ML18039A233

HOVs to

valve factor requirements

established

in industry testing which the

NRC

had previously reviewed.

These

comparisons

were performed to determine

if the licensee's

AVFs were reasonable.

0

4l

b.

Observations

and Findin s

14

1.

Sco

e of MOVs Included in the

GL 89-10 Pro ram

A reduction in the scope of MOVs included in the Browns Ferry GL 89-10

rogram was questioned

during Inspection 50-260,

296/95-19.

The

icensee

had reduced the scopes for Units 2 and

3 from 56

MOVs each to

36 MOVs each,

based

on a re-evaluation -of the functions of the

This reduction in scope

was identified for further review as Inspector

Follow up Item (IFI) 50-260(296)/95019-01,

Reduced

Scope of Valves in GL 89-10 Program.

In a letter dated

October

7,

1996. the

NRC informed the licensee that it

had concluded that the criteria used

by TVA in re-evaluating

the safety

functions of the Browns Ferry

MOVs were unsatisfactory

and

may have

resulted in inappropriate

removal of MOVs from the Browns Ferry GL 89-10

rogram.

An assessment

for the valves

removed

was enclosed with the

etter.

The

NRC requested

the licensee to re-examine the safety

functions of the Browns Ferry

HOVs consistent with the

NRC assessment

and to provide any appropriate corrections

to the

GL 89-10 program.

The

licensee

was asked to notify the

NRC of the findings of the re-

examination

and the actions taken

as

a result of those findings.

In a letter dated January

6,

1997. the licensee

provided

a response to

the NRC's request for re-evaluation of the safety functions of the

MOVs

removed from, or not included in, the Browns Ferry GL 89-10 program.

The licensee

stated that

15

MOVs were being added to the individual

GL 89-10 sub-programs

for Units 2 and 3.

The licensee

also stated that

plant procedures

were being revised to require that the applicable

system,

or train, for 18

MOVs (per unit) be declared

inoperable if the

valve was taken out of its normal (i.e.. safety) position for testing.

The licensee

noted that the applicable

Browns Ferry Technical

Specification Limiting Conditions for Operation would govern until such

testing

was completed

and the valve was returned to its normal position.

The letter identified the involved

MOVs and the dates

when these

commitment actions would be implemented.

The addition of 15 valves to

each unit's sub-program

was to be completed

by January

31,

1997.

The

procedure

changes

regarding operability during testing f'r 18

MOVs were

to be completed

by February

21,

1997.

During this inspection,

the inspectors verified that the 15

MOVs listed

in the licensee's

letter of January

6.

1997,

had been included in the

GL 89-10 sub-programs

for Browns Ferry Units 2 and 3.

The GL 89-10 sub-

programs for Units 2 and 3 at Browns Ferry each contained

34 gate valves

and

17 globe valves.

The inspectors

reviewed

a sample of the related

licensee's

procedures

to determine whether the associated

system or

train would be declared

when any of the identified 18 MOVs

were placed in their non-safety positions.

The inspectors

found that

the licensee fai'led to provide the appropriate

procedural

requi rements

for several

For example,

Surveillance Instruction (SI) 3-SI-

4.5.B. l.c(II), Revision 9, did not declare Unit 3

RHR Loop II inoperable

when

MOV 3-FCV-74-66 was cycled.

Similarly, SI 2-SI-4.5.B.l.c(II),

Revision 21, did not declare Unit 2

RHR Loop II inoperable

when

MOV 2-

FCV-74-66 was cycled.

The licensee's

failure to revise plant procedures

in accordance

with its January

6,

1997 letter is

a deviation from a

commitment (260,296/97-11-04).

2.

Determinations of Settin

s and Verifications of MOV Ca abilities

'I

Switch Settin

s

The licensee controlled the operation of MOVs in its GL 89-10 program

.

through

a combination of torque

and limit switches.

The torque switch

was bypassed

in the closing di rection for 95 to 98K of stroke length as

confirmed

by diagnostic data.

For opening,

the torque switch was

bypassed

for the entire stroke.

The licensee calculated the thrust and torque requirements

for MOVs in

the

GL 89-10 program using standard

industry equations.

The predicted

thrust requirements

for gate valves were calculated typically assuming

a

.0.4 valve factor with a 20K safety factor .

For globe valves,

the

licensee

assumed

a valve factor of 1.0 for closing and 1.2 for opening.

The licensee typically assumed

a stem friction coefficient of 0. 15 and

margin for potential

reduction in thrust output under

dynamic conditions

(referred to as rate of loading) of 10K for gate valves

and

15K for

globe valves.

'For MOVs with roller-screw stem nuts, the licensee

based

the stem friction .coefficient on manufacturer's

information.

