ML20006E289

LER 90-001-00:on 900112,reactor Scrammed Because MSIVs Were Less than 90% Open.Caused by Component Failure & Configuration of Condenser Vacuum Sensing Lines & Instruments.Valves replaced.W/900207 Ltr

ML20006E289
Person / Time
Site:	Hatch
Issue date:	02/07/1990
From:	Hairston W, Tipps S GEORGIA POWER CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
HL-960, LER-90-001-01, LER-90-1-1, NUDOCS 9002220524
Download: ML20006E289 (10)

Text

& , ,. 1 ,w -

, a, ,

dekwg a Piueer'Compas-

{,'[ 'ti t-. '3'-

, jht ,l, 333 Pedmont Annue '

, hf- '*-

c. Atfanta. Gevaa 30300

,, Tetophone 404 520 3195 -

+' ~ >

"' i MMng Acidnss g - 40 ltwirm Corner Pmkway .

, Pp:;t Offce Du 1295 .

~

D>msngham, Alabama 35201:

, - letphnew 205 WJ $081.-

=

t% > + n 1HW se: tem ,

k ' W. G. Hairston,'lli >

SenorVco h er;wed'-

3.; . , . Nogedf. Oswatg>ns ;

T, HL-960' y , 000197-February 7, 1990 l U.S.: Nuclear Regulatory Commission VATTN: . Document Control Desk:

Washington..D.C. .20555 i

PLANT HATCH.- UNIT 2 NRC= DOCKET 50-366

-OPERATING LICENSE NPF 1 LICENSEE: EVENT REPORT

!b COMPONENT FAILURE AND INADEQUATE DESIGN CAUSE GROUP I ISOLATION AND SGM N Gentlemen: :

In accordance with the requirements =of 10 CFR 50.73(a)(2)(iv), Georgia-Power Company

-is ; submitting the enclosed Licensee Event Report (LER)~

concerning7 the. unanticipated actuation of some Engineered Safety Features-(ESFs). LThis' event occurred lat Plant Hatch - Unit 2.

Sincerely,

, n

~

. nl. k.

W. G. Hairston, III JJP/ct y i

m

Enclosure:

LER 50-366/1990-001

,m

• c:

~

(See next page.)

w a-q i

s- ' 9002220524 900207 SIy"l PDR ADOCK 05000366 g "

.. S PNV ;jg ,

s.

tr F J

'I ri:

'N

~

e. i LCicorgid Potverla8k l U.S.- Nuclear Regulatory Commission-February 7 -1990- .

Page Two 1 l

's I c: Georaia' Power Company

~

,' LMr. H. C.- Nix,. General Manager - Nuclear Plant

.Mr.-J. D. Heidt, Manager Engineering and Licensing - Hatch i

GO-NORMS

.U.S. Nuclear Regulatory Commission. Washinaton. D.C. 1 L Mr. L. P. Crocker, Licensing Project Manager - Hatch 1

. U.S. ' Nuclear Reaulatory Commission. Reaion II 1

l. Mr. S. D.:Ebneter, Regional Administrator Mr.1 J. E.'Menning, Senior Resident Inspector - Hatch ,

i l

I I

L o i la t

r l-l 000197 0

l 5 .

.} "

term,308 U.S. NUCLE AR kt1ULaTORY COMMISSION APPROVED OMS NO 2180 0104 m'2 8 8 8'c LICENSEE EVENT REPORT (LER) f ACILITY NAMA 01 CSCRET IJUMBER (2) PAGE(38 PLANT HATCH, UNIT 2' ois;oioioi 3;6 6 1 ilorl 0 18 I11L8144 COMPONENT FAILURE AND INADEQUATE DESIGN CAUSE GROUP I ISOLATION AND SCRAM

$VtNT DATE (Si Lt R NUMet R (si REPORT DATE (7) OTHif4 F ACILITill INVOLVED (4) 80 MONTH VG AF; f e.CeLit v N AMEs DOCKE I NUM8tRI3)

MONTH DAY VfAR YEAR j (" b ((y DAY 01610i0i0 1 i ; .l w ~~ ~

Ol 1 1l2 9 0 9l0 0l0 l1 0l 0 0l2 0 l7 9l0 o iss o ioloi i i THis REPORT is sv3Mitiso PunsvaNT TO THE REOulRtMENTs Or to CsR 6 <caece . er more er rae reir ..,i 01:

