An improved method for reactor coolant pump abnormality monitoring using power line signal analysis

Abstract

An improved method to detect the reactor coolant pump (RCP) abnormality is suggested in this work. The monitoring parameters that are acquired from power line signal analysis are motor torque, motor speed and characteristic harmonic frequencies. The combination of Wigner-Ville Distribution (WVD) and feature area matrix comparison method is used for abnormality diagnosis. For validation of the proposed method, the test was performed during cool-down phase and heat-up phase in nuclear power plant (NPP) by cross-comparison with RCP vibration monitoring system (VMS). Using pump internal inspection results, the diagnosis prediction is verified. (c) 2005 Elsevier B.V. All rights reserved.