The Blogfaddah links to a study that suggests the U.S. could convert entirely to solar energy by laying down newfangled nanotech solar panels over 2% of the continental land mass. As N.Z. Bear notes, even 2% is an awful lot of land:
According to this page, the total land area of the U.S. is 3,537,379 square miles. Take away Alaska and Hawaii to get the continental U.S., and you are left with 2,959,005 square miles. Two percent of that is…
Fifty-nine thousand, one hundred and eighty square miles. That’s 59,180.
For perspective: Over half of the fifty states are smaller in area than 59,180 square miles. The closest in size to that number are Iowa (55,869), Michigan (56,804), and Georgia (57,906).
So: who’s for paving over Georgia?
Since I er, live there, allow me to be the first to say, “Not on my back yard.” More to the point, that kind of study is one of the reasons why engineers refer to solar power as “alternative energy for people who never had to take a physics class.”
Several years back, I did some back-of-the-envelope calculations on how many solar panels it would take to replicate a single nuclear plant. I assumed an orbiting platform using panels equivalent to the ones on the International Space Station, and 100% collection and transmission efficiency–even though the latter is physically impossible. Here’s what I came up with, at my old proto-blog:
[T[he Space Station solar panels are fairly large as such things go, about 8,740 square feet in area. At 64 kilowatts of power for every 8,740 square feet of solar panel, you'd need about 136 million square feet of panels to generate a single megawatt of electricity. Just for comparison's sake, the Farley Nuclear Power Plant near Dothan, Alabama, can produce as much as 1,776 megawatts of power. That kind of output would require approximately 242 billion square feet of orbiting solar cells--about 8,700 square miles, an area larger than the state of New Jersey.
Bear in mind, terrestrial solar plants would receive far less energy than any orbiting platform, you're getting into Monty-Python-very-silly territory by suggesting that the current state of solar technology is anywhere near capable of providing electrical power at any reasonable scale. I'm not saying they can't be improved, or that it isn't worth doing research on better collection methods, but right now, counting on solar power is as crazy as deciding to pave over Georgia... or orbit New Jersey.
UPDATE: In a comment, Lee Valentine of the Space Studies Institute at Princeton corrects my math:
[Y]ou have made an error of a factor of one thousand in your BOTE calculation. The area required to deliver one magawatt (10^6 watts) is 136,000 square feet. That would be a square about 370 feet on a side.
It appears the mistake was in confusing megawatts with gigawatts. A megawatt is one thousand kilowatts, NOT one million Kilowatts. (kilo=10^3, mega=10^6, giga=10^9) One million kilowatts (1 Gigawatt) is enough to run a moderate sized city. Unfortunately, the error means that the estimate of the area required for a 1.778 gigawatt plant is too large by a factor of 1,000. The area is actually about 8.7 square miles, while the sun shines. In actuality, you would need several times that area to compensate for clouds and haze and night. The correction factor for New Jersey is about eight, so about seventy square miles of solar cells would be required. Just the size of Newark, not the whole state.
Dr. Valentine is right, of course, and I both thank him for the correction and humbly apologize for never going back and checking my five-year-old calculations (damn those envelopes!). I should note that orbiting Newark would still be a rather significant effort.