Dear Administrator Jackson,

Major advancements in the science of lifecycle greenhouse gas (GHG) analysis have occurred since the Environmental Protection Agency (EPA) finalized the expanded Renewable Fuel Standard (RFS2) nearly three years ago. As a result of these analytical improvements and the availability of more robust data, the lifecycle GHG analyses of corn and sugarcane ethanol conducted for EPA’s final rule are now obsolete and do not reflect current production practices and efficiencies. Indeed, improved modeling and better data show that the corn ethanol process is more efficient and producing less GHG emissions today than EPA assumed would be in the case in 2022. Accordingly, we encourage EPA to initiate a process to revise its lifecycle GHG analysis of corn and sugarcane ethanol to better reflect the current state of the science and data. In the pre-amble for the RFS2 final rule, EPA acknowledged that lifecycle GHG analysis is an evolving science, and that updates to the Agency’s analysis would be undertaken as better data and methodologies became available. Further, EPA recognized that technology adoption and efficiency improvements in biofuel production may also necessitate periodic reassessments of the RFS2 lifecycle analysis. For example, EPA wrote that it “…recognizes that as the state of scientific knowledge continues to evolve in this area, the lifecycle GHG assessments for a variety of fuel pathways will continue to change.”1 The Agency further stated that it “…plans to continue to improve upon its [lifecycle] analyses, and will update it in the future as appropriate…”2 and “…the Agency is also committing to further reassess these determinations and lifecycle estimates.”3 Given EPA’s commitment to update its analysis to reflect the most current data and studies, and in recognition of the breadth of new information available, we believe that the time is now for the Agency to re-evaluate its lifecycle assessments and acknowledge the new science and data discussed in this letter. Over the past three years, important new information has become available regarding ethanol plant energy use and related GHG emissions. Three new versions of Argonne National Laboratory’s Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET) model have been released since EPA published its final rule in March 2010. 4 GREET1.8d, released in August 2010, significantly revised ethanol plant energy use based on an extensive peer-reviewed and published survey of ethanol plants conducted by the University of Illinois at Chicago.5 According to GREET1.8d (and the 1 75 Fed. Reg. 14,765 2 75 Fed. Reg. 14,677 3 Id. 4 EPA used both GREET1.8b and GREET1.8c for parts of its RFS2 lifecycle analysis. 5 Mueller, S. (2010). 2008 National dry mill corn ethanol survey. Biotechnology Letters, 32, 1261-1264. Summary available at: two subsequent versions of GREET), dry mill ethanol plants producing dried distillers grains use, on average, 27,576 BTU/gallon of natural gas thermal energy. Dry mill plants producing wet distillers grains use 16,435 BTU/gallon on average. These values (based on real-world data) are considerably lower than EPA’s final rule assumptions regarding 2022-era natural gas use by dry mill ethanol plants. A comparison of the current GREET model’s ethanol plant energy use values to EPA’s assumptions for 2010, 2015 and 2022 is shown in Figures 1-2 of the attachment. Additionally, a key assumption in EPA’s analysis of ethanol plant energy use was that it would take until 2022 for 70% of dry mill plants to adopt corn oil extraction technology. However, it is estimated that roughly two-thirds of the dry mill industry has already adopted corn oil extraction technology today. In its July 2011 proposal for 2012 RFS volume requirements, EPA acknowledged the rapid adoption of corn oil extraction, noting that the Agency “expects that the percentage of dry mill ethanol facilities using some form of corn oil extraction technology will increase to 60% by 2013.”6 Yet, this revised assumption is not reflected in EPA’s lifecycle analysis