Generator - Xfmr Neutral Circulating Current

[jghrist]

Our client has a 2000 kW diesel generator connected to the 34.5 kV system through a 2500 kVA, 480 volt grd-wye to 34.5 kV delta, transformer. The generator is wye connected with the neutral connected to the transformer neutral, grounded at the transformer only. There are no phase-to-neutral connected loads on the 480 volt system.

There is a considerable amount of neutral current, even though there is no zero-sequence connection to the 34.5 kV system and no phase-to-neutral load. Current measurements at the transformer were:

X1 - 663A, X2 - 665A, X3 - 630A, X0 - 81A

The third harmonic neutral current is 75% of the fundamental (60% of total rms). There is little other harmonic current. There is very little harmonic current in the phases (1% third, 2.5% fifth, 1.7% seventh).

We are attributing the third harmonic current to circulating current between the generator and the transformer, similar to third harmonic circulating current between two paralleled generators.

Could unbalanced cable impedances account for fundamental neutral currents?

The neutral current is proportional to load current, so that it is about 300A at full load (2400A phase current). Do you see any reason to be concerned with this level of neutral current? Ground fault protection is provided with a CT in the single ground connection, and will not see the circulating neutral current.

 

[rbulsara]

My two cents:

1. I would guess this is because of the gen winding is not 2/3 pitch.

2. I would say, if there is no possibility of false ground fault trip AND you are not planning to run the gens in parallel with the utility for extended period of time, (beyond few seconds for the soft transfers), 5% (of FLA) circulating is not harmful.

3. I would look to provide a reactor in the neutral circuit (between the utility and generator), if this has to be dealt with for extended paralleling as it would cause additional heating.

4. Do you have GF function in the trip units on breakers that connect the generator bus to the utility bus? If so check their settings. They may not be tripping now because they are set higher, but then you have to make sure that in actual load conditions they will not see more circulating current than their setting (1200A most likely). The breaker trip units as you know most likely rely on vector sum of three phase CTs.

5. Just curious, there should not be any circulating current between the generators when not paralleled with the utility.

6. Some fundamental circ current would be due to mismatched impedance of the sources and between phases, and/or voltage differences.

 

[dpc]

This magnitude of neutral current does not strike me as much of a concern. The generator probably has a third harmonic voltage that exceeds the transformer's due to the winding pitch. You do use up a bit of the rating of the generator and transformer (as square root of the sum of the squares).

If there are no line-to-neutral loads, maybe the neutrals don't need to be tied together.

 

[jghrist]

The generators are used for peak shaving so will be operated at full load connected to the utility for 2-3 hours at a time. The only loads on the 480 volt are jacket heaters, fans, etc.

GF protection is by a separate overcurrent relay and the CT has been moved to sense ground current only (connection between the neutral and ground).

Circulating current seems proportional to load current and there is practically no load when not paralleled with the utility.

There are three generators, but they are not paralleled on the 480 volt side, only on the 34.5 kV bus. The neutral current was measured with only one generator operating or with all three operating.

Good point, dpc about the need for tying the neutrals together. I'll have to check on that. Why not leave the generator neutral ungrounded and just ground the tranformer neutral?

 

[dpc]

Yes, I have done it that way in the past, just to avoid the circulating currents. The neutral point can't go too far as long the transformer is always grounded. Of course, if there are 277 V loads, that won't work.

If the generators are matched, their 3rd harmonic voltages will be roughly equal so current will flow from them to the transformers.

 

[waross]

You may have gathered from some of my previous posts that I am not a fan of wye/delta transformer connections. I may have found another reason to dislike them.
When wye delta distribution transformers have the wye point tied to the neutral, almost any imbalance will cause circulating currents in the delta. As the delta tries to balance the voltages by transfering power from two phases to the the third phase the result is neutral currents on the wye side. You may also have heavy circulating currents in the delta that are contributing to heat build up.
In distribution systems this is avoided by floating the wye point. In distribution systems this brings up switching surge issues, but if these are Generator Service Units, you will avoid the switching issues.
I would prefer to see the ground on the generator. This is grounding the system at the source. Grounding the transformer neutral and floating the generator is, in effect, grounding the system at the load. (The transformer is the load on the generator).
Respectfully

 

[davidbeach]

A generator step-up unit transformer (GSU) should always, if at all possible, be a delta on the generator side, and usually a wye on the high side. In this case, if the utility required the high side delta, the transformer should have been a delta-delta; but it's a bit late for that now. You've probably done the best you can do by looking at the transformer neutral-ground bond for ground fault sensing. I'd be leery of a reactor in the neutral as the voltage drop across the reactor may cause more problems than the reduction in current will solve.

