When it comes to biofuels we have choices. We can do it poorly, using short-run approaches with no potential to scale, poor trajectory, and adverse environmental impact. Or we can do it right, with sustainable, long-term solutions that can meet both our biofuel needs and our environmental needs.

We do need strong regulation to ensure against land-use abuses. I have suggested that each cellulosic facility be individually certified with a LEEDS-like “CLAW” rating, and that countries which allow environmentally sensitive lands to be encroached be disqualified from CLAW-rated fuel markets.

We think a good fuel has to meet the CLAW requirements:

C — COST below gasoline
L — low to no additional LAND use; benefits for using degraded land to restore biodiversity and organic material
A — AIR quality improvements, i.e. low carbon emissions
W — limited WATER use.

Cellulosic ethanol (and cellulosic biofuels at large) can meet these requirements.

Environmentally, cellulosic ethanol can reduce emissions on a per-mile driven basis by 75-85% with limited water usage for process and feedstock, as illustrated later. Range, Coskata, and others currently have small-scale pilots projecting 75% less water use than corn ethanol, with energy in/out ratio between 7-10 EROI (though we consider this a less important variable than carbon emissions per mile driven).

Sustainable land use

The question about biomass production that arises first is about land use: how much will we need? What will it take? Is it scalable? For conservatism, I assume CAFE standards in the U.S. per current law, though I expect by 2030 to have much higher CAFE and fleet standards (hopefully up near 54mpg or a 100% higher that 2007 averages), which will dramatically reduce the need for fuel an hence biomass. Yes, this would include lighter vehicles, more efficient engines, better aerodynamics, low-cost hybrids, and whatever else we can get the consumer to buy that increases mpg.

There are many approaches to production of feedstock for biofuels. To make a material impact in replacing gasoline, major feedstocks need to collectively produce more than a hundred billion gallons in the U.S. — preferably more than 150 billion gallons, to replace gasoline. Replacing gasoline and replacing diesel involve different technologies and markets. The focus here is principally on gasoline replacement in America’s cars and light trucks, though I do briefly touch upon diesel feedstocks.

I believe that a sustainable biofuel needs yields of at least 2,000 gallons (ethanol equivalent) per acre in the long run, to meet the world’s oil replacement needs on a manageable amount of land (with the exception of winter cover crops that use no additional lands). I believe, as estimated in my papers elsewhere, that 2,500 gallons of ethanol equivalent per acre annually is a reasonable assumption. (Assuming corn grain yields of 140 to 170 bushels/acre, typical of the mid-Western corn belt, and 2.8 gallons of ethanol from a bushel of corn — the range in ethanol production from corn is only 392 to 476 gallons/acre.)

Chemical and water inputs, and the effect on biodiversity, should be minimal, if any. Cost should be below that of oil. Feedstock production should not materially increase the land under annual cultivation or affect food security, but should enhance energy security, reduce poverty, and increase rural incomes. None of the “food/feed crop” based biofuels (corn or sugar based) or classic biodiesel sources (vegetable oils) come close to these targets.

Is such a fantasy possible? Yes! Part I covers sources of biomass, Part II will cover agronomy practices for yield, biodiversity, water, and chemical efficiency, and Part III discusses the reasonableness of yield assumptions that lead to 2,500 gallons per acre. My calculations later show that if we can increase engine and automobile efficiency materially at the same time, we will need no additional land for biofuels.

Renewable feedstocks

Currently there are two primary feedstocks for the production of renewable biofuels to replace gasoline: sugar from sugar cane (primarily used in Brazil) and starch from corn (the source of most U.S.-based ethanol). In Asia and Africa, tapioca, potatoes, and other starch crops are being used (sadly!). Amongst feedstocks, there has been significant discussion regarding both corn stalks and wheat straw. I am not a huge fan of wheat straw or corn stalks, though they are possibilities.

In my opinion, cellulosic ethanol plants need to reach production levels of 100m gallons per year per plant to achieve economies of scale. (Expensive fuels don’t sell! A local conversion plant near the field and distributed supply would be ideal, and I continue to investigate technologies that might make this possible). That would dictate feedstock needs of around 1,000,000 tons per year, per plant. In the short and medium term, at biomass yields of 10 tons/acre (by 2030 we expect about 20-25 tons/acre), 100,000 acres of land would be needed per cellulosic ethanol plant, or 40,000 acres by 2030. With yields of approximately 2 tons/acre, the usage of either corn stalk or wheat straw would effectively quintuple land usage and substantially increase transportation distances and costs — hence my skepticism.

