Generator Rotor Fault Detection

Very quickly I would imagine (increased current, heat and failure). I wouldn't want to be any where near one if it developed a turn-to-turn fault while under load.

Why do you ask?

If such a fault were to happen (on one pole), the electromagnetic field would be affected, creating an effective imbalance both in the mechanical and electrical components. If the AVR continous to support the increased current drawn to maintain the stator terminal voltage, then there will be increased vibration, pole-shoe temperatures and possible generator bearing failure as a consequence of the increased vibration.

I think the best way is to monitor the AVR current. Vibration analysis would give you an indication but requires that you have initial good values of the whole system to compare with. Really depends on how the AVR is set up to handle reactive power for the generator.

What about monitoring the excitation current, voltage and generator output MW to see if the relationship changes. As long as the resistance of the rotor winding does not change appreciably, the relationship should be fairly constant.

If there are turns shorting in the rotor, I would expect to see the excitation current increase and the excitation voltage drop while the MW out remains about the same or the excitation voltage drop and the excitation current remain about the same while the MW output decreases some.

I am sure I have seen an alarm for this as well as a rotor ground fault alarm on a control system I once worked on but don't really know how it was done.

16MW is a lot of power. Since a turn to turn winding fault will likely be across adjacent turns of the same phase and thus be a shorting loop of a small part of an output loop, it maybe very difficult to detect. Complicating your search will be output load changes beyond your control. One possible idea would be to monitor the DC current applied to the rotor and compare this with the changes in the AC output current of all three phases. Fluctuations in the rotor current that don't correlate with output current load changes might implicate an insulation breakdown. Particularly if a higher rotor current correlates with the voltage of the one phase with a short in it. But this may just be an artifact of the high noise environment.

The thermal effect suggested by others of this short maybe your best bet.

I thought a fault in the rotor would affect all of the phases roughly equally since the rotor provides the excitation field for all of the windings in the stator.

Please correct me if I am wrong.

Thanks!

TT3

You caught me. I misread and thought the original question was about a short in one of the three stator windings. A short in the rotor winding would affect all three output phases. (Maybe somebody should mark my previous answer as off topic. ) A rotor short would make the generator less efficient, but measuring the applied power for delivered power at 16MW would be tricky at best. If the short is an intermittent you might be able to notice an output power dip that corrolates with rotor current spikes.