He said 'virtually infinite', which I think is reasonable if you take that to mean 'more than we could ever use'.
Already today there are large scale solar development projects that, if scaled, offer the potential to supply enough electricity for a whole continent. eg:[1].
Tom Murphy calculates that with the energy needs increasing only "2.3% more per year" the whole solid earth would have to be covered with the solar panels in 275 years. 2.3% per year is 10 times more energy needed in 100 years, 1000 times more in 300 years.
If you don't agree please point to the errors in his calculations. Start from here:
"The irradiance at the Earth's surface to approximately 1000 W / m2 for a surface perpendicular to the Sun's rays at sea level on a clear day."
Then calculate that most of the surface is not perpendicular, factor the whole land area on Earth. I come to around 3e16 W falling on the land area of Earth. World energy consumption in 2008 was 1.5e13 W according to Wikipedia and the growth in 2010 was 5% which is much faster need growth -- approx 10 times in 50 years == 100 times in 100 years == 10000 times in 200 years!
So if we cover all the Earth land area with 100% efficient solar panels, we get only 3e16/1.5e13 = 2000 times more energy than we use now. With 5% per years growth, we'd need more than that in around 150 years.
Your first link . . . we seem to be underperforming the logarithmic regression line solidly for the last forty years. Does it really make sense to pretend not to notice that? I guess if you want to sell the claim that solar is going to run out in two hundred years . . .
> If you don't agree please point to the errors in his calculations.
acqq nails it -- there's an upper bound, and in the meantime you have to put all those panels somewhere. Get ready to write off a lot of land to scale up.
(Get ready to write off a lot more if this is how you're planning to power the magic desalination plants that will give us infinite water).
