ML18005A214

system

then

automatically started

due to the engineered

safety feature circuitry

sensing

that both main feedwater

pumps were tripped and the auxiliary

0

3

pump control switches

were in the "auto" position.

The two

motor-driven auxiliary feedwater

pumps

started

as required

and the

control

room operators

stabilized the plant parameters.

The cause of

the

event

was

investigated

and

a

Work Request

(WR/87-BDIP1)

was

initiated to replace

the

broken valve

stem.

The "B" main feedwater

pump was started

and the plant heatup continued.

Failure of the

ERFIS Plant Computer

On September

27, at 2:03 p.m., the licensee

declared

an Unusual

Event

(UE)

as

a result

of the

loss

of the

Emergency

Response

Facility

Information

System

(ERFIS) plant computer.

The

system failure was

attributed

to

the

data

disks

failing

on

both

the

"A" and

"B"

computers.

This failure

had

no

immediate

impact

on

the

safe

operation

of the plant.

The control

room notified the state,

local

and

NRC officials.

The

computers

were

repaired

and

the

UE

was

terminated at 3:02 p.m.

Unusual

Event - Vent System for the Reactor Coolant

Loop

On October 9,

the

NRC Duty Officer was notified of an Unusual

Event

(UE) which had, been declared

by the licensee

at 6: 10 a.m.

The

UE was

declared

as

the result of an apparent

malfunction of valves

on the

reactor

coolant

system

head

vent

system

(RCSHVS).

The

valves

malfunctioned

while

conducting

an

Operations

Surveillance

Test,

OST-1043,

"Reactor

Coolant

System

Vent Path

. Operability,

quarter'ly

Interval".

The

OST

allowed

opening

only

one

valve at

a

time to-

prevent

a

flow path

out of

the

reactor

coolant

system

(RCS).

However,

due to the design of the valves

and the test

sequence,

the

test

resulted

in the

inadvertent

actuation

of the

two valves

in

series,

allowing reactor coolant to vent to the pressurizer

relief

tank (PRT) and to the containment

atmosphere.

Throughout

the event,

the plant was operating at approximately

91 percent

power (Node 1).

The inspectors

interviewed the personnel

on shift at the time of the

test,

and

other

licensee

personnel,

and

evaluated

the

licensee's

preliminary incident report

and established

the

sequence

of events.

The

valves

discussed

below

are

identified

on

site

drawing

CPL-2165-S-1301.

Each of the valves are of identical design

and are

solenoid

actuated,

pilot

operated

globe

valves

manufactured

by

Target-Rock.

The

normal

expected

cycle closing time for each valve

is'bout

two seconds.

The chronology is presented

in the following

paragraphs.

Inservice

inspection

(ISI) personnel

requested

a test of selected

valves

( 1RC-900,

901,

902

and

904)

to quantify

suspected

valve

degradation.

Two other

similar functioning

valves,

1RC-903

and

1RC-905 did not have

suspected

degradation

and,

ther'efore,

were

not

tested.

The control

switches

for these

valves

have positions for

"shut pull to lock",

and "open".

When not in "shut pull to lock",

the

switch

spring

returns

to

a

neutral

position.

The

control

switches for all six. valves were in the "shut pull to lock" position

which is

normal for Power

Operations

(Node 1).

OST-1043

is the

routine surveillance

procedure for measuring

valve response

time,

and

it was

used

to test

the

selected

valves

~

Permission

to perform the

test

was granted

by the shift foreman,

as required

by the procedure.

At approximately

5:00 a.m.

on October

9,

was satisfactorily

cycled.

A few 'seconds

later,

was

opened

and the operator

observed that the posi;tion indication light for valve 1RC-904,

which

was

in

the

"shut

pull

to

lock" position,

also

gave

an

open

indication.

1RC-900 was quickly shut

and

1RC-904 immediately closed.

The

operators

continued

with the test,

since

they

were

able

to

immediately shut both .1RC-900

and

1RC-904.

