less of this pollutant than gasoline” are by themselves meaningless and often misleading unless the goal and scope of the study cited in support of this statement are presented. (See Box 5-1 for a description of the importance of care when reporting results from LCA studies.)
A common problem is confusion over two different approaches of LCA—attributional and consequential—and their appropriate use when evaluating biofuels. Attributional LCA, the more traditional form, traces the material and energy flows of a biofuel supply chain and seeks to attribute environmental impact to a biofuel based upon these flows. Consequential LCA, on the other hand, considers the environmental effects of the cascade of events that occur as a result of a decision to produce or not to produce a given biofuel. Many differences between these two approaches of LCA arise because of their distinct applications (Ekvall and Weidema, 2004; Ekvall and Andræ, 2006). Attributional LCA makes use of process-specific or average data, while consequential LCA uses marginal data. Attributional LCA does not consider the market-mediated effects of a given biofuel, such as environmental effects caused by changes in crop or petroleum prices as a result of biofuel production. Consequential LCA, similar to a cost-benefit analysis, includes market-mediated effects. In essence, attributional LCA takes as a given the total environmental effect of all human activities and seeks to assign responsibility for a portion of the effect to a particular biofuel. Consequential LCA also takes as a given the total environmental effect of all human activities, but it assigns to a particular biofuel the change in total effect caused by a decision and the resulting action of whether to implement, expand, or contract biofuel production. As such, attributional LCA is useful in improving efficiency along a biofuel supply chain, and consequential LCA is appropriate in the evaluation of policy and regulation.
Both attributional and consequential LCA make use of knowledge of biofuel supply chains, but conducting the latter is far more complicated as it requires marginal data and modeling of market-mediated effects (Kløverpris et al., 2008; Finnveden et al., 2009). In addition, consequential LCA requires preparation of two alternate scenarios (that is, scenarios that represent “yes” and “no” to a decision) whereas attributional LCA requires only one scenario be described (that is, an actual or a projected scenario). Similarly, when measuring the direct environmental effects of supply chains themselves, attributional LCA can rely on actual, measured data, whereas consequential LCA requires that at least one set of data be estimated: When evaluating policies already fully implemented, one set would have to be estimated (that is, the scenario that did not occur) and when evaluating policies with future effects, two sets would have to be estimated (that is, the scenarios for both the “yes” and “no” to a decision). In total, the uncertainty surrounding the results from consequential LCA is compounded compared to attributional LCA, complicating its use in policy decisions, even where LCA is mandated such as in the Renewable Fuel Standard as amended in the Energy Independence and Security Act of 2007 (RFS2).
This discussion of LCA methodology is important to understanding the environmental effects of biofuels. To date, a large number of studies have used attributional LCA to evaluate individual biofuel production streams and the biofuels industry as a whole. Such studies are helpful for assessing the environmental performance of biofuel supply chains, but they do not consider the broader range of effects from increased biofuel production, such as the effects mediated by markets. Only studies that specifically estimate the environmental effects resulting from the marginal increase in fuel production caused by RFS2 are appropriate for assessing the environmental effects of increasing biofuel production due to its implementation. Studies that have used consequential LCA as a means of quantifying the marginal impact of increased biofuel production are sparse and much needed. In this chapter, results using both methods are presented, with the caveat that what might have been found under one set of circumstances may not hold under other conditions.