So with the help of two buddies, Friend, a Merck vice president and pioneer in digital gene arrays, drew a back-of-the-envelope plan for an energy storage system that extends the life of battery banks. In 2004, they started rigging up a Rube Goldberg contraption that uses solar panels and electrolyzers to generate hydrogen and allows Web-based monitoring of its proton-exchange-membrane fuel cell.

I've been tinkering with something very similar to this for a few years now. This was my original concept as well, that you could use hydrogen as a storage medium for backup electricity. It actually does work. I wasn't using a lab grade or even industrial grade electrolyzer, and I was able to make a good yield of hydrogen from water. Even filtered rain water will do, although the pH value of rain water vs potable water can take the yield down about 20%. Storing hydrogen is not trivial, but it can be done. I did use a very simple fuel cell to convert it back to electricity just to see that it would work. It does, but the yields were not great due to the fact that the fuel cell was not sophisticated. The point really was that I generated a usable amount of electricity from nothing but sunshine and rainwater, which is about as sustainable as you're going to get.

I've talked to several 'experts' about this setup, and everyone I've talked to says that deep cycle gel batteries are a better storage medium than hydrogen. The down side to batteries is that they need to be vented appropriately (you'll have the same problem with a high quality fuel cell as they need induction air) and you'll have to replace them in about 3-5 years, depending on how many cycles they're given. The cost outcome ends up being still in battery's favor as the duty cycle for high end fuel cells is still not that great (5-8 years).

All in, as an investment, it still doesn't work out. You're spending well over $1/KWH, where a >$1 rate is just break even. In TVA-Land, this is ludicrous. Still, if your conscious tells you that this makes sense, then the pricing isn't so outrageous that you couldn't do it. I believe that systems like this will get more efficient in the next 5 years and end up being about 2x what you'd spend for a grid tied system. This is very similar to buying a tankless water heater. They last longer, are more energy efficient, and only cost about 2x what you'd be spending anyways (including duty cycle).

I live in an urban neighborhood. So there are other constraints I'm dealing with (historic zoning, space issues, codes, and complexities of actually maintaining the system). But if this were for a rural cabin I'd do it in a second.