Detecting changes in the condition of process instruments

By custom, the output from an instrument is a signal representing a measure of the process variable, and instrument performance is usually specified in terms of accuracy, linearity and rangeability. The process operator typically uses the specified performance criteria to assess the operational condition of the instrument based on the premise that degradation in performance essentially reflects malfunction on the part of the instrument. Current developments in intelligent measurement systems place emphasis on local sensor validation schemes to incorporate systematic means for detecting and isolating instrument effects from plant effects. This paper describes a method for detecting and isolating changes in a flow-measuring system using wide-band and unconditioned signals taken directly from the flow-sensing element