The shift foreman

was not

'consulted

about the opening of 1RC-904.

The shift foreman

was not- in

the control

room at this time.

Due to the

unexpected

flicker of the

open light for 1RC-904

and the

prompt closing of 1RC-900,

a response

time for valve

was not

obtained,

and at 5:03 a.m. it was cycled again to obtain the response

time.

When

was cycled full open,

both

and

1RC-905

opened.

During an attempt to close

the valves

the control

switches

were put in the

shut position momentarily

and the valves

remained

open.

( 1RC-904

and

1RC-905 did not close

because

they

had

opened

from upstream

pressure,

not the control

switch).

Up to this point,

the evolution was being carried out by two senior reactor operators.

Approximately

30

seconds

after

opening

1RC-900,

the shift foreman

returned

to

the

control

room,

unaware

of

the

event.

Shortly

thereafter,

a third operator

became

involved

as the first operators

tried to close

the three

valves.

The third operator

stated

that

would close if the control

switch

was

held in the "shut"

position until the valve indicated

closed.

This

was

done

and all

three

valves

closed;

and

1RC-905

closed

immediately after

was observed

to be closed.

The licensee's

best

estimate

of

total

time the valves

were

open is

two minutes,

based

on computer

stored

data.

Small

changes

in pressurizer

level

and pressure

were

observed while the valves were open.

Shift personnel

then discussed

the event in order to determine if it

was

safe

to continue

testing;

this

discussion

lasted

about

ten

minutes.

Throughout the discussion,

the shift foreman

was unaware of

the results

from the first cycle of 1RC-900,

which

had occurred at

5:00 a.m.

The items discussed

were the

need to proceed

and quantify

the response 'time of 1RC-900

and the balance

of the untested

valves,

the confirmed ability to close

a valve by holding the control switch

in

the

"shut"

position,

and

the

fact

that

OST-1043

was

the

established,

approved

procedure for quantifying valve response

times.

The shift foreman

concluded that it was

safe to proceed

with valve

testing

because

of the

demonstrated

ability to close

the valves.

Further consultation with the next level of plant supervision

was not

done prior to proceeding with the testing.

The

remaining

valves

were

tested

commencing

at

5: 15 a.m.

and

concluding at 5:21 a.m.

Each

time

a vent valve

(1RC-900,

901,

or

902)

was cycled,

both block valves

( 1RC-904

and 905) indicated

open.

The results

from the test were as follows:

was fully opened;

and

905 open lights came

on.

The

operators

immediateTy

closed

after full

open

,indication;

and

905

immediately

closed.

The closing

stroke time was 5.2 seconds.

'RC-901

was fully opened;

and

905 open lights'came

on.

The

operators

immediately

closed

1RC-901

after

full

open

indication

and

905 closed

immediately.

The closing

stroke time was 1.0 second.

1RC-902

was fully,'opened;

and

905 open lights came

on.

The

operators

immediately

closed

1RC-902

after full

open

indication;

and

905 closed.

The closing stroke time was

9.7 seconds.

At approximately 5:20 a.m.

the licensee

completed

a quick assessment

of the

amount

of reactor

coolant that

had

been

lost during

the

testing.

The

amount

was

estimated

to

be

about

150 gallons;

a

subsequent

assessment

in the licensee's

preliminary incident report

concluded

the loss to be approximately

200 gallons.

The majority of

the

loss

was

attributed

to the'econd

cycling of

(at

5:03 a.m.).

The shift

foreman

reviewed

the

Emergency

Plan

procedures,

and

concluded that declaration

of an Unusual

Event was appropriate.

His

decision

was based

on the fact that the Technical Specification limit

for reactor

coolant

system

leakage

had

been

exceeded

during that

interval.

The

declaration

was

made

and

terminated

effective

6: 10 a.m.

State

and local officials were notified in accordance

with

the applicable

procedures.

The licensee

investigation

of the event

began during the morning of

October

9.