M00519) j 20 402th) 20 406tel X .0.13i.H2H.o 73.7 uni n 20 406teH1 HH $0.38(sH11 [ lo.73teH2Hel 73.7 Hel 7

~ 0 01 110 i 0 20 40siainH.n m3 i.Hn

.One H2H..d _ g,,gs-g,;

Hi gey,r , j 20 406teH11(ad 60.73tel(2HH 50.73(sit 2HvetiHAl. J664, a 20 406teH110e 60.73te H2 HM 80.73teH2HveuMSI 20 406teH11tvl 60.73teH2 Hod 50.73teH2Hal LICENSEE CONT ACT FOR THIS LER H2)

NIME TELEPHONE NUMBER Astt A CODE Steven B. -Tipps, Manager Nuclear Safety and Compliance, Hatch 9 l12 3 6i 7l-i 7l 81Si l 1

COMPLETE ONI LINE FOR E ACH COMPONENT F AILURE Dt3CRIBED IN THis REPORT H36

- COMPONINT

"" A C-CAUSE COMPONtNT CAUSE Sv 81 t M g {0 PR SY ST E M Ma%AC R E,P O NPR X- Di 6.

i RTY i i H03 i i i7 N i , i i i i i X BGI  ; R;L iY G ;0 ;8 l 0 Y i i i i i ; i .

SUPPLEMENT AL REPORT ERPECf t0 (14) MONTH DAY YEAR SU8 Mis $10N

^ " "

Tl vti v, ,,,_ ,ey,e, ,,,crio sU,wwoN oe rs, 7 NO 06 0li 910  ;

A. T R AC T a,-, R, , -. , s e,e,e-a,e+ a, . . , e .~e eve..en.o . n ei j

On 1/12/90 at approximately 1610 CST, Unit 2 was in the Run mode at an approximate power  !

of 2436 CMWT (approximately 100% of rated thermal power). At that time, the reactor l scrammed because the Main Steamline Isolation Valves (MSIVs) were less than 90% open. The  !

MSIVs had-isolated on a Group 1 Primary Containment Isolation System (PCIS) signal which  !

resulted from a false low condenser vacuum signal. The High Pressure Coolant Injection'

!. (HPCI) system automatically initiated and injected on low reactor water level as  !

. required. 'Following water level recovery, HPCI injection valve 2E41-F006 closed i automatically on high water level; however, it could not be re-opened when Operations personnel subsequently attempted to start HPCI manually. The Reactor Core Isolation Cooling system 'and two Control Rod Drive system pumps were used to control water level-  :

following -the failure of valve 2E41-F006 to open. 3 The root causes of the scram are component failure and the configuration of the condenser  !

i vacuum sensing lines and instruments. The disc of root isolation valve 2N61-F588D separated from its stem isolating the common sensing line for vacuum switches 2821-N056C l and D. Consequently, these switches then sensed a low condenser vacuum and, because they  !"

input to the:A and B trip systems respectively of the isolation logic, the MSIVs isolated. .The cause of valve 2E41-F006 failing to open appears to be component failure.

The heater strip of a thermal overload relay in the valve motor's local starter failed causing an open circuit to the motor. Further information will be provided in an update to'this report.

1 Corrective actions for this event included replacing the root isolation valves with a new l type of valve, replacing the thermal overload relays in the local starter, and 1 reconfiguring the tubing for the vacuum switches.

1 r.- 3ee

~ ~ ~^"

3 4 M * '

l

[hk 00AC P4 MSA '

U,$.NUCLtA3 El1ULATORY CoacMeme04 UCENSEE EVENT REPORT RER) TEXT CDNTINUATISN Arraivto o= No. sm-o*

l. g t

IXPtRES: 8/31/W :

Q9 ,

FAceL4rv NAtet Hl, DOCKET NUGADER 2) Ltm NUndSER tel PA06 (3) y vida -

" h M 6 'W.?:

y PLANT. HATCH, UNIT .,

015l0lolol31616 9l0 -

' 010 l1 - -

0 10 ' 0 12 0F 0 18 rixt n w. wwe r asw. inn .

i PLANT AND SYSTEM IDENTIFICATION r

. General. Electric - Boiling Water Reactor . .

