ML20085C433

Trip Rept:Onsite Analysis of Human Factors of Event at Monticello on June 6,1991

ML20085C433
Person / Time
Site:	Monticello
Issue date:	06/06/1991
From:	Meyer O EG&G IDAHO, INC.
To:	NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD)
Shared Package
ML20085C436	List: ML20085C433
References
EGG-HFRU-9804, NUDOCS 9110010201
Download: ML20085C433 (22)

Text

,*

b CC I

[GG-lif RV-9804 1 RIP REPORT:

ONSITE ANALYSIS Of TliE liUMAN FACTORS OF AN EVENT AT MONTICELLO ON JUNE 6, 1991 (HI-H1 IRM SCRAM)

Orville Meyer-Onsite Team:

iJohn Kauffman, NRC/AEOD

• George Lanik, NRC/AE00 Orville Meyer; lNEL

• Team Leader Published August 1991 1'Y A

_s a

A - -4 c ; A 4,. ;; ;

IdahoN'atiodilEngineeringLab' oratory:

X, lEG&G'Idahoi:Inc.?,-'s

~ 1 P. O. Box 1625 Idaho Falls, ID 83415

' ga tu, ;+

s Prepared for the Office for the Analysis and Evaluation of Operational Data

- U.S. Nuclear Regulatory Commission,

Washirigton, D.C.

20555 Under DOE Contract No. DE-AC07-761001570

..-~~~n

(,9//coloaol-m m

\\

Q9au + n 1

l EXECUllVE

SUMMARY

The Office for the Analysis and Evaluation of Operational Data (AEOD) of the U.S. Nuclear Regulatory Commission (NRC) has initiated a program to study human performance during operating events. As part of this program AE00 formed a team to conduct an onsite analysis of an event at the Monticello nuclear generating station. A Hi-Hi intermediate range monitor scram occurred while control rods nere being inserted to shut down the reactor from 6% power.

The reactor power decrease left the decay heat unable to sustain the small, miscellaneous steam loads that were online.

The resultant reactor cooldown of approximatelyl3-1/2*F per minute. caused an unexpected increase in reactor power and a Hi-Hi scram on range 7 of the IRMs before the operator could switch to range 8.

The study of human performance focused on the actions of control room operators during the period leading up to and following the Hi-Hi scram,- The study was based on data from discussions, plant logs, the statian's scram report, and interviews with control room operators and other station staff.

Idaho Natienal Engineering Laboratory provided assistance as part of the AE00 program to study human performance during operating events.

The reactor had returned to power operation on June 5,1991, after a refueling outage but was shut down by a spurious scram at 1:47 p.m. the same day. The reactor was restarted at approximately 10:00 a.m. June 6 and at 3:00 p.m. was holding at 6% power,= 930 psig, to perform a calibration of the IRM

. r' nuclear instrument channels.'

n i

The, temperature dfl tlie; tailpipe' fr'omithe'H safety relief valve began to.

increase 'during the reactor heatup, which 'iridicated a small leak'across the seat of tiie safety relief valve. The temperature continued to rise while the pressure was'being held at' 930' psig for the:lRM calibration and exceeded e

170* F. The temperature was approaching 200*F, where the plant's operating procedures call for taking the plant to cold shutdown. The manufacturer's data states that the safety relief valve may fail open at 230*F.

iii

l The plant management decided to shut down the plant at the completion of the intermediate range monitor calibration and replace the H safety relief valve.

Management and the control room operators were aware and concerned about the probability and consequences of a failure of the H safety relief valve. At the completion of the intermeciate range monitor calibration the shift manager directed the operator at the controls to shut down the reactor without delay.

The shift manager provided no amplifying instructions to the operator at the controls and there was no briefing of the control room l

operators on the plant status or on the plan of action.

1 The operator at the controls began promptly to notch insert the control rods.

He did not observe that when the reactor power was decreased below the I'

level at which all turbine bypass valves (TBVs) were shut that the reactor pressure was now decreasing because the reactor power was below the level j

required to sustain the miscellaneous steam loads. _ Below this power level tha reactor power would decrease after each rod notch insertion and then-start to increase due to the reactor cooldown. -The operator did not realize that tho' tendency for power to increase was due to the cooldown.

Neither the shift i'

manager nor the lead reactor operator were overseeing the OACT or mor.itoring reactor pressure. The operator at the con'.rols continued to notch insert control rods for 20 minutes until the main steam line low pressure alarm was received.

