [1 - 4 ; {

L . c ., 3:

GULF STATES UTILITIES COMPANY l ama moeur,o mn uen e, w, u enec4au mS.e.uano l Aht A CQ(jf h'44 (M {lL44 346 l@

i i  !

January 31,1990 g RBG-32236 -

File Nos. G9.5, G9.25.1.3 i L

t' s

)-

I U.S. Nuclear Regulatory Commission i Document Control Desk-j Washington, D.C. 20555 Gentlemen:

River-Bend Station - Unit 1

! Docket No. 50-458 l i Please find enclosed Licensee Event Report No.89-036, Revision 1 for River Bend Station - Unit 1. This supplemental report is being submitted to provide the results of GSU's' safety assessment concerning valves which were energized contrary to the plant fire hazards analysis.,

Sincerely, '

f, . I. Wf u 'J. E. Booker

~

Manager-River Bend Oversight purP River Bend Nuclear Group gk

> SCtt hh -

3/TFP/ W/DCH/CKC/pg cc: U.S. Nuclear Regulatory Commission 611 Ryan Plaza Drive Suite 1000 Arlington, TX 76011 NRC Resident Inspector P.O. Box 1051 St. Francisville, LA:70775 INPO Records Center 1100 Circle 75 Parkway Atlanta, GA 30339-3064 l

'9002120299 900131 PDR ADOCK 05000458  !!!

g S PDC .'.

, l' '

l

9 g ses v4=we63 . new6.ni.,ca ca oveo one no siewin

• :: n e, sic

, UCENSEE EVENT REPORT (LE]

. .u r, iii oocui uu . m ..e #

RIVER BEND STATION o I s I o l o l 0 Lt is in 1 loFlc [7 Various System Motor Operated Valves Found Energized Contrary to Plant Fire !!arard Analyses Due to Failure to Implement' Design Documents 1

. v ., wn i.i ti . u== . i., n.on n <>> or i. ..eiuv... . 6v.. .I J wo=v. cav vea. vaa. "n',;;;.'6 $;;fll wo=t- onv vea. ' *c'uv v =awu pocaat =vves.isi j 0 15101010 1 I I g

, a I g g l l 1 10 o I7 9 9 9 19 01316 o 11 0l1 3 l1 910 oisioio,o, i l

, . . e ..v im . v.c v.v . ...v..... n i.... i ,e . ., - .. , ,,,ui an* 3 . .. . . . . = ? i.H . n n., .

. se maimui w mi.im es.n.H H.i n ni.i 1 io io . ini.:

. l .,

. n n.n..,

_ g; ,;, g ..

30 W.eH1 Heli N tt.elaini es.tp. sigh nH., J.s., d M W 46H1Hel 90 ?SieH3Hei te.ttleH3H.ehHet l 90 .MeeHt H.I 98 ?S .Gitual te tSteH3H.,

L OCt.ggg t CO.,f.et po. T . L t. (t3

=ist TELOP o=t =ww.E.

a.la cod 4 L. A. England, Director - Nuclear Licensing 51014 31 81 11 -1 41 h 4 5

. tin . u.. ...i.e ..,..6un ...e...... , . .. n <n, Civst 8vittw coupo%ght "'*g g 'C' "$'g0 (,'n'[,' Caust systtw Co.*>oment 8'Nyg at pg.o.Aegs p

I I I I I l l l l 1 1 I I I -

1 I l l l l l l l l l l l l

.u t. c.i .. e en o n.,

pou- ear _via.

'E'I

~1vii,,,.,,-.,uerr,osv..,,.n.ure, 1~o i , ,

.=v . .c t ,o . . . ... . .. . . . . -,, ,. , n . ,

At 1400 hours0.0162 days <br />0.389 hours <br />0.00231 weeks <br />5.327e-4 months <br /> on 10/17/89, with the reactor at full power in Operational Condition 1, it was reported to the shift supervisor that various motor operated valves in the plant were energized, contrary to the assumptions contained in the plant fire hazards analysis. Because these valves were not de-energized, this event is reportable as a condition that is outside the design basis of the plant.

