Towards a better consideration of the environmental footprint of complex industrial systems through an LCA approach based on operating scenarios.

Summary Life Cycle Assessment (LCA) is commonly used in companies as an environmental assessment tool. But it is also associated with a certain number of limitations, such as the lack or quality of the data collected. This problem is amplified when dealing with complex industrial systems, characterized by a large number of subsystems and components and an extremely long and uncertain life cycle. Different methods exist to precisely manage data uncertainty, but they are generally very time and resource consuming. When the objective of LCA is to guide the research of improvement projects, a high accuracy of the results is not necessarily necessary. We propose here a scenario-based LCA approach allowing to model with little information different possible operating trajectories of a complex industrial system. From the identification of particularly uncertain and relevant life cycle elements to be integrated (maintenance, spare parts, accidents, upgrading, life cycle modulation, end-of-life options), operating scenarios are built from an adapted tool, the SRI matrix (Stanford Research Institute). It can be used to quickly generate a set of four scenarios (worst case, best case, probable future and a different alternative) to understand the extent of possible environmental impacts depending on the operator's decisions and natural or political-economic hazards. The simulation of these scenarios allows to make the decisions taken on the basis of the LCA more reliable. The whole approach has been applied at Alstom Grid on AC/DC conversion stations for the primary aluminium industry.