At that time the operator at the control stopped control ro.1 insertion to check the reactor cooldown rate. Reactor cooldown continued to add reactivity; andtheresultingpowerincNasereachedtheHikHi(RMtripM-The~mainstdas isolation valves were subsequently closed to limit reactor cooldown and the reactor conditions werkstab111z6d.

4 g-The operational strategy chosen by the shift manager for reactor shutdown was to notch insert controlerods rather than to go to shutdown on the mode-switch.

This required monitoring and control of any reactor cooldown and compensating for its effect on reactivity. There was a low level of awareness of this requirement on the part of the shift managor, the lead reactor cperator, and operator at the controls, The lack of a ciew briefing on the plant status and plan of action contributed to this low level of task iv l.

2

f awareness.

Since neither the shif t manager or the lead reactor operator vere monitoring reactor conditions, the lack of monitoring and control of the reactor pressure by the og rator at the control continued until the main steam line low pressure alarm was reached. At that point insufficient time remained to avoid an inter.nediate range monitor Hi-Hi scram.

i 4

?

\\

.) L i

f 4

I

1. k V

ACKt40WLEDGEMEtnS We express appreciation to the Monticello staff for their cooperation in providing the information necessary to analyze the human performance during the operating event. We particularly thank the operators who were on duty during the event for their cooperation during the interviews.

h 1

.]

4 Ii r

U b.l,

Y'_;'$ ; l ?l -

!?

l 4

?

-~b

g,

.jf i _ _

+

. -l.4_', ! ',;.

, j j.,

e

{$1:

d k_\\ n jy -

~

p yl? !; y

'.4g Qgyp pp a -

as 1'

- apa.

s

, c.7, J

.g : :

y

$ k(' '

h9

'I i

+

A

4ik,

).' c

• 16 i

4

,e vi

-.-.r CONTENTS-

- EXECUilVE

SUMMARY

111 ACKNOWLEDGEMENTS vi ACRONYMS

..............-..............~.... ylli 1.

INTRODUCTION

..............e.............

I-e 1.1 Purpose.:.r.

1 1.2 Scope.................-...............

.1 1.3 On s i t e An aly s i s.... _.. _.,. - _. _-........ : '

3 L. r z. t..

1 2.

DESCRIPTION 0F -THE EVENT ANALYSIS............:..... '.

2 2.1 Background--............................

2-2.2. Timeltide of the Event i

t

. x. g.y..t.w.f.;.l.g L.n.g i. v.g.;

. 3i ; w

..4 m. 2

..~,.,y s

t 2.3 Analysis 1,...,.... >... e. VM...... J. ~. -.. -. N R.-

101 W v

x 2.3.1-Operational._ Strategy for Reactor Shutdown 10 2.3. 2 Task Awarenes s......... -............

11 2.3.3 Comand,: Control.. and Comunications..........

12 2.3.4 Stress Management 13 a

2.3.5 Man. machine Interface 13 L

2.3,6: Training......-.-..~........,.._........

13 l'

2.3.7. Event Evaluation...-. s -... -..-..:..

.:...t.e.

14 4

~ 3.

SUMMARY

OF HUMAN FACTORS OF THE EVENT s :.

.~.e.,.,..

.s.c.

15'-

.3

.7,

a a

.y L

dFIGURE L, 4 4 msg w,

E

• ~ y,

=,..,. -,

.< ~

u s.

= =..,.,

.m

-i,-

n.

w w 1.-

~ Monticello Control Room-Staffing,oJune 6JI991,~ 3f:00111i00 p.m.C M

shift.,. E.,,,......_....;,.;.;;s.g.... 3. _ _. ;. m...., ;.

~

,4_w a

-aggy yn s

_ge q@gg;g r 'N.

4 ;;;ygg;:gy Jgg j gqg ggg gg y.

g,

,erm 3, 7#g f7 y#

- l.

, ey '

e

. - t u

u<

u e

s w

'w a,

m 6,

ge-

k. t p

a f J.

v.

(

),

A-l-

i ii; jlk,fi; > j

+x 9,

_)

l l

vii f

~

~

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

ACRONYMS AE0D Of fice for the Analysis and Evaluation of Operational Data BWR boiling water reactor EHC electro - hydraulic control INEL Idaho National Engineering Laboratory IRM intermediate range monitor LR0 lead reactor operator MSIS main steam isolation system MSly main steam isolation valve NI nuclear-instrument NRC Nuclear Regulatbry Commission 0ATC operatorlat~tNkcodtr'ols' reactor operator.

im SH shift manager SRV safety relief valve SS shift superintendent TBV turbine ~bypassvalvej r

1 I

5 #

4

.I 5-4, i y 't

(

_3 I,,

. L.

r s

-f ~. ' -.*'

s_

u p*

y y

A

, q ;.,- -D

(+;

q.,

.c<}g e

+

.d*,

y

, 7p"y p -

viii

1.

