Alternate Energy: Ethanol
This month I am going to write about alternate energy. I am a chemical engineer by profession will be exploring alternate energy from this perspective. I will explore one particular technology, the production of grain alcohol (ethanol) for use (blended with gasoline) as a motor fuel. I will focus particularly on ethanol from corn because it is a fully-developed mature process that produces automotive fuel on a large scale that is currently used in millions of vehicles. The U.S. produced about 4.5 billion gallons of fuel ethanol in 2005, equivalent about 2% of gasoline consumption, after adjustment for the lower energy content of ethanol relative to gasoline.1
Figure 1 shows a schematic for a state-of-the art ethanol from corn process.2 Corn is milled, mixed with hot water recycled from the DDG recovery unit (see below), enzyme (a-amylase) is added and the mixture heated to 190° F. The amylase breaks down the large starch molecules into smaller ones, converting the corn mash into a syrup. Urea (a yeast nutrient) is added and the mixture sterilized by heating to 230° F for 20 minutes.
The solution is then cooled to 140° F and another enzyme (glucoamylase) is added. This enzyme further breaks down the starch to simple sugars, mostly glucose. This solution is cooled and yeast added. The mixture is fermented at 93° F for two days, during which the sugars from the starch are converted into ethanol (about 12% by volume) and carbon dioxide (CO2). The carbon dioxide is passed through water (scrubbed) to remove ethanol vapors and vented to the atmosphere. The ethanol-containing scrubber water is recycled back into the beer column (see below).
After the fermentation is complete, the fermentation beer is sent to the beer column, where ethanol is recovered. The beer is heated, driving off dissolved CO2 and some ethanol vapor. This CO2 and ethanol vapor stream is also passed though the scrubber and the ethanol scrubbed from the CO2 is recycled back into the beer column. After this degassing step, the beer is distilled in the beer column producing a vapor phase containing 41% ethanol and a distillation "bottoms" phase containing water and beer solids, but less than 0.1% ethanol. The vapor phase is sent into the ethanol purification unit and the bottoms into the DGG recovery unit.
The ethanol purification unit consists of another distillation column and a molecular sieve dryer. The distillation column produces a 96% ethanol stream and a >99.9% water stream. The water stream is fed back to the scrubber and into the liquefaction reactor. The ethanol stream is fed into the molecular sieve unit, where it is dried to >99.9% ethanol. This dry ethanol stream is mixed with gasoline denaturant to produce the 95:5 ethanol-gasoline mixture that is fuel ethanol.
The DGG recovery unit consists of a six-effect vacuum evaporator and a decanter centrifuge. Two product steams and a side stream are produced by this unit. The products are a slurry containing insoluble beer solids at a 35% concentration that is discharged from the centrifuge and a concentrated syrup containing 55% dissolved solids that is discharged from the evaporator. The side stream is the hot water stream used to make up the fermentation mix in the liquefaction reactor.
The two product streams are fed into a gas-fired rotary drum dryer and dried to 9% moisture. The product, called distillers' dried grains with solubles (DDG), is rich in protein (27%) and fat (9%).2 It is sold as a high-quality animal feed. The process is essentially zero discharge, all the water is recycled back into the process. The overall mass balance is one bushel of corn (47.6 lbs. dry weight) yields 15.8 lb. of carbon dioxide, 16.8 lb. of DDG and 2.6 gallons of fuel ethanol (16.5 lb. of dry ethanol).
Figure 1. Process Schematic for Ethanol from Corn Process2
The energy balance for ethanol production is a matter of some controversy. Typical energy input for a state-of-the-art process like the one I have described is about 34,000 BTU per gallon of fuel ethanol.3 Average ethanol input for actual plants in operation, most of which are older and not state of the art, is considerable higher at about 52,000 BTU/gal.3 Many plants are old or small and do not use molecular sieve dryers for ethanol purification, relying on a third distillation step that consumes more energy than the dryers. Older plants also do not feature all the energy recycling and can require 80,000 or more BTU to produce a gallon of fuel ethanol. As the energy content of ethanol is about 84,000 BTU/gallon, these old plants sometime consumed more energy to make the ethanol than what was contained in the product. Even the state-of-the art process requires energy equal to about 40% of the energy in the product to manufacture fuel ethanol. The effect of plant efficiency has affected the conclusions of studies looking at the total energy efficiency of corn to ethanol conversion.
