ML17291A132

in the

SSPVs of Control Rods 02-19

and 14-55, respectively.

These

failures resulted in reactor

operation for 7 weeks with a significantly

degraded

control rod system

and the failure of Control Rod 06-39 to insert

during surveillance testing conducted

on March 26,

1994.

9406150024

940606

PDR

ADOCK 05000397

8

PDR

'

4

~

The supply system failed to adhere to procedures

requiring the performance of

prompt operability assessments

(POAs)

when finding physical

evidence of

degraded

equipment.

On December

4,

1993, the licensee

found the time for SON

of Control Rod 14-55 to have degraded

by approximately

a factor of three

and

did not perform

a

POA.

On February

7 and

17,

1994, the supply system

found

significantly degraded

in the

SSPVs of Control Rods 02-19

and

14-55, respectively,

and failed to perform POAs.

The supply system failed to adequately

trend rod control system performance

and failed to implement

an effective preventive maintenance

program in

response

to industry information associated

with Buna-N components

in the

control rod SSPVs.

Weaknesses:

This inspection identified additional

examples of the following problems that

the

NRC has previously identified at the supply system:

failure to take

prompt corrective actions

when finding degraded

safety-reTated

components,

inadequate

evaluation

and corrective action for problems identified by

industry communications,

inadequate

management

oversight,

poor communications

between organizations,

and the failure to adhere to procedures.

Summar

of Ins ection Findin s:

~

Apparent Violation 397/9415-01

was opened

(Section 6).

~

Apparent Violation 397/9415-02

was

opened

(Section 6).

Attachment:

~

Persons

Contacted

and Exit Meeting

~

T

DETAILS

1

PLANT STATUS

At the start of this inspection period, the plant was operating at

approximately

68 percent

power with five control rods fully inserted.

Control Rods 54-47,

54-15,

30-31,

and 10-15 were fully inserted

because

operators

had

declared

these

rods inoperable

when the control rods exhibited slow

SOM during

surveillance testing performed

on March 26,

1994.

Control Rod 30-03 was fully

inserted

on March 26,

1994, since it could not,be

scram time tested

due to a

failed transponder

card that was located in the"testing circuitry.

Licensee

craftsmen

were replacing the Buna-N components

(diaphragms), of the

SSPVs for

the four control rods that had exhibited slow SOM.

Because

Control Rod 06-39

had failed to insert

on March 26,

1994, licensee

engineers

were testing the

diaphragms that

had

been

removed

from Control Rod 06-39.

At the conclusion of

the period,

the reactor

was shut

down in Mode

5 (Refueling)

and the Buna-N

diaphragms of all the

had

been replaced.

2

CONTROL ROD SYSTEM DESCRIPTION AND DESIGN BASES

The control rod system is one of the reactivity control

systems for the

WNP-2

boiling water reactor

(BWR).

The reactor control rods are positioned

by

hydraulics.

The control rods

can

be positioned in small increments

or rapidly

inserted into the core

(scrammed).

Rapid control rod insertion is achieved

by

the following means:

(1) automatically,

the reactor protection

system

deenergizes

the

SSPVs of each of the

185 control rods which redirects

the

pneumatics of each individual control rod;

(2) automatically,

the reactor

protection

system energizes

the backup

SSPVs which redirects

the pneumatics of

all control rods;

and (3) automatically,

when reactor level reaches

-50 inches

or reactor pressure

reaches

1076 psig, the anticipated transient without

scram automatic

rod insertion

scram system which redirects the pneumatics of

all control rods.

Each of the individual control rod SSPVs,

which the

licensee

procured

as guality Class

1 components,

has four normally energized

Automatic Switch Company

(ASCO) solenoid operated

valves that have internal

components

made of Buna-N rubber.

The system also contains

two normally

deenergized

backup

SSPVs consisting of two ASCO solenoid valves,

which were

procured

as guality Class

2 components.

The internal

components

of the backup

scram valves are

made of Buna-N rubber.

The anticipated transient without

scram-automatic

rod insertion

system

uses

valves that do not contain

Buna-N

components.

The design basis of the reactivity control

system is to provide sufficient

nuclear reactivity control devices

(control rods) to control excess reactivity

in the core

and to provide For adjustments

of the control rods to permit power

generation.

The safety functions of the reactivity control

system are to

provide sufficient excess

negative reactivity to keep the reactor shut

down

and to provide sufficient rapid insertion of control rods (reactivity) so that

no fuel

damage results

from any operating transient.

3

CONTROL ROD DRIVE (CRD)

SYSTEM TS AND BASES

The following TS are applicable to the control rods while in operational

conditions

one

and two:

TS 3. 1.3. 1

All control rods shall

be operable.

TS 3. 1.3.2

The maximum scram insertion time of each control rod from the

fully withdrawn position to notch Position 6,

based

on deenergization

of the

'SPVs

as time zero, shall not exceed

7 seconds.