INTRODUCTION 1.1 Purpose The Office for the Analysis and Evaluation of Operational Data (AE00) of theU.S. Nuclear.RegulatoryCommission(NRC)hasinitiatedaprogramtostudy human performance during operating' events. As part of this program AEOD formed a team to conduct an onsite analysis of an event at the Monticello' nuclear generating station. A Hi-ili intermediate range monitor (IRM) scram occurred while control rods were being inserted to shut down the reactor from 6% power. The reactor power' decrease left the decay heat' unable to sustain the small, miscelluieous_ steam loads that were on line. The resultant reactor ~

cooldown of approximately 3-1/2'F per minute caused an unexpected increase in reactor power and a.Hi-Hi scram on range 7 of the IRMs before the operator could switch to range.8..

W 5;gl'L g.

y 1.2 StopA The human performance stut r focused on the actions of control room operators during the perind 4 ding up to and following the Hi-Hi scram.

The study was based on data om discussions, plant logs, the station's scram report, and interviews with control room operators and other station staff.

Idaho National Engineering Laboratory-(INEL) provided assistance as part of the AE00 program.

k

[ #'f m

'p

?

/

,,,,z,8 g

f tl..

y

)-

1.3 h 0nsite Analysi6 m

TNe;onslikan ysjsthamc'onNst'e$1of Y y

-John"Kauffinan, NRC/AE00.

George 3Lanikg NRC/AE0D (team. leader); -

g y

o Orville Meyer, INEL.

1.

The team was at the Monticello site on " June 13 and 14,1991.

I

2.

DESCRIPTION OF Tite EVENT ANAL.YSIS 2.1 lkckaround The Monticello nuclear generating station is on the Mississippi River near Monticello, Minnesota, and is owned and operated by the Northern States Power Co. The reactor is a boilir.g water reactor (BWR) 3 with a Mark I containment with a nominal rating of 545 MWe.

The reactor was in its 20th year of operation, having entered commercial operation on June 30, 1971.

~

The reactor had returned to power operation on June 5, 1991, after a refueling cutage but was shutdown from a spurious scram at 1:47 p.m. the same I

day.

The reactor was restarted at approximately 10:00 a.m. June 6 and at 3:00 p.m. was holding at 6% power, 930 psig, to perform a calibration of the l

IRM nuclear instrument;(NI) channels.

a '

W

}

s The temperature of'the' tailpipe from safety relief valve (SRV) H began to increase during the reacto-heatup, which indicated a small leak across the seat of the SRV. The temperature continued to rise while the pressure was being held at 930 psig for the IRM calibration.

It exceeded 170*F and was approaching 200*F, where the plant's operating procedures call for taking the plant to cold shutdown. The manufacturer's data stated that the SRV might fall open at 230'F..

[

L The plant manageddecIjded to take the plant to cold shutdown after the:

l-IRMcalibration'wascoEplsied.7 When the calibration Was conipleted-the shift 2

manager (SM) directed the operator,at the controls (0ACT) to shutdown the reactor. andheLOACT began;pjomptly to insertithe_ control l rods'. lht' thatitime, the elect'rolhydrauncoN6olf(EHC)]afcontrolling' reactor' pre's'sdre with 'one

~

turbine bypass valve (TBV)Lapproximately half open and with miscellaneous steam' loads' being carried Lalsoidhe T8V closed automatically as;the reactor -

l power was decreased.

Further rod insertion and the further decrease of reactor power caused the reactor pressure and temperature to decrease since decay heat was less than the steam loads. This sequence continued until the IRMs were ranged down to range 7 and reactor pressure was approaching 700 psig.

2 l

. ~ -. - - -

At that time the OACT stopped control rod insertion to check the reactor cooldown rate.

Reactor cooldown continued and put the reactor on a moderate positive period.

The resulting power increase reached the lii-Hi trip on range 7 of both A and B IRM channels and before the OACT could take action, a reacter trip occurred at 4:39 p.m. with all control rods fully inserted.

The main steam isolation valves (MSIVs) were subsequently closed, to limit reactor cooldown and the reactor conditions were stabilized.

The Monticello control room staffing is shown in figure 1.

