3 ms·
It is a good write-up. I partially agree with some of your points. I have some alternatives that I think might be useful. 1-2. I'm not bothered by net metering
by rickydroll 2y ago
It is a good write-up. I partially agree with some of your points. I have some alternatives that I think might be useful.
1-2. I'm not bothered by net metering. I don't think symmetrical metering is appropriate, and the difference between buying and selling price should be used to fund infrastructure contributions. I suspect it'll just get used as dividends for shareholders.
3. That's where a reserve charge would be appropriate. In case I'm using the wrong term, a reserve charge is a fee assessed to pay for excess capacity on the grid in case you cannot provide it yourself.
4. Variability is an important issue. To me, accept the variability of lots of small-scale producers because any one of them going down has a much smaller impact on the grid. A single hailstorm can take out a utility-scale solar installation. If rooftop solar is distributed over a greater geographic region and is much less vulnerable to localized damage.
However, I think we're missing an opportunity to improve local power resilience by emphasizing and planning for virtual power plants constructed out of rooftop solar and local batteries. In my perfect world, municipalities would have rooftop solar on every city-owned building where it's practical. There would be multiple battery sites distributed on municipal properties. City residents would be encouraged to put in their own rooftop solar and house batteries to contribute to the virtual power plant run by the municipality.
There would still need to be a reserve charge to pay for the grid, distributed among all the municipal residents, and then a usage charge for what they actually consume.
We used to have municipal power companies, and some communities around here still do. While they don't generate power, they do a much better job handling lines and restoring service after weather-based outages. We should look to the past to learn how to build a distributed future for renewable energy power systems.