I want to add to this that I routinely see solar plants compared on a cost basis with other forms of energy by using the nameplate capacity. Which is, hmm. Rat

I want to add to this that I routinely see solar plants compared on a cost basis with other forms of energy by using the nameplate capacity. Which is, hmm. Rather than impute motive, since I'm sure motives vary, I'm going to talk about why this doesn't work. Classic heat plants (coal, diesel, nuke, doesn't matter) get around 90% of the nameplate. Specifically they're running 90% of the time, and producing at the full capacity while running. That percentage is called the capacity factor. Because for classic generators the capacity factor is high (hydro can vary a lot based on water available in the reservoir), nameplate capacity, which is what the plant yields under ideal conditions, is usually what we talk about. The problem is that the nameplate capacity of solar is what you get on a perfectly sunny day, with the sun shining directly on the panel. What you want in order to assess cost is the nameplate capacity multiplied by the capacity factor, which is the averaged amount of power you can get out of the plant given real-world conditions. For solar, this can push 30% in an ideal location like Arizona, and be as low as 13% in a not-ideal location like Minnesota. Wind can push 50% capacity when well installed, but it is intermittent in an even less predictable way than solar. If the wind stops in the middle of the night, all wind and solar generation put together is bupkis. We need nuclear. We could do without all of the other carbon-free electrical generation by use of nuclear energy. I don't think we should, mind you, solar in particular has a big advantage in that it's just about the only generating source which comes in small modules, so we can chip away at generation by adding whatever's affordable and build up over time. But next time you hear that solar is cheaper, see if you can check the numbers and determine if the claim is being made on the basis of nameplate capacity. If it is, multiply that cost by four.