Oiagnostic

error and torque switch repeatabi lity were accounted

for in the switch

setting cal'culations.

The licensee established its assumptions

for

determining predicted thrust

and torque requirements

based

on test data

principally from Browns Ferry and other

TVA nuclear

power plants.

The inspectors

noted that thrust settings for a few MOVs at Browns Ferry

had .not been

updated in the controlling setting drawings to reflect the

results of the dynamic tests

performed earlier in 1997.

The licensee

acknowledged that its

MOV setting drawings

needed to be updated.

Accuracy of the

MOV setting drawings will be re-evaluated prior to the

NRC closing its review of the Browns Ferry GL 89-10 program.

For some

MOVs, the licensee

assumed

run efficiency in the closing

direction when predicting the torque output capability of its actuators.

The inspectors

noted. that the actuator manufacturer

is preparing

new

guidelines that might affect the acceptability of the licensee's

use of

run efficiency.

Licensee

personnel

stated that they were aware of this

situation

and would address

the capability of the affected

MOVs when the

new guidance

was issued.

Oesi n-Basis

Ca abilit

At Browns Ferry, the licensee

had dynamically tested

8 gate valves

and

7

globe valves in Unit 2,

and

5 gate valves

and

7 globe valves in Unit 3

16

as part of its GL 89-10 program.

The test data for each of these valves

was used to establish its valve factor,

stem friction coefficient,

and

rate of'oading.

Using these,

the calculations

and settings

for. each of

the dynamically tested

valves were revised.

For

MOVs tested at partial design-basis

differential pressure

conditions,

the licensee extrapolated test data

from test conditions to

design-basis

conditions in evaluating thrust

and torque requi rements.

Appendix

E of MMDP-5 provided guidelines for in-plant differential

pressure testing of MOVs at Browns Ferry.

The inspectors

noted, that the

guidelines did not clear ly address test conditions that would provide

sufficient contact load to assure reliable data.

The inspectors

did not

identify any cases

where this resulted in unsatisfactory

data.

The

licensee

agreed that clarification was appropriate

and indicated that

additional

guidance for test setup conditions,

such

as prepared

by the

Electric Power

Research

Institute, would be provided to ensure reliable

extrapolation of dynamic test data.

For the non-dynamically tested

valves,

Browns Ferry still relied on the

general

valve factor assumptions

(e.g..

0.4 for gate valves) established

by the licensee's

corporate office for determining the predicted thrust

and torque requi rements.

It had not formally assessed

the validity of

these

assumptions

based

on test information.

The inspectors

found that

some of the results

from tests

performed at Browns Ferry did not support

the assumptions.

For example.

dynamic tests

completed at Browns Ferry

in 1995 (subsequent

to previous

NRC GL 89-10 inspections)

on

MOVs 3-FCV-

75-09,

3-FCV-75-37,

and 3-FCV-74-71 revealed

gate valve factors of 0.6

or greater.

The inspectors

also noted that the licensee

had not

performed

a documented

evaluation of the test data to verify the

adequacy of the program assumptions

for stem friction coefficient and

rate of loading effects.

The licensee's

fai lure to evaluate

dynamic

test data in relationship to its

GL 89-10 program assumptions

could have

resulted in MOVs being incapable of performing their safety functions.

The inspectors

found that the issue of the licensee's

failure to assess

the validity of its assumptions,

as discussed

in the preceding

paragraph,

had been identified in a licensee

self-assessment

conducted

in May 1997.

In the self-assessment

report dated

November 7,

1997, the

licensee

determined that the Browns 'Ferry GL 89-10 program was not

adequate

to support

an

NRC inspection of GL 89-10 implementation.

One

of the recommendations

in the self-assessment

report was that

justifications and analyses

were needed to strengthen

the basis for the

assumed

valve factor and other appropriate factors.

This was identified

for. resolution in Browns Ferry Problem Evaluation Report

(PER)

971770

(initiated on November

12,

1997).

The

PER stated that the licensee's

corporate

engineering office had issued

a white paper that concluded

that

a valve friction factor of 0.6 should

be used for gate valves that

could not be dynamically tested.

The

PER also indicated that the valve

factor guidance provided in Design Standard

DS-M18.2.21

was inadequate

and that .an evaluation of the

MOVs in the

GL 89-10 program was required

to ensure the abi'lity of these

MOVs to perform their safety functions.

Using

a valve factor of 0.6 for gate 'valves

and 1.2 for globe valves,

'

0

17

the licensee

reviewed the capability of its GL 89-10

MOVs to perform

.their safety functions.

In some cases,

the licensee relied on actual

dynamic test data

or bypass of the torque switch to justify the

operability of'ts

The licensee

was currently completing its

plans for justifying the design assumptions

for valve factor,

stem

friction coefficient and rate of loading effects.

