Zero Sequence Currents through single phase to ground connected load

[samreye]

I work for a PUD in Washington and we had two feeders out of one substation that tripped at the same time. One was faulted, the other wasn't. I'm now reviewing the causes of sympathetic tripping and my notes (from an SEL class I took a couple years ago) state that zero sequence fault current from a phase to ground fault can flow through single phase to ground connected load. There's no further explanation in the notes as to how this can happen. Can anyone provide an explanation? I understand how the zero sequence current can flow through a three phase bank with grounded Y on the high side and delta on the low side, but struggling with the single phase connections.

 

[stevenal]

http://www.selinc.com/techpprs/6061.pdf

shows a single feeder example on page 18. Guess you could put the load on the unfaulted feeder. Zero sequence current through a load, though, sounds like load current to me. I think your unbalance would need to be pretty severe on the unfaulted feeder prior to the fault for the fault to have an effect. Sure you didn't go circuit to circuit?

 

[dpc]

Are you saying you had induced currents on the non-faulted feeder due to mutual coupling?

 

[musicguy1800]

samreye,
Could you describe in a little more detail the events that occurred? What type of relay tripped each feeder?

 

[musicguy1800]

and also where was the fault location?

 

[samreye]

Thanks for the quick response and link to the SEL paper. Here's more details on the fault. The faulted feeder had some UG cable that went bad, so we believe it was a single line to ground fault. The cable was approximately a half mile from the substation. The feeder relays are Westinghouse Type CO relays, so there's no fault data available. Additionally, all targets dropped (all three phases plus ground), but we have reclosing on the circuit and the targets hadn't been reset in a while, so there's no way of knowing which targets were associated with the bo cable fault. As for the other feeder that tripped at the same time, the ground relay had the only target that dropped, which to me suggests that we had 3Io fault current coming back on the unfaulted feeder. There are no three phase grounded Y - delta distributions transformers on that feeder though, hence the inquery into single phase grounded distribution transformers as a path for zero sequence current.

 

[jghrist]

I assume that the ground relay that tripped was fed by the residual from the phase CTs. You should check to see if the CT neutral is grounded in more than one location. Return current from a fault on one feeder could flow through the other feeder's ground relay if there are two ground points.

 

[stevenal]

Do the circuits share some structures on the overhead portions? Did the (presumably) unfaulted feeder successfully reclose or keep tripping to lockout? UG fault causes overhead conductor movement. Circuit 1 A phase contacts circuit 2 B phase. To each relay it looks like a line to ground fault.

 

[samreye]

Yes, the two circuits share structures for a few spans out of the substation. Four over two construction with a shared neutral. I originally dismissed a circuit to circuit fault due to conductor movement because the presumable unfaulted circuit ground relay tripped. But if a phase to phase fault between circuits looks like a ground fault to the relays, then that's certainly a possibility. The presumably unfaulted circuit did successfully reclose after 10 seconds, so what ever caused the relay to operate was no longer present when the breaker closed. As for the CT grounds, drawings indicate that it's grounded at the CT, which is actually fairly close to the relay in the metal clad switchgear. Thank you for the information and suggestions.

 

[jghrist]

Drawings do not always reflect the true state of CT wiring, but if it is in metal clad switchgear, this makes induced currents in the CT neutral a lot less likely.

 

[Deansharafi]

We had an occasion where a trip on a feeder caused another feeder to trip, and we found that the CTs were grounded at more than one point. Suggestion: Remove all earths from the CTs on the feeder that tripped and make sure that CT sircuits are not grounded (megger or voltmeter check), then re-assert the ground only in one point. Most of the times you will find the reason in the secondary system rather than primary.

 

[jbartos]

Suggestion to samreye (Electrical) Jan 27, 2004 marked ///\\I work for a PUD in Washington and we had two feeders out of one substation that tripped at the same time. One was faulted, the other wasn't. I'm now reviewing the causes of sympathetic tripping and my notes (from an SEL class I took a couple years ago) state that zero sequence fault current from a phase to ground fault can flow through single phase to ground connected load.
///Yes, this is correct. The single phase fault will inject current into ground. Any single phase load of the same faulted phase to ground fault will provide the path to the zero sequence current back to the source of the zero sequence current. Often, it is easier to find the answer in textbooks that cover this topic, e.g.:
William D. Stevenson, Jr., "Elements of Power System Analysis," 3rd Ed., McGraw-Hill Book Co., 1975, Chapter 13 "Unsymmetrical Faults"\\ There's no further explanation in the notes as to how this can happen. Can anyone provide an explanation?
///Try to get a good textbook since the literature that you are reading is presupposing that you have a pertinent background.\\ I understand how the zero sequence current can flow through a three phase bank with grounded Y on the high side and delta on the low side, but struggling with the single phase connections.
///This needs background covering Unsymmetrical Faults.\\\

 

[jghrist]

Theoretically, part of the fault current can flow back through line-ground connected loads on the unfaulted circuit. Practically, the load impedance is so much higher than the source impedance that this current is negligible.

 

[gwgroup]

If the load is drawn between phase and neutral of the transformer, there may be other restrictions than those given by the rated current of the winding. Subclause 8.1 of IEC 60076-1 requires that the neutral terminal shall always be dimensioned for the recognised earth fault current and for continuous load current through the neutral when specified (this is a normal condition for distribution transformers).