An analysis of the effects of changes in Premise databases on the environmental impacts of key products
摘要
Premise has been quickly adopted as a major tool for performing prospective LCAs (pLCA). Premise databases are generated by combining the ecoinvent database and data from climate change (CC) scenarios implemented in Integrated Assessment Models (IAMs). Both the extent of changes made with Premise and the focus on CC raise a question on how they affect LCA results. We aim to give practitioners better insight into how changes may affect results when using Premise databases.
MethodsOur analysis of Premise consists of several steps. We begin by generating three Premise databases, one for 2020, and two for 2050 assuming different future changes: one that represents a continuation of current trajectories and policies (Base) and one that represents assuming ambitious decarbonization strategies (RCP 2.6). We analyze these databases based on 9 industrial products: Electricity, Heat, Road Transport, Steel, Aluminium, Copper, Concrete, Organic Chemical and Inorganic Chemical. We calculate results for each of them for 16 impact categories, although we focus on CC as the main driver of changes in these databases. We then assess impact change the different databases for each product. We do a contribution analysis where we analyze the contribution of 15 large groups of processes to these products.
Results & discussionFor CC, we find that impact changes for most products are related to impact reductions in electricity production and increased use of biomass and Carbon Capture and Storage. We find that Electricity, Heat, Road Transport, Steel and Concrete are primary contributors to their own respective impacts. Other products are instead mainly affected by these groups. Furthermore, the contribution from the Ore & Minerals, Energy resources, Chemicals and Transport product groups is much higher when we include Indirect contributions (contributions ‘connected through’ another group, e.g. production of fuel for Transport), these product groups connect to impacts elsewhere in the system. In other impact categories we find much more varied results, though the uncertainty around these results is high, specifically ‘land use’ shows large changes in impact.
ConclusionsOur approach can highlight products or product groups of interest in large databases like Premise, which can aid in modeling and finding errors. We suggest Premise to be used primarily for assessments of climate change impacts and practitioners clearly communicate versions and settings to make their work reproduceable. Despite us taking a critical stance toward some aspects of Premise, we still recommend using it for pLCA.