Solar hydrogen

One of the biggest problems in moving to a renewable energy portfolio is the transportability question. In fixed plant, you can afford to have acres of solar capture material, for instance, but not in a car or plane (yes there are exceptions!) In addition, we need to store the energy for nighttime or cloudy day use. So batteries, ultracapacitors, underground pressurized caverns, water columns, etc. are all being explored for storage. One method getting some attention is cracking water to make hydrogen and oxygen, the hydrogen being a useful fuel, which burned makes pure (ish) water. It has a great mass energy density around 120MJ/kg and and marginal volumetric density of 8.5MJ/l in liquid form (gasoline by comparison is 45MJ/kg and 31MJ/l.) So we would need 4x the gas tank size, and at high pressure, to move us the same distance. Of course, engineering the car for better mileage could make up a lot of that difference. And for fixed installations, this doesn't necessarily present a problem.

So I was intrigued when this landed on my inbox. What IS good is they have apparently made cracking hydrogen a much less energy intensive process. This is important. But they have committed a foul in claiming the technology is inherently 'solar'. This is common now as solar energy is such a hot field that any way to connect ones work to that field is a way to raise its profile. However, any energy source will work fine as long as it produces electrons. Perhaps it is enabling for a home fuel cell driven by solar power, but it really is 'electron agnostic'. As opposed to this finding. Here, unlike the claims in the MIT finding, the technology is inherently solar. And it yields hydrogen. It may not be strictly photonic technology, but it is still quite interesting.

The hydrogen production question is being approached electrochemically as above, as well as biologically. But I wonder if this is fool's gold. The volumetric density is still low, much lower than hydrocarbons. Using this for a plane fuel for instance might be prohibitive. If we are doing something with water and CO2, maybe we should be making long-chain hydrocarbons [sub req]. They can be pumped straight into our existing energy infrastructure, refined like any oil, and transported without a massive investment in new handling and tranport mechanisms. Of course, if they are solar driven (e.g. algae), they will also have the scale problem

Solar scale

I wanted to spend a few minutes talking about the scale of the problem. It is important to consider this when making arguments about the technologies involved. Lets start with an assertion that is useful: it would take 10,000 sunny sq miles to replace current electrical demand with photovoltaic generated by silicon cells at 20% or so efficiency. This number is reasonable though I'm sure there are arguments. Current electrical usage in the US is about 20% of total energy usage (again, round number) so for all energy at 20% efficiency, we are talking about 50,000 sq miles of silicon (32 M acres). Another point to make is that 50k sq-mi is today's number, if we are talking about a renewable portfolio that really addresses the problem in say 30 years, we have to factor in the growth rate of 1.9%pa (hopefully reduced by conservation).

Now one myth going around is that we could do that with Walmart stores, maybe throw in some other big box stores. Walmart according to Wikipedia has about 500M sq ft of stores in the US. This is 18 sq miles. Off by three orders of magnitude. In other words we would need to have 1000 Walmart-sized chains completely converted, oh by the way, all in sunny climates like AZ. How about roads? Roughly 4M miles or roads by an average width of 50ft would give us 20,000 sq miles, getting close. But a road is a LOT easier to make in bulk than a solar cell. Even with oil and asphalt prices going up.

My point isn't that 50,000 square miles is a lot (it is), but that it is a large area to cover with manufactured materials. In the end, all forms of solar energy require capturing solar flux. Most of the really scalable technologies seem to have an efficiency problem which would double or more the amount required, and since half the cost of the silicon systems is in balance of system costs, the BOS cost would essentially double as well. Not to mention the power transmission infrastructure and power storage infrastructure needed to make it really viable. So efficiency is key, you can't just 'make it up in volume'. The flip side it that IF you can make it economically attractive in comparison to existing tech, with carbon tax/C&T and/or by reducing the cost, then you have a HUGE market.

BTW, this applies to all solar driven technologies. For instance corn ethanol is fundamentally solar energy, but at much less efficiency. Biofuels from algae are also solar derived, and suffer the same scale problem. Although we are talking about much simpler scalable technologies since they are 'self assembling' of a sort.

So maybe the approach is different. We really do need to be looking for a solar roofing material, or road material, or parking lot cover. And we need to conserve. Alot. That is where solid state lighting, OLED displays and other optoelectronic technologies can help.