Not much point in grounding the transformer wye-point if you don't also connect the generator neutral. Grounding of the wye-point is intended to stabilize the voltages across the windings, but there has to be somewhere for current to go from that point, the connection back to the generator being that place, for the stabilization to happen.

Live with currents on this project, next time connect the generators to a delta winding.

 

[rbulsara]

I should have read the op more carefully regarding the step-up trasnformer.

Yes, I agree with dpc that not interconnecting the neutrals would work for 3 wire load. If the same transformer is used as normal power source when generators are not running, only logical point of neutral grounding will be at the transformer.

 

[waross]

Hi jghrist;
I think that you are in the best position to decide where to ground or if to run without a ground, but the main thing is to break the neutral connection.
respectfully

 

[stevenal]

Perhaps impedance grounding might be used at the transformer and generator. Neutrals are not directly tied together, only through the two impedances. Still a path for ground fault current from either source, but ground current during normal operation will be very small. I'm involved with a project that uses grounding resistors in this manner. Residual current is un-measurable.

 

[jghrist]

Davidbeach raises a good point about stabilizing the voltages across the windings. I think that this may be more important than reducing the circulating neutral current. 300A will not cause that much additional heating and efficiency is not real important in a peak shaving unit.

I think the choice of Delta-wye transformers was dictated by availability and cost. These are rebuilt transformers.

It would be interesting to do the math to see if this much circulating fundamental neutral current can be attributed to unbalanced phase impedances. I did not expect any significant neutral current with no zero-sequence connection to the primary and no Ø-N load. Even the amount of third harmonic in the neutral is surprising considering that there is only 1% third harmonic in the phase current.

 

[ttind]

WAROSS please look at my post on "corner grounded delta" I have a big current unbalance in my motors from this circuit.

 

[matrixabc]

If you Parallel a generator with the grid you must not connect the star point of the alternator winding to earth, other wise you will get circulating current through the parallel earth and neutral paths, we would fit an neutral earth contactor. open in parallel and closed at all other times.

 

[davidbeach]

matrixabc, that would not be an issue here, the delta winding on the transformer blocks any circulating currents between the utility and the generator.

 

[waross]

It may be a primary unbalance problem.
If the current drawn by the three HV phases is unequal, the three low voltage currents will be unequal, hence a neutral current. One of your generator phase windings and one of your transformer phase windings will be passing a normal neutral current.
If your HV currents are not equal, chances are very good that your HV voltages are not equal. That will give a phase shift on the transformers. That will shift the position of the neutral off of the neutral point and also cause neutral currents.
Are you connecting to a long line with no transpositions?
respectfully

 

[waross]

PS. These currents will usually increase with the load.
respectfully

 

[jghrist]

The connection is to a 34.5 kV system with a total of 38 miles of line, close to a 100-34.5 kV transformer. The unbalance in the 34.5 kV primary will not cause secondary neutral currents because the transformer is connected delta-wye with the wye on the 480V generator side.

 

[waross]

An voltage unbalance on the HV delta side may cause phase shifts. The phase shifts and unequal voltages will reflect to the wye winding as unequal voltages to neutral. When you connect this to a generator with equal voltages to the neutral, there will be a voltage difference somewhere which will be made up by neutral currents.
I would float the transformer wye point. I assume that this is a Generator Step-up Unit and is permanently connected to the generator. The permanent connection will avoid the switching surges common on floating wye connections. If you want a system ground put it at the transformer neutral. Solid, low impedance, high impedance, floating, whatever you prefer but leave the transformer neutral unconnected. That will allow the transformer neutral to comply with phase shifts and or unequal phase voltages without neutral currents.
respectfully