The lead reactor operator (LRO), reactor operator A (RO-A), and the 0ATC were on 8-hour shifts and at the time of the reactor scram were on a 3:00 to 11:00 p.m.

l shift. The SH and the shift superintendent (SS) were on 12-hour shifts and were on a 7:00 a.m. to 7:00 p.m. shift.

R0 8 had been held over from the 7:00 a.m. to 3:00 p.m. shift to close the main generator high voltage disconnect switches in preparation for p;tting the generator on line.

Due to.the need to clear bird nests that had been built on the disconnects during refueling, R0-B-was in the switchyard most of the time period covered by this analysis.

The SM, the LRO, the OATC, and the nuclear engineer were in the control room during the 20 minutes precedi n the scram.

R0-A was in the back area of the control room reviewing shutdown procedures during these 20 minutes, i

The onsite analysis team interviewed the OATC, the SM, R0-B, the nuclear engineer, the site superintendent, and the lead initial licensing training instructor. The SS and the LR0 were not available for interview.

- 2. 2 Time line'of the Event j

The following; event time _line was; developed from in'terviews wlth the SM, the 0ATC, the nuclear engineer, andLthe lite' superintendent'and from tha~ '

l station's scram report, control room logs, the IRM chart recorder, and the proce'ss computer; data records.n W

l l

l l

^ M i........ - _........

i b

Shift manager j

a i

b i

Shift superintendent

• Nuclear engineer b

a Lead reactor operator

~

l I

b C

e Reacter operator A Reactor operator (operator at the controls) c Reactor Operator B E01sti:

h a.

Senior reas. tor operator license b.

Present in the control room during control rod insertion leading to the scram c.

Reactor operator license (The site specific job position titles for the R0, the LRO, ar.d the OACT are not used in this report.)

x t

/

1

.[

's

q

/y a ;q q,.

Figure 1. Mc iticello control room staffing, June 6, 1991, 3:00 to 11:00 p.m.

shift.

4

~

r.

M5491 i

i The reactor had been refueled (one third of core was new 1: 47 p.m.

fuel and returned to power but stranned from a spurious signal a few hours after return to full power).

A SRV lifted once and reseated after the scram.

Operators noticed this because usually several SRV lifts occur after reactor scram frem full power.

(This was evidence of low.

decay heat, but there was no specific transfer of this.

f information to other shifts.)

v.

E6/11 The reactor was restarted.

The plan was to bring reactor 10:00 a.m.

o to 3:00 p.m.

to 930 psig and increase power to 1-1/2;TBVs open (~8%)l ',

f p

c.

. power and hold t'o spin the main turbino :at no. load for af ; JT balance test.

Increasing reactor pressure and power took longer than anticipated due to low incremental (notch) worth of the control rods.

When the reactor pressure was atl600 psigi an elevated < $M tailpipe temperature on SRV H was noted by the maintenance';,

.superintenden't and was rep +orted to'the SH.M *. ' y M 2 M -

^

r

.o.

y qg, o

k., < ; K $

Shift change occurred for non-supervisorytoperators whop;hy g 3:00 to a

4:14 p.m.'

' were' on"ari '8-hour ' shift.$ LSupervisorylol erst' ors,; incl'udihg[@Qf

~

i the SM and the SS, were on _a 12-hour shift from 7:00 a.m.?

M

,_ M y 4 y g.gQhgsg to 7:00 p.m.

+

Reactor power was stabilized for calibration of the IRM NI channels.

Steam loads consisted of one\\TBV at-44% opsnM and miscellaneous loads including the steam air ejectors for the main condenser.

5

~ _ -..

.. ~ -..

. _.,.. - --~

f t,.

j The SRY H tailpipe temperature continued to be observed by

- r the SS and the SM.

(This indication is read from behind the control-room panels.) A temperature of 170*F and r

increasing was secorded.

(At 200*f the station operating procedures call for placing the plant in cold: shutdown.

At 230*F the manufacturer's literature states-th+at the (SRV.

[

n

~

e aa w*

x o @j y UWW may fail open.)

y

s. c An informal conference was' held in thErearif the control

~

m.~

room..TheSS,theSM,1theisitesuperistendent,fsdltheE

. plant managerjiscusskih'eiplant stMs,;espscidlly[ibb

~

~

x - w a,m amwn

<w

,+

. plan to conduct the turb~ine-balance testWersus;the,jylskE F-.~

~

as wg un xm of SRV H failing open. There was also theJultimate need?to(#

E

,m m sms.

go to cold; shutdown %w_d replace the TheJsitel' n o

,~.

