Storage as the answer for wind and solar?
The wind doesn't blow all the time, the sun doesn't shine at night, and its local intensity can be reduced by clouds and weather. Often the argument is made buy those who push "green energy" that this isn't much of a problem because we can produce extra energy when possible and store it for when these sources produce little or no energy. But how well is that going to work, how much would it really cost. I'll do a few quick back of the envelope calculations, with data from a couple of quick searches. Not a perfect answer, but it should give a general idea of the magnitude of the problem.
For every $700 it pays for a compressed air system, the utility gets 1 kilowatt of electricity, supplied for more than 20 hours, enough to run one coffee maker all day [source: EAC, NSTAR]. Pumped hydroelectric costs more -- $2,250 per kilowatt.
For power that lasts minutes to hours, lithium-ion batteries cost $1,100 per kilowatt (or coffee maker), flywheels cost $1,250 per kilowatt, flow batteries cost $2,500 per kilowatt, and high-temperature batteries like sodium-sulfur cost $3,100 per kilowatt [source: EAC]. And storage in supercapacitors costs even more.
So lets say you need to store 100 GW/hours (5 gigawatts for 20 hours, more than 12 because some nights are longer and because you want to have extra in case you need it, after all your talking about solar providing virtually all the electricity in the country, so presumably some areas only have solar). Storage will probably go down in price lets assume its cost one half as much as the current price.. The compressed air system could then provide 1 kw for those twenty hours for $350. 5 GW would cost 5 million times as much or $1.75 bil just for the storage capacity.
At half the current price the cheapest storage would cost $350 for 1 kw for 20 hours, $350 per 20kwh is $17.5 per kw/hour.
"Actual electricity generation in 2007 was 4,157 Terawatt hours"
Lets try to scale that up to cover the electricity needs of the whole US (which I'm assuming, despite evidence to the contrary, does not grow over time)
4157 terrawatt hours, divided by 365 (2007 was not a leap year) Is 11,389 gigawatt hours. Cut that in half (I'll assume that we get no clouds or other interruptions during the daytime and only have to worry about nights), and you have 5700 gw/hour (rounded off since the reality won't be that precise, and giving more exact calculations would be false precision).
5700 GW/hours at $17.5 per kw/hour would be about a hundred trillion dollars.
But we typically use a bit less electricity at night so lets cut that in half. Now its about 50 trillion dollars.
Lets say technology improves in such a way that the costs goes down more than I thought, so cut the cost by a factor of 5 (meaning the total reduction is to 10% of the initial price), that brings the cost down to $10tril dollars, and that doesn't include maintenance, or spare capacity, or the cost for the solar plants, or the cost for additional distribution. Those would probably add trillions more. Lets say the total cost is $20tril. Assume we can reasonably apply $100bil a year to the effort (that seems high but I'm assuming we are making it a major priority), ok then it only takes 200 years to get it done.
Lets cut it in half again as a generous fudge factor. OK, it will take us a century.
And that doesn't include margin for increasing needs in the future.