Uncertainty analysis at the catchment scale for higher-tier BPR and REACH risk assessments
Characterising and communicating uncertainty is an important step in the evaluation of chemicals. Due to the difficulty in validating models against field data, catchment-based scenarios are often used as benchmarks to test and compare products, using worst-case assumptions. This leads to a disconnect between the risk assessment’s exposure estimates and the likely real-life exposure from actual product use, which makes it difficult to progress into higher tiers, measure uncertainty and review potential benefits from mitigation measures and alternative practices.
There is a common saying, that all models are wrong but some are useful (George E. P. Box). Current regulatory catchment models are useful in providing a standard for substance and product evaluation. However, they are not useful in simulating actual product use and environmental exposure.
A computer model was designed to expand from standard regulatory catchment models to a more complex framework that enables the investigation of alternative behaviours. One example of the use of the model would be to visualise potential daily variation in product use across a catchment. Some variability can be simulated in behaviours within individual buildings, the use of a Box-Muller transform for parameter values which are derived from a range of experimental results, and/or the consideration of alternative treatment options.
Graphical outputs from such models can be a powerful communication tool: by allowing easy visualisation of likely variations, they help risk assessors and risk managers to develop an understanding of the potential uncertainty attached to a range of scenarios and products. This, in turn, would facilitate moving away from benchmark regulatory scenarios and towards more realistic risk assessments that better simulate the link between product use and environmental exposure.
The slides from this presentation are available for free download.