.. ww 2

.s

~

,x, s g g endentjls g g tle:fn gthe gan g rg Q $

7ap o

.&w g-?op,erati,ons supportigroupF which consisEs;o,f two off3 u 53ywy 1

a m w*

• W i p. pn b M u r m

y, ym.m

' SMsandEhreeoff-duty 1SSs).

M u

.The calibration of4the IRM NIs:was' completed with 'thb "

4:14 p.m.

+

+

c reactor power levelicaleulated to be 6.2%.1

[

~

g Be.

[i

-.N 4

.. + 2 L 4W '

Y n,

?&W

~

The pl+ ant managementiconference. agreed;tofshutLdown}the d,

wn n

+

mn. c+

reactor and replace thef H SRV.. #Th+e finalidecision wasy.a@%

& adnn nunwionalftarget' wast tok

?

-.b y

v w ; umy 1

madelbyithe ~p;lant' manager' &The o

, s.9(theidrywell!)mwu n > per a

nw nuw gm he@w.

o u

gi

~

N ntrob r &y,midnigh,t M @3 fbeiin A*.n he % e w q a n wwas not O% u g%

A

%.m m

tSRVa rihe co a.w 1specifica oom crew ef 7,

p y, p m x g o p<and;the management.dec. % q y g

w.nagg.

m plant status tsion. eThe w c about t g m 4 m p g gund mitionlofumyyy p n,, w dhe; al E re iLR0 andjthe^0ATC.wereenware punqumwp amm zug y &conwwpa&QrQ 37 y

e* ;,g:m the H:SRV,.and M t {he pun laanagene

'conferencesisiithey J

• 1 contro,l roomiconcerned th :p~ x y ;s e n y -w n w. y >the d,

m.

9Gme ISRV4p@@dh%Q%, M,e_ossibility ofgfailur j$gg

[

.7 w

s,

.m

$$p M*

The SH mentally considered,the option!of,plac!pdg the mode; qp,

4

+

switch in the shutdown position,e p m e mwhich would have ' ~y' ng

(

mk ~ ag uggwa x g

~

7 j

initiated a manual scram of the control rods and thereby eliminated the time required to muually insert the rods, 6

F-s.

4yie-s e T

9 f

y i:n'..-4---1 e--A +

vg "'1T

• lt-t*

1T--"-Pg' Fe it-r--

^^$'"

3

.m one rod notch at a time.

The station's normal operating q

procedure for reactor shutdown did not provide this as an option for a reactor shutdown and plant policy was to avoid the resultant wear on the control rud mechanisms.

The SM discarded the option without voicing it.

TheSHdirectedthe'LR0'andt$e'0ATCtoshu't'down'thei 4:19 p.m.

reac;or "without delay." The'0ATC began promptly to prepare to Ssert control rods, one notch at a time'per the prescribad, detailed sequence.

.TheSM,theSS,andtheLR0revkewedanddiscussedact' ions for proceeding to cold shutdown and preparations for replacing the H SRV.

The SS entered the plant to initiate the time consuraing process of de-inerting_ the1 dryOklljtoy

\\ +- *"

z i

vV.

l permit entry and access'to the HLSRVL hThe LR0' began.to

-f y

+

review the administrative control actions needed' e? '

s

.~....

to transfer from the normal startup procedure to the r.ormal shutdown _ procedure.

1 R0-B tad returned to the controlLroom'after closing the e,

high voltage disconnect switches. HeiwasfdirectedtoIre-

-open(theswitches1nddreturnedtoihsswitchyard.h,Iw r, j y,.

gg y-

.;c. y

< The.0ATC began notch jinsertionsloficon.g,

N.

o n ;.

z, responded to!the decreasing 2 reactor power.m y,

trolirods,dThe 4:20 p.e.

g ggya ; m,e e by incrementalm M v

+

w n.

. yg ;

.w w n ;a n ~ w, m ; m m,,

s closing movements oflthe IBV;after each rod notch

~,

x a.a w,x w ay. ngreac, ma m o; tant,g m y g a.

so

.x insertion and thereby ma niained! con,hej3V po s

tor g 7

< e

' n _;mne. geyA grgejn ;t 2;

p.

The iridicat change

-pressure.

notobservedby.th'e.0ATC..[(Itlisnotfreadableby,the0ATC.

a -

wh'en heisTseated atjthe100k5lpaneld$\\insertingirbdsjs%g

,p, t

t.

4 f'

7 i

4:30 1/2 p.m a$ partly open 1GV was now fully closed by the a'itomatic

'ar pressure control function of the LHC.