Better data and information are also available regarding emissions related to corn production. Specifically, newer data from the USDA National Agricultural Statistics Service is available regarding fertilizer use and energy use related to corn farming. This updated agricultural data is included in the current version of the GREET model, as well as the latest version of the Forest and Agricultural Sector Optimization Model (FASOM), the model used by EPA to estimate domestic agriculture emissions. Further, new research by Clay et al. demonstrates that no-till corn production systems (which account for about 30% of U.S. corn acres7 ) can result in a large carbon sequestration credit (approximately 10-15 grams of CO2-equivalent/mega joule [g/MJ]) that is not reflected in EPA’s current analysis. 8 In addition, research is under way to estimate the impacts on corn ethanol lifecycle GHG emissions of replacing a portion of traditional animal feed ingredients with corn stover that is harvested as a “co-product” of corn grain production. While predictive land use change (LUC) analysis remains highly uncertain, the methods and data associated with LUC estimation have substantially improved since EPA finalized the RFS2. These improvements have resulted in corn ethanol LUC emissions estimates that are much lower than EPA’s estimate for the RFS2. The improved LUC emissions estimates primarily result from better data and enhanced understanding of: the types of land most likely to be converted, the most likely location of predicted conversions, crop yields on newly converted lands, crop yield responses to changes in prices, carbon stocks and emissions from land conversion, the effects of animal feed co-products on land use, and crop switching/cross-commodity effects. New and improved methodologies for accounting for land use emissions over time (i.e., “time accounting”) have also been established.9 Important revisions have been made to Purdue University’s GTAP model, which was used by EPA to “cross-check” its LUC results from the FASOM/FAPRI framework. Specifically, improvements were made to the model’s energy elasticities, treatment of distillers grains, land conversion factors for new cropland, treatment of endogenous yield for cropland pasture, handling of cropland switching, and availability of cropland pasture and CRP. The result of these improvements was a reduction in estimated 6 76 Fed. Reg. 38,863 7 8 Clay, D., Schumacher, T., Clay, S., Chang, J., Gelderman, R., Carlson, C., Reitsma, K., & Janssen, L. (2012). Corn Yields and No-Tillage Affects Carbon Sequestration and Carbon Footprints. Agronomy Journal, 104, 763-770. It is notable that this study examines the carbon sequestration credit of no-tillage systems only; the emissions and sequestration impacts of other conservation tillage practices (e.g., strip-till, reduced till, etc.) have not been examined. 9 See, for example, Kloverpris, J. & Mueller, S. (2012). Baseline time accounting: Considering global land use dynamics when estimating the climate impact of indirect land use change caused by biofuels. Int J Life Cycle Assess, online Sep. 11, 2012. LUC emissions for corn ethanol from 30 g/MJ to 14.5-18.2 g/MJ.10 Subsequent work by Purdue researchers lowered corn ethanol LUC emissions further to 12.9-17 g/MJ.11 Meanwhile, LUC modeling conducted in 2011 by the International Food Policy Research Institute (IFPRI) for the European Commission estimated corn ethanol LUC emissions at 10 g/MJ.12 IFPRI utilized the MIRAGE model for this research. In a report released in May 2012, researchers at Argonne National Laboratory and University of Illinois Chicago built upon Purdue’s recent GTAP work to develop a Carbon Calculator for Land Use Change from Biofuels Production (CCLUB) that is included in the newest version of the GREET model. 