Summary

This chapter presents an analysis of the uncertainties of damage costs, all the more important because their uncertainties are large. Two methods for the analysis of uncertainties are presented. One is the customary Monte Carlo approach; it is general and powerful, but opaque because it produces only numbers. As an alternative we present an analytical approach that is suitable for multiplicative models, in particular the “uniform world model” (UWM) for damage costs; it has the advantage of being transparent and easy to modify if one wants to test different assumptions about the various sources of uncertainty. We show results, based on a literature review of the various sources of uncertainty in the steps of the damage cost calculation. We find that the uncertainty of damage costs can be characterized, with a sufficiently good approximation, by a lognormal probability distribution with multiplicative confidence intervals around the median estimate μg (a random variable has a lognormal distribution if the distribution of the logarithm of the variable is normal). The width of the confidence intervals is given by the geometric standard deviation σg, such that the 68% confidence interval ranges from μg/σg to μg σg. For the classical air pollutants (PM, NOx, SO2, VOC) we find that σg is approximately 3; for toxic metals we estimate that it is about 4 and for dioxins and greenhouse gases about 5. We also present a simple method for the uncertainty of the sum of damage costs due to different pollutants, for instance the damage cost of a kWh of electricity.