IR 05000317/1990024

Insp Rept 50-317/90-24 on 900904-07.No Violations Noted

ML20062B844
Person / Time
Site:	Calvert Cliffs
Issue date:	10/02/1990
From:	Roxanne Summers NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
To:
Shared Package
ML20062B842	List: IR 05000317/1990024 ML20062B836
References
50-317-90-24, NUDOCS 9010260161
Download: ML20062B844 (21)

Text

,

.

-

,

,

.

s U.S. NUCLEAR REGULATORY COMMISSION-Region I Report No.:

50-31//90-24 License No.:

DPR-53 Licensee:

Baltimore Gas and Electric Company Post Office Box 1475 Saltimore, Maryland 21203 Facility:

Cahert Clif f s Nuclear Power Plant, Unit 1 Location:

Lusby, Maryland Inspection conducted:

September 4-7, 1990 Inspectors:

R. Summers, Project Engineer _, DRP A. Howe, Resident Inspector W. Baunack, Senior Reactor Engineer, DRS

- l S. Sun, NRR E. Trager, AE0D Approved by:

^% m %

lo O

-

Robert J. Summers, Acting Chief

/Date Reactor Projects Section No lA

'

j

-

Inspection Summary:

This inspection report documents the special team inspection of-licensee activ-ities associated with the unexpected voiding'in the reactor vessel head during an RCS draining evolution on August 30, 1990, at' Unit'1. The report includes:

a sequence of events; an assess operator actions and procedural, ment of ' safety significance; an assessment of adequacy; a review of associated instrumenta-tion including modifications, maintenance and training-activities thereon; and, an assessment of the licensee's investigation. activities.

Results:

40 violations were identified.

The Team had three concerns regarding staff-support, operator performance, and temporary' modifications..These' issues are more fully addressed in the Executive Summary which follows.

-

'

,

4

"

9010260161 901009 PDR ADOCK 05000317 O

PDC

'I

,

,,

[e - }

.

'

.!

'

.

,

.

.'

-EXECUTIVE SUMMARY

'i Special Team Inspection' Report No. 50-317/90-24 Sequence of Events The Team had two observations on the ' development' of the sequence,,'which is applicable to both the licensee's and the Team's process,

!

1.

It wGs very difficult to recover raw information during ~ he operator t

-

interviews since they were not conducted immediately af ter the ~ event.

ii.

Logkeeoing in the Control Room made it difficult to reconstruct the: time

!

line of activities leading up to,- during and immediately af ter the' void

occurrence.

The most probable cause of the void formation-was the collection of entrapped-air in the RCS into the reactor vessel head and-controllelement drive penetra,

l tions, which, when depressurized on _ August 30 ino support of RCS draining --

)

-

expanded until the void was detectable' by the Reactor Vessel Level Monitoring:

{

System (RVLMS),

Safety Significance

The Team determined both plant or.d personnel safety significance s as follows:

1.

The formation of the void alone could not= have -resulted'in either direct core uncovery or the loss of shutdown cooling function and RCS mixing without additional indications thatothe Team? believed - would - have ' been -

properly acted upon by the operating personnel-

'

.

.

ii. The venting of the void was ~ conducted at _a time.when.its composition ^ had not yet been determined.

In addition,. since the venting /was into ;the;

-

containment atmosphere, adequate precautionaryf measures. were not.taken

.

y using plant chemistry, radiological controls, or other engineering support i

to protect personnel in the containment.

q The Team was concerned with the lack, of adequate support prior to the venting.'

!

evolution (Section_2).

~

'

o Operator Performance and Procedure Adherence

j i.

There was an apparent difference in operating philosophy regarding proced-

.

-

ure adherence between the two shifts involved during-the ; draining'

)

evolution.

The Team was concerned with the use of " journeyman knowledge"f as applied

_

to the venting operation on August. 30L as. it, related to the personne1'

<j safety considerations. mentioned previously-(Section-3).,

t'

..

[

.

"-

,

i n.

.(I.

'

,

'

l'

I ;l f

i,i-i'

>

t

,,

'

,,

,

'Y K

_, _ _

L W, i

'

'

'

..

.

.

.

.'

.

Executive Summary (Continued)

&

11. There appears to be a need to correct certain prot 9dures.

The RVLMS alarm response procedure does not account.for the condition of the plant at the time of the void formation (it' assumes accident ~condi-tions only).

Various operating procedures show different' ways of venting-_the vessel-head area.

At least one method, to the quench tank,-does not appear to work.

Unless conditional upon certain operating. parameters, a single correct method should be. developed and the appropriate procedures changed to reflect that approved process.

Contributing Factors The Team was concerned-that the operators were either not' aware of a Temporary Modification to RVLMS (or of its effect), which had been made prior to the RCS draining evolution (Section 4).

Licensee Assessment

,

,

The Team found that the licensee's investigation' team' (Significant : Incident

-

Finding Team (SIFT) was a90ressively pursuing ' the event'.

The SIFT findings were consistent with those developed by the NRC. Team (Section 6).

.

--'t I

i k

' i,

!',

,

s

g '.

,-

j

"

"f

,

,

,p

-

. - - -

-

~ -

,

..

n-m

-

-

,w

,

,

m

>

.

.

-- a~

.

.

~..

.m

..

,'.

,/ \\

-!

.. : 4

,.

.

.

.

.

,.

,.

+

-..

,

- }:

)

.,

41

%

w

,

.

.....

g

.

TABLE OF CONTENTS y

a

.

Pye

-:

e

1.0 S e q ue n c e o f E v e n t s................. -........ -.. >......... :.....

~ 1:

j

.

.

.

1.1 N a r r a t i v e ' S umma ry...........,............_......... :....... c 11"

'

!

.

.

1.2 T i m e L i n e........ ~.................... '.......-... :.......

- 3.-

[

..-

2.0- Safety Significancen.............................-............

A j

-

.

.

.

3 s

,

3.0 Operator Performance and Procedure Adequacy.....WA._'..

4..

83

t 3.1-Background..........................................

..-

8-

.

...*

<3.2 Operator Performance..................................