(At this time the sum of the decay heat and recctor power was equal to the miscellaneous steam loads.)

The 0ATC did not notice that the 10V was now closed. The e

TSV position lights are in the son.e panel location as the 1BV position indicator and do require the 0ATC to deliberately look away from the C0-5 panel, but they would have been visible to the OA10.

1he SM observed that the IUV was closed but did not i

recognize the significance, which was that further t

reductionr. fn reactor power would cause reactor heat to be i

less than the miscellaneous steam loads and that.this.

t would cause the reactor pressure and temperature to decrease.

I 4:30 1/2 p.m.

The OM" cantinued to notch insert control rods after the iPV was closed. The reactor pressure began to decrease and the negative temperature coefficient of reactivity began to add reactivity in direct proportion to the amount ofreactortemperaturedecrease(coefficientwasestimated later at -1 x 10 delta-K per 'F).

After each rod notch l

l insertion, the IP>i indication decreased but began to increateaga.injueto,lhe:reaciorcooldown.

Theilendency

. ~

. ~.

forthel,RHindicationioinbreaseagainpromptedthe0ATC I toLinsert yotLi iotherfr @ )o(ch.1 ThE 0ATf didlnot realize' a

t o

i the meanirig 'o'f 'the tendeNyfof'the.!RM jhdication' tb it.e.rease and that it was caused by the reactor cooldown.

m

pp.

.M

.' MQma f

The above sequence of rod notch insertion and IRM response continued for,eight minutes and the OACT succeeded in lowering reacior power icvel f rom range 9 to 7 on the IPJ4.

8 l

l 1

Hain steam line low pressure alarm

'4:33 p.m.

+

l 4

1ho 0ACT noticed that reac*or pressure was decreasing 4:38:30 p.m.

+

toward 700 psig and recalled that reactor cooldown rate i

should be checked before decreasing reactor pressure from normal operating pressure of,930 psig.to below 600 psig.

o The OATC discontinued rod notch insertion to check l,1 cooldown rate but did not realize the effect that the L

cooldown was having on reactivity, g

Reactor power _tr.:reased due to continueh reactor cooldown; I

An alarm and half scram trip occurred on th_e IRH B.

channels. The 0ATC moved to reset the B half scram trip,

' believing this to be caused by a spurious level spike.

e n.

f I

_f

,* 4 The SHslattent. ion wa,s directed to the OATC's actions at-

"t n ;

.,he C025hnel.by the alarm; The SM and the LR0'had not been monitoring the decreases in both reactor power and g

O pressure.

P The continued reactor power increase tripped the A IM n

4:39 p.m.

+

channels half scras trip and _the reactor, scranned from half-scram trips on' both the A:and B IM ch'annels.. All rodsdere~ inserted..The operators entered emergency!

M

~

~

. t

. operating procedure C-4lforlreshonseito reactor. scram.

[

1..

,.mt.

w

. a.,

,um

. m +a

w. g 1

a:

.+

w

,+.

. -j e>

11.

.,..A.

.. ? !.

a1 i

r 4

~

, a

.:+4 isted an investigation of the cause of the Hi-The SM-init a

wv

.c w

y Q,

ay mj e : w n>HillR$; scra% g

~

1. y p/ nap 4

p m

uw p

p

.x t

u a.

Operators completed immediste' actions"forfemergency,

1 4:50 p.m.-

- - @ r peratihg" procedure Cr4"and reset the reactor-scram W a &W !!

e 5:05 p.m.

The RO-B had completed closure of the high voltage.

disconnect switches and returned to the control room.

L I

L 9

a Per the suggestion of R0 8, the SM directtd that the M51Vs 5:?0 p.rn, e

be closed to prevent exceeding the lechnical Specification limits in reactor cooldown.

Reattor was at approximately 447'f, 425 psig.

2.3 Analy111 2.3.1 OpnA110nal.51rJLt.t9Lf_RT_fka012r_ihuldQWD Many BWRs turn the mode switch to the shutdown position after power is reduced to less than ~12% and thereby scram the control rods so inat reactivity control is ensured regardless of variation in pressure, temperature or xenon without operator action.

This stratcgy has the disadvantage of possible wear of the control rod mechanisms and of the need to re-vent the scram accumulators.. The Monticello plant management had decided not to employ this strategy'and the shutdown procedure did not provide this as an~ option to the notch by notch insertion of the control rods.

