ML17342B296

Rel i abi 1 ity

Data Collection, Storage,

Analysis and Display

Regulatory

Guide

1'.97 Variable Availability

5.0

6.0

v.o

8.0

Persons

Contacted

Exit Interview

Licensee Actions on Previousl

Identified Findin

s

Glossar

of Acron ms and Initialisms

Assessment

of Radiolo ical Releases

Source

Term

Section

5. 1.1 of the site Radiological

Emergency

Plan (Plant Release

Pathways)

discussed

the potential plant effluent release

points.

The

principle release

point was the plant vent.

The following potential

release

sources

were directed to the plant .vent:

containment

purge

systems

for

both

units;

gas

decay

tanks;

auxiliary building

ventilation

system;

Unit

4

spent

fuel pit ventilation;

radwaste

building ventilation system;

and laundry facility ventilation system.

The plant vent monitoring system consisted

of noble gas monitors,

and

cartridges. for analysis

of particulates

and iodines.

= The noble

gas

monitor readings

were available in the Control

Room (CR) ind from the

Emergency

Response

Data Acquisition and Display System

(ERDADS).

The

Unit 3 spent fuel pit area

was separately

vented.

The exhaust

flow

wa's

monitored

for

noble

gas,

particulates,

and

iodines.

This

information

was available

from the

Control

Room

and

the

terminals

in the

and

The

steam jet air ejectors

were

provided

with gross

radioactivity

monitors.

. Information

for

a

release

through

the

atmospheric

dump valves

was obtained

from the

main

steam line monitors.

Appendix II of Procedure,

1302,

"Core

Damage

Assessment"

established

precalculated

relationships

between

cor'e

damage

and

the following:

post accident

sampling results;

hydrogen

concentration;

containment

high

range

radiation monitors;

core exit thermocouples;

and reactor

vessel

level.

This

procedure

was

based

upon

the

"Westinghouse

Owner's

Group

Post

Accident

Core

Damage

Assessment

Methodology',"

Revision

1 dated

March

1984.

The plant specific

implementation

of

this methodology

was described

in

a licensee

consultant's

document

entitled

"Turkey Point

Units

3

and

4

Core

Damage

Assessment

Guidelines" (Control

No. 03-1050-1084,

Job

No. 1050-029).

For

both the manual

(EPIP 20126)

and computer-based

dose calculation

models

'there

were

four methods

available

for determining

release

rate.

The

four

methods

consisted

.of

grab

samples

(non-PASS),

effluent monitors,

containment

high

range

radiation

monitors,

and

default values.

Grab

samples

were obtained

from assigned

plant grab

sample

locations.

NUREG-0737,

Item II.B.2 specified that licensees

conduct

radiation

and

shielding

design

reviews

of

spaces

around

systems that

may contain highly radioactive materials

as

a result of

an

accident.

Where

access

is

required

following

an

accident,

licensees

should provide for adequate

access

for personnel

through.

design

changes,

increased

temporary

or

permanent

shielding,

or

through postaccident

procedural

controls.

The dose rate criteria for

initial

areas

are

those

of

General

Design

Criteria 19.

Plant

personnel

could

not identify any

study

designed

to determine

the

accessibility of normal

grab

sample

locations

under

severe

accident

conditions,

or the feasibility of collection of highly radioactive

accident

samples

at

the

grab

sample

locations.

Effluent monitors

were located

on the plant vent,

main

steam line, Unit 3 spent fuel

building vent

and air ejectors.

No study could be located

by plant

personnel

identifying the effect of containment

and

system

shine

on

the

plant

vent

monitoring

system

(SPING-4).

For

a

LOCA type

accident,

the

containment

high

range

radiation

monitor

and

the

Technical

Specification

containment

leakage

rate

could

be

used to

determine

radioactive

material

release

rate.

The

method

for

determining

release

rates

was

from default

values

based

on

the

accident analysis

presented

in the plant's

FSAR.

The primary method for obtaining

source

term data in the

TSC was by

phone

communication

with the

Control

Room

(CR).

The

Emergency

Operations Facility (EOF) obtained the data via a dedicated line from

the

Technical

Support

Center

(TSC).

This

information

was

also

available

in the

CR,

TSC,

and

on

ERDADS terminals.

The

upgrade effort will include the

computer

dose

assessment

model part

of ERDADS and will provide automatic

data

input from the

parameters

maintained

on

During

a review of dose

assessment

on

ERDADS, it was noted that many values- were

incorrect

and carried

no warning flags to inform the user.

During

the

inspection

(February

24,

1988),

the

following

values

were

identified as incorrect;

however

no user warning system

was available

to detect

such data:

Unit 3 containment

high range

monitor reading

(53-43,500 R/hr); plant

vent flow (0.23 CFM);. and

wind

speed

and

direction from the South

Dade Tower 60

m elevation.

Based

on the above findings, the licensee

agreed to evaluate

and take

appropriate

action

on the following items:

Prior

to

integrating

the

dose

assessment

code

into

ERDADS,

determine

the reliability of ERDADS data which would be used for

dose

assessment

calculations

under the proposed

upgraded

ERDADS,

and devise

a method to inform the

user

when

data

points

are

incorrect

or

out

of

service

(IFI

50-250/88-01-01,

50-251/88-01-01).

Evaluate

the effect of containment

and system

shine

on the plant

vent monitoring

system

(SPING-4)

measurements

and reliability

during

accident

conditions

(IFI

50-250/88-01-02,

50-251/88-01-02).

Evaluate

the accessibility of the sampling station

used to take

samples

for source

term'evelopment

after

an

accident

(IFI

50-250/88-01-03,

50-251/88-01-03).

Dose Assessment

The calculational

methodology for the computer

dose

assessment

model

was

very similar to the

manual

method

contained

in

EPIP 20126,

"Offsite

Dose

Calculations"

with the exception

of the atmospheric

transport algorithms.

This methodology

was described

in the user's

guide

and

in

"FPL

Class

A

Emergency

Dose

. Calculation

Computer

Program,"

Rev.

12/87.

An outline of the program testing

conducted

in

an effort to verify and validate the

PC-based

dose

assessment

program

was

found in

a

FPL inter-office correspondence

memo to file

11580

(84) dated

November

6,

1984.

The documentation

was

stored

in nine

filing envelopes

identified- as

"Dose Assessment

Program

IBM PC."

A

review of the

documents

revealed

that

they did not contain

the

following:

(1)

a description

of the

methodology

used to verify and

validate

the model;

(2)

a verification of the equations

upon which

the

computer algorithms

were derived;

(3)

a verification that

the

basic

equations

were correctly

entered

into the

computer

language

used 'to write the code;

and (4)

a complete explanation of significant

differences

encountered

when code output was compared to NRC, State,

and other

dose

assessment

models.

