In this work package of the eteam project different aspects of the tools were evaluated which may lead to an uncertainty of the model estimate, i.e. a deviation of the estimate from the actual exposure value at a workplace. For this purpose a qualitative approach as suggested by the WHO and the corresponding REACH guidance document about uncertainty assessments was used. In order to assess the overall uncertainty, a matrix was developed for each model, which includes a list of all possible sources of uncertainty and a categorisation of different aspects (e.g. transparency and uncertainty of the knowledge base). The different aspects were then discussed and compared. As a result it was shown that for each tool a number of possible sources of uncertainties exists, however, it has to be taken into account that it always depends on the model and the situation itself which sources of uncertainty are dominant and how uncertain the final exposure estimate may be.

In combination with the results of the between user reliability evaluation (BURE) it can be summarised that a clear scenario description including unambiguous parameter definitions is essential for a successfull exposure assessment, since wrong input parameters can easily cause deviations of the result up to three orders of magnitude. Furthermore, indications could be drawn from BURE, which input parameters may have major influence on the overall uncertainty, e.g. dustiness, RMMs and the type of setting, and thus, should probably be improved concerning vagueness and level of detail.

A lower between user variability can be supported by decreasing the vagueness of model input parameters and implemented tool guidance but also has to be accompanied by further measures, e.g. information about the exposure situations, model user trainings. As a consequence, the results of the model uncertainty described in this work package should not be interpreted isolated but only in combination with the other work packages of the eteam project and within the regulatory context.