ML20070B472
| ML20070B472 | |
| Person / Time | |
|---|---|
| Site: | University of Virginia |
| Issue date: | 06/21/1994 |
| From: | Mulder R VIRGINIA, UNIV. OF, CHARLOTTESVILLE, VA |
| To: | NRC |
| References | |
| NUDOCS 9406300171 | |
| Download: ML20070B472 (3) | |
Text
.
l'** SCH00L OF ,
. ENGINEERING @
' June 21,1994 & APPUED SOENCE NUCLEAR REACTOR FACHJIY l Department of Mechanical I Director Division of Reactor Licensing Aerospace & Nudcar Engineering U.S. Nuclear Regulatory Commission University of virginia Washington D.C. Charlouesvme, VA 22WB-2442 20555 MW2-5440 FAX: 804-9M2-5m j
Subject:
Non reportable event at the University of Virginia Reactor Facility [ Reactor Docket Number 50-62]: Failure to follow a step in an experimental procedure leading to the non-availability of only the automatic reactor shutdown function associated with that experiment. (Note: All reactor safety scrams were available and unaffected, hence reactor safety was never adversely alTected).
Dear Sir:
On Friday June 17,1994, during the completion of the " Pre-Startup Checks for the Gas Cooled Niineral Irradiation Facility (511F)" checklist,it was discovered that the experiment's scram bypass switch position check had not been performed correctly during the previous three days. As a result, automatic reactor shutdown would not have occurred were the carbon-dioxide gas coolant flow to the mineral bed in the experimental facility to have been interrupted. h should be noted that the experimental scram associated with the h11F is intended to keep ?he material being Irradiated in the h11F from self-annealing (an experimental, not safety, requirement) and that this scram is not required for reactor safety. Also, since all license (Technical Specification) required scrams were operational during the entire time, reactor safety was not adversely alTected.
l The N11F scram was hypassed late on hionday, June 13,1994 to permit a scheduled surveillance check of the low poolwater scram circuits. The next morning, the reactor supervisor completed the reactor startup checklist, as well as the checklist associated with the N11F. Step 6 in the A11F checklist states:" Insure that the Experiment Scram is not bypassed when starting the reactor ." This is done by checking visually the position of the key-operated scram switch and signing-off (initialing) the checklist. The bypass switch for the 511F scram is located in the reactor room and not in the reactor control room. Rather than checking the switch position visually, which would have required a short walk, the supervisor took a short-cut and considered a lit annunciator light at the console to indicate that the switch was in the on-line position.
l Unfortunately, the supervisor also must have also misread the lit light position, for the annunciator light corresponding to the 511F scram cannot light up when the switch is in the bypass m6de. Annunciator lights at the console sometimes will light when the reactor is in a shutdown and secured condition because the scrams are set and may trip with electilcal line disturbances. Summer thunderstorms occurring overnight when the facility is closed usual ly cause these noise signals and corresponding annunciator lights 1 9406300171 940621 PDR ADOCK 05000062 S PDR '
l O 1
e *
, p -
(511F Scram, page 2, cont.)
to corne on. The supenisor reasoned that a " trip light on" condition can only occur when the switch not hypassed, and that the annunciator light must be indicative of an i unbypassed switch position and the compressor in the off state. Accordingly, he started the gas cooling system for the experiment, cleared the lights on the scram annunciator panel, initialed the A11F checklist, completed the daily checklist and permitted the reactor to be operated.
The improper checks by the reactor supervisor occurred on Tuesday, June 14 and, similarly, again on Thursday, June 16. On Wednesday, June 15, the checklist was completed by another operator and through miscommunication with a trainee, the step was initialled by the SRO as having been completed, when in fact it had not. The operator performing the checklist on Friday morning did the proper check and discovered the N11F bypass switch in the bypass mode.
This event parallels a similar occurrente at the 5 51W Georgia Tech reactor in February of this year. Their event also involved a failure to follow procedures properly, and resulted in the unavailability of some experimental scrams. In the licensee and NRC review of that incident, it was concluded that little safety significance was involved. The smne conclusion can be reached for the situation being reported here by us.
UVAR operation was put on hold last Friday until the investigation and consultations are completed. As regards the operation of the N11F last week, its lead shield temperature was procedurally checked and recorded every hour. During the time that the 511F scram was not available the shleid temperature did not vary from its normal value with the UVAR at power. This periodic temperature check somewhat off-set the scram unavailability. The operators were intenlewed separately, and as a group, about the process used to complete checklists. The problem appears to have been limited to this one item in the N11F checklist and is not a generic problem, all other checks having been carried out as required.
We are thoroughly embarrassed by this situation. The Georgia Tech incident was made known to us in N1 arch by way of a copy of the NRC Inspection report on that incident, and also by an article in the TRTR Newsletter. Both documents were read and signed-oft on by all U.Va. reactor operators. There is no excuse for this step in the 511F checklist not having been completed correctly.
Given the seriousness with which U.Va.'s reactor management views the proper conduct of reactor operations, our reactor supervisor has been suspended from carrying out Ilcensed duties involving the operation of the reactor. Given his prior regulatory compliance record, consideration will be given to permanently reassigning him to another position on stali not involving operation of the reactor nor management of the reactor stalT. The senior operator who has in the past substituted for the reactor supenisor will be named as acting reactor supenisor. ,
i l *-
(MIF Scram, page 3, cont.)
UVAR management has been striving to improve its operating record. Cicarly, we are dependent on the quality of our operators, their training and the seriousness with which l
they take on their license responsibilities. In addition to the actions specille to this lucident, we are in the process of hiring additional operators. On Friday, when the procedural failure was discovered, we were interviewing an SRO from a research reactor '
which will be decommissioned. We expect that the addition of new people to our staff who have a background in research reactor operations will be very beneficial.
Following further consultations with the Chair of the U.Va. Reactor Safety Committee, who has been appraised of the situation, and with the NRC, the reactor director will release the UVAR to service. This is justified since there are no problems with the facility, its operating procedures, or the training of the remaining staff that would impede a return to routine operations. I will be available, if necessary, at (804) 982-5440.
Sincere)S(
l'\
/
b i
'n' Robert U. Mulder, Director M
U.Va. Reactor Facility &
Assoc Qt/ / County of lPd d. Prof. :' ofN Nuclear Engineering 1mnonwea!ih of Virginia j m.ctry certify that the attached document is a true arxl o cx t copy of a
, presented before
..,d..ff. a ~,.,
o ..
run tNs l!_d da/ of __b1v 19 t:y... .o.-.md'.cd
..o ~. .
.,Nb t I
.--,n (C . .aua.,,
.caumm opret, _
I 19 N -
cc: Mr. Stewart Ebneter, NRC Region 11 Regional Administrator Document Control Desk, NRC, Washington D.C.
U.Va. Reactor Safety Committee Commonwealth of Virginia. Department of Emergency Services American Nuclear insurers