« Hybrids Accelerate in BC | Main | To Cut Or Not To Cut? »

March 02, 2005

Everybody Loves Biofuels

Finally, something that both ends of the political spectrum can agree on:  biofuels.

This Oregonian article describes an alliance between a rural rightwinger and an urban lefty, both of whom have fallen for fuels made from grain and oilseeds.  The rural legislator likes them because they boost farmers' income; the urban legislator likes them because they reduce fossil fuel consumption.

I'll leave my thoughts about food-based fuels for a later post.  But in the meantime, it seems like the article omits an important biofuel--ethanol that's made from crop and wood wastes, rather than grain. It's called cellulose ethanol, and there's a much more energy potential from that than there is from food crops.  A Canadian company, Iogen, is currently planning to build a cellulose ethanol plant in Idaho that will use crop waste as a feedstock.  (For more on this topic, see the energy section of this year's Cascadia Scorecard.)

Posted by ClarkWD | Permalink


It's my understanding that ethanol takes as much fossil fuel to produce than is obtained through burning it. In other words, it's a highly inefficient process. I'm happy to stand corrected on this one.

I drive a biodiesel car so I'm not trying to throw stones at biofuels, to give you some context.

Posted by: Dave | Mar 2, 2005 4:54:12 PM

Dave -- although I said I'd post on food-based biofuels later, I guess now's as good a time as any.

First, I believe that the figure you cite is for corn-based ethanol, not cellulose ethanol. Cellulose ethanol has a much better energy balance.

Also, it probably comes from research by Cornell scientist David Pimentel. Most researchers who've looked at the issue disagree with him, saying that his input and yield figures are out of date. (And for the record, I have no reason to believe that the researchers who are critiquing Pimentel's work are simply shilling for ethanol producers. I think they're sincere.)

But, at heart, you're right. At the very best, you only squeeze a little "net energy" from turning corn into ethanol, after accounting for all of the energy that goes into fertilizers and tractor fuel for the corn, and the ethanol distillation process itself. And most of the positive "net energy" comes from considering so-called "co-products" -- e.g., the stuff that's left over from the distillation process that you can use as cattle feed, offsetting some of the fossil fuels that would have gone to produce feed from scratch.

Perhaps the best that ethanol has to recommend for itself is that much of the energy input is from coal (ethanol distillation apparently uses electricity, which mostly comes from coal), but the energy output is transportation fuel. So you use some of America's abundant coal reserves to offset some petroleum imports.

Big whoop.

So basically, corn-based ethanol uses a lot of land, fertilizers, and pesticides, for a small net energy gain. And it also requires a bunch of federal subsidies and regulations to make it economically feasible -- money that would be better put to other purposes.

So if you can't already tell, I hold a dim view of corn-based ethanol. But, again, cellulose ethanol is another matter entirely.

Posted by: Clark Williams-Derry | Mar 2, 2005 8:12:11 PM

You guys are doing a great job on this blog.

I do have some concerns about the EROEI for cellulose ethanol - do you have references outside of the company website? (EROEI = Energy return on energy invested)

Sure there are critics of Pimental. On the other hand, has anyone ever made money, or, positive net energy on any biofuel, when government subsidies or recycling are taken in to account?

Not to my knowledge, and believe me I'm looking.

I am glad Iogen is trying to do this - but I'd like to see the numbers before getting on board.

Posted by: Jon S | Mar 2, 2005 8:28:20 PM

Jon S. -- Why, I'm glad you asked.

According to this report:
cellulose ethanol has a net energy gain of 60,000 BTU per gallon. (The report cites a study published in 2000, which I haven't reviewed.)

Ethanol contains about 76,000 BTU per gallon. So that means, what, a gross return of 76K BTU for 16K BTU invested -- or an EROEI of ~4.75.

(I think this is how you do an EROEI calculation -- if I'm wrong, it's just because I'm not sure, not because I'm trying to be misleading. And, of course, this is at best a rough estimate, espceially since I don't know precisely how the 60,000 BTU figure is calculated, and how it compares with other EROEI studies.)

With these caveats, my opinion is that the EROEI of cellulose ethanol is positive enough for it to be worth doing, or at least taking a very serious look at.

