摘要

The separation of light hydrocarbons (LHCs) by conventional distillation process is highly energy inten-sive. In this work, a novel method was proposed to optimize heat pump assisted distillation process with multiple columns. The MINLP models were formulated in Python to determine the superstructure of the heat and work exchange networks. A genetic algorithm was utilized to find feasible solutions for HPD with minimum total annualized cost (TAC). With the optimal parameters, 36.9% of TAC was reduced after retrofitted by HP. The establishing of HPD with pumparounds was conducted at high DTminand it would contribute to the energy reduction of proposed HPD process. The improved optimization procedure for the DWC process was developed and a high-performance DWC-HPD process was further established with an energy reduction of 60%. MINLP models as well as improved HP/DWC optimization procedure show impressive advantages in the cost and energy reduction of the conventional distillation process.