Operations initiated the required firewatches on the valves and associated raceways, or the valves were de-energized. All valves have been reanalyzed and procedures are being revised to require the valves to be de-energized or provide justification to leave the valves energized. These actions ensure that a fire in any area in the plant would leave at least one method of safe shutdown unaffected.

Therefore, there was no significant impact on the health and safety of the public as a result of this event.

l L

l i

I .._ _ . . _

F I

• esas pere sea y a noc6 gam c a;v6'soav CommaID* Der

, UCENSEE EVENT REPORT (LERI TEXT CONTINUATION enaovio ou. =o me.m.

me. . , .

,mm ... aoeset ausen m 6.u.m . . . .

"*" "hle 'h*,.#

RIVER BUND STATION Olilol018lalslR 81 9 --

Ol 31 6 --

dl 0 12 0' Ob son n m . m mm a e r n ,. m m m REPORTED CONDITION At 1400 hours0.0162 days <br />0.389 hours <br />0.00231 weeks <br />5.327e-4 months <br /> on 10/17/89, with the reactor at full power in Operational Condition 1, it was reported to the shift supervisor that various motor operated valves (MOVs) (*V*) in the plant were energized, contrary to the assumptions contained in the plant fire hazards analysis (FHA). These valves are listed in Tables 2 and 5 of design Specification 240.201, " Fire Analysis and Evaluation Criteria", I and are shown to be assumed to have power removed during plant l operation. The list of valves affected consists of thirteen valves in I the residual heat removal system (*B0*), three in the fuel pool cooling system (*DA*), one in the reactor core isolation cooling system (*BN*), two in the standby service water system (*KG*), and one

]

in the main steam system (*TA*).

Because these valves were not de-energized as assumed in the FHA, compliance with General Design Criteria 3 of Appendix A to 10CFR50 was not assured; therefore, this event is being considered reportable under 10CFR50.73(a)(2)(11)(B) as a condition that is outside the design basis of the plant.

JNVESiljAllpN A detailed study of the FHA was prompted by the investigation of an

, earlier condition involving improperly installed fire walls at the Division I remote shutdown panel (reference LER 88-009). Comparison of the FHA requirements to the valve lineups and station operating procedures (SOPS) noted that the valves listed as ' remove power' in the FHA were instead energized. The reasons why the FHA requirements were not reflected in procedures and operational practice is unknown.

A review of the USAR and the original design criteria, and conversations with the Architect / Engineer (AE) Stone and Webster Engineering Corporation, demonstrated why these valves were shown to be de-energized. Two of the valves, 1E12*MOVF009 and 1821*MOVF019 were considered potential LOCA pathways due to a fire in the main control room or in the remote shutdown panel, in the case of the IE12*M0VF009 valve, the concern is to prevent a fire-induced opening of the low pressure shutdown cooling residual heat removal (RHR) (*BO*) system to the reactor vessel at operating pressure. This high/ low pressure interface valve is identified in section 9.5 of the USAR. An enable switch (*IS*) on valve IE12*MOVF008, the outboard isolation valve (*ISV*), is used to protect the RHR system from spurious actuations generated in a fire in the main control room. This switch is located in the Division I remote shutdown room, along with the controls for 1E12*MOVF009. A fire assumed to affect the remote shutdown room could potentially open both inboard and outboard valves, over pressurizing the RHR system with vessel pressure. A review of the operating requirements for the IE12*MOVF009 valve showed that plant operation in Operational Conditions 1, 2, and 3 with the valve de-energized is acceptable.

.,.4,... .u. .o ....... m ...