TS 3. 1.3.4

The average

scram insertion time of all operable control rods

from the fully withdrawn position, for control rods arranged

in a two-by-two

array,

based

on the deenergization

of the

as time zero shall not exceed

any of the following:

I

Position Inserted

From Fully Withdrawn

Average

Scram Insertion

Time (Seconds)

45

39

25

5

0.430

0.868

1.936

3.497

The following sentences

describe

the bases for these

TS.

The requirements for

the various

scram time measurements

ensure that any indication of systematic

problems with rod drives will be investigated

on

a timely basis.

The

occurrence of scram times longer than those specified

should

be viewed

as

an

indication of a systemic

problem with the rod drives and, therefore,

the

surveillance interval is reduced

in order to prevent operation of the reactor

for long periods with a potential

serious

problem.

The number of control rods

permitted to be inoperable

could be more than eight allowed by the TS, =but the

occurrence of eight inoperable

rods could be indicative of a generic

problem

and the reactor must

be shutdown for resolution of the problem.

4

INDUSTRY EXPERIENCE

(92904)

4. 1

NRC Information Bulletin

IB

78-14

"Deterioration of Buna-N

Com onents

in ASCO Solenoids"

On December

19,

1978, the

NRC issued

IB 78-14 to notify licensee's

operating

BWRs about the thermal

aging of SSPV

Buna-N components.

This IB required that

licensees

with an operating license:

(1) review Buna-N material applications

in control rod scram systems,

determine the time since installation,.and

describe

a schedule for replacement,

both in response

to this IB and for

periodic maintenance;

(2) report the results of that review;

and (3) describe

the basis for the replacement

schedule

and describe

any proposed

replacement

time in excess of 3 years.

For all

BWR licensees

with construction permits,

which applied to WNP-2 in 1978, this IB was provided for information only.

The licensee initiated

a plant tracking log (PTL) item to track the

development

and implementation of a preventive maintenance

task to perform

periodic replacement

of Buna-N components

in SSPVs.

In an interoffice memorandum,

licensee

management

determined to replace

Buna-N

and 0-rings of SSPVs at

a rate of 25 percent

per year, starting in

April 1986.

In an inspection

conducted during March 1-5 and March 15-19,

1982,

the

NRC closed this IB based

on the licensee's

interoffice memorandum

plan to initiate

a 25 percent

replacement

program.

During this inspection,

the inspectors

found that the licensee

did not implement this plan.

This

deviation from a commitment is not being cited due to the

age of the issue,

the complete refurbishment of the

SSPVs in 1994,

and the supply system's

intent to replace

the

Buna-N components

with Viton components

in 1995.

The inspectors

found that the licensee

replaced

the Buna-N components

of

10 percent of the control rod

SSPVs in 1986.

Following this replacement,

the

licensee

changed

the preventive maintenance

scheduling

method.

As a result of

this change,

the planned

replacements

in 1987,

1988,

and

1989 did not occur.

In Hay of 1989,

when the

SSPVs of two control rods failed due to hardened

Buna-N components,

the licensee identified that they had not implemented the

specified replacement

program.

Therefore,

the licensee

decided to replace the

Buna-N components

of all the

SSPVs in Hay of 1990.

The licensee

did not

notify the

NRC that they had failed to implement the

25 percent

replacement

commitment.

The licensee

also failed to notify the

NRC that they did not plan

to implement

a 25 percent

replacement

following the

1990 refurbishment,

since

they had qualified the

Buna-N components

to a service life of 5.5 years

based

on their operating experience.

4.2

NRC Information Notice 86-78

"Scram Solenoid Pilot Valve Rebuild Kit

Problems

On September

2,

1986, the

NRC issued this notice to alert licensees

to

potential

problems with kits used to refurbish

This notice described

quality control

problems

associated

with the

SSPV rebuild kits.

The licensee

took no action in response

to this notice.

The licensee

noted that,

in

May 1986,

38 SSPVs

had

been rebuilt and were properly functioning.

The system

engineer

noted that close attention would be given to the control rods with

the rebuilt SSPVs.

4.3

NRC Generic Letter

GL 91-015

"0 eratin

Ex erience

Feedback

Re ort

Solenoid-0 crated

Valve Problems At U.S. Reactors"

This

GL dated

September

23,

1991, required

no specific action or written

response

by the licensee.

However, the

NRC expected that the licensee

would

review the information presented

in the

GL and consider actions

as appropriate

to avoid similar problems.

The

GL specifically mentioned control rod failures

in

BWRs due to hardening of Buna-N components.

On October

22,

1991,

the licensee

performed

an initial screening

and

determined that the

GL was applicable to WNP-2.

The licensee

completed

the screening

on September

2,

1993,

by assigning

actions to evaluate

the

\\I

-6-

recommendations

of NUREG 1275,

Volume 6, "Operating Experience

Feedback

Report,

Solenoid-Operated

Valve Problems."

The actions

were being tracked

by PTL 44726.

This

PTL item was

open at the time of the failure of Control Rod 06-39 to insert.

The

NRC inspectors

found that the lack of timeliness

of this review had

been previously discussed

in NRC Inspection

Report 50-397/94-08.

The failure to consistently

perform detailed timely

assessment

of industry experience

is

a continuing problem.

