residential transformer neutral current

If the load is balanced, both in magnitude and power factor, there will be no neutral current. Calculations to support this would be vector addition of the currents.

Heavy residential loads tend to be connected line to line and so do not, cannot cause neutral current.
This would be;
Electric heating.
Clothes drier.
Hot water tank.
Oven elements.
Stove or cook top elements, on the higher heats.
Larger air conditioners.
120 volt loads may be connected to either of the hot lines.
Some of the loads are resistive and the resulting neutral current is the difference between the line currents on line 1 and line 2. The power factor is 100% with resistive loads.
The challenge is the other appliances which have power factors of less than 100%
This will include all 120 volt motors;
Fans, 120 volt air conditioners, 120 volt pumps, refrigerators, freezers, some fluorescent lights, (any light with a ballast is a candidate for a less than unity power factor.)
A rigorous solution would require the current and power factor (and the resulting phase angle of the current) of each device to be determined and then added vectorily.
A field approach may be to measure the current on each line with a clamp-on ammeter and take the difference.
Then measure the neutral current. If the neutral current is greater than the current indicated by the simple calculations, assume that there are some appliances running on 120 volts and less than 100% power factor. It is a fact of life, no correction is required. You will find this situation in virtualy every house in North America which has a refrigerator and/or other motor driven appliances.

Three phase systems:
In those houses and apartments running on 2 phases of a 3 phase system the neutral current may be estimated by assuming that the line currents are at 100% power factor and then adding the line currents vectorilly. This is fairly simple and may be done with the help of pythagoris' theorem.
A meter check of the actual neutral current will probably show a discrepancy, and the rigourous solution will be similar to the rigourous solution for single phase, with the phase angles on line 2 adjusted to include the normal 120 deg. phase displacement between line 1 and line 2.
It has just taken me a lot more time to say the same thing as jghrist said.
respectfully