Detection of Broken Impeller in Submersible Pump by Estimation of Rotational Frequency from Motor Current Signal

Abstract

With the growth of automation in industry, it has become necessary to monitor the condition of machines and machinery systems to avoid sudden failures which may be catastrophic. There are various condition monitoring techniques available such as wear debris analysis, thermography, vibration analysis, motor current signature analysis (MCSA), etc. for fault detection of machines. Among these, MCSA is a new and emerging technique for fault detection of a submersible pump and other machineries. Apart from proper signal selection, the proper signal processing technique is also important for machinery fault diagnosis. A number of signal processing techniques such as FFT, STFT, and CWT are used for detection of faults by various researchers. In presence of fault, the speed of rotor changes. That will be hidden in the current signal in the form of modulation. Hence, instantaneous frequency (IF) estimation technique is an important technique, which can extract the hidden information. The present work attempts to detect the broken impeller in submersible pump using MCSA. Zero-crossing technique and frequency-domain-based IF estimation technique has been adopted to estimate IF from motor current signal and that has been used to detect the fault in submersible pump. Using these techniques, it is observed that the speed fluctuation increases when there is a defect or level of defect increases in impeller.