Marginal and average considerations in LCA and their role for defining emission factors and characterization factors
Reinout Heijungs
Abstract
Abstract Introduction In the literature on LCA, one often finds the terms “marginal” and “average,” often in combination with words like “data,” “process,” “emission,” or “characterisation factor.” However, the meaning of these terms appears to differ between sources. This paper aims to clarify the situation. Critical analysis We review the various definitions and interpretations of the terms “marginal” and “average” in economics, LCI and LCIA. We also study the role of various related terms, such as “linear” and “incremental.” It turns out that the term “marginal” is used for characterizing processes in some sources and for characterizing the data that describes processes in other sources. These two interpretations are shown to differ substantially in a hypothetical example. We also note that the situation in the LCIA literature differs markedly from that in the LCI literature. Conclusion and discussion We propose to distinguish three concepts, marginal, average, and average marginal, and offer verbal definitions, mathematical equations, and a numerical example with a graphical interpretation. We also draw an agenda to research the implications for the attributional-consequential debate, the development of databases and software, and several other topics. This may also help to bring more insights in the continuing controversy on consequential versus attributional LCA.