Mathematical modelling of the decanter for olive oil extraction

The present paper concerns the development of a mathematical model of an innovative horizontal centrifuge for solid-liquid separation for olive oil extraction process. This to know the trend of the operative parameters related to the perfromance characteristics of the machine. A Megala 690 decanter was used for trials in which 2- and 3-phase configuration were analyzed. Different olive paste mass flow rates were considered to evaluate the related extractability and oil loss in pomace values. The experimental data set was divided in two subsets to train and test a 2nd-order polynomial mathematical model. Trained models shown a good generalization capability for both extractability and oil loss values, giving RMSE values ranged 0.549-0.693 and 0.215-0.812, respectively. The application of mathematical models to predict the decanter's behavior could be a useful tool to set the right mass flow rate properly in order to maximize the extraction performances.