. i

.ne v s aucts.a ...uutoav ca ,

• • • J

, UCENSEE EVENT REPORT ILER) TEXT CONTINUATON **aovio om. =o ae-ea semes s no  ;

eno w ,nams m soonst wuamen m ,,, .,,, ..., .  !

m. 'umt- == )

RIVER BEND STATION o l$ {c le le l 4l g g g l9 --

01316 -

01 1 01 3 0' Ol7 an . . .= =. mu w n l Valve 1821*MOVF019 is the outboard isolation valve for the main steam '

isolation valve (MSIV) (*IS*) drains. A postulated control room fire could cause this valve to open spuriously, along with valves IB21*MOVF016 and 1821*MOVF085, dumping steam to the main condenser .

(*CDV*). When this scenario is considered with a loss of offsite ,

power, as required by the FHA, steam may be vented to the this atmosphere as main condenser vacuum is lost. ,

The remaining valves were listed as ' power removed' by the A/E because the divisional separation of the raceway was never verified to meet 10CFR50 Appendix R standards. As a result of this event, GSU has identified all affected cabling, including associated circuits, and has evaluated their separation and fire area locations. This evaluation shows that all valves listed have adequate Appendix R separation. This analysis considers the effect of a fire anywhere in the plant, with the requisite loss of offsite power.

The three spent fuel pool cooling (SFC) (*DA*) valves require '

administrative control as a fire may effect either the inboard or outboard division, possibly causing a loss of fuel pool cooling to the containment fuel pools. This situation could only occur during refueling, and assuming worst case conditions, 1.25 hours2.893519e-4 days <br />0.00694 hours <br />4.133598e-5 weeks <br />9.5125e-6 months <br /> is required to heat the pool to the upper limit of 150 degrees F. This is more than adequate time to manually reposition the valves if required. >

A review of previously submitted LERs by River Bend Station revealed six previous LERs related to design requirements not being reflected in plant operating procedures. LERs86-066 and 87-026 identified fire doors which were not listed on the appropriate surveillance '.est procedures (STPs), LER 88-010 identified a secondary containment door which was not listed on the appropriate STP, LER 89-003 identified a breaker which was not listed on the appropriate STP, Additionally, IER 86-059 identified that a design modification to the low pressure coolant injection line that changed the location of the piping high point vents was not reflected in the appropriate STP and LER 87-030 identified that the appropriate area temperatures were not being monitored in tne reactor plant component cooling areas as required by the Technical Specifications. However, none of these events were related to ensuring that the assumptions of the FHA were properly reflected in the plant operating procedures.

CORRECTIVE ACTION Engineering identified two possible methods to satisfy the requirements of the analysis and the plant Technical Specifications (TS). One method was to de-energize the M0Vs as assumed in the FHA.

The second method was to treat the valves and the associated electrical raceways as having missing installed fire barriers, and to take' the action prescribed in TS 3/4.7.7, instituting a roving firewatch in the affected areas. The latter method was adopted and

.uao.oi +. . m m g....

.ne ann ve wma ...wa ,.  !

• • • UCENAEE EVENT REPORT (LERI TEXT CONTINUATION maovie o e no ow.oie.

sa es e n a

.as e n, seean ei ,,,,,,,,,,,,, , , , , ,

.... atm.*6 -

n t; RIVER BYND STATION # lilo jo le l al gl p gj g -- djg - ol1 ol4 0F ob veu . .se a amewon the required firewatches were initiated. Two valves, 1E12*MOVF009 and 1B21*MOVF019 were in areas not accessible to firewatch personnel and ,

were therefore de-energized. The choice of methods considered the effect on Operations personnel due to a number of de-ehergized valves, including the effect due to lit annunciators, and the additional requirement to enter the auxiliary building to de-energize the valves.

Operations will change the valve lineups to show valve IE12*MOVF009 de-energized until reactor pressure is reduced below 135 psig reactor pressure, which is the system pressure for shutdown cooling piping.

Valve 1821*MOVF019 is required for startup and is used in some operational transients. Therefore, the status of this valve was changed to closed and de-energized. However, administrative controls have been implemented to allow station personnel while at local motor l control centers to open the valve when needed. This will enable positive control of this pathway in the event of a control room fire.