Discussions

with the

cognizant

licensee

representative

responsible for maintaining the code revealed

that it was not subject to any established

software control procedure

that

ensures

adequate

documentation

and

performance

testing of all

modifications to the code prior to a

new version

being

accepted

and

placed

in service.

For example, modifications were

made to the code

prior

to

the

1988

Annual

Emergency

Preparedness

Exercise.

These

modifications were tested

and the

new version of the code

was placed

in service.

During the exercise,

both the licensee

and the

State

(running

the

same

version

of the

code)

experienced

problems with

software

performance.

Both 'the

manual

and

computer

dose

assessment

models

calculated

thyroid dose

to the adult thyroid instead

of the

more conservative

child thyroid.

However, the protective action

recommendations

based

on thyroid dose

started

at 2.5 rem instead of 5 rem.

This appeared

to be consistent with the approximate

2 to

1 sensitivity of child to

adult thyroid dose indicated in EPA Publication

EPA-520.

Offsite dose calculations

were governed

by Emergency

Procedure

20126.

This procedure

covered

two methods for estimating

doses.

The primary

method,

which was

used after the

and

EOF were activated,

was

a

compiled

BASIC computer

code written for execution

on

an

IBM PC.

In

the

PC

method,

atmospheric

transport

and diffusion were estimated

using

a combination of segmented

plume

and puff models.

The

second

method

was

a

manual

method

that

used

a

set

of reference

tables

contained

in

EP 20126 to determine

normalized

dose

rates.

In the

manual

method,

atmospheric

transport

and diffusion estimate at 1, 2,

5,

and

10 miles were based

on the output of the

PC code.

Estimates

at 3, 4, 6, 7, 8, 9,

11,

and

12 miles were based

on computations

made

with

a straight-line

Gaussian

model.

These

computational

methods

were appropriate

for emergency

dose calculations.

A short user's

guide to the

PC dose

computation

method

was contained

in Appendix

B of

EP 20126;

more comprehensive,

but still incomplete,

documentation

of the computer

code

was contained

in an Environmental

Sciences

Corporation

(ESG) report dated

February

28,

1984.

Portions

of this report

and

a listing of t'e

computer

code

were

examined

during the appraisal.

The equation

given in the report for diffusion

of puffs differs in several

respects

from the standard

Gaussian

puff

equations.

Some

of -the

differences

appeared

to

be

based

on

assumptions,

while others

appeared

to be errors.

Diffusion equations

derived from the computer

code were consistent with the equations

in

the report.

A

number

of errors

were

found in the

code

in file

XOQ.BAS, dated April 7,

1987.

The errors

found in the

code

had off-setting effects

with a net

result that the puff-model estimate

of concentration

and therefore

dose

was

generally within

a

factor of

2

or

3 of the

correct

estimates.

The errors that were found cause

the code to be suspect.

One

area

was

noted

where

the

model

deviated

from published

NRC

guidance.

Regulatory

Guide

1, 145 generally limits the

reducti,on of

concentrations

due to building wakes to no more than

a factor of 3.

It appeared

that

the building

wake

correction

to

the

diffusion

coefficients

may result

in

decreases

in concentration

and

dose

estimates

of slightly more than

a factor of 3 at

a distance of -1.mile

in extremely stable conditions.

The

dose

computation

methods

have

not

been

fully verified

or

validated.

Some computational

checks of the computer

code

have

been

made.

However,

these

checks

started with the equations

in the

ESC

report.

The equations

were not verified.

Records

code

development

and verification that

were

examined

were not complete.

The records did not contain

a history of

the

changes

made to the

code.

There

were letters

that

described

problems

encountered

and indicated that the problems

were corrected;

however,

the letters did not describe

the

changes

that were

made to

the code to fix the problems.

Remarks

contained

in the computer .code

indicated that the code

had

been modified'as

recently

as

the

summer

of 1987.

The recent modifications were not documented.

The

above

detailed

review

and

discussion

of the

Class

A

Dose

Assessment

Model disclosed that

no procedure

or administrative policy

was

available

to control,

modify,

and

ensure

maintenance

of the

subject

computer

model.

Accordingly, failure to provide

required

control constituted

the apparent violation defined below.

Apparent violation (50-250/88-01-04,

50-251/88-01-04):

Contrary to

Criterion

V of Appendix

B to

10

CFR 50,

Procedure

QP.5. 1 of the

Turkey Point Quality Assurance

Manual

and Section 8.4 of Emergency

Procedure

20126,

the

licensee

failed to establish,

implement

. and,

maintain

a written software procedure

and or administrative policy to

control the computer

based

dose calculation model.

Additionally, the licensee

agreed

to evaluate

and take appropriate

action

on the following:

Complete validation

and verification documentation

for the dose

assessment

computer

code

(IFI 50-250/88-01-05,

50-251/88"01-05).

2.0

Perform

new

comparison

with

NRC

and

State

models

once

the

'ompute}

code

has

been corrected,

validated,

and verified.

All

significant differences

should

be

determined

and

the

reasons

documented (IFI 50-250/88-01-06,

50-251/88-01-06).

Meteorolo ical Information

Onsite

meteorological

data

was

provided

by

primary

and

backup

meteorological

systems.

The

primary

system,

was

located

approximately

0.75 mile from the reactor

and measured

wind direction

and wind speed at .a height of 10 mete~s.

This data provided primary

meteorological

input to dose computations.

The standard deviation of

wind direction

was

computed

from the wind direction

signal

and

was

used

as

a secondary

indication of atmospheric stability.

The backup

system

was

located

about

6.5 miles

from the

reactor

and

includes

measurement

of wind direction

and wind speed at

a height of 60 meters

and

redundant

ins'trumentation

for measuring

the

temperatures

and

temperature

difference

(delta T)

between

10

and

60 meters.

These

. temperature

differences

were

the

primary

method

of

estimating

'tmospheric stability, while the

60 meter

wind data

were backups

to

the data

from the primary system.

The

backup

system

also

included

measurements

of

dew point,

solar

radiation,

and

precipitation.

Homestead Air Force

Base

provided

an additional

backup

source

for

meteorological

data.

The

instruments

in both

systems

were

well'xposed

and

provided

data

generally

representative

of atmospheric

conditions at the plant.

Signals

from

the

meteorological

instruments

pass

directly

to

instrument

sheds

located

near the bases

of the towers where they were

conditioned,

displayed,

and

recorded.

Selected

signals

were

transmitted to the

Land Utilization Facility and the plant.

The

instruments

and

tower's

were protected

from lightning,

and the

instrument

sheds

appeared

to

have

adequate

environmental

control to

permit

the

instrumentation

to

operate

reliably.

Instrument

electrical.

power

was

obtained

from

normal

AC

lines.

Diesel

generators

provided

a source of emergency

backup

power.