Energy Industry Identification System codes are identified in~ the text as' (EIIS J' Code XX).

SUMMARY

40F EVENT- ,

L On 1/12/90 at approximately 1610. CST, Unit 2 was in the Run mode atLan ' approximate'

.po'wer of 2436 CMWT .(approximately 100% of rated thermal power). At that' time, the reactor. scrammed because-the Main Steamline Isolation Valves (MSIVs, EIIS Code:SB) ,

t were less than -90% open. The MSIVs had isolated on a Group 1 Primary Containment.

L

' Isolation System (PCIS, EIIS Code JM)~ signal which resulted from a sensed low condenser vacuum. Vacuum switches 2B21-N056C and D, which provide input- to the At s

' and-B trip systems of the PCIS Group 1 isolation-logic, respectively, sensed low '

condenser vacuum after root isolation valve 2N61-F5880 failed and isolated the switches'E common'. sensi ng :1_ine. - During normal scram recovery activities, the HighL Pressure' Coolant Injection (HPCI, EIIS: Code BG) system; automatically' initiated and -

-injected'on low reactor water level as required. Following water level recovery,-

HPCI injectionivalve 2E41-F006 closed automatically on high water level; Lhowever, it could not be re-opened when Operations personnel subsequently attempted to start -

HPCI manually. The Reactor Core Isolation Cooling (RCIC, EIIS Code BN) system and two Control. Rod' Drive (CRD, EIIS Code AA) system pumps were used to control water level following-the failure of valve 2E41-F006 to open.

The.: root causes of the scram are component failure and the configuration of the- I

. condenser vacuum sensing lines and instruments. The disc of root' isolation valve-2N61-F588D separated from its stem isolating.the common sensing line for vacuum .

switches 2B21-N056C and D. These switches then sensed a low condenser vacuum and, '

because they input to the A and.B trip. systems respectively of. the isolation logic, the MSIVs isolated. The cause of valve 2E41' F006 failing to open; appears to be W component' fail ure. The heater strip of a thermal overload relay in the valve motor'sLlocal ~ starter failed causing an open circuit to the motor. Further

, . information will be provided in an update to this report.

l l- Corrective actions for this event included replacing the root isolation valves with L' .a new type of valve, replacing the thermal overload relays in the local starter, and

reconfiguring the tubing for the vacuum switches, retaining the acceptability of the i design relative to single ~ failure criterion and reducing the potential for

' unnecessary full- Group 11 isolation logic actuations.

l l

l-i 4 NIC FORM 366A . *U.S. GP0y 1968 520 589 U00 4 C$h:

1. .. .. wr ,_;

f gestC Po4n asea U.S. NUCLt?4 [EIULAfoAY ConsMt0060N I"l

- LICENSEE EVENT REPORT (LER) TEXT CONTINUATIDN APP oveo ove No. mo-oio4  ;

ExPimfS 4/31/0B -

9Acatify haast 41) DOCOLEY NUhASSR (2) ggggyggggggg p4gg (3) -

= "W.T 71*.7 F

PLANT HATCH, UNIT 2 o ls lo lo j o l 3l6 l 6 9 l0 -

0l 011 -

0 l0 0 l3 or 0l8 ifAT (# mom spece 4 reewed use ochstomer NAC Fenn JORWs> (17)

DESCRIPTION OF EVENT On 1/12/90 at-approximately 1508 CST, Unit 2 was in the Run modo at an approximate power of 2436 CMWT (approximately 100% of rated thermal power). At that time, a Group 1 PCIS signal was received in-the B isolation logic trip system. Operations and Instrument and Controls (180) personnel investigated the unexpected one-half Group 1 isolation signal. They discovered relay 2A71-K680 in the PCIS logic was de-energized. This relay de-energizes when vacuum switch 2B21-N0560 trips on low -

condenser vacuum. It appeared at the time that relay _2A71-K68D had de-energized due to a failure of vacuum switch 2B21-N056D; chart recorder 2N21-R602 showed ~no indication of actual loss of condenser vacuum.

i At approximately 1610 CST, before I&C technicians had gone to examine vacuum switch L 2B21-N0560, the A isolation logic trip system actuated when vacuum switch 2B21-N056C L tripped and a full Group 1 isolation signal resulted. The MSIVs began to close in i

response to the full Group 1 isolation signal as designed. When the MSIVs closed to :

> less than 90% open, a full Reactor Protection System (RPS, EIIS Code JC) trip signal was generated as designed and the reactor scrammed.