And here's one another thought re Pimentel's energy calculations. People who disagree with Pimentel tend to use average yields and inputs for for the entire U.S. corn crop. But if ethanol subsidies were ended and demand for corn fell, the land that would go is likely to be the marginal cropland -- the stuff which has the lowest average yields, highest input costs, or higher risk of crop failure. (This ain't necessarily true, of course -- with a fall in corn demand, some farmers with high-value corn cropland might decide to get into some other crop. But overall, if the aggregate demand for ag. products drops, then the stuff that is most likely to go out of production is the stuff that's the least economically productive.) So perhaps looking at the marginal cropland is the right way to go -- and that means lowballing yield estimates, or highballing estimates of energy-intensive inputs.

In other words, Pimentel's EROEI figure could be close to the mark for the specific land that would go out of production if ethanol subsidies ceased. I just haven't seen this argument made anywhere else, so I don't know whether these numbers have actually been run.

Posted by: Clark Williams-Derry | Mar 3, 2005 10:56:23 AM

My understanding is the energy energy in-energy out equation these days is something like 0.9 for petroleum gasoline and 1.3 or so for ethanol. That’s significant in itself, but as you note, the end game is clearly cellulosic ethanol. We believe that as demand grows nationally the Midwest will no longer be able to meet it with corn, and as it continues to grow grain will no longer be able to meet it, and cellulosic comes of age. As with the wind industry, you have to walk before you can run – a new industry can’t go from zero to 60 to instantaneously replace an incumbent – to evolve it needs real world experience. With ethanol the environmental gains have and will continue to grow with time if we provide consistent and growing market pull.

And – hey – what about those fossil subsidies! If you’re going to diss ethanol subsidies you oughta be fair and compare. Whether or not to include the Iraq invasion and occupation is a tough call, though…

Posted by: rhys roth | Mar 4, 2005 3:46:18 PM

USDA's current estimate is that corn ethanol has a net energy balance of 1.67 -- it's gone up since they last ran the numbers, both because of gains in crop yields per energy input, and improvements in ethanol production.

See, eg.,

originial study here:

Most of the positive net energy balance can be attributed to co-products -- stuff left over after the ethanol is produced that can be, say, fed to cows, offsetting other energy expenditures for cattle feed.

Still, if you look at ethanol alone, and ignore the coproducts, USDA says that you get a little more energy from the ethanol than you put in for fertilizer, transportation, and distiallation.

Now, I think the figure of 1.67 may be high for two reasons:

First, it doesn't include the energy costs of constructing ethanol plants. (Ethanol critic David Pimentel does try to include construction energy costs, but apparently uses figures that are too old to be relevant.)

Second, the USDA numbers look at the average yield & input numbers for major corn producing states. But arguably the best thing to do is to look at the marginal cropland -- the land that has low yields, high costs for energy-intensive inputs, or is most prone to drought or floods. That's the land that's likely to go out of production if demand for corn were to fall.

If you look at the numbers in the pdf above, there's big variation by state in crop yields and energy costs. The variation within states is likely to be just as large. So the marginal acre of corn -- the acre that would go out of production if ethanol subsidies stopped -- likely has a much lower energy balance than the average acre.

(I hope I'm not repeating myself here...)

Posted by: Clark Williams-Derry | Mar 4, 2005 3:50:44 PM


A few questions:

1. What’s the rationale for excluding the co-products? If the energy in goes to produce not only fuel but other products that would otherwise require additional energy to make… seems reasonable to count that in some way.

2. Whether or not to add the energy it takes to construct the ethanol plant – yes, that makes sense if they also include the energy costs of constructing the refinery and other infrastructure when tallying how much energy goes into producing a gallon of fossil gas – do you know if they do? (by the way, what’s your best number as to energy in-energy out for fossil gasoline? It seems that a lot of discussion of ethanol’s energy costs don’t compare it to fossil gasoline’s energy cost – don’t you have to for it to make sense?)

3. As for choosing marginal lands over average yields-inputs for the calculation, I’m not certain that is valid in terms of real-world economic decision-making, but it could be. It could be good to get David Morris’ view on this point if you haven’t already. He lives in corn country and pays close attention to ethanol. He’s VP at the Institute for Local Self-Reliance – why don’t you run that by him: dmorris@ilsr.org.

Posted by: rhys roth | Mar 4, 2005 3:53:58 PM

Good questions.

1. I don't think there's a particular rationale -- it's just a point that the USDA researcher went to some length to emphasize in his study, that corn ethanol now has a slightly positive net energy balance even excluding coproducts -- that is, if you use 1 unit of fossil fuel energy, you get 1.06 units of ethanol energy out, excluding coproducts.

That used to not be the case -- in previous studies, I believe, the coproducts were essential to make the net energy balance positive.

