Optimal Design of the Distillation Process Using the Artificial Neural Networks Method

Distillation is the most used separation operation in chemical industries; the important disadvantage of this process is the high energy consumption without reaching a high level of purity of bioethanol. The objective of this study is to optimize the operating conditions such as the reflux ratio, feeding tray position and column pressure of the distillation process in order to reduce the operating energy consumption and cost. The optimization was performed by the principles of central composite design (CCD) and the Artificial Neural Networks method (ANN) coupled with the desirability function. The optimal neural network configuration for the operating energy and cost has one hidden layer with tree neurons. Two models are developed and used to control and optimize the operating conditions of the distillation process. The operating energy consumption and cost of distillation column were reduced to about 50 %.