Ryberg, Morten1; Vieira, Marisa D. M.3; Zgola, Melissa4; Bare, Jane5; Rosenbaum, Ralph K.1
1 Department of Management Engineering, Technical University of Denmark2 Quantitative Sustainability Assessment, Department of Management Engineering, Technical University of Denmark3 PRé Consultants B.V.4 Quantis International5 U.S. Environmental Protection Agency
When LCA practitioners perform LCAs, the interpretation of the results can be difficult without a reference point to benchmark the results. Hence, normalization factors are important for relating results to a common reference. The main purpose of this paper was to update the normalization factors for the US and US-Canadian regions. The normalization factors were used for highlighting the most contributing substances, thereby enabling practitioners to put more focus on important substances, when compiling the inventory, as well as providing them with normalization factors reflecting the actual situation. Normalization factors were calculated using characterization factors from the TRACI 2.1 LCIA model. The inventory was based on US databases on emissions of substances. The Canadian inventory was based on a previous inventory with 2005 as reference, in this inventory the most significant substances were updated to 2008 data. The results showed that impact categories were generally dominated by a small number of substances. The contribution analysis showed that the reporting of substance classes was highly significant for the environmental impacts, although in reality, these substances are nonspecific in composition, so the characterization factors which were selected to represent these categories may be significantly different from the actual identity of these aggregates. Furthermore the contribution highlighted the issue of carefully examining the effects of metals, even though the toxicity based categories have only interim characterization factors calculated with USEtox. A need for improved understanding of the wide range of uncertainties incorporated into studies with reported substance classes was indentified. This was especially important since aggregated substance classes are often used in LCA modeling when information on the particular substance is missing. Given the dominance of metals to the human and ecotoxicity categories, it is imperative to refine the CFs within USEtox. Some of the results within this paper indicate that soil emissions of metals are significantly higher than we expect in actuality.
Clean Technologies and Environmental Policy, 2014, Vol 16, Issue 2, p. 329-339