4.4

General Electric

GE

Service Information Letters

SILs

4.4.1

GE SIL Number

128

GE issued

SIL Number 128, Revision 0, "Preventive Maintenance for Control

Rod

Drive (CRD) Scram Pilot Valves,"

on March 31,

1975.

The SIL discussed

failures,

due to diaphragm cracking, of SSPV

Buna-N components.

The purpose

of the SIL was to recommend

a maintenance

program for scheduled

replacement

of

all

SSPV diaphragms.

GE recommended

that licensees

establish

a program

so

that all

were replaced after 3-4 years of initial service.

The SIL noted that the initial operating service life was due to the longer

storage

time of the rubber parts prior to initial plant operation.

GE also

stated that the replacement

of these

components

could be performed while the

reactor

was operating at high powers.

Because

WNP-2 was in construction,

the

licensee

took no immediate action.

GE issued

Revision

1 to this SIL on January

30,

1976.

This revision notified

customers of the unacceptability of performing diaphragm replacements

at power

since the replacement

conflicted with GE fuel preconditioning

recommendations

for minimizing CRD movements

at high power levels.

Because

WNP-2 was in

construction,

the licensee

took no immediate action.

GE issued

Revision

1, Supplement

1, to this SIL in August 1978.

This

supplement .notified customers

of the delayed

CRD insertion of two control rods

at

an operating

BWR.

The purpose of the SIL was to discuss

the cause of the

occurrence

and to recommend

a preventive

program to help in preventing future

occurrences.

GE recommended

the following:

(1) establish

a preventive

maintenance

program to replace all core assemblies,

diaphragms,

and associated

parts in all SSPVs,

backup

scram valves,

and scram discharge

volume vent and

drain pilot valves at periodic intervals,

and (2) rebuild the

periodically to assure

that the Buna-N parts

do not exceed

a combined shelf

life-service life of 7 years.

Because

WNP-2 was in construction,

the licensee

took no immediate action.

The licensee initiated operation evaluation report

(OER)

75003 to track the development

and implementation of a preventive

maintenance

program for the

GE issued

Revision

1, Supplement

2, to this SIL on March 2,

1984.

This

supplement notified customers

to discontinue

using "Loctite" on the acorn nut

on the solenoid housing.

In this supplement,

GE reemphasized

their

recommendation

that the Buna-N components for the

CRD SSPVs should

be replaced

periodically to assure

that the

age of these

components

would not exceed

7 years.

The licensee

incorporated this recommendation

in OER 75003.

On July 2,

1984, the licensee

closed out all actions associated

with this SIL

by revising their scheduled

maintenance

system to include

a preventive

maintenance

task that would begin replacement of Buna-N components,

in

25 percent of the

SSPVs,

in April 1986.

4.4.2

GE SIL Number

575

'E issued

SIL 575;

"CRD Slow Start of Motion (SOH)," on October 27,

1993.

This SIL described

the failures

and offered recommendations

for preventing

recurrence of CRD slow SOH.

The purpose of the SIL was to recommend

actions

which would reduce

the probability that slow scram

SOM times would occur.

GE

stated that there

was

no safety concern

because

scram time surveillances

would

identify any slow

SOH drives.

GE provided the following recommendations:

(1) routinely check the scram air header

pressure

to be wi.thin 70 to

75 pounds;

(2) pay close attention to degradation

of the

SOH times and, if

unusual

SOM times are detected,

replace

the Buna-N components;

and (3) Buna-N

components

should

be limited to approximately

4 years of plant service

and,

before extending the service life of Buna-N components

beyond the recommended

3-4 years,

evaluate

the components

based

on plant specific operating service

conditions to determine whether they will perform acceptably.

On November

14,

1993,

the licensee initiated t)ER 75003G to consider the

recommendations

of this SIL.

On January

21,

1994, the licensee

closed the

OER

after having implemented actions to address

GE's recommendations.

In their

response

to this

OER, the licensee

noted that their program was working

because

surveillance testing

had identified that Control Rod 14-55 had slow

SOH on December

4,

1993.

However, the licensee failed to note that the delay

in replacing the

SSPVs of Control Rod 14-55 did not conform to GE's

recommendation

to replace the Buna-N components if unusual

SOH times were

detected

and that this delay resulted

in operating in violation of TS 3. 1.3.4.

The licensee

implemented

the other recommendations

of this SIL.

4.5

Conclusions

During the period of 1986 through

1989, the licensee failed to implement the

commitment to perform an annual

replacement of the Buna-N components

in

25 percent of the control rod SSPVs.

Further,

the inspector concluded that

the Supply System

was not timely in evaluating

GL 91-15

and did not

effectively implement all the recommendations

of GE SIL 575.

5

CONTROL ROD 06-39

FAILURE TO INSERT

(92901)

5. 1

Licensee

Immediate

Res

onse to Control Rod 06-39 Failure to Insert

At 1:43 p.m.

(PST),

Saturday,

March 26,

1994, with the reactor at 75 percent

power, Control Rod 06-39 failed to insert when

WNP-2 operators initiated

a

scram signal during surveillance testing.

