Evaluation of Selection Methods for use in Life Cycle Impact Assessment

Today very few LCA studies include ecotoxicity and human toxicity in the impact assessment and if they do it is typically highly incomplete. The reason for this seems to be that in many cases an extremely high number of chemical emissions from the inventory potentially contribute to the toxicity related impact categories and only for a small part of them there are characterisation factors provided by the applied impact assessment method. This calls for a method that is able to select/prioritise those chemical emissions that contribute significantly to the toxicity related impact categories. Such a method is called a selection method and its overall aim is to focus the effort on significant chemical emissions when Life Cycle Impact Assessment is done on toxic releases. Today experience from application of the few existing selection methods is very sparse and the need for research within this area therefore seems urgent. This paper will present the result of a comparison between different selection methods (e.g. CPM-selection and priofactor) including a partial order ranking method called Hasse diagram technique. Furthermore a characterisation method (EDIP) is included in order to compare the ranking of the selection methods with the results of a characterisation. The data used for this comparison comprises a test set of around 80 different substances covering all relevant combinations of different substance properties through representatives of different substance groups, i.e. non-dissociating organics, dissociating organics, amphiphilics, metals and other inorganics. This test set has been developed within the EU project OMNIITOX for a structured comparison of characterisation methods and selection methods. The comparison includes an identification of differences in ranking between the different methods and an analysis of the causes to the observed differences.