ML19281B108

Summary of 790221 Telcon Re Victoreen Area Monitor Model 845 Sys Failure,Caused by Faulty Detector Reed Chopping Switch

ML19281B108
Person / Time
Site:	Pennsylvania State University
Issue date:	03/05/1979
From:	Levine S PENNSYLVANIA STATE UNIV., UNIVERSITY PARK, PA
To:	Briggs L NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
References
NUDOCS 7904210071
Download: ML19281B108 (2)

Text

.

1$th T H E P ENNSYLVANI A S T ATE U NIVERSITY UNIVERSITY PARK, PENNSYLVANIA 16802 Collere of Engineering Area Code 814 lircucale Nuclear Rextor 865 6351 5 March 1979 Mr. Larry Briggs Nuclear Regulatory Commission 631 Park Avenue King of Prussia, PA 19406 Re : Victoreen Area Monitor, Model 845, System Failure

Dear Mr. Briggs:

1 The purpose of this letter is to summarize the conversation held between Dr. S. H. Levine of Penn State and Mr. Ralph Architzel of the NRC on February 21, 1979 regarding the referenced subject.

The present radiation monitoring system on the reactor bridge is being replaced with a new Victorcen Area Monitor System, Model 845.

The equipment had been installed on the reactor bridge and was being tested. During the initial test period the Victoreen area monitor system failed. It was observed that the system was indicating a reading of less than 1 mR/hr under conditions of reactor operation that should have produced a reading of 8 to 10 mR/hr. An internal source check of the detector model 847-1 produced no response. Investigation of the failure pointed to a failure in the detector reed chopping switch. A phone call was placed to the Victorean Instrument Company's Service Department which at that time advised us of a production problem of the reed switch located in the ion chamber detector amplifier assembly. We discovered that after operation of an unknown period, the iron involved in the construction of the reed chopper would become permanently magnetic and then lock the reed in a closed state. A change in production material and construction has been made to eliminate this problem. We were also advised that an effort was made to recall and correct all of the detectors; however, since they did not know when the problem started, some units with this problem may still be in the field.

The Victorcen equipment involved in this failure was installed but still in a testing program. It was therefore not connected to any alarms nor was it being relied upon as an area monitor. It is intended, however, that this equipment will in the future be used as a replacement for existing monitoring equipment. The two detector assemblies that we have for this type monitor are being returned to the manufacturer for correction of the above described failure.

There are many components in a new safety system that can fail when the system is first placed under test - such items as resisttrs, capacitors, solid state components, etc. Normally, systems are designed to fail safe or 7904210 67/

s ,

Mr. Larry Briggs Nuclear Regulatory Commission page 2 be redundant so that their failure will not result in any hazard. Neverthe-less, the requirement of reporting of defects and noncompliances as delegated in 10 CFR Part 21 of the NRC Rules and Regulations is not clear.

To be safe when any component fails in any safety system, it should be reported to the NRC. We are not sure the failure reported in this letter is required by 10 CFR 21, but we are notifying you as a result of my telephone conversation with Mr. Architzel.

Sincerely yours,

/

Samuel H. Levine Director SHL:m cc: I. MaMaster W. Witzig F. Remick