Plant

procedures

provided

for daily

inspections

and

periodic

calibrations

of the

meteorological

instrument

systems.

Records

indicated that calibrations

were conducted quarterly.

Meteorological

data availability from the instruments

was excellent.

Meteorological

data

were available

in the Control

Room from strip

chart recorders

and via ERDADS.

They were

made available to the

and

EOF via

and telephone

communications.

,EP 20126 requ'ires

the

use

of

15-minute

averaged

meteorological

data.

To

obtain

meteorological

data

appropriate

for use

in dose

assessment,

average

wind directions,

wind speeds

and stabilities

have

to

be

estimated

from the strip chart recorders.

Biases

and errors

in estimation of

average

meteorological

conditions

were well

documented.

Given

the

width of strip charts

in the

Control

Room, it was unlikely that

averages

could

be

,estimated

with sufficient

accuracy

for

dose

assessment

applications.

The

meteorological

data

available

through

the

were

not

averaged.

Current data reflected

the

instantaneous

variations that

were

common

in the

atmosphere.

Estimation of averages

by trending

the data

on

ERDADS was

no better than estimation of averages

from the

strip charts

in the Control

Room.

As

a result,

the meteorological

data

in

ERDAOS

were

not

adequate

for

use

in

dose

=assessment.

Further,

on

several

occasions

during

the

Appraisal,

the

. meteorological

information displayed

by

ERDAOS were clearly wrong.

For example:

uncharacteristically

wide variations in wind direction

were noted during relatively strong winds;

a wind speed of more than

40

mph

was

displayed

when

the

actual

winds

were light.

Air

temperatures

greater

than 100'F were displayed.

~

During

the

Appraisal, it

was

determined

that

the

~ range

of the

temperature

difference

instruments

used

to determine

atmospheric

stability was -5

F to +15

F.

The temperature

difference (delta-T)

range

shown

on the Control

Room analog strip chart recorder

and strip

chart was -5'F to +5 F.

Control

Room personnel

believed

the posted

range to

be correct:

The

range of the meteorological

instruments,

however,

was

changed

to -5

F +15

F

on or about

October

7,

1987.

Nemos

dated

September

9,

1987,

and September

29,

1987, described

the

intent to change

the. range

and

the anticipated

date of change.

The

Control

Room strip charts

and strip chart

recorder

were neither

changed

nor Control

Room staff informed that

a delta-T

range

change

had occurred.

As

a consequence

of the missed

communication,

Control

Room personnel

could

not

properly

determine

temperature

difference.

Therefore,

atmospheric

conditions

associated

with

poor

dispersion

and

high

potential

doses

in the event of a. release

could not be distinguished

from atmospheric

conditions

a'ssociated

with good dispersion

and low

potential

doses.

The errors

associated

with improper determination

of

the

temperature

difference

could

result

in

significant

underestimates

of

potential

doses.

For

example,

an

actual

temperature

difference of 1.4'F translates

to

an

F stability class

using Table

2 of

EP 20126.

On the strip chart

in the Control

Room

this temperature

difference

would appear

as

a line at about

32% of

full scale. 'his

indication

would

have

been

interpreted

as

a

temperature

difference

of -1.8'F

(A stability)

by

Control

Room

personnel.

The difference

in

dose

rates

estimated

for

A and

F

stability

classes

is

a function of distance.

At

a distance

of

1'ile, meteorological

and respective

dose

assessment

calculations

disclosed

that

dose

rates

estimated

for

A stability

were

approximately

a factor of 150 lower than the dose rate estimated for

F stability.

At 3 miles, the estimated

dose rate

was approximately

a

factor of 115 lower.

At 9 miles,

dose

rates

were

about

a factor of

90 lower.

A

As discussed

above, offsite dose

assessment

calculations

based

upon

inaccurate

delta-T ranges,

and inaccurate

methods

for obtaining the

time-averaged

(15 minute increments)

meteorological

data could result

in significantly nonconservative

dose

estimates

in the

event

of

a

radioactive

material

release.

Based

upon these findings, inspection

disclosed

the

apparent

violation

defined

below.

The

findings

addressed

were

contrary

to

10

CFR 50.47(b)(9)

which

requires

provisions. for adequate

methods.,

systems,

and equipment for assessing

and

monitoring

actual

and

potential

offsite

consequences

of

a

radiological

emergency

condition.

Additionally, Emergency

Procedure

'P

20126,

"Offsite

Dose

Calculations,"

requires

that

15 minute

averaged

meteorological

data

from primary or

backup

meteorological

towers

be

used to obtain statistically valid meteorological

data to

calculate offsite doses.

Apparent

violation

(50-250/88-01-07,

50-251/88-01-07):

As

a

consequence

of the failure to promulgate. the change

in delta-T range,

offsite

dose

assessment

could

be

inaccurate

and

result

in

significantly nonconservative

dose

estimates

by

up to

a factor of

150.

Additionally, the method

used for obtaining

15 minute averaged

meteorological

data

was not sufficiently accurate

to be

used in dose

projection

and assessment

calculations.

3,0

Technical

Su

ort Center

3.1

Re viator

Guide 1.97 Variable Availabilit

and Sufficienc

The

Florida

Power

and

Light Company

received

a

Safety

Evaluation

Report

(SER)

for

conformance

to

Regulat'ory

Guide

(RG)

1.97 'on

March 20,

1986.

The report concluded that

FP8L either

conformed to,

or was justified in deviating from, the guidance of

RG 1.97 for each

post-accident

monitoring variable,

with the exception

of "plant and

environs radiation" (i.e., portable radiation monitoring equipment).

The

SER therefore

approved

the installed

sensors

at Turkey Point.

As

with all

RG 1.97 reviews, this approval

was for installed instrument

loops

and did not include

an evaluation of electronic transmission

of

these

variables

to

data

acquisition

systems

for the

Emergency

Response

Facilities.

0

The licensee

report of the

implementation

of

RG 1.97

was available

and

was

used

as

an appraisal

information resource.

The Turkey Point

SPDS,

which consisted

of a subset of the software running

on the

same

MODCOMP computer

system

as

ERDADS,

received

a

post-implementation

audit by

NRC in March 1987.

To date,

NRC has not issued

an

SER for

that audit.

This lack of an

SPDS

SER appears

consistent

with the

reliability and data error issues

raised during this

ERF Appraisal.

8

0

Plant variables

were transmitted

to the

TSC and

EOF via the

ERDADS was

one of the

software

routines

which runs

on the

MODCOMP

computer

system which also hosts

the safety parameter

display

system

(SPDS).

Of'the

RG 1.97 variables

approved in the March 1986

SER, the

only notable

parameters

which were not data linked to the

ERDADS were

several

of the

area

radiation monitors

(ARMs).

The

ARMs at Turkey

Point have

a history of problems requiring continual

maintenance.