13 The CCLUB estimates corn ethanol LUC emissions at 8-9.1 g/MJ. Most recently, Kim, Dale, and Ong estimated corn ethanol LUC emissions at 3.9-8.6 g/MJ using a new allocation method that more accurately assigns LUC emissions among the various drivers of conversion.14 Thus, based on newer data and improved methodologies, the independent estimates of corn ethanol LUC produced since the RFS2 was finalized have generally trended in the range of 8-18 g/MJ. This compares to EPA’s net LUC emissions estimate for corn ethanol of 28.4 g/MJ.15 A comparison of major LUC estimates for corn ethanol is presented in Figure 3 of the attachment. Because the FASOM/FAPRI modeling system used by EPA is not readily available to stakeholders, it is unclear whether these models have been similarly updated to reflect more current data and advanced scientific understanding of LUC. In addition, we continue to believe the Agency’s LUC analysis significantly exaggerates land conversion because it attempts to isolate the potential land use impacts of increasing the production of only one biofuel at a time while holding other biofuel volumes constant. Because EISA requires increasing volumes of various biofuels simultaneously, EPA should have based its LUC estimates on the scenario results that simulated concomitant increases in the various biofuels required by the Act. This issue was the subject of a letter from me to Administrator Jackson dated Aug. 4, 2010. 16 The letter showed that allowing the models to increase all biofuels simultaneously resulted in corn ethanol LUC emissions of 10.8 g/MJ, a 62% reduction from EPA’s current estimate. While we did receive a response to the letter from EPA, we do not feel the issue has been sufficiently resolved. We continue to believe it is both possible and necessary to base land use estimates on the scenario that most closely resembles reality (i.e., simultaneous volume increases). When some of the abovementioned improvements are combined into one analysis, the results are striking. A recent study by Wang et al. incorporated many of the aforementioned improvements regarding ethanol plant energy use, corn farming energy use, and land use change.17 As a result, the authors found “…U.S. corn ethanol at present, on average, results in a life-cycle reduction in GHG emissions of 24% relative to the emissions associated with gasoline. Dry milling plants have larger reductions…(emphasis added)” Further, the researchers concluded, “Our results are in contrast to some of the studies published in the past three years for two major reasons: updated data to reflect technology improvements over time 10 Tyner, W., Taheripour, F., Zhuang, Q., Birur, D., & Baldos, U. (2010). Land Use Changes and Consequent CO2 Emissions due to US Corn Ethanol Production: A Comprehensive Analysis, Final Report. Available at: 11 Taheripour, F. & Tyner, W. (2012). Induced land use emissions due to first and second generation biofuels and uncertainty in land use emissions factors. Agricultural & Applied Economics Association’s 2012 Annual Meeting, Seattle, Washington, August 12-14, 2012. Available at: 12 Laborde, D. (2011). Assessing the Land Use Change Consequences of European Biofuel Policies, Final Report. Available at: 13 Mueller, S., Dunn, J., & Wang, M. (2012). Carbon Calculator for Land Use Change from Biofuels Production (CCLUB): Users’ Manual and Technical Documentation. ANL/ESD/12-5. Available at: 14 Kim, S, Dale, B.E., & Ong, R.G. (2012). An alternative approach to indirect land use change: Allocating greenhouse gas effects among different uses of land. Biomass & Bioenergy, 46, 447-452. 15 “International Land Use Change” (30.3 g/MJ) + “Domestic Land Use Change” (-1.9 g/MJ) = 28.4 g/MJ. See Table V.C-1, 75 Fed. Reg. 14,788. 