ML20197J368

Forwards ORNL Rept Evaluating Significance of Failures & Degradations During 840920 Facility Startup.Method of Analysis Compared to Reliability & Risk Assessment Branch Rept.W/O Encl

ML20197J368
Person / Time
Site:	Trojan File:Portland General Electric icon.png
Issue date:	03/05/1985
From:	Burdick G NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
To:	Thadani A Office of Nuclear Reactor Regulation
References
NUDOCS 8503120576
Download: ML20197J368 (17)

Text

'

V};.,J,11 n lu,,,am,:o. gam 2oi.cm

}

(,f,y

^ ^

_[ ;

pg , 3 i.3. ' , , ,

, . 1, HEMORANDUM FOR: Ashok Thadani, Chief # '" -

Reliability and Risk Assessnent' ttranctr jM .

" - ~ ~ ~ ~ ~

Division of Safety TechnologV Office nf Nuclear Reactor Regulation FROM: Gary R. Burdick, Chief Reactor Risk Branch Division of Risk Analysis and Operatinns Office of Nuclear Regulatory Research SU3 JECT: RESULTS OF AN EVALUATION OF THE SIGNIFICANCE OF TliE PULTIPLE FAILURES OCCURRING AT THE TROJAN NilClEAR POWER PLANT WHILE UNDERGOING STARTUP SEPTEM3ER 20,19M Attached for your infornation is an ORNL report cvaluating the significance of a series af failures and degradations occurring during startup of the Trof an Nuclerr Plant en September 20, 1984. This evaluation was undertaken as a result of a suggestion by a member of the precursor program research review aroun that this sequerce of events could represent an important precursor.

The failures and degradations at Trojan included control room operator errors, maintr nance errors, and eauipment failures er degradations. Examples of the naintenance errors and equipment failures as described in the LER include:

"iscalibration of several time delay relays in the control circuitry for the diesel-driven AFUS pump. These relays were not covered by maintenance or celibration procedures.

The turbine-driven AFWS pump had a failed suction pressure transmit-ter arid insufficient snubbino of sensing lines causing control oscil-lations.

An enerqcncy power DG failed to start dua to its crankesse pressure switch being tripped. The LER notes that while this trip would nornally be bypassed--the present configuration would only have provided bypass af ter 15 seconds of DG operation, hence the tripped pressure switch would have precluded DG start even under emergency conditions.

The "A" nain steam isolation valve bypass valve was found groundid.

, N. ,

r------ - - -

T-

• "**~'* 8503120576 850305 Nj RES SUBJ R2912.01 l

ue.... .,,.,...n- OFFICI A L RECORD COPY

Ashok Thadant

• The no':L pr.acursar analysis nothnd evaluates aper 3+inval i"cidante urino ona nr mre of fcur initiating event typ.as and encrosonndino event trees dep ct iein i possiblo nutennat niven the initiatian avaats. The i ni t i a t ir.n avante co.siderid i.re:

Srell LOCf. initistnr I0F" ini'ia*nr Steari li'a brack iiitiator LonP initia+nr.

The ihnve tupra of ini+iators have $?cr. shown b> nast PR?s te contributa 'he mst tc ccre trl+ frecurncy. Thesa can cnntribute the rest because (al thr v are the erst frequer.tiv occurrinq iritiators, nr (h) +ht availabla safcquards rouipnent tn nitigate these initiaters can havn a ralativalv hioh probabili+v of r ailira. L'hile the event trees for 311 er the ahnvn initiaters are to scue degree effected bv the Tro.ian incident, only twn evant trees are significant1v affected by the reported failures or degradations. Those .'re the event 'roas for the LOOP ard LOFlf initiators.

Mter OR4L had counleted and suhnitted the attachad analysis, and whila we were saakinc additional background it.foraation on these even+s, we learned that P.P.A4 had also inslyzed these events. In comparino the two analyses, we nnte that both .'nal/scs used sinilar postulated initiator events to evalucte the Troian events. The ':P.AB analvsis deesn't usa event trees as such, but does previde an equive' ant are. lysis hv discussing two nost probabla scenario types which thev consider cnuld result from the sequence of equipment 'ailures or degradations described in thn LEP. Tho OPliL analysis postulates generally similar types of scanarios as RRA4 in their evaluation.