Cornell ecologist David Pimentel is one of the harshest critics of the rationale for producing ethanol from corn. He argues that the total energy required to produce a gallon of fuel ethanol is 131,000 BTU, much more than the energy contained in the ethanol produced.4 In his analysis he considers the energy required to grow the corn used to produce ethanol as well as the energy used to produce the ethanol from the corn. Pimentel uses a value of 74,000 BTU/gal as the energy input for producing ethanol from corn.3 As described above, state-of-the-art plants only use 46% of this amount of energy to produce ethanol and even the average plant today does better than this.3 Older plants have efficiencies similar to those Pimentel employs and it is likely he employed data corresponding to these plants in his analysis.
Pimentel also considers the energy consumed in growing the corn. In a 2001 report Pimentel reports per acre energy requirements of 140 gallons of fuel (18.6 million BTU) to produce 127 bushels of corn yielding 328 gallons of ethanol.4 He also reports a total energy requirement (including the energy to grow the corn and to process it to ethanol) of 131,000 BTUs per gallon of ethanol.4 A 1995 paper by Lorentz and Morris3 gives more of details underlying Pimentel's estimate. Pimentel estimates 16.4 million BTU per acre are required to produce 115 bushels of corn yielding 288 gallons of fuel ethanol, for an energy input of 57,000 BTU/gal, which when summed with the 74,000 BTU estimate for ethanol production from corn, yields the 131,000 BTU figure. Since the 131,000 BTU/gal figure in 2001 is the same as in 1995, it is reasonable to assume that Pimentel's 2001 paper does not reflect a re-analysis of process energetics. Thus, although he acknowledges a 13% increase in corn yield from 115 bushels/acre to 127 bushels/acre, he sticks to his 57,000 BTU/gal estimate of energy required to grow corn.
If we focus simply on Pimentel's worst-case 57,000 BTU/gal figure for the energy used to grow the corn, we note that the ethanol is not the only valuable product produced by the corn-to-ethanol process. Recall that DDG (6.4 lbs. per gallon of fuel ethanol) is also produced. This animal feed can substitute for the use of grain and contributes an energy credit on this basis. Pimentel gives a figure of 32,000 BTU for the energy value of the DDG produced along with one gallon of fuel ethanol. This figure should be subtracted from 57,000 BTU energy cost of the corn to obtain a net energy cost of 25,000 BTU per gallon of ethanol from the corn. When this figure is added to the state-of-the-art value of 34,000 BTU/gallon for ethanol production from the corn, we obtain 59,000 BTU/gal of energy consumed to produce ethanol using Pimentel's estimates for agricultural energy efficiency with the efficiency achievable in a modern plant. If the efficiency of the average plant in operation is used, the total energy required rises to 77,000 BTU/gal. Thus, even using Pimentel's pessimistic figures for corn growing, ethanol from corn does produce net energy. Other workers have provided estimates for the energy requirements for growing corn that are lower than Pimentel's,5 but in every case, the amount of energy required to produce ethanol (49,000-83,000 BTU/gal after by product credit)5 is a fairly large fraction of the 84,000 BTU/gal energy content of the ethanol product.
This question of efficiency is important because it affects the basic economics of ethanol from corn. Right now, the cost of producing ethanol is less than the market value of the ethanol produced. A government subsidy makes ethanol production profitable. Thus, we cannot point to the fact that ethanol is being made commercially as evidence that the practice is economically sound. Without the subsidy, the industry would collapse. This is the reason for the academic energy efficiencies discussed above.
The consensus from the efficiency debate is that ethanol does contain more energy than is required to produce it, at least if the ethanol is made using a modern plant. That is, producing ethanol with modern processes consumes less fossil fuel than the ethanol displaces, meaning that if the price of fossil fuels rise the price of ethanol will rise more slowly. At some point the price of fossil fuel should reach a point where ethanol is cheaper.
