Exploring alternative normalization approaches for life cycle assessment

PurposeThis work investigates the use of alternative approaches to normalization in life cycle assessment (LCA) and shows the relevance of the normalization step in the interpretation of the results of life cycle impact assessment (LCIA) by testing the use of five alternative normalization sets.MethodsFive normalization sets are applied and compared. The five sets are based on the following: (i) a production-based approach at global and (ii) at EU level; (iii) a consumption-based approach at EU level based on process-based LCA; (iv) a consumption-based approach at EU level based on environmental extended input/output; and (v) a planetary boundaries-based approach. The five normalization sets are applied to the environmental impacts of 144 products, and the resulting normalized impacts are aggregated into a single score by adopting two alternative weighting sets to investigate how the adoption of different normalization and weighting sets can affect the interpretation of LCIA results. The relative contribution of each impact category to the single score is derived and the ranking of impact categories is compared for each normalization and weighting option.Results and discussionThe relative contribution of the impact categories to the aggregated score of a product is significantly affected by the choice of the normalization set and to a lesser extent by the application of different weighting sets. The main benefits and limitations of each normalization approach presented are discussed together with their implications on the interpretation of the results deriving from the application of each set.ConclusionsThe dominating role of the normalization step on the interpretation of the results emphasizes the need to choose the most suitable normalization set according to the goal and scope of the study and to make sure that normalization references are based on comprehensive inventories of emissions and resources, well aligned with the impact assessment methods used in terms of coverage and classification, to avoid the risk of biased normalization. Future research needs for developing more robust and comprehensive normalization sets are identified.