Licensed operators

concluded that

the control rod had malfunctioned

because

the positions of Control Rod 06-39

switches

and valves

had

been verified as correct.

At 1:52 p.m., the operators

manually inserted

Control Rod 06-39, deenergized

its electrical

components

and declared it inoperable

in accordance

with WNP-2

TS 3. 1.3. l.a. l.a.

The shift manager

(SH) notified supply system

management

of

the malfunction;

however,

he did not notify the

NRC resident

inspector of this

failure.

The

SH initiated problem evaluation request

(PER)

294-0235 to document the

failure of Control Rod 06-39 to insert.

He initiated emergency

Work

Order

HS 28 to remove,

disassemble,

and rebuild CRD-V-117 and CRD-V-118, the

SSPVs for Control Rod 06-39.

At 3 p.m., the

SH decided to perform scram time

testing

on 20 of the

185 control rods to eliminate the concern over

a generic

failure mode.

Subsequently,

the plant manager

(PM) concluded that scram

testing of all

185 control rods

was necessary

to ensure

acceptable

operation.

At 8: 16 p.m.,

on March 27,

1994, the 'operators

completed

the scram time

testing of the

185 control rods.

The testing identified that four control

rods

(54-47,

54-15, 30-31,

and 10-15) were slow on

SOH, but within the time

limits of TS 3. 1.3.4.

Additionally,'perators

had also fully inserted

Control Rod 30-03 because it could not be scram time tested,

due to a failure of a

transponder

card in the control rod testing circuitry.

At approximately

6 p.m.,

March 26,

1994, the system engineer

determined that

Control Rod 06-39

had failed to insert because

the Buna-N exhaust

diaphr agm,

an environmentally qualified (Eg) component of SSPV CRD-V-118,

had failed due

to hardening.

The system engineer characterized

the failure as brittle

cracking of the exhaust

diaphragm of CRD-V-118, with the crack going

approximately

180 degrees

around the interior circumference of the exhaust

diaphragm,

and hardening without cracking of the pressure

diaphragm.

At

8 p.m., the licensee

concluded that the failure of the exhaust

diaphragm could

have

caused

the failure of the rod to scram

due to the air leakage rate into

the exhaust

being greater

than the rate being exhausted

through the

vent.

On March 27,

1994, at 3:58 p.m., the repair of the

SSPVs of Control Rod 06-39

was completed

and operators

declared that control rod operable.

At 8: 16 p.m.,

operators

completed

the scram time testing of all

185 control rods

and

increased

reactor

power to 88 percent.

Based

on the guidance

provided in licensee

Plant Procedure

Manual

(PPH) 1.3. 12B, "Operability Evaluation," Revision 0, the

SM considered that,

with the exception of the four rods which were slow on

SOH and the one rod

that could not be tested,

all the control rods were operable

because

the rods

had met surveillance

requirements.

Paragraph

3.7 of PPH 1.3. 12B states that,

when qualification is called into question,

performance of surveillance

requirements

alone

may not

be sufficient to determine operability, unless

the

surveillance testing adequately verifies the capability of the equipment

by

accounting for the degraded

condition.

5.2

Licensee

Followu

Res

onse to Control Rod 06-39 Failure to Insert

At

1 p.m.,

on March 28,

1994, the

PH met with resident

inspectors

and

discussed

the failure of Control Rod 06-39 to insert.

The

PM confirmed that

he considered that the control rods were operable

because

the control rods

had

repeatedly

met surveillance

requirements

during the testing performed over the

weekend.

The

PM discussed

the safety significance of the event

and the supply

system's

planned actions to prevent event recurrence.

The planned actions

included replacing the Buna-N components of the

SSPVs for the five control

rods that had

been declared

inoperable.

The

PH considered

that the failure

of Control Rod 06-39 to insert

and the slow SO) of the other four control

rods

had limited safety significance

because

the safety analysis

assumes

one

rod does not insert

and that

up to eight control rods

can

be slow without

adverse

impact to fuel.

He also noted that the backup

scram valves would

probably have functioned, resulting in the insertion of all control rods

(including 06-39).

At 4:45 p.m., the

PH agreed to

a

10 a.m.,

March 29,

1994,

conference call with the

NRC to discuss

the failure of Control Rod 06-39 to

insert

and licensee corrective actions.

On the evening of March 28,

1994, after having examined the diaphragms of the

SSPVs that

had

been refurbished,

the licensee

decided to rebuild the

SSPVs of

all

185 control rods (six at

a time) with the reactor operating at reduced

power.

At 10 a.m.,

March 29,,1994,

the licensee,

NRC Region V, and the Office of

Nuclear

Reactor Regulation

(NRR) had

a conference

telephone call to discuss

the events of March 26,

1994, licensee

actions to prevent recurrence

of these

events,

and supply system

and industry experience with Buna-N components

in

control rod SSPVs.

The participants

discussed

continued

power operation with

degraded

SSPV diaphragms,

the operability of the control rods,

and the

replacement

of the

while at power.