In a letter dated

December

15,

1997. the licensee indicated it would revise its GL 89-10

program by January

30,

1998, to address

issues

such

as the adequacy of

the design assumptions.

The inspectors

considered

the licensee's

failure to evaluate test data to assess

its GL 89-10 program design

assumptions

to be

a violation of the requi rements of 10 CFR 50 'ppendix

B, Criterion III, "Design Control."

The inspectors verified that the

violation was not repetitive.

This non-repetitive,

licensee-identified

and corrected violation is being treated

as

a Non-Cited Violation,

consistent with Section VII.B.1 of the

NRC Enforcement Policy (50-

260,296/97-11-05).

Another example is identified in a subsequent

paragraph of this report section.

The inspectors

found that the licensee

had incorrectly evaluated

the

operability of'everal

MOVs in Problem Evaluation Report

(PER)

940066

(initiated on March 22,

1994).

In the evaluation,

the licensee

had

relied on the "stall efficiency" of MOV actuators to evaluate the

operability of the

The actuator

manufacturer

had provided

a value

for stall efficiency for'se by licensees

in evaluating potential

structural

damage in the event of an actuator

motor stall condition.

The manufacturer,

Limitorque, specifically stated that the use of stall

efficiency .was not reliable for predicting actuator output capability.

For example,

see Limitorque Haintenance

Update 92-1 (issued in 1992) or

Application Guide for Motor-Operated

Valves in Nuclear

Power Plants,

(NP-

6660-D, dated

March 1990) prepared

by the Electric Power

Research

Institute.

A current evaluation using the appropriate efficiency showed

that the valves

had been operable..

However, the inappropriate

use of

stall efficiency in predicting actuator output could have resulted in

the licensee

not recognizing

an inoperable condition of a safety-related

MOV.

The inspectors

considered

the licensee's

use of stall efficiency

to represent

a violation of the requirements

of Criterion III of 10 CFR 50, Appendix B, Design Control.

The inspectors

found that the licensee

had recognized that use of stall efficiency .was inappropriate

and that

evaluations

subsequent

to

PER 940066 did not use stall efficiency.

For

example,

the licensee

evaluated the operability of its GL 89-10

MOVs in

its recent

PER 971770 using pullout or run efficiency (rather

than stall

efficiency) together with consideration of the applicable motor curves.

The inspectors verified that the"violation was not repetitive.

This

non-repetitive,

licensee-identified

and corrected violation is being

treated

as

a Non-Cited Violation, consistent with Section VII.B.1 of the

NRC Enforcement Policy (50-260,296/97-11-05).

Another example is

identified in a preceding

paragraph of this report section.

0

MOV De radation

18

The inspectors

found that the licensee

had not provided clear guidance

for addressing

potential degradation of HOV performance.

The licensee

agreed that its guidance could be clarified and initiated steps

for this

action.

3

M~QV T

di

GL 89-10 suggested

that

MOV data

on failures

and corrective actions

(including repairs,

analyses,

as-found deteriorated

conditions, etc.)

be

periodically examined at least every two years or after each refueling

outage to establish

trends in MOV operability.

Currently, the licensee

implemented this trending through requi rements in Procedure

MHDP-5.

Until'evision 0 of MMDP-5, became effective on August 20.

1997, the

trending requirements

were specified

by Site Standard

Practice

SSP-6.51,

"Program Plan for Generic Letter 89-10," Revision 3

~ effective January

10,

1996.

MMDP-5 requi red issuance of an

MOV trending report at the

completion of every testing cycle (18 months or end of each refueling

outage).

The report was to include (1)

MOV signature analysis report

and (2) maintenance

history (corrective

and preventive).

SSP-6.51

had

required

a similar report at the same periodicity.

The inspectors

found

that the licensee

had not prepared the trending report required

by the

rocedures.

The licensee's

failure to prepare the trend report required

y procedures

is

a violation of 10 CFR. 50. Appendix B, Criterion V,

"Procedures"

(50-260,296/97-11-06).

4.

Periodic Verification of MOV Ca abilit

In Section 3. 13 of MMDP-5, the licensee stated that the purpose of

periodic monitoring of MOVs was to ensure

continued

MOV ability to

function under

design-basis

conditions,

and to identify isolated or

generic problems with MOVs or the overall program.

MNDP-5 also noted

that periodic verification should identify deterioration of MOV

components

before significant degradation

occurs.

MMDP-5 indicated that

the licensee

was working with the owners groups to share test data.

In

a letter to the

NRC dated

March 17,

1997, the licensee stated that it

was participating in a joint owners group program in response to GL 96-

05, Periodic Verification of Design-Basis Capability of Safety-Related

Motor-Operated

Valves.