The nr:L and RRA3 a.nalvses have several siqnificant differences in the equip-ev.p t fnilura rates used and in thr. assumptions regarding possible recovary of failed or degraded coulomnt. However, in addition to these differences, tScro are also dif forarces of the analvsis nethods as perferned by ORfit and P. RAP..

"c Mt nvaluated the likalibcod that additional failures or actions could have nceurred in a enetinuous sequence at the time of the cbsarved failures result-irn ia

• h r- scen rin + hey costols+ed. In contrcst, ODW.'s analvsis indica +rs

' hee cansider it was nniv (cr+uitcus tha' *hase v.*rinus 04uinmnt failurM.

danrar9tinns, er m18d.justncnts wore disenvarad at startun. !n nther words, b.'d ' he cperator performed r nere centrolled and correct startup, +he saf a+ w equipnant wouldn't have been challenced and benea the degrsdations and failures likolu wnuldn't have been discovered. OP.ftL's earlysis assesses that situation t.nd the pntantial vulncrahility of the plant to initiatina events which could occur in the 1-mnth interval post startup until subscouent equiprannt surveil-lance tasts wro performad. The CRr:L anal" sis assums thesa problems would have r.11 been dotected imd ccrrected et the next surveillance test. ORf L I i i I i

"" m > .j. ..l.....

"N' .

.l... ..... . . .. ... .

...... .... . ........... ............. ..*.**.**a .. .....******* **.*******. .*..****

Ashok Thadani 3 notes, howevcc, that if this assurvtion is wrnoa then +hrir estimatiurs o' event seriousness cc.uld br- nnne >rser. a tiva. .

In enryaring the twr, ana lvsrs, m. r,ote that the RRAC auar,*itrtive assessneat dif'ars bv acpropinstely an nrder of mgritudo frna the ORNL snalysis. T_iig RRA3 cntditinnal probr.bility of core dar:aon associated with +his event was 10 nr less. The att 1. reno' + assestas this serins of ev~ ts as having .*n anproxinatolv 10'fched 00':

condi+1onsl pre.hability of core dwqc. The ..r.t.achrd +abins 6: this renn surmrize thr diterences batween the two a n a l vses .

ISI Gary R. Curdick, Chief Reactor Risk Branch Division o' Risk Analysis and operations Office of Nuclear Rrc,UlatorV 90scarch Attachrxn+s:

As stated cc: C. Trarr.;11 G. Edison G. Ilnlahan 0 '<arrr.t t D. 7tonann K. Seyfrit II. Sharcn Distribution:

RES Central File - R-2912.01.04 CIRC /CHRON RRB RDG/SUBJ Manning Reading

[ Baranowsky Reading l Burdick Reading l

1 i L n b L i

. . .u y, i RRB,: D,RAO:RE,5 nR,B,j L.S I C DRAO:RE5

,ky, ,G. ,k, . . .

pd,,..f.M.an.n.i.n9.:cs,h,,P,BJ,)!.85 o~"> . 3!. . 3!!!!.85. . .

m ,,,,.o,

- o,w- .i.3!.5!.85$W-

~ ~

o .

T 4

.\

Loss of Main Feedwater Evaluation RRAB ORNL i

Postulates loss of diesel driven AFW pumps (0.5 Pr.), Postulates AFW & secondary heat removal failure loss of turbine-driven AFW pumps (0.5 Pr.), loss of (0.34 Pr.), failure to implement feed and biced non-safety grade AFW pump (0.01 to 0.1 Pr.), failure (0.34 Pr.), challenged by one-half month's to depressurize and utilize main feedwater pump expected LOFW initiators (0.005 to 0.05 Pr.), and failure to implement feed and bleed (0.01 to 0.1 Pr.) Conditional Pr = 1.4E-3 Conditional Pr = 1.3E-7 to 1.3E-4 Rationale Rationale Evaluation is based on actual or potential failures Evaluation is based on reported failed or of equipment that was or could have been cal'ied degraded safety equipment for estimating upon during this incident. plant susceptability to loss of feedwater initiator events dering a 1-month Takes credit for the possibility of secondary assumed susceptability. This evaluation system depressurization and using condensate assumes the failed or degraded equipment would pumps for steam generator feed. have been caught during the next monthly test.

