Ethanol from Corn: Neither Renewable nor Reliable

Though the process of producing ethanol was until recently widely believed to use more energy than it created, farming and ethanol conversion practices have improved, causing more to argue that it is a sustainable and more secure alternative to gasoline. For instance, a particularly optimistic study conducted by the USDA – one widely cited by ethanol proponents – estimates that for every unit of energy used to produce ethanol from corn, 1.34 units are created. The study, like others preceding it, estimates this “net-energy value” by assuming the energy used in the production process is derived from the typical mix of energy sources like coal, natural gas, and diesel. Though this approach is consistent with current production practices, it is misleading from a policy-analysis perspective because it ignores the fact that ethanol is promoted as a renewable source of fuel. For instance, the Renewable Fuels Act of 2001 provided large subsides with the aim of boosting the production of domestic ethanol to more than 28.4 billion liters by 2012. More recently, a bi-partisan group of Midwestern senators introduced the BioFuels Security Act that proposes a new renewable fuels standard that calls for the production of 227 billion liters of ethanol and biodiesel by the year 2030. After signing the Energy Policy Act of 2005, which provided more subsidies for ethanol producers, President Bush said: “The bill includes a flexible, cost-effective renewable fuel standard that will double the amount of ethanol and biodiesel in our fuel supply over the next seven years.” Clearly, the promise of a “renewable” automobile fuel is a major driving force behind support for ethanol. Nonetheless, there has been little or no discussion of how much gasoline ethanol could displace if it were produced in a sustainable fashion. Furthermore, though reliability is obviously a central component of an energy security policy, policy makers and researchers have paid little attention to the likelihood of an ethanol supply disruption relative to that of petroleum.

HOW SUSTAINABLE IS ETHANOL?

If the objective of promoting ethanol is to rely more on domestic fossil fuels, then perhaps it would be more efficient to directly use natural gas and liquefied coal to power cars – compatible vehicles have been operating for years on US roadways and we would not disrupt the food supply. If, however, the objective is to power cars with a sustainable, domestically produced fuel – the objective publicly promoted by the U.S. government – then the modeling approach used to analyze the policy should assume ethanol is produced in a sustainable fashion. A simply way to do this is to create a balanced energy model where a portion of the ethanol output is fed-back into the production and distribution process to make up for the energy used to farm, distill and transport the ethanol. It is important to note that this approach is not meant to be literal; for example, ethanol would not typically be burned locally at distillation plant to power the process. Nonetheless, this approach simplifies the traditional analysis that mixes renewable and fossil fuel sources, making judging the relative merit of ethanol as a renewable energy source ambiguous. Therefore, below, we convert energy contribution from fossil fuel sources to equivalent amounts of ethanol, and subtract the ethanol from the gross production values.

Virtually all ethanol produced in the U.S. comes from corn. Farmers grow the corn that converts solar energy into chemical energy. Then, the harvested corn is transported to a distillation plant where it is converted into ethanol. Finally, the ethanol is trucked to fueling stations. Table 1 shows the USDA’s estimate of the energy required at each stage of this process to produce one litter of ethanol. In particular, approximately 6.0 mega joules per liter (MJ / L) are required at the production stage, 0.6 MJ / L are required to transport the corn to the ethanol plant, 14.4 MJ / L are required to operate the ethanol plant, and 0.4 MJ / L are required to truck the ethanol to fueling stations. A liter of ethanol contains approximately 23.6 mega joules (MJ) of energy. The USDA study adds an “energy credit” of about 3.8 mega joules (MJ) to account for energy contained in co-products. The logic behind the energy credit is that co-products (mostly cattle-feed) have economic value and would require energy to produce if they weren’t produced during the ethanol process. This is despite the fact that if enough ethanol were produced to actually displace significant amounts of gasoline, the supply of co-product would exceed the demand and thus energy would be required to dispose of the excess co-product. Nonetheless, the inclusion of this energy credit brings the gross energy output of ethanol to near 27.4 MJ / L. After subtracting the 21.4 MJ required to power the process, the net energy remaining for automobile fuel is approximately 6 MJ/L, or, put another way, for every liter of ethanol produced 0.256 liters could be delivered to fueling stations.

IS THERE ENOUGH CORN?

Using the net-energy yield reported in Table 1, we can calculate how much corn would be required to displace just 15% of our gasoline consumption. The estimate requires some assumptions regarding how much corn the country can produce. We assume that the number of metric tons of corn harvested per hectare is 9,400, equal to the 2006 average and the highest level on record. Further, we assume the number of hectares harvested is approximately 30,400,000. This is equal to the 2005 harvest; the second highest level on record. These assumptions imply a total harvest of 28,450,000 metric tons. Finally, consistent with the USDA study, we assume that a metric ton of corn produces 4,000 liters of ethanol.

