Methodology for a combined uncertainty analysis and data quality rating of existing graphite datasets in context of battery LCAs

Abstract

Battery raw materials have been identified as one of the main contributors to the overall environmental impact of batteries, however the primary data related to their production remains scarce and is subjected to high uncertainty. In particular, the commercial life cycle inventory datasets related to refining and processing of graphite as an anode material for lithium ion batteries are subjected to in-transparency and are difficult to be traced back. The existing Life Cycle Assessment (LCA) studies on traction batteries make use of these old and partly incomplete graphite datasets without consideration of underlying uncertainties and their effects on the overall environmental impacts. This paper addresses this issue by providing a comprehensive overview of the conventional graphite production processes, a summary of existing graphite datasets, along with their sources and presenting the underlying reasons for the major deviations. Subsequently, the quality of nine different Life Cycle Inventory (LCI) graphite datasets is discussed based on evaluation performed according to EPA data quality guidelines, whereby data gaps are identified and datasets are ranked according to an established data quality rating (DQR) methodology. Furthermore, the paper provides a common base for LCA results of existing graphite production datasets which were remodeled and assessed in Brightway2 (BW2) using background data from a ecoinvent. The results of DQR and LCA reveal major discrepancies in the datasets for the production of battery grade graphite. A Monte Carlo (MC) analysis is also provided to identify uncertainties stemming from technospheric flows, generating result bandwidths. A subsequent case study on state of the art lithium ion battery cell is also presented revealing that the choice of a certain graphite dataset could have a significant impact on the overall LCA results of a battery. The study highlights the need of the graphite datasets to receive more attention in the future.