Credits both condensate recovery and feed and bleed as unrelated operations The probability of failure to depressurize and use condensate pumps and failure to accomplish feed and bleed cannot be quantified as totally independent operator actions, they are correlated by available time.

LOOP Evaluation RRAB ORNL 4

Postulates that plant trip causes grid loss Two sequences postulated.

(1E-3 Pr.) and the one remaining emergency power DG fails (1E-2 Pr.) - Degraded emergency DG operates (0.93 Pr.),

degraded AFWS fails (0.34 Pr.), HPI fails '

Conditional Pr = 1E-5 (0.34 Pr.), challenged by one-half month's expected LOOPS.

- Degraded emergency DG fails (0.071 Pr.),

degraded AFWS consequently failed (1.0 Pr.),

challenged by one-half month's expected LOOPS.

Conditional Pr = 2E-4 Rationale Rationale Evaluation based on assumption that the plant trip Evaluation is based on reported failed or degraded occurring in this incident could upset and cause safety equipment for estinating plant susceptability grid loss and the probability that the one to loss of offsite power (LOOP) initiators during a operable diesel generator fails. 1 month assumed susceptability. This evaluation assumes the failed or degraded equipment (including miscalibrated sensors) would have been caught during the next monthly test. ,

1 i

post omct nox v CAK RfDGE NATIONAL LABORATORY OAm MIDGE.TENNEsstt 37sai OptmAf tD 97 ManTm uAmitTTA tutmGv SYSTEMS WC October 30, 1984 Mr. Frederick M. Manning Division of Risk Analysis and Operations U.S. Nuclear Regulatory Commission 5650 Nicholson Lane Rockville, Maryland 20852

Dear Fred:

ASP Assessment of September 20, 1984 Trojan Event You requested we take a " quick lock" at the September 20, 1984 reactor trip event at Trojan and assess its impact using the existing ASP meth-odology and data base described in our 1980-81 report,NUREG/CR-3591.

This event would have ranked well within the upper one-third of 1980-81 ranked events, had it occurred in that time frame.

Some notes on the event analysis assumptions and results are attached, along with a package of standard ASP documentation for the event. I hope you will find the information useful. If you have further ques-tions, please feel free to call Joe Minarick at FTS 626-4671 or 615-482-9031.

Sincerely yours, l

M __

J. R. Buchanan, Director Nuclear Operations Analysis Center JRB:JWM:ap Attachments:

As noted cc/att: Richard Barrett, NRC

~

TROJAN REACTOR TRANSIENT f On September 20, 1984, Trojan experienced a reactor trip and spurious safety injection during startup. Numerous components failed to operate correctly, including both auxiliary feedwater pumps, a diesel generator, and the main steam isolation valves. Event details are documented in the attached ASP Precursor Description and Categorization sheets and event trees.

The likelihood of potential severe core damage was estimated for the event based on the existing ASP methodology. The actual e>ent occurred at low power during a startup following refueling. Because of the lack of decay heat, substantial time would have existed during the actual event, if core cooling had been initially unavailable, before core damage could have occurred.

For this assessment, instead of incorporating the observed teactor trip and inadvertant safety injection (with subsequent loss of feedwater) into the event model, the assumption was made that the next test which would have identified the inoperable equipment would occur approximately one-half month after startup (in actuality, some of the component failures may not have been detected during normal surveillance testing). A potential loss of feedwater (LOFW), loss of offsite power (LOOP), small break LOCA and steam line break were then postulated to occur in the 15-day period

~

with a probability determined from the industry-average initiating event frequencies developed in NUREG/CR-3591.

