Life Cycle Assessment of Lithium-Ion Batteries

Sammanfattning

The rapid growth of electric vehicles (EVs) has increased the importance of understandingthe environmental impacts associated with lithium-ion battery (LIB)production. Life cycle assessment (LCA) is widely used to evaluate these impacts,but previous studies report large variations in results due to differences in methodologicalchoices, such as system boundaries, data sources, and modelling assumptions.In addition, module and pack assembly processes are often inconsistentlyand insufficiently represented in existing literature. This thesis investigates howexisting LIB LCA studies differ regarding methodological approaches, data gaps,and reported environmental impacts, while also suggesting how the modelling ofmodule and pack assembly processes can be improved. The study was carried outthrough three main steps: a state-of-the-art review, benchmarking of selected LIBLCA studies, and the development of a model for module and pack assembly. Thebenchmarking compared studies across several parameters, such as system boundaries,functional units (FUs), electricity mixes, life cycle impact assessment (LCIA)methods, transparency, and environmental impacts across several impact categories.In addition, process mapping and data collection were carried out for module andpack assembly to develop a more transparent and structured modelling approachfor these stages. The results show large methodological variations across the reviewedstudies and the differences strongly influenced the reported environmentalimpacts and often limited comparability between studies. In some cases, variationwithin the same battery chemistry was larger than the variation between differentchemistries, indicating that methodological choices can influence results as much asthe battery chemistry itself. Cathode production was identified as the most commonenvironmental hotspot. The analysis also highlighted significant challenges relatedto data availability and transparency. Many studies relied heavily on secondarydata, while representative industrial primary data remained limited. The findingsof this thesis demonstrate the need for improved transparency, consistent methodologicalapproaches, and better access to representative primary data in order toincrease the robustness and comparability of future LIB LCAs.