The licensee related that

they had reviewed the scram times of the control rods

and believed that they

could predict the condition and failure of the

based

on scram

times.

The licensee

stated that they planned to have

an operability

determination

and

a basis for continued operation

(BCO) completed

by March 30,

1994.

The

PM also committed to scram testing the control rods each

week and

to immediately shut

down the reactor if another control rod failed to insert

during the

SSPV diaphragm testing

and replacement.

From Harch 29-31,

1994, the licensee

continued to rebuild SSPVs,

to collect

and organize data to incorporate

in the

BCO,

and to hold Plant Operating

Committee meetings to discuss drafts of the

BCO.

The licensee

found that the

diaphragms installed in the

had

come from procurements

in 3 separate

years

1983,

1987,

and

1990.

Furthermore,

th'e licensee

found that they

h'ad

not maintained records to be able identify which lot diaphragms

were in which

valves.

A technical

expert from GE arrived

on site to assist

the licensee

in

assessing

the degradation of the control rod SSPVs.

On March 31,

1994,

the licensee,

NRC Region V, and

NRR had

a conference

telephone call to discuss

the licensee's

progress

in developing

and

-10-

documenting

an operability determination

and

BCO.

The licensee

discussed

their assessment

of the data that had

been recently collected.

The licensee

noted that,

instead of refurbishing the four SSPVs of each control rod, only

the exhaust

diaphragms of each

CRD-V-118 valve would be replaced.

When

performing this replacement, if the exhaust

was found to be brittle,

the pressure

diaphragm of the CRD-V-118 valve would also

be replaced.

The

licensee related that the

BCO had not been completed,

but would be available

in draft for discussion

on April 1,

1994.

On April 1,

1994, the licensee,

NRC Region

V, and

NRR had

a conference

telephone call to discuss

the draft

BCO.

In the draft

BCO the licensee

concluded

the following:

(1) data taken

and reviewed

showed that when severe

degradation

is present,

there is increased

scram time;

(2) catastrophic

failures have not resulted in failure to scram,

but rather increased

scram

time;

(3) shelf-life and storage

are minimal contributors to the degradation,

and installed service in the plant under elevated

temperature

and stress

are

major contributors to degradation;

(4) the failure of Control Rod 06-39 was

an

unusual failure of two or more diaphragms

in the

SSPVs;

(5) based

on scram

time testing,

there were

no SSPVs

and control rods which would not perform

their safety function.

With respect

to operability, the

BCO stated that successful

testing

demonstrates

that the currently installed elastomers

are meeting their design

basis function.

However, all remaining diaphragms

were being replaced

on

an

expedited basis.

The

NRC participants

had the following comments with respect

to the

BCO:

(1) it did not appear that time to scram

was

a reliable

means of determining

the extent of degradation

and, without understanding

the extent of

degradation, it would be difficult to predict the remaining useful life of the

diaphragms;

(2) the

BCO stated that it took the failure of two diaphragms

to

cause

the failure of a control rod to insert but that the licensee

could not

reproduce this in testing

and the pressure

diaphragm of CRD-V-118 for Control Rod 06-39

had not failed, it was only stiff and brittle; and (3) the licensee

believed that the degradation

was temperature

not support the assumption.

The licensee

stated that these

comments

would be factored into the final

BCO.

The

PH reiterated that

he believed all the control rods would function if

required.

On April 8,

1994,

the licensee,

NRC Region IV, the Walnut Creek Field Office,

and

NRR had

a conference

telephone call to discuss

the final

BCO.

The

BCO

discussed

the previous

NRC comments.

The licensee

stated

they believed that

the failures of the diaphragms

was

a generic industry issue

and that they

would continue with their root cause

evaluation.

They statused

the diaphragm

replacement,

stating the effort would be completed

by April 12,

1994.

At the close of the inspection period,

the licensee

was continuing with their

root cause evaluation.

To date,

they had identified that all the severely

degraded

appeared

to have

been procured in 1989, that there

appeared

to be manufacturing

and chemical differences

between

the diaphragms

procured

in 1989

and those

procured earlier,

and that the configuration of the

failure of the exhaust

was critical in determining the control rod

insertion time.

5.3

Conclusions

The licensee

was not timely in developing

a

BCO.

The licensee's initial

operability assessment

appeared

to rely heavily on control rods meeting

surveillance

requirements

and not on data collected

from the disassembly

and

examination of the

SSPVs of Control Rods 02-19,

14-55,

and 06-39.

The

licensee

refurbishment of the

Control Rod 14-55 inoperable,

was generally well controlled.

6

FOLLOWUP OF CONTROL ROD SURVEILLANCE TESTING AND REPAIR

(92701)

6. 1

NRC Followu

Ins ection of the Failure of Control Rod 06-39 to Scram

and

the Slow

SOH of Four Other Control

Rods

On March 28,

1994, during the resident

inspector's daily plant walkdown, the

inspector learned that Control Rod 06-39 failed to automatically insert, that

Control Rods 54-47,

54-15, 30-31,

10-15 were slow on

SOH,

and that Control Rod 30-03 could not be scram time tested

due to a failed transponder

card.

