ML20212D397
| ML20212D397 | |
| Person / Time | |
|---|---|
| Site: | Pennsylvania State University |
| Issue date: | 10/27/1997 |
| From: | Huges D, Sears C PENNSYLVANIA STATE UNIV., UNIVERSITY PARK, PA |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 9710310087 | |
| Download: ML20212D397 (2) | |
Text
s
.y PENNSTATE College of Engincenny FO L Breareale Nudear Reactor Building Radiation sdence and Engineering Center he Penn9hania State University University Pu ( PA IM02@01 October 27,1997 U.S. Nuclear Regulatory Commission Document Control Desk Washington,DC 20555 Re: Reponable Occurrence: Violation of Tech Spec 3.2.6 License No. R-2, Docket No.50-005
Dear Sir or Madame:
The PSPR staff was unable to contact our Project Manager, Marvin Mendonca , on 24 and 27 October i*>97. Instead voice mail was left requesting a retum call. This letter is to confirm the informatior, that was to be given to Marvin Mendonca regarding a noncompliance with a Limiting Condition for Operation (LCO). The specific section of Technical Specifications (TS) with which we were in noncompliance is section 3.2.4, Table 2b, Minimum PSBR Interlocks. On 24 October 1997, the semi-annual channel test of the Lop Power interlock was being performed as n: quired by TS 4.2.4.b. This panicular interlock is reqmred to prevent pulsing if the log power is greater than 1 kW. This time the surveillance was being performed by CCP-4, Rev.1, our new procedure for vedfying the interlock requirements of TS 3.2.4, Table 2b. The previous method used an intemal ramp test signal as input and the historical trends from the console to verify the setpoint. The ramp test signal is additive with the core fission detector signal. The new method uses a signal generated extemal r' rom the instrumentation which is injected prior to the preamp with the detector removed. The power readout of the console is used to record the log power at which the interlock activates, We believe the new method is more precise and accurate because of the removal of the detector signal noise and the use of a static input as opposed to the ramp in aut which eliminates error due to computer acquisition time delay. It was discovered that the Tardware interlock did not prevent pulsing until 1.04 kW. There is no indication that the setpoint varied fmm the last surveillance. It is assumed that the new method's increased precision and accuracy revealed the noncompliance whereas the previous method did not. The software interlock, which is a duplication of the TS required hardware interlocks did indeed prevent entering pulse mode at 1 kW. However, TS 3.2.4 specifically refers to the reactor safety system and we do not take credit for the software interlocks. Having the hrdware actuate above a level of I kW does not meet the definition of operable,i.e. capable of performing its intended function. Thus, this event is a reportable occurrence.
Since the reactor was already shutdown and the surveillance was being performed with a DO NOT OPERATE tag in place, no further actions were needed to shutdown the reactor.
AP-13 (97-14) was initiated to change the setpoints for the I kW interlock so as to assure it is functioning properly. Since there is no reason presently to initiate a pulse or square wave from f
near 1 kW, it was decided to set the interlock setpoint conservatively. This will provide more ;
assurance regardless of drift or any other effects that the interlock will meet TS. The hardware -
setpoint is to be set to 0.8 kW and the software setpoint to 0.75 kW. 33, m - 1 L
The reactor is to remain shutdown until it is determined that we are and will remain in full compliance with TS. It is expected that we can return to operation sometime 28 October 97.
9710310087 971027 ADOCK 05000005 PDR S PDR lg f{
,; c.:
U.S. Nuclear Regulatory Commission
- Page 2 October 27,1997 -
It shouki be noted that the log power noise bandwidth at 1 kW is greater than 4% and that the L hysteresis is peater than the noise bandwidth. This means that if we were operating at 1 kW
- _ - nominal, the interlock would have activated within seconds just due to the normal noise level and L remained activated until the nominal log power was reduced to less than 0.8 kW.
The basis for the 1 kW interlock is to minimize the fuel temperature prior to a pulse to ensure the L Safety limit of the fuel (1150 *C) is not exceeded. The core temperature rise due to the pulse L
reactivity is additive to the core temperature prior to the pulse. De Safety Analysis Report, .
Chapter IX Section F, Reactivity Accident , demonstrates that as long as the cold clean core reactivity excess is $7.00 or less the Safety Limit will not be exceeded even if the core is operating at 1.15 MW and a pulse occurs with the maximum remaining core reactivity excess. De core temperature at steady state operation at 1.04 kW is within 2 C of ambient. At no time has our -
operation been outside the initial conditions of the Safety Analysis Report, Chapter IX Section F, Reactivity Accident analysis. This event is not of safety significance.
This letter is being faxed today,27 October 1997, to Marvin Mendonca and Thomas Dragou 1. A 14 day written report, as required by TS 6.62, will be submitted.
SL rely, D
wtb 1 W el E. Hughes(gfor C. Frederick Sears)
Manager of Engineering 1 DEH:CFS/ldb4062.97
- pc: T. Dragoun,- NRC, Region I M. Mendonca, NRC, Headquarters E. Klevans -
G. McMurtry R. Erickson J. Mahaffy -
W. Witzig i:
-