Turn to turn fault detection and classification in stator winding of synchronous generators based on terminal voltage waveform components

In this paper, a novel method is presented to detect and classify turn-turn faults (TTF) in stator winding of the synchronous generators on the basis of resulting harmonics contents in the terminal voltage waveforms. Analytical results by using Decision Tree (DT) show that this algorithm is practicable using only the first harmonic of residual voltage and only two harmonic component values. Simulations in Maxwell software are done using Fuji´s technical documents and data sheets of an actual salient pole synchronous generator (one unit of an Iran´s hydroelectric power plants) and all of related parameters (such as B-H curve, unsymmetrical air gap and pole saliency, slot-teeth effect, and so on) are considered to obtain a comprehensive model, without any simplifier assumption.