ML20148C412

Insp Rept 50-412/88-03 on 880201-03.No Violations Noted.One Item Remained Unresolved at End of Insp.Major Areas Inspected:Total Loss of Control Room Annunciators on 880128

ML20148C412
Person / Time
Site:	Beaver Valley
Issue date:	03/04/1988
From:	Lange D, Pullani S NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
To:
Shared Package
ML20148C367	List: ML20148C363 ML20148C412
References
50-412-88-03, 50-412-88-3, NUDOCS 8803220379
Download: ML20148C412 (6)

See also: IR 05000412/1988003

Text

.

.

U.S. NUCLEAR REGULATORY COMMISSION

REGION I

Report No. 50-412/88-03

Docket No. 50-412

License Nc. NPF-64 Priority - Category C

Licensee: Duquesne Light Company

Post Office Box 4

Shippingport, Pennsylvania 15077

Facility Name: Beaver Valley Unit 2

Inspection At: Shippingport, Pennsylvania

Inspection Conducted: February 1-3, 1988

Inspectors: /d ct ni 22688

S. Pullani,fenior Opfations Engineer date '

Approved by: . -

M J-V'ff

D. LangV, Chief, BWR Section date

Inspection Summary: Inspection on February 1-3, 1988 (Report No. 50-412/88-03)

Areas Inspected: Special inspection to follow up the total loss of control

room annunciators event on January 28, 1988.

Results: No violations were identified. One item remained unresolved at the

end of the inspection.

$$[ $O0 k

G

.

.

.

DETAILS

1. Summary of Event

At approximately 1908 hours0.0221 days <br />0.53 hours <br />0.00315 weeks <br />7.25994e-4 months <br /> on January 28, 1988, with the unit in cold

shutdown (Mode 5), the control room experienced a total loss of all

annunciators. A small fire was discovered in an annunciator logic cabiret

(Bay 5) two floors below the control room and was immediately

extinguished by the two operators who had been dispatched to investigate.

Plant parameters were available throughout the event at the control room

instruments and the Safety Parameter Display System (SPDS). Due to the

sustained loss of annunciators (greater than 5 minutes), the licensee

declared an ALERT in accordance with the Emergency Preparedness Plan and

made all required notifications. The damaged solid state cards were

removed and the annunciation system was restored to all but a few known

annunciator windows. Following the restoration of the system, the ALERT

was terminated at 2245 hours0.026 days <br />0.624 hours <br />0.00371 weeks <br />8.542225e-4 months <br />.

2. . Description of Event

About half an hour prior to the event, the Gaseous Waste System Trouble

annunciator flashed in and out with ne message on the CRT or the printer.

This annunciator is fed from Multiple Input (MI) Card 65, where the fire

apparently started (see Section 3). The next indication of a problem was

all existing annunciators which were in "red" went to "white" with all

horns sounding. The operators were unable to control the annunciators for

about 30 seconds. Then the annunciators returned to normal. All annun-

ciators were acknowledged and horns silenced for about 15 seconds. Then

the annunciators went out again. The above sequence of events occurred

between approximately 1905 and 1908 hours0.0221 days <br />0.53 hours <br />0.00315 weeks <br />7.25994e-4 months <br />.

Two operators were dispatched to the 707 feet elevation of the Control

Building (2 floors below the Control Room) 1 investigate the problem

when, at 1908 hours0.0221 days <br />0.53 hours <br />0.00315 weeks <br />7.25994e-4 months <br />, a fire was discovered in Annunciator Logic Cabinet

IHA-LOG-M (Bay 5). The Control Rocm then assembled the emergency squad

and dispatched to the location of the fire.

At 1915 hours0.0222 days <br />0.532 hours <br />0.00317 weeks <br />7.286575e-4 months <br />, the fire was reported extinguished by the two operators

investigating the problem. Damage was reported in a number of cards in

l the logic cabinet (Bay 5). Notes: (1) No early warning fire detection

j alarms were received and no automatic CO2 discharges occurred in the room.

,

Subsequently, the CO2 system was placed in lock out by the emergency squad

'

to prevent its unnecessary operation. (2) One of the two operators found,

of the three invertors which powers the Annunciator System, the #1 and #3

invertors in tripped condition and the #2 invertor making roise and its

lights flickering in synchronism with the noise oscillation. Therefore,

he tripped the #2 invertor.

I

,

!

1

k

-

_ _ _ _ _ _ _ _

r

..

3

-

.

At 1920 hours0.0222 days <br />0.533 hours <br />0.00317 weeks <br />7.3056e-4 months <br />, an ALERT was declared in accordance with the Emergency

~

Preparedness Plan (EPP), due to sustained loss of annunciators for greater-

than 5 minutes. Notifications were immediately commenced. I&C personnel

were called in to repair.the damaged components. At 1940 hours0.0225 days <br />0.539 hours <br />0.00321 weeks <br />7.3817e-4 months <br />, the NRC

was notified in accordance with 10 CFR 50.72.

At 2245 hours0.026 days <br />0.624 hours <br />0.00371 weeks <br />8.542225e-4 months <br />, the ALERT was terminated. No radiological releases

occurred. The control of the plant was not compromised as all control

. board and vertical board instrumentation and controls remained available '

to monitor plant systems.

