Energetic, economic and environmental (3E) multi-objective optimization of the back-end separation of ethylene plant based on adaptive surrogate model

Ethylene separation is an important part in olefin production process, but it brings about high energy consumption and emissions. Therefore, the energetic, economic and environmental (3E) multi-objective optimization of ethylene separation process is of great significance for the sustainable development of olefin industry. Regarding the high computation cost of conventional optimization approach based on process simulation model, a novel optimization framework named multi-objective adaptive surrogate model assisted optimization (MOASO) is introduced. In this framework, adaptive sampling based on sparsity and IGGD improvement is proposed, aiming at promoting the accuracy of surrogate model successively as optimization proceeds. In addition, a modified non-dominant sorting genetic algorithm-II (NSGAII) embedding a density-based local search operator is developed for evolutionary optimization. The MOASO is applied to typical test functions and a practical ethylene separation process. The results show that the proposed framework has highly acceptable optimization performance and computational efficiency, reducing 72% of the calculation burden. Compared with the actual operating condition, the energy consumption and CO2 emission can be reduced by 3.5 × 104 GJ/year and 3.81 × 105 kg/year while increasing the annual gross profit by 4.36 × 105 $/year under a typical optimized condition. In general, the proposed framework can be used as an effective tool for multi-objective optimization, from which insights for clean production and sustainable development in complex large-scale chemical processes could be gained. •A novel Multi-Objective Adaptive Surrogate model assisted Optimization framework (MOASO) is developed.•MOASO shows statistically superior performance with regard to convergence and uniform distribution of Pareto solutions.•3E objectives for an ethylene plant are expected to be improved via using MOASO.