Probabilistic health risk assessment of carcinogenic emissions from a MSW gasification plant

Abstract

Health risk assessment due to the atmospheric emissions of carcinogenic pollutants (PCDD/Fs and Cd) from a waste gasification plant is performed by means of a probabilistic approach based on probability density functions for the description of the input data of the model parameters involved in the assessment. These functions incorporate both the epistemic and stochastic uncertainty of the input data (namely, the emission rate of the pollutants) and of all the parameters used for individual exposure assessment through the pathways of inhalation, soil ingestion and dermal contact, and diet. The uncertainty is propagated throughout the evaluation by Monte Carlo technique, resulting in the probability distribution of the individual risk. The median risk levels nearby the plant are in the 10(-8)-10(-10) range, ten-fold lower than the deterministic estimate based on precautionary values for the input data; however, the very upper percentiles (>95th) of the risk distribution can exceed the conventional 10(-6) reference value. The estimated risk is almost entirely determined by the Cd exposure through the diet; the pathways arising from PCDD/Fs exposure are without any practical significance, suggesting that the emission control should focus on Cd in order to reduce the carcinogenic risk. Risk variance decomposition shows the prevailing influence on the estimated risk of the Cd concentration at the emission stack: thus, for a more accurate risk assessment the efforts should focus primarily on the definition of its probability density function.