I haven't thought things through -- there might be a reason to exclude the extra .61 units of energy embodied in the coproducts. But for the time being, I take the USDA number -- 1.67 -- at face value (with some caveats that it may be an overestimate.)

2. I'm just not familiar enough with the literature to know whether refineries, pipelines etc. are included in gasoline energy balance calculations. They should be, obviously.

I've seen a few estimates that suggest that the net energy balance for gasoline is between -15% and -26%. (That is, you start with one unit of energy in the crude in the ground, and you get .74 to .85 units of usable energy in the gasoline after pumping, shipping, and refining it.) That number may vary, depending on the quality of the crude & how hard it is to get out of the ground.

To me, the better comparison isn't energy balance but carbon balance--apparently, a lot of the energy used to produce ethanol comes from electricity, much of it from coal, which could make the net carbon benefits a little lower than the net energy benefits. But I haven't seen an estimate of this that I trust -- or, really, I've seen too many estimates with widely different values, so I'm not sure what I trust.

3. That could be right. If subsidies for ethanol were dropped, you could see high-value cropland switched to some other commodity besides corn, and low value cropland staying in corn. (In which case, the average energy balance of ethanol might fall a bit.)

But having worked for a long time on farm subsidy issues in DC, the economic consensus that I absorbed was this: farm subsidies artificially raise demand for commodities; without subsidies, apparent demand would fall; if demand fell, production would fall; and the first land to leave production would be the marginal cropland -- land that was too wet, too dry, too cold, too hot, or too exhausted to farm profitably without subsidies. It seems to me that this is the land for which the net energy balance is pretty bad -- the land for which yields are low and inputs are high.

This consensus may be dated or wrong or not subtle enough, but it's the assumption that I was working from.

Posted by: Clark Williams-Derry | Mar 4, 2005 3:57:16 PM

Oh, and after looking throught the USDA study, I've found a possible reason to be skeptical of the USDA's treatment of coproducts.

There's a big question about how much of the energy inputs (fertilizer and fuel on the farm, energy in the ethanol conversion plant) to assign to ethanol per se, vs. how much should be assigned to the various coproducts--cattle feed, corn oil, and the like.

Do you split up the inputs by market value of the end products? By the energy content of the end products? By the weight of the various end products? Do you assign energy values to the coproducts--say, corn gluten feed & corn oil--based on how much it would take to produce those same products through other means?

The study chooses a simple, and defensible, method--it assigns energy inputs based on the weight of starch vs. non-starch in the corn. Starch accounts for 2/3 of the average dry weight of a kernel of corn, so the study attributes 1/3 of the total energy of farm inputs, and transporting the corn to the ethanol plant, to the coproducts. And the study assumes (for reasons that I can't quite decipher) that between 36% and 41% of the energy from the ethanol production process is attributable to the coproducts.

Now this is fairly transparent, and probably defensible. But its also debatable. Many of the coproducts--cattle feed, for example--could probably be produced through far less energy intensive means. So it could be cheating, sort of, to claim that more than a third of the energy for inputs and in ethanol plants is attributable to the byproducts, rather than to the ethanol itself.

(Sort of like driving 50 miles to buy a special bottle of wine, and picking up some milk while you're there -- and attributing half the energy cost of the trip to the milk. You could have gotten the milk for a lot less energy by walking to the corner store...)

So, these factors--the lack of accounting for energy costs of ethanol plants and other physical capital; assuming yields and inputs for average, rather than marginal, cropland; and (potentially) assigning too much energy to the coproducts rather than the ethanol make me think that the net energy balance figures cited by USDA may be a little too high.

Posted by: Clark Williams-Derry | Mar 4, 2005 9:03:15 PM

Thanks for all the insights - I will follow up on the cellulose stuff. I would agree with your final analysis that the USDA numbers are high; I don't think you squeeze positive energy out of corn ethanol.

Your EROEI math was fine, by the way, there is no set way of doing it. The basic idea is what matters.

Posted by: Jon S | Mar 4, 2005 11:17:27 PM

Corn and other grain-derived ethanol are transition feedstocks to cellulosic ethanol, and should be understood in this context. Subsidies for grain ethanol are building the market for ethanol overall. Going right to cellulosic is indeed asking the industry to run before it can walk. Since cellulosic ethanol represents processing challenges not faced by grain ethanol, it now costs more. But eventually cellulosic will cost less because feedstocks are so much cheaper and more abundant. Subsidies help the industry reach that point.