A

summary of discussions

with the licensee

indicated that overall

ARM

reliability

was

approximately

80% (i.e.,

at

any

given

time,

approximately

20% of the

ARMs were out of service).

The licensee

had

a long term plan to replace

the

ARM system.

This work should improve

the availability of ARM data .on the

Based

on

the

above

findings,

RG 1.97 variables

availabili.ty and

adequacy

appeared

. to

be

adequate

to

respond

to

a

simulated

emergency.

.A.l

~d

As discussed

in Section 3.1

above,

parameter availability via the

Has satisfactory

in terms

of data

being. entered

into the

system.

The

system

was

included

in

routine

~ station

survei1 lance

checks,

3.1.2

Manual

Data

~

3'.3

The backup

system provided for transmitting variables to the

TSC and

the

was

through

dedicated

telephone

communicators.

The

reliability of the telephone

systems

between

the Control

Room,

and

appeared

satisfactory.

Each

system

had

a battery

or

redundant

power supply backup.

Status

boards

in both the

TSC and

were satisfactory.

System

85

telephones

provided 'by

AT&T had the

standard battery backup.

Telephones

maintained

by FP&L were provided

with redundant

AC power supplies.

I~lAd

Based

on the above findings, the licensee

agreed

to evaluate

and take

appropriate

action

on the following:

ERDADS .provided

no

concise

display of containment

isolation

status.

While all isolation

valve positions

appeared

to

be

input to the

system,

there

was

no rapid

method,

on

a single

screen,

to assess

their collective status

( IFI 50-250/88-01-08,

50-251/88-'1-08).

ERDADS displayed

numerous

erroneous

values

such as:

condensate

pumps off when. they were actually running; containment radiation

at 5.6

E+7 mr/hr; operator

aid displays

alarmed

in orange

and

red;

and Unit 4 Reactor

Coolant

Pumps off at

27% power;

ARMs

9

indicated

out of

scan

or were displaying

spurious

data (IFI

50-250/88"01-09,

50-251/88=01-09).

3.2

Functional

Ca abilities

3.2.1

Power

Su

lies

The primary source

of all

480

and

120'VAC powe'r to the

was

a

480/277

VAC Distribution

Panel

(DP-76)

which

had

two

separate

incoming

power supplies.

The

normal

supply to DP-76 is the 13.2

KV

Florida City line (offsite power).

The alternate

supply

was

from

load

center

3G

which

receives

power

from

Turkey

Point

Unit

2

non-vital

C Bus.

C Bus can

be supplied

from Unit 3 sources:

offsite

power,

main turbine,

and the Unit

1

and

2 (fossi l units) cranking

diesel

in the event of loss of power from Unit 3.

With the exception

of telephones,

all equipment,

lighting, and vegti lation loads in the

TSC were, supplied

from the

scheme

described

above.

The microcomputer

based

dose

projection

system

had

an

additional

level

of

power

redundancy

provided by a 750 ampere

hour uninterruptable

power-supply

(UPS).

The dose projectio'n

UPS should provide power to the computer

for about

6 to 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months in the event of loss of other

power sources.

TSC lighting had

an additional

level of redundancy

provided

by two

~ wall mounted, battery

powered

emergency

flood light units.

In the event of. a loss of the

13.2

KV Florida City line,

an auto

transfer

switch would align to the alternate

3G load center.

This

auto transfer

switch

was

power seeking,

and

normal

seeking after

a

30 minute

delay.

Therefore,

all

equipment

was

supplied

by

reliable

redundant

power.

The

inspector

noted

that this transfer

switch

was

not

included

in

any

preventive

maintenance

'program to

insure that operability was tested.

The

inspector

observed

tests

of all

the

above

power

systems

and

inspec'ted

their physical

condition.

A problem

was

noted with the

battery

powered lighting units.

The electrolyte levels in both units

were at approximately

50% of required.

This condition was believed

to have

been

caused

by

a lack of

a preventive

maintenance

program

for the

charging

units

which

caused

the units

to

remain

in fast

charge,

thereby boiling off the electrolytes.

Based

on the

above

review,

the licensee

agreed to evaluate

and take

appropriate

action

on the following:

Addition of the

TSC emergency

backup lighting and the automatic

transfer

switch

to

the

preventive

maintenance

program

(IFI

50"250/88-01-10,

50-251/88-01-10).

3.2.2

~

TSC Data Anal sis

The

ERDADS and the status

boards

were the primary sources

of data for

managers.

The

has

a

14 hour1.62037e-4 days 0.00389 hours 2.314815e-5 weeks 5.327e-6 months

historical

trending

10

e

3.3

capability, with the user able to select individual points

and obtain

either

a tabular or graphic trend display covering

a period from 0.5

to

14 hours1.62037e-4 days 0.00389 hours 2.314815e-5 weeks 5.327e-6 months in the past.

Emergency

action

levels

(EALs) for

Turkey Point

were

included in

Emergency

Procedure

20101,

'-'Duties

of the

Emergency Coordinator."

Turkey

Point

EALs did

not

contain

as

many

specific

parameter

setpoints

or trigger points

as most

EALs, but where

included,

there

was agreement

with the setpoints

used in the

ERDADS system.

TSC Habitabilit

The

TSC was constructed

of 12" concrete filled block walls and

a 10"

concrete

slab roof.

The facility was equipped with a

HVAC filtration

system.

Shielding

calculations

performed

by Bechtel

(Calculation

No. M08-117-10,

Rev. 0,

Job

No. 5177-117,

dated

June

3,

1980)

indicated that personnel

would not receive

a radiation

exposure

in

excess .of

5 rem whole body, or 25 rem to the thyroid for the duration

of an accident.

Air radionuclide

concentrations

would

be moni.tored

during

emergencies

by

an

Eberline

AMS-13

(Beta

Continuous

Air

Monitor/CAM) which

samples

the aj r intake

down stream of the

system

filters.

A review of the

procedures

was

conducted

to determine

who

was

responsible

for establishing

and monitoring

TSC habitability,

and to

determine

the

guidance

provided

for

conducting

habitability

measurements.

Emergency

Procedure

20105 specified that the Emergency

Coordinator

(EC) was respon'sible

for determining

when the

TSC should

be

evacuated

based

on

habitability

information.

However,

the

procedure

did not specify

who, was responsible

for providing the

EC

with "this

information.

None of the

emergency

plan

implementing

procedures

(EPIPs)

were found to contain

information or guidance

on

TSC habitability.

Discussions

with licensee'ersonnel

determined

that

the

Radiation

Team

Leader

in

the

was

responsible

for

establishing

and

monitoring

TSC habitability and providing the

EC

with

appropriate

information.

[The

guidance

for

conducting

habitability surveys

was

located

in Sections

4.8

and 8.4 of Health

Physics

Procedure

HP 91,

"Emergency Radiation

Team

Response

Onsite."