The valves are currently tagged out. Revision of the valve lineup and procedural changes have been completed. The three SFC valves require I administrative control under worst case conditions to be aligned to supply cooling to the upper pools within 1.25 hours2.893519e-4 days <br />0.00694 hours <br />4.133598e-5 weeks <br />9.5125e-6 months <br /> to limit the upper pool temperature to 150 degrees F. This caution will be added to the pre-fire strategy for the area in the fuel building containing these valves by February 28, 1990. A change document revising Tables 2 and 5 of the FHA has been completed.

SAffTf_ ASSESSMENT As identified in the investigation, the majority of the valves and cabling were found to have sufficient separation to enable re-energizing. In the cases of SFC, a~ fire could potentially cause either division to close while the plant was in a refueling cycle and fuel pool cooling to the upper containment pools would be required.

The closure of these valves would be noted with any gradual rise of pool temperature. Assuming a fire did affect these valves, upon detection of increasing pool temperature, valve positions could be determined and repositioned if necessary. Alternate methods exist to provide water to the upper pools (water from fire protection system

(*KP*) for instance). Therefore, pool temperatures would not be expected to exceed design levels, i.e. operator actions would prevent any overheating in the upper pool. For all valves except 1E12*MOVF009 and 1821*MOVF019, analysis demonstrates that a fire in any area would leave at least one method of safe shutdown unaffected.

Valve 1821*MOVF019 is the outboard drain and containment isolation valve serving the inboard MSIVs. This valve, in combination with valves 1821*MOVF016, 1821*MOVF085 and 1821-MOVF021 provide a pathway to the main condenser for condensate and warmup. The piping and instrumentation diagram (PID) shows that valves F016, F019 and F085 are normally open wb11e F021 is normally closed. The only area where the control circuits for all these valves may be potentially affected

. o.o u m m =

g....

w maa o e mucanneovmo , ca

• • • UCENSEE EVENT REPORT (LERI TEXT CONTINUATION awaoup owe ao mo-om axe, ass nee passuTV maan He D0Catt seveSet st' Les muutee les past tai

's.a "tdu." "'Jf.#

RIVER B'END STATION 0 l6 l0 l0 le l4 l gl g gl9 --

D b 16 -

0 11 o l; M nb von se . .ar a an w nn by the same fire event is in the main control room. The concern is that a fire in the main control room may cause the F021 valve to spuriously open, providing a path for steam release to the condenser, ,

Note that the contribution of valve 1821-A0VF033, in parallel with F021, is neglected due to the small (1") line size. The other drains are 3" lines. Steam release to the condenser is of no consequence with condenser vacuum and circulating water operational. Assuming the main control room fire has also damaged these systems, eventually steam will be vented from the condenser through the air release valves (ARVs). A loss of offsite power (LOOP), required to be considered for Appendix R, would have the same effect on the condenser.

The potential effect of steam release on offsite dose through the path outlined above has been evaluated. Several cases were considered as follows:

. Steam release continues unabated for the 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> period until cold shutdown.

. Steam release is terminated by operator action at 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />.

. Determination of the time required to close the path before 10CFR part 20 limits were exceeded.

. Using the maximum iodine levels, the analysis showed that at a 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> duration, offsite dose calculated was less than 1% of the 10CFR 100 limits, and only slightly above those in 10CFR20. Closure of the release path at less than I hour and 39 minutes would limit release to the normal operation limits proscribed by 10CFR20. The results were much less than accidents previously identified, such as a steam line break outside the containment. Another case considered in the calculation was the offsite dose after 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> based on the actual expected iodine levels, rather than the maximum. This analysis shows the calculated offsite dose is less than the 10CFR20 limits.

Any release would be detected by monitoring equipment either inside or outside the plant. The effect on reactor water level would be small but noticeable. These indications would guide the operators to manually close this leakage path. A PRA conducted on this scenario as outlined above indicato that the probability of the event is estimated at 1.9E-04 over the time the valve was energized.

l In summary, the probability of this postulated event was low over the duration of time that the valve was energized. Furthermore, the offsite dose calculation based on the actual expected iodine levels l

provided an offsite dose less than the 10CFR20 limits. Therefore, GSU concludes that this postulated event did not adversely affect the health and safety of the public.

g.... .u a o.o io o. sow.

