My reading of the user manual and the 2008 IJLCA paper is that USEtox should be used to identify the top toxics from a system. The large uncertainties associated with characterization factors means that results less than three orders of magnitude or so apart are essentially the same. In these cases I am thinking about publishing a list of the top toxic releases (those contributing more than 0.1% of the total impact) and the part of the life cycle where they enter the environment. Is this in line with the recommendations by the USEtox team?
If this is correct, then under what circumstances should numerical results be published?
Greg
Thank you for your question!
For sure, uncertainties need to be taken into account and results rely on the level of available information regarding uncertainties. Hence, it really depends on what you want. Generally, all chemicals assessed within USEtox™ and results flagged as recommended have similar uncertainty and a ranking can be achieved based on this. Results flagged as interim, however, have higher uncertainty and it is recommeded to additionally perform a full sensitivity study for those. In contrast, there is no need to assume or state that characterization factors that are 3 orders of magnitude apart are essentially the same, especially not when you look at chemicals of a similar class of substances within organic compounds.
So far, there is no general opinion of the USEtox™ team that needs to be taken into account, but interpretation of results is rather case study-specifc and up to the user apart from interpreting the units of characterization factors.