3 ms·

Momentum, installed base, established interests, changing uses. Foolishly reductionistic model: at the instant that solar becomes defacto cheaper, if everyone

Momentum, installed base, established interests, changing uses. Foolishly reductionistic model: at the instant that solar becomes defacto cheaper, if everyone immediately changed, you'd need at least [total world power consumption / solar panel production rate] years to change over. 2022 world electricity consumption was about 25000 TWH. That envelope-backs to about 2800 GW if the sun is always directly overhead everywhere and there are no transmission losses. In 2022, we installed about 228GW of solar panels. Assuming constant production (HAH!) that says it should take more than a decade to replace the existing demand. Which will of course stand still for us. Apply whatever multiplication factors you like about what percentage of the year a given real solar cell will generate power; about how quickly folks are convinced to change; about the construction of projects which were planned before this price threshold was passed; about the impact of using grid power to do transportation work which was previously done by pumping dead dinosaurs... The confounding factors abound. But that's an ignorant sideline answer to why. :)

cesarb 3y ago
> if the sun is always directly overhead everywhere [...] Apply whatever multiplication factors you like about what percentage of the year a given real solar cell will generate power A simple but useful back-of-the-envelope estimate would be 25%. Half of the day, there's no sun, so you have 50%; the other half, it increases gradually until the middle of the day, then decreases gradually until the sun sets. Simplifying this to a linear increase and decrease, if you plot it into a graph you'd have a pair of isosceles right triangles, which cover half of the area, so you have another 50% over that first 50%, and the end result is 25%. (IIRC, the real result for the single-axis trackers you'd find on grid-scale power plants is something like 30%, showing how good that simple back-of-the-envelope estimate is.)

cesarb 3y ago

> if the sun is always directly overhead everywhere [...] Apply whatever multiplication factors you like about what percentage of the year a given real solar cell will generate power A simple but useful back-of-the-envelope estimate would be 25%. Half of the day, there's no sun, so you have 50%; the other half, it increases gradually until the middle of the day, then decreases gradually until the sun sets. Simplifying this to a linear increase and decrease, if you plot it into a graph you'd have a pair of isosceles right triangles, which cover half of the area, so you have another 50% over that first 50%, and the end result is 25%. (IIRC, the real result for the single-axis trackers you'd find on grid-scale power plants is something like 30%, showing how good that simple back-of-the-envelope estimate is.)