16 Available at: 17 Wang, M.Q., Han, J., Haq, Z., Tyner, W.E., Wu, M., & Elgowainy, A. (2011). Energy and greenhouse gas emission effects of corn and cellulosic ethanol with technology improvements and land use changes. Biomass & Bioenergy, 35, 1885-1896. and detailed simulations of a few critical issues in modeling the LUCs of corn ethanol.” This study is critically important because it demonstrates that average corn ethanol reduces lifecycle GHG emissions by more than 20% today, and that average GHG savings will be far greater in 2022 than EPA’s assumption of 21%. Recent studies have also improved the understanding of the GHG impacts of sugarcane ethanol produced in Brazil, where sugarcane area (harvested) has increased 55% since 2006. Analyses by Seabra et al.18 and Macedo et al.19 examine fossil energy use and GHG emissions related to sugarcane ethanol production. Based on these studies and others, the GREET model was updated to include N2O emissions from filtercake and vinasse, as well as supplemental fertilizer inputs due to increased sugarcane collection. Additionally, a December 2011 study by Tsao et al. found that existing estimates of sugarcane ethanol lifecycle GHG emissions tend to underestimate emissions from sugarcane field burning.20 The authors found, “…even in regions where pre-harvest field burning has been eliminated on half the croplands, regional emissions of air pollutants continue to increase owing to the expansion of sugar-cane growing areas, and burning continues to be the dominant life-cycle stage for emissions.” Further, they concluded, “Accounting for this effect leads to revised regional estimates of burned area that are four times greater than some previous estimates.” On the subject of land use emissions, Lapola et al. found sugarcane ethanol expansion to be a chief contributor of potential LUC emissions in Brazil. 21 The results of the Lapola et al. paper stand in stark contrast to EPA’s analysis, which attributed only 4.8 g/MJ of LUC emissions to sugarcane ethanol. Further, Adami et al. found 70% of the land directly converted to sugarcane production in Brazil from 2005 to 2010 was previously in pasture; this is contrary to EPA’s modeling results that assume cane expansion would be largely offset by reduced corn acres in Brazil.22 While the new studies, model updates, and improved data described above are the most noteworthy advancements in the science of ethanol lifecycle analysis, they represent only a portion of the work that has occurred since EPA finalized the RFS2. Clearly, the understanding of ethanol’s GHG impacts has progressed rapidly and meaningfully over the course of the last three years. Thus, we believe there is ample scientific justification for EPA to revisit its analysis. Updating EPA’s lifecycle analysis of corn and sugarcane ethanol is important for several reasons. First, it demonstrates EPA is honoring its commitment to periodically revisit its lifecycle analyses and make changes as warranted by scientific advancements. EPA has been a leader in the field of biofuels lifecycle assessment, and initiating a process to update the RFS2 analysis ensures that the Agency maintains an active and relevant role in the scientific discussion around biofuel lifecycle GHG accounting. Second, an effort by EPA to update its analysis will enhance the public’s understanding of corn ethanol’s lifecycle GHG impacts and serve to inform debate on future biofuels policies. In addition, updated analyses of corn and sugarcane ethanol will allow for fairer comparisons of the two fuels moving forward. Finally, updating EPA’s analysis would help ease the Agency’s workload and reduce the backlog of petitions for new pathways. We note that four of the 13 pending petitions for new ethanol pathways are for new corn starch ethanol pathways. These petitions likely were submitted so that the facilities could expand capacity without needing to use the prescriptive “advanced technologies” 18 Seabra, J., Macedo, I., Chum, H., Faroni, C., & Sarto, C. A. (2011). Life cycle assessment of Brazilian sugarcane products: GHG emissions and energy use. Biofuels, Bioproducts and Biorefining, 5, 519–532. 19 Macedo, I., Seabra, J., & Silva, J. (2008). Greenhouse gases emissions in the production and use of ethanol from sugarcane in Brazil: The 2005/2006 averages and a prediction for 2020. Biomass and Bioenergy, 32, 582–595. 