. 9)

{

3.3 P ro c e d u re. Ad e q u a cy..............._........... _.,. f....._...,.

'10?

-

,

~

4.0 Reactor Coolant System level Instrument'ation.-....'.v........

'12 <

..

.

.

<>

'V 5.0 Reportability...............................................

13s

-

';

.

l

w 6. 0 Li c e n s e e ' s. I n.v e s t i g a t i o n........................i.. -..........

c14 L

.

7v 7.0 E x i t I n t e r v i e w...................... =.................. :. 2....,

14;

,-

,

.t Table 1 - List of Acronyms i'

%

,

,

Table 2 Reactor Coolant Systems Levels.

i{

.

,

,t x

.

.

'y

.

Figure 1.- Reacto'r Coolant Sy:; tem Vent' PathLandL Level :Indicationsj

~

U

'

,

a

,1-t

'.

'%

_

_

lh ;

.

-f

.E]

d s

lt

'.i '

,

.

,

!

'(

,

g

'n

>

if

i Mt

' to

!E

, c

.

-. :. i

! 1

,

';,'

<

.l

- Q.

<

.i

- -

,

H

]

'

'

.

!l -

, _

']

i

.g.

j

'.?

!

'i

.,

,

,

.h

\\,

$

,

>

f-

'

,

n,.

s

' Y ag

.t-

,

.

/!

I I

f(

g

< i q'..

,p-l

,

i

,..

'.

i; v/ f I',:s. g

,

4 Y, '

,'l V:. ; [ - ! y'q i

,

, > > :

,

t i

,

,

• .,

I r

.

>

. !

,'

.f-h' r,

N

'

'

v-

.'g-'o &.a.,

. j'

U d,

Cm

,

..g j

',,

- :

i i

i

-

1 -

"

m,

,

,,

, -

m 4.:

.., o r

<

,

'

!

4'

.j

i *;

' ' ,' \\ ) ) (. -

. _I k 3 s

!

..-

t i

i

'

.

? i

'

_

f i .

<(n.

f. )

,ji

,O

.,.,.,

[ i'

i

'

.{

...

)

-

i

,;.;

i, ' j W. 1. I t

.jr,

-3.;

'yy'

i

,

,,l l [.' M ! ')

'i J[<

( g ;. ;l ""

'

r

,

.

--

$

ey

,

f 3. ;.,p

', ' +

< h.

.,

(( % Gy pie d.v f b. f, %.,; c-M 31,tH.a.e M,

,i,

'L~, '.,1

' /

Ae

'

@s l j *<

s

.

.

.l

~ :. m

,

,,

.r-

. : 20,d :-

l D.a

>

>

o

,

<

L i '

MJ s y> W u

r -

v

' (}/,

,

,

,I,,, ; ' l '

i) t

A3, ( :, ', ' ra!.

,

+v

l;,]

'jrt i

'

'

.

~/g ~

3:

' N,y- }Y ( y 9'.,.

&,i i(.

!

h.

I jfh: /h, !

, $,

[, ',, 5,., [

g j.: i

, N, jiY

,/

j; h:

'

'.

,,

.-

,

-

.

~

.

-

.

DETAlj.S 1.0 Sequence of Events Refer to Table 1,

List of Acronyms, Table 2, Reactor : Coolant. System Levels, and Figure 1, Reactor Coolant System Vent Path and Level Indica-tions, for supporting details.

'

1]

1.1 Narrative Summary On August 30, 1990, the operators commenced draining'the Unit 1 RCS-in oraer to support the replacement of the No.12A RCP seal package, j

The Unit had been in a cold shutdown condition since April 1990 for a-

scheduled maintenance outage.

Prior to that, the Unit had = L'een 't I

power for approximately.1 week.

The Unit.had been. previously main-i tained in -a shutdown condition for an ' extended period of m time.

'

Therefore, at the time of this event, there was very, little decay -

heat in the core which could'have' contributed to the significance of the event.

li hmedictely - prior to thefdraining evolui;on, the RCS was11n the cold

!

shutdown-condition, with a steam bubbleLin the Pressurizer (Pzr)' and z

decay heat removal in' service. About 3 weeks earlie the RCS had been filled and -vented per operating procedure to= suppoi a maintenance

,

on the Salt Water Systems ~. At that time, the operators _did not com-plete venting of the Control Element-Drive; Mechanism penetrationsLdue'

to personnel dose considerations. Plant management was already aware at that time, that the No. 12A RCP seal needed: replacement-and,there-fore the normal RCS fill and vent was not. necessary.

This" later.

.

contributed to the unexpected void formationf on August 30,-1990.

During the current maintenance outage, -other RCP sealsT hadi been

rep" ed.

This activity requires operation during reduced inventory.

i Thc

.censee had _ previously ' implemented ' procedures for. such opera-j tion.

The operators were familiar with the special problems' asso -

-

ciated with reduced inventory operations and' were : preparing to

)

implement this ' procedure in. support of. the ' August 30,: 1990 draining i

evolution.

Reduced -inventory. control procedures, training) and O

familiarity of operation by the. control room' staff was Lreviewed by.

j the resident inspector in Inspection Report No. 50-317/90-09; in_ May j

1990. At that time, the licensee's' actions were'found acceptable, in; l

that they had procedurallyL' incorporated Generic Letter 88-17.

'

.the recommendations 'of-l

,

0 f

-

d

'

o

-

-

.

.

.

.

g

_,

}-

' j{:

Ok 'i

'

-

.

,

'

.:

.

.

.

.'

,

'

Although the intent of the RCS' draining was to reduce level to the

~

,

'

top of the hot leg (~39.08 feet in the Refueling Pool Level indica-tors), the operators were surprised when the 1st RVLMS alarm (~48.581

[

feet) occurred at about 0537 hours0.00622 days <br />0.149 hours <br />8.878968e-4 weeks <br />2.043285e-4 months <br />.

(Please-refer to the attachmenL for reference level descriptions).

-

This alarm indicates that' level should be equivalent to the bottom of the Pzr.

