> That's not true at all, as people use more electricity, aggregate usage gets higher and the grid eventually needs beefier interconnections, beefier transformers...
To a degree, but not in a linear fashion to the amount of power delivered, and certainly not on a unit-power delivery basis.
>The delivery charge would actually be fairer based on peak usage, as it has more relevance to the sizing of the grid.
Very true, and also very common in commercial power billing, usually called a "demand charge" where you pay a specific tariff based of that peak of usage. Unfortunately, it can create really spikey and hard to manage bills, so I accept the argument that it's not appropriate for residential use. Thus, a flat charge on your meter size. On many commercial bills, the demand charge can exceed 75% of the total bill!
Even demand charges often do a bad job of matching utility costs. If I run a church and my peak demand is consistently reached at 11 AM on Sunday morning, guess what, I'm way overpaying relative to my cost to the utility.
The problem with high meter charges (or charging based on panel amperage/etc.) is that it doesn't do a good job matching utility costs either. Let's say I have a 2-story house and split it into a duplex. I add a second panel on the top floor and live on the bottom. Did my second panel double costs for the utility? Absolutely not. If I build a carriage house in my backyard with a 200 amp panel, does it cost the same to the utility as if I put a 200 amp panel on my new house in my far-off exurb? Absolutely not.
That's not to get into the distributional effects of these kinds of changes. Any kind of base-rate increase will absolutely hammer the poor in order to save lots for suburb mcmansions. This may be economically more efficient, but good luck selling it.
If you'd want the transmission charges to match the cost to the utility you should have a base charge consisting of basically the cost to the utility of maintaining your customer relationship after the initial hookup cost, and another part to match the utility O&M cost of the grid distributed over all the customers. Then on top of that a time-varying per-kWh charge for the electricity transferred. This would probably in most cases be pretty cheap, except when the grid (either the utility grid as a whole or the local part that you're connected to) starts to become overloaded; in that case scarcity pricing would apply which would presumably be very high. This would incite customers to reduce usage during scarcity, or give the utility funds to invest in grid expansion.
Similarly for the energy price, that should match the wholesale price. Though see the $10k bills some people on a wholesale price plan got during last year's Texas freeze for why such an idea might not be so popular in practice, theoretically beautiful as it may be.
To a degree, but not in a linear fashion to the amount of power delivered, and certainly not on a unit-power delivery basis.
>The delivery charge would actually be fairer based on peak usage, as it has more relevance to the sizing of the grid.
Very true, and also very common in commercial power billing, usually called a "demand charge" where you pay a specific tariff based of that peak of usage. Unfortunately, it can create really spikey and hard to manage bills, so I accept the argument that it's not appropriate for residential use. Thus, a flat charge on your meter size. On many commercial bills, the demand charge can exceed 75% of the total bill!