A 1999 cost estimate2 gave a value of $0.88 per gallon for the cost of manufacturing ethanol from corn. Of this cost, $0.68 was for the corn and $0.19 were for the utilities and the gasoline denaturant. A credit of $0.29 was given for the value of the DDG. Rising fossil fuel prices will have a proportional impact on utilities cost and the cost of gasoline; doubling the cost of fuel will double the cost of utilities and denaturant from $0.19 to $0.38. Pimentel's 140 gallons of fossil fuel to produce 328 gallons of ethanol implies an ethanol cost fuel cost sensitivity of 0.43. That is, if the cost of fuel were to double, the cost of corn would increase by 43% from $0.68 to $0.97. Because ethanol has 2/3 the energy content of ethanol of gasoline, the cost is multiplied by 1.5 to put it on a gasoline-equivalent basis, which allows a direct comparison to the price of gasoline. Thus the $0.88/gal cost given in the 1999 study is equivalent to a gasoline cost of $1.32 per gallon.
Fuel costs in 1999 were about $1/gal, while the corresponding ethanol cost was $1.32. As fuel costs rise, the cost of ethanol should rise as shown in Figure 2 by the red line. At a fuel cost of about $2.20/gallon (in 1999 dollars) ethanol cost should be about the same as gasoline on an energy-equivalent basis. At fuel prices above this level (corresponding to about $2.65 in today's money) ethanol produced from state-of-the-art plants, like the one costed in the 1999 study, should be profitable without a subsidy.
For the average plant in operation, the utilities cost in the 1999 study was increased by 63% to reflect their greater energy consumption. This situation is shown in Figure 2 as the blue line. Break even is seen at about $4.70 (in 1999 dollars). Thus, if ethanol proponents are correct and the plants and farming practices are as efficient as ethanol boosters say, we should expect fuel ethanol production to become profitable without a subsidy when gasoline prices reach the $3-5 range. On the other hand, if critics like Pimentel are correct, then the price sensitivity of ethanol to fuel costs is greater than one and ethanol will always be more expensive than gasoline no matter how high gas prices rise.
Figure 2. Projected impact of rising fuel prices on ethanol costs
Assuming Figure 2 is approximately correct and peak oil leads to steadily rising fuel prices, ethanol from corn will become economic and producers will be motivated to increase production capacity. Already, something like 18% of the corn crop is being used to produce fuel that meets only 2% of transportation needs. Were the entire crop used for ethanol, it would only provide about 11% of current needs. Long before this level of production were reached, corn demand by ethanol manufacturers would start to drive up corn prices.
Approximately three-fifths of woody plant tissues is cellulose and hemicellulose, both of which consist of sugars than can, in principal, be converted into ethanol. Cellulose is a long chain of glucose sugars like starch, except it has a different structure. Enzymes that break down cellulose (cellulases) were not available commercially twenty-five years ago when work on alternate energy first began in response to the 1970's oil shocks. Amylases have been used commercially for making beverage ethanol from grains for centuries, which is one reason the existing ethanol fuel industry in this country uses a foodstuff like corn, as opposed to lower value woody materials like waste paper. Today commercial cellulases are available.
Hemicelluose is a large molecule made up of mostly five-carbon sugars (glucose has six carbons). It is readily broken down by a short exposure to dilute acid at an elevated temperature. The five carbon sugars are harder to ferment. In fact twenty-five years ago no yeasts that fermented five-carbon sugars to ethanol were known. In the 1980's a number of yeasts that did so were discovered; I did my thesis work in xylose fermentation with one of them6 (xylose is the most abundant five carbon sugar). Today genetically engineered versions of yeast and bacteria that ferment both five and six carbon sugars have been developed.
