An Expeditious Algorithm for Identification and Classification of Rotor Faults in Salient Pole Synchronous Generators

Abstract

Providing secure electricity depends on the reliable operation of the synchronous generators in power plants. Interturn short-circuit (ITSC) and dynamic eccentricity (DE) faults are among the most prevalent fault types in wound field synchronous generators (WFSGs). Among the various signals used to detect faults, such as current, voltage, and vibration, the stray magnetic field provides higher sensitivity for early-stage ITSC and DE detection. The introduced detection methods rely on comparison of various signals frequency spectrum's from fault and healthy generators. However, obtaining a frequency spectrum of a healthy synchronous generator that has been operating for decades in a power plant is almost impossible. A unique pattern is introduced using fast Fourier transform that can be implemented on a low-end microcontroller and ideal for Internet of Things solutions. The efficacy of the proposed pattern is examined by finite element modeling and validated on a 100 kVA WFSG setup in the laboratory and field test in a power plant with a power rating of 22 MVA.

An Expeditious Algorithm for Identification and Classification of Rotor Faults in Salient Pole Synchronous Generators