Approaching infinity

Very informative post. A bit embarrased that I didn't know that what I was actually was looking for are fuel cells. Also, I weren't very specific in my first post but the case I've been thinking about is wind energy. The thing about wind energy is that, at least according to what I've read (Smart Power Generation), you would have to be able to store the excess electricity of 3 windy days in a row, an absolutely enormous amount of energy. Now that is probably a bit of an extreme case as you would probably cut back on other energy production in areas with lots of wind energy during windy days but that is still a lot of electricity to store. Obviously cost/energy stored is important here, and according to the previous mentioned book using batteries to store this amount of electricity would be way too expensive. Do you think fuel cells made of cheap materials could in this case be viable in the abscence of any hydropower that could be used as a balancing factor?

I have been thinking of different ways of storing energy in times of electricity overproduction, in particular electrolysis as it can form relatively good quality energy (chemical energy) that isn't too volatile. Electrolyzing H20 and CO2 into methanol is a way of doing this, but if you want to store electrical energy instead of chemical energy it's not very efficient to just burn the methanol. What would be more efficient is if you could reverse the electrolysis process to turn the chemical energy immideately into electrical energy again, (theoretically) a much more efficient transformation. I have been searching for how high efficiency these energy transformations have but I can't seem to find any numbers of this. Without any numbers on how efficient these reactions might be it's impossible to say if it can be viable. So essentially what I'm actually searching for is an reversible electrolysis reaction that uses relatively cheap materials and is efficient in both directions, but failing that a general picture of how efficient electrolysis is right now and how efficient it could theoretically/reasonably become will more than suffice.