Condition monitoring and fault detection of a compressor using signal processing techniques

Abstract

Several techniques developed for performance monitoring and fault detection of machinery in the petrochemical industry is presented. The main objective is to perform condition monitoring and fault detection of a compressor based on the collected vibration and other on-site data from a Bently Navada data acquisition system. The analysis techniques developed, configuration and functions of the computer program are also briefly introduced. The techniques can be divided into four parts: (1) time-domain analysis, (2) frequency-domain analysis, (3) orbit analysis, and (4) trend analysis. The program developed has been tested using on-site data from a compressor at Suncor Inc.