The investigation

involved the

manager-operations,

the

operations

'supervis'or

and the shift foreman,

who was recalled to the

site.

The investigation

included

a review of the

logs

and

records

from the event,

interviews with selected

members

of the shift crew

and

a technical

evaluation of the response

of the valves.

The

investigation

concluded

that

the

actions

by the

operators

.to

proceed

with OST-1043 resulted

in

a challenge

to the capability to

maintain adequate

water inventory for the reactor coolant

system

and

that

prompt

measures

were

required

to

address

this

incorrect

personnel

decision.

The

oncoming shift (at 6:00 p.m.

on October

9)

was

briefed

by plant

management

on the

event

and the errors

made

during the testing of the

RCSHVS valves.

These

presentations

were

made

by the plant general

manager,

the manager-operations,

and the

operations

supervisor.

Each shift attended

similar briefings prior

to assuming

a watch.

The shift

foreman

involved in the

incident

was

counseled

on

the

seriousness

of the

incident

by the

manager-operations,

operations

supervisor,

and separately

by the plant general

manager.

The shift

foreman

was also required to assist

in the incident investigation

and

to participate

in the briefings for shift personnel.

The licensee

also initiated

a task force to investigate

the operation

of the

valves.

Shortly after contacting

Target-Rock,

the

valve

manufacturer,

the

licensee

learned that the behavior of the block

valves

( 1RC-904

and

905)

had

been

observed

at other plants.

A

technical

paper

was

published

by ASME, addressing

the likelihood of

this

valve behavior.

The

paper

is titled,

"Spurious

Opening

of

Hydraulic-Assisted,

Pilot Operated

Valves - An Investigation of the

Phenomenon",

and

was

published

as

ASME Publication

81-BVP-39

in

April 1981.

The licensee

determined

that this information

had not

been available to the Shearon

Harris site staff prior to this event.

As

a result of the technical

review, it was determined that actuating

1RC-904 before operating

1RC-900,

901,

902 or

903 could result in the

formation of an air pocket

above

the plug in the valve.

The air

pocket

could

cause

the valve to

come off of its seat

when

shocked

with a

sudden

pressure

surge

from the

opening of an

upstream

valve.

Also the control circuit for 1RC-904 would allow the valve to remain

in the

open

position if the

control

switch

was

in the

"normal"

position,

when the valve

came off its seat.

A technical

evaluation

concluded

that

by physically rotating the

valve

body

so that

the

solenoid

was below the horizontal,

could prevent the formation of an

air pocket.

Subsequently

the licensee

changed

the

sequence

of valve

testing

in

OST-1043

to caution

operators

to

open

the

downstream

valves

last

(OST-1043

Temporary

Change

07016,

dated

October 19).

Because

of the

consequences

of valve fai lure,

the licensee

is also

pursuing ISI relief and

a Technical Specification

change to decrease

the testing

frequency in order to avoid cycling of these

valves while

the reactor

coolant

system is pressurized.

The licensee's

plans

are

to

complete

the

procedure.

changes

and

design

changes

prior to

start-up

from the

current

outage.

The

licensee

has

conducted

on-shift training

so that shift personnel

are

aware of the technical

details

on the valves'nadvertent

operation.

The event

was caused

in part by the actions of the shift personnel

in

carrying

out

their

responsibilities

as

licensed

operators.

Specifically,

Technical Specification 6.8. la

requires

that

plant

operations

be

controlled

in

accordance

with

administrative

procedures.

The

plant

procedure

OMM-001,

Rev.

3,

"Operations

- Conduct

of Operations",

step 3.2.3.4

requires

that the plant

be

operated

in

a safe

manner at all times.

The action to proceed

with

the testing of RCSHVS valves after the 5:03 a.m.

event

on October 9th

was

not in accordance

with the requirements

of OMM-1 as described

above.

This is

a violation, "Operators Manipulating Valves Which Were

Known

to Operate

in an Unsafe Manner", 50-400/87-37-01.

