Methods of dealing with co-products of biofuels in life-cycle analysis and consequent results within the U.S. context Michael Wang a,n , Hong Huo b , Salil Arora a a Center for Transportation Research, Argonne National Laboratory, Argonne, IL 60439, USA b Institute of Energy, Environment, and Economics, Tsinghua University, Beijing, 100084, China article info Article history: Received 10 November 2008 Accepted 24 March 2010 Available online 21 April 2010 Keywords: Biofuels Life-cycle analysis Co-products abstract Products other than biofuels are produced in biofuel plants. For example, corn ethanol plants produce distillers’ grains and solubles. Soybean crushing plants produce soy meal and soy oil, which is used for biodiesel production. Electricity is generated in sugarcane ethanol plants both for internal consumption and export to the electric grid. Future cellulosic ethanol plants could be designed to co-produce electricity with ethanol. It is important to take co-products into account in the life-cycle analysis of biofuels and several methods are available to do so. Although the International Standard Organization’s ISO 14040 advocates the system boundary expansion method (also known as the ‘‘displacement method’’ or the ‘‘substitution method’’) for life-cycle analyses, application of the method has been limited because of the difficulty in identifying and quantifying potential products to be displaced by biofuel co-products. As a result, some LCA studies and policy-making processes have considered alternative methods. In this paper, we examine the available methods to deal with biofuel co-products, explore the strengths and weaknesses of each method, and present biofuel LCA results with different co-product methods within the U.S. context. & 2010 Elsevier Ltd. All rights reserved. 1. Introduction Biofuels are being promoted on a global basis because their use may potentially reduce greenhouse gas (GHG) emissions and help achieve energy security by reducing the transportation sector’s use of and reliance upon petroleum (RFA, 2007; EC, 2009a, 2009b; BRDi, 2008; CARB, 2009; EPA, 2009). The pursuit of the various biofuel types and production pathways with true GHG and energy benefits requires examination of the energy use and emission burdens of the whole life cycle of biofuels, covering agricultural chemical production, farming of biofuel feedstocks, biofuel production, and biofuel utilization in motor vehicles. A traditional life-cycle analysis (LCA) is conducted by following each biofuel life-cycle step to derive the total energy use and emission burdens for a given amount of biofuels produced and used. Furthermore, biofuel LCA energy use and emission results are usually compared to those of baseline fuels to be displaced with biofuels so that the relative merits of biofuels can be assessed (Wang, 2008). Several major LCA models are available and studies have been completed with them (see Delucchi, 2003; Brinkman et al., 2005; CONCAWE/EUCAR/JRC, 2007; (S&T) 2 Consultants Inc., 2008). The seemingly straightforward LCA for biofuels is data intensive. Major efforts are continuing to collect the best data to reflect current practice and future trends of biofuel production. Besides, the current debate on the relative energy and environmental benefits of biofuels focuses on indirect changes (such as indirect land use changes) at the global scale as well as direct changes at the local/regional scale that could be caused by large-scale biofuel production (Searchinger et al., 2008). Major regulatory efforts began recently to tackle these complex issues within the LCA context (CARB, 2009; EPA, 2010). In recent years, LCAs for biofuels have been examined closely for their general approach regarding LCA methodology selection between the so-called attributional LCA and consequential LCA (see Ekvall and Weidema, 2004). Traditionally, LCAs for transpor- tation fuels have followed the attributional LCA approach, through which individual processes of a fuel cycle are identified (especially with technology advancements), and the energy use and emission burdens of a given process are allocated among different products. The approach has been developed from conventional engineering analysis of system designs and perfor- mance. It has been advocated and used by many (ISO, 1997; Wang, 1999; Delucchi, 2003; (S&T) 2 Consultants Inc., 2008; CONCAWE/EUCAR/JRC, 2007). On the other hand, the consequen- tial LCA approach takes into account the effects of processes that are directly involved in the generation of a given product and all the indirect effects, such as the secondary and tertiary effects of introducing the product to the marketplace. Historically, con- sequential LCAs were conducted with economic input–output models within an economy (usually within a country), but they Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/enpol Energy Policy 0301-4215/$ - see front matter & 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.enpol.2010.03.052 n Corresponding author. Tel.: + 1 630 252 2819; fax: + 1 630 252 3443. E-mail address: mqwang@anl.gov (M. Wang). Energy Policy 39 (2011) 5726–5736