Toxicity impacts evolution in the Environmental Footprint methods: an example-based interpretation of the results

In life cycle impact assessment (LCIA), assessing human and freshwater toxicity impacts is particularly challenging. The Environmental Footprint (EF) LCIA method, recommended by the European Commission, results from an effort to address known limitations of available impact assessment models and to enhance transparency of input data. This paper aims to provide an example-based guidance for the interpretation of toxicity impacts with the EF3.1 method, embracing aspects related to life cycle inventory (LCI), LCIA and completeness of coverage of characterization factors (CFs). This study compares the impacts on freshwater ecotoxicity (ECOTOX), human toxicity cancer (HTOX_c), and human toxicity non-cancer (HTOX_nc) characterised by means of EF3.1 and its previous versions EF2.0 and EF3.0. The EF versions refinements entail: new calculation principles, new data sources and the extension of the substance coverage (e.g., inorganics). To guide the interpretation of the results, CFs were tested on a case study encompassing 165 products in 5 areas of consumption: food, mobility, housing, household goods and appliances. The comparison between EF versions shows differences in terms of final LCIA characterised result obtained and of number of characterised elementary flows. Results show that the development of the toxicity methods in EF3.0 and EF3.1 addressed a number of previously identified shortcomings. Main contributing substances and underpinning processes changed among method versions. All drivers of toxicity impacts were analysed individually. Some of the most contributing substances were in common across products and areas of consumption, suggesting that there is ‘background toxicity’ directly associated with recurring processes related to raw materials production (e.g., ferrochromium and copper) and landfill waste treatment (e.g., steel slag). A general interpretation workflow for EF toxicity impacts has been defined based on the case study. LCIA contribution of organics and inorganics should be assessed separately. A correct interpretation shall entail reliability checks at both inventory (e.g., modelling of emissions, time and geographical representativeness) and CFs levels (e.g., quality scores of input data), including the study of individual elementary flows Moreover, the analysis of uncharacterised flows brings insights on possible underestimations of impacts.