Quote from vhehn:
and versions of all three would be feasible if we could figure out a way to store the energy they can produce.
Nature provided the best storage of all: fossil fuels, i.e., stored ancient sunlight. Unfortunately, it took tens of millions of years to accumulate and we're using it up at accelerating rates.
The solar flux incident on the earth's surface can be converted by solar cells and stored in batteries or hydrogen -- but there are two fundamental constraints: mean incident solar insolation at the earth's surface and conversion efficiency.
I've seen ballpark figures on how much solar energy is incident upon the earth, e.g., enough in one day to power human cities for a year (I'm just guessing here), but it's another matter altogether to capture that energy. We already sequester a large % of solar energy indirectly in farming and fishing.
Then there's the conversion efficiency of solar cells, which is ultimately limited by the energy band gaps in semiconductors. I'm guessing again, but it probably can't go much beyond 20-30%, which would be a phenomenal improvement over the 12% of today's typical cells.
As for batteries and hydrogen for storage, battery capacity is again limited by those same chemical properties, and hydrogen is a net energy loser -- i.e., it takes more energy to break up the source molecule (e.g., CH4, H20) than is contained in the hydrogen itself (H2). The only advantage of hydrogen is that it can be used for transportation.
Uranium is fantastically energy dense, but the required isotope is very diffuse in nature and must be mined and enriched, which of course takes energy. Some speak of 'peak uranium' in the same sense as peak oil, meaning that there's some maximum production rate based on world reserves and extraction rates.
Just pointing out that there are some real fundamental limits here. Rubber bands and bell jars can also be used to store energy.