Reducing epistemological uncertainty in life cycle inventory

Life Cycle Assessment is very dependent on data of good quality. LCA data should ideally be accessible, relevant and reliable. Common sources of data for LCA-practitioners are databases of different kinds. The use of the information in databases for LCI introduces an epistemological uncertainty—since the system where the data is to be used (the unit process in question for the LCA) may differ from the system where the data was generated (the unit process in the database). Information about data, metadata, is often used to provide knowledge about the applicability and relevance of data, but the practitioner has limited possibilities to evaluate any differences between the required data and the available data. An analysis is made in this article where the problem is treated from two sides, the compiler of information (the database compiler) and the receiver of information (LCA-practitioner). To structure the relevance aspect of LCI-data, a methodology is suggested where the most frequent LCI-emissions are considered to have three different phases; primary emission mechanism, conversion in process, and end-of-pipe measures. An analysis is made for each of these stages with its governing parameters, which regulates the amount of the emission. A table is presented with all the common LCI emissions with their governing parameters. The methodology enables the practitioner to evaluate differences between unit processes and also gives information about the spread in emission factors between different variants of a unit process. The methodology is then applied to cement production, showing how the knowledge of governing parameters should be handled. A set of questions is compiled which elucidates the epistemological uncertainty for cement production.