Patching Life Cycle Inventory (LCI) data gaps through expert elicitation: case study of laundry detergents

Abstract One approach to address the unavailability of Life Cycle Inventory (LCI) data is the use of a substitute or proxy dataset for certain processes. Although proxies have been regularly used in Life Cycle Assessments (LCAs), there has been no effort to (1) establish any form of guidance for the selection of products/processes as substitutes to fill LCI data gaps and, (2) estimate the associated uncertainty for the proxy choice and use. We explore these issues via a study of Cumulative Energy Demand (CED) in chemical manufacturing in laundry detergents. CED was chosen as the target environmental impact due to its less complex nature of modeling. Expert opinions were elicited through a survey of global LCA practitioners querying best proxy choices to substitute for missing data for five functional chemical groups in laundry detergents. The uncertainty associated with the use of proxies, referred to as the “proxy deviation parameter” (PDP), was quantified for each of the five groups, to be used along with the proxy guidance criteria. The results of the study imply the following: Firstly, this method of proxy selection to address data gaps is feasible. Secondly, the method is more robust and transparent than existing methods of selecting proxies since it provides expert guidance for selecting proxies, and enables the quantification of the associated uncertainty. The study also found that more than 50% of the experts consistently indicated that they would choose a different proxy if total environmental impact were considered instead of CED.