1A-SA Electrical

Bus Blackout

On October

10 the plant experienced

an automatic start of the

1A-SA

emergency

diesel

generator after loss of the

1A-SA electrical

safety

bus.

The diesel

generator

automatically

started

and picked

up the

required

"A" train loads.

When this event occurred the plant was in

Cold Shutdown

and

had just begun

a plant outage.

Implementation of a

modification

had

required

personnel

to deenergize

one

of the

two

parallel

lines

for the

"A" start-up

transformer

to

allow

the

installation of

a

new, electrical

cable to the

switch yard

from the

relay

cabinet.

While

performing

the

work

inside

the

cabinet

(hammering

and drilling), licensee

personnel

inadvertently

jarred

protection relays causing

a trip of the remaining breaker for the "A"

. start-up transformer,

resulting in an "A" bus blackout.

The licensee

restored

offsite

power

and

secured

the

diesel

generator.

All

equipment

functioned

as

designed

with the

exception

of the

"A"

emergency

service water screen

wash

pump which did not start

when the

diesel

generator

assumed

the

bus

loads.

The

licensee

is

investigating this event

and

the

inspectors will follow up

on thi-s

event during

a future inspection.

Failure of the

RHR Suction Valve Interlock

On October

15, while the plant was in Mode

5 (Cold Shutdown),

the "B"

train suction,

1RH-1

and

1RH-39,

valves

in each

RHR loop suddenly

closed.

The running

RHR pump was stopped

by the operators

to prevent

the

pump from being damaged.

The first event of this type

happened

at about

7:40 a.m

,

and

the

second

event occurred at 9:22 p.m.

The

RHR suction valves were shut for only fifteen minutes during the last

event,

and for less

than five minutes

during the first event.

The

inspectors

reviewed the

TS Action Statement

3.4. 1.4. 1,

observed

the

plant condition,

which

was

Cold Shutdown,

and determined

that the

licensee

complied with the applicable action statement.

The licensee

reported

these

occurrences

to the

NRC duty officer on October

16 as

a

requirement of 10 CFR 50.72,

using

NUREG 1022

as

a guide.

The first

event

was reported to the

NRC duty officer as having occurred

due to

an electrical

spike caused

by the technicians

who were performing

a

test

on the interlocks for the two

RHR suction valves.

The test

was

being

conducted

to satisfy

the

requirements

of Surveillance

Test

OST-1071,

Rev.

0,

Step

7. 1.

This

test

required

lifting the

electrical

leads in the

process

instrumentation

control

cabinets

to

.

prevent

the automatic

closure of residual

heat

removal

(RHR) valves

RH-l,

RH-2,

RH-39

and

RH-40.

The

inspectors

interviewed

the

technicians

and were informed that as of October

23 the cause of the

valves'losure

was

yet

to

be

determined,

and

is still

under

evaluation

by the

licensee.

The

licensee

plans

to

complete this

evaluation

prior to returning

the

plant to

power operation.

The

technicians

also

stated

that they

had

repeated

all of the

steps of

OST-1071,

starting

from Step

7. l.c,

and

as

part of the further

evaluation

they will rerun the entire Step 7. 1.

The next time the

two

RHR suction

valves

closed,

at 9:22 p.m., the

cause

was attributed to the failure of the test instrument which was

being

used during the: OST.

The instrument

was

found to

have

weak

batteries,

which

had

apparently

discharged

during

the

test's

duration.

The discharged

batteries

caused

the test instrument output

to be low.

The low output signal

allowed the closing circuit for the

"B" train

RHR pump suction valves to actuate,

and the valves closed.

After the operating

RHR loop

was returned

to service

the

OST

was

continued.

The

instrument

was

replaced

with one that

had fully

charged batteries

and the test

was completed successfully.

One violation was identified in the areas

inspected'.

Monthly Surveillance

Observation

(61726)

The inspectors

witnessed

the licensee

conducting maintenance

surveillance

test activities

on safety-related

systems

and

components

to verify that

the

licensee

performed

the

activities

in

accordance

with

licensee

requirements.

