EEbeat

I read with interest your editorial about the electric grid. (August Viewpoint)
Your observations on the distribution grid are to the point, as far as I can tell. I would like to broach the subject of the customer hookup to complete the picture. I have reasons to believe major issues and inefficiencies exist there.

1. Industrial hookup
In general, it is clean, as is should be. The power provider places a transformer outside the premises, that reduces the mult-kilovolt input to the 110-480V three phase used inside the premises according to electric codes
The interface between the power provider and user is a switch. When it is turned off, firefighters etc. must be sure, all equipment is without power on the premises
It is usually a 4+1 wire (star) configuration: 3 wires carry the phases, the 4th carries a phase, when one fails. The +1 is a protective ground, not carrying much current. Its function is to prevent shock and trip a ground fault interrupt contactor turning off all power to a device. It is a clean setup.

2. Domestic hookup
It is a bastardized, unclean configuration. The power arrives at your neighborhood as 3+1; 3 phase, plus a steel cable physically carrying them, and additionally serving as a safety, distributed ground wire. It is grounded via a ground rod at every (or so) pole. When a lightning hits nearby, these carry the induced spikes to ground, minimizing equipment damage
The uncleanness and inefficiency comes from the way the homes are hooked up to it. According to code, homes are not served with three phase. There is no contemporary excuse for this legacy limitation. A single phase transformer is hooked up to one phase via a single wire. The other end of the transformer is terminated to the safety ground. The current return to the main transformer in the sub-distribution switching center (sometimes miles away) is then via the safety ground steel wire and the actual high impedance ground. This is again a legacy configuration, that outlived its primitive excuses. (It saves 1 wire out of the correct 4+1 configuration) I presume it is a national code, since in a number of states I visited, all use it. Steel wire is a high resistance, inefficient current carrier. In my researches I did not find the actual chapter and verse in the national code, so I cannot provide it. I hope you have better sources.
Applying entirely rough guesstimates, I think 2 – 10% losses, depending on the local and seasonal circumstances. Verifying the existence of this is straightforward: Connect a voltmeter between the ground rod under a pole transformer and the ground a few meters away. You will see 60Hz voltage drop
While it verifies the existence of the inefficiency, it is insufficient to estimate the magnitude.

Best regards, Levente