After writing this post on switchgrass and this one on Brazil and Ethanol on how I thought biofuels were the future, I came upon this article about a study by Tad W. Patzek that claimed more fossil energy is used to produce ethanol from corn than is produced in the ethanol. So I decided to bite the bullet and read his full report (.pdf) so I could better understand what he was saying and see if this would change my faith in biofuels.
The report was a bear. I can't recommend it to read. It was tough to understand where his numbers came from and how he was going from one step to another. And anyone who writes in a scientific report
Thus, the environment defined here cannot be dismissed summarily as the raving of a green lunatic.(his italics not mine) makes me a little leery on the results. So I checked with some other sources (really easy to find tons of ethanol energy reports with Google) and I have a decent grasp on it now. For this post I will go with Mr. "I am not a raving green lunatic"'s numbers because I know where to find his numbers and all the reports are within 20% or so of each other.
There are lots of things that are worthy of discussion here, but I am just going to look at energy losses/gains (ignoring other environmental issues like soil degradation and by product gains like corn syrup and oil). I think it makes more sense to look at this in two parts: creation of the corn and turning the corn into ethanol.
1) Creation of the Corn
To over simplify this step: you take your corn seeds, use some fertilizer, take some energy from the sun and create lots of corn.
Energy inputs (fertilizer+diesel fuel) + energy captured from sun = energy in corn
27 GJ of energy inputs (fossil fuels of some type)+ 57 GJ from sunlight = 84 GJ of energy in corn per ha (hectare = 2.4 acres)
Brief Aside 1: How good is corn as a way to capture solar energy?
About 20 TJ/ha of sunlight falls over 120 days, so only (57GJ/20TJ=) .285% of solar energy is turned into starch. This certainly makes the solar cells 14% efficiency seem impressive. Extra energy is absorbed by the corn that is not turned into starch but still this only gets you to .7%.
Brief Aside 2: Would it make sense to go "organic" and not use fertilizer?
I guess this depends on how you look at it and what you are comparing to. It is safe to assume that the average corn yield in the U.S. has increased 5-fold over the last 70 years. For example, in Indiana (Nielsen, 2002), the average corn yield was 30 bushels per acre in 1930 and 156 bushels per acre in 2001. The steadily improving yield resulted mostly from the increased fertilizer use and better corn genetics.
Ignoring the corn genetics, the farmers are currently using about 10 GJ/ha of energy in fertilizers. But, their crop energy yield has gone up 5 times from 16.8 GJ to 84 GJ. So this "investment" of 10 GJ in fertilizer energy lead to (84-16.8=) 67 GJ in new corn energy. So I think the fertilizer is a good energy investment.
2) Turning Corn into Ethanol
To over simplify this one (a better explanation here), you take the corn starch, put it in big heated vats with yeast, get ethanol and some non-energy byproducts (corn oil, corn protein, corn syrup), and distill the ethanol to 96% purity.
From an energy perspective (sticking with our per ha reference from before):
84 GJ of energy in corn + 48 GJ in fossil energy = 70 GJ of ethanol
So in total here you are going from 84GJ+48GJ= 132 GJ and getting 70 GJ in return, or a loss of almost 50% energy.
The first part of this energy loss is that taken by the yeast as they ferment the glucose in the corn grain into industrial beer. This is about a 16% energy loss (84GJ of corn going to 70GJ ethanol).
Brief Aside 3: How does the biochemistry of fermentation work?
This page has all sort of neat information on the biochemistry of the transformation and here is the chemical reaction:
C6H12O6 (glucose) + 2 Pi + 2 ATP + 2 ADP--> 2 C2H5OH (ethanol)+ 2 CO2 + 4 ATP +heat + biomass
All that work, and all the yeast get for their trouble are two lousy ATPs. But, that is where the energy "loss" goes. They also estimate that:
Approximately - 95% of sugar is converted into ethanol.The second part of the energy loss in this step is in the use of fossil fuels (mainly coal and natural gas). The fossil fuel energy to heat the vats and transport the fuel and other stuff (which I don't completely understand) takes almost 70% (48GJ/70GJ) of the energy that you get in the ethanol. Even if you assume that no fossil energy is needed to create your corn, this step still kills you. To make ethanol more viable there need to be serious reductions here.
Approximately - 1% is converted into cell material
Approximately - 4% is converted into other end products (other metabolites).
Some ethanol is lost from the wine to the atmosphere
Brief Aside 4: Can Genetic Engineering come to the rescue here?
One interesting idea is using a little genetic engineering to create better yeasts. Craig Venter and his boys over at Biological Energy Alternatives (IBEA), are looking into this (when he isn't traveling around to exotic locations, scuba diving and capturing and sequencing new species of bacteria).
This article has research creating a yeast that allows noncooking and low-temperature-cooking fermentation systems that have succeeded in reducing energy consumption by approximately 50%.
Total Energy Balance:
When you put the two steps together (per ha):
27 GJ of energy inputs (fossil fuels of some type)+ 57 GJ from sunlight = 84 GJ of energy in corn
84 GJ of energy in corn + 48 GJ in fossil energy = 70 GJ of ethanol
You get 27 GJ fossil fuel in creating the corn + 48 GJ fossil fuel energy creating the ethanol = 76 GJ fossil fuel inputs and 70GJ of ethanol output. You lose 76-70 = -6GJ or -6/70 = 7% energy on the whole thing. This number changes depending on what report you look at and what assumptions are used but even in best case scenarios only a 10% gain.
Brief Aside 5: What if you look at ethanol as a way of "driving on coal"?
Another way to look at this is that you are converting low value energy sources like coal and natural gas into a liquid energy that can easily be used in cars. In this way, it is similar to techniques that turn coal or natural gas into synthetic gasoline.
As this breakdown shows: 212,674 BTU of ethanol takes 137,750 BTU of Coal (64%), 28,547 BTU of Natural Gas (13.4%), and 30,000 BTU of diesel and gasoline (14%). The entire process doesn't create any energy but rather transforms coal and natural gas into a higher value energy source.
If you support energy independence for the US, you can look at this as a way of converting US coal into a fuel that Americans can drive. On the other hand, if you find coal to be dirty, cause air pollution, and give off more carbon dioxide than other fossil fuels, then the idea of your "clean corn" ethanol actually being 64% dirty coal isn't so comforting.
I had liked the idea of ethanol because I thought of it as a way to harness solar energy and turn into a liquid fuel to power our existing vehicles. Instead of collecting the solar energy with solar panels that would make the land an eyesore(and then somehow turn that electricity into a form that could be used in a car), we could dot the land with corn stalks.
Turns out that once you account for the fossil fuel usage to create ethanol very little (if any) of the energy is solar. Instead it is coal and natural gas. I can no longer be excited about ethanol, as it is currently being made from corn, from an environmental standpoint.
What would it take to make me like ethanol in the future?
If the distillation fossil energy could be cut in half (which seems very possible given the research into it), then the ethanol would have a positive energy balance of 27% and would capture 19GJ of sunlight energy per ha. It would go back to a way of harnessing sunlight for a portable fuel.
But, at 27%, or lets go 25% to make number easy, you would need 1/.25 = 4 times as much land to produce a given amount of energy. So the estimates I had before about how much land would be need to devote to fuel crops to fuel all the cars in the US in a sustainable manner would need to be quadrupled.