Operators also 'had operable Pzr level instrumentation.

which indicated that the Pzr was nearly full at the same time. The operators treated the RVLMS alarm as valid = even though they: hadfcon-

,

flictir.g Pzr level indications., Appropriate actions were taken:;to

'

resolve the conflicting indications and the ' operators quickly; and -

correctly identified the void in the reactor-vessel head. They-also j

appropriately determined that the void' was a= bubble of non-condens--

'

,

able gases, i

The condition' was evaluated by the' operating' staff and. determined to-be not : safety significant since the void was notLincreasing in size

,

-

and that' level was far above the top of the core asLwell as'above..the-i level in which.reducedt inventory controls c were established.- The-

'

operating staff then continued to drain the RCS,until such time that venting of ~ the vessel head : could be achieved.-

The venting : was finally achieved by a second operating crew about: Sh hours after the

,

condition 'was recognized.

The operators did not1 understand why: the

"

-

void formed or why the RVLMS-alarms indicated as they.did. --They were also unsuccessful in quickly establishingnan.-appropriatef vent path,-

which was probably attributable to inaccurate procedures.;' Although;

-

operations department management was aware o'f L the problemL and ' the:

-

.

potential solutions early' on the ' day-shif t' of August 30,1 1990, sit was not clear. that other station management or. engineering : support were:

>

made aware of ' the condition -until af terJ it Mas corrected byr the operating crew.

'

The' STI found that the most likely lowest level-in: the. reactor vessel

-

was between 45.9 feet and 48.6 feet '(Refueling Poo1Lle' el); At worst v

case, - assuming a single,. unidentified ~ instrument L fail' re, the ' level; u

could have been as low as ~42.5 feet. - For thi.s worst case condition,-

there would have,been about 3.4 feet ofm water! above 'the top' of the

hot leg' and about 8.8' feet of.-water above the t' p of the core..Forf

-

o the most ' likely conditions,. there would L have-been : at least ' an addi'-

'tional 3.4 feet of vater above those fstated.

-

I l

!

sg

'

l

,

j

>

a q

>

r

.

'

t:..

. 1

J

' '.

,

t i

}

!

'

i

r

'e

s

,

,

.

.

.

.

i 3'

Had the void continued to enlarge, three more RVLMS alarms would have~

!

occurred prior to the bubble entering the ~ top of the Hot Leg.-

The bubble would mo.t likely have traveled along the ~ top of ~.the Hot Leg until it encountered the Pzr surge line; The contained volume of -

i water in the Pzr would have been released into the-RCS at that time, which would have also been indicated in the Control--Room. With these various indications, the operators ~ would have had sufficient, time to respond to a decreasing RCS -level prior to a loss of SDC.

Based on discussions with operations prrsonnel ' and a, review of. var-ious records, the STI determined that-the' void. formation was not' the

'

result of an inadvertent draining of the. RCS or due to steam bubble,

3:

formation..The' most likely cause was-due to.'.n accumulation of non-condensable gases.

The cause of this accuaulation' could. 'not ' be definitively identified; however, was 'most likely1due to entrapped.

air pockets -in the -RCS swept into the CEDM penetrations under-pressure.

1.2 Timeyne Init'.a1 Conditions:

(as of 0000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> 8/30)

Unit 1 in mode 5

--

RCS temperature @ 120 F i

--

RCS pressure @ 165 psia:

.

--

Pressurizer bubble collapsed / pressurizer. solid -

--

Pressurizer cooldown in ~ progress using auxiliary spray from<

--

number 11 charging pump-Shutdown cooling in service using number 11 PSI, pump

--

'

Operators in the process of entering. reduced RCS inventory' con-

--

~

L ditions per OP-5 " Plant Shutdown From Hot Standby To. cold Shut-

-

down" in order to replace the No.12A RCP seal.

l RVLMS status:

.

The channel A 10 inch light (indicates the lowest level) wasilit and-the reactor vessel water level low' annunciator' was on -and -had been j

-

previously acknowledged by the operators. 3A maintenance request' had

.,

been written for this light = in. June 1990. The' remaining' channel A

"

and.all..w el B lights were-out. ~ The two highest 11evelsrin? channel-B were

'r red via a temporary - modification. - Thh modification

would caua both level. lights. to illuminate if ; the. highest sensor-

'

became uncovered.

Operators were' not aware of' the system: response from this modification' until after the RCS' drain;was ; complete.

The

.

-

RVLMS Channel A Trouble and-the Channel B Trouble annunciators were.

both bypassed.

.

.h t

'

,V

',

y

<

,

,

,

'

'

'

'

t

,

,

n i

'

.

<

.

.

4

.

Date Time 8/30 0240 Pressurizer cool down complete. Pressurizer water tempera-ture ~ 190'F.

No. 11 charging pump was secured.

,

0340 RCS depressurization to 25 psia completed. Letdown secured.

per OP-5.

,

0520 Cross connect of SDC to CVCS purification completed per OP-5 and 01-2D.

~0033 Commenced RCS drain per OP-5... Drain was ' started by diver-ting' SDC flow via the CVCS purification to the waste pro-cessing system.

~0535 Opened pressurizer vent -. valves RC 215 and 216 per OP-5.

This vented the pressurizer ~ to the ' containment atmosphere via a vent rig.

0537 Within 2 minutes of apening pressurizer vents the RVLMS alarm lights illuminated on the channel A 185" level and.

channel 8 185" and 153" levels.

The' channel'A 10" light ~

went out shortly' af ter.

No annunciators were received as

-

they ware either bypassed or already in; the alarm state.

Pressurizer level was ~330 inches.

,

i 0537 to

~0600 Operators thought. that illumination. of the. RVLMS.. lights was unusual so several action's were taken:

,

Control room operators reviewed plant status including SDC parameters, pressurizer level,.RCS temperature;.

-They concluded that !DC was not in jeopardy :and;that a steam:

bubble in the reactor. vessel' head was very unlikely.

The alarm response procedureL was reviewed but guidance aas.not applicable for the - conditions of 'the plant : ati that ' time.

RCS drain rate was calculated via' timing of waste. system

,

integrator change and calculating flow.