It should be noted that the state with the most active state-level subsidy program, Minnesota, has used it to create an ethanol industry almost entirely owned by farmer coops. Minnesota caps producer credits in such a way that they capitalize the plant, which is enough. Now, 80% of Minnesota ethanol is produced without state subsidies.

We don’t like to use the term “industrial policy” in the U.S., even though we employ industrial policies all the time. Ethanol subsidies represent a policy that has founded a successful industry. Whether or not this has been the absolute best way to spend the money, it has been productive. And let’s not underestimate the realpolitik importance of a renewable energy program that actually has broad and deep bipartisan support. If we were not spending the money on ethanol, it is not very likely we would be devoting it to solar photovoltaics or fast trains. This money is actually going to build a real renewable energy system. If it wasn’t it would probably be buying smart bombs.

What would happen if ethanol subsidies were suddenly removed? The only fair way to pose this question is to ask what would happen if fossil subsidies were also removed at the same time. That would level the playing field significantly, and in that framework, what would happen to ethanol and ethanol cropland demand? I don’t know if anyone has even tried to model this. If ethanol subsidies alone went away, you probably would see corn acreage go out of production, and more petroleum burned. Not an ideal result environmentally or any other way (except if you’re an oil company).

The ethanol industry certainly has the environmental impacts of agriculture, though increasingly corn is cultivated no-till so erosion problems are lessened. The proper comparison must be to oil wells, pipelines, tankers and refineries. Can anyone seriously propose they would rather live next to a drill field than a corn field, or an oil refinery rather than a biorefinery? Of course sustainable agriculture proponents need to work for improvements in cultivation of fuel crops, as they do for food and fiber. The point is that ethanol is already more sustainable than petroleum fuels, and as efficiencies improve and cellulose becomes the feedstock, ethanol will be exceedingly more sustainable.

The energy balance study I like to quote is Argonne National Laboratory’s.
www.transportation.anl.gov/pdfs/TA/58.pdf It shows that ethanol is a 10% blend with gasoline reduces greenhouse gases by 2% and fossil energy use by 3%. Near future cellulosic ethanol in the same mix cuts greenhouse emissions by 6% and fossil use 6-7%. Of course, we want to get to higher blends fast, because this is where the big environmental gains are to be found.

But should we be using cropland to grow fuel? Absolutely, yes. We should also be using it to grow chemicals, materials and everything else we now manufacture from petroleum. Otherwise we will continue to pump up ancient biomass in the form of petroleum and dump millions of years worth of accumulated photosynthetic energy into the atmosphere in a few generations with impacts we all agree are unacceptable. Is there enough land? In the context of dramatic efficiency improvements, yes, and that is an important context for all discussions of future food, fiber, fuel and materials needs. Otherwise, every rock we roll up the hill will roll back down on us.

Posted by: Patrick Mazza | Mar 7, 2005 4:40:05 PM

Couple follow-ups on your insightful points, Clark:

I’m wondering if the energy balance data you cite for petroleum might be a case of apples to oranges. In your example of one unit in the ground yields .74 unit of gasoline – does that factor in how many units of outside energy it took to explore for, extract, transport, refine and deliver (plus to construct the refineries, etc.)? It may well, but it sounded like it could just be how much of the original energy survives all the production processes, which would be different that the ethanol calculation.

That applies to net carbon balance, too, of course. On ethanol’s net carbon balance – the goal of renewable fuels legislation (where this whole discussion started) in the region is in-state production. So we’d want to know what the carbon balance would look like in MT-ID-OR-WA rather than in the Midwest, yes?

I don’t want to defend corn subsidies in the least, though I see now that it has some complex relationship to the energy balance of ethanol. I do want to suggest that subsidies for petroleum be part of any discussion of subsidies that benefit ethanol.

Posted by: rhys roth | Mar 8, 2005 10:19:39 AM

I'm glad you folks are there. Where I live, we've got a legislator who's proposing a mandatory 10% ethanol in all the gas, all year 'round, in an area where nearly all of the farmland has been mushrooming into subdivisions faster than you can blink. Traffic and car issues have become a nightmare on the Front Range, particularly in the winter.

I'd like to believe this would work, but I keep hearing really contradictory info. I do have one question of this group, if you don't mind. Do you think that the major agribusiness players will push out ethanol-producers if fossil-fuel subsidies go down or if oil and gas start getting scarcer? Will we just end up giving different subsidies to the same big corporations?

Posted by: Eva Kosinski | Feb 13, 2006 8:00:36 AM