Most controlled shutdowns of the Monticello reactor from power operations had occurred with sufficient decay heat to operate the steam driven air ejectors for the main condenser, This permits the operators to leave the reactor pressure control in automatic with modulation of IBV position controlling the reactor pressure. The operator _ can then insert control rods without immediate concern about reactor cooldown or heatup until after the reactor hasf been driven sdcEiticA1Vaffightficant margin. If decay he'a't is at a lower value, the steaddriven air eN$torsiould be secured and th'e motor

• driven, mechanical vacuum pump started to permit continued ' automatic pressure- @ o._

~

control of a TBN After thks scram [ tiie'Mohicello" nuclear ~cndi' ce'ri. !

1. '4

?

n estimated that the decay heat level had been 2 to 3 MW thermal versus typical levels of.15_ to 20 MW thermal after other controlled shutdowns from power operation.

However, the fact that the decay heat was very low was overlooked '

by the operators and the information about only one SRV lifting after the previous scram was not passed on and evaluated.

10

l This event prompted the Monticello station staff to contact other OWR operators to seek information on their strategies and procedures for controlled shutdowns from power operation. Monticello planned to re evaluate l

their strategies and procedures after receipt of this information, in the interim a note was added to the shutdown procedure and a memorandum was issued to Operations personnel.

2.3+2 To Awarantu The decision by plant management to bring the reactor to a cold shutdown condition to prevent a failure of SRV H presented the control room crew with a l

number of related tasks. Under the direction of the SM, R0 B was sont to the switchyard to open the high voltage disconnects, the SS went into the' plant to prepare for entry into the drywell, the L.R0 directed his attention to transferring from the startup procedure to the shutdown procedure { and thel DATC was-told to shut down the reactor.

1he 0ATC proceeded to notch insert ~s control rods-almost immediately after receiving the order to shiit down the' reactor.

The absence of any-pause by the SM and the 0ATC to review the requirements of the task before actually inserting control rods indicates a low level of task awareness. Conversely, a pause to review the task before beginning the task is a method for ensuring a sufficient level of task awareness. The unawareness of the plant staff _ of the_ effect _of the low level of decay heat upon reactor control tasks at low power was the basic factor leading to'the unintended scram. However, there were other factors that permitted this-to progress to the actual scram.

j The unawareness of t'he possible effects of decay heat on reactivity l management indicates alow:levellof. recall during thiEevent ofLthelbasic,

~

I operational theory of reactivity management of a BWR [Inlthe'shodieml reactivity of'a BWR is determined by power and pressure, which determines the--

void-faction,-byltemperature of the water, and bytthe controlfrodst$ Allfofh %

these variables need to.be monitored by the operator in order to control and manage reactivity, m

Therefore, full awareness of the reactor shutdown task should result in l

monitoring and control of the reactor pressure, regardless of decay heat 11'

levels.

In this event it could have permitted reducing the cooldown rate while the control rods were being inserted and thereby restoring full control of the reactivity.

This relatively mild event provides a clear illustration of the significance of the need for awareness of the requirements of the reactivity management task.

2.3.3 D n gnd. Contr01. m LCr mgniS/J1001 The decision by plant management to shut down the reactor was made by the management conference in the rear of the control room and was translated almost imediately into actions by the operators. Since the tailpipe temperature of the H SRV was still rising towards the 230*f. temperature, where=

failure was expected, there was a need for early initiation of cooldown if failure of the SRV was to be avoided, liowever, placing the mode switch in I

shutdown and dumping steam to the condenser to cooldown t,he reactor. wo_uld have-(

been a simple, very controllable option.

ThewearonthecontbolIrodho I

mechanisms could have been balanced against the risk of SRV failure'by tNe' y

control room command if the command thought the risk justified the shutdown.

i The absence of the command and communication function of briefing the crew on the plant status and the options was compounded by a low level of control of the OATC's action. The OATC was given a broad command-

• $ hut down the reactor"- with no amplifying-instructions on control of. pressure,; steam.

Ioads, and reactivity. There was no feedback of information from the 0ATC to either the LR0 or the SM and little_ direct overvjew of the 0ATC by either, g, y" The reactor-pressure and temperaturelindicators1were providing direct %.

7#

4 indication' of the reactivity addition because of.the cooldown.' Other. signals 3,.

of an unusual conditisn. occurred [ including the iBV p;osition ind@atohndj qg j

I.

lights, the MSL low pressure al' arm', and poweilncreases after'each' rodN insertion. Steam loads could have been reduced-to reduce the cooldown: rate.

More effective monitoring lof the plant statusiand the 0ATC'slactions;by the#S

~

LP.0 or the SM coule have prevented the scram without any knowledge of decay heat levels.

