Simultaneous Product Loss and Energy Minimization for the Cryogenic Separation System of an Ethylene Plant

The integrated cryogenic separation system of an ethylene plant consists of a multi-stage cold box subsystem and a de-methanizer column subsystem. It separates H2 and CH4 from the charge gas from cracking furnaces through integrated liquefaction and distillation operations. Since the majority of charge gas needs to be liquefied for separation, tremendous energy is consumed in this system; meanwhile, it is also observed that the main product of ethylene can easily enter the fuel gas stream during the cryogenic separation process, which causes significant product loss. Thus, to optimize the operation of the integrated cryogenic separation system is very important to the profitability of an ethylene plant.

In this paper, a general methodology is developed for simultaneous product loss and energy minimization for the cryogenic separation system of an ethylene plant.  Firstly, the simulation model for the integrated cryogenic separation system is constructed and validated.  Secondly, the sensitivity analysis is conducted based on the simulation model to identify the most influential process variables for the ethylene product loss and energy consumption. After that, quantitative relations between those   influential process variables and various manipulated variables, such as the column tray efficiency, heat exchanger efficiency, and compressor power, are modeled with the help of artificial neural network techniques.  Finally, an optimization model based on these surrogate models are developed and solved by an intelligent optimization algorithm. The developed methodology can quickly provide multi-objective optimal solutions that well balance the ethylene product loss and energy consumption of the studied cryogenic separation system.