Vinod Joins the Cellulosic Ethanol Bandwagon

Time once again to revisit Fat Knowledge's favorite topic: ethanol. (See articles here, here, here, and here). This time super VC and founder of Sun Microsystems Vinod Khosla is throwing his weight in the ring. From the NY Times:

"I am convinced we can replace a majority of petroleum used for cars and light trucks with ethanol within 25 years," he said. He has already invested "tens of millions of dollars," he said, in private companies that are developing methods to produce ethanol using plant sources other than corn.

What is his motivation? From The Economist:

Like many very rich men, he now wants to improve the world: “Just starting another Sun doesn't do it for me any more.” As an engineer turned venture capitalist, Mr Khosla has a healthy respect for the power of new technologies to create disruptive innovations. And the free marketeer in him clearly relishes the prospect of really taking on the big, rich and well-entrenched firms that dominate the oil industry.

He even created a very informative 102 pg PowerPoint slide with his sales pitch (isn't there a law against creating PowerPoint slides that long?). So Vinod thinks that Cellulosic Ethanol is the way to go. Is it? Lets take a look at the pertinent questions.

1) What is cellulosic ethanol?
Cellulosic ethanol is ethanol that is generated from cellulose rather than corn or sugar. Examples of cellulosic materials are paper, cardboard, wood, and other fibrous plant material. The probable feedstocks for cellulosic ethanol in the US include corn stover and stalks from other grains and perennial grass (e.g. switchgrass and Miscanthus).

2) Does it require more energy to produce than it gives out?
One of the big debates about corn ethanol is whether it has a positive energy balance. Depending on assumptions corn ethanol has either a slightly negative or slightly positive energy balance. This means that it is not really creating energy as much as it is transforming coal and natural gas into fluid energy that can be used in a car.

Cellulosic ethanol on the other hand is much more strongly positive. Vinod's PowerPoint slide puts it at 4-8 times:

Energy Balance (Energy OUT vs. IN)
“Corn” ethanol numbers ~1.2-1.8X
….but reality from non-corn ethanol is…
Sugarcane ethanol (Brazil) ~8X
Cellulosic ethanol ~4-8X
Petroleum energy balance at ~0.75

This EIA paper refers to another paper that says it a net energy balance of more than 60,000 Btu per gallon. They state a gallon of ethanol has 76,000 BTU, that looks to me like it takes 16,000 to create 76,000 or a balance of 4.75.

This report puts it at 162% increase in energy but it is over 10 years old and might be out of date.

This article states:

So, according to the U.S. Department of Energy, for every one unit of energy available at the fuel pump, 1.23 units of fossil energy are used to produce gasoline, 0.74 of fossil energy are used to produce corn-based ethanol, and only 0.2 units of fossil energy are used to produce cellulosic ethanol.

So it is much better than corn, definitely positive, probably somewhere around 5 times right now, but also lower than sugar cane.

3) How much land would it take?
In this University of Illinois report (.pdf) they were able to get 35 dry tons of Miscanthus per hectare which is about 15 tons/acre.

Iogen is able to turn a ton of straw into 80 to 85 gallons of ethanol. Other estimates have 100 gallons a ton.

So, 1 acre could yield 1,500 gallons of ethanol based on these figures. If you are assume that it is 5 times in its energy balance, and you wanted to make the whole thing use just its own energy, 1/5 of the energy would go to running the conversion and you would be left with only 4/5 or 1,200 gallons an acre. But, this may only be possible yield on the best land, so other land might yield considerably less.

4) Can it make a serious dent on oil imports and consumption?
According to the EIA, the US uses 420 million gallons of gasoline a day for motor vehicles. For a year that works out to 630 million gallons * 365 days = 230 bil gallons. Since ethanol has less energy per gallon it takes 1.5 gallons of ethanol for each gallon of gasoline.

To replace all that gasoline it would take 230 bil * 1.5 / 1,200 gallons/acre = 191 million acres. The US currently has 700 million acres in range and crop land, so this is in the realm of the possible.

The US has 39m acres in Conservation Reserve Program, which is previous farm land that farmers are being paid not to farm. If they could get the 15 tons/acre (might not be realistic as this was probably marginal farming land) this could account for 20% of US transportation fuel.

