Peer review of RFS2 life-cycle analysis complete

By Erin Voegele | August 10, 2009
Report posted on Aug. 11, 2009 at 11:00 a.m. CST

The U.S. EPA recently announced that the peer review of the life cycle analysis component of the second stage of the renewable fuels standard (RFS2) is complete. The peer review documents are now available on the EPA Web site.

According to the EPA, the agency decided to initiate this independent peer review in an effort to respond to stakeholder concerns, and to ensure that the agency makes decisions based on the best available science.

The formal independent expert peer reviews were conducted in four primary areas:
1. The use of satellite data and land conversion greenhouse gas (GHG) emission factors in land use modeling
2. The methods used to account for the variable timing of GHG emissions
3. The modeling and data used to account for GHG emissions from foreign crop production
4. How the models utilized by EPA are used together to provide overall lifecycle estimates

The review was completed using independent, third-party contractors, which selected highly qualified peer reviewers. The EPA provided these contractors with a description of the expertise that is required of the reviewers and examples of experts that fit the qualifications. The agency also provided the contractor with the names of reviewers that were recommended by stakeholders. The contractors then independently developed their own list of expert reviewers and screened candidates for possible conflicts of interest, bias, contractual relationships to the EPA and for overall perception of independence for from the agency.

According to documents posted to EPA's Web site, the selected peer reviewers are recognized as leading experts in their respective fields, including life cycle assessment, economic modeling, remote sensing imagery, biofuels technologies, soil science, agricultural economics, and climate science.

In order to complete the peer reviews, the contractors provided the reviewers with materials for review and charge questions that were developed by the EPA. The contractors then conducted meetings and teleconferences as needed to facilitate the process. Peer reviewers worked independently of one another and were not asked to reach a consensus.

Five peer reviewers assessed the land use modeling component of the life cycle analysis. These reviewers included Holly Gibbs, Stanford University; Richard Houghton, Woods Hole Oceanographic Institution; Rattan Lal, Ohio State University; Jason Tullis, University of Arkansas; and Brian Wardlow, University of Nebraska.

In general, the reviewers agreed that the approach taken by EPA and Winrock International was scientifically justifiable, especially given existing data and technology constraints. The reviewers highlighted several problematic areas that are a part of the satellite imagery analysis, but agreed more strongly with the EPA's approach regarding the emissions factor analysis.

The main areas of concern highlighted by the reviewers in regard to the satellite imagery analysis include the 3-year time period of the two data sets that were chosen, as well as the error associated with each of the data sets, the coarse resolution of the satellite images and the change detection analysis performed on the two data sets from 2001 and 2004. Reviewers also expressed concern over the role of land categorized as mixed' or other' in the Winrock satellite data, the methodology for projecting land use patterns caused specifically by biofuel production, and the evaluation or error and uncertainty associated with the satellite imagery analysis.

Five different peer reviewers assessed the methods used to account for the variable timing of GHG emissions in the proposed rule for RFS2. These reviewers include Joseph Fargione, the Nature Conservancy; Ralph Heimlich, Agricultural Conservation Economics; Elizabeth Marshall, World Resources Institute; Jeremy Martin, Union of Concerned Scientists; and Kenneth Richards, Indiana University.

While most of the reviewers generally agreed that the approach taken by the EPA to calculate these factors was scientifically objective, according to the peer review documents released by the EPA, the reviewers' opinions of appropriate time frames varied. The reviewers disagreed on whether the EPA should use an impact time frame or a project time frame, and what the duration of those time frames should be. Some of the researchers also questioned the appropriateness of weighing emissions by applying a discount rate, noting that discount rates are primarily applied to monetary units, not physical units such as a carbon emissions.

Four peer reviewers were chosen to evaluate the modeling and data used to account for GHG emissions from foreign crop production. These reviewers include Elizabeth Boyer, Pennsylvania State University; Kenneth Cassman, University of Nebraska; John Freney, Commonwealth Scientific and Industrial Research Organisation; and Arvin Mosier, USDA Agricultural Research Service.

The reviewers found that the methodologies used to evaluate these impacts in the RFS2 rulemaking are generally a good first approximation of changes in GHG emissions. They also noted that several areas should be upgraded, including the correction of typographical errors and data tables, verifying specific emission factors used in the models, and using additional models to reduce the uncertainty in the estimation of nitrous oxide. While the reviewers found the methodologies used by the EPA generally acceptable, they also suggested that fertilizer use data and pesticide use data should be improved.

The four reviewers were split on how the models should deal with future yield increases. Two saw extrapolation in recent yield trends as the best approach, while the others questioned the premise that yields will continue to increase.

Four different reviewers were chosen to evaluate how the models used by the EPA are used together to provide overall life cycle estimates. The reviewers include Martin Banse, Agricultural Economics Research Institute; Timothy Searchinger, Princeton University; John Sheehan, University of Minnesota; and Michael Wang, Argonne National Laboratory.

In general, these four reviewers agreed that the EPA's approach of linking partial equilibrium models was preferable to using a general equilibrium model, such as the Global Trade Analysis Project (GTAP). This is because the reviewers agreed that no existing model can comprehensively simulate the direct and indirect effects of biofuel production domestically and internationally. The reviewers also emphasized that the partial equilibrium models used by the EPA have both positive and negative qualities, but despite problematic areas within the current approach, most of the reviewers believe the existing approach to be more reasonable than relying on the GTAP model.

Wang also suggested that the EPA include the forestry sector in the analysis. In addition, both Wang and Sheehan suggested that the 2005 baseline for carbon emissions from fuel stipulated by the Energy Independence and Security Act of 2007 is inappropriate because it underestimates GHG emissions from petroleum fuels. This is due to the fact that it is predicted that petroleum fuels will increasingly come from unconventional sources and that the growth of global demand for petroleum could generate unanticipated indirect effects in the petroleum sector.

On August 7 the Renewable Fuels Association released a statement regarding the completion of the peer review analysis for the RFS2 proposed rulemaking. According to the statement, the RFA asserts that the EPA stacked the deck against biofuels in its process to peer review its indirect land use change analysis conducted for the RFS2.

The RFA said that several noted anti-ethanol and anti-agriculture activists, including Searchinger and Fargione, were among those participating in the peer review process. "EPA has asked the foxes to guard the hen house on this issue," RFA President Bob Dinneen said. "By adding lawyers and advocates to a scientific review panel, EPA bureaucrats have made a mockery of the Administration's commitment to sound science. These reviews absolutely cannot be viewed as objective or unbiased. Many of these reviewers have repeatedly and openly demonstrated unabashed and politically-motivated biases against biofuels in the past, which immediately casts a long shadow of doubt over the legitimacy of EPA's peer review process."

"This is a perversion of what the peer review process is supposed to achieve," Dinneen continued. "It's little wonder EPA waited until Congress left town for the August recess to release these reviews. EPA cannot feel comfortable that they are getting complete and unbiased feedback based on the panel they have assembled here."