In

reviewing the event against

the

WNP-2 TSs,

the resident

inspectors identified

that the licensed

operators

had incorrectly logged that the facility entered

limiting condition for operation 3. 1.3. l.a. l.a instead of 3. 1.3. l.b. l.b.

This

error had

no safety significance;

however,

the error indicates

a weakness

in

operator understanding

of the TSs

and incomplete

management

oversight.

The

operators

corrected

the log entry on March 29,

1994.

At

1 p.m., the resident

inspector

met with the licensee to further discuss

actions that the licensee

planned with respect to the failure of Control Rod 06-39 to insert

and the slow

SOH of the other four rods.

The

PH stated

that the

SSPVs for the four slow rods would be rebuilt prior to declaring the

rods operable.

Additionally, he stated that

he considered

that the other

180 rods were operable

because

these

rods

had consistently

passed

surveillance

tests

and

showed

no history of being slow.

The inspector identified that,

even though the licensee

considered

that the control rods were operable,

the

licensee

did not perform

a

POA.

PPH 1.3. 12B, "Operability Evaluation,"

Revision 0, required

a

POA be performed that included documenting the problem,

the basis for considering

the equipment operable,

and the

BCO.

The failure to

perform

a

POA is

an apparent violation of 10 CFR Part 50, Appendix B,

Criterion

V (Violation 397/9415-01).

6.2

Followu

of Previous

Control

Rod Deficiencies

6.2. 1

Control Rod 14-55 Slow

SOM

On March 28,

1994, at

5 p.m.,

when, performing followup inspection of failure

of Control Rod 06-39 to scram,

the

NRC resident

inspector learned

from a

-12-

licensee materials

engineer that, during surveillance testing performed

on

December

4,

1994, Control Rod 14-55

had exhibited increased

SOM time.

On

August 8,

1993, during previous surveillance testing,

Control Rod 14-55

inserted in 0.28 seconds

and

had

a 2X2 array

SOH of 0.31 seconds.

On

December

4,

1993,

Control Rod 14-55

SOH time was 0.72 seconds

and the

2X2 array

SOH time was 0.41 seconds.

Even though Control Rod 14-55 exhibited

an approximately three times slower insertion

speed,

the

2X2 array

SOH was

only 0.030 of a second

from exceeding

the

TS limit, and the control rod had

never previously exhibited increased

insertion time, the licensee

did not

initiate a

PER.

The inspector

found that the licensee

did not initiate a

PER

because

the

2X2 array average

SOM time did not exceed

TS limits.

Also, the

licensee did not establish

an increased

surveillance

frequency to minimize the

chance of exceeding

the

TS limit.

Early in January

1994,

the licensee

wrote Maintenance

Work Request

DL93 to

replace the

SSPVs of Control Rod 14-55.

On January

7,

1994, prior to

replacing the

SSPVs,

the licensee

again performed surveillance testing

on

Control Rod 14-55.

That testing identified that Control Rod 14-55 inserted in

1.20 seconds

and that the

2X2 array

SOM time was 0.58 seconds,

which exceeded

the

TS limit of 0.43 seconds.

The licensee,

therefore,

declared

entered

the appropriate

TS limiting condition for

operation, initiated

PER 294-0014,

and replaced

the SSPVs.

In PER 294-0014 the licensee

stated,

"following removal of the SSPVs,

they

will be retained for disassembly

and inspection to determine possible failure

modes

and

any generic implications for valves remaining in service."

On

March 9,

1994,

the licensee initiated the following PTL actions associated

with this

PER:

(1) by April 1,

1994,

perform

a material

analysis of the

elastomers

from Hydraulic Control Units 02-19

and 14-55;

(2) by April 1,

1994,

perform

a review of scram time history to determine

SSPV rebuild impact to

Refueling Outage

9;

and (3) by April 15,

1994, evaluate

the need to rebuild

the

SSPVs in Refueling Outage

9 vice Refueling Outage

10.

The inspector

learned that the licensee

planned to have

GE perform

a failure analysis of the

6.2.2

Control Rod 02-19 Air Leakage

The inspector also learned

from the materials

engineer that

on January

31,

1994,

a licensee

system engineer

had identified air leakage

coming from the

vent of the

SSPVs for Control Rod 02-19.

The engineer

documented this

deficiency in

PER 294-0067.

In this

PER, the engineer,

based

on discussions

with GE, concluded,

"the cause of the problem

was likely due to air leakage

past the exhaust

diaphragm of SSPV CRD-V-117, but could have resulted

from

leakage

by the pressure

diaphragm of solenoid valve CRD-V-118."

The licensee

also concluded that the air leakage

did not prevent the control rod from

performing its intended

scram safety function.

It was understood,

however,

that the control rod may not meet its

SOH limits.

The system engineer

recommended

rebuilding the

SSPVs at the next available opportunity.

-13-

On February

1,

1994,

the licensee

performed

a prompt operability assessment.

