>>>Uhm, we *do* have 49 other states, and some don't have as much sunshine :)
>
>Yes, but transmission wires don't care if they have sunshine. ;-) Little old New Zealand shifts power across the Cook Strait so I'm sure the US can string a few wires to where power is needed even more easily than you shift water where it is needed.
>
>>>Don't get me wrong - renewable energy has good supplemental and regional value. You'll never hear me say it can't be part of the solution. But too many advocates don't accept the limits (nor the economic pitfalls, like the huge negative impact on real estate)
>
>In the middle of Arizona? One solar farm in the middle of nowhere theoretically can power most of the country.

From http://www.eia.gov/electricity/annual/html/epa_01_02.html , annual US electrical consumption is about 3.7 million GWh. That's

3.7 x 10^6 GWH, or
3.7 x 10^12 kWh

From http://www.azsolarcenter.org/images/articles/az/solmap.gif , Arizona gets roughly 7,000 Wh per square inch per day, or 7 kWh per square inch per day, or 7 x 365 = 2,555 kWh per square inch per year.

So, with perfect efficiency and no transmission losses US demand could be supplied by

3.7 x 10^12 / 2,555 square inches of solar farm, or 1.45 x 10^9 square inches, which is a square about 38,000 inches each side, or 3,173 feet, or just under 1km each side. So, basically a square kilometer. Less than I would have thought.

Maximum PV efficiency is currently about 30%, commercial quite a bit less. Solar thermal can be better, especially at scale. Maybe a combined cycle might near 50%.

Even if overall efficiency was only 5%, you'd still need only 20 square kilometers.

Now, just to get the cost per kW down ...

UPDATE: whoops, looks like I misread the legend on the solar insolation chart. It should be about 7 kWh per square meter per day, not per square inch. That gybes with a global average of 6 kWh/m^2/day , as stated in https://en.wikipedia.org/wiki/Solar_irradiance#Earth .

Since there are about 1,550 square inches in a square meter, then all the results would need to be scaled up by that factor. So:

Perfect efficiency - about 1,550 square kilometers or a square a little under 40 kilometers on a side. Divide by your efficiency value of choice.

>>>bottom line...nuclear energy is (relatively) the safest and most efficient form of large-scale energy
>
>Spoken like a true Frenchman. ;-) Seriously, you guys have whole deserts you could use for power without anybody noticing except Coyote and Road Runner.