20 Tsao, C-C., Campbell, J.E., Mena-Carrasco, M., Spak, S.N., Carmichael, G.R., & Chen, Y. (2011). Increased estimates of airpollution emissions from Brazilian sugar-cane ethanol. Nature Climate Change, 1325, Advanced Online Publication, available at: 21 Lapola, D., Schaldach, R., Alcamo, J., Bondeau, A., Koch, J. Koelking, C., & Priess, J. (2010). Indirect land-use changes can overcome carbon savings from biofuels in Brazil. Proceedings of the National Academy of Science, 107, 3388-3393. 22Adami, M., Friedrich, B., Rudorff, T., Freitas, R.M., Aguiar, D.A., Sugawara, L.M., & Mello, M.P. (2012). Remote Sensing Time Series to Evaluate Direct Land Use Change of Recent Expanded Sugarcane Crop in Brazil. Sustainability, 4, 574-585. designated by EPA in the RFS2 final rule. If EPA updated its lifecycle analysis and found that average corn ethanol results in greater than 20% GHG savings today, corn ethanol producers seeking to expand capacity would not need to implement the “advanced technologies” or file a petition for a new pathway. Thank you for your consideration of this information and we look forward to further interaction with the Agency on issues related to lifecycle analysis and RFS2 implementation. Sincerely, Bob Dinneen President & CEO Cc: Gina McCarthy Chris Grundler ATTACHMENT FIGURE 1. Corn Ethanol Plant Natural Gas Energy Use (Dry Mill, DDGS), Comparison of GREET Values to EPA Assumptions FIGURE 2. Corn Ethanol Plant Natural Gas Energy Use (Dry Mill, WDGS), Comparison of GREET Values to EPA Assumptions 31,614 30,512 28,660 27,576 25,000 26,000 27,000 28,000 29,000 30,000 31,000 32,000 2010 2015 2022 GREET1_2012 EPA: RFS2 Final Rule Argonne BTU/gallon Corn Ethanol Dry Mill Natural Gas Thermal Energy Use (DDGS) 18,842 18,185 17,081 16,435 15,000 16,000 17,000 18,000 19,000 20,000 2010 2015 2022 GREET1_2012 EPA: RFS2 Final Rule Argonne BTU/gallon Corn Ethanol Dry Mill Natural GasThermal Energy Use (WDGS) FIGURE 3. Corn Ethanol LUC Emissions Estimates, 2008-Present 1. Searchinger, T., Heimlich, R., Houghton, R.A., Dong, F., Elobeid, A., & Fabiosa, J.(2008) Use of U.S. croplands for biofuels increases greenhouse gases through emissions from land-use change. Science, 319, 1238-1240. 2. U.S. EPA (2009). Notice of Proposed Rulemaking: Changes to Renewable Fuel Standard Program 3. California Air Resources Board (2009). California’s Low Carbon Fuel Standard, Final Statement of Reasons. 4. U.S. EPA (2010). Regulation of Fuels and Fuel Additives: Changes to Renewable Fuel Standard Program; Final Rule. 5. Hertel, T.W., Golub, A.A., Jones, A.D., O’Hare, M., Plevin, R.J., & Kammen, D.M. (2010). Effects of US Maize Ethanol on Global Land Use and Greenhouse Gas Emissions: Estimating Market-mediated Responses. BioScience, 60, 223-231. 6. Tyner, W., Taheripour, F., Zhuang, Q., Birur, D., & Baldos, U. (2010). Land Use Changes and Consequent CO2 Emissions due to US Corn Ethanol Production: A Comprehensive Analysis, Final Report. 7. Laborde, D. (2011). Assessing the Land Use Change Consequences of European Biofuel Policies, Final Report. 8. Mueller, S., Dunn, J., & Wang, M. (2012). Carbon Calculator for Land Use Change from Biofuels Production (CCLUB): Users’ Manual and Technical Documentation. ANL/ESD/12-5. 9. Kloverpris, J. & Mueller, S. (2012). Baseline time accounting: Considering global land use dynamics when estimating the climate impact of indirect land use change caused by biofuels. Int J Life Cycle Assess, published online. [Value shown is from Hertel et al., corrected for time accounting] 10. Kim, S, Dale, B.E., & Ong, R.G. (2012). An alternative approach to indirect land use change: Allocating greenhouse gas effects among different uses of land. Biomass & Bioenergy, 46, 447-452 104 63.3 30 28.4 27 10 9.1 11 14.5 3.9 0 10 20 30 40 50 60 70 80 90 100 110 Searchinger et al. [1] U.S. EPA [2] CA Air Resources Board [3] U.S. EPA [4] Hertel et al. [5] Tyner et al. [6] Laborde [7] GREET1_2012 CCLUB [8] Mueller & Kloverpris [9] Kim et al. [10] Feb-08 May-09 Dec-09 Mar-10 Mar-10 Jun-10 Oct-11 Jul-12 Aug-12 Nov-12 g CO2e/MJ CORN ETHANOL LUC EMISSIONS ESTIMATES, 2008-PRESENT 18 8.6

November 30, 2012
The Honorable Lisa P. Jackson
Administrator U.S. Environmental Protection Agency
Mail Code 1101A 1200 Pennsylvania Avenue,
N.W. Washington, D.C. 20460