I happened to stumble upon this gem of economic wisdom on an amazon.com bulletin board:

Nothing Could be Simpler than accomplishing a full recovery of American
economic power in very short order.

1. Finance a crash program to electrify road transportation.

2. Build solar and wind farms, car and truck batteries, hydrogen fuel
sources and fuel cells, and electric cars and trucks, to turn sun and
wind into electricity, and water into hydrogen, as though they were
planes, tanks, ships and synthetic rubber-- and this was 1942.

3. This will end unemployment, high gas prices, unwanted carbon dioxide, oil imports, and all fear that we are too old to learn new tricks.

The responses, quite rightly point out that at no point does the proponent explain how his "simple" solution will be financed, but that is only half the problem.

Now, let us suppose that we have a limitless amount of money, or even that studies show we can quickly and simply convert from petrochemical transportation to electric. It doesn't matter. What I want to do is simply remove the funding from the equation.

Let us also assume that conservatives and liberals alike decide that this is important enough to ignore the free market. Or else the government finds a way to "encourage" this that leaves it free market enough to appease the majority. Or even that oil mysteriously runs out one day and we have to do it as an emergency measure. In other words, let us also ignore the fact that it really isn't the job of government to tell us how to get from one place to another or how to produce electricity.

 So, if we ignore the fact that it is not the role of government and ignore the costs, the question remains, how exactly will this work?

Well, wind is pretty unreliable and not really feasible for most of the nation, so let us look strictly at solar energy. The earth's surface, over  a twenty four hour period, sees an average of 164 Watts per square meter.The current maximum efficiency of solar cells is 28%. Even if we assume technology will improve, we also have to allow for additional losses when dealing with large scale production. If we have many square kilometers of cells, cleaning, breakdowns, and simple transmission losses, not to mention the losses of putting energy into and out of storage will cause a decline in efficiency. So, I think it is somewhat optimistic to even estimate 20% efficiency. However, for the sake of argument, let us allow 20% efficiency and say we can produce about 30 W per square meter.

The US in 2005 used 29000 TWh, or 29,000,000,000,000,000 Wh. There are 8766 hours per year. That means that each hour we need to produce 3.3 TW. Since a square meter can produce (given our assumptions) 30 W, that would mean we need 110,000,000,000 square meters. As 1 square kilometer is 1,000,000 square meters, that would translate into 110,000 square kilometers. In other words, to maintain our power consumption in even 2005 terms would require an optimistically 20% efficient array of solar cells 332 km on a side (208 miles).

That doesn't even include the storage batteries or the wires, or any space between the cells to allow people in to clean or service the cells. And cleaning will be big business, as to maintain even that optimistic 20% efficiency, we need spotless cells and need to replace broken or defective cells immediately. To be realistic, we probably need to at least double that to allow for transmission lines, space for people to travel between them, storage batteries, to allow for redundant backup cells and extra production to both support increasing consumption and to fill in for broken or dirty cells.

So, if we somehow can convert everything to electric, and if hydrogen driven cars do not require any more energy input that oil cars, either for production of hydrogen or when it is used for transportation, and assuming our optimistic assumptions about cell availability and efficiency hold true, then we simply need a grid of solar cells about 160,000 square miles, a little larger than the state of California.

See how simple it is?

Sadly, whenever someone proposes "transitioning" to "green energy" this is what they mean. Since nuclear power is largely dismissed by populist environmentalists and politicians (there are some environmentalists who recognize how clean it is, but they seem still a minority), we are stuck with plans that use solar energy in various forms. However, solar energy is quite diffuse. Even at 100% efficiency, it would require covering something like Washington state in collector grids just to maintain our consumption of 3 years ago.

Unless we can tap stored energy, in the form of petrochemicals, or other chemical or nuclear sources, we will need to dedicate a lot of our land area to solar collectors, or else drastically curtail our energy consumption, and accept the loss of wealth that entails.

POSTSCRIPT

Whenever I use precise numbers, I figure that there is every likelihood I made an error somewhere. So, please, if someone spots an error, let me know in the comments. However, even if I am off by a magnitude of 10, it still is a plan that requires just an absurd amount of land to produce even a modest amount of energy. It is why fully solar homes are rarely air conditioned, and rarely have many electrical appliances. Solar energy is "clean" (though panels often use a lot of very toxic and often scarce resources), but it is not a highly concentrated form of energy. That is why most homes use it for trivial functions like heating water or supplementing the power grid, not to power everything in the house, as well as the electric car. And there is just no way it is ever going to power the factory next door.

ADDENDUM

For those environmentalists who decry the ecological damage of oil or coal, I have to ask, how is covering California in solar panels going to produce less environmental disruption than our present methods of producing energy?

UPDATE 08/03/2008

I noticed something that may be seen as a mistake but was intentional. When I doubled the size, what I actually did was double the length of each side of the gird, which quadrupled the size. As I am allowing for wiring, access, etc. it seemed a better idea to extend each side rather than just double the grid. If you want to simply double the grid size, we have an area half the size of California, still unrealistically huge.

And, I must remind anyone who quibbles with almost doubling each side (208 mi to 400 mi), that my assumptions are quite unrealistic from real life examples. In experimental large scale solar generator prototypes, the area allowed for maintenance, as well as redundant backup systems is far more than 4 times the minimum. For a reliable real world system, we probably would end up covering everything west of the Rockies, if not more.