What causes 120 Hz second harmonic current on A and B phases of 3-phase Wye system?

Does anyone know what could produce 120 hz current on two legs of a 3 phase system? (120 hz is being produced on a and b phase, not c) We know that it's not a drive misfiring and we know it isn't a transformer saturating.

Details:
The service entrance transformer is a Wye – Grounded Wye winding configuration. The load bus is commercial facility, nothing out of the ordinary. One hint is that we only see elevated levels of the 2nd harmonic current on two of the three phases; with an average of 8% Distortion level as expressed as percentage of the fundamental, with fundamental current measuring an average of 100 amps. This is measured at 480V; at the clients main switchboard.

This was measured over a two day period; and I am beginning to question whether or not a recent client lighting retrofit with high efficiency ballast might be the source; whereas they do have some drives, these are not energized during the predominant period of measurement; but when AC Drive is online and working, approximately 15 minutes out of two days profile, it provided slightly higher 2nd order current as well. I scoped the drives, there is no problem with pulse firing sequences; as that was the first place I looked given historically having seen 2nd order current produced by a misfiring SCR on a six pulse drive. But that is not the case here.

Thank you for your help, any feedback would be much appreciated!!!!
-Ashley

I have seen this several times from electronic transformers and or electronic drives.
In all cases it resulted in damaging filters in other or there own equipment when parallel cabling resulted in building up in harmonics.
As a result the capacitors in the filters lost fast there life time.

P.S. A good surge protection helps to control the problems.

The 2nd harmonic of 120 hz comes from the fundamental frequency which is 60 hz. To eliminate this 120 hz. 2nd harmonic frequency is by passing the frequency across the transformer by using capacitor. This can be done first by knowing the inductive reactance XL of the transformer. Then look for a capacitor value that its capacitive reactance Xc must be equal to XL. In order the circuit must be in resonant to 60 hz.

The 2nd harmonics in the transformer occurs due to the capacitance between turns of the transformer coils

Thank you all so much for your questions and assistance in the Power Quality problem; We have identified the source of the 120 hertz current.....when reviewing current spectrums as compared from one phase to another, we noticed that the 2nd order current was only in elevated levels on two of the three phases in the system. This pointed us in the direction that the source of the current is single phase loads. When we went through system panel schedules, the only harmonic producing load that remotely met the profile of what we were looking for was a high efficiency fluorescent lighting; the client had just recently had a lighting retrofit and replaced HID lighting with fluorescent style ballasts. We asked the client to monitor the second harmonic current on all three phases, and then cycle the lighting panel on and off. When they did so, the two phases that expressed elevated harmonic current at 120 hertz dropped off significantly, while the third phase never moved. When lighting panels were turned back on, both phases harmonic content jumped back up to recorded levels.

These are 277 volt lights; and apparently the client’s electrical installer only used two phases of the system when powering them; no engineering support was used for project install and design. In this case, it might have been for the best; as had all three phases had lighting evenly distributed across them, we might still be wondering as to what the source of this unusual negative sequence component could be. Now that the ballasts have been identified, and approximately 6 amps of 2nd harmonic current are recorded in the system, what would your comments be in next step? Client is soon to add approximately 16 motors and AC Drives to the system; and I am concerned that the negative sequence current poses counterclockwise torque and heating in motor shafts. Would this concern warrant another lighting change out in your opinion? At what levels of 2nd order / negative sequence current would one feel action was needed to protect 480V motor loads in parallel to system lighting circuits?

Thanks,
Ashley