RCS Edrain rate flow was calculated to be approximately '85, gpm.JThe pres-

'

surizer had about a 2" to 2-1/2"' per' minute drop in level

-

which also corresponds -to 85 gpm flowrate. Operators-were.

confident that-the expectedidrain. path was theJonly. drain path for-the RCS.

I

t

<

,l

'

'

'

y-q-

.;

,

.

.

.

,

.

.

S-

.

f I

O_ ate Time

,

8/30 0537 to

~0600 The plant watch supervisor. (PWS)' took. local RVLMS cabinet readings.

Alarms were determined to.be valid (i.e., high

-

delta T's).

It teck about 15 to 20 minutes for the PWS -to complete this task.

.

Operators discussed concerns. about level indications an'd:

-

the ' possible presence of a void in. the head composed. of.

non-condensable gasses. The operators were also concerned ~

with operating - with the RCS (Pzr) solid.

They: decided to

-

watch RVLMS and continue ' draining. Although not wr_itten in:

logs, the operating shift had decided to secure,the drain-down if additional RVLMS alarm lights illuminated and indi-cated that the void was becoming larger.

0655 The oncoming shift supervisor saw RVLMS lights and -was very concerned with this condition.

He wanted assistance from the system engineer to evaluate this. condition..The'oncom--

,

ing and offgoing shift supervisor agreed to'stop-draining in order to evaluate the-situation, to allow the system

engineer to see current conditions, and to-allow conduct of

'

a shift turnover, pressurizer level'was at1~ 190"'when the drain was secured.

0700 to 0800 Shift turnover conducted; GS-N0 l and. AGS-N0 Ltour control room during this time.

~0820 Operators vented the reactor vessel 1 head-to quench tank in.-

accordance with OP-5. 'This-was unsuccessful-in removing-the void.

There was no flow.(either liquid or gas) since

,

there was no change noted. in, pressurizer-level, quench' tank level, or o.uench tank pressure. Operators noted that there.

was about. a 7 psi overpressure in.the quench tank.. Since-the RCS was vented to atmosphere through the _ pressurizer, they determined that there was no driving' head to. move ~ any void gas to the tank.

_.

Operators placed refuel level instrument' in' service per

-

OP-5.

This instrument senses level directly from the hot leg.

As a part-of placing the ~ instrument in : service, a

'

. tygon tube standpipe, which also: indicates RCS -level, was placed in service and various instrument levels were com-pared.

The refuel level instrument : the ' tygon tube.and

,

pressurizer levels all 'were within 0.1. foot of. each other,

-

corresponding to thel then current ipressurizer level,of.

-

about 190". The tygon-tube was valved -.out of service and periodically placed in. service..when needed for comparison.

,

S j

___g

'

.

-

.

.

.

Day Time 8/30-0820 Operators decided-to recommence draining.

The AGS-N0 con-curreA Drain would_ be stopped.1_f any more RVLMS lights or unusual conditions were noted.

-

-

0900'

System engineer was in the control room. He reviewed RVLMS-indications. He determined that the _ alarms were valid.

He:

'

informed operators that.the 153" light _ on channel _B is an engineering - problem.

~0915 Commenced draining the RCS.

~0920 Operators vented vessel head via valves, RC 274 and.275, to the emoty refuel ' pool floor in containment.- The - head was -

vented for a total of about.15 minutes. SomeJgas and water-flow was seen in the tygon tube connected to the end of the-vent ' path.

RCS drain'ing was in. progress. - The operators-decided that with the pressurizer level at about 130 inches, at that time, that there was insufficient head i to completely move _ the void through the small vent line. Air-

,

samples for radioactivity.were - taken In?the_ area of othe vent outlet to ?the refuel pool floor during ventin;,

$ ample results - were confirmed at about 110:40 a;m.

No

,

appreciable activity was found.'

~1030 The RCS was drained to a pressurizer level of 340" per OP-5 so' that the steam generator tubes could be ' blown" free of-water using plant-air. (Oecreasing the Pz Alevel was neces-

,

sary to, provide suf ficient free vo'lume. In: the pressurize _r to receive the water. from the 50 tubes.-)

The SG ' tube.

.

" blows" were done in accordance with OP-5.

.

~1105

. Blew steam generator tubes twice per 01-1A and OP-5.1 Pressurizer level was increased > 260".

Operators then vented the head again to the refuel: pool via

'

. RC 274 -and 275 for about 5 minutes. RVLMS < lights went out.

1.ocally, gas was noted to be venting-for most of Sf minutes

-

j{

before water was observed :1n the) tygon tube.

,

..

Operators completed draining-the RCS to reduced: inventory; l

operations without further complication:.

8/31 Licensee formed a team.to evaluate possible voidfin reactor vessel

..

head.

!

.

,

/

k,

{

.))

.d-; '

'

'

,

-

,

,

,

.

.

-

i 2.0 Safety S_ignificance-

'

The STI was direct;d to assess and evaluate the significance of-the-drain.

down incident. The Team coaridered the likelihood of various hazards that might result from the conditions that were present or might have evolved.'

'q The hazards considered are as-follows:

a.

Core uncovery:

Found to be-highly unlikely.. The void _ in the head ~

region of the reactor vessel was a relatively small constant volume.

The temperature of the reactor coolant (118 F) made it likely that'

the volume consisted of non-condensables, probably air swept from the~

steam generator tub _es during. a previous fill evolution, 'but not-

,

vented from the CEDMs.

b.

Loss of shutdown cooling (SDC): Found to be' highly unlikely,.because the volume o# the void was not' increasing. The RVLMS alarm and other.

levels were ss follows:

Mcati,oj!

Elevation

,

1st RVLMS alarm 48.58'-(Same as bottom of pressurizer)

2nd RVLMS alarm 45.91'

3rd RVLMS alarm *

42.49'

4th RVLMS alarm 39.08' -(Same as) topiof. hot-leg)

The first RVLMS ' alarm had indicated on channel A andcthe first and l

second-alarms had indicate on channel B.

Assuming-the second channel

'

B alarm' was valid, both: the third andLthe fourth' alarms would have had to indicate before reactor vessel water level would be down to

~

-

l the top of the RCS hot legs.

l c.