In addition, there is value to plowing corn stalks and wheat straw under to improve the soil. Winter cover crops like legumes and winter rye (no biomass-optimized winter cover crops have been developed, but grasses are a good candidate), grown on row crop lands during their idle period during winters, can yield 3-5 tons/acre with no additional land usage and may actually improve land ecology where row crops are grown anyway. To quote Prof. Bransby, a renowned agronomist from Auburn University, in a personal communication:

Regarding water and fertilizer needs of cover crops: The answer is that no irrigation is needed, and fertilizer needs are about 30% of the fertilizer requirements of corn. Also, there are multiple benefits from cover crop/traditional crop rotations (compared to traditional crops with no cover crops), including better soil protection/less soil erosion, improved soil organic matter, better water holding capacity, suppression of crop pests, etc. Provided this is done with conservation tillage practices, there should be no serious negative environmental impacts. …

It is reasonable to assume that winter cover crops can be grown on the same land that our summer traditional crops are grown, and summer cover crops can be grown on land where traditional winter crops (mainly winter wheat) are grown. As far as I know, most of this land is currently idle/fallow at the time when these cover crops would be grown. From the USDA National Agricultural Statistics website the 2007 acreage (in millions) for our major traditional crops is as follows: corn, 93; soybeans, 63; cotton, 11; sorghum, 8; winter wheat, 44; Total = 219. At a modest estimate of 3 tons/acre/year, this would provide 657 million tons of biomass annually. With research and genetic improvement, I believe the yield could be increased to 5 tons/acre within 10 years, for a total of 1.1 billion tons/year. Acreage for all annual crops is 317 million. For various reasons, it is unrealistic to assume that 100% of land in traditional crops could be planted to cover crops to produce biomass. Maybe 70%?

To be conservative, in my estimates I have scaled that down to 50% of forest waste and annual land for winter cover crops. My estimates suggest that any feedstock transportation beyond about 50-75 miles (preferably under 30 miles) will reduce its competitiveness, unless the crop is essentially free (like winter cover crops), in which case a maximum 100 mile radius might make sense.

Energy crops and winter cover crops will reduce the need of substantial transport infrastructure for biomass — and answer critics’ questions about infrastructure: If these plants were distributed around the country it would substantially reduce need for infrastructure. If most of the biofuels are not concentrated in the Midwest, smaller pipelines will suffice. Biomass crops will be widely distributed and will minimize the need for infrastructure.

To compete with $50/barrel oil (which we are unlikely to see again without significant reduction in demand), I believe feedstock cost based on current conversion efficiencies (which are subject to improvement), delivered to the factory, has to be below $50/ton of dry biomass (plus or minus 25% depending upon feedstock type). As such, I limit (in my estimates) potential incremental land using feedstocks to crops that yield over 10 tons/acre in the mid-term — effectively, “energy crops." I should also note that a number of “biomass concentration” approaches are being investigated that may ultimately reduce biomass transportation costs even further, but are currently in early research stages. For example, one approach is the production of “bio-oil” at small-scale localized biomass pyrolysis units. This bio-oil can then be transported to a centralized facility for conversion and up-grading to a liquid fuel or used as-is for applications like home heating oil. (See: Kior).

Yield Density
Source: David Bransby & Ceres (PDF)

As discussed earlier, I estimate feedstock costs need to be under $50 per ton delivered within the next decade (and lower in the short run) to compete with $50/barrel oil. Switchgrass and miscanthus-like grasses (C4 photosynthetic grasses) and certain trees are the most likely feedstocks to provide our liquid fuel requirements in the long run. Tree crops developed for the paper pulp business will also make for good crops. Many client paper mills are going out of business, and these communities are crying for local economic stimulus and jobs. Given these prices, biomass has the potential to substantially increase farm income and reduce the need for farm subsidies.

Waste feedstocks?

While I believe that energy crops will meet most of our feedstock needs, I have invested time and money in the potential of waste feedstocks, as I think they can make a material impact and reduce the above cited biomass needs by an additional 10-20% or more. Promising waste feedstocks include municipal sewage, even municipal solid waste — the paper, wood, construction waste, even lawn clippings that are brought to a landfill. Something that has been a problem (especially with disposal) may soon become an opportunity!

There is sufficient municipal waste to produce tens of billions of gallons of ethanol. The waste is available in large enough quantities (in most major cities) to justify waste-specific plants and actually has a negative cost (usually a tipping fee). I’m also intrigued by the possibility of using farm organic waste. One of my favorites is a proposal (LanzaTech) to take all the waste carbon monoxide from steel mill flue gases (already collected and piped, available to go into a process) to make ethanol. There is enough carbon monoxide coming out of today’s steel mills to produce over fifty billion gallons of ethanol! Forest waste could be treated similarly, and is discussed below.