With the MSIVs' fully closed, the reactor was isolated from the condenser (EIIS Code SQ) and reactor pressure began to increase. Safety Relief Valves (SRVs, EIIS Code JE) 2B21-F013A, D, E, and H lif ted to relieve pressure as designed. This action, in.

conjunction with the high pressure portion of the Low Low Set (LLS) arming logic having been fulfilled, armed LLS logic and LLS SRVs 2B21-F0138, F, and G lifted to control reactor pressure in the LLS mode. Reactor pressure peaked at approximately 1117 psig and LLS maintained pressure between 850 psig and 990 psig thereafter as ,

designed. 1 As the SRVs actuated to relieve reactor pressure, water inventory was -lost to the Suppression Pool (EIIS Code BT) as expected. Reactor water level decreased to  !

approximately minus 40 inches relative to instrument zero (to approximately 10.4 feet above top of active fuel). PCIS Group 2 and 5 isolation signals were received and all Group 2 and 5 Primary Containment Isolation Valves closed as designed.

Additionally, both Recirculation Pumps (EIIS Code AD) tripped, HPCI automatically started and injected to the vessel (RCIC had been started manually by Operations personnel prior to reactor water level reaching RCIC's automatic injection point),

and both trains of the Standby Cas Treatment (EIIS Code BH) system started.

With HPCI and RCIC injecting water into the reactor, water level increased to the ~

high reactor water level setpoint (approximately 52 inches above instrument zero)

. and HPCI and RCIC tripped per design. Reactor Feedwater Pumps A and B (EIIS Code SJ) also received a trip signal although they had stopped injecting water into the reactor vessel earlier in the event; with the MSIVs closed, no steam was available to drive their turbines. As the LLS SRVs continued to cycle to maintain reactor pressure between 850 psig and 990 psig, water level again began to decrease. i Operations personnel attempted to start HPCI manually to control reactor water level; however, HPCI injection valve 2E41-F006 could not be opened from the Main Control Room using its Remote Manual Switch. Operations personnel proceeded to use i

RCIC and both CRD pumps to recover and maintain reactor water level. HPCI was used l in the' Full Flow Test Mode to assist in controlling reactor pressure.

i y ,,, _ u... m , n.. - .. -

i SIAC P.e'ai ageA - U,8. NUCLEA3 IEIULATORY COMM80840N LICENSEE EVENT REPORT (LER) TEXT CUNTINUATIDN" . APPRovio oms No. 3 noaios l IXPIRES: 8/31/W D j PACILITY tsAMt H) DOCKET NV,SER (2} ggy NUMB (R (Gl . PA04 (3) -

- A- w m:.n l 1

' PLANT' HATCH,= UNIT 2- 0 l5 lo lo lo l 3l6 l 6 9 l0 -

0l 0l1 -

0 10- 0 l4 OF 018 1nX1n . ,*.o M amer ane,mn g -- At approxima.tely 1704 CST, the unit was in a stable condition _ in the. Hot Shutdown .1 mode. . Reactor water level was being maintained at the normal level of- '!

approximately 37. inches above instrument zero, Reactor. pressure was being '

maintained at 'approximately 800 psig. j 3

CAUSE.0F THEL EVENT The root causes of the scram are . component failure and the configuration of the- ,

, condenser vacuum sensing lines and instruments. The disc in root ! solation valve i

L-

_2N61-F588D separated from its stem as a result of excessive wear due to vibration-  !