The licensee's

plans for MOV periodic

verification were found to be adequate for GL 89-10.

The

NRC may re-

assess

the licensee's

long-term

MOV periodic verification program as

part of its review for GL 96-05.

I

5.

Walkdown

The inspectors

conducted

.a plant walkdown of MOVs to- assess

their

condition.

A previous walkdown described

in Inspection Report 50-259,

260, 296/95-53,

reported that cover bolts were loose or missing

on

several

The inspectors

did not observe

any similar examples

during the current inspection.

The inspectors

noted

a significant

difference in the quality of stem grease

applied to various

MOVs. This

0

1

0

19

was apparently

due to the licensee's

past

use of different stem greases.

A future closeout inspection for GL 89-10 will examine whether the

licensee's justification of'tem friction coefficient and rate of

loading effects are applicable to the different greases

or a possible

mixture of the greases.

The inspectors

al'so observed the presence of a

plastic cap

on an actuator

grease relief line for one

HOV (2-FCV-74-

106).

This was not

a

HOV.

The licensee

acknowledged

the

inspectors'alkdown

findings and stated that appropriate action would

be taken.

6.

Motor Brakes

Inspection

Report 50-259,

260, 296/95-53 indicated that motor brakes

had

not been

removed from HOVs 3-FCV-73-34, 40,

and 44.

During the current

inspection,

the inspectors

reviewed documentation

attached to BFPER

940038 which verified that the brakes

had been

removed.

c.

Conclusions

Implementation of GL 89-10 at Browns Ferry was not sufficiently

developed to permit the

NRC to complete its review.

The inspectors

did

not identify any immediate operability concerns with GL 89-10 .HOVs at

Browns Ferry.

The inspectors

identif'ied a violation involving the licensee's

failure

repare

a trend report required

by procedures

and

a deviation from a

icensee

commitment to revise procedures

to assure

the applicable

system

or train for certain

HOVs would be declared

when the

MOVs

were placed in thei r non-safety positions during operation.

In

addition,

a non-cited violation resulted

from (1)

a failure to evaluate

HOV test data to verify that it supported

GL 89-10 program design

assumptions

and (2) improper use of actuator stall efficiency.

A summary of the principal actions to be performed

by the licensee to

address

the issues identified in this inspection

and complete

implementation of GL 89-10 is as follows:

Preparation of test-based

justifications for the valve factor,

stem friction coefficient,

and rate of loading assumed

in the

design calculations for each

HOV (including MOVs with

special

features

such

as roller screw stem nuts, different stem

greases.

or mixed stem greases).

(Refer to NCV 50-260,296/97-11-

05)

Updating of calculations

and

HOV setting drawings

based

on the

justified assumptions.

Implementation of trend reporting based

on the periodic reviews

recommended

by GL 89-10.

(Refer to VIO 50-260,

296/97-11-06)

E2

E2.1

20

~

Correction of procedures

that did not implement the licensee's

commitment to assure that the applicable

system or train for

certain

HOVs would be declared

when the

HOVs were

placed in their non-safety positions during operation.

(Refer to

DEV 50-260.296/97-11-04)

In a letter

dated

December

15.

1997, the licensee

committed to revise

the Browns Ferry GL 89-10 program by January

30,

1998, to address

the

design input issues identified during this inspection.

Further, the

licensee

committed to revise its

MOV design ca1culations to reflect the

GL 89-10 program design input revisions

by Harch 31,

1998, for Unit 3,

and August 31.

1998, for Unit 2.

The

NRC will re-inspect the

program at Browns Ferry following notification from the licensee that

the issues

described in this report are sufficiently resolved to permit

verification that the intent of GL 89-10 has

been met.

The licensee

stated that

a status letter would be provided to the

NRC by March 2.

1998, indicating its progress

in completing the commitment actions.

Engineering

Support of Facilities and Equipment

Unit 2 Reactor

Scram

Due to

EHC Leak and Reactor

Protection

S stem Rela

Problem

Ins ection Sco

e

37551

93702

The inspectors

monitored

and reviewed the licensee's

investigation

and

corrective actions following a Unit 2,reactor

scram.

Observations

and Findin s

After the Unit 2 reactor

scram

on October 28,

1997, the licensee

appointed

an Incident Investigation (II) team headed

by an engineering

manager to determine the cause of the scram.

Some initial information

indicated that

a scram air header

pressure

problem was involved.

Later

on October

28, the licensee

simulated the same conditions which existed

when the scram was received

and inspected

the air system.

No problems

were found.

The low pressure

was subsequently

determined to have

occurred after the scram

as expected.

