trat1208
Electrical
- Jan 22, 2009
- 15
Are there any good relay guys out there that can give me some comments on the following.
At one of our facilities, we have a large cross compound unit connected generator, HP and LP approx 300MW / 20kV each. It is high impedance grounded with the neutral transformer with secondary resistor connected to the neutral of the HP. The HP was rewound recently and since then we have gotten intermittent alarms from our SEL300G relay due to operation of the 64G2 element (set to alarm, not trip). The 64G2 is set up to do 3rd Harmonic Diff between the sum of phase 3rd Harm volts and neutral transformer 3rd harmonic volts. It is my understanding that such an issue should probably be expected due to the rewind because the 3rd harmonic generated is dependent on the physical properties of the generator windings.
The engineer at the site followed the SEL procedure for collecting no load and full load 3rd harmonic data in order to reset the relay. He found the data to be inadequate for establishing new settings. Basically what he found is that the phase 3rd Harmonic increases fairly steadily with load, but the Neutral 3rd harmonic does not increase proportionally.
Question 1.
If there was no error collecting the 3rd Harmonic data, would I be correct in concluding that this data suggests that we have a ground fault on the stator close to the neutral? My understanding is that phase and neutral 3rd Harmonic volts should rise in proportion to one another over the MW range of the generator. If neutral 3rd Harm volts stay flat while phase 3rd Harm volts go up, there must be a ground near the neutral that is suppressing this neutral voltage.
Question 2.
Assuming that there is this ground at the neutral, there of course is no hindrance to operation unless there develops a second ground somewhere in the generator, generator bus, or main and aux transformer delta windings. The caveat is that such a second ground would likely lead to catastrophic damage because the high impedance grounding is short circuited.
Are there any non-invasive methods of confirming that this ground at the neutral exists while the generator is online? I'm guessing no.
Question 3.
All the literature that I have found discusses this 3rd Harm voltage diff scheme in a tandem generator application, not cross-compound. For my cross-compound application, does this scheme detect ground faults in just the HP (which is connected to the neutral transformer) or will it also sense ground faults in the LP (not connected to neutral transformer).
In the Mozina paper "15 Years of Experience with 100% Stator Ground Fault Protection . . .", he says that the 3rd Harmonic undervoltage scheme will not detect a fault in the lower portion of the non-neutral transformer connected generator, and it would seem logical that this would similarly hold for the 3rd harmonic diff scheme. However, I was doing the circuit analysis and I seem to come to the conclusion that it will detect faults at the neutral and phase portions of each generator. It's very possible that I am missing something though. I believe that a traditional fundamental frequency neutral overvoltage scheme will detect ground faults in the top 90% of each generators' windings and by similar logic I think that that would extend to this 3rd harm diff scheme as well - but again I might be wrong and I could be missing something.
Can anyone confirm the limitations on the protection offered by this 3rd harm voltage diff scheme in cross-compound application?
Thanks
At one of our facilities, we have a large cross compound unit connected generator, HP and LP approx 300MW / 20kV each. It is high impedance grounded with the neutral transformer with secondary resistor connected to the neutral of the HP. The HP was rewound recently and since then we have gotten intermittent alarms from our SEL300G relay due to operation of the 64G2 element (set to alarm, not trip). The 64G2 is set up to do 3rd Harmonic Diff between the sum of phase 3rd Harm volts and neutral transformer 3rd harmonic volts. It is my understanding that such an issue should probably be expected due to the rewind because the 3rd harmonic generated is dependent on the physical properties of the generator windings.
The engineer at the site followed the SEL procedure for collecting no load and full load 3rd harmonic data in order to reset the relay. He found the data to be inadequate for establishing new settings. Basically what he found is that the phase 3rd Harmonic increases fairly steadily with load, but the Neutral 3rd harmonic does not increase proportionally.
Question 1.
If there was no error collecting the 3rd Harmonic data, would I be correct in concluding that this data suggests that we have a ground fault on the stator close to the neutral? My understanding is that phase and neutral 3rd Harmonic volts should rise in proportion to one another over the MW range of the generator. If neutral 3rd Harm volts stay flat while phase 3rd Harm volts go up, there must be a ground near the neutral that is suppressing this neutral voltage.
Question 2.
Assuming that there is this ground at the neutral, there of course is no hindrance to operation unless there develops a second ground somewhere in the generator, generator bus, or main and aux transformer delta windings. The caveat is that such a second ground would likely lead to catastrophic damage because the high impedance grounding is short circuited.
Are there any non-invasive methods of confirming that this ground at the neutral exists while the generator is online? I'm guessing no.
Question 3.
All the literature that I have found discusses this 3rd Harm voltage diff scheme in a tandem generator application, not cross-compound. For my cross-compound application, does this scheme detect ground faults in just the HP (which is connected to the neutral transformer) or will it also sense ground faults in the LP (not connected to neutral transformer).
In the Mozina paper "15 Years of Experience with 100% Stator Ground Fault Protection . . .", he says that the 3rd Harmonic undervoltage scheme will not detect a fault in the lower portion of the non-neutral transformer connected generator, and it would seem logical that this would similarly hold for the 3rd harmonic diff scheme. However, I was doing the circuit analysis and I seem to come to the conclusion that it will detect faults at the neutral and phase portions of each generator. It's very possible that I am missing something though. I believe that a traditional fundamental frequency neutral overvoltage scheme will detect ground faults in the top 90% of each generators' windings and by similar logic I think that that would extend to this 3rd harm diff scheme as well - but again I might be wrong and I could be missing something.
Can anyone confirm the limitations on the protection offered by this 3rd harm voltage diff scheme in cross-compound application?
Thanks