.in the -vacuum sensing line from- the condenser. The keeper ring and stem shoulder had worn!to the point where the disc fell off the stem. When the disc fell off the stem, it isolated the vacuum sensing line which feeds vacuum switches L 2821-N056C and D. Over an unknown period of time, the isolated line lost vacuum, l probably from small ~1nleakages through various sources. The line lost vacuum to the point where, at approximately 1508 CST, switch 2B21-N056D tripped. Due to slight differences in _ actual trip setpoints (the as-found trip. setpoints- of the two switches were within procedural tolerances, but slightly different), the line had to further lose vacuum to actuate the second vacuum switch; therefore, switch 2B21-N056C did not _ trip until approximately 1610 CST.

p 'The configuration of the condenser vacuum sensing lines and instruments relative l

to plant-reliability considerations resulted in an unnecessary full Group 1 isolation and the subsequent scram. Two vacuum switches were fed by one of two sensing lines off the-condenser; two more vacuum switches were fed by the other - ,

sensing line. Each-vacuum switch, in turn, provided input to one of four channels L' in the Group 1 isolation logic. These four channels- (A1, A2, B1, and B2) are ,

divided'into-two trip systems-(A and B). One of the two channels in each system -

must. actuate to trip: the system and both systems must trip to generate a full Group 1 isolation signal and cause the MSIVs to close. Vacuum switch 2B21-N056D lprovided input to the B2 channel (the B trip system) and vacuum switch 2B21-N056C '

provided input to the A2 channel (the A trip system). With the two switches fed by the same sensing line providing inputs to channels in different trip systems, a single failure can cause false trips in both systems and a full Group 1 isolation signal. This design was implemented to meet single failure criterion should one of the two sensing lines fail in the Lbility to sense low condenser vacuum; however, it also resulted in an increased potential for unnecessary full isolation logic actuations. It should be noted that the low condenser vacuum (pressure)

- switches for the turbine (EIIS Code TA) trip logic are arranged the same way as V the switches for the Group 1 -isolatio'1 logic. The' trip logics also are the same.

/T " Because a turbine trip will result in a reactor scram, the design of the turbine trip switches and logic is also inadequate.

I 9

i NKC PORM 3664 *U.$. CFOs 1,84 SN-589 00 0 0 (t9 80 -

J

b[

senC form 300A U 8 NUCLE AA EEIULATE3Y COMMISSION 1" '

LICENSEE EVENT REPORT (LER) TEXT CONTINUATION' cPzoveo ous wo. ano.oto4 EXPIRES: 8/31/95 :

P ActLITY hAMt Hi DOCOLET NUMBER (2) LER NUMSER (8) PA05 (3) -

vtan Stu t AL -

mt e N PLANT HATCH, UNIT 2 o is lo lo lo l 3l6 l 6 9 l0 -

0l0l1 -

0 l0 0l5 oF Ql8 TEXT (# more apose 4 mewed use sewsow MC form 3054 W (17)

The cause of HPCI injection valve 2E41-F006 failing to' open appears to be

, component failure. The heater strip of a thermal overload relay in the valve motor's local starter failed. The heater strip (a metallic strip which is part of the thermal overload: relay) fused as attempts were made to open the valve. This created an open circuit to the motor. No current could reach the motor; therefore, it could not be energized to move the valve. The design configuration was reviewed and found to be in accordance with the applicable regulatory. guidance and industry standards. The review of the design configuration confirmed that, under the condition experienced, the thermal overload heater strip would be predicted to fail before other components in the motor circuitry. Investigation is: continuing into the cause of the failure'of the heater strip.

REPORTABILITY ANALYSIS AND SAFETY ASSESSMENT This report is required per 10 CFR 50.73 (a)(2)(iv) .because an unplanned actuation of the RPS and Engineered Safety Features (ESF) occurred. Specifically, the RPS-was initiated automatically on MSIVs less than 90% open. The other ESFs which activated during this event were the PCIS valve Groups 1, 2, and 5; the HPCI System; LLS; and the Standby Gas Treatment System. This report also is required per 10 CFR 50.73 (a)(2)(v) because the HPCI system did not function as designed following initial recovery of reactor water level. The injection valve's motor control circuitry failed thereby preventing HPCI from being used for continued reactor water level control.

The RPS automatically initiates a reactor scram to ensure the radioactive materials barriers (such as fuel cladding and pressure system boundary) are maintained and to mitigate the consequences of transients and accidents. The MSIV closure scram is provided to limit the release of fission products from the nuclear system. Automatic closure of the MSIVs can be initiated as a result of E various conditions. One of these is low condenser vacuum. Lcw condenser vacuum L indicates a possible leak in the condenser. Closing the MSIVs prevents potential loss of reactor coolant and potential release of radioactive material from the nuclear system process barrier.