In this assessment

the licensee

concluded that Control Rod 02-19 was operable

and

recommended

that the

SSPVs for Control Rod 02-19

be replaced

in the near

term.

The licensee

stated

in the assessment

that

GE believed the most likely

source of air leakage

was past the seat of the CRD-V-117 exhaust

and

not past the CRD-V-118 pressure

diaphragm.

The licensee further stated that

GE felt, based

on the recent

scram time for Control Rod 02-19, that the

were reaching

normal end-of-life for the diaphragm material

(Buna-N) in the

SSPVs for this control rod.

Even though Control Rod 14-55 had

exhibited slow rod motion in January

1994

and the cause

was determined to be

the Buna-N components

reaching end-of-life, the licensee

concluded that the

diaphragm seat

leakage

would not interfere with normal

scram time or

contribute to increased

rod insertion time.

The licensee did not scram time

test Control Rod 14-55 to verify their assumption.

The licensee initiated the following PTL action associated

with this

PER:

by

March 18,

1994,

perform

a material

analysis of the elastomers

from Hydraulic

Control Units 02-19

and 14-55 to support rebuild in Refueling Outages

9 or 10.

6.2.3

Disassembly

and Inspection of the

SSPVs of Control Rods 02-19

and

14-55

On February 6,

1994, the licensee

received

GE's cost estimate for the

disassembly

and failure analysis of Control Rod 14-55 SSPVs.

Because

the

estimate

was high, the licensee

performed the failure analysis

themselves.

On

February 7,

1994, after having replaced

the CRD-V-117 and CRD-V-118 valves for

Control Rod 02-19, licensee

engineers

disassembled

and inspected

those

The engineers

found that the exhaust

diaphragm of CRD-V-118 had cracked

180 degrees

around its outer circumference,

and the pressure

diaphragm,

while

intact,

was stiff and very brittle.

Even though the findings of this

disassembly significantly differed from the conclusions

the licensee

reached

in

PER 294-0067

and the licensee

committed to perform

a generic

impact

assessment,

the licensee

did not revise the previous operability assessment

or

perform

a new prompt operability determination.

PPM 1.3.12B, "Operability

Evaluation," Revision 0, requires

a prompt operability determination

when

finding physical

evidence of degraded

or nonconforming

components

that have

undergone

physical

change

from a previous acceptable

condition or has failed

testing requirements.

The failure to perform

a prompt operability

determination is

an apparent violation of 10 CFR Part 50, Appendix B,

Criterion

V (Violation 397/9415-01).

On February

17,

1994, the licensee

disassembled

and inspected

the

SSPVs of

Control Rod 14-55, which had

been replaced

on January

7,

1994.

The licensee

found that the CRD-V-118 exhaust

was hardened

and cracked

(having

a

360 degree

crack around its inner circumference),

and the pressure

operability determination

was brittle and stiff.

The licensee

did not initiate a prompt operability

determination to assess

the impact of finding severely

degraded

components.

PPM 1.3. 12B, "Operability Evaluation," Revision 0, requires

a prompt

when finding physical

evidence of degraded

or

nonconforming

components that have

undergone

physical

change

from a previous

acceptable

condition or has failed testing requirements.

The failure to

perform

a prompt operability determination

is

an apparent violation of 10 CFR Part 50, Appendix B, Criterion

V (Violation 397/9415-01).

The inspector learned that the licensee

engineers

considered

that

a new or

revised operability assessment

was not necessary.

The engineers

considered

that the hardened

were confirmation of their*assumptions

in

PERs

294-0014

and 294-0067.

The inspector

considered this logic incorrect,

since the earlier assumptions

did not consider

complete failure of the exhaust

and believed the air leakage to be originating from the exhaust

diaphragms of CRD-V-117.

The inspector

also learned that equipment

qualification personnel

had not been

informed of the failed diaphragms,

even

though the engineers

were fully aware that these

components

were qualified

components.

On one occasion,

one of the engineers

involved with the diaphragm

failures asked

an equipment qualification engineer questions

about

Buna-N

qualification but did not mention that the

had failed.

The

inspector visually examined the failed diaphragms of the

SSPVs for Control Rod 14-55.

The diaphragms

were

so severely

hardened that any torsional or

twisting force resulted

in the diaphragm cracking.

The inspector also noted

the circumferential

cracking of the exhaust

diaphragms.

The inspector

considered that the degraded

and failed Buna-N components

(the diaphragms)

were

a significant condition adverse

to quality.

The inspector also learned that the licensee's

senior nuclear

managers

were

aware of the failures of the

SSPVs for Control Rods 14-55

and 02-19.

The

senior managers,

however,

had not seen

the diaphragms.

Because

they had not

seen

the diaphragms,

the managers

were not personally

aware of the extent of

degradation.

On March 4,

1994, the Senior Management

Review Group met to

discuss

budget

and planning.

In this meeting the Senior Management

Review

Group reallocated

resources

to perform the

SSPV refurbishment

in Refueling

Outage

9 vice Refueling Outage

10.

The

PM decided to begin the refurbishment

on about April 4,

1994.

