Principal Component Analysis based on ANOVA Test for Multishaft Centrifugal Compressor Fault Detection and Diagnosis

In the present paper a Fault Diagnosis system for a multishaft Centrifugal compressor included in an Integrated Gasification and Combined Cycle section of a refinement plant is illustrated. The Principal Component Analysis (PCA) technique, a multivariable data driven analysis method, has been adopted for monitoring the chemical process performances. The contribution of the paper is a new ANOVA-based procedure for the determination of the number of principal components. Both single and multiple faults which may cause errors in the sensor readings and/or in the process actuators have been considered. Trial results proved the goodness and the effectiveness of the proposed diagnoser on detection and isolation of faults assuring a satisfactory promptness as well.