CEDM damage:

Found to be unlikely.

The plant was shutdown with all control element assemblies fully inserted.

Thi s, wa s a potential equipment problem rather than a' safety problem. The licensee hadinot vented the CEDMs earlier, but they intended tot completei this. step

~

,

l prior to starting up the plant following.the RCP seal replacement.

'

d.

Uncontrolled drain of the. reactor coolant system (RCS): Found_to be I

l unlikely. The oper'ators detertrined the decrease in pressurizer level i

.

was consistent with the rate of -drainage from the. RCS. - The operators stated -they would have terminated' the ' drain ' down if, additional RVLMS

~

alarms had indicated that the void was getting larger.

~

e.

Steam void:

Found to be highly Eun11kely. The. VS was at '118 F-and the pressurizer'had.been coolsd down to <190F.

1,s described in Sec.

tion 1, there was very ~11ttla decay heat since the' plan't.had;operatad'

>

for only ~1 week in Apri.li1990 and was previouslylin an extended out-:

age and shutdown cool _ing was in service-providing: forced :flowL-cooling.

s b

if

,

.,

M,,

'

e s

.

.

-

t

,

i

'

.-

-

.

.

.

.

r

.

8-

.

'f.

Radioactive gas release:

Found to be possible.

The void was vented

!

.

before the air samples were counted. _ However, the: plant had been!

shutdown for about 15 months except for a one week -period _ in-April'

,

1990.

In addition, venting of the reactor vessel head was done to the bottom of the-refuel pool, where no personnel were present and_

j where the gas would be diluted substantially.

There was' also' the

'

possibility, although not -probable, that the-gas could have been

'

toxic.

The identity of' the. gas though, was not determined either

<

empirically or ~by testing before the gas was vented to the contain--

ment atmosphere.

Considering the probability anrJ severity of. these potential hazards,. a determination was made + hat the drain dowri incident ' had little Lor no

-

signi'ficance frnm a reactor-safety. viewpoint; but may have had 'the: poten-tial for a personnel safety problem as noted in (f) above. The. team con :

cluded that the failure to assess the potential.- hazards prior? to' venting-

'

was a weakness in the response-to the event.

The. licensee should respond

'

to this in writing.

3.0 g e stor Per' emance anet Procedure Adequacy

'

3.1 Backarcutnf The shif t or13anization at Calvert Cliffs. consists cf a shif t. super-intendent, who has overall responsibility. for operations, a control room supervisor, who directs contrsol room operators ifor both units,

_

.,

and senior react.or operators assigned to various jobs.

During x this

>

e ven t,, the involved personnel included the :SS, the CRS, and the -SRO -

and RO that were dedicated-to the. drain down operation. ~

g Although the shift ler.jth is 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />, the operators' generally work'

for :10 hours to r rmit an overlap ~ of shifts and. other activities.

' The oncoming shif t routinely becomes familiar with 'the-~ state ?of the plant and operations in progress. - For example,;in this. incident, the'

crew that was oncoming for the 0800-1600- shift. began system walkdowns around 0700 hours0.0081 days <br />0.194 hours <br />0.00116 weeks <br />2.6635e-4 months <br />.

~

'

The shift supervisor on the first shift (1200-0800)- had many-years.

experience in! that position -and had worked 1 severaliyears with that

..

same crew.

Th-SS was working his sixth day in?a row, the first-

'

three of'which he had worked 12 hour1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> shif ts.

The:CRS and-the dedi a cated SR0 and RO on the first shif t - had :lessT experience Lin those'

~

.

positions-(4 months, 8. months,:and 4 months, respectively). On the:

second shif t, the =SS and ' the dedicated ' SRO and R0"hadl2-years,19

'

months, and 7 years'in'those-positions, respectively.

<

The Team' reviewed ' factors - scch ' ah operator 1 experience, ; and, incident'

time ' and duration, to: determine: whether? theya may have influenced-

?

performance during this 1.1cident.

It? was! determined that any such

l influence was not apparent.

'

r l

'

i l'

.I*'

I

/

t

A !

.f.'

..d.--

y W.j,

,

,

=!

"(

i

,

Vp:.-

-

.

.

,

.

3.2 03rator_Pgfgmance The Team evaluated the actions of the operators in response to:the _

void formuletion.

This evaluation' was based on an examination - of plant oper4 ting drta such-as recorder traces, alarm printoutst and.

operator logs. -Interviews and the written statements of control room and, other ' licensee personnel were also useu for. this evalu ation.

The SS on the.first (1200-0800) shift stated his actions were influ-

.

,

enced by his concern for working near solid plant conditions, :1.e.,

without a steam : bubble in the pressurizer.

He said he had been

involved in a plant overpressurization in the past that. occurred dur-ing solid plant operations. He ;also stated he felt verbatim compli -

ance with the draining procedure was required, so that--he had to drain to less than.170" before venting to the quench tank.

He also said the RVLMS indications :seemed questionable.

He thought that the, void was an air. bubble, as had happened previously-on" Unit 2.

The operators checked the change in - pressurizer -level"during - the ' drain down was consistent with the flowrate. out of the 'RCS.

The 'SS halted -

-

the drain down when requested to do so by the operators from the oncoming shift.

The decision to continue draining when t'he void. ;was' initially

'

recognized was based on information avajlable at the time. The oper-ators judged that. the void was.not an immediate threatito nuclear safety.

They were concerned with present' solid planticonditions and wanted to reduce pressurizer level to below 170" andf exit this condi-i tion.

The decision to recommence the - drain ~ uas ' concurred on -by1

.

operat'ans management after estab'11shing -additional._ meansJ of RCS '

level indication.

In addition, they had;information from the: RVLMS system engtneer on the problem with that system (153" and 185" channel B level alarms were jumpered together) and had! installed the refuel level instrumentation to support the other Pzr 'and RCS level data.

• The venting of the void gases on the second: shift was Ldone. without first obtaining addittorial information onithe gases'. through chemical and radiological testing.

The team noted the followihg strengths in. operator performance:

1.