Two conditional probabilities of potential severe core damage were estimated for the event. These estimates utilized different sets of re-covery likelihoods for the observed failed equipment to provide information on the sensitivity of the results to differing recovery assumptions, as follows:

Estimate 1 e Auxiliary Feedwater - recovery class R2 (failure appeared recoverable at the failed equipment in required period following the assumed initiating event; recovery from the control room did not appear possible

- probability of failing to recover = 0.34).

e MSIV closure - recovery class R1 (failure did not appear recoverable in the required period following the assumed initiating event, either from

the control room or at the failed equipment - probability of failing to recover = 0.58).

e Degraded Emergency Power System - recovery class R1 for unavailable diesel generator.

Estimate 2 o Auxiliary Feedwater - recovery class R3 (f ailure appeared-recoverable in the required period following the assumed initiating event from the control room, but the recovery was not routine - probability of failing to recover = 0.12).

e MSIV closure - recovery class R3.

e Degraded Emergency Power System - recovery class R1 for the unavailable diesel generator.

~

Utilizing the above sets of recovery assumptions resulted in the following core damage probability estimates:

Estimate 1: PSCD = 1.7E-3

~

Estimate 2: PSCD "

In both cases, the dominant contribution resulted from LOFW sequences.

., followed by those for LOOP. Contributions from main steam line break sequences (those impacted by the MSIV failures) were an order of magnitude below the contribution from LOOP sequences.

Based on the above estimates, the Trojan event can be ranked with respect to other precursor events. These two estimates would rank the Trojan event either fifth or eighth among 1980-1981 precursors, had it occurred in that time period.

A review of precursors which occurred at Trojan between criticality and 1981 was also made. Three precursors had been identified at Trojan.

Two of these involved multiple MSIV failures (on April 14, 1979 and April 11, 1980).

PRECURSOR DESCRIPTION AND DATA NSIC Accession Number:

Date:

Title:

Loss of Feedwater plus Safety System Failures at Trojan The failure sequence was:

See next page.

4 Corrective action:

Repairs were made after cooldown to cold shutdown.

Design purpose of failed system or component:

(a) the AFW pumps provide SG cooling on loss of main feedwater (b) the diesel generators provide ac power on loss of offsite power (c) the MSIVs close to prevent excessive cooldown in the SG/RCS and to prevent excessive inventory loss from the SGs l

1 l

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

1. At 8:18 a.m. on September 20, 1984, Trojan was at 15% full power in the process of restart after a lengthy refueling outage (%S months).

The operators were in the process of paralleling the pain turbine generator to the grid.

2. It is believed a faulty load controller or steam bypass system problem in the steam generator (SG) secondary side was responsible for a high-high steamline flow signal coincident with a low-low SG T,y signal or a low SG pressure signal which caused a turbine trip. The turbine power level was higher than the operators believed due to a stuck power recorder (MW).

3. These signals plus the turbine trip then resulted in a reactor trip from 15% of full power. The reactor trip signal was initiated by an inter-l mediate range high flux monitor (IRM). A safety injection (SI) by the ECCS subsequently occurred. The IRM was not supposed to cause a trip prior to exceeding 25% of full power. The operator was not aware of the IRM setpoint. Also, the rod blocks had not been reset to a lower level

- and the trips had actually been set to 35%. The ECCS was initiated either by the SG high flow and low temperature signals or by the low pressure signals. The SI signal only lasted for one cycle or less l

(16 as) but this time was sufficient to cause actuation.

4. Following the plant trip, several safety systems failed to perform as required.

(a) the diesel driven auxiliary feedwater pump (1 of 2 ESF AFW pu=ps) failed to start on demand. The failure was due to an incorrect zero time delay on the low oil pressure relay (normal delay time in 25 sec. to allow the normal operating condition of 17 psi at 600 rpm to develop). The relay tripped immediately upon the

AFW diesel start signal since the oil pressure is initially low at startup and requires time to reach its normal operating value.

The surveillance test would not have detected the incorrect setting. A high crankcase pressure signal failure may also have

~

been involved.

(b) The steam driven auxiliary feedwater pump (the second ESF AFW pump) initially started but tripped after 7 minutes due to a failed low suction pressure transmitter. This occurrence created a total loss of the safety-grade AFW pumps. The pump was manually

~

restarted but exhibited erratic behavior, i.e., oscillating flow, and was shut off for unsatisfactory performance. The oscillations were due to a misadjustment in the discharge pressure transmitter snubber. A third motor driven non-safety-grade startup pump was started and provided adequate flow to the SGs. The pump is not normally connected to the ESF buses.