These observations

included witnessing

selected

portions of

each

surveillance,

review of the surveillance

procedure

to

ensure

that

administrative

controls

were

in force,

determining

that

approval

was

obtained prior to conducting

the surveillance

test

and

the

individuals

conducting

the

test

were qualified in accordance

with plant-approved

procedures.

Other

observations

included

ascertaining

that

test

instrumentation

used

was

calibrated,

data

collected

was

within the

specified

requirements

of

Technical

Specifications,

any

identified

discrepancies

were properly noted,

and the systems

were correctly returned

to service.

The following specific activities were observed:

During this

inspection

period

the

licensee

requested

exigency

on

a

Technical

Specification

change

request

for *Survei 1'lance

Requirement

4.8. 1. 1.2.f. 11.

As verified by the licensee's

procedure

OST-1824,

this

surveillance

currently

ensures

that

during

a

load

rejection

test,

emergency

diesel

generator

voltage

does

not

exceed

a

maximum of 7590

volts.

The

licensee

requested

that this

surveillance

requirement

be

changed

to reflect

a

110 percent

value of the voltage at the start of the

test.

The original specification stipulates that "verifying the generator

capability to reject

a load of between

6200

and

6500

kW without tripping.

The generator

voltage shall not exceed'590

volts during and following the

load rejection".

This surveillance

is performed

to verify that

a

load

rejection of the diesel

generator

does not generate

an overspeed trip and

that

the

voltage

regulator

functions

correctly,

ensuring

the

diesel

generator

is available

immediately following a load rejection.

The 7590

voltage limit is

based

on

110

percent

of nominal

generator

starting

voltage

of

6900

volts.

During

the

most

recent

test

conducted

on

October

13,

the

licensee

experienced

problems with meeting

the

maximum

voltage limit of 7590 volts.

With the generator

connected

to the grid,

voltage

of the

generator

must

be

maintained

greater

than

system

grid

voltage to ensure

proper operation

of emergency

diesel

generator

within

its design

limits.. Therefore initial generator

starting

voltage

was

greater

than

the

6900 volts (approximately

7350 volts) due to low system

grid demand (higher

system grid voltage).

Starting at this higher initial

voltage of 7350 volts the

maximum voltage

reached

7850 volts,

exceeding

the

TS limit of 7590. volts'.

However,

the

voltage

did not

exceed

110

percent

(8085 volts) of the initial starting voltage of 7350 volts.

The

licensee

reviewed this specification

and requested

that the

NRC approve,

a

change

to

TS 4.8. 1. 1.2.f. 11 which would limit the generator

voltage to

less

than

or equal

to

110 percent of the voltage

on the generator

at the

start of the test.

The licensee

has provided

a detailed analysis

to the

NRC for review prior to 'approval

of this

TS

change.

The

resident

inspectors will continue

to monitor the status of this change

request

in

future inspection periods.

No violations or deviations

were identified in the areas

inspected.

Monthly Maintenance

Observation

(62703,

62700,

37700)

The inspectors

reviewed the licensee's

maintenance activities during this

inspection

period

to verify the

following:

maintenance

personnel

were

obtaining

the

appropriate

tag

out

and

clearance

approvals

prior to

commencing

work activities,

correct

documentation

was available for a'll

requested

parts

and material prior to use,

procedures

were available

and

adequate

for the work being

conducted,

maintenance

personnel

performing

work activities were qualified to accomplish

these

tasks,

no maintenance

activities

reviewed

were violating any limiting conditions for operation

during the specific evolutions,

the required

gA/gC reviews

and

gC hold

points

were

implemented,

post-maintenance

testing

activities

were

completed,

and

equipment

was

properly

returned

to

service

after

the

completion of work activities.

The inspectors

obtained

copies of several

Plant

Change

Requests

(PCR)

and

the applicable

Work Request

(WR) for each

specific

PCR.

