The USDA recently updated the numbers on the energy balance of corn ethanol in 2015 Energy Balance for the Corn-Ethanol Industry. Today returning guest Todd “Ike” Kiefer scrutinizes the numbers in the report and he raises some critical questions about the data and methodology.
Previously Mr. Kiefer wrote an article critical of the Navy’s efforts to promote biofuels in a periodical that is sent to Congress and top military leaders. The article was entitled Energy Insecurity: The False Promise of Liquid Biofuels (discussed here). He also wrote a guest article here in the past called EPA’s Sleight of Hand on Cellulosic Fuel Rule Change. His biography can be found at the end of the article.
I would remind readers that while I may agree with much, if not most of what Mr. Kiefer writes, these are his opinions. I have not taken a close look at this USDA paper myself, so it is possible that we could have a difference of opinion on some element(s) of the analysis. I don’t know that to be the case, but until I read the paper myself I offer up that caveat.
A Critical Review of the USDA’s 2015 Energy Balance for Corn Ethanol
Todd “Ike” Kiefer
This just published 2015 USDA paper on corn ethanol energy return on investment (EROI) is the latest in a string written by the same author, and in a series published under the imprimatur of the USDA Office of the Chief Economist. The author, Paul W. Gallagher, cites his own previous work 8 times, as well as referencing papers of the preceding lead author in this series, Hosein Shapouri. This paper represents a curious mixture of consistency and inconsistency. The author was kind enough to post the supporting calculation spreadsheets online. This is the most fruitful place to explore. Examination of these raises some serious considerations.
The first thing of note is that the headline number — the EROI of corn ethanol — while hailed by the Secretary of Agriculture as having made great strides, is in fact little changed. Shapouri’s 2001 USDA paper found the EROI of corn ethanol to be 1.67:1. Gallagher’s 2015 paper finds the 9-state average to be 1.5:1 without credit for co-products, and 2.2:1 with credit. While this might appear mildly interesting in proportion, it is totally insignificant in context of the energy needs of a post-industrial economy, and in comparison to other energy sources.
Rome achieved an EROI of 1.8:1 2000 years ago using wheat-fed slaves and oxen-fed alfalfa — right in line with corn ethanol today. This EROI, typical of agricultural age civilizations, was the best human civilization achieved for millennia. However, beginning in the 17th century with the invention of the steam engine, leading-edge civilizations quickly advanced to an energy intensity and quality of life that is sustainable only with much higher returns.
The U.S. economy since the 20th century has slipped into recession whenever any significant portion of its energy supply has dipped below a 6:1 EROI, as happens with some predictability in the multi-decadal cycle of the global crude oil market. A net EROI greater than 10:1 empirically seems to be essential to economic growth in modern western nations with their extensive energy consumption overheads. Today, the U.S. economy is sustained by EROIs of approximately 15:1 (refined petroleum), 28:1 (natural gas), 30:1 coal electricity, 75:1 nuclear electricity, and > 100:1 hydro-electricity. Price is inversely correlated with EROI, and the low price for natural gas today — about $2 per million BTU — is the fruit of a wave of capital investment in technology that has not only opened up previously inaccessible shale resources, but has done so with an 85:1 EROI in the Marcellus.
So, in the context of 10:1 to 100:1 EROI alternatives, and a 6:1 EROI threshold to sustain current quality of life, a 2:1 EROI is woefully inadequate. Any investor would be a fool to put his money into a 2:1 proposition when there are 85:1 returns to be had. He could certainly not afford to do so in a competitive environment and hope to survive when others are pursuing the higher returns. It would also be unconscionable to intentionally pursue a course of action that would revert his own household and his nation to the energy poverty of the agricultural age, and undermine all the benefits that accessible and affordable energy brings, such as clean indoor heating and cooking, purified water, sewerage, waste disposal, refrigeration, modern healthcare, and all the mundane miracles of transportation we enjoy today.
So the question for the paper really is much more than the decimal place accuracy of the current anemic EROI of corn ethanol — it is the plausibility of its predictions for future order-of-magnitude progress. That question hinges on whether there is room for improvement in the yields of bushels of corn per acre and in the yield of ethanol per bushel of corn, without investing proportionately more energy to achieve them. The answer is no and no. It is telling to find in Gallagher’s compilation of USDA survey data that the average yield of corn continues its slow, asymptotic approach to 180 bu/acre-yr, and the yield of ethanol per bushel has been stuck at 2.7-2.8 gallons since 2001. Predictions of breaking the 500-gal/acre-yr yield wall are not happening for corn ethanol.