The refurbishment

was to be performed

one control rod

at

a time, with the intent of minimizing the impact

on power production.

10 CFR Part 50, Appendix B, Criterion XVI, requires that measures

be

established

to assure

that conditions

adverse

to quality,

such

as failures,

nonconformances,

and defective material

and equipment

are promptly identified

and corrected.

In the case of significant conditions

adverse to quality, the

measures

shall

assure

that the cause of the condition is determined

and

corrective action taken to preclude repetition.

The identification of the

significant condition adverse

to quality, the cause of the condition,

and the

corrective action shall

be documented

and reported to the appropriate

levels

of management.

The licensee

did not promptly identify and correct the

condition of failed and degraded

Buna-N components

in the

SSPVs to prevent the

repetitive failures.

This is

an apparent violation of 10 CFR Part 50,

Appendix B, Criterion XVI (Violation 397/9415-02).

6.2.4

Control Rod 54-35

and 58-35 Air Leakage

On May 3,

1989, with the reactor shut

down

and

a half-scram inserted,

the

licensee

found air issuing from the vents of the

SSPVs of Control Rods 54-35

-15-

and 58-35.

The system engineer wrote Plant Problem Material Deficiency

Report 289-0309 to immediately rebuild the

SSPVs of Control Rods 54-35

and

58-35

and rebuild all the control rod

SSPVs in 1990.

When the licensee

disassembled

the

SSPVs,

they found that the valves'una-N

0-Rings were

extremely brittle, resembling bakelite.

The licensee

concluded that

fragments of the hardened

0-Rings prevented

internal

components

from moving,

causing air leakage

and valve malfunction.

The licensee

noted that the Buna-N

diaphragms of these

SSPVs were flexible.

The craftsmen

noted that the

diaphragms of the

SSPVs for Control Rod 54-35 were covered with a powdery

substance

that resembled

copper or brass dust.

The licensee

concluded that

the powdery substance

most likely originated

from the air supply tubing;

however,

a confirmatory analysis

was not perfonhed;

The licensee

concluded that the root cause of the

SSPV failures

was that

previous corrective actions

had not been

implemented.

Specifically, the

25 percent

replacement

schedule of SSPV Buna-N components

had not been

implemented.

The corrective action to prevent recurrence

was to replace

the

Buna-N components

in all

SSPVs in 1990.

The inspector considered this action

imprudent

and untimely.

The licensee

did not evaluate

a larger sample of

SSPVs to determine

the extent of the problem.

The action was untimely because

deferring the replacement

until 1990 resulted

in exceeding

the qualified life

of the Buna-N components.

The licensee

did not implement

a corrective action

to periodically replace

the

SSPV Buna-N components,

which was the root cause

of the event.

In 1990,

when the

Buna-N components of the

SSPVs were replaced,

the work

request did not include instructions that required the craftsmen to document

the condition of the

Buna-N components

upon replacement.

The system enginee}

kept approximately

15 of the diaphragms.

These

were flexible.

The

engineer,'owever,

did not record from which valves the diaphragms

were

removed.

Without traceability, the licensee

could not determine length of

service life.

6.2.5

Conclusions

The licensee failed to adhere to procedures

and perform prompt operability

assessments

on several

occasions

when finding safety-related

components

significantly degraded.

On three occasions

the licensee failed to take prompt

corrective action to prevent recurring failures of safety-related

components.

7

Eg

OF SSPV

BUNA-N COMPONENTS

(92903)

7. 1

E

Histor

of Control

Rod

SSPV Buna-N

Com onents

The inspector reviewed the

SSPV environmental qualification documentation

(WNP-2 gID number

16700,

Revision 6)

and discussed

this material with the

licensee

equipment qualification personnel.

The licensee

described

the

following qualification history of the

SSPV

Buna-N components:

(1)

From January

1976 to August 1978, the licensee,

based

on

GE SIL 128,

-16-

Revision

1, considered

Buna-N component service life to be

3 to 4 years

and

total life (service life plus shelf life) 9 years.

(2)

From August

1978 to December

1978,

the licensee,

based

on

GE SIL 128,

Revision

1, Supplement

1, considered

Buna-N component life to be

7 years.

(3) In December

1985,

the licensee,

based

on the supply system's

engineering

standard,

determined that,

when

Buna-N components

were properly stored,

shelf

life had little effect on natural

aging.

(4) In June

1987,

based

on a

GE letter to Vermont Yankee,

the licensee

again

concluded that shelf life had little affect

on natural

aging

and that service

life was 5.6 years.

(5) In July 1990,

the licensee,

based

on WNP-2 operating experience

(the

Buna-N component failure described

in paragraph

6.2.4),

revised service life

to 5.5 years in their qualification documents.

The inspector

noted that in 1990 the licensee

had used operating experience

as

the method for determining qualified life of the diaphragms for the

10 CFR 50.49

and

NRC Branch Technical

Positions permit qualifying equipment

using operating experience;

however, qualification by operating experience

is

a lesser preferred

method.

The inspector

noted that the licensee

determined

that the qualified life of the Buna-N diaphragms

used in the