Processes that convert woody plants into ethanol have been developed and tested at the laboratory and pilot scale. At least one company has demonstrated ethanol production from woody materials at industrial scale.7 Should gas prices reach the $5 level and ethanol fuel become profitable without a subsidy, US corn resources will soon be fully tapped. If half the current crop were to be used for ethanol production, it would displace only about 6% of current gasoline consumption. The USDA estimates that approximately 300 million tons of woody biomass is potentially available today, with no changes in crop cultivation or forest management practices.8 This material would be able to produce fuel ethanol displacing another 11% of current gasoline use. Thus, ethanol has the potential for replacing about one-sixth of current gasoline consumption. With realistic yield increases and deliberate management of agriculture and forestry to maximize production, the USDA estimates the available resource could more than triple to a billion tons by mid-century,8 enough to displace 36% of current consumption.
These developments are not going to occur with gasoline prices remaining at current levels. Last May I presented a crude model that suggested that a price rise to $10/gallon in today's money would be accompanied by more than a 70% reduction in fuel consumption.9 This suggests that the volume of gasoline consumption in a world where ethanol is economic might be a third of today's volume. In this case, corn will be sufficient to displace about a fifth of future gasoline use and current wood resources another third. By mid-century, all gasoline could be replaced by ethanol, if necessary.
Ethanol is not the only viable liquid fuel option. Pyrolysis oil is another option as is hydrogen produced from coal and water, or from water by nuclear-generated electricity. It is clear that we will not run out of fuel for our cars. It is equally clear that fuel will be considerably more expensive than it is today. Any replacement for petroleum will require a larger drain on our other principal energy sources: coal, natural gas and nuclear. Recall that although ethanol production does produce a net energy gain, ethanol production still requires a large input from the other energy sources that gasoline production does not. Since natural gas production may also experience a peak sometime after that for petroleum, energy production will require more intensive use of coal and nuclear energy, both of which require more capital to produce cleanly and safely. Thus, not only gasoline, but all forms of energy will become much more expensive in the future.
1. Zullo, Luca, "Fuel Ethanol Technology and Markets Beyond the Renewable Fuel Standard", Cargill Inc., (http://biomass.ucdavis.edu/pages/forum/3rd/LZullo.pdf).
2. McAloon, Andrew, Frank Taylor, Winnie Yee Kelly Ibsen, and Robert Wooley., Determining the Cost of Producing Ethanol from Corn Starch and Lignocellulosic Feedstocks, National Renewable Energy Laboratory, October 2000 (http://www.public.iastate.edu/~brummer/papers/Ethanol production.pdf).
3. Lorentz, David and David Morris, "How Much Energy Does It Take to Make a Gallon of Ethanol?", Institute for Local Self Reliance, August 1995 (http://www.greatchange.org/bb-alcohol2.html).
4. Segelken, Roger, "Ethanol fuel from corn faulted as 'unsustainable subsidized food burning' in analysis by Cornell scientist" Cornell University Press Release, Aug. 6, 2001 (http://www.news.cornell.edu/releases/Aug01/corn-basedethanol.hrs.html).
5. Shapouri, Hosein, James A. Duffield and Michael S. Graboski, Estimating the Net Energy Balance of Corn Ethanol, USDA Agricultural Economic Report Number 721, July 1995 (http://www.ethanol-gec.org/corn_eth.htm).
6. Alexander, Michael A., Thomas W. Chapman and Thomas W. Jeffries, (1988) "Continuous Xylose Fermentation by Candida shehatae in a Two-Stage Reactor", Applied Biochemistry and Biotechnology 17: 221-229. (http://www.rmmn.org/documnts/pdf1988/alexa88a.pdf).
7. Brown, Stuart, F., Biorefinery Breakthrough, Fortune, February 13, 2006. (http://www.iogen.ca/news_events/iogen_news/2006_02_13_Biorefinery_Breakthrough.pdf).
8. Perlack, R. D, L. L Wright, A. F. Turhollow, R. L. Graham, B. J. Stocks, and D. C. Erbach, Biomass as Feedstocks for a Bioenergy and Bioproducts Industry: The Technical Feasiability of a Billion-Ton Annual Supply, Oak Ridge National Laboratoy, (http://feedstockreview.ornl.gov/pdf/billion_ton_vision.pdf).
9. Alexander, Michael A., "Oil Prices, the Kondratiev Cycle and Peak Oil", Safehaven, April 2006.