(c) The "A" emergency diesel generator did not autostart and could not be started manually due to failure of the high crankcase pressure switch. The "B" DG was available.

(d) The MSlVs in all four (4) steam lines failed to close on demand.

The SI close signal (16 ns) was of too short a duration to actuate the system. The valves were manually closed.

5. The plant was placed in a cold shutdown condition, i

l 1

6

• Plant .et 15%

! power ase.I Inad controller /

steam Iyp. ass rr en results in inarivurtent RT/Lufv H5IVs Alv ECCS* EPS* I'" * ""

• I

• I RT and Safety CI"*** Scscre Injectfon J. -

e ..re ti eme ce

-g (30 dk(DI 1

g 8:o

)k / 5, Ah M*

W cS U  :

S (D)

- too V Y ar 1/ ,/////// /1 rvrau r  ;;nse

o* *

oeee

> ATWS 20***

Actual occurrence for loss of m:in feedwater and safety system fattures at Trojan

• Response not required to mitigate actual initiating event. .

• • Bleed and feed assuecd effective for decay heat removal.

i **arlant in startup from refueling and core decay heat level inw. Longer tlian .everace t Imc re r to.1 av.a!!able for recovery.

g e _

c n I 1 3 4 5 6 7 a 9 0 1 2 1 _

e 1 1 1 1 u

q .

eo sN l

a i

t e e m nr eeea g

s s s e

s s s s t t vr m o o e e o o e e e. o e e e s oeoa N N Y Y N M Y Y R N Y Y Y y PSCD s

y mr t e g e T n 4- 4- 4 f

i a

gel nr E e e s oooo 6 s. c.

LCC d g 2 2 s n a

n eo r ri ut 4 e sc t hse 4- - a gej G t e w irn

• e.

NPi 3'. d .

L

o. c e e

f es l e n ol i reae a ol Vr 3 m u 3 -

TYns - G RR oo s 9 f OOil o PPtC A 2 2 s d

s

. e o d l n

Ya Rm r

O e PD

+

0 f o

- t d - s a

n l e

en ra

- v ej yryo

_ r erm t o et a e nr l edR l vn i T I

idot s ece ueee AFSM (

A 4 V s$,' 5 s f t oa 3

.js.'e*  : :, '

es ce r nr t

o eu cy ul al 5 qi rr - e a tY E Sf

/. l 3 3 r -

e e -

G G f

o t

a L *e 4 T w h L m end sie yo - m oee l MF f &1

s. 1 P i T

S

. E 1 1i! 1 I  ! I

!' i

d . .

i 8.oss of Tiit tilne [mer- Auulliary PORY Ptev er Illah I.on g rotential $cquence offette Cenerator gency Feedwater and Demanded PonV isola- Pressure Tere Scecre No.

Power Ituns 8.sch Power Secondary tion Valev Injection Core Core 9 1 and Assumes Meat Removal Closure Cooling Damage House Lo. ids I

No 1 Na 2 No )

04 1 (CD) j (to) Yes 4 2.iG-3 3E-3 Yes

.ot75/ y c 3.I G -3 5

he %O bars No 6

l l No 7 r '

(F)

rge,)

/,

' (0)  ;

s Yes 8

/,

,p I.36-3

'I Yes 9

+ 3*I No 10

+

.o4 y,, gg i 1.9G-3 No \1 7.ie-a.  !  !(F)

- Yes 13 I'*

P w 1.on E- 4 EsTsenTe i Sequence of interest for a postulated LOOP for the loss of main feedwater and safety system failures at Trojan 9 6

s Small Reactor Auxiliary Illgte Low Pressure Potential Sequence 1,0CA Trip Feedwater and Pressure Recirculation Severe No.

