ML20198Q604
| ML20198Q604 | |
| Person / Time | |
|---|---|
| Site: | Breazeale Nuclear Reactor |
| Issue date: | 11/05/1997 |
| From: | Randy Erickson PENNSYLVANIA STATE UNIV., UNIVERSITY PARK, PA |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 9711120278 | |
| Download: ML20198Q604 (3) | |
Text
._ _
i ..
(814) 86L9580 PENNSTATE Rodney A.ErkLson L"C""
he Penn9h ania State Unnenity vice President for Rescarch 304 old Main Dean of the oraduate school Unhersity PmL PA 1680L15N November 5,1997 Document Control Desk United State Nuclear Regulatory Commission Washington,DC 20555 Re: Reportable Occurrence: Violation of Tech Spec 3.2.4, Table 2b, Minimum PSBR Interlocks License No. R-2, Docket No. 50-5
Dear Sir or Madame:
This letter is to inform you of a reportable occurrence at the Pennsylvania State University Breazcale Nuclear Reactor (PSBR) on Friday,24 October 1997. This letter is being submitted in accordance with Sections 6.5.2 and 6.6.2 of the PSBR Technical SpeciGcations (TS). The initial verbal report of this occurrence was made that evening by Dr. C. Frederick Sears, Director, Radiation Science and Engineering Center via telephone voice mail to Marvin Mendonca of the Non-Power Reactors and Decommissioning Project Directorat . A second telephone voice mail to Mr. Mendonca was made on Monday,27 October 1997 and confimled in writing that s'une day with copies faxed to Mr. Mendonca and Thomas Dragoun, Region 1. Please note that our letter of 27 October 1997 referenced TS 3.2.6 in the subject; the correct reference is TS 3.2.4, Table 2b.
PSBR TS 3.2.4, Table 2b, Minimum PSBR Interlocks requires a Log Power interlock which will
" prevent pulsing from levels above 1 Kw".
Descrintion of Event:
On Friday,24 October 1997; CCP-4, (Rev.1), Interlocks Channel Check, was being performed.
This periodic (semi-annual) surveillance procedure is utilized for verifying the various interlocks specified by TS 3.2.4, Table 2b. CCP-4 had just undergone a major revision to improve its level of detail, data recording, and precision. During the testing of the log power interlock which prevents pulsing from a power level above 1 Kw,it was found that the software interlock (Protection, Control & Monitoring System, PCMS) functioned at 1 Kw but the hardware interlock (Reactor Safety System, RSS) did not actuate until 1.04 Kw. The actuation points were reproducible. TS 3.2.4 specifically refers to the RSS; thus the i Kw interloc t did not meet the defm' ition of operable,i.e. capable of performing its intended function.
The reactor was in shutdown with a DO NOT OPERATE tag in place for the performance of several surveillance procedures when this event occurred. The Director ordered the reactor to remain shutdown until the issue was resolved.
A setpoint change request (97-14) was initiated utilizing AP-12, Change,in order to assure the interlocks would function properly regardless of minor instrumentation drifts, settings, or method
- of measuring the actuation point. The changes were implemented and verified over the next several working days.
t On Tuesday,28 October 1997, during a scheduled operator training session, this event and its causes were discussed. On Wednesday,29 October 1997, final veriGcation was completed, a reactor checkout was performed, and further discussion regarding the event was held with the -
hcQdCI Director. The DO NOT OPERATE tag was thea cleared and the reactor was returned to service.
An Equal Opportunity Univenity 9711120278 971105 *
- hlbb.
- 'a kh. ldl PDR ADOCK 05000005
- a - l 8 PDR
]
Review of Event:
Two reasons were fotmd for the RSS 1 Kw interlock actuation occurring above 1 Kw. The first is the lack of detail and precision in some of the maintenance and surveillance procedures. This area is presently being addressed by procedure upgrades. These upgrades include more detained instructions, more information being recorded, and as-found as well as as-left data. The Revision I upgrade of CCP-4 was largely responsible for our finding this particular deficiency.
