Update [2007-9-25 15:12:2 by Tom Philpott]:In the 24-hour lag time between finishing this piece and its posting, I had an email exchange with Keith Smith of the University of Edinburgh, one of the authors of the study discussed below. I’ve modified the post to add information I got from Smith.

By all accounts, biofuels deliver startlingly modest reductions in greenhouse gases. In a relatively generous assessment of the environmental benefits of ethanol and biodiesel released last year, University of Minnesota researchers credited corn-based ethanol with 12 percent less net greenhouse-gas emissions than gasoline, while finding that soy-based biodiesel emits 41 percent less.

But here’s the catch: It takes so much corn to produce a gallon of ethanol, and so much soy to produce a gallon of biodiesel, that the net GHG advantages are likely to be almost nil. The U of Minn researchers write [emphasis mine]:

[I]f one replaced a total of 5 percent of gasoline energy with ethanol energy, greenhouse gas emissions from driving cars would be a bit more than a half percent lower (5 percent times 12 percent).

Whoa. In 2006, U.S. ethanol producers burned through 18 percent of the corn harvest to offset 3 percent of gasoline use. What the Minnesota study is telling us is that we could increase corn ethanol production by two-thirds (to achieve a 5 percent offset) — burning through 40 percent of the corn crop — and still only reduce greenhouse gas emissions by just a bit more than a half percent.

By my calculations — based on 2006 output of 4.8 billion gallons of corn ethanol — it would take about 8 billion gallons to achieve that 0.5 percent drop in GHG emissions. Ethanol production is subsidized by a dizzying array of public programs; the most direct one is the $0.51 per gallon blender’s tax credit for using the stuff.

To gain that razor-thin GHG advantage, the Treasury would be out some $4 billion ($0.51 times 8 billion) per year, just from that one form of public support. This is sound public policy? In a rapidly warming world, dropping $4 billion on corn ethanol seems clinically insane as a strategic use of the public purse.

And — finally coming to the point of this post — a new study [PDF] has emerged declaring that even that comically paltry GHG benefit may be spectral. Biofuel use may actually increase GHG emissions.

The study, published in the journal Atmospheric Chemistry and Physics, was performed by American, German, and British researchers, and included the Nobel laureate Paul Crutzen. I have no information on funding.

It claims that biofuel production emits far more nitrous oxide, a greenhouse gas some 296 times more potent than carbon dioxide, than has normally been assumed.

The source is artificial fertilizer, a potent source of nitrogen. When farmers apply it to soil, a certain amount of it — between 3 percent and 5 percent, according to the study — enters the atmosphere as nitrous oxide.

In our email exchange, the University of Edinburgh’s Keith Smith, one of the study’s authors wrote that, “if you increase the inputs of [fertilizer] through expanding biofuel production, you can expect more N2O [nitrous oxide] pro rata.”

As a result, corn-based ethanol releases, at worst, 50 percent more GHGs than conventional gasoline, the study concludes. At best, corn ethanol emits 10 percent less GHGs than gasoline, or less than half of the advantage (22 percent) calculated by the U of Minn researchers.

Biodiesel from rapeseed — the preferred feedstock in Europe — is even worse, the study says. It releases as much as 70 percent more GHGs than conventional fuel. Ouch.

I asked Smith precisely how his study’s methodology differ from that of the U of Minn researchers. He replied:

Assessments like this one you mention are calculating the net gain when the fossil fuel energy needed to produce a biofuel is subtracted from the energy value of the fuel — as your figures show, 78% (i.e. 100-22%) or 59% (100-41) of the “gain” is cancelled out by these energy costs. Our assessment of the global warming implications is largely additional to these considerations, becasue most assessments hitherto have ignored the N2O emission issue, or only allowed for PART of it — we have caclculated a bigger percentage conversion of fertiliser N to N2O than previously accepted.

Smith’s study claims that cellulosic ethanol made from agricultural "wastes" — e.g., Archer Daniels Midland’s big plan to use "corn stover" — would "cause unfavourable or low-gain impacts on climate." Cellulosic ethanol from switchgrass and other perennial grasses would be much better, though, the report states.

Even so, as I reported a couple of weeks ago, the USDA — cellulosic ethanol’s research great sponsor — is becoming increasingly dubious about the viability of cellulosic technology.