What, you disagree? Allow me to bombard you with some inconvenient truths.
1000L of ethanol supplies 20GJ of energy. But how much does it cost to make the ethanol? To grow the corn required to make 1000L of ethanol, you must spend 14GJ of energy on an array of tasks and resources ranging from fueling the tractors and irrigating the land to making fertilizer and moving your crop. Pretty good so far, right? You get 20GJ by spending 14GJ. Not so fast. Ever try to dump a bag of frozen corn in your gas tank? Trust me, it doesn't work too well. Don't ask.
Once you have the corn, you have to crush it, ferment it, and distill the ethanol, and that costs energy. How much energy? That depends on who you ask. Some estimates (Pimentel, 1991) go as high as 20GJ, others as low as 11GJ (Marland, 1991), with several others (Keeny and DeLuca, 1992; Morris and Ahmed 1992; Shapouri and Duffield, 1995) falling somewhere in between. So, that gives us a range. Depending on who you talk to, it takes 34-25GJ of energy to make 20GJ worth of ethanol. This means that for every GJ you spend on making ethanol, you get 588 - 800MJ back. Remember, it is customary to leave solar input out of these equations. And yes, I linky. It's a pdf, see table 8 on page 30 (I converted from the given kcal values).
But....but...soybean biodiesel...that's good, right? It's better, but not great. Every GJ of soy biodiesel costs 1.27 GJ of energy to grow and prepare (Pimentel and Patzek, 2005; sorry, no linky here; too lazy to find it online). However, unlike maize-ethanol, soy-biodiesel has usable byproducts (soymeal). When the energy value of that is factored in, you get 1GJ of biodiesel that costs just over 1GJ to make. Sure, it's OK, but what's the point? You're literally spinning your wheels.
Of course, there is some variance from study to study, but the general concept holds: It is either pointless or counterproductive to use biofuels from crops. Thus reads the excellent argument against biofuel use. It's a feel-good proposition, but that's it. In the latter case it does nothing, and in the former, worse than nothing. Furthermore, this takes up arable land in a world where arable land is rapidly becoming a scarce and valuable resource to feed an ever-increasing population. The use of biofuels is counterproductive. Using current technology.
Enter the recent Science paper. It's one of those smack-your-forehead why-didn't-I-think-of-that kind of things. They took farmed out land; sandy, nitrogen poor soil that had been overfarmed and was now good for little more than pastureland. There are no commercial crops that can be grown on it without heavy fertilization and preparation. But "commercial crops" are only a very small percentage of plant life. Enter the native grasses. These guys tilled it, sowed randomized combinations of 1, 2, 4, 8, or 16 different grassland plant seeds, watered them, and then maintained them with "low input" techniques. That is, they were weeded once each season (less in the higher diversity plots), and let go. They planted a grassland.
They let it go for 10 years, burning the grass off each spring after taking a sample for above ground biomass production and then replanting. Energy out based on total biomass was 68 GJ/yr/ha. Energy required to plant, grow, maintain, and harvest was 4GJ/ha/yr. That's a ratio of 17:1. The ratio for corn is a little less than 5:1 (reference in first link).
According to the Science paper, if you were to simply burn this stuff in a furnace as a coal supplement and use it to run a generator, you'd net 17.8GJ of electricity/ha. That means that spending a GJ of energy gives you back 4.4GJ. And that's just burning it. Ethanol conversion is about the same as just burning it, but if you convert it to biohydrocarbons, you net 28 GJ. One GJ of energy for the low, low price of 140MJ.
BUT THAT'S NOT ALL!
When they started the experiment in 1994, they measured the carbon content of the soil. They did it again in 2004, and found that in each hectare, there was an additional 4000kg of carbon in the soil. Yes, friends and neighbors, this stuff sequesters carbon. And remember, they were burning off the tops of the plants, all the carbon that was in the stems and leaves was burned back into the atmosphere, just like what would happen if the plants were harvested, converted to biofuel and burned. Just without the intervening steps.
So, how does it work? Why is it so damn cheap to make it this way? Simple : Entropy rules. You can maintain an ordered state, but it will cost you a good deal of energy, and that's what traditional monoculture agriculture is. Weeds creep in, pests eat the crops, plants die and compete for resources. The ordered state of the farm is constantly under attack, and the farmer has to expend vast quantities of energy to keep entropy at bay. For making foodstuffs, this is necessary and good. If you let chaos reign, the natural progression of the ecosystem will take over and the foodstuffs will be gone in no time. But this doesn't happen in the wild. In a grassland (or any ecosystem), there is a balance, a dynamic equilibrium that will act to maintain itself. Polyculture systems simply capitalize on this concept. A Fischer-Tropsch reaction doesn't care if the carbon in the chamber comes from soybeans, corn, switchgrass, or coal. A coal furnace cares even less. So, you can grow a mixture of hardy crap plants, the kind of plants you try to stave off in the monoculture you have growing in your front yard. You can even throw in the bugs that are sitting on the plants at harvest time.
Basically, you have an artificial ecosystem. Who needs fertilizer when you can put in legumes, which fix nitrogen right out of the atmosphere? Who needs pesticides when the plants themselves have been fighting the same parasites and herbivores for the past few million years? Who needs herbicides in a weed garden? Who needs to irrigate when you can plant grasses that are adapted to the normal precipitation in the region?
But is this practical? Can it be done? Let's run a few numbers.
Energy consumption on a national or global scale is measured in quads. One quad is 1 quadrillion BTUs, or about 1 billion GJ (1055055900, to be exact). The world uses about 421 quads each year (2003 data). This translates to 4.4 x 1011 GJ worldwide. Using the most efficient method, biohydrocarbon production, this requires 4.4 x 1011/28 ~ 1.6 x 1010 ha = 1.6 x 108 sq km of land. The world has a land area of 1.48 x 108 sq km. We're a bit short.
If every square inch of land was usable for this (which it's not), and was used for growing weeds, (which ain't gonna happen) we'd still be short by about 3.36 x 108GJ (0.31 quads).
OK, how about just using it as a supplement? Would that work?
By the above calculations, even if we assume we could get to soil in Antarctica, grow it in the Sahara, and harvest it in the Himalayas, we need about the same percentage of the Earth's land area as the supplement percentage. That is, if we supplement 10%, we must commit 10% of the Earth's land area to the cause. That's an area roughly equivalent to the entirety of the US and Mexico, plus about half of Canada.
But let's take it a step farther. About 15% of that 421 quads already comes from clean or renewable sources of energy like nuclear, wind, geothermal, etc. So to fulfill all of Earth's current needs, minus those sources, we have to supply 357 quads, requiring 1.36 x 108 square km. To hit the 10% target, now we just need all of the US, all of Mexico, and about a quarter of Canada.
Biofuels are not the answer. Unless, of course, you want to try the sea.