A paper has just appeared in the science magazine Nature, which reports that appreciable yields of isobutanol (IB) and other higher chain alcohols (1-butanol, 2-methyl-1-butanol, 3-methyl-1-butanol and 2-phenylethanol), can be produced by fermenting glucose with genetically modified E.coli. Such higher-chain alcohols have a greater energy density than ethanol and are much less prone to absorb water from the atmosphere, being consequently less corrosive toward engine parts. Liquid fuels should be applauded, since they can be handled within largely existing infrastructures and they do not require the fabrication of new engineering on a staggering scale to do this, unlike hydrogen. Since such liquid fuels can be burned in diesel engines, rather than requiring fuel cells for which there is insufficient platinum to provide more than a small percentage of the current number of the 600 million vehicles on the world's roads, this tried and confirmed means is quite adaptable for the purpose.
However, the question of scale remains. The overall reaction for the conversion of glucose to IB may be represented thus:
C6H12O6 --> C4H10O (IB) + 2 CO2 + H2O.
Since the respective molecular weights of C6H12O6 and IB are 180 and 74, if the process were 100% efficient, we might expect a yield of 74/180 = 0.41 g IB/g of glucose. The actual yield is found to be 51%, and so we get 0.41 x .51 = 0.21 g of IB/g of glucose.
The world uses 30 billion barrels of oil each year. Taking the accepted conversion factor of 7.3 barrels per tonne of oil, this amounts to 4.1 x 10^9 tonnes.
There is a difference in the proportion of oil that is used to run transportation and for other purposes, and e.g. the US uses more of it for heating-oil. It is reckoned that 68% of US oil goes for transportation while the value is closer to 72% in Europe (UK). I shall therefore assume 70% for transport as a world average. Hence, 0.70 x 4.1 x 10^9 = 2.87 x 10^9 tonnes of oil are used to underpin world transport per year.
The heat of combustion of IB is 36.0 GJ/tonne (compared with about 30 GJ/tonne for ethanol) and it is reckoned that on an oil equivalent basis, crude oil can be costed-in at 42 GJ/tonne. Therefore we would need to produce (42/33) x 2.87 x 10^9 = 3.65 x 10^9 tonnes of it annually. Since 0.21 g of IB can be made from each g of glucose, we therefore need (1.0/0.21) x 3.65 x 10^9 = 1.74 x 10^10 tonnes of glucose.
If we assume a good yield of 16 tonnes of "sugar" per hectare from beet or cane (corn sugar yields are nowhere near this), we need 1.74 x 10^10/16 = 1.09 x 10^9 hectares of arable land to grow it on, or 1.09 x 10^7 km^2. That's 10.9 million km^2 and should be compared with the total of 14.9 million km^2 there is available over the entire surface of the Earth.
Hence we may deduce that the enterprise would require 10.9/14.9 = 73.0% of all of it. That's three-quarters! Since we still need to grow food for a rising world population and it is reckoned that without oil and gas to make fuels and run modern "industrial" agriculture that 14.9 million km^2 can only support 3 billion people or less than half the world's current population, the idea of using this technology or indeed any other means for turning food crop-land over to crops for biofuels on a large scale just seems crazy.
In pointing out such matters of scale I have been accused of autarky, when suggesting the dearth of land upon which the UK might be self-sufficient to some extent in terms of fuel production, as we are quite a small set of islands. For example, similar reasoning suggests that to produce the IB equivalent of 60 million tonnes of oil that we use each year just for transportation (and another 23 million tonnes is used for heating and as a chemical feedstock for industry) would require 179,000 km^2 of arable land or nearly three times the 65,000 km^2 we have altogether. "Surely you can just import it all from elsewhere," is the general theme, but the sums above show this simply doesn't stack-up on a world scale. It is striking how annoyed some people become when they are presented with hard numbers that fail to support their pet modes of energy salvation; that some technology will snatch us from the jaws of death at the eleventh hour: yet, this seems to me increasingly unlikely.
The US uses one quarter of the world's recovered oil, and so if 68% of that goes for transport, this amounts to: 0.25 x 0.68 x 30 x 10^9 barrels/7.3 barrels/tonne = 698.6 million tonnes oil equivalent. To produce this would need:
(1.0/0.21) x 698.6 x 10^6 x 42/33 = 4.23 x 10^9 tonnes of sugar, grown on: 4.23 x 10^9/16 tonnes/ha = 2.65 x 10^8 ha = 2.65 million km^2 of arable land. This amounts to (2.65/14.9) x 100 = 17.7% of all arable land on earth, or about one sixth of it. For interest, the total area of arable land in the US amounts to about 1.83 million km^2, to take an autarkic view, and so even the US could not be self-sufficient in fuel to any degree using this kind of technology, clever though it is since higher alcohols are usually only produced in minute quantities during fermentation processes.
Interestingly, these higher alcohols are also known as "congeners" and are thought partly responsible for the well known hangover if we drink too much in the way of alcoholic beverages, rather than the ethanol itself. It appears we are due for the mother of all hangovers in consequence of consuming too much energy, and the only cure will be relative abstinence.
(2) "Efficient Biofuel Made From Genetically Modified E.Coli Bacteria." http://www.sciencedaily.com/releases/2008/01/080106202952.htm