The

health

physics

procedural

reference

was

not

found

in

the

emergency

procedures.]

(See

IFI

50-250/88-01-11.,

50-251/88-01-11

below.)

The

TSC ventilation

system

consisted

of

a

normal air h'andling unit

with an associated

chiller unit and

an

em'ergency train consisting of

a series

of

HEPA and charcoal filters with an

1100

CFM air handling

unit.

Upon

TSC activation,

the ventilation system would

be manually

switched to the emergency

mode, which places

the HEPA/Charcoal train

in series with the normal train.

The system

then draws

a combination

of outside

and recirculated

air through

both trains to provide

a

filtered, pressurized

environment

in the

The

TSC ventilation

11

system

performed satisfactorily, with all fans

and dampers

operating

properly to maintain

a positive pressure within the

After observing

several

tests of the system,

the inspector discussed

with the licensee

the following items:

(1)

The normal

TSC air handling 'unit control circuit was wired such

that both the

condenser

(chiller) unit

and

the

fan cycle

by

thermostatic

control.

With the

system

in the

emergency

mode,

. stopping the normal

fan (with the emergency train still running)

reduced

the

positive

pressure

in

the

Because

no

differential

pressure

meter

was installed, it could

not

be

determined

quantitatively

how

much

positive

pressure

was

maintained.

The licensee

might consider

adding

a step

to the

emergency ventilation system startup

procedure

to switch the fan

mode switch on the pillar mounted

TSC thermostat

to "ON" versus

"AUTO."

This would keep the normal

mode fan from shutting off

during

emergency

operation

of the

system.

Another alternative

would be to rewire the control circuit to

keep

the normal-fan

running when the

system

was in the emergency

mode.

(2)

HVAC Construction

Drawing 5610-A-57

showed

a differential

pressure

ressure

indicator

(DPI-6406)

installed

to

monitor

the

difference

in positive

pressure

between

the

and

outside

atmosphere.

This indicator was not installed in the system.

The inspector

reviewed the two TSC ventilation

system test procedures

approved for use at Turkey Point.

The following findings addressing

the

subject

test

procedures

were

identified:

(1) Procedure

O-OSP-301.2.,

Emergency

Ventilation

System

Operational

Test,"

was

never

performed

and

was not included in the

station preventive

maintenance

scheduling;

(2) Procedure

0-OSP-301. 1,

E

rgency Ventilation System Filter Performance

Test," contained

no

acceptance

criteria

for

the

various

filter

and

adsorb

me

er

penetration

tests.

The latter procedure

has

been

performed,

but was

not included in the preventive

maintenance

prog'ram.

Based

upon

the

above findings,

the licensee

agreed

to evaluate

and.

take appropri'ate

action

on the following:

Ensure

that

EPIPs

reference

supporting

plant

procedures

((IFI

50-250/88"01-11,

50-251/88-01-11).

Install

a

DP indicator

so that positive pressure

in the

TSC can

be

verified

during

emergency

ventilation

system

operation

(50-250/88-01-12,

50-251/88-01-12).

I

1

d

P ocedure

O-OSP-301.2

"TSC Emergency Ventilation System

(IFI

Operational

Test,"

in the preventive

maintenance

program

<

50-250/88-01-13,

50-251/88-01-13).

12

Provide acceptance

criteria for the various penetration

tests

in

Procedure

0-OSP-301. 1,

"TSC Emergency Ventilation System Filter

Performance

Test,"

and place this procedure

in the. preventive

maintenance

program (IFI 50-250/88-01-14,

50-251/88-01-14).

3.4

Data Collection

Stora

e

Anal sis

and Dis la

At the time of this Appraisal,

the

licensee

was

in the

process

of

upgrading the

Details of the

ERDADS upgrade

were classified

as proprietary.

Findings described

in this section

are

based

on the

evaluation of existing computer, systems

and their use to support

ERF

functions.

3.4.1

Methods of Data Collection

Real-time

data

acquisition,

display,

and

storage

to

support

ERF

functions were

performed

by

ERDADS for Turkey Point Units

3

and

4.

is

a

distributed

system

which

included

the

processors

,identified and discussed

below.

Plant Environmental

Data

S stem

PEDS

.The

PEDS

was the

ERDADS Host computer

system.

PEDS

was

based

on

a

MODCOMP 7870 "Classic II" with 2 megabytes

(MB) random access

memory

(RAM),

a

300

MB hard disk,

a 800/1600 bits per inch (bpi) magnetic

tape drive.

Other computers

in the distributed network

send data to

this

system

via

a Satellite

computer.

The Plant

Data Concentrator

(PDC) computers

communicated directly with the

PEDS.

There were two

identically configured

PEDSs at the Turkey Point Plant.

The

PEDSs

communicated

with

each

other

via

a

high

speed

data

link

(approximately

600,000 bits/second).

PEDSs directly controlled

ERF

display cathode

ray tubes

(CRTs)

and printers.

Satellite

Com uters

The satellite

computers

were

MODCOMP Model 7870s with 1.5.MB

RAM but

had

no hard disks.

There

'was

a satellite

computer for each

Turkey

Point Unit to collect data

from special

purpose

processors

and pass

that data to the

PEDSs.

Satellite

computers collected data

from two

Eberline computers,

a

CLIMATRONIX meteorological

computer

system,

and

from two Digital Data Processing

System

(DDPS) computers

and transmit

the

data

to both

the

PEDSs.

Using this

redundance

configuration,

plant computer

system reliability was enhanced.

Plant Data Concentrators

PDCs

were

based

on

MODCOMP

MODACS 3

and

MODCOMP 7821

processors.

PDCs

acquired

analog

and digital

data

from plant

sensors

and

transmitted that data to the satellite

computers just described.

The

following table lists the sensors

monitored

by

PDCs for Turkey Po'int

Units

3 and 4:

13

Unit ¹

Anal og

Sensors

Digi

ta1'ensors

Total

Sensors

570

527

1097

Eberline

These

were

INTEL 8085 based

microcomputers with 16 kilobytes (KB) RAM

and

32

KB

ROM (read

only memory).

Each Turkey Point Unit had

an

Eberline

system to acquire radiation monitor data

and transmit it to

the

PEDS via the Satellite computer

systems.

vali.fied Safet

Parameter

Dis la

S stem

SPDS

These

were

Combustion

Engineering

INTEL 8085 drive microcomputers.

There

are

2 QSPDSs,

one for each

Turkey Point Unit.

Their purpose

was

to monitor critical

core

parameters

and

to display

them

on

request.

Meteorolo ical

Com uter

This

was

a

CLIMATRONIX processor

with 4 sets

of 128

KB RAM.

Its

function

was

to

collect

meteorological

data,

perform

required

computations,

and transmit it to the

Meteorology is fully

discussed

in Section 2.0,

above.

