I had given-up on the idea that producing hydrogen by fermentation to run all the world's transport is at all feasible. I remain to be convinced that it is, as I stated in an early posting (Feb. 26th, 2006) "Biohydrogen from Sugar - a Preposterous Idea," on the basis that "We would need an area of land more than twice the size of the U.K. to grow enough crops to replace our current demand for liquid petroleum fuels by bio-hydrogen, and hence the concept is utterly preposterous." The other problem is that to fill the huge volume of reactors for fermentation would require 150 cubic kilometers of fresh water, which is more than the total volume available for every man, woman and child in the UK.
However, I was sent a an early press-release of a paper which reports on the greatly enhanced production of hydrogen, in yield and rate, that might be achieved by immobilising hydrogen-producing bacteria onto the surface of an electrode, passing a current and thereby stimulating the proton and electron generating activity of such "exoelectrogenic" bacteria using a small applied voltage. The cell is described in , and is impressive, in comparison with simply having hydrogen producing bacteria swirling around in a stirred fermentation vessel. Naturally, there are additional resource demands incurred by this more sophisticated technology, which employs a cathode made of carbon cloth onto which a platinum catalyst is supported. The anode chamber was filled with graphite granules and a graphite rod was inserted into the granules.
Bacteria from a soil or waste-water source were inoculated and enriched on a specific substrate using a phosphate buffer and nutrient medium. High yields of hydrogen were obtained from glucose and also from its commonly encountered fermentation products, e.g. acetic acid, butyric acid, lactic acid, propionic acid and valeric acid, meaning that by a change in applied voltage it might be possible to produce hydrogen from these too, and thus rendering the process of fermentation overall more efficient in respect to hydrogen production.
By looking at some rough numbers, it is possible to gauge the likelihood of the technology being adopted on the large scale, in order to match the amount of oil we currently get through in terms of fuel.
The reactor volume is given as 14 mls (anode chamber) and 28 mls (cathode chamber), making a total of 42 mls, from which 1.1 m^3 of H2 is obtained per day. The cathode has an area of 1 cm^2 and is made of carbon cloth on which 0.5 mg of Pt has been deposited.
To match 60 million tonnes of oil, we need about 6 x 10^9 kg of H2 (6 million tonnes). 1 kg of H2 is 500 moles, and occupies a volume of:
500 moles x 24.5 litres/mole/1000l/m^3 = 12.25 m^3.
Hence, 6 x 10^9 kg of H2 has a volume = 12.25 m^3/kg x 6 x 10^9 kg = 7.35 x 10^10 m^3
The reactor produces 1.1 m^3 of H2 per day x 365 days/year = 401.5 m^3/year.
Therefore we need: 7.35 x 10^10 m^3/year/401.5 m^3 H2/m^3 (reactor volume)/year =
1.83 x 10^8 m^3 reactor volume.
[This is a huge improvement over the 1.5 x 10^11 m^3 for a "free" fermentation process, and implies a factor of 800 less in terms of water required. However, in some of the fermentations, water is a reactant but even so, we still need much less than 1% of the comparable quantity of water to run it].
How much platinum is required? 0.5 mg/cm^2/42 mls of reactor cell volume in total.
1.83 x 10^8 m^3/42 x 10^-6 m^3 x 0.5 mg = 2.18 x 10^3 tonnes of Pt = 2180 tonnes. This is equal to the world output of new platinum for 14 years, and that is just to fit the UK's needs, let alone the rest of the world! Thus we have hit the first resource bottleneck.
We would also need 50g Pt/fuel cell x 33 million cars on UK roads = 1650 tonnes of new Pt for fuel cells in which to "burn" the hydrogen, making 3830 tonnes of Pt required in total, or 25 years worth of the world output of the metal.
How much land would be needed to grow the sugar crop? Let's assume that the technology can be adapted to extract 100% of the hydrogen in a sugar C6H12O6 (including the acids etc. that it produces in a first fermentation) which is pretty optimistic:
C6H12O6 ---> 6CO2 + 6H2 + 6 "O" (in an unspecified chemical form).
MW = 180 12
So, we need 180/12 x 6 x 10^9 kg H2 = 9 x 10^10 kg = 9 x 10^7 tonnes of glucose.
If we assume a yield of 19 tonnes of "sugar" per hectare, and an efficiency of 80% to extract the hydrogen, we need:
100/80% x 9 x 10^7/19 = 59.21 x 10^6 ha of arable land = 59,210 km^2 which is 91% of the total of 65,000 km^2 there is altogether. So, we couldn't grow any other crops for food, and while it represents a considerable improvement over unassisted fermentation of sugar into hydrogen, it is still impractical on the grand scale of our transportation requirement.
How much generating capacity would be needed to run the system, by applying a voltage to the anodes?
The average is 300 mW/m^2 of electrode surface.
1 cm^2 corresponds to 42 mls of reactor volume, and the total reactor volume is 1.83 x 10^8 m^3.
Hence the total electrode area is: 1.83 x 10^8/42 x 10^-6 x 1 cm^2 = 4.36 x 10^12 cm^2, and since 1 m^2 = 10^4 cm^2, this amounts to 4.36 x 10^8 m^2.
Thus, the power needed is: 4.36 x 10^8 m^2 x 300 x 10^-3 W/m^2 = 1.31 x 10^8 W = 131 MW, which is not too bad, about 13% of the output of a typical power plant.
How much graphite is needed?
Anode chamber has a volume of 14 mls. If we assume spherical particles, their volume is:
4/3 x pi x (4.54/2 x 10^-3)^3 = 4.9 x 10^-8 m^3. To find the overall volume they occupy, it is helpful to imagine each one occupying a cube of side 4.54 x 10^-3 m (4.54 mm), for which the volume is:
(4.54 x 10^-3 m)^3 = 9.36 x 10^-8 m^3. The total anode volume is (14/42) x 1.83 x 10^8 m^3 = 6.1 x 10^7 m^3, of which, (4.90 x 10^-8/9.36 x 10^-8) x 6.1 x 10^7 m^3 = 3.19 x 10^7 m^3 is graphite. There is a graphite electrode inserted too, which occupies some of the internal space of the cell, but assuming the volume just determined and a density of graphite of 2.25 tonnes/m^3, this amounts to:
3.19 x 10^7 m^3 x 2.25 tonnes/m^3 = 7.2 x 10^7 tonnes or 72 million tonnes of graphite.
As a means to replace oil for transportation, the technology could not be scaled-up sufficiently for the task, certainly not to fuel the entire world's transport. The above figures only refer to the UK, and should be multiplied by around 20 to meet the needs of ca 600 million road vehicles as there are reckoned to be altogether. This would mean that 3830 tonnes x 600 million/33 million vehicles = 69,636 tonnes of Pt would be required, and yet the metal is recovered at a rate if about 150 tonnes per year, implying it would take 464 years to install the lot, using electrohydrogenolysis with fuel cells. This quantity is actually close to the reckoned world reserve of Pt, and so we all of that would need to be turned-over for this purpose, and none for jewelry, scientific apparatus or catalytic convertors to keep the internal combustion engine powered vehicles running "clean" while they were phased out by the new "hydrogen" technology.
It is an interesting paper, and the authors may be correct in their assertion that the technology might still prove useful for local fertilizer production, say, even if a full-scale transportation system based on hydrogen is never implemented (which it never will be). However, the scale even of this will be likely be very small, for the simple facts of limited resources and the otherwise massive engineering requirements.
 S.Cheng and B.E.Logan, "Sustainable and efficient biohydrogen production via electrohydrogenolysis," PNAS, 2007, Early Edition.