Artificial Intelligent Modeling and Optimizing of an Industrial Hydrocracker Plant

Abstract

The main objective of this study is the modelling and optimization of an industrial Hydrocracker Unit (HU) by means of Adaptive Neuro Fuzzy Inference System (ANFIS) model. In this case, some data were collected from an industrial hydrocracker plant. Inputs of an ANFIS include flow rate of fresh feed and recycle hydrogen, temperature of reactors, mole percentage of H₂ and H₂S, feed flow rate and temperature of debutanizer, pressure of debutanizer receiver, top and bottom temperature and pressure of fractionator column. The network was employed to calculate the flow rate of gas oil, kerosene, Light Naphtha (LN), and Heavy Naphtha (HN).Unseen data points were used to check generalization capability of the best network. There were good adjustment between network estimations and unseen data. Finally optimization was performed to maximize the production volume percent of gas oil, kerosene, HN and LN and to identify the sets of optimum operating parameters in order to maximize yields of mentioned product. Optimum conditions were found as feed flow rate of 90.9 m³/h, reactor temperature of 378.4 °C, hydrogen flow rate of 54.31 MSCM/h and LN (feed vol.%) of 9.34.