The inspectors

reviewed

information relating to previous

scrams

which supported the conclusion

that the scram air system

had performed

as expected.

The investigation continued

and early on October 29. the II team

developed

a postulated

cause of the scram.

Additional troubleshooting

and, diagnostic testing enabled the team to conclude the following

sequence

of events

associated

with the scram:

A half scram was present

on the A side of the Reactor

Protection

System

(RPS)

due to the g1 turbine control valve (CV) being closed

for repair of an electrohydraulic control fluid leak.

The 5A-K14B relay, associated

with the Bl RPS ci rcuitry,

unexpectedly

de-energized.

The team found this through

a detailed

0

21

review of computer information involving relay conditions at the

time of the scram.

The 5A-K14B relay has contacts

in the

RPS

circuitry which serve to de-energize

some of the scram solenoid

ilot valves

(as well as other functions). It is located at the

ottom of a long sequence of normally shut contacts,

any of which

open to initiate

a scram signal.

Two of those contacts

are from

one

GE HFA relay connected to the

CV fast closure circuitry which

senses

CV closure

as

a function of EHC pressure

at the valve.

The

parallel relay (5A-K14F) in the Bl RPS circuitry did not de-

energize.

The licensee

performed testing which showed that,

on

decreasing

voltage, the 5A-K14B relay contacts

began to chatter or

cycle at

a voltage value above that at which the relay (and the

5A-K14F relay) de-energized.

The 5A-K14B relay which is

a

CR105

contactor,

was replaced

and tested.

The relay will be examined

for more information on the failure mechanism.

The other

RPS scram.

solenoid pilot valve relays were also tested

and no others

were

found to display the characteristics

seen

on the 5A-K14B.

The licensee

also conducted cycling of the turbine control valves

and monitored pressure

indications at the point that

a turbine

control valve fast closure signal

(decreasing

EHC pressure)

would

be sensed.

The testing

showed that cycling of the valves

caused

notable

(300-400 psig)

downward spikes in

EHC fluid pressure.

Subsequently,

the licensee

indicated that modification of the

EHC

'ystem to reduce these osci llations (installation of orifices) was

being considered.

~

Despite testing,

the licensee

did not find any problem with the

turbine fast closure relay which has two contacts in the Bl RPS

circuitry to the 5A-K14B relay.

It is postulated that the

momentarily sensed

decreased

EHC fluid pressure

caused the relay

contacts to open just long enough to de-energize

the sensitive

5A-

K14B relay.

~

The conditions resulted in about half of the Unit 2 SSPVs getting

an "electrical" scram signal

(channel

A t, scram

and the one relay

in the Bl channel).

This would cause the Group I and

IV rods to

get

a scram signal.

This theory also conforms with an Senior

Reactor Operator's

(SRO) observation that the Group I and

IV rod

groups

scram lights appeared to have actuated

before the other two

groups.

~

A "8" RPS side backup scram valve was energized

by, the 5A-K14B (in

conjunction with the half scram) which vented air off of the

control rod drive air header

and

a full scram was actuated.

The resident

inspectors directly observed portions of the

troubleshooting

and corrective actions.

Workers replacing the 5A-K14B

relay were attentive in their work and were utilizing good stop-think-

ask-and-act

techniques.

The inspectors

examined recorder

traces

which

supported

the licensee's

conclusions

regarding

EHC fluid pressure

oscillations

and relay performance.

Portions of testing of some of the

0

22

other

RPS relays were observed.

The inspectors

noted that the pressure

switches set point for turbine fast closure input to RPS were

conservatively

above the value specified in the Updated Final Safety

Analysis Report

(UFSAR).

One of the inspectors

attended

the restart Plant Operations

Review

Committee meeting

and noted that the II team

made

an effective

presentation

regarding the electrical

equipment

performance which caused

the scram.

The inspector

questioned if General Electric

(GE) had any

additional information regarding the

EHC pressure

spike theory and the

parallel

RPS relays not being actuated.

At the time, it was not known

if additional information was available,,but

management

indicated that

it was being pursued.

The II team also

recommended

several

long term

action items including development of stronger procedural

guidelines for

repairs of a control valve servo while at power.

Prior to restart,

the

licensee tested the other scram contactors

and did not identify any

problems.

A few days later,

a

GE engineer

noted that Service Information L'etter

(SIL) 508,

Scram Contractor Coil Life and Maintenance,

addressed

a

similar issue.

The SIL was issued in February

1990 and addressed

an

unexpected

scram at another Boiling Water

Reactor

which had been

caused

by some scram contactors

dropping out during very short duration scram

signals.

The inspector obtained

a copy of the SIL and the licensee's

documented

actions in response to the SIL.

The circumstances

involved

in the scram described

in the SIL appear to be very similar to the

Browns Ferry scram.