Vinod sees 900m tons of crop being harvestable from stovers, winter crops and converting soybeans to switchgrass. That would create 90 billion gallons of ethanol a year or about 40% of total transportation consumption.

If these numbers are feasible then it would make a serious impact.

5) Is it economical?
How do you answer this one? You are looking at two industries: agriculture and energy that are full of market distorting subsidies and non market externalities. So it really depends on what you put into your calculations. Should the subsidies that farmers are given to produce crops be added in? How about the subsidies given to oil producers? How about the tariff on importing ethanol from Brazil? What about the cost to society of carbon emissions? What about the cost of maintaining our large military in order to keep the oil lanes open? What about the funding of the new technologies to create ethanol more efficiently?

But, given everything that is currently in place, Vinod puts the break even at scale likely to be ~$35/barrel. Now that oil has hit $70/barrel this looks like it can compete.

From the Rocky Mountain Institute:

Such firms as Iogen and Novozymes have been developing enzymes, and "smart bugs," that can turn biomass such as corn residues (leaves, stalks, and cobs) into sugars that can then be converted into ethanol. Historically, the biggest cost component of this technology was the creation of enzymes. Earlier this year, though, in combination with the National Renewable Energy Laboratory, Novozymes announced a 30-fold reduction in the cost of enzyme production in laboratory trials. Expected benefits from this process include low energy requirements, high efficiency, and mild process conditions. A pilot plant exists in Ontario and another is planned in Hawai'i. The first commercial-scale enzymatic reduction hydrolosis plant is scheduled to be built and operational by Iogen within two years, producing ethanol at a targeted cost of $1.30 per gallon.

From Iogen:

Straw looks a lot like hay, but has fewer nutrients and therefore less value. A ton of baled straw goes for about $40 and yields 80 to 85 gallons of ethanol, making the price of the main raw material about 50 cents a gallon. By contrast, with oil at $70 for a 42-gallon barrel, the raw material for gasoline is about $1.67 a gallon. Iogen chose to use straw because the technology to bale it is available. Brian Foody, the president of the company, said other wastes could be used, including the entire corn plant, not just the kernels, which are used in existing ethanol plants. Even old newspapers are an option.

and another from Iogen:

Last December the bipartisan National Commission on Energy Policy released a report, Ending the Energy Stalemate , that analyzed the potentials of various alternative fuels, including both types of ethanol (which is just an industrial grade of alcohol). Only cellulosic ethanol got a decisive thumbs-up. By 2020, the commission predicts, its production cost could be less than 80 cents a gallon.

So $1.30 a gallon to maybe $.80 a gallon. That would be competitive if true (and who knows what assumptions, subsidies or externalities they are figuring in). With better enzymes for breaking down the cellulose and creating the ethanol and better crop yields, these prices could go down as well.

5) Other objections:

a) Shouldn't we just remove the $.54 a gallon tariff on Brazil ethanol and import it from them and other tropical nations where they can create ethanol more efficiently from sugar?

I think you can make a strong case for this. Of course you also have issues with them potentially cutting down rain forests in order to grow sugar cane. And if you think energy independence is important, this doesn't help either. From the NY Times:

But Brazilian officials and business executives say the ethanol industry would develop even faster if the United States did not levy a tax of 54 cents a gallon on all imports of Brazilian cane-based ethanol. The expansion of sugar production, for example, has come largely at the expense of pasture land, leading to worries that the grazing of cattle, another booming export product, could be shifted to the Amazon, encouraging greater deforestation.

Industry and government officials say such concerns are unwarranted. Sugar cane's expanding frontier is, they argue, an environmental plus, because it is putting largely abandoned or degraded pasture land back into production.
For each unit of energy expended to turn cane into ethanol, 8.3 times as much energy is created, compared with a maximum of 1.3 times for corn, according to scientists at the Center for Sugarcane Technology here and other Brazilian research institutes.

"There's no reason why we shouldn't be able to improve that ratio to 10 to 1," said Suani Teixeira Coelho, director of the National Center for Biomass at the University of São Paulo. "It's no miracle. Our energy balance is so favorable not just because we have high yields, but also because we don't use any fossil fuels to process the cane, which is not the case with corn."