Di ital Data Processin

S stem

This was

a

NOVA 840 minicomputer

based

system with 48

KB RAM and two

2.5

MB hard disks.

This was

a plant computer

system installed in the

1970s

that

was

normally not

used

to provide

data

to the

Currently, this system

was

used to generate

flux maps

on

demand

and

to calibrate nuclear plant instrumentation for 100% rated

power.

ERDADS Documentation

Review

Nine

licensee

documents

describing

ERDADS and peripheral

computers

were

reviewed

during

the

ERF Appraisal

process.

Most

of

the

documents

reviewed

had not been

updated

since

1983.

Based'pon

the above findings, this portion of the licensee's

program

appeared

adequate.

Data Dis la

s in the

There

were three

RAMTEK 9400 graphic display

CRTs with RAMTEK GK-120

keyboards

in

the

The

RAMTEK

CRTs

were

controlled'y

RAMTEK-9400M/I graphics

generators.

Users

could. select

the

Unit

desired

and display safety parameters

or parameter

sets of interest.

On request,

a color hard copy could be generated.

14

Typical

displays

included

a graphical

representation

of

a plant

system

with real-time

updated

parameters.

In several

cases

trend

graphs

of safety

parameters

were

shown

on

the display

to provide

additional

information about

how critical parameters

were changing.

The

ERDADS Operator's

Manual that explained

use of the display

system

was

reviewed.

The

manual

provides'xamples

of typical

safety

parameter

displays.

The

ERDADS Operator's

Manual,'owever,

was

incomplete.

It was

also

noted that

there

was

an

absence

of the

command

description

"GET" (load

a display with data

from

a

user

selected

time frame

within the

past

14 hours1.62037e-4 days 0.00389 hours 2.314815e-5 weeks 5.327e-6 months )

and

"PEU" (print

engineering

units).

The

current

manual,

dated

May

1985,

is

a

preliminary document.

ERDADS display functions

were

noted to

be

impaired

by lockups

and

lengthy delays

during

ERF appraisal

review activities.

Delays of up

to

10 minutes

on

some

reports

were

observed.

During the course

of

the Appraisal, inspection

team

members

evaluating

the

ERDADS system,

observed

licensee

personnel

using

the

system

during

the

annual

exercise,

and

requested

licensee

computer

systems

personnel

to call

up specific displays.

. Over the

4-day Appraisal,

the

system

was

observed

to "lock-up" on numerous

occasions

(at least

6 times), both

with inspection

team

personnel.

and

licensee

personnel

operating

the

RAMTEK terminals..

The

EOF terminals at the

FP&L Corporate office

in, Miami were

locked

up for the 2.5 hour5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months

period

during

which. the

inspection

team

observed

the

annual

exercise

from that location.

While FP&L management

maintained that these

system failures were

the

result

of

system

operation

by inexperienced

personnel,

the fact

remained that the

system

locked

up frequently while being

used

by

licensee

personnel

during

the

annual

exercise,

and

by computer

systems

personnel

assigned

to assist

the

NRC inspection

team.

As described

in Section 3.2. 1,

power for the

ERDADS terminals in the

TSC'as

redundant

and reliable.

Although the

data

terminals

locked

up

on

numerous

occasions

when various display

commands

were

being

processed,

the

power supply transfer tests

conducted

at the

request of the inspection

team did not cause

ERDADS failure.

During

both

normal

to

emergency,

and

emergency

to

normal

power switches

('break before

make operations),

the terminal

screens

went blank for

about

1 second

and

then

returned

to the requested

display

mode with

no operator reset

being required.

Based

on the

above

review,

the licensee

agreed to evaluate

and take

appropriate

action

on the following:

Correct,

update,

- and

provide

current

procedural

manual

documentation

for

operation

of

the

system

(IFI

50-250/88"01-15,

50-251/88-01-15).

15

Reduce

system

lock-up (unavailability).

Perform

load

tests

to

identi fy

problem

sources

(IFI

50-250/88-01-16,

50-251/88-01-16).

3.4.2,1

Time Resolutioa

ERDADS computers

read,

analyze,

and store to hard disk data

from 2194

analog

and digital

sensors

(Turkey

Point

Units

3

and

4).

The

sampling

rate for data acquisition

varied

between

every

second

and

every

60 seconds,

dependi,ng

on the critical classification

of the.

safety

parameter

monitored.

The

data

rate

was

considered

low to

moderate

speed.

ERDADS also collected meteorological

data

from the

CLIMATRONIX computer

and

produced

a

meteorological

display

on

request.

The data acquisition

tasks

were

assigned

a high priority

and

even

when display

tasks

were

observed

to lock-up,

the

system

continued

to collect

and store plant

sensor

data

without apparent

data loss.

3.4.2.2

Si nal Isolation

I

At Turkey Point both optical

and transformer isolation devices

were

used to provide isolation sufficient to meet

NUREG-0737,

Supplement

1

criteria.

This

was verified

by letter "Safety

Parameter

Display

System

(SPDS),

Implementation

Plan

and

Parameter

Selection

Report,"

submitted to the licensee

by the

NRC.

3.4.2 '

Data Communications

data

communications

capabilities

were

reviewed.

Error

checking

and correcting

was reported to be done

by modem firmware and

operating

system

software.

Data

communications

between

MODCOMP

processors

used

high

speed

data

links

(approximately

600,000 bits/second).

MODCOMP operating

system

software,

MAXNET, was

- reported to be used to drive the computer to computer data

exchange.

MANNET

used

cyclical

redundancy

checking

to detect

communication

errors.

3 '.2.4

Processin

Ca acities

3.4.2.5

The

ERDADS and peripheral

computer

systems

were configured to support

plant safety monitoring

and reporting

needs.

ERDADS processing

was

based

on multitasking

to allow several

software

functions

to

be

processed

concurrently.

Data acquisition

and storage

tasks

were high

priority tasks

and continued

to execute

even

when other tasks

were

locked

up (e,g., display tasks).

Data Stora

e

Ca acit

Date

storage

had

been

functionally implemented

to meet

NUREG-0696

criteria.

Licensee

personnel

interviewed reported that at

any time,

14 hours1.62037e-4 days 0.00389 hours 2.314815e-5 weeks 5.327e-6 months

of historical

data

was

available

to

provide

trending

16

information

on critical plant parameters.

On demand plant parameter

data

can

be continuously stored via

a magnetic tape.

This data will

continue

to

be

stored until the

tape

storage

process

is operator

halted or until another

tape is mounted.

The data

storage capability to disk file was demonstrated - after the

ERF Exercise.

Display functions

locked

up

on

ERF display

CRTs for

approximately

2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months ,

as

discussed

in'ection '3.4.2,

above.