-Quick. recognition of the void indication on RVLMS'and the deter-r mination that the void was not' a steam bubble,t causediby somel unknown drain of the reactor vessel, nor lthe inadvertent intro-

_

duction _of a-gasito the RCS from some other activity.

?

1-

• s

'

'

e

,

,

.

,.

-

.t

,

,

'

i

-

[

'

i_i. The operators clearly understood that' if the void increased in -

size, there was a potential threat' to lose shutdown-cooling..

Therefore operators' monitored void indication and were: prepared

to take action if growth was indicated.

.,

iii. The operators were prepared to stop draining the RCS.if another RVLMS alarm occurred. They determined that there was no nuclear safety problem at the time.

The team noted the Qllowing weaknesses in ciperator performance:

1.

Operators vented the' reactor vessel: head without understanding.

the root cause of the void..No ef fort was made' to understand the. chemical and radiological. content of the gas prior to vent-

,

ing. Adequate personnel safety precautions were not taken-.while-venting this unknown gas.

-

11. The operatorsused " journeyman knowledge" to -perform the valve manipulations to vent the reactor vessol" head rather thane per-forming this evolution in accordance with an approved procedure.

'

This weakness is. discussed in greater detail' in? Section. 3.3 below.

iii. The operatorsL were not aware of the effects of a temporary mod-ification on the RVLMS

"6" channel indicationin response ~ to a-void in the reactor vessel-head..This confused.the operators-when the first-indicator was lit on channel

"A" and the the.

first two were lit on. channel

"B".

The Team noted :that - the cause of this seemed to' be~ procedural? rather thant an1 indicator of poor operator performance (this is further discussed. in Section 4 of.the report).

The Team alsou noted..that-subse-quent to the void. problem,. operator, training - on temporary modifications was performed.

3.3 Procedure Adequacy The Team evaluated the adequacy of the prccedures associated 'with

!

.

this event. ' These procedures 'and operato'r usez of them were reviewed.

The procedure in use.,at the ~ time --of the void formation was OP-5

" Plant Shutdown from Hot Standby, to Cold Shutdown". OP-5 refers the.

operators to other procedures such: as operating instructions (OI's).

The operators also referenced the annunciator iresponse' procedure for the " Reactor Vessel Level Low" annunciator.

i

,

i s

.!3

!

,

.

,

- 4 h

I

.

()

.

,

,

'

,

,

^

.

,

,

The Team noted that both OP-5' and 01-1A section VIII, for steami

-

generator (S/G) tube blowing, contain steps to vent the,. reactor vessel head,? voids formed while blowing water from the 5/G tubes.:

i The linL,ps dit hr between the procedures in that 01-1A' vents : the '-

'

head directly to de containment atmosphere via manual, valves while.

i OP-5 directs a vent of the head and the pressurizer to the -quench tank via solesoid operated valves.

The operators attempted a vent

~

per OP-5 but surmised that this-was not effective since the reactor-

,

coolant system was. vented to atmosphere. and the-~ quench tank-was 'at about 7 psig using a ~ nitrogen overpressure.

The Team. observed thatu 01-1A does not contain precautions for possible radioactive releases:

~

to the containment.

The Team did assesu + hat these. precautions:

should be normally contained.in the special t rk permit for-the job -

'

performed. OP-5 and 01-1A both indicate that the: refuel. levelfindi--

cation system should be in service before the step is reached in OP-5

.

l that actually places it in service.

This forces the operator to use-his judgement in determining ; when to place this indication inL ser--

vice. The " Reactor Vessel Level ; Low" annunciator response '. procedure indicates RCS draining as a lpossible cause. for ' the alarm : but the actions provided are applicable sto void _ formation in the ' vessel head-as a result of an accident where a steam void was formed. Thus this?

procedure ' did not provide guidance to the operators'-- during this event.

The Team noted that.there-were minor procedural' inadequacies

. '

which contributed to the problems tha,t. the operators :experiencedi while responding to this event. 1The SIFT also noted e these same problems.

The licensee should evaluate this more!~ thoroughly and-

,

correct the inadequacies.

There was an apparent difference in operating philosophies between the two shif ts involved in _ the application of' ; the ' administrative.

requirements of CC.' J00."Use of. Procedures". OneE shift felt that a.

change-to. OP-5 would be needed to ' skip thosei stepsanecessary ; to

-

immediately vent the reactor vessel head,- yet the othbrJ shift felt-

'

-

that they could skip steps' in.0P-5 'and vent =the headt to Lthe quench -

s

tank. When this did not work that-shift then used " journeyman know-'

i ledge" to perform actions not stated in,0P-5. to; vent the head to ' the

containment atmosphere.

In CCI-300 ' " journeyman. knowledge" - i s Enot specifically defined and is reserved forithe " simplest off manipula-

!.

tions".

Guidance in CCI-300 indicates that;" actions: that" alter the

.

'

operability of equipment which) affect plant' safety,n plant availabil-

!

ity, or radiological barriers, shall' be' performed-in 'accordance with

.

,

approved' procedures".- The Team is concerned withethe, application _ of L

" journeyman knowledge" for this venting _ operation since adequate -

precautionary measures' were not taken to. use-- plant; chemistry, radio-logical controls, or' other engineering ' support Lto" protect personnel.

in the containment from the potential adverse consequences"of ' venting:

'

an unknown gas from the reactor vessel head.

.

b t

't

,f y

l

!

'l y

)

'

[

.-

'

.

.

.

Conclusion The Team concluded that initial operator' recognition and verification of the void formation was a strength.

However, the Team was con-cerned with the decision to vent the reactor vessel head without

.

first understanding the cause and thus not ensuring': adequate person-nel safety precautions for venting an. unknown gas.. The: Team also concluded that the use of." journeyman knowledge", as applied to 'the venting operation on August 30'due-to the previously mentioned poten-tial personnel safety considerations, was a ' weakness in the licen-

see's response to this event and that the licensee should respond to this weakness in writing.

4.0 Reactor Coolant Sy_ stem level Instrumentation

,

Pressurizer (Prz) level is Indicated in the control room lby.three level '

'

'

instruments.

Prz level is also recorded from ' any; onefof, these instru-

,

ments.

