Multiple linear regression based congener profile correlation to estimate the toxicity (TEQ) and dioxin concentration in atmospheric emissions

The formation of polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/F) is governed by complex chemical reactions with complex kinetic models. The simulation of Municipal Solid Waste incinerators, or other industrial thermal processes, is a powerful tool that can be used to optimize and control them, and reducing the number of components to simulate is mandatory for a proper use. In this study it has been determined that only the formation of 3 of the 17 main PCDD/F congeners can be considered as linearly independent. This fact has been used to establish linear regression models that are able to estimate properly the total amount and toxicity of a sample considering only the amount of 1,2,3,6,7,8-HxCDD, OCDF and 2,3,7,8-TCDF. All models have been validated using new samples performing a close approach to the real values provided by complete analytical studies. The average relative error is 3.5% and the maximum relative error is below 9% for these new testing samples. The main goal of our investigation is to build a dynamic simulation process of a MSW facility and include the dioxins formation on it using these models.