L The MSIVs have position switches installed on the valves. These switches provide

, RPS trip signals. If the MSIVs were to close suddenly, this could cause a rapid pressure increase in the reactor vessel. This pressure increase would affect the  !

reactor vessel (due -to the pressure increase) and result in a positive reactivity '

insertion (due to void collapse). The MSIV closure scram anticipates the neutron '

flux scram and the high pressure scrams. In this event all of the three RPS L scrams $SIV closure, neutron flux and high pressure) were operable and the MSIV .

E position scram functioned as designed to terminate power production prior to the l other variables (pressure and neutron flux) exceeding their trip setpoints.

l

)

romu seen u.s. cro, tees.szo-m,cou ro

h N8tC PoiJ 334A U.S. NUCLEAQ Kt1ULATOM COMMISSION b  ?"

LICENSEE EVENT REP 3RT (LER) TEXT CONTINUATION Armoveo o e No 3t:o-oio.

EXPIRE $: B/31/N P ACILITY NAM 4 {1) DOCK 61 NUMBER (2) LER NUMetR ($p Pact (3) naa "50;'." -

0 #,70 PLANT HATCH, UNIT 2 015 lo lo lo 1316 l 6 9 10 01Ol1 -

0 10 Ol6 OF Ol8 ftXT fr more Swee 4 sccused, use esistooner NAC Form M W (17)

Following the scram, reactor water level was restored via the automatic initiation of HPCI and the manual = initiation of RCIC (RCIC was initiated manually prior to its automatic initiation setpoint). The SRVs operated in their relief and, later, LLS modes to control reactor pressure. Consequently, reactor vessel pressure was

. maintained well below vessel design pressure and ' vessel level did not decrease below approximately 10.4 feet above the top of the active fuel.

The HPCI system is provided to assure that the reactor is adequately cooled to limit fuel-clad temperature in the-event of a small break in the nuclear boiler system causing a loss of coolant.which does not result in rapid depressurization i of the reactor -vessel . . The Automatic Depressurization System ( ADS, EIIS Code JE) is a backup for the HPCI system. Upon ADS initiation, the reactor is depressurized to a point where either the Low Pressure Coolant Injection (LPCI, EIIS Code B0) system or the Core Spray (CS, EIIS Code BM) system can operate to maintain adequate' core cooling.-

- In this event, the HPCI system was rendered inoperable following successful automatic initiation when its injection valve failed in the closed position when a manual re-start of HPCI was attempted. The LPCI pumps and their associated equipmsnt, ADS, and both loops of CS were operable. Based upon the Unit 2 Final Safety Analysis Report (FSAR), either loop of the CS system or the LPCI system can supply sufficient cooling to the reactor for any rupture of the nuclear safety boundary up to and including the Design Basis Accident (DBA).

Based on the above _information, it is concluded that this event had no adverse 4 impact on nuclear plant safety. The above analysis is applicable to all- reactor power levels.

4 NXC DOMM 344A 'U.S. CP0s 19,0 520-58000070 0 43)

~4

=ac .6 sena vs wuctuni:vtatoRv coMuission F ~

~ UCENSEE EVENT REPORT (LER) TEXT CENTINUATION umoveo oue no me.o*

IXPlRES. 4/31/88 .

f act4tiv NAMG til (80Cetti NUMeta (2) ttR NUMcER 16) PAGE (3)

YEAR Nm Naa l i

PLANT HATCH. UNIT'2 01510 lo l o l 316 l 6 9 10 01011 -

010 0 17 of 0l0 TEXT w ouro apose 4 #seured esse esidennet NRC Form JE4W (17) i CORRECTIVE ACTIONS i Root isolation valves 2N61-F588B and D (in the sensing lines for Group 1 isolation j logic vacuum switches) and 2N61-F061 and F064 (in the sensing lines for turbine trip-

' logic vacuum switches) were removed and replaced. They were replaced with gate valves

'that are not susceptible to the failure mode of valve 2N61-F588D. The gate valves' disc and stem are one piece; therefore, sensing line vibration can not cause the disc to rotate and wear any disc retaining parts as was the case with the failed valve. The valves also were installed upside down so any catastrophic failure of the disc / stem will not result in tne disc falling into and isolating the sensing line. Additionally, the disc'was removed from root isolation valves 2B21-R462A and B, and 2N61-F009 and F010 to prevent any sensing line vibration from causing the disc to separate and isolate its vacuum sensing line. Two isolation valves still remain in each sensing line, i.e., the new gate valve and the instrument isolation valve. (This latter valve is located on the instrument rack and is not subject to effects from sensing line vibration.) Design Change Request (DCR) 2H90-003 was developed and approved, in accordance with plant administrative controls, to allow for these changes to the valves in the condenser vacuum sensing piping. The new valves were installed and the discs removed in the other isolation valves on 1/14/90.

