# 3 phase AC: Can I terminate at main service junction to mitigate line loss? Harmonics?

Suppose I am running very long 3 phase wye distribution from a generator. At the end of my run I will have great line loss and low voltage so i must increase the voltage at the source. This will then cause an over voltage condition early on in the run. Suppose I terminate my run by splicing the end back into the initial breakout point so that I will have a loop. My initial analysis is that it would effectively double the size of the conductor, reducing line loss, but my conductors would be split between two uneven lengths (close to the source, I would have one very short leg, and one very long leg, while at the farthest point i would have an equal length split between two conductors) Assume both linear and non-linear loads.

1. Would the load equalize between the uneven conductors?

2. What kind of effects would this have on the line loss?

3. Would this introduce any harmonics into the system?

4. What other factors should i could consider.

If you understand this, would you mind giving me some insight on how to calculate and prove this myself?
Thank you.

You are describing a ring circuit where each load on the circuit is fed from both sides. Your analysis is correct.

1. Would the load equalize between the uneven conductors?

The currents will vary to balance the voltage drops along the line, as you suspect.

1. What kind of effects would this have on the line loss?

You will improve line loss at every load point.

1. Would this introduce any harmonics into the system?

No. Harmonics are caused by non-linearities in current draw by non-linear components such as rectifiers or saturated (overloaded) transformers, for example.

1. What other factors should i could consider.

Safety. What happens if the loop breaks and you don't detect it. There are plenty of precedents for this, however, such as doubling up cables for a single load. simulate this circuit – Schematic created using CircuitLab

Figure 1. Equivalent circuit.

You should be able to simplify the analysis to the circuit of Figure 1 for each phase. If neutral current is high further analysis of that will be required.

There's a sample calculation at The IET forum which may be of help. The Wikipedia article Ring circuit may also be of interest.

Yes, a ring distribution would potentially reduce resistive losses. The wiring still has to be fully rated for the loads (the loop might break in a storm, after all).

1. The loads won't equally draw from both limbs of a loop, but
2. loss wouldn't be minimum if they did.