3. Apparent Cause of Event

At the time of this reporting, the licensee is still investigating the

root cause of the event. A possible cause is a power supply problem or

degradation of insulation which caused arcing and fire. From the physical

appearance of the fire damage, the problem appears to be initially caused

by a component failure resulting in arcing and fire in MI Card 65. The-

failed component appears to be one of the four "Varistors" labelled VR3 in

Card 65. Varistors are used to protect the cards from surge over-voltage.

The licensee identified six cards around the location of the fire which

were severely damaged (i.e., Cards 52, 53, 54, 64, 65, and 66). These cards

were subsequently replaced. The most severe damage was in,[ard 65,

especially around Varistor VR3 which had completely (vaporMed. This is ---

why the failure of Varistor VR3 is suspected as most probable root cause.

The insulation of coils of nearby Relays RSS and RS6 were found completely

burnt. Approximately 10 cards around the location of the fire experienced

smoke damage which were subsequently cleaned out and tested

satisfactorily. All annunciators were completely returned to normal

operation at approximately 0700 hours0.0081 days <br />0.194 hours <br />0.00116 weeks <br />2.6635e-4 months <br /> on February 1, 1988.

4. Licensee Corrective Actions

The licensee's immediate corrective acticas included:

1. Dispatching emergency squad to combat fire and de-energizing the

annunciator bay.

2. Calling out I&C personnel to repair annunciator circuits.

3. Utilization of control board and vertical board instruments and computer

to monitor plant systems.

4. Declaration of ALERT to assemble additional support.

At the time of this reporting, the licensee has not finalized their course of

long term corrective actions, pending completion of their event evaluat'on

to identify the root cause.

.

-

4

With respect to the licensee corrective actions, the insp3ctor had several

questions. The licensee's response to these questions are discussed

below:

1. In response to the inspector's question on the histo"y of failures of

Varistors, the licensee stated that this event is an isolated case.

Although there are approximately 730 cards in the system, with up to

4 Varistors in each card, this is the first time it failed resulting

in a fire. The licensee also verified with General Electric, the

manufacturer of the Varistor, (Note: Electro Devices Inc., the

manufacturer of the Annunciator System, is now out of business) that

no known similar failure history existed. The licensee also stated

that to the best of their knowledge, they do not have Varistors in

other plant systems although other type or make of voltage surge

protectors have been used and that there was no known failures of a

similar nature. The licensee will further verify the usage of

Varistors in other systems.

2. In response to the inspector's question regarding a surveillance

testing program for the cards, the licensee stated thit the

annunciators are a non-safety related (Nol-Class 1E) system and their

failure rate is small and as such they have no existing surveillance

testing program. They do not plan to have a program unless a high

future failure rate dictates such a program.

3. The inspector noted that there are several fuses rated 1.0 or 1.25

amperes in the dc circuits downstream of the 3 invertor breakers

which in turn are 20 or 30 amperes. The inspector questioned why

these fuses did not blow during the event before their upstream

breakers tripped (see Section 3). The licensee clarifed that these

fuses are slow blow type design, and are in series with the invertor

breakers. Blowing of the fuses will have same effect as tripping of

the breakers and would not have prevented the event.

The NRC vill follow up on: (1) the licensee's investigation of the root

cause of the problem (Section 3), (2) the long term corrective

actions (Section 4) and (3) a complete resolution of the inspector's

question as discussed in Item 1 above. They collectively constitute

an Unresolved Item, pending completion of licensee's actions to

resolve these issues (50-412/88-03-01).

5. Unresolved Items

Unresolved items are matters for which more information is required in

order ascertain whether they are acceptable, violations, or deviations.

An unresolved item is discussed in Section 4. of this report.

I

.

'

5

6. Exit Interview

The inspector met with the licensee representatives, denoted in

Attachment 1, at the conclusion of the inspection of February 3, 1988.

The inspector summarized the scope and findings of the inspection at that

time.

The inspector and the licensee discussed the contents of this inspection

report to ascertain that it did not contain any proprietary information.

The licensee agreed that the inspection report may be placed in the Public

document Room without prior licensee review for proprietary information

(10 CFR 2.790) .

At no time during this inspection was written material provided to the

licensee by the inspector.

,_ C-

.

.

S

ATTACHMENT 1

Persons Contacted

1. Duquesne light Company (DLC)

• P. Bienick, Assistant Superientendant, Engineering

• J. Crockett, Senior Manager, Nuclear Operations

• J. Forney, I&C Engineer

L. Freeland, Nuclear Operations Supervisor, Unit 1

• E. Haley, I&C Foreman

• R. Hecht, Director, I&C

• W. Lacy, Plant Manager

• L. Lipchick, Senior Licensing Supervisor

• J. Maracek, Senior Licensing Supervisor

• A. Mizia, Supervisor, QA Operations

• J. Nazar, Design Engineer, NED

• T. Noonan, Assistant Plant Manager

B. Sepeiak, Licensing Engineer

• J. Sieber, Vice President, Nuclear

• G. Svaranowic, Fire Protection Engineer

• D. Szucs, Senior Engineer, Licensing

• C. Trasada, I&C Engineer

• R. Wittschen, Licensing Engineer

2. Nuclear Regulatory Commission (NRC)

• J. Beall, Senior Resident Inspector

S. Pindale, Resident Inspector

• Denotes those present at the exit meeting.

(

~ _ ~