From Newsweek (.pdf):

Southern countries growing suger cane, on the other hand, can get up to 5 times as much biofuel from each acre of land. "Without too much effort, producing ethanol from sugar cane in developing countries like Brazil and India could replace 10% of global gasoline fuel," says Lew Fulton, biofuels expert at the International Energy Agency. Malaysia, Indonesia and Australia are well positioned to join Brazil as global suppliers of sugar-cane ethanol.

b) Shouldn't we be using our land for food, or if not that returning it to nature?
As the population goes from 6 to 10 billion, should we really be setting aside vast chunks of land to create crops to fuel our vehicles? Vinod has this odd slide (p 26) where he points out that we could turn all of the land used to export crops to other countries and turn that into ethanol producing land. Instead of feeding those in the 3rd world we should feed our vehicles. I don't think I agree with that.

I think there is a case for this, but at the moment we have extra land that is not being used for agriculture that could be used to grow switchgrass.

c) Wouldn't it be even better to use solar panels and electric cars?
If we used solar panels, it would take about 1/50 the amount of space to capture the same amount of energy from the sun as we do with cellulose ethanol. This means we could use much less land and leave the rest for nature. On the other hand, maybe it is better to have 50 acres of switch grass vs. having 1 that is full of inorganic photovoltaic cells.

Another issue with solar is that it is much more expensive right now than ethanol. Battery technology has also not advanced far enough to have electric powered cars. Maybe in 10 years when new batteries become available, but for now ethanol will allow a car to travel much further on one tank than a battery pack will allow.

d) Would it be better to just burn the grass as electricity instead of turning it into ethanol?
Another interesting idea. But, once again if we are looking for a way to power our vehicles, we need better batteries before this is viable. On the other hand, if we are just trying to generate the most amount of usable energy per acre this might be a way to go.

In the U of I report they calculate:

If only 20% of the over 11 million hectares currently devoted to agriculture in the state were to produce Miscanthus, they could provide 145 TWh of electricity.

This works out to 1.6MWh/ton or 77MWh/ha or 31MWh/acre. If we compare this with our 1,200 net gallons/acre number from above, we get 1,200 gallons * 76,000 BTU/gallon = 91m BTU. There are 0.29307107 watt hours in a BTU, so that gets us to 26Mwh. So, from this very rough calculation we see that ethanol has slightly less energy than what could be produced in electricity. I would guess that you would actually get quite a bit more electricity than ethanol in terms of BTU. But, given the extra range and cheaper price of an internal combustion engine over a battery powered electric car, we are still going to need liquid fuels and this loss of energy is probably worth while.

e) Is it better than biodiesel or hydrogen?
I haven't been able to find any good comparisons between biodiesel and cellulose ethanol. So I don't know. It is possible to convert ethanol into biodiesel, so maybe it has more to do with the type of engine you are trying to run it on.

As for hydrogen, Vinod throws the smackdown on hydrogen on slides 52 and 53. Basically it would be a lot more expensive to use hydrogen and it is going to take at least 15 years to get there. It is not clear where you get the hydrogen from, how you store it, or how you transport it. It is going to take a major infrastructure change to sell hydrogen rather than a fluid. And on a well to wheels analysis is isn't clear that hydrogen is more efficient.

f) Other dissenters:
This article and discussion over at the oilddrum.com doesn't believe that the yields on switchgrass will be able to amount to those stated earlier. He questions whether switchgrass would really be much better than corn for generating corn in terms of price or energy yield.

David Pimentel1 and Tad W. Patzek state in this paper that cellulose ethanol takes 45% more fossil energy than it produces. Not really clear how they derived this number. Most of the loss is due to steam and electricity use. But, other sites say that part of the cellulose is burned to generate the steam and electricity, so it is not clear to me how they can state it loses 45% of its energy.

Overall, I think the case for cellulosic ethanol is pretty compelling. The key question is how well the actual yields compare to those estimated here, and how the costs work out is it ramps up. Definitely one to watch (and possibly invest) in the alternative fuel arena.