The SIL recommended that preventive maintenance

be

performed in the contactors

and that thei r service life be reviewed.

Documentation indicated that the Browns Ferry Unit 2 scram contactors

had been replaced in 1988 and thus would not be expected to be

approaching

end of service life.

The inspector

reviewed portions of

Procedure

EPI-099-RLY001,

and concluded that it addressed

the preventive

maintenance

recommended

in the SIL.

Licensee

Event Report

(LER) 260/97-007

was submitted to the

NRC on

November 25,

1997,

and addressed

the scram.

The

LER included

description of corrective actions to reduce

EHC pressure

decreases

during control valve movement through

EHC system modifications.

After startup of the unit, several

other incidents of control valve

servo leaks occurred

on Unit 2.

(A total of six servo leaks occurred

on

Unit 2 since startup after the refueling outage).

The leaks were

identified through careful monitoring of the

EHC tank levels.

In each

case,

power

was reduced

and the servos

were replaced.

The licensee

revised the repair

procedures

to include recommendations

from the II

team.

The inspector s verified that TS requi rements

were met during the

work activities which included installation of jumpers around pressure

switches with inputs into the Reactor

Protection

System.

The licensee

formed

a team which continued to investigate the cause of

the leaks.

The investigation included extensive testing of postulated

causes

and working with the servo vendors.

Subsequently,

the licensee

0

4i

23

determined the leaks were due to vibration or pressure oscillation

induced

movement of a servo port plug which was not secured sufficiently

to prevent the movement.

The plug movement

(a few mils) resulted in

degradation of an o'ring seal

and consequently

EHC oil leaked out of the

servo.

The leaking servos

had been replaced during the most recent

refueling outage with rebuilt servos.

The licensee indicated that the

servos

are rebuilt by two vendors which utilize different means of

securing the plug.

The plugs

on the leaking servos

were secured

in a

manner that .appeared to allow slight movement of the plug.

.At the close

of this report. the licensee

was attempting to procure replacement

servos of the type which seem to allow less plug movement.

For the

interim period, the licensee

has revised work instructions for servo

replacement

such that the plug will be more securely fastened.

Unit 3

also has the same type of servos installed but no leakage

has

been

detected.

Those servos apparently were from a different batch than the

Unit 2 servos.

The licensee is continuing to investigate the magnitude

and significance of vibrations .in the Unit 2

EHC system.

Work

instructions

have been

developed to install

a pressure

transducer

on

a

Unit 2 control valve during the next replacement that is performed.

The

device will monitor

EHC fluid pressure

at the control valves.

Problem Evaluation Report

(PER) 971714 described

the investigation

and'orrective

actions

regarding the

EHC leaks.

The

PER noted that

additional

EHC system modifications. including installation of

accumulators

and control valve manifolds,

are planned

as part of the

Power Uprate Project.

These modifications are referenced

in GE Turbine

Information Letter 1123,

issued

November

14,

1992.

The modifications are

intended to reduce

EHC fluid pressure oscillations.

Operations is

continuing to closely monitor

EHC reservoir fluid level

on both units.

The inspectors verified that the enhanced

monitoring is being properly

performed.

Conclusions

The inspectors

concluded that the l.icensee's

incident investigation

team

performed well in investigating the equipment

issues

which caused

the

scram.

The team studied available

information, developed

a theory of

the cause,

and in a prompt manner,

performed electrical testing which

supported the postulated

cause.

The investigation into the

EHC fluid

leaks

was also thorough

and determined the cause of the leaks in a

reasonable

time, period.

RCIC Lube Oil Particulate

Levels Increased

Ins ection Sco

e

37551

The inspector

reviewed data

and developed

a time line to assess

the

licensee's

actions in response to -high particulate levels in the Reactor

Core Isolation Cooling (RCIC) system lube oil.

Oil samples

indicated

a

articulate count which was in the action range

as identified by the

icensee's

procedures

and documentation in the vendor manual.

l

b.

Observations

and Findin s

24

On October 20.

1997,

a lube oil sample

was taken from the

RCIC system

following work which had been performed during the Unit 2 outage which

ended October

18,

1997.

Oil analyses

results

were issued

from the

licensee's

Central

Laboratory on October 22,

1997, which identified

a

particulate count of ISO 20/15.

On November 3,

1997, the licensee initiated a work request

C337743 to

sample the oil again to verify the high particulate content

and change

the oil if necessary.

On November 6,

1997, the sample

was taken to

verify Central

Laboratory results.

Browns Ferry laboratory determined

the particulate count,to

be high and that

an oil change

was required.

On November 21,

1997, the licensee

changed the oil in the

RCIC system.