Data

sto'rage

via disk file continued

even during lock-up problems (this

was demonstrated

by historical trend plots). It also

was

shown that

historical data files were not lost on

ERDADS re-boot.

3.4.2.6

3.4.2.7

Model

and

S stem Reliabilit

and Validit

A

The paper,

"Setpoint Analysis Documentation

for SAS," September

22,

1982,

was

reviewed

and

found

to

contain

documented

algorithm

descripti.ons.

Algorithm

descriptions

were

not

checked

for

correctness

during this Appraisal.

Discussions

with the licensee

revealed that

ERDAOS analog

as well as

digital sensors

were redundantly

sampled.

Reliabilit

of Com uter

S stems

Manual logging of. computer

system unavailability was performed

by the

licensee for the

OOPS but had not been

done since

mid-December

1987.

Also, availability data for ERDAOS computers

was not compiled.

Based

on the

above

review,

the licensee

agreed to evaluate

and take

appropriate

action

on the following:

Availability for all computer

systems

supporting

ERF functions

should

be tracked

and

improvements

made

in availability where

appropriate (IFI 50-250/88-01-17,

50-251/88-01-17)

.

3.4.2.8

Manual

S stems

Manual

data

entry :.for

special-

purpose

computations

by

EROADS

computers

were reported

by licensee

contacts

to be:

(1) loading

new

tables for each fueling cycle;

and (2) calorimetry verification.

3.4.2.9

Environment Control

S stems

Air conditioning was reported

by licensee

personnel

to be functional

in the computer

room.

The air conditioning system

was reported to be

set to maintain

ambient

temperature

at about

80 degrees

Fahrenheit.

Inspection confirmed these reports.

17

4.1

Emer enc

0 erations Facilit

Location and Habitabilit

The Turkey Point

EOF is located in Miami, approximately

24 miles from

the plant site;

hence,

there were

no habitability -requirements

to be

met.

4.2

4.2.1

Functional

Ca abilities

Data Anal sis

Ade uac

The

EOF used the

same

ERDADS system for data acquisition

'as that used

in the

Status

boards

in the

EOF were more oriented toward dose

assessment

than

those

in the

TSC,

since

the

was

the

primary

facility for this function.

A large working area

was provided for

the

FPKL and

Florida Division of

Emergency

Management

personnel

involved in dose

assessment.

4.2.2

~Bk

EIIF

4.2.3

~

Not applicable to Turkey Point

~R1i II I I

The single

EOF,

which. was located

24 miles from the plant site,

was

provided with a reliable backup

power supply. If normal

power to the

EOF was lost,

a 750

KW gas turbine generator,

located in an adjacent

building, automatically starts.

The turbine generator

output breaker

automatically

closes

in approximately

40 seconds

to provide power to

the entire building in which the

EOF is located.

The

inspector

reviewed test

and maintenance

records

for the gas turbine generator

and

found

them satisfactory.

The

inspector

reviewed

documentation

that

showed that the turbine

was test

loaded

monthly

and

was

more

than

capable

of

assuming

the full electrical

load

of the

building.

The

EOF dose

assessment

computer

has

a

UPS similar to that

installed in the

4.3

Data Collection

Stora

e

Analysis

and Dis la

The

same

computers

supporting

ERF activities also supported

the

These

systems

and details

of their

functions

have

been

previously described.

EOF display

CRTs (4

RAMTEK 9400s)

are tPe

same

as the

TSC and allow users

to view Units

3 or 4 on request.

The

items

that

were

reported

in

Section

3, 1

above,

as

needing

evaluation

and action

in. the

TSC also hold for the

There

was

a

concern

about communications

between

the

ERDADS at the plant site

and

the

EOF display

CRTs.

According to licensee

documentation

reviewed,

some ll devices

are serviced

by a single

9600 baud conditioned line,

plus

modems

and multiplexers,

between

the plant

computers

and

the

18

This

appears

to

be

a large

number of devices

serviced

by

a

single

9600 baud line.

The licensee

agreed to consider

the following items for improvement:

Installation of

a

second

conditioned line with required

modems

and multiplexers

between

the. plant computers

and the

This

will provide backup capability and will speed

up .ERDADS display

performance at the

Acquisition of

a dedicated

computer

to

process

only display

functions.

This will speed

up display functions at the

CR,

the

and

the

EOF,

and

reduce

the overall

processing

load

on

4.4

Re viator

Guide 1.97 Variable Avai labilit

The data availability in the

EOF was essentially

the

same

as that in

the

TSC, with the

ERDADS system

and dedicated

communicator/recorder

personnel

serving

as

primary

and

backup

sources

of

data

(see

Section

3. 1, above).

5..0

Persons

Contacted

P. Bailey, Nuclear Energy Senior Analyst

"C. Baker, Plant Manager

D. Baker,

Engineer

"E. Baker,

Land Utilization

"A. Byrnes,

OSC Supervisor

G. Casto,

Emergency

Preparedness,

Corporate

G.

Dam, Site

ATILT Telephone

Representative

V. Edwards,

Senior Technician

R. Fisher,

Principal Specialist

"R. Fritchley, Assistant Training Superintendent

R. Gouldy, Corporate

Licensing

F. Guendelsberger,

Computer

System

Engineer

J. Harley,

I&C Foreman

"R. Hart, Licensing Engineer

G. Hollinger, Licensed Operator Training

"P. Hughes,

Health Physics

Supervisor

"J.

Kappes,

Maintenance

Superintendent

L. LeGarde,

Emergency

Preparedness

Coordinator, Site

, J. Maisler, Corporate

EP Manager

BE Maxwell, Electrical

Maintenance

Engineer

J.

Odom, Site Vice President

"H. Paduano,

Manager,

Nuclear. Energy Services

~L. Pearce,

Operations

Superintendent

~R. Steinke,

Chemistry Supervisor

"J. Strong, Assistant Superintendent,

Electrical

"H. Taylor,

Emergency

Preparedness

Technician,

Site

19

"D. Tomaszewski,

I&C Supervisor

"J. Walls,

gA Auditor

J.

Webb, Operations

Maintenance

Coordinator

G. Winters,

TSC,

HVAC System Engineer

B. Mood,

18C Technician

Nuclear Regulatory

Commission

"R, Brewer, Senior Resident

Inspector

"T. Decker,

NRC-RII

"E. Williams,

EPB/NRR

6.0

"Attended exit interview

Exit Interview

~G

The

inspection

scope

and findings'ere

summarized

on

February 25,

1988,

with those

persons

indicated

in

Paragraph

5

above.

The

inspector

described

the areas

inspected

and discussed

in detail. the

inspection

findings herein.

No dissenting

comments

were

received

from the licensee.

Although proprietary material

was reviewed during

the inspection,

such material

was neither

removed

from the site

nor

entered into this report.

