Here is my point, suppose a kWh costs 10 cents retail. This represents an average cost of maybe 1 cent of profit, 1 cent of admin costs, 3 might be transmission

Here is my point, suppose a kWh costs 10 cents retail. This represents an average cost of maybe 1 cent of profit, 1 cent of admin costs, 3 might be transmission costs, and 5 might be generation costs. In a 1-1 net metering scheme, I generate 1 kWh in the distribution grid, it goes to my neighbor, the utility bills my neighbor 10, and credits me 1 kWh. When I draw that 1 kWh, the utility loses 2 cents on average. But they also saved 8 cents by selling my kWh to the neighbor. The summing works like this: 10 from my neighbor to the utility 0 from me to the utility -10 for the utility to provide me that kWh 8 to the utility for avoiding the cost of generating and distributing the 1 kWh that went from my system to the neighbor. The net to the utility is 8 cents. Ouch. But we are talking about the average kWh. It might cost the utility 3 cents to buy an off-peak kWh and 8 cents to buy an on-peak. If that is the case, they are already losing 2 cents for every on-peak kWh they are selling me but making up for it with off-peak (I am likely to use more off-peak anyhow). In this case, netting might look like the following: 10 from my neighbor to the utility 0 from me to the utility -8 for the utility to provide me an off-peak kWh 11 to the utility for avoiding the cost of generating and distributing the 1 kWh that went from my system to the neighbor. The net to the utility in this case is 13 cents. 13 cents is better than 8 cents but it is worse than the 20 cents they would have made if I just consumed without generating. Whether you credit at wholesale or do a .75-1 netmetering or whatever depends on evaluating the stranded costs and whether they are reasonable. Insofar as you look at average kWh prices and not the value of actual production from solar, you inevitably overstate the stranded costs.