> To provide baseload-ish power, you need multiple installations to cover that different areas receive power at different times in a stochastic fashion and you

> To provide baseload-ish power, you need multiple installations to cover that different areas receive power at different times in a stochastic fashion and you need backup plants just in case your multiple solar farms are shady or the wind stops in multiple places. Absolutely, and the different areas you need to cover have to be very far apart, most likely distributed over multiple countries. This is covered in the excellent Without Hot Air: http://www.withouthotair.com/c26/page_187.shtml http://www.withouthotair.com/c26/page_187.shtml "Between October 2006 and February 2007 there were 17 days when the output from Britain’s 1632 windmills was less than 10% of their capacity. During that period there were five days when output was less than 5% and one day when it was only 2%." So if you build half your capacity in solar and half in wind, you'll find edge cases where it's night and the wind isn't blowing, and you're suddenly generating 1% of capacity overnight. This is aggregated over the UK, which is a very small country, but it still highlights the magnitude of the problem. There are quite a few potential solutions, none of them good; country scale storage (a Tesla can power an average American home for more than a day). Solar thermal, where if you store enough molten salt you can generate at night. Triaging all power usage with a smart grid, so most stuff just turns off. A damn near worldwide grid with HVDC. Keep an entire grid worth of fossil fuel plants in reserve. But in any case, it's not nearly as simple as matching wind MW for MW with coal, something which seems to get missed a lot, especially when discussing Germany. (I feel like a broken record recommending this book all the time, but it really is fantastic for hackers interested in energy).