Table 2 reports the percentage of the all-time-high harvest in the U.S. that would have to be devoted to ethanol production in order to displace 15% of our current gasoline consumption. Based on the Bureau of Transportation Statistics, 15% of our annual gasoline consumption is over 98.6 billion liters. Since, according to the Department of Energy, flex-fuel vehicles typically get about 20-30% fewer miles per liter when fueled with E85, this means that a minimum of 123.3 billion liters of ethanol would be required, assuming a 20% efficiency loss. According to the USDA study, the net energy (including their energy credit) used to farm, distill, and transport one liter of ethanol is near 17.6 MJ. A liter of ethanol provides a heating value of about 23.6 MJ, so we will need to withhold 0.744 L of ethanol to cover the production energy for every full liter produced. That leaves 0.256 liters for sale to the customer. The USDA study assumed the ethanol plant can yield 10.14 liters of ethanol for every bushel of corn. This converts to a yield of 2.5 kg/L where one bushel weighs 25kg. Since only 0.256 liters is left for sale out of every liter produced, we will actually need to farm 2.5/0.256 = 9.77kg of corn for every liter sold to the customer. Or inverting this, we will net 0.103 liters of ethanol for every kg of corn harvested. Consequently, 1,203 million tons (123 billion liters ÷ 103 L/t) of corn, or over 423% of the all-time high harvest, would be needed to displace 15% of our gasoline consumption. In fact, if we devoted 100% of all corn to producing ethanol, we could displace only about 3.5% of current gasoline consumption.

IS ETHANOL RELIABLE?

Data from the National Agricultural Statistics Service show that since 1960 total corn harvests have increased from about 102 to 267 billion kg. Over the same period, the total number of squared meters harvested has fluctuated around 275 billion, meaning that production gains are almost entirely explained by yield increases. However, researchers have observed that the year-to-year percentage gain in yields has steadily declined over the same period. The rate peaked at between 3% and 5% in the early 1960’s and was less than 1.5% in 2001 – a growth rate that is not expected to even keep up with food demand. Researchers predict that even under the best-case global warming scenario, corn yields are likely to decline by 22% in the short-run. What is more worrisome for an energy security policy that would rely to some extent on a reliable supply of corn, is that researchers believe U.S. corn yield variability is escalating, and the most productive farmers face a higher risk for catastrophic losses due to increased sensitivity to weather conditions.

The point of this discussion is to emphasize that there is little reason to expect corn yield variability to decline. If we assume it will stay constant, we can use historical data to estimate what sorts of ethanol disruptions we can expect in the future. We can then compare corn yield variability to variability in oil imports to see which is more reliable. The period we consider is 1960 through 2005 – a period that included, among other oil shocks, the Six-Day War, the Arab oil embargo, the Iranian revolution, and the outbreak of the Iran-Iraq War. The first step is identifying which distribution best fits the empirical data so we can calculate the standard deviation – a common measure of variability. Using observations for the annual change in corn production and oil imports we use the Kolmogorov-Smirnov test to rank the fit of alternative distributions. The Kolmogorov-Smirnov test is a widely used statistical method used to identify which distribution best fits a set of empirical observations. Table 3 reports the results. The distribution that best fits the corn data is the logistic with mean 3.3% and a standard deviation of 11.9%. The distribution that best fits the oil data is a logistic with mean 5% and a standard deviation of 6.8%. The 90% confidence intervals suggests that in 1 out of every 20 years we can expect corn yields to decline by 31.8%, while we can expect oil imports to decline by 14.9%. Thus, based on history, by displacing gasoline with ethanol we exchange geo-political risk with yield risk and history suggests that yield risk is about twice as high.

A WEAKER SUPPLY RESPONSE?

Relying on ethanol exposes the economy to an entirely new risk – an undesirable link between ethanol supply disruptions and ethanol demand shocks created by their joint dependency on weather. In the case of gasoline, there is no obvious link. For example, during a particularly hot and dry summer the demand and price for gasoline increases as we drive longer distances to escape the heat, spend more time on congested roads, and use our air-conditioning more often. But, the hot weather does not increase the cost of producing gasoline, so increases in the price of gas have an unambiguously positive impact on the supply of gas. The relationship is illustrated in Figure 1. D₀ and S_1,A are, respectively, the demand and supply curves for gasoline. The two curves conceptually illustrate that the demand for gasoline decreases and the supply of it increases as price increases and vice versa. The intersection of the two curves indicates the price where producers are willing to supply the same quantity that is demanded – the market equilibrium. Suppose Q₀ and P₀ are the equilibrium quantity and price of gasoline, respectively, before, say, a heat wave. In response to a heat wave, the demand for gasoline shifts outwards to D₁. The market supply curve, which depends on the marginal cost of producing gasoline, does not shift since the marginal cost is not affected by the heat wave. The result is the new equilibrium (P_1,A, Q_1,A). In the case of ethanol, as with gasoline, a heat wave shifts the demand curve out to D₁. But, because corn yields are especially sensitive to rainfall shortages during July and high-temperatures during August, the heat wave also shifts the supply curve back as lower corn yields, or increased input costs, increase the marginal cost of producing ethanol. The result of the correlation between demand shocks and supply shortages is to weaken the supply response relative to that of gasoline. For example, a supply-curve shift to S_1,B, increases the equilibrium price, relative to the case where marginal costs are not affected by weather, from P_1,A to P_1,B and reduces the equilibrium quantity from Q_1,A to Q_1,B. The actual strength of the weather-created link between fuel demand shocks and the price of corn is unknown, but the relationship should be well understood before framing an energy security policy around ethanol.

