[ChuckMcM]

Pardon me for being a bit pedantic here but not all bio fuels require farmland.

Bacterial based processes (generally e.coli) work in digestor/reactors), algae based fuels simply need access to sunlight (can be grown in tubes attached to the side of buildings). Further some starter crops can use land that is not deemed arable by farming standards.

[dredmorbius]

The scale at which algae would have to be produced to provide even a small fraction of existing petroleum consumption would mean you'd be growing it on rather more area than you've got affixed to building facades.

For the US alone, sever tens of millions of acres, at 15,625 square miles (40,468 km^2) per 10 million acres.

If you were, say, build your 10 million acres of algae grow tanks along the length of US Interstate 80 (2,899 miles), they'd extend 2.6 miles to either side of the highway, for its full length. And that's about 25% of what would be required to replace present oil consumption, so figure on extending your tanks out another 7.8 miles on either side, or string tanks along Interstates 10, 40, and 70 as well (mind the routing around the Great Lakes). Oh, and algae don't freeze well, that's going to cut into your growing season.

Plus you've got to provide water, nutrients, stirring, and keep pathogens and scavengers off the crop.

It's ... a pretty significant undertaking.

[ChuckMcM]

You could grow algae in clear tubes across 15 square miles of the Mohave desert and not consume a single acre of arable farmland.

[dredmorbius]

Where are you going to get the water from?

If it's seawater, how are you going to return it to the ocean. You can't dump salt water on land -- it poisons the soil. Even freshwater evaporating in dry climates leads to salinization.

What's that going to cost you in energy inputs (pumping costs)?

Though the thought occurs to me that the Salton Sea in southern California might make a possibly suitable grow region.

[ChuckMcM]

If we've posited a desert based system, I'm going to guess it is closed loop with respect to water (which is to say the water never leaves the system as anything except perhaps evaporation during post processing.) So the initial water budget probably comes from the aquifer or is imported. Once the system is running an interesting question is how much water would it lose over time.

And while that is an interesting question, it doesn't change the situation that no farmland was harmed in the process. :-)

[dredmorbius]

For a system comprising tens to hundreds of millions of acres, incidental and transpiration water losses are going to be nontrivial.

Hermetically sealing such a system would be ... prohibitively expensive.

You'll be pumping water in from somewhere.