The USDA-predicted breakthrough for cellulosic ethanol has similarly failed to materialize. Those who’ve studied and attacked the chemistry in detail, like Codexis, have had to concede that the energy penalty of hydrolyzing lignocellulose is greater than the energy benefit of its cultivation compared to starch or oil-yielding plants. Others have reconfirmed the absolute fossil-fuel dependence of modern intensive agriculture, and have realized that the charted equivalent-energy prices of biofuels and fossil fuels, while highly correlated, can never cross. Biofuels have a built-in premium to fossil fuels that is their debt to them for being the feedstock for their fertilizers and pesticides and the energy for their cultivation and processing. Unfortunately, Poet and Abengoa and DuPont did not do their thermodynamics and energy balance homework before charging ahead with quarter-billion-dollar cellulosic ethanol plants that are now years behind schedule and failing to ramp up to their promised capacities because the insiders know they will lose money on every gallon. These facilities, without some new massive subsidy regime, will soon join those of Cello, KL Energy, Range Fuels, and KiOR in conspicuous disuse.
The only other facilities currently claiming cellulosic ethanol Renewable Identification Numbers (RIN) from the EPA are doing so because of EPA bad faith in administering the Renewable Fuel Standards (RFS) program. The agency unilaterally decided in 2014 to redefine “cellulosic biofuel” to include landfill methane and biogas that have nothing to do with cultivated feedstock. They also decided to redefine the world “cellulosic” to mean “75% cellulosic,” and to apply it only to the inputs, not the outputs. Thus, operations like Quad County Corn Processors can run milled corn kernels through a second cycle and count the tiny amount of ethanol produced from residual starch missed in the milling process but leached out in the secondary process as “cellulosic” because 75% of the material shoveled in was corn bran and only 25% was corn starch. It matters not to the EPA if all of that hemicellulosic material remains inert throughout the process and comes out the far end unconverted, and that the only ethanol produced is from starch; they still get cellulosic RINs from the EPA. Such deceitful and desperate measures from the EPA are still two orders of magnitude too small to inflate the current production of cellulosic ethanol to match its 2005 predictions.
So, are the small EROIs in the USDA paper correct, or does corn ethanol have a chance of becoming a meaningful player in energy? When the specific numbers and calculations in the spreadsheets are examined, some things pop out to someone familiar with EROI protocols. The first thing of note is that the author favored use of lower heating values (LHV). This is contrary to standard EROI practice of using higher heating values (HHV) because of their superior utility for comparisons and conversions between energy sources. When Shapouri published in refereed journals he used HHVs exclusively (presumably based on peer review feedback), but when he published for the USDA, he used LHVs. Using HHV is particularly critical for the “embodied energy” values which are the core of EROI computations. Shapouri collected some rough LHV embodied energy estimates from a single source named Stokes in 2003 that Gallagher continues to use today. There is quite a bit of variability in the literature for the embodied energy in fertilizers and pesticides and fossil fuel inputs. For this reason it is difficult to criticize an author’s specific choice for these values. However, it was observed that the embodied energy for liquefied petroleum gas (used for drying corn) was roughly equal to its LHV. This is obviously an error, as the embodied energy of every refined fuel should be significantly higher than its HHV, let alone its LHV.
In looking at the compiled USDA survey data for energy inputs, we find that natural gas use per acre for the most recent data set (2010) is mysteriously low compared to preceding years, decreasing from well over 200 cu ft per acre for every previous reporting period to only 34 cu ft per acre. The raw data inputs in the spreadsheets are all suspiciously rounded to multiples of ten, unlike other inputs, as if they are guesses or estimates. This may be a typographical error, or a survey error. Either way, the unusually low value serves to inflate the most recent EROI calculations.
The most disturbing observation is that 1/3 of all the energy inputs required to grow corn were set aside right from the start. The rationale is apparently that, since only 66% of the corn plant is starch, only that amount of the input energy should be allocated to the resulting corn starch ethanol. This is an indefensible gimmick, as one is obligated to grow the entire corn plant to get the corn kernel starch. The proper way to account for the non-starch portion of the plant is to count all energy inputs at the front end of the process, and to give appropriate energy credits for any valuable co-products at the end of the process. For example, every gallon of corn ethanol produced by dry milling is accompanied by 5.3 lb of distillers dried grains and solubles (DDGS) and 2.15 lb of distillers wet grains (DWG). These co-products have HHV energies that can be counted in the EROI computations. These are obligate co-products, not optional ones, and they need to be treated holistically within the EROI calculations, not be somehow segregated out of them.