Summary

When we assume the ethanol production process is fully renewable, it would take all the corn in the country to displace about 3.5% of our gasoline consumption – only slightly more than we could displace by making sure drivers’ tires were inflated properly. There are also ethical considerations. In particular, the United States is responsible for over 40% of the world’s corn supply and 70% of total global exports. Even small diversions of corn supplies to ethanol could have dramatic implications for the world’s poor, especially considering that researchers believe that food production will need to triple by the year 2050 to accommodate expected demand. Furthermore, ethanol would not necessarily be a more reliable source of fuel. By displacing gasoline with ethanol we are displacing geo-political risk with yield risk, and historical corn yields have been about twice as volatile as oil imports. Finally, because large temperature increases can simultaneously increase fuel demand and the cost of growing corn, the supply response of ethanol producers to temperature induced demand shocks would likely be weaker than that of gasoline producers.

READINGS

■ “Post-Green Revolution Trends in Yield Potential of Temperate Maize in the North-Central United States,” by Donald Duvick and Kenneth Cassman. Crop Science, Vol. 39 (1999).

■ “Biomass as an Energy Source for the Midwestern U.S.,” by Dennis Keeney and Thomas DeLuca. American Journal of Alternative Agriculture, Vol. 7 (1992).

■ “Constraints on the Expansion of the Global Food Supply,” Henery Kendall and David Pimentel. Ambio, Vol. 23 (1994).

■ “Crop Scientists Seek a New Revolution,” by Charles Mann. Science Vol. 283 (1999).

■ “Variability and Growth in Grain Yields, 1950–94: Does The Record Point to Greater Instability?” by Rosamond Naylor, Walter Falcon, and Erika Zavaleta. Population Development Review Vol. 23 (1997).

■ “U.S. Agriculture and Climate Change: New Results,” by John Reilly. Climate Change Vol. 57 (2003).

■ “Estimating the Impact of Climate Change on Crop Yields: The Importance of Non-Linear Temperature Effects,” by Wolfram Schlenker and Michael Roberts. Available at SSRN:http://ssrn.com/abstract=934549, 2006.

■ “The Energy Balance of Corn Ethanol: An Update,” by Hosein Shapouri, James Duffield, and Michael Wang. Published by the U.S. Department of Agriculture, Economic Research Service, Washington, D.C., 2002

---

[1] James Eaves is an Assistant Professor in the Department of Finance. He can be contacted at James.Eaves@fsa.ulaval.ca.

Stephen Eaves is the Vice President of Eaves Devices. He can be reached at stephen.eaves@eavesdevices.com.

TABLE 1
Ethanol’s Net Energy Value The USDA’s estimate of the net energy gained from producing ethanol from corn.
	Mega joules per liter
Energy in one liter of ethanol	23.4
USDA Credit	3.8
Grow corn	(6.0)
Transport corn to plant	(0.6)
Operate plant	(14.4)
transport to ethanol to fueling stations	(0.4)
Net Energy	6.0

TABLE 2
All for Almost Nothing The table reports the percentage of the all-time-high corn harvest required to displace 15% of annual gasoline consumption. The last row reports the percentage of gasoline consumption displaced by devoted all corn to ethanol production.
15% of annual gasoline consumption (millions of liters)	98,647
Required ethanol (millions of liters ) a	123,308
Net liters / metric ton	103
Required corn (millions of tons)	1,203
% of all-time high harvested	423%
% of gasoline demand displaced with 100% of all-time high corn harvested	3.50%
a Assuming a 20% efficiency loss

TABLE 3
The Reliability of Corn Relative to Imported Petroleum Summary statistics for the annual percentage change in corn yields and oil imports
	Mean	Standard Deviation	90% confidence interval
			Lower	Upper
Corn Yields	3.3	11.9	-31.8	38.6
Petroleum Imports	5.0	6.7	-14.9	24.9

Who is a Revolutionary?

Greetings, Members and Visitors to the Salon

I was asking myself this question the other day: Who is a Revolutionary? Now to answer it, I suppose one must ask "What is a revolution?" I define revolution to be a break with prior continuity. Thus, the invention of the car is a revolution, but Ford's ability to sell many, cheaply is not.

My favorite example (for the moment!) of a Revolutionary is Karlheinz Brandenburg. And who was he, you might ask? He is responsible for the greatest revolutionary change in music in the past 20 years. Yes, he was the driving force behind the MP3 compression format. This format is the reason that many of us listen to music on computers, not CDs (which were not much of a revolution), gave Napster something to do, and made the iPod a useful idea, etc.

Wikipedia is another example, for its impact on the way we understand and share knowledge.

What do you think? Who is your favorite Revolutionary? Why?