05000298/LER-1981-006, Forwards LER 81-006/03L-0.Detailed Event Analysis Submitted

Forwards LER 81-006/03L-0.Detailed Event Analysis Submitted

ML20004C360
Person / Time
Site:	Cooper
Issue date:	05/14/1981
From:	Lessor L NEBRASKA PUBLIC POWER DISTRICT
To:	Seyfrit K NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV)
Shared Package
ML20004C361	List: 05000298/LER-1981-006, Forwards LER 81-006/03L-0.Detailed Event Analysis Submitted 05000298/LER-1981-006-03, /03L-0:on 810414,during Normal Surveillance Testing of Scram Discharge Header Cms,Three of Four Channels Installed Failed.Caused by Component Failure.Marginal Crystals Used in Transmitter Replaced by Manufacturer
References
CN55810280, NUDOCS 8106030253
Download: ML20004C360 (3)

LER-1981-006, Forwards LER 81-006/03L-0.Detailed Event Analysis Submitted

Event date:
Report date:
2981981006R00 - NRC Website

text

o.,

f cooper NUCLEAR STATION MM Nebraska Public Power District

• h*M O R M;"E #$" "

• ~

W s. w CNSS810280

'cA May 14, 1981 G

9/'\\k

/

e Mr. K. V. Seyfrit. Director L i U.S. Nucles Regulatory Commission kk8

.l _

N 7/

Office of Inspection and Enforcement Region IV q

/

/

611 Ryan Plaza Drive

/

/

Suite 1000 4

1o r

Arlington, Texas 76011

Dear Sir:

This report is submitted in ac'cordance with Section 6.7.2.B.2 of the Technical Specifications for Cooper Nuclear Station and discusses a reportable occurrence that was discovered on April 14, 1981. A licensee event report form is also enclosed.

Report No.:

50-298-81-06 Report Date:

May 14, 1981 Occurrence Date: April 14, 1981 Facility:

Cooper Nuclear Station Brownville, Nebraska 68321 Identification of Occurrence:

Conditions leading to operation in a degraded mode permitted by a limiting condition for operation as delineated in IE Bulletin 80-17.

Conditions Prior to Occurrence:

Steady state power operation at approximately 91% reactor power.

Description of Occurrence:

During routine surveillance testing of the Scram Discharge Header Constant Monitoring System three of the four installed monitors failed to function properly. A subsequent failure is also de-scribed in the analysis of this occurrence.

Designation of Apparent Cause of Occurrence:

The Constant Monitoring System is an ultrasonic level measurement

/

device for determination of water level in the Scram Discharge Header. The apparent cause of the occurrence is component failure.

5 The manufacturer has stated a marginal group of 1 MHz crystals had been installed in the transmitter. These same crystals turned up

//

l defective in their commercial line of flow devices that uses sire

(

ilar circuit design.

8106030253 5

., k-

- Mr. K.:V. Seyfrit' May:14,vl981-Page 2.

' Analysis of Occurrence:

The' Constant Monitoring System is an ultrasonic measurement device that detects water in the Scram Discharge Headers. The system' consists of four channels, two per discharge header. 'Any one of the'four transducers will initiate a common annunciator in the Control Room and : start a SDV _ level recorder. 7.his system is backed up by a single float type level switch on-each header..The float-type level switches have respective' annunciators in the Control

~ Room.-

In.the event either. Scram Discharger Header fails to drain properly or fills without being detected by the Scram Discharge '

Instrument Volume, these instruments will-detect the loss of fre.e

. volume to allow suf ficient time for : full scram capability. -. Failure of -these instruments to function could allow the Scram Discharge Header to fill undetected, thus preventing.a full scram.

_ Subsequent to repairs made to the CMS on April 16, 1981, the.

^

reactor was manually scrammed in preparation for a refueling outage on April 21, 1981. At this time, two of four channels and one float level switch failed to detect water level. However, two of-three Control Room annunciators alarmed and the SDV level recorder started. The two channels that did detect the scram properly were on the header furthest from the instrument volume. The float. level suction on this header did not function properly. The float switch

~

that worked properly was on the other header where both channels of the CMS failed. This float switch failure is related to the rate.

of filling of the SDV header. When the SDV header fills rapidly, the float switch does not respond. During subsequent filling oper-ations at slow rates, the float switches operated properly. The apparent cause of the two channels not responding during the scram was a bonding problem in the transducer itself-and a logic problem in the circuit board. This bonding problem was not identified during the initial surveillance testing on April 14, 1981.

7 This event presented no adverse consequences from the standpoint of public health and safety.

Corrective Action

i In response to the initial failure, air testing of the SDV header -

j was initiated each eight hours. The discrepant crystals were changed out in all circuit boards and crystals of another vendor were selected by the designer of the CMS.

I i

b 1

a-

J 9, < ;? '

.. Mr. K. V. Seyfrit May_14, 1981 Pag e 3.

Subsequent to the second faih re following the scram, a bonding

' problem was identified within the transducer. - The original' con-i s truction of the transducers had five bonds from the piezo. material to the-pipe. The transducers were redesigned, eliminating three of the bonding surfaces and; the bonding epoxy was ~ changed to_ one -

having a greater resistance ao high temperature..The bonding agent is now rated' in excess of the system design. temperature.

In addi-

' tion, there are now-four installed spares that can be utilized should another transducer l'ailure o: cur. The logic was modified to raise the setpoint to 2" rather that. Ih".

This provides better reliability and does not change the operator actions as previously submitted to.the NRC. Testing required by IE Bulletin 80-17, Supplement 4, except for: full scram testing, will be completed prior to start-up.

- Sincerely.,

A w%

L. C. Lessor Station Superintendent

_ Cooper Nuclear Station 1LCL:cg Attach.

.e.