These

documents

were reviewed to ensure that the

PCR had

been generated

in accordance

with

the

appropriate

engineering

document,

evaluated

technically

and

administratively

by the authorized

groups

and that personnel

implementing

these

PCRs

were qualified in accordance

with the licensee's

procedures.

Specific details of each

PCR reviewed are

as follows:

PCR-2171

Main

Feedwater

Pump

Total

Discharge

Head

Reduction.

Secondary

perturbations

have

caused

numerous

oscillations

of high

discharge

pressure

'and

low suction

pressure

trips of the

pumps

and

subsequent

plant trips since the start-up of the Harris plant.

These

instabilities

in the

system

have usually resulted in poor

control

valve

response

and operation

which in

some

cases

lead

to trips of the

main

pumps

(MFP)

and

condensate

booste~

pumps.

The

licensee

and

pump

vendor

have

reviewed

these

10

problems

in detai

1

and determined

proper operation of these control

valves

could

be obtained

by reducing

the discharge

pressure

of the

NFP.

The licensee

issued

two WRs for each

NFP,

87-BCJQl

and 87-BCJQ2

for

MFP

1A

and

87 BCJRl

and

87

BCJR2

for

MFP

1B.

These

WRs

implemented

PCR-2171 which required machineing the

MFP implemented to

reduce

the total discharge

pressure

by approximately

15 percent

in

total

discharge

head.

The inspectors

witnessed

the disassembly

and

reassembly

of the

NFPs

as authorized

by

WR 87-BCJQ2

and

87

BCJR2 for

NFP lA and

1B respectively.

The licensee

plans

on completing

post

maintenance

testing

and design verification prior to declaring

the

main feedwater

system

and control valves operable.

PCR-2173

- Heater

Drain

Pump

(HDP) Impeller Removal.

Due to the

previously identified secondary

plant pressure

response

problems

the

licensee

investigated

actions

necessary

to

make

the

heater

drain

system

more reliable.

Field investigation of these

pressure

problems

found that

by throttling the

manual isolation valve upstream of the

HDP level control

valve provided

a pressure

drop (approximately

80

pounds

per

square

inch)

reducing

the

likelihood

that

a

sudden

pressure

change

could

adversely

affect

the

heater

drain

system.

Discussions

between

the

licensee

and

vendor

(Ingersoll-Rand)

concerning

pump operability

and design capabilities

found that this

80

psi

drop

was

approximately

equal

to

one

stage

of the

HDP.

Therefore,

the

vendor

recommended

that

the

pump

be

modified

by

removing

the last

stage

which would reduce

the total discharge

head

by the necessary

pressure.

Removal of the sixth stage

from the

pump

necessitated

installing

a flanged volute extension

which would direct

discharge

flow out of the

pump

to

the

discharge

piping.

WR

documentation

to implement

and perform the necessary

work activities

were issued,

performed

and

are

in the review process

at this time.

These

WRs are identified

as

87-BCBNl and 87-BCBS1 for HDP

1A and

1B

respectively.

The

inspectors

witnessed

the

removal

of the sixth

stage

and

replacement

with

the

required

spool

piece.

Final

acceptance

'of this modification

by the licensee will be

performed

prior to end of the outage.

PCR-1829 - Feedwater

Recirculation

Oi ifi'ces.

This

PCR modified the

NFP recirculation lines

and logic for the main feedwater flow control

valves

(FCV).

The licensee

found that the downstream

side of the

NFP

feed

line orifices

were

experiencing

cavitation

due

to

a

sudden

pressure

drop prior to entering

the

main condenser.

'This condition

lead

to recirculation

flow control

problems

with the

system

as

designed.

The licensee,

was in the process

of implementing

changes

to

the

NFP

lines

and

control

logic

as

follows;

installation

of

additional

restrictive orifices

downstream

of the

FCV which will

reduce

flashing

upstream

of the orifice.

These orifices will also

limit the

amount

of recirculation

flow returning to the

condenser

when the

FCV's are full open.