Sampling following the oil change indicated

a high level of particulate.

The licensee

changed the oil and sampled for particulate levels several

times in an, attempt to decrease

particulate levels to within the normal

range.

Analyses of the oil sample taken from the RCIC system following

a 15 minute run on November 23.

1997,

showed

a particulate level of ISO

18/14 which placed the system in the Alert Range.

The Alert Range

requi res that the lube oil be filtered or replaced at the next refueling

outage if the alert limits persist.

On November

23,

1997, the licensee

declared the RCIC system operable.

Chemistry Instruction (CI) CI-130, Diesel

Fuel

and Lube Oil Monitoring

Program,

Revision

5

~ Attachment 3, describes

the RCIC lube oil

particulate guidelines.

The procedure

describes

three ranges:

Normal

Range

(ISO 16/13), Alert Range

(ISO 18/15),

and Action Range

(ISO

>18/>15).

The procedure states that the particulate counts which exceed

the alert range require immediate corrective action to be taken to

restore the oil to normal or alert levels.

The procedure further states

that the HPCI/RCIC system

may remain "functional" for a period not to

exceed

30 days while resampling / filtering / replacement activities are

in process.

The licensee

interpreted the procedural

requirements to

mean that the oil should

be changed within 30 days.

The RCIC,oil was

changed within thirty days of the date when the oil analysis results

were issued

from the licensee's

Central

Laboratory on October 22.

1997.

The licensee is in the process of developing guidance for the lube oil

program on site.

In inspection report 259,260,296/97-08.

the inspector

concluded that

weaknesses

in the licensee's

lubrication oil analysis

program permitted

the incorrect type of lubricating oil to be added to a second

EDG

several

months af'ter it had been installed in a different

EDG.

c.

Conclusions

Al.though the licensee's

actions to decrease

the oil particulate levels

in the

RCIC system were adequate,

the

RCIC oil particulate level issue

represents

another

example of the difficulties that the licensee is

II

E8

E8.1

25

experiencing with the lube oil analysis

program.

The licensee

has not

determined the cause of increased particulate but has determined it is

not due to excessive. bearing wear.

Miscellaneous

Engineering

Issues

(92902)

Closed

Ins ection Fol,low u

Item 50-'260 296/95-19-01:

reduced scope'of

valves in GL 89-10 program.

This item involved the licensee's

removal

of a number

of. valves from its GL 89-10 program.

The licensee's

actions

to resolve this item are discussed

in E1.3.b.1

above.

The licensee's

.

actions in returning valves to its program were satisfactory,

except

as,

addressed

by the violation described in that section of this report.

R2

R2.1

IV.

Plant

Su

ort

Status of Radiation Protection

and Chemistry Facilities and Equipment

Radi oacti ve Effluent Honitor Probl ems

Ins ection Sco

e

37551

71750

During review of operating logs, the resident inspectors

had noted

discussion of a problem involving the radioactive effluent monitor (0-

RH-90-130).

The inspector

questioned

resolution of the problem and the

'bsence

of a Problem Evaluation Report at

a subsequent

Management

Review

Committee

(HRC) meeting.

Several

days later, the problem occurred

again.

The inspectors

questioned

plant management

regarding the

problems

and the oper abi:lity status of the monitor .

The licensee

formed

an investigation team.

The inspectors

reviewed the regulatory

and

procedural

requirements

for the monitor,

examined the detector

and

associated

piping,

and monitored the licensee's

investigation.

b. Observations

and Findin s

The resident inspectors

noted that the October 21,

1997. operating

logs

described

an incident in which the radwaste

operators

had secured

an

effluent release after it was noted that the radiation monitor was

indicating an activity level which dropped to less than the background

level recorded before the release

began.

In accordance with the Offsite

Dose Calculation Manual, the monitor is required to normally be operable

during releases.

Releases

are permitted to be continued with an

inoperable monitor if compensatory

measures

are completed.

The release

was subsequently

restarted.

At a Management

Review Committee meeting,

the resident inspector questioned

the

resolution of the problem and .the

.absence of a Problem, Evaluation Report

(PER).

Subsequently,

maintenance

personnel

reported to the inspector that the problem had been attributed

to demineralized water remaining in the detector after

a cleaning

evolution.

C

0

26

On October 27, the problem occurred again.

The monitor was not declared

and two releases

were made without compensatory

sampling

performed.

PER 971713 was initiated on October

28 to address

the issue.

After HRC review of this

PER on October 29. the inspector

questioned

the

operability status

of. the monitor since releases

were being

made without

completion of compensatory

measures

required

by the Offsite Dose

Calculation Manual

(ODCH).

The inspectors

also questioned

plant

management

regarding resolution of the problems.

