Resolution of Generic Safety Issues: Issue 3: Set Point Drift in Instrumentation (Rev. 1) ( NUREG-0933, Main Report with Supplements 1–34 )
This issue is identified in Appendix D of NUREG-05724 and is one of the key observations made after the ACRS requested its members and consultants to make comprehensive reviews of all LERs issued during the years 1976, 1977, and 1978. Data collected over this 3-year period showed that 10% of all LERs were related to drift in the set points of instrumentation beyond Technical Specification (TS) limits. This amounted to an average of 258 LERs for each of the 3 years considered.
An unplanned change in the set point of an instrument (set point drift) will alter the actual value of the measured parameter at which a particular action is to occur. Drift in the set point of an instrument could result in delay in the initiation of a safety function.
The solutions proposed in NUREG-05724 were as follows: (1) where set point drift is due to component failures, make the necessary repair, recalibrate, and restore the instruments to service, and (2) increase the margin between the selected set points and TS limits in order to accommodate inherent instrument inaccuracies that result in set point drift.
Because of the qualitative judgment that the safety significance is very small and, because of the difficulty in trying to quantify the risk importance of this issue due to the fact that the drift problem would permeate all of the accident analyses, no attempt was made at quantification.
While set point drift might result in some small degradation of safety function, it would not prevent the initiation of a safety function or result in a major degradation because of the following reasons: (1) safety instruments are redundant and diverse, (2) the measured variables change rapidly, and (3) drift is periodically detected and corrected. Since all safety channels are redundant, the instruments in both channels would have to drift at the same time in order for the delay to occur. But the redundant instruments are identical, sense the same variable, and operate under the same conditions and, therefore, the amount of drift would not be independent. The number of LERs reporting simultaneous drift in redundant channels was not given in NUREG-0572.4
Most safety functions are initiated by diverse signals emitted from different instruments. Drift in any one of these instruments would be expected to be independent of drift in the remaining instruments. Therefore, instrument drift would be expected in many (but not all) cases to result in some delay in the initiation of a safety function.
Most of the process variables used to initiate safety systems have high rates of change during accidents that would result in significant consequences. Therefore, the time delay in reaching an instrument trip point is insensitive to set point.
Drift is detected and corrected during regular surveillance of safety instrumentation. Thus, the degree of drift is limited. Repair and recalibration of instruments that have drifted beyond TS limits are actions that are required of licensees in order to keep their plants in operation. In fact, the strict NRC reporting requirements for these events produced the large volume of LERs that was the basis for consideration of instrumentation set point drift as a generic safety issue. To some degree, this reporting requirement is an incentive to licensees to reduce or allow for instrument set point drift.
In order to address the establishment and maintenance of set points for individual safety-related instrument channels, the Instrument Society of America (ISA) issued Draft F to ISA S67.0449 on May 22, 1979. This standard was adopted by the NRC and was incorporated in the Proposed Revision 2 to Regulatory Guide 1.10550 which was issued for public comment in December 1981. In February 1986, the staff published Revision 2 to Regulatory Guide 1.105 endorsing the guidance of ISA S67.04-1982. No revisions to the SRP11 were required. Thus, this issue was RESOLVED and no new requirements were established.903