l ,

Secondary injection and LPR/IIPI Core llent Removal Croiss-Camnect Dimar,e t

4 No 1 2.LE-4 Yes 3 4.0E-4 .

No 4 f.9E-3/y Yes 5 Q 3a w 1.ss-4 Yes 6

' '#~

3.4.s-s 'I(F) 7 s -- Y e s 1

P: 4.SE-9 I

]

ESTam4TE l i

Sequence of interest for a postulated small break LOCA for the loss of main feedwater and safety system failures at Trojan l

g I

e n ' s s e ,* t 3 5

• n 9 :s I l s 5 '  ? t f ' I i i u

g I g l i t 3 M s.,

l a

i t e e

• nr g s s s s eeea t vr m o s

o oN o s

e s

e o o

s e

e o " e e e s oeoa PSCD M g Y e

N N Y v h N T' N s y N ' i T Y o

l g e a h l 4 A 4 t gsl e ro 4

e s n eeo s c. ra LTC s. L. s. j2 oj

c. z I3 z f o II 2 r e ' kT r

vs u

5 a

eo 3 et 0l e ra PC g o. b 0 C s 0

6 ee d

e nr u i u don ll eTi i vn aeent 0 puer t

8 ma af POliCf a 8 e 0 0 t s J

sei e t td d s ai ey drcn et A o 4 t s rn i 4 - a iect ucii

- t 1 y qnrJ t s t s t eooJ 3 '. t e s ;a B A f B sf oa ps n

• eo ri ad ut s ci \*

n e sei tgejrt t - 43 ra en rns - i G o s PI ( o. f r u S.

L e

. c tt

,r sa r e et ew rd le a. ee s d t e AF s e ue M

[

(,

i nf a

r M. e$ h' * .

n rn e fi d ood 3, oa e tt e 4 m r

i era c at t

) e e sel e p cf penof .

no et esf psGI E

( s p e

's r s' , ' , / s.

' u

/ '

q r 5 e o - 1 S t e c e G ar e c (- E T

B S 1 n

n o eea vne k

%,M t.

m e

TeLti r ee t T S

iL E

'1

9 e# ,

f o-e S

v

• N M. d e d N e P O = M

, - - N.

F

• b O ME

.e WW I e *-

• E o

== = > =.e e a 0 0 e

e

• O O e

t e

t O' O t

t t

O W Q 4 > > M *~

8.muO E E E E > > E E > > > h 1 l e g x a

?

Q Eka T T. T. a OO g 9 w a 4

S' u u 4 $ 5

__g n s

'g eO m= b 3 ** W ev U T.

m$E

- oe

• C ska o f 7 3

y e e w -

9 M o7 C

. "S *4 -r w 0-> c 8

m. m. u hO fge:c W w.

W W

O a

1 .

. . O

,1 ~

u o .

, a e *8 O 4 e5

>> h wc

.t. e* ,I -1 ME "O C b mc

.e 4 0 ** a

" ("* m. wa 5 ($m2 m w $ o=

_i Ns s i .aA\ \W s th

. . m -- .

D'

• g g y

6 0 C w'J

• • & O e= tfl 3d t C" *4 25 W WC J (/) W 4

= -r

• is O e i ' ;= J e*S e F

.3 i a  %*

e eg p$.

m

o. g y

4

, -- e, _. - , g n-__.--- - , . , - - - - - -, -.

d .

j -

I. css of Tier b ine Emer- Ava i l l.s ry roav riiety or Stich I.nn a retential Sequence Offsite Generator gency Feedwater and Demanded PoRV leala- Pressure Term Severe No.

Power Runs Back Power Surondary tien Valve Injection Core Core 9 and Assumes Neat Osmoval Closure fooling Damage Itause Lo.edo l No I 1

Na 1 i

I No )

ong 1 F(CD)

$ kCD) Yes 4

• 1.M-3 838*3 N7 k M o hr3 5.lE-3 Yes 3 No 6 I d p

No 1 jr l -(F)

.  :(c ,)

,' (D) j  ;

Yes 8

' / s ,

q, ;g l3G-3 Y Yes 9

• 3*%

i No 10 foq Yes 11

! . gg,3 No 11 1le-2  !!(F) fee

- I) 80 P 1.as .4 ESTtmmre 2.