In addition, a_ reactor vessel Llevel monitoring system (RVLMS) is J

also available.

RVLMS has two channels ( A and B). - Each chann'el has _eight heated Junction thermocouples-that are: sertically or,iented in: the reactor vessel head area

.

to provide eight discrete level indication ? points ' above the top of the fuel (185", _153", 112", 71", 50", 29",

.19",,

a nd.10").

If J1evel drops.

.

below one of the points,. an Lindicatird -light will illuminate on the' con-

~

trol. room panel and the reactor vessal water level' low annunciator.' will-alarm.

If level continues - to. dror, additional' lights'Qillilight as level i

-

goes below the sensor.

During' reactor vessel : drain-down-an Hadditional

-

refueling level indication and tygon tubing is placed into service.

The t

i latter two indicate level above the bottom of-the reactor coolant system l

'

-

hot leg.

During.this occurrence, good agreement was achieved.between the-t three Prz level channels, the refueling level indication iand_ the tygon

'

tubing. Only the. RVLMS indication 1was ever questio ed, consequently, the-

'

remaining discussion is limited to the RVLMS.

Inspection findings indicate all operators'have receivedLtraining: adequate-to understand the. operation of ~the system. A detailed system description 1.

providing system operatinn and system design characteristics.has be'en pro-

)

vided for plant personnel' use.= A system engineer-responsible for main--

!

taining and monitoring,thefsystem performance has-been assigned. ; Training was provided to instrument technicians responsible for the : system. main-

-

"

'

'

tenance.

Both the system engineer and the instrument. technicians' were found to ;be knowledgeable of; RVLMS. 'Also, adequate. operating' procedures have.been provided, 'with the _ exception -that. thel reactor > vessel: water level i

low alarm response procedure did -not-. provide; guidance-for thisiunforeseen

'

occurrence..This is under review by the licensee.

l

,

l-l l

(

'

..

~?

I'

, ',

,

)

>

>

.

,

-

~,

,Oi

.-

,

!

/

.,;

,,,

,

..

.-

<

.

.?

-

13'

l

,

Anthough operators did respond appropriately to the RVLMS indication, they expressed a lack of confidence in the system.

This lack of confidence

>

resulted from the' many spurious indications received and the_ length of time the RVLMS trouble alarms had been ~ indicated.

A through review of maintenance records and discussions with personnel indicate-frequent main -

tenance is required. RVLMS work requests are generally 'given a low. pri-t ority, which results in the trouble alarms being. displayed in the control'

.

room for long periods-of time.

The-troubles experienced were'found to be-t generally associated with only single sensors a'nd consequently - af fected

only one of eight indications in a channel. On no occasion, was; an entire channel found to be inoperable.

At the time of the void formation, a' temporary modification was-install'ed on the B channel of the RVLMS.to correct a failed. thermocouple-in a sen-

sor.

This modification had the effect. of causing the 'second low level'

i lite. to illuminate along with the first.- This was judged to be a"conser-vative modification. However, the operators were not_ aware of-the modifi-cation or of its effect and were consequently confused by the' indication.

In spite of the conflicting. alarms on RVLMS', the operators Lresponded con-servatively by believing that the' lowest indicated level was= valid. This

.

,

modification had been installed ~1n accordance -with CCI-117I, Temporary Modification Control.

This calls into question 'the adequacy; of CCI-1171

,

to be able to keep operators informed of the effects -of: installed imodifi :

cations, particularly modifications which will be4 installed for - long periods of time.

Conclusion Reactor coolant system level indications, ' including RVLMS, appear to: have adequately indicated level. Excepting the effect of the temporary modifi-cation, adequate training as well,as operating procedures for RVLMS were-provided to operators and technicians. Operators expressed a lack of con--

fidence in RVLMS due to frequent indications of problems.

Considerable

RVLMS maintenance has been required, however, no' total-' channel failure has-occurred. The confusion caused by the temporary modification to 'RVLMS was

'

a concern of the Team.

The Team' concluded ~that the licensee's. controls for temporary modifications was a. weakness which contributed to this event and that they should respond to this weakness in writing.

5.0 Reportability

.)

The inspectors reviewed the licensee's : event reporting p'rocedure,10 CFR -

50.72 and NUREG-1022, Supplement 1 and' determined. that L the ; unexpected bubble or void 'in the RCS vessel head was not' iminediately reportable.

This is based on the fact that the. event did not involve:

entry into an emergency condition; a technical-specification. violation; 'an "unanalyzed'F

plant condition (reduced RCS inventory operation's are analyzed with appro'-

!

.priate procedures.avai.lable); an event that' 'alone could have prevent'ed, j

!

<

'

.

.

,

4f SM

. p

,

. [i

'

i i

<

'o-

.

.

,

j, i

.,

.

,

,

,_b

-

'

fulfillment of' a safety function; or an unplanned release of r'adioactiv-ity. Although the operators did not expect void, formation in the reactor -

pressure vessel head at the time' and could notl Immediately ' determine. the

'

cause when RVLMS indicated :the; void, the operators 'did not dismiss the indication and maintained the/ plant'.in a condition -where its -principal =

safety barriers were not seriously degraded.

6.0 Licensee's Investigation

,

On Friday morning, August 31,'1990,- 'the NRC: Restilent! Inspectors began to gather information about the unexpected void.: formation during; the.RCS draining on August-30,.1990. -The licensee initiated a problem. repert, at 9:00 a.m. on August? 31, 1990- to initiate a review-of the event to deter-mine what, if any, additional. - actions. were-necessary.

Later on August 31, 1990, af ter additional communications between? NRC Regional land d

licensee management, the - station' management initiated a ' Significant :.Inci-dent Finding Team (SIFT)..The' SIFT. process was: found :to-be Ja thorough assessment of the event.

Portions of _ the SIFT' investigation' of-the' event-were observed and the SIFT presentation to the -.POSRC was attended by the-

NRC Team.

The conclusions, sequence of events and; recommendations for

corrective actions by the SIFT were in. substantial. agreement with that of the NRC Team.

7.0 Exit Interview

,

i

...

