CO2 emissions from crop residue-derived biofuels. [Letter to the Editor]

Authors: SHEEHAN, JOHN - Colorado State University; Adler, Paul; Del Grosso, Stephen - Steve; EASTER, M - Colorado State University; PARTON, WILLIAM - Colorado State University; PAUSTIAN, KEITH - Colorado State University; WILLIAMS, S - Colorado State University

Submitted to: Nature Climate Change
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/2/2014
Publication Date: 10/29/2014
Citation: Sheehan, J., Adler, P.R., Del Grosso, S.J., Easter, M., Parton, W., Paustian, K., Williams, S. 2014. CO2 emissions from crop residue-derived biofuels. [Letter to the Editor]. Nature Climate Change. 4:932–933. doi:10.1038/nclimate2403.

Interpretive Summary: The claim by Liska et al. that corn stover-derived ethanol can be worse than gasoline has generated lots of media interest, but offers little value to the research community or to policymakers. They have merely demonstrated that if you model an irresponsible and unsustainable scenario, the results will look irresponsible and unsustainable. No one who has given serious thought to crop residues for biofuels would find their proposed across-the-board 6 Mg ha-1 collection rate in the US Corn Belt at all reasonable. A decade ago, Sheehan et al. used soil carbon and life cycle assessment (LCA) modelling to show that corn stover for ethanol would only make sense if farmers simultaneously adopted conservation tillage practices and constrained removal rates to account for local yield, soil, climate and topographical conditions. Numerous other field and modelling studies have shown that soil carbon levels can be maintained with conservation tillage and moderate stover removal. In contrast, Liska et al. applied a very simplistic soil carbon model that ignores important variables such as soil moisture and soil texture — making regional extrapolations highly questionable — and doesn’t allow for varying management practices. This is an important shortcoming, as farmers can reduce their tillage intensity with stover harvest, saving money, without compromising yields. Figure 1a illustrates these weaknesses when net changes in soil carbon emissions are modelled with DayCent (the analysis was using the DayCent model version used in the most recent US national greenhouse-gas emissions inventory: www.epa.gov/climatechange/ ghgemissions/usinventoryreport.html) for the Mead, Nebraska site in Liska et al.’s study. The results for 50% stover removal at this site are consistent with the 6 Mg ha-1 collection scenario used by Liska and co-authors. With a continuation of existing tillage practices, soil carbon losses over the first 10 years average 0.43 Mg C ha-1 yr-1, comparable to the value of 0.47 Mg C ha-1 yr-1 reported by Liska and co-workers. Adoption of conservation tillage reduces soil carbon losses to almost zero, while including a cover crop can yield net soil carbon increases. The problems with the scenario of Liska et al. do not end at the farm. They took another extreme position by ignoring the carbon savings associated with coproduct electricity in the conversion facility. Figure 1b puts their assumptions for the farm and the biorefinery in an LCA context. Non-soil carbon emissions are taken from a recent LCA by the US Department of Energy’s National Renewable Energy Laboratory. Here, we focus on the 50% collection scenario in Fig. 1a. Combining the assumptions of unchanged tillage and no electricity credit made by Liska et al., we too can generate a carbon footprint that is slightly worse than that of gasoline. Adoption of conservation tillage and proper accounting of the coproduct credit brings the footprint of ethanol below the Renewable Fuel Standard requirement for advanced biofuels. Adopting cover-crop strategies could make ethanol almost carbon neutral. The study by Liska et al. is symptomatic of a broader problem in the realm of LCA. Had they followed International Organization for Standardization standards10 and engaged stakeholders in the design of this study, instead of unilaterally making extreme and unsustainable assumptions, they might have ended up evaluating more useful scenarios. This is a mistake all too commonly found in the LCA literature.

Technical Abstract: The claim by Liska et al. that corn stover-derived ethanol can be worse than gasoline has generated lots of media interest, but offers little value to the research community or to policymakers. They have merely demonstrated that if you model an irresponsible and unsustainable scenario, the results will look irresponsible and unsustainable. No one who has given serious thought to crop residues for biofuels would find their proposed across-the-board 6 Mg ha-1 collection rate in the US Corn Belt at all reasonable. A decade ago, Sheehan et al. used soil carbon and life cycle assessment (LCA) modelling to show that corn stover for ethanol would only make sense if farmers simultaneously adopted conservation tillage practices and constrained removal rates to account for local yield, soil, climate and topographical conditions. Numerous other field and modelling studies have shown that soil carbon levels can be maintained with conservation tillage and moderate stover removal. In contrast, Liska et al. applied a very simplistic soil carbon model that ignores important variables such as soil moisture and soil texture — making regional extrapolations highly questionable — and doesn’t allow for varying management practices. This is an important shortcoming, as farmers can reduce their tillage intensity with stover harvest, saving money, without compromising yields. Figure 1a illustrates these weaknesses when net changes in soil carbon emissions are modelled with DayCent (the analysis was using the DayCent model version used in the most recent US national greenhouse-gas emissions inventory: www.epa.gov/climatechange/ ghgemissions/usinventoryreport.html) for the Mead, Nebraska site in Liska et al.’s study. The results for 50% stover removal at this site are consistent with the 6 Mg ha-1 collection scenario used by Liska and co-authors. With a continuation of existing tillage practices, soil carbon losses over the first 10 years average 0.43 Mg C ha-1 yr-1, comparable to the value of 0.47 Mg C ha-1 yr-1 reported by Liska and co-workers. Adoption of conservation tillage reduces soil carbon losses to almost zero, while including a cover crop can yield net soil carbon increases. The problems with the scenario of Liska et al. do not end at the farm. They took another extreme position by ignoring the carbon savings associated with coproduct electricity in the conversion facility. Figure 1b puts their assumptions for the farm and the biorefinery in an LCA context. Non-soil carbon emissions are taken from a recent LCA by the US Department of Energy’s National Renewable Energy Laboratory. Here, we focus on the 50% collection scenario in Fig. 1a. Combining the assumptions of unchanged tillage and no electricity credit made by Liska et al., we too can generate a carbon footprint that is slightly worse than that of gasoline. Adoption of conservation tillage and proper accounting of the coproduct credit brings the footprint of ethanol below the Renewable Fuel Standard requirement for advanced biofuels. Adopting cover-crop strategies could make ethanol almost carbon neutral. The study by Liska et al. is symptomatic of a broader problem in the realm of LCA. Had they followed International Organization for Standardization standards10 and engaged stakeholders in the design of this study, instead of unilaterally making extreme and unsustainable assumptions, they might have ended up evaluating more useful scenarios. This is a mistake all too commonly found in the LCA literature.