Sequence of interest. for a postulated LOOP for the loss of main I

' feedwater and safety system failures at Trojan l

1

' +

1 i

1 1

9 1 ,

l Auxiliary lii gh Low Pressure Potential Sequence i

Small Reactor Recirculation Severe No.

i LOCA Trip Feedwater and Pres sure i Secondary injection and LPR/IIPI Core Cross-Connect Damar,e IIen t Removal i

4

.I i

-No I Yes 2 2.s e-4 Yes 3 L.oc-4 ,

No 4

~

\

190-3 /gr Y"5 5

' for 144 W ,,g ,,4 I

i Yes 6

- see-4 3.6s-5 t i(F) j I ' Yes 7 52.

I P s 1.&G-9 i EsTimais 7.

i i

i I

( .

Sequence of intereet. for a postulated small break LOCA for the loss of main feedwater and safety system failures at Trojan i

t s

e n , n s s . - s s.

, y s s - t i

e. , '- t a s , ,

i i i i i s i i i p

e.,

ss l

a .

i t e e nr g eeea , s

• s s s s s e

e s

s o , e o e e s t vr m oeoa PSCD uM ve n o w ,,, y e

1 e . oN u

e v s 3 Y N v s ) v o

l g - e n .

h i -

9 9- t gei nr o 9- 1- - n oeo 9 e. e ra tTC e 6 t. II t.

. oj

s. gI 2

I3 a a f o 13 t r e kT r

. u a vs 3 et eo 3 - ra 0i PC -

e b o o s g

e.. c ee d

e nr u iu don ll eTi i Y n R eent t

ma af O pb o e s.

FOICR a. o e o t m d

se t

e t d ds ai ey drcn et A o t s rn i 4- a .

i ect uci6 4- l y qnrJ c e ut eooJ pCBA s. i. t e sf e s oa n

1l ps eo ri ad ut sc) n h sel geje 4- 4- ra e rn s s 4 o .

MPI H 4 3 f r

e. 3 C

o. liI e

t. tt

) r sa re et ew rd ls a. ee aJ m e ue h A I

t e nf AF [ I i

,e f r I d,,:;h I

(,

n s -

- d rn ood I.

e fi oa e ti e 1 m r

a mra c at t

)

n e

. a cf uael ienof g

e F P no et esi ( s e

_ NSGit , , r , :

u a v'e '-ss t.

q r e o 2 S t m c e 5-ar ec RS s s.

t.

e 3 T

- b a ees ene h

/E To m i

s a r _- ' i Ttt s

'0 h 1 E

. ~

i

,e

. \

CATEGORIZATION OF ACCIDENT SEQUENCE PRECURSORS NSIC ACCESSION NUMBER:

LER NO:

DATE OF LER:

DATE OF EVEN1: September 20, 1984 SYSTD1 INVOLVED: Main feedwater, AWS, EPS, MSIVs COMPONENT INVOLVED: MSIVs, IRM, A NPs, DG, SG Instr.

CAUSE: Equipment failures SEQUENCE OF INTEREST: LOOP, LO W , MSLB, LOCA ACTUAL OCCURRENCE: LOW REACTOR NAME: Trojan DOCKET NUMBER: 344 REACTOR TYPE: PWR DESIGN ELECTRICAL RATING: 1130 MWe REACTOR AGE: 8.7 years

- VENDOR: Westinghouse ARCHITECT-ENGINEERS: Bechtel OPERATORS: Portland General Electric LOCATION: 30 mi NW of Portland, Oregon" DURATION! NA PLANT OPERATING CONDITION: Startup after refueling outage (15% of full power)

TYPE OF FAILURE: inadequate performance; b failed to start; (c) made inoperable;

(d)

DISCOVERY METHOD: Operational event COMMENT:

s

- _ . - -- - . - - - , _, _ , , , ,_y__,,,_r__,_,,  %-__.,.__ _ _ . .y-. ___., , , , _ ~ ,,.-. _ _.. ._-.-- , m-m-____-,-~_

)

)

}

i f

/ .

f&cv a &z

/ (,llA A j' Q~.irIC ci g 27dC

.