"' v .ucaan ...u w ,so

=== e men

• • • UCENSEE EVENT REPORT RER) TEXT CENTZUATION macao o.e *o so-me toian swa

=vn ==== n, ooene, =ase. ei o . ,,u . i., ...e is n'a "htMP W11:

RIVER BNND STATION 0lSjol0l9lal sl g dg --

013l6 -

01 1 Ol 6 0F Ol7 nn, . . sn.=se a mmw nn Valve IE12*MOVF009 operates in tandem with IE12*MOVF008 to provide containment isolation to the RHR normal suction path. These valves also protect the low pressure (200 psig) RHR system from the high pressures (1050 psig) in the reactor pressure vessel (RPV). This pair of valves is protected in the main control room by an enable / disable switch for the F008 valve located in the Division I remote shutdown room. During power operation, there is no power to the circuits for F008 in the control room, although position indication is provided. A fire in the main control room could only affect the F009 valve, preserving the high/ low pressure interface. However, these valves also are both controlled on the Division I remote shutdown panel and a fire in this area could potentially affect both valves, causing them to spuriously open. During power operation, this would flood the RHR system with vessel pressure.

Exposing the RHR system to RPV pressure would certainly cause extensive damage, particularly to thermowells, instrument taps and

. pumps. Damage would occur downstream of F008, as the piping between the valves is designed to RPV pressure. A PRA was performed to determine the effect this event would have on core damage frequency (CDF). This analysis conservatively assumes that: a) RHR piping ruptures in the auxiliary building and cannot be isolated, and b) RHR loops A, B, and C are lost due to flooding, pipe rupture or loss of function due to the original fire. Reactor water inventory would be

, maintained by a combination of HPCS and LPCS. Over the long term, additional water is available from the service water or fire protection water systems injecting into the vessel or suppression pool. The estimated CDF given this scenario is 5.8E-08, lower than the total RBS CDF of 5.0E-06. A LOOP is not a precursor to this event, but is evaluated in the response of the plant.

This PRA is also conservative by not considering the effect of the protection afforded by the Reg Guide 1.75 separation provided in the shutdown cabinet, separating one division from another. Doors in the Division I room are locked, so routine entry is not possible. The room is provided with smoke detection and response to any fire detected would be rapid due to the proximity of the control roou and the fire brigade equipment locker. There is no equipment in the room that requires servicing with flammable or combustible liquids.

An interfacing system LOCA would be a severe accident, but recoverable. The CDF derived from the analysis indicates that the safety significance is low, and is necessarily conservative. There have been no fires at River Bend Station in the areas discussed. The health and safety of the public was not at risk from this postulated low-probability event.

With .the exceptions of valve 1B21*MOVF019 and valve IE12*MOVF009, I engineering evaluation shows all valves listed have adequate Appendix R separation and/or can be administrative 1y controlled. The F009 and I

.ue yo m .s. m .s.

g....

ys.{. g,. ., .p,, . . . - -

e -

p '!.a ; ', + 1 asac cer, seen v s aucttin astutatong :- r A ***! " . UCENSEE EVENT REPORT (LER) TEXT CCNTINUATON u aovie oue =o m .oie.

o ,

exmu e,we usion man m . socaat muussa m ... .,,, . . , , ,

nn- "tL11 ' "'11.D RIVER DE'ND STATION 0l6l0l0lej41SlB 81 9 -

01316 -

011 0l7 or 0 17 i t =, . . ae a muw im F019 valves.have been de-energized and are subject to the corrective-

. actions. described in this report. No fires have occurred in the -areas ={

under consideration. There-was no significant impact- on the health >

and:-safety of the public as a-result of this event.

N_0TE: EnergyL Industry >

the text as (*XX*) Identification System Codes are identified in

. l

+

{

8i 3

-i P

4 i

i

. i, r

o

.u.-,....

7;~ .

. - . .