Licensee Action on Previousl

Identified Findin

s

a

~

(Cl osed)

Inspector

Fol 1 owup

Item (IFI)

250,

251/87-03-01:

Failure

to

noti fy

the

State

within

15 minutes

following

declaration of the Site Area Emergency.

b.

The inspector

reviewed the change

in the order of notifications

made

in

EP Procedure

20101

where

the State

Warning Point was

listed

to

be

notified first.

This

was

completed

in

the

procedure

revision

dated

April 7,

1987.

Training

on

this

procedure

revision was completed

in December

1987.

The inspector

observed

the notification of the State during the

Exercise

on February

23,

1988,

and found the notification to be

timely and accurate.

(Closed)

IFI 250,

251/87-03-02:

Failure'to consistently

inform

plant staff, via public address

system,

of emergency

status

and

.

associated

directives during the simulated

emergency conditions.

The inspector

reviewed the step

changes

8.4.3, 8.5.3,

and 8.6.3

of Procedure

EP 20101

dated

April 7,

1987,

that

keyed

the

Emergency

Coordinator

to

use

the

Public

Address

System.

Training Brief 204 issued

on June

18,

1987,

emphasizing

the

use

of the Public Address

System for the

1987 Emergency Coordinator

training was also reviewed.

20

The

inspector

observed

the

use

of the

Public

Address

System

during

the 'ebruary

23,

1988,

annual

Exercise.

The

announcements

were noted to be timely and informative.

c.

(Closed)

IFI

250,

251/87-13-01:

Factoring

into

each

year'

January

update

of the

Emergency

Response, Directory information

, on individuals disqualified per

EP 2021.

The

inspector

reviewed

documentation

indicating

that

all

responders

listed

in the

Emergency

Response

Directory dated

December

31,

1987,

were verified against the

Emergency

Response

Personnel

training report.

d.

(Closed)

IFI 250,

251/87-13-02:

Development

of .an

auditable

system

for

documenting

all

weaknesses

and

deficiencies

identified in drill and exercise critiques

and for tracking the

corrective action for each

such finding to ensure

completion.

The inspector

reviewed

the

gA and

NRC findings listed

on the

Commitment Tracking

Program

(CTRAC).

Drill critique items

are

tracked

as stated

in letter

PTN-EP-87-049.

The tracking systems

appeared

to provide

an appropriate

means to follow deficiencies

and weaknesses

identified in drill and exercise critiques.

Glossar

of Acron

ms and Initialisms

ARM

CAM

CRT

'OPS

DP

EC

KB

PEDS

PEU

.

ROM

SPOS

Area Radiation Monitor

Continuous Air Monitor

Cathode

Ray Tube

Digital Data Processing

System

Differential Pressure

Emergency Action Level

Emergency Coordinator

Emergency

Preparedness

Implementing Procedures

'Emergency

Response

Data Acquisition and Display System

High Efficiency Particulate Air Filters

Heating, Ventilation and Air Conditioning

Ki 1 obytes

Plant Data Concentrator

Plant Environmental

Data

System

Print Engineering Units

qualified Safety Parameter

Display System

Random Access

Memory

Regulatory

Guide

Read Only Memory

Safety Evaluation

Report

Safety Parameter

Display System

Uninterruptable

Power Supply

APPENDIX A

Apparent Apprai sal Violations

1.

Failure

of

licensee

to

establish,

implement,

and

maintain

a written

software

procedure

and/or administrative

policy to control

the

computer

based

dose calculation

model (50-250/88-01-04,

50-251/88-01-04).

2.

As

a

consequence

of the licensee's

failure to promulgate

the

change

in

delta-T

range,

dose

assessments

could

be

inaccurate

and

result

in

significantly nonconservative

dose

estimates

up

to

a

factor of

150.

Additionally,

the

method

used

for

obtaining

15

minute

averaged

meteorological

data

was

not sufficiently accurate

to

be

used

in

dose

projection

and assessment

calculations

(50-250/88-01-07,

50-251/88-01-07).

APPENDIX B

Appraisal

Open

Items

2.

3.

5.

O.

7.

8.

9.

10.

Determination

of reliability of

ERDADS data

used

for dose

assessment

calculations

under proposed

upgraded

ERDADS prior to integration of dose

assessment

code into

ERDADS ... (IFI 50-250/88-01-01,

50-251/88-01-01).

Evaluation of effect of containment

and

system

shine

and plant

shine

on

plant vent monitoring

system

(SPING-4)

during accident

conditions (IFI

50"250/88"01-02,

50-251/88"01"02).

Evaluation

of the accessibility

of the

sampling

station

used

to take

samples

for

source

term

development

after

an

accident

(IFI

50-250/88-01-03,

50-281/88-01-03).

Completion. of, validation

and verification documentation

for the

dose

assessment

computer

code (IFI 50-250/88-01-05,

50-251/88-01-05).

Performance

of

new

comparison

with the

NRC and state

models

following

computer

code correction, validation,

and verification.

Determination

and

documentati on

of

al l

si gni ficant

differences

(IFI

50-250/88-01-06,

50-251/88-01-06).

ERDADS provide

no concise display of containment isolation status.

While

all isolation valve positions

appeared

to

be entered

into system,

there

was

no rapid method,

on

a single

screen. to assess

their collective status

( IFI 50"250/88-01-08,

50-251/88-01" 08) .

Display of

numerous

erroneous

ERDADS values,

e.g.,

condensate

pump off

while actually operating,

containment radiation at 5.6

E+7 mR/hr; operator

aid displays

alarm in orange

and red colors, Unit 4

RCP off at

27% power,

ARM(s)

indicated

out

of

scan

or

display

of

spurious

data

(IFI

50"250/88-01-09,

50-251/88-01-09).

Addition of TSC emergency

backup lighting and automatic transfer switch to

the preventive

maintenance

program (IFI 50-250/88-01-10,

50-251/88-01-10).

Assurance

that

EPIP references

supporting plant procedures

are documented

therein (IFI 50-250/88-01-11,

50-251/88-01-11)

.

Installation of DP indicator

so that positive pressure

with the

TSC can

be

verified

during

emergency

~ ventilation

system

operation

(IFI

'0-250/88-01-12,

50-251/88-01-12).

11.

Inclusion

of

Procedure

O-OSP-301.2,

Emergency

Ventilation

System

~ Operational

Test," in preventive maintenance

program (IFI 50-250/88-01-13,

50-251/88-01-13).

(0

12.

Provi sion

of

acceptance

criteria

for

var ious

penetration

tests

in

Procedure

0-OSP-301. 1,

Emergency

Ventilation.

System

Filter

Performance

Test,"

placement

of subject

procedure

in

the

preventive

maintenance

program (IFI 50-250/88-01-14,

50-251/88-01-14)

.

13,

Updating

and

providing procedural

manual

documentation

for

the