Empiric model for the prediction of packaging waste pyrolysis yields

The yield and characteristics of packaging waste pyrolysis products depend on the composition of the input material. The aim of this study is to predict the yield of the different pyrolysis fractions (organic liquid, aqueous liquid, gas, char, inorganics) as a function of the input waste composition. Nine real municipal packaging waste samples and four mixtures of pure materials prepared by the authors have been pyrolysed in a 3.5dm3 semi-batch reactor at 500°C. The pyrolysis yields obtained in these experiments, together with some data about the pyrolysis yields of specific materials taken from the literature, have been used as raw data for developing the prediction model. The model parameters have been obtained by means of multiple linear regression of the experimental data. The accuracy of the predicted values is influenced by the nature of the specific sample; the predicted values are more accurate when mixtures of pure materials are studied than when real complex samples are considered. Anyway, the predicted values are acceptable enough to be a useful tool for designing industrial processes. Additionally, the model is easily used since it only requires a few composition data.