Now to the numbers. How much biomass can convert to biofuels without subsuming other uses for land and biomass? More than enough! There are four principal sources of biomass and biofuels I consider

  1. energy crops on agricultural land and timberlands using crop rotation schemes that improve traditional row crop agriculture and recover previously degraded lands
  2. winter cover crops grown on current annual crop lands, using the land during the winter season when it is generally dormant (while improving land ecology)
  3. excess forest product that is currently unused (about 225 million tons, according to the U.S. Department of Energy), and
  4. organic municipal waste, industrial waste and municipal sewage.

For the U.S., the world’s most oil intensive economy, my calculations show that a small dose of vision, two decades of agricultural development, and process technology that is in pilots today, with less than 5% of our annual crop and timberlands, could more than supply our biofuel needs to replace all of our light-vehicle gasoline usage by 2030. The table below shows one of many possible scenarios:

Cellulosic Ethanol Production Estimates (Billions of Gals) Tons (M) Biomass Needed Winter Crop Biomass Acres (Millions of Ac.) Biomass Yield (tons/ac) Forestry Tons Excess Biomass (Millions) Biomass Yield (tons/ac) Biomass needed from dedicated crop/timber land (millions tons) Crop/timber land needed (millions of acres) Crop/timber land needed – assuming yields at 75% of projected levels
2015 3.7 36.5 4.1 3.4 22.4 10.9 0 0.0 0.0
2020 27.9 259 42.9 3.8 81.3 15.4 14 0.9 1.3
2025 94.8 861 150.0 4.2 98.9 20.5 132 6.4 8.6
2030 166.8 1,517 150.0 4.6 113.0 24.5 708 28.9 38.5
1. Cellulosic production is assumed initially to represent ethanol demand not met by corn — by 2030, it is equal to the numbers of gallons of ethanol equivalent needed to replace all light-vehicle gasoline usage. I assume the mileage discount for ethanol vs. gasoline declines from 25% in 2020 to 15% in 2030.
2. Biomass from waste production is not explicitly modeled here — I believe this has the potential to meet 10-20% of biomass need.
3. Current CAFE laws are assumed to reduce gasoline demand. Additional ICE engine efficiency/higher CAFE could substitute for higher efficiency on ethanol assumed above.
4. Yield projections (tons per acre) are based on fertile land. The usage of degraded land will result in lower yields. Yields projections (gallons/ton) go from 90 tons per gallons today to 110 tons per gallon in 2030.
5. I assume that the primary source of dedicated land for energy crops will be cropland, but commercial reduction in today’s forest resource usage (i.e., more paper mill closures) could be offset by using it for biofuels, while also reducing the amount of cropland needed.
6. If winter over crops do not provide any biomass, I project energy crop usage of 57M acres in 2030 — furthermore, if yields are 18 tons/ac instead of 24, I project total land use at 76M acres.
7. I believe that replacing diesel may require an additional 18m acres in cropland, but it is not included here.

While my projections above are based on my most likely scenario, I’d like to lay out a more optimistic vision. In this scenario (as of 2030), I project usage of 70% of annual crop land for winter over crops (approximately 225M acres instead of 150M acres), and 70% of the excess forestry biomass (170M dry tons). Early experimental data has shown that other biofuels may produce yields equivalent to 150 gallons of ethanol equivalent biofuels per ton (as opposed to the 110 projected in the table above), long before 2030 (based on data disclosed confidentially to me).

In this (optimistic) scenario, all of our light-vehicle transportation needs would be met without using any devoted energy cropland! Going further, about 20% of our corn production today is used for ethanol — in this scenario, the 20% of 90M acres (18M acres) devoted to corn for ethanol today could be “released” and utilized for other purposes! And that’s only the beginning — one of our investments is working to improve the mileage efficiency of the standard ICE (Internal Combustion Engine) by 50-100% for ethanol and gasoline, dramatically reducing biomass needs. Increased CAFE standards will help too. Additional degraded land can be recovered if our 10 year by 10 year biomass crop rotation scheme is followed (described in Part II).

In combination with the other factors listed above, we are confident that biomass needs will not be a limiting factor by 2030.

It is worth noting that unless we dramatically reduce carbon emissions and stop global warming, millions of acres of land will be “dislocated” from its current uses and must be figured into the “net land use” equation. Though many technologies will contribute to displacing oil-based fuels, I don’t believe any other technology is pragmatically likely to achieve as large a reduction in emissions from transportation fuels.

A recent Booz Allen Hamilton study noted that worldwide, there is up an additional 6 billion acres of rain-fed land available for agricultural production. (Clearly there would be opportunity cost associated with this land use.) Farmers will make more money. We will sell less subsidized crops — an issue over which the Doha round of trade talks have broken down, as developing countries demand fewer agricultural subsidies in the west. (Organizations like Oxfam now oppose the dumping of subsidized US food crop in Africa, where agriculture is often the only means of income generation.) We will import less oil and export fewer crops, allowing farmers in poor countries to make a living (helping reduce third world poverty) while we in the U.S. improve our trade balance.