The reactor cooldown was actually not brought under control-l until 40 minutes after the scram.

1 12

~

l 1

It,e f ailino open of an SRV while the reactor is at normal operating pressure is a significant transient due to either the uneven heating of the torus or the rapid reactor vessel cooldown if other SRVs are opened to equalize torus heating.

ihe major objective of the control room command was in this event to reduce reactor pressure to prevent the SRV 11 from failing open. This objective was met, liowever, scraming would have been more direct and if the SH or the LR0 had been monitoring reactor pressure the decrease in reactor pressure would have indicated that a manual scram was appropriate.

2.3.4 11rc1Llituntm3D1 The perceived risk of failure of the SRV H created a significant level of stress among the control room operators, llowever, of more direct significance is that the absence of a command bi:efing on the plant status and the plan of action permitted this stress to interfere with task awareness and.with command, control, and comunications.

2.3.5 tian-mashine IqtitfEg A transition from low power operation of a BWR to a hot shutdown condition by means of rod notch insertion requires the operator to monitor reactor pressure and cooldown rate and to monitor IRM levels and range switch settings while executing a prescribed, rather complex sequence of rod notch insertions. The arrangement of indicators and controls makes it difficult for a single operator to do all of this. Assistanceinmonitoringthejindicators is often provided the OATC at BWR plants.

lt is standard' practice to ' provide' ' y a nuclear engineer to monitor the startup of the_ reactor. Consideration should be given to.stmilar. assistance d0 ring' reactor. shutdown'b/ rod Iotch insertion.

2.3.6 ILLihjR9

~

n 1.'

There is obviously a training aspect to all of the observations above in 2.3.1 to 2.3.5.

Training objectives and content are means for improving operational strategies, awareness of reactivity management principles, comand 13

e i

i o

and control, management of stress, and adapting control practices to i

circumvent potential problems with the man machine interface.

2.3.7 Event Evaluation i

Shutdown of the reactor by mu.ual insertion of control rods from a condition with appreciable steam loads but with a low level of decay is 'not a common operation so there may be an opportunity for an improved understanding I

of BWR operations through an interchange of information between Monticello and other BWR plant operators.- The Monticello staff expected to initiate an interchange of infomation.-

t' The Monticello staff method of event recall and assessment by the operators is based on a control room crew group discussion. ThereLare advantages to de briefing;the operators as a crew in a group.

H6 wever,_there i'

is a specific benefit 1'f[the operators first write individual statemenis of their personal knowledge and perspective of the everh.' This avoidflost$gan i

individual's particular observations and opinions because of the possible D

development of a " group think' version of the event.

li The Monticollo station uses a unique application of the Human Performance Evaluation System. Monticello has a Human Performance Task Force, which has a representative from each major grou'p of the station staff. Each individual I

may refer events, probleres, or suggestions'to:the t, sk force.; The site -

J e

superintendent is.the chairman of the taskJorce.J The deliberations of the g.

task, force are private.1[AfterI keysk drce realhda consensus (tNeyLpublis@

t

+

n m.

their. findings and recommendations. The recommendations are eithsrI implemented by.the h

$N:k'tsNhk$$skl7

.[

forceforfurtker~r[siatlon'5111ntmanageNentorNefered esolution[ '

Yj M

h 'N [#

x,, gg - p yw; jy <

n a +,

14

. ~ - -

3.

SUf', MARY Of HL"Wi i AC10R5 Of lHL lVUiT The operational strategy for reactivity control during shutdown of a BWR with a low level of decay heat must prevent miscellaneous steam loads from causing an unintended cooldown of the reactor and provide corrective actions if cooldown becomes excessive.

The operational strategy used for cooldown during this event was the initiating factor leading to the unintended scram.

A low level of conscious awareness of the reactivity control task requirements permitted the reactor cooldown to continue until the negative temperature coefficient over-rode the control rod insertion and precipitated a lii-Hi IRH scram.

These were several indications that could have permitted the reactor cooldown to have been detected, its affect on the reactivity control understood, and the cooldown stopped or its rate decreased..This did not happen because of a low level of conmand, control, and communications with respect to the reactor control task. An unnecessary level of stress among the control room operators exacerbated the command, control, and connunications problem.

The man machine interface for the 0ATC makes it difficult for the OA1C to monitor the basic reactor conditions of pressure, temperature, water level, power level, and rod positions, it is suggested that providing a;.istance, such as is provided by a nucicar engineer during startup, be considered.

i l

l l

15

-.