The monitor was

declared

inoperable (administrative decision)

on October

29 and

compensatory

measures

were completed for all releases.

An Incident

Investigation (II) team was formed .to investigate the problems.

After some investigation,

the team identified that the problem was

caused

by water leaking out of the detector

housing

volume through

a

closed drain valve.

Maintenance

personnel

found an accumulation of crud

on the valve seat which allowed leakage

through the valve.

Drainage of

water out of the volume decreased

the shielding between the detector

and

the chamber walls.

When

a release

was started,

the volume was refilled

with low activity water and the radiation levels sensed

by the detector

wer e reduced.

The valve was replaced. and the licensee

subsequently

conducted testing .which supported

the postulated

cause.

During a

subsequent

release,

the decreased

radiation levels were not observed.

The investigation

team concluded that the RH-130 monitor had, in fact,

been functional despite the water draining problem.

The team also

verified that the

ODCM compensatory

measures

had been completed for all

periods in which the monitor had been inoperable.

The inspectors

reviewed portions of Technical

Instruction O-TI-45, Liquid Process

Radiation Monitors.

This procedure

determines

the alarm setpoints

as

required

by the

ODCH.

At Browns Ferry, 'liquid radwaste

batch discharges

are controlled by procedure O-SI-4.8.A, Liquid Effluent Permit.

Representative

samples

are analyzed prior to the discharge

and the

monitor serves

as

an independent

check during the discharge.

The

inspectors

noted

numerous

conservatisms

were applied, in the set point,

determinations

for 'RH-130, including conservative

assumptions

regarding

condenser circulating system (dilution) water flow.

After review of the

procedures

and discussions

with the team, the inspectors

concluded that

the decreased,

background effects observed

did not adversely affect the

.operability of the monitor.

On December

2,

1997, the .proposed corrective actions for

PER 971713 were

reviewed, by the Management

Review Committee

(HRC).

Initially, the

actions did not address

the inspector's principle concern that observed

abnormalities

associated

with the RM-130 were not fully understood

and

radioactive material

releases

were permitted to continue without

compensatory

measures.

During the

HRC meeting, the Site Vice President

directed that

a corrective action

be added which addressed

this concern.

The inspector discussed

the concern with the Operations

Manager

and

verified that the concern would be adequately

addressed.

\\

ll

0

c.

Conclusions

27

Unexpected indications

on the radioactive effluent monitor

recorder

were

not thoroughly investigated prior to discontinuation of compensatory

actions

required for an inoperable monitor.

The licensee

subsequently

completed

a detailed

investigation which identified that

a leaking

valve had caused

the unexpected

indications.

The investigation also

concluded that the monitor was not inoperable

and that regulatory

requirements

were met during, the period.

Xl

Exit Meeting Summary

V.

Mana ement Meetin s

The resident inspector presented

inspection findings and results to

licensee

management

on December

12.

1997.

Other formal meetings to

discuss

report issues

were conducted

on November

21,

and

December

3,

1997.

The licensee

acknowledged

the findings presented.

Proprietary

information is not included in this inspection report.

PARTIAL LIST OF

PERSONS

CONTACTED

T. Abney, Licensing Manager

J. Brazell, Site Security Manager

R. Coleman, Acting Radiological Control

Manage

M. Cooper,

Corporate

Component

Engineering

Manager

J.

Corey, Radiological Controls

and Chemistry

Manager

T. Cornelius,

Emergency

Preparedness

and Planning

C. Crane, Site Vice President,

Browns,Ferry

R.

Greenman,

Training Manager

J.

Johnson,

Site Quality Assurance

Manager

R. Jones,

Assistant Plant Manager

G. Little. Acting Operations

Manager

D.

Nye, Site Engineering

Manager

J. Schlessel.

Acting Maintenance

Manage

K. Singer,

Plant Manager

INSPECTION

PROCEDURES

USED

IP 37551:

IP 40500:

IP 62707:

IP 71707:

IP 71714:

IP 71750:

Onsite Engineering

Licensee Self-Assessments

Maintenance

Observations

Plant Operations

Cold Weather Preparations

Plant Support Activities

0

(

0

IP 9Z901:

IP 92902:

IP 93702:

TI 2515/109:

28

Follow up-Plant Operations

Follow up-Maintenance

Prompt Onsite Response to Events at Operating

Power Reactor

Implementation of Generic Letter 89-10

ITEMS OPENED

DISCUSSED

AND CLOSED

OPENED

~T

e

Item Number

IFI

260,296/97-11-01

260.296/97-11-02

URI

260,296/97-11-03

DEV

260,296/97-11-04

NCV

260,296/97-11-05

260,296/97-11-06

CLOSED

T~e

Item Number

.260/96-06-02