The second reason involves using derived or calculated variables and setpoints vice using the actual measured parameter. A review was made of past surveillances dating from 1991 when the present console and instrumentation were installed. Several methods for conducting the surveillance had been employed over that period. In many cases little or no specific information was recorded other than that the interlock functioned. In most cases criteda were specified for actuation of the interlock but the test methods would have made it difficult to determine the exact actuation )oint.
In all cases the basic intent of the interlock were met. Where actual numbers were secordec they differed from time to time and represented various intermediate test points which were believed to be directly relatable to the parameter ofinterest. During investigation of this event we found measurable differences at various test points of what was previously believed to be the same signal. These differences are well within the tolerances of the various ponions of the circuit but they explain minor variations between setpoints and actual signal level.
Previously our surveillance procedures had calculated intemal circuit voltages on signal levels bared on an assumption of perfect correlation throughout an instrument chain. We are now switching to using either an input signal (such as pulses per second for the source range interlock) or the final output (such as indicated power level for the 1 Kw interlock). Thus the intermediate actuation points ar a<! justed to produce the desired result based on the specification rather than on a calculated intemal voltage.
Based on the review of CCP-4 for the period since 1991 (the time period for the present console) it appears that the reactor has always been prevented from entering the pulse mode for power levels greater than 1 Kw. This prevention has been assured by the combination of the RSS (hardware) and its backup, the PCMS (software). Only for the past two months is there any indication that the RSS interlock might not have been activated prior to exceeding i Kw.
The review of CCP-4 records indicates the RSS 1 Kw interlock setpoint was adjusted on 13 December 1996 to a value calculated to be approximately 0.9 Kw. On 24 October 1997 the setpoint was found to be as set in December 1996. The indicated power level, which is now used for determination of power related setpoints, was found to be 1.04 Kw at RSS interlock actuation.
If all voltages representing log power throughout the wide range channel were identical, the interlock would have activated at an indicated power of about 0.9 Kw.
For most at the period since December 1996 we know that the log power, which actuates the 1 Kw interlock, was reading higher than actual power (this is conservative). In the July through September 1997 perio<1 we expended considerable effort to improve the alignment and calibration of the wide range channel of which the log power is one component. During that tin.e we achiev-d a better alignment between the log power (low range) and the percent power (high range) portions of the channel. By September we had achieved a smooth, linear alignment over the entire range from startup to power operation.
Tims it is possible that if challenged over the past two months the RSS might not have prevented entering the pulse mode between 1.00 and 1.04 Kw. It is noted that the PCMS would have
. prevented such operation. In addition, our procedures require pulsing to be done from initial power levels below 1 Kw; we typically use 0.10 Kw as the starting power. At no time have pulses been initiated from power levels of 1 Kw or greater.
This event is not of safety significance.
s .-%
Conective and Preventative Actions:
The immediate corrective actions were to maintain the reactor in SHUTDOWN and to adjust the interlock setpoints to assure appropriate actuation. Since the interlock surveillance procedure was in progress we were able to verify that all other interlocks met the TSs.
On a longer term we are continuing to upgrade our maintenance and surveillance procedures to incorporate what we have Icamed and to increase the thoroughness and precision which we believe is appropriate to assure safety as well as compliance with TSs.
We also continue to incorporate our experiences into our training sessions so as to prevent such events and to detect where our past practices may not have produced our desired level of safety assurance. Our Safeguards Committee has been informed of this event and is aware of our ongoing efforts to improve our procedures and their implementation.
As we continue our efforts to improve our procedures there may be other reports. These should be viewed as signs of healthy improvements rather than as a sign of developing problems.
If you have any question regarding this matter, please call Dr. Sears at 814-865-6351.
Sincerely, Dr. Rodney A. Erickson Vice President for Research Dean of the Graduate School RAE:CFS/ldb4064.97 pc: M. Mendonca, NRC Headquarters T. Dragoun, NRC Region 1 G. J McMurtry E. H. Klevans '
W. F. Witzig J. H. Mahaffy -
C. F. Sears Subscribed to the sworn before me on this 4 M_ - day of hem 491997, Notary Public in and for Centre County, Pennsylvania.
'I N NNarial Seal
%nnie L Durns. Notary Public .
3;oge Boro. Centre Cm,'y
- ssion Expires Nov 22 m
_ _ - _ _ _ _ _ - _