Enabling streamlined life cycle assessment : materials-classification derived structured underspecification

Abstract

As environmental footprint considerations for companies gain greater importance, the need for quantitative impact assessment tools such as life cycle assessment (LCA) has become a higher priority. Currently, the cost and time burden associated with LCA has prevented it from becoming more prevalent. While several streamlining approaches have been suggested, questions regarding the effectiveness and efficiency of the streamlined results are still of concern. The streamlining method of probabilistic underspecification has shown initial success in its ability to reduce LCA efforts while simultaneously increasing certainty in the final impact assessment. Probabilistic underspecification streamlines LCA by prioritizing targets of more refined data collection and by implementing the use of underspecified surrogate data within LCI analysis. This thesis concentrates on further developing and improving the streamlining methodology of probabilistic underspecification through refinement of the materials classification systems for polymers and minerals and through additional case study analysis. The classification system allows for a better understanding of the relationship between the degree of materials specificity and the uncertainty in the resulting impact values. Additionally, the resulting polymer and mineral classifications were combined with existing materials classifications to conduct an alkaline battery case study in order to test the effectiveness of the streamlining method. The material classifications created through this research provide a logical and practical approach to underspecification while maintaining consistent and reasonable levels of uncertainty. Furthermore, the case study analysis showed that the streamlining methodology significantly lowered LCA burden by systematically reducing the number of product components requiring full specification. This research provides further evidence that probabilistic underspecification may provide a promising LCA streamlining method among a set of such strategies that can significantly reduce LCA efforts while maintaining the accuracy of the overall impact assessment.