An exit interview was conducted on September'7; 1990, with representatives

"

of the licensee staff. The Team presented its ' findings 'atithet _timek No written materials were provided to the. licensee.

f I

'l

-

i i

V l/

.{

<

,-

'

(

i J

, e.]

'r

'

' f

_

'

,

>

'

l l{

\\,

'

'

'

!

it

,

,

,.i

,,..

<

-

.

.o

,

.

,,

w[h

,,

i' '

.._

,1

a

.

<

'

y y,

.., -.

y

.,

,

,

,

.,3

,

..,

' 4-

.

.e

-

...

,

,

i

,

n y

<

c

>

,;

_

m

.;

'l TABLE l'

!

'#

-

,

-Aeronyms

- <

.

)

AGS-N0-

- Assistant General-Supervisor,: Nuclear Operations j

BG&E Baltimore. Gas and Electric. Company;

'

tl CCI'

Calvert Cliffs Instruction _._

'{

'

CEDM Control Element Drive Mecha'nism

_

'

'

'

'

-CR0

' Control Room Operator

+

CRS Control Room Supervisor-

,

,

,

CVCS Chemical.and Volume' Control System <

GS-N0 General Supervisor, Nuclear Operations

,

,

LPSI Low Pressure Safety Injection

'

OIL

_ Operating Instruction-

'

M

'

OP Operating Procedure

.

..

'

-

POSRC.

Plant Operations Safety Review Committee-

PWS Plant-Watch. Supervisor

.

,

'

.

,

Pzr Pressurizer-

,

RC Reactor Coolant-

'

RCP Reactor Coolant Pump,

'

.,

,

RCS.

. Reactor Coolant System

.t

,

R0 Reactor Operator:

RVLMS ReactorVesselLevelMonitohingSystem,

,

-

'

SDC

' Shutdown Cooling

SG Steam Generator

'

SIFT

- Significant Incident _Findin'g Team

,

,

SR0 Senior Reactor Operator d

. SS Shift Superintendent t

'

'

g.

.

' STI-Special; Team Inspection

' T-Mod Temporary Modification'

t

,

.

.

._,b

.)

i,

'p t

'i

,

(

.y r

e r

-..

-

'

'\\

I is

--!

.m.

-

.'. l

.

~~

'

Lij

!

?

s

-

-

-

o

.

'

'

b;

N

'if

/s O

j( j

.lu

' (

(h!

' il

.'.

,

,

a ',.8w y,Nf.Il I

\\,'i -

' ' '

-

', _ _

.

,

Ct

._

~

i " M('

,

u M

's it

' d,f

'

'

'

'

,.,;.

.1

.

.

l

,

,,,

,.

.-

- p,, x;c ~

_.,

<,

,

>

,

<"~

jQ)g; '

6d c

' '

'

n _ :.., ',) 3-,

• <,

w

'

ag

,

ll

!:

t t

.

I

%

'

,

. :

,

., <

,

r

.,.

x;

<

.-e

,

.

6).

g

.

'

.

.y.

,

.

,

Reactor. Coolant Systen Levels i

..

..e..

,

' Pressurizer RefN11ngPool.

'

'

<

t Level Level

• BBM Top of Pressurizer 81.00 ft.-

Indicated Top of Pressurizer'

360 inches 78.00-ft.

j Top of Refueling Pool 252 inches 69.00 ft.

-

Normal Refueling Pool' Level 228 inches 67.00 ft.

,

UGS Deck Plates 204 inches 65.00 ft.

-

Thimble Support Plate 156 inches 61.00 ft.

ICI Installation 108 inches:

57.00 ft.

TECH. SPEC. Refueling Pool Level 104.4 inches-56.70 ft..

Core Support Barrel Lift Rig Install 84 inches.

55.00 ft..

(

Bottom of Pressurizer & lst RVLMS Alarm 0-inches

48.58 ft.

_185 inches'

i t

2nd'RVLMS Alarm 45.91 ft.

.153 inches

,

Top of UGS-45.00 ft.

Reactor Vessel Flange 43.83.ft.

3rd RVLMS Alarm 142'.49 ft.

112 inches il

.

'

Bottom of S/G Primary Manway L40.33 ft.

d Top of Hot Leg, 4th RVLMS' Alarm & RCP Seal Maint.

39.08 ft.

71 inches

-

,

Nozzle Dam Installation'

37.58~ft.

.

Middle of Hot Leg & 5th RVLMS Alarm 37.33 ft.

50 inches

.

Minimum level to Drain 2CS

~36.83 ft.

Bottom of Hot Leg & 6th RVLMS Alara 35.58 ft.

'29 inches i

7th RVLMS Alarm 34.74 ft..

19, inches:

M

,

'

Last RVLMS Alarm 33.99 ft, 10: inches

,y.

Tcp of Cwe 33'70 ft.

.

,

I,

-

_

l '

f ty

,

'

,

,

ge

,

-<-

j

- -

,.

.

,,

-

.

.

-:

_

RCIO5SV ggcf.

s TO GUENCH TANK

'

-kCIO65V

- !

'

Om

?

..

.

RC200RV4 b

RC2OIRV

!

~o-c PS6466CV

.

,

te y ',.

'

RC216

!

78'

/

\\

'-RC215

'

~

f

)

RC2761 :

.

I

~

_r_

l_

-

O FGl22 l

'PZR RCIO4SV-: ;-

%TO VENT

.

t

.RCIO3SV.1 [

1;;RC275 1:;RC274

"

-

.._

.

1 :- RC I Oa.-

X RCIO3-

.N

/

"

-

!

M4858'

-

,

EliQ 4591 S/G

,

'

'

.

,

'

,

!

i

.

180f4249'

m

.

-

-

,

~

R/V.

_

_

~j

.,..

,

.L.,-

~

~

_4TH 3900'

~

3733'5Iti

~

.

..

.HOTJ-LEG

~

-_

,

I78'JINDICATED TOP LOFiPRE$SURIZER -

l 33.70'

-

i4658'ilSTi RVLMSEALARM..

TOP OF' CORE'

--

.

45.91' 2ND L42.49' 3RD 7.3 5T

-

.

-