FCV logic was changed

from "open"

and

"modulate" to that of "closed"

and ".auto".

This logic change will

allow opening of the

FCV when the

NFP flow drops to 4300

gpm and will

close

the

FCVs

when

NFP flow reaches

8600

gpm.

Due to the

new

~

8

~

I

0

~ 1 I J

~

0'6

~ *\\-

orifice sizing,

when the

FCV's are full open

4300

gpm of

MFP flow

will return to the condenser,

yet still maintaining

MFP flow at 8600

gpm.

This

4300

gpm will prevent

FCV cycling

excessively

while

ensuring that

MFP minimum flow requirements

are

met.

The licensee

conducted

these

changes

in accordance

with

WRs 87-BELK1, 87-BELK2,

87-BELN1 and 87-BELN2.

No violations or deviations

were identified in the areas

inspected.

6

~

Onsite Nuclear Safety Committee

(40700)

With the

plant

in

Mode

4

(Hot Shutdown),

operating

on

RHR

shutdown

'ooling, the plant experienced

a reactor protection

system actuation.

The

event

occurred while the licensee

was implementing

PCR-2292, modification

on auxiliary

control logic.

Review of this

PCR

was

performed

and 'approved

by the Plant Nuclear Safety

Committee prior to technicians

attempting

to install

these

modifications.

When'nstrumentation

and

Control technicians

removed

power from Card

8 in Process

Instrumentation

and

Control

(PIC) Cabinet

4,

bistable

P-13 (input for turbine

power

greater

than

10 percent)

deenergized

enabling

the

associated

turbine

trip/reactor trip to activate

causing

the plant to experience

a reactor

trip.

Licensee

personnel

informed

the inspectors

that it was believed

that

the

input to the turbine trip/reactor trip (first stage

turbine

pressure)

would fail

low on'oss

of

power

and

not activate

the

P-13

bistable.

Subsequent

licensee

investigation into the electrical

drawings

found that the

P-13 bistable

would in fact initiate when the

power for

Caid

8 in

PIC

4

was

removed.

NRC

inspectors,

performing

a

special

inspection

during

the

week of October

19,

1987,

obtained

a

copy of

PCR-2292 for review as

a part of this inspection efforts.

The results of

this review will be documented

in a subsequent

inspection report.

No violations or deviations

were identified in the areas

inspected.

7.

AFW Logic Deficiency (36100)

NRC personnel

evaluated

an event which the licensee

reported in accordance

with 10 CFR Part 21.

This Part

21 event

was identified in September,

1987

during evaluation

of an unrelated

event,

discussed

in Inspection

Report

50-400/87-34.

The

issue

dealt with

a

design

review of

a

postulated

accident

in which the loss of a vital

DC bus

(1B-SB) was concurrent with

loss of preferred (offsite) power.

The licensee's

evaluation of this item

determined

that this'ituation,

of its self,

did not result

in

any

unanalyzed

design condition.

However, while performing this analysis,

the

licensee

found three other accident

scenarios

which needed to be evaluated

and resolved.

These three

scenarios

are

as follows:

Scenario I:

The loss of vital

DC bus

1B-SB coincident with a loss

of off-site power causes

loss of one motor driven

AFW

pump and the turbine driven

AFW pump.

12

Scenario II:

The inadvertent

actuation of a relay causes

isolation

of AFW flow to one intact steam generator.

Scenario III:

The loss of vital

DC bus

1B-SB coincident with a loss

of off-site

power

allowed

the

Engineered

Safeguard

Feature

Actuation System to isolate

AFW from all three

steam generators.

After performing engineering

design

reanalyses,

the licensee

submitted

a

letter

on

September

22,

1987, outlining the results

of these

reviews to

the

NRC.

Details

of this letter identified that

two of the

three

scenarios

( I 5 II) were

encompassed

by the

FSAR,

Chapter

15 acceptance

criteria.

In scenario III,, the licensee

implemented wiring changes

to the

auxiliary