SIMULATION AND MULTI-OBJECTIVE OPTIMIZATION OF COLD-END SEPARATION PROCESS OF AN ETHYLENE PLANT

Abstract

A plant for producing ethylene and other valuable chemicals consists of steam cracking, compression and separation sections. The last section consists of a number of distillation columns for separating cracking products. As many of them operate at cryogenic temperatures, ethylene separation is highly energy intensive. To maximize profit, it is important to reduce losses and energy requirements, which are conflicting and require multi-objective optimization (MOO). In this study, cold-end separation of an ethylene plant is simulated in Aspen Hysys and validated with available industrial data. Using this model and elitist non-dominated sorting genetic algorithm, MOO of the cold-end separation is studied. In recent years, hybrid systems of distillation and membrane separation have been analyzed for improving efficiency of olefin-paraffin separation. Hence, retrofitting distillation columns in the existing cold-end separation of an ethylene plant with membrane units is investigated by optimization for multiple objectives.