Sympathetic Tripping Distribution Feeder

[WiringBoy]

Please see the link below;

Hi Your opinion would be appreciated:

http://www2.selinc.com/techpprs/6061.pdf

I have read this paper few times and this line on page 5 puzzles me:
"If the nondirectional ground overcurrent protection for Relay 1 is lowset,
this protection could operate undesirably."

When the fault happens on the adjoining feeder and the ground fault flows the neutral of un-faulted feeder, why direction is important because neutral current travels in the un-faulted feeder in the same direction with or without fault.

Could you please help what am I missing ?

 

[cranky108]

Not many details here.

It could be a large zero-sequence source on the feeder, or a large motor load, or Cooper 4C reclosers also had this problem. It could be mis-wiring of the breaker controls, or loads of other things.

It is not unheard of for heavly motor loaded circuits to trip for a fault on an adjacent circuit. The solution is a ground directional relay.

 

[DTR2011]

From PSRC info:

Sympathetic tripping. The phenomena where a breaker trips for a fault on a nearby circuit, usually caused by current inrush on a circuit after the faulted feeder breaker opens and the system voltage returns to normal. (PC37.230)

 

[dpc]

If you have a distribution transformer with a grounded wye primary and a delta secondary, it will source ground fault current, even for faults on another feeder from the same substation. This can cause a nuisance trip of the ground fault element on the unfaulted feeder. The solution is to apply directional torque control so that the ground element does not respond to faults on other feeders. The relay is not measuring neutral current, it is computing the imbalance of three outgoing phase currents.

There are other possible zero sequence sources, as cranky108 suggests.

 

[waross]

Hi dpc;
A question, If the distribution transformer has a three legged core will the phantom delta source ground fault current?
Thanks
Bill

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[davidbeach]

As I read it, (although not in great depth) the issue is voltage depression during the fault causes refrigeration compressors to stall. If the fault that causes the depression is an SLG fault, it will only be loads served by that phase on other circuits that see the voltage depression. Then, when the fault is cleared, all those stalled compressors see normal voltage, resulting in drastic increase in current. Since it was only one phase of the load that sees this increase, and the distribution circuit makes no distinction between ground (fault) and neutral (load), the ground relay will respond. Cold load on steroids with no active cold load pickup.

 

[dpc]

It's cheating to actually read the referenced paper. I prefer to bend the question to match answers I might actually know.

Bill,

If the transformer is grounded wye on both sides, this doesn't occur - the zero sequence current passes through the transformer and it doesn't source any. If the secondary is ungrounded wye (I think this would be unusual), I think you could get the effect you mentioned, but I expect the current would be fairly low.

We've seen this problem with some large 480 V delta secondaries.

 

[waross]

Thanks dpc

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[LiteYear]

Hi WiringBoy,

I don't think anyone has attempted to address your question yet. I'm far from a symmetrical components expert but perhaps I can shed some light.

When the fault happens on the adjoining feeder and the ground fault flows the neutral of un-faulted feeder, why direction is important because neutral current travels in the un-faulted feeder in the same direction with or without fault.

I think some of the problem might be that you talk of neutral current and the paper is dealing with zero-sequence current. In essence they are the same thing, but while neutral current is a physical, measurable quantity, zero-sequence current is a man-made, calculated quantity. Depending on how it is calculated, I think it might not match the intuition you would otherwise get from imagining the neutral current at a certain point. In particular, the paper uses a sequence connection diagram under ground fault conditions to determine the zero sequence current. Since the sequence diagram is only a representation of the circuit under particular conditions, the directions and points of reference may not match up perfectly with the neutral current in the original circuit.

So if Relay 1 determines zero-sequence current using the same method as the sequence connection diagram, then perhaps it will register a reverse zero sequence current. As long as the calculation method is consistent, it will all work out in the end. Now I don't have the sequence network analysis expertise to figure out if the method is actually valid, but it's plausible I suppose.

That being said, I've run a bunch of different simulations to see if it's actually possible to get the sum of the phase currents (which is how I would imagine zero-sequence current is actually calculated) to reverse under fault conditions. I've convinced myself that that is not possible, given the scenario described by the paper. So at this stage I'm willing to suggest that the author has got themselves confused with their sequence diagram, and that in actual fact you will not get an actual physical relay to measure any reverse direction zero-sequence current. As you say, neutral current flows in the same direction with or without the fault.

 

[dpc]

Liteyear,

The direction of the zero sequence current depends on the location of the fault. You are assuming the fault is downline on the feeder the relay is supposed to be protecting. If the fault is on another feeder, and there is a ground source on another feeder, the fault current flows in the other direction on the feeder with the ground source.

 

[LiteYear]

dpc, I was just going off the diagram in the paper, which I've copied here.

Is that the situation you're describing? The fault is on one feeder, there is a ground connected load on another feeder, and the relay on that feeder will see some extra zero-sequence current due to the voltage sag. I still can't see how it would be in the reverse direction to the normal zero-sequence current due to the load.

 

[slavag]

It's standard situation with low setting of ground element.
Don't forgot about capacitance current.
Next important point, it's measurement ground current: from three phases or zero-sequence current transformer.

solutions:
1. increase setting
2. directional ground fault.
3. using of zero sequence current transformer.

 

[WiringBoy]

Thanks Liteyear, I appreciate the effort. I still did not get it.Something is missing for sure. Hopefully somebody can give more explanation.

 

[cranky108]

Consitering that ground relays can also trip for unbalanced single phase load, it is worth looking into the normal balance of your single phase load. If the unbalance plus the additional current drawn from stalled motors is enough, the problem may not be solved by a directional relay.

That said, be sure your fix is a fix for the real problem. Otherwise it could look bad. If you have electronic relays on the unfaulted circuit, I suggest you review the recordings.