The vacuum sensing lines (3/8 inch stainless tubing) were reconfigured such that each of the four sensing lines off the condenser now has one Group 1 isolation logic vacuum switch and one turbine trip logic vac.uum switch. The new arrangement is single failure proof and prevents spurious trips due to single failures. The sensing lines were reconfigured on 1/14/90. Design Change Request 2H40-003 alco provided for

'reconfigun tion of the vacuum sensing lines.

The thermal overload relays in the motor's local starter for HPCI injection valve 2E41-F006 were removed and replaced under a Maintenance Work . Order. The valve was functionally tested using the Motor Actuator Characterizer (MAC) test equipment to ensure the torque switch settings were correct and the valve was functioning properly.

As-found torque switch settings were acceptable and the valve stroked open and closed properly. HPCI was declared operable and returned to service at approximately 1950 CST on 1/14/90. Engineering continues to investigate the cause(s) of the heater strip

. fail ure. Further corrective actions will be taken, as necessary, based on the results of the investigation. An update to this LER detailing the cause(s) of the heater strip failure and corrective actions taken will be submitted by approximately 4/1/90.

The Unit 1 vacuum sensing lines, root isolation valves, and Group 1 isolation and turbine trip logics were examined to determine if similar problems existed. It was found that the root isolation valves, and the sensing line and logic arrangements were f the same as Unit 2. Unit 1, therefore, is vulnerable to a spurious reactor scram in the event of certain failures in any one vacuum sensing line. Design Change Request lH90-009 has been generated to implement a change to Unit 1 similar to the change implemented on Unit 2. The Unit 1 design change will be implemented prior to startup i from the Unit 1 Refueling Outage currently scheduled to begin 2/17/90.

L 1

l (L

g,o= m. u cm,n.e * -

r

'I lenC Polm 3 Bed :' U.S. NUCLE A3 s.EIUt? TORY ColusMISSION I* > ? * /

. LICENSEE EVENT REPORT (LER) TEXT CONTINUATION APPaovio oMe No siso-oio4 h: EXPIRE S. t!31/98 FacettTY NAME til DOCKlf NUMB 8R Gl .- LER NUI/SER 18) PAOS (3) vtAR My ',^ 4 I "y 58,y,"  ;

(

PLANT HATCH, UNIT 2 o is lo l0 lo l 3l6 l 6 9 10

-~ 0l Ol l:

0 10 0l8 oF Ql8 TEXT t# muere guese e soeweg use sWeseens/ NRC pone 354's)(th ADDITIONAL INFORMATION

[ .1.. Previous Similar Events:

4 There was one similar event in which the reactor scrammed due to MSIVs less than-90% open.- This event was: reported in LER 50-321/1988-009, dated 6/20/68.- In that event, the MSIVs drifted closed from loss of air due to an incorrect instrument-air system valve lineup. Corrective actions taken for that event would not have

prevented the event described in this LER because the causes of the MSIVs closing-are different.

2. Failed Components Identification:

a. Master Parts List Number: 2N61-F588D Manufacturer: Hancock Root Cause Code: X l Model . Number: 5500W- EIIS Component Code: RTV I Type: Root Isolation Valve l Manufacturer Code: H037 i EIIS System Code: DL

.l Reportable to NPRDS: No j

'3 r _ b. . - Master Parts List Number: None Manufacturer: General Electric Root Cause Code: X F Model Number: CR 124LO28 EIIS Component Code: RLY Type: Thermal 0verload Relay Manufacturer Code: G080 EIIS System Code: BG Reportable to NPRDS: Yes

3. Other Affected Equipment:

L L No systems other than the RPS, LLS,' PCIS, and HPCI were affected by this event.

l l

l l -

N7.C PORM 3eeA. 'Y 8. GPor 1,68-520-589:00070

, G43) ' .

_