|
(23-05) AMI: Assessment of the Median Impact on Ecosystems based on non parametric tools. Payet, Jerome*,1, Pennington, David1, Jolliet, Olivier1, 1 DGR- GECOS - EPF Lausanne, Lausanne, Switzerland ABSTRACT- Current methods in Life Cycle Assessment (LCA) for toxicological impacts on ecosystems are derived from regulatory-orientated risk assessment (RA) procedures. Improvements are needed to facilitate comparative assessment. The AMI method (Assessment of the Media Impact) offers a toxicological measure to facilitate the comparison of chemicals in applications such as LCA. Using non-parametric tools, the AMI method is the first to assess the median response of species and the associated confidence interval based on bootstrapping. Several points will be outlined in the presentation: 1 - The main differences between the toxicological measures in many RA approaches in LCA: RA is often conservative, whereas LCA is a comparative approach; RA must provide a policy-based threshold level, whereas LCA is based on best-estimates often using linear low-dose extrapolation . RA commonly works to identify a small number of highly toxic chemicals requiring further assessment, often at a local scale, while LCA deals with the emission quantities of many high as well as low toxic chemicals at regional, continental, and global scales; 2 - A non-parametric approach can better facilitate calculation of characterisation factors without the need for assuming a toxicity distribution amongst species (species sensitivity distribution). A characterisation factor based on the median response of species, with the associated uncertainty measures based on bootstrapping, avoids the assumption of a simple parametric distribution such as log-normal, log-logistic, or log-triangular. 3 - The reliability of the extrapolation from acute to chronic: an extrapolation from acute EC50 to Chronic EC50 is more reliable, as the variability due to the experimental design is avoided in comparison to extrapolation to a NOEC. Key words: life cycle impact assessment, aquatic ecosystems, Uncertainty, risk assessment |
