Ready-to-use and advanced methodologies to prioritise the regionalisation effort in LCA

¹ CIRAIG, École Polytechnique de Montréal, P.O. Box 6079, Montréal, Québec H3C 3A7, Canada
laure.patouillard@polymtl.ca
² IFP Energies nouvelles, 1-4 avenue de Bois-Préau, 92852 Rueil-Malmaison, France
³ UMR 0210 INRA-AgroParisTech Economie publique, INRA, Thiverval-Grignon, France
⁴ CIRAIG, Department of Strategy and CSR, ESG UQAM, Montreal ( Qc), Canada

Received: 21 September 2015
Accepted: 15 January 2016

Abstract

The spatial dimension is an important aspect in life cycle assessment (LCA). Indeed life cycle processes, and therefore also elementary flows, are most likely to be geographically scattered due to global supply chains. The environmental impacts related to an elementary flow can be different across the globe depending on the spatial variability of ecosystem sensitivity. Integrating spatial dimension seems to be a promising way to increase LCA results’ reliability by reducing spatial uncertainties. LCA regionalisation refers to the integration of the spatial variability that really exists to improve result representativeness and reduce spatial uncertainties. As regionalisation requires additional effort for LCA practitioners, it is necessary to ensure the interest of such an effort and to prioritise it. This work proposes two methodologies to prioritise the regionalisation effort in LCA depending on the decision context and the type of study. They allow the selection of impacts, processes and regionalisation aspects that LCA practitioners need to focus on. The first one is a ready-to-use methodology and the second one involves more advanced tools based on spatial uncertainty analysis. The proposed methodologies are stepwise. The ready-to-use methodology involves tools that are already used by LCA practitioners and is based on impact contribution analysis. Regarding the advanced methodology, most relevant impact categories and processes to be further investigated for regionalisation are selected according to their impact contribution and their uncertainty level by using Monte Carlo simulation and regression analysis. The regionalisation effort is estimated depending on the confidence level required for the case study. The relevance and limits of each methodologies are investigated. Results underline the importance to take into account both impact and uncertainty contributions to select processes that need to be regionalised. The proposed methodologies highlight the need to focus on the goal and scope at the early stages of a study in order to clearly identify the intended audience and its requirements regarding study quality. A dialogue with the decision maker during the study would be beneficial. Those approaches can be relevant for other types of uncertainty but should be adapted.