ML20079D591
| ML20079D591 | |
| Person / Time | |
|---|---|
| Site: | University of California - Irvine |
| Issue date: | 07/09/1991 |
| From: | Geoffrey Miller CALIFORNIA, UNIV. OF, IRVINE, CA |
| To: | Kirsch D NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
| References | |
| NUDOCS 9107180171 | |
| Download: ML20079D591 (2) | |
Text
6 UNIVERSITY OF CAllFORNIA, IRVINE
(
- Q n ratti 3 . nu s . inusi . Ltw Asu t is . nniawoi . 4s na uo . s4s s u4st i c o (gg},-
>4N14 aann4n4 . s4N14 c ut?
v _y DrlwRTMLNT or cHLMis1RY IRvlNI, cALHokNIA W?!?
Nt'CLLAR f.iActOR f AclLITY SLTE RvlSOR DR O E MILLER TLL' i 714: 84tA49 oR 714 8460A2 U.S. Nuclear Regulatory Commission, July 9th 1991 Region V, 1450 Maria L:me, Suite 210 Walnut Creek, CA 94596 Attention: D.F. Kirsch, Chief, Reactor Safety Branch 8d-3N Qodd: " SM - License R-116 Licensee Renon of Abnonnal Occurrence - UC IrvinentekEMcInr
&norted in Accordanca witb " actor TechnicalSEcci6catinnt Secjipn 6 7 c.3 G':ntlemen:
On July 3rd 1991, a Senior Reactor Operator was starting oper.ition to calibrate the SlilM control rod on our TRIGA reactor when she noticed that an unusually short period was generated when only a modest period should have resulted from the amount of SillM rod inserted. On an attempt to repeat the insenion cycle, she noticed that the REG rod, which should remain stationary et an intermediate position during the SillM movement was slowly rising, this creating the unusual reactivity insertion. She scrammed the reactor and brought the matter to the attention of the Reactor Supervisor. Wifa all other control rods inserted, and thus the teactor shutdown by over
$6.00, the Supervisor operated the REG rod and confirmed that even when the UP button was released the rod continued its upward movement in a leproducible manner. This constituted a
" uncontrolled or unanticipated change in reactivity" and so is an Abnomial Ocntttence as dcEntd in Section 1.13.e. of the Technical Speci0 cations for the facility.
At no time did the period exceed the non-required PERIOD SCRAM setting of 3 seconds, nor did the power level exceed any scram setting. In fact power was less than 1 kilowatt during operation, since all control rod calibrations are necessarily done at power levels below this value.
Since for a TRIGA reactor it is well established that pidsed reactivity insertions of well over the i
full value of the REG rod ($2.92) are permissible and safe, TilERE WAS NO SAFETY lIAZARD h'OR ANY DECREASE IN SAFETY MARGINS AS A RESULT OF TI!!S INCIDENT. The failure is purely one of annoyance to standard operations since control rods are supposed to stay vhere they are put! All functions of the rod drive were nonnal (i.e: response to UP or DOWN button press, SCRAM, and " auto" drive down features) except for the upward drift when supposed to remain stationary, l
Suspicion as to cause immediately fell on the drive UP or DOWN switches because of a warning memo received from the reactor manufacturer (2-15-89).110 wever, extensive tests and thoreugh
- exammations of switching circuits failed to confirm this was the cause. Attention was next given to the control rod circuitry at the reactor bridge location. On previous occasions, failure had been
! noted of a 220 ohm 2 watt resistor in the balancing circuit for the two motor windings. This has l been replaced at least twice before i, the lifetime of this facility, each time as a result of inspections l conducted while the reactor was not in operation, so not in those cases, resulting in an Abnormal Occurrence situation.
9107180171 9107o9 {4'y $
gun noccuosoog6 i o
s l
The resistor had a measured resistance of 450 ohms instead of 220 ohms showing aging as a result of overheating. Replacing the resistor with a new one did not cure the problem, the REG nxi drive still responded in the inappropriate way, with continual upward drift.. Careful funh:r inspection and diagnosis revealed that there had developed a leakage current pathway on the circuit board allowing current to flow through the resistor, bypassing the series variable resistor provided i for balance adjustment, and bypassing the drive DOWN motor winding, thus giving preference to the UP direction. In nonnal DOWN operations, this whole section is designed to be shoned and so does not affect nonnal drive down functions. As this pathway did nat take ah the current, nor was it dicgnosable by h,v vo'tage rc.;ist....cc naa.surements, it w as noi readily "rparent as the cause of the problem.
On July 8th 1991, the resistor was replaced with a 200 ohm 5 watt resistor (TRW PW5 style) ,
mounted so as to be well cooled by air, and bypas3:ng the suspect (browned) portion of the circuit mounting board. Full normal nx! drive function was obtained, and the variable potentiometer could again be used to obtain proper balance of up and down rod movement to allow for the weight of the fuel follower on the nxi.
It is postulated that the ciruit board carbonization occurred as a combined result of resistor heating in normal use and soldering and unsoldering activities carried out during this and fonner resistor replacement effons. The new arrangement and higher wattage resistor will provide better cooling and longer time to failure.
This incident was reponed by telephone to the NRC Emergency number at approximately 5:20 pm on 7/3/91 in accordance with Technical Specifications 6.7.b.3.
Sincerely yours, 0.Q I
h ,
George E. Miller Reactor Supervisor ec: Reactor Operations Committee members Senior Operatur, Dr Patricia Rogers Document Control Desk, U.S. NRC, Washington, D.C. 20555 General Atomics, TRIGA Division, P.O. Box 85608 San Diego, CA 92138 5608