The EPA today held a public hearing on the RFS2 Rule and will be holding a Workshop on Lifecycle Greenhouse Gas Analysis for the Proposed Revisions to the National Renewable Fuels Standard Program tomorrow. These fora are intended to solicit feedback from stakeholders in the rulemaking and provide information on how the EPA developed its lifecycle model.
BIO submitted testimony to the EPA asking that the agency maintain flexibility in the rule to reconsider the greenhouse gas reductions associated with biofuels as the science of life cycle assessment, particularly methods of accounting indirect land use change, matures.
At a U.S. House Agriculture Subcommittee hearing on May 6, several experts discussed the considerable uncertainties that remain in the current models for ILUC that the EPA has employed.
Joe Glauber, chief economist of the USDA, listed some of the sources of uncertainty, including crop yields and production growth over time, economic and non-economic limits to land use change, and assumptions plugged into the models at the start. Glauber also noted that predicted U.S. plantings of principal crops for 2009 remain well below historical trends and that increases in corn and soybean acres are offset by declines in wheat and cotton acres.
As Glauber points out, conversion of forests is one of the key factors in the life cycle greenhouse gas estimates of the EPA’s analysis and many of the published studies on ILUC. Yet, there is considerable uncertainty over whether forest would be the primary source of new cropland.
Bruce Babcock, director of the Center for Agricultural and Rural Development at Iowa State University, pointed out that there is little evidence that U.S. biofuel production has caused deforestation either in the United States or in Brazil. Rather than conversion of forest or pastureland, “U.S. cropland has increased primarily through a reduction in CRP acres and through increased double cropping of soybeans after wheat.” Similarly, expansion of soybean production in Brazil has been accommodated by increased stocking rates on pastureland. It is increases in cattle production in Brazil that have led to deforestation for pastureland in the Amazon.
Babcock concluded, “The precision with which models can estimate emissions associated with market-induced land use changes is low.”
The fact is that the model employed by the EPA attaches heavy penalties for greenhouse gas emissions to production of biofuels based on assumptions that worldwide agriculture is a zero-sum game in which use of crops for one purpose leaves a shortfall in crop availability elsewhere that can only be made up by converting forest to agriculture. When the conclusions of the model are driven by the assumptions, they should be seen as logically invalid.