In accordance with the principles of sound EROI computations, and in compliance with ISO 14044 that governs environmental lifecycle assessments, there are two ways to properly deal with co-products. The choice of method depends upon whether the co-products are economically superior or inferior to existing alternatives already in the market. If they are superior, then they get full credit for their HHV energy yield in the EROI calculation because they will naturally displace the alternatives. However, if they are inferior, then one should substitute the embodied energy of the alternative commodities already in the market, because the co-product could and would be alternatively generated in this superior way if not for forced displacement. It turns out that all the co-products of corn ethanol (DDGS, DWG, corn gluten meal, corn gluten feed, corn oil) have equivalent commodities already in the market place that are fully substitutable at lower cost in energy and money. All corn ethanol co-products can be displaced more economically by abundantly available corn meal, soy meal, soy oil, and urea. A proper EROI calculation for corn ethanol would use the energy to make these displacing commodities as the value of the energy in corn ethanol co-products. This realistic approach to EROI is part of what is called a “consequential lifecycle assessment” that more accurately portrays real world costs and benefits.
When this reviewer adjusted the spreadsheet calculations to reflect HHVs instead of LHVs, and substituted some more rigorous embodied energy values, and applied proper co-product credit computations, the ultimate EROI results came out about the same for the base cases without co-product credits: 1.5:1 for dry mill and 2.2:1 for wet mill. However the values with co-product credits were only 1.8:1 for dry mill and 2.4:1 for wet mill — nothing close to the 4:1 claim for Iowa, or the 6:1 minimum to be marginally useful to an industrial-age economy.
Burning stover for biorefinery process heat does not substantially improve EROI, because it is merely a substitution of one energy source for another, and its removal negatively affects the mass-balance of nutrients in the soil, and the reduction of ground cover also accelerates the loss of additional nutrients and moisture. Using stover and/or DDGS as a substitute for fossil fuels affects a calculation called the fossil fuel ratio (FFR), but there is no significant energy saving that would affect EROI. There is no free lunch. Also, anyone who has seen large-scale stover-burning ethanol plants realizes their emissions are an EPA nightmare. They operate at lower thermodynamic efficiencies and naturally belch black smoke particulates, volatile organics, sulfur, nitrogen oxides, and heavy metals like a Carnegie steel mill from the turn of the last century. They also produce fly ash and bottom ash that must be dealt with. They require essentially the same emission controls as a coal plant, and this further reduces efficiency.
Clean and cheap natural gas is likely to remain the processing plant fuel of choice, and there is no evidence of the fantasy of corn farms and ethanol plants powered only by corn and ethanol becoming a reality. The low EROI of corn ethanol is the reason for this non-migration away from fossil fuel energy; the non-migration is not the reason for the low ethanol EROI. To get the highest ethanol EROI possible, one has use the highest EROI energy inputs available, and the fossil fuel alternatives are far superior to ethanol in this regard. The EROI Gallagher and Shapouri and others are really computing in these papers is not a pure corn ethanol EROI at all, but rather a hybrid fossil fuel-ethanol EROI. A true corn ethanol EROI would be for the process of making tomorrow’s batch of corn ethanol using only today’s supply of corn ethanol and nature’s inputs of sun and nutrients. Making corn ethanol this way is a non-sustainable energy death spiral. It has a huge negative energy balance without high EROI injects from fossil fuels in the form of fuels, fertilizers, pesticides, machinery, and grid electricity. Likewise, using high EROI energy sources directly as fuels is more efficient and beneficial to the economy than using them as a crutch to subsidize low-EROI alternatives. The nation’s huge ongoing “investment” in low-return corn ethanol represents a huge opportunity cost of energy and money and other resources that could be spent far more productively elsewhere.
This USDA paper would have benefited from more peer review before the fact, but the author’s courage and integrity to share his data and computations online is to be applauded. We need more of this full disclosure from the USDA and EPA and DOE.
Todd “Ike” Kiefer graduated from Annapolis in 1988 with an undergraduate degree in physics, and then earned master’s degrees in strategy and military history at the Army Command and General Staff College in Fort Leavenworth KS. He retired as a Captain after a 25-year military career as Naval aviator and electronic warfare expert (EA-6B Prowler pilot). He has been deployed eight times to the Middle East and Southwest Asia, and served twenty-two months on the ground in Iraq. He commanded Al Asad Air Base in Al Anbar Province Iraq, and spent three years as Pentagon strategic planner on Joint Staff. His most recent assignment was three years as CJCS Chair and faculty instructor of strategy, leadership, and warfighting at the US Air Force Air War College in Montgomery AL.