Modeling, Simulation and multi-objective optimization of industrial, low-density polyethylene reactor

Abstract

The production of low-density polyethylene (LDPE) at high-pressure using tubular reactors is an important commercial process despite many developments in low-pressure processes such as gas phase and slurry polymerization. The tubular and autoclave reactors at high-pressure currently account for almost 25% of the world-wide polyethylene production. Therefore, even a small improvement of these processes will have a significant impact on the polyethylene industries. In this study, a comprehensive model for an industrial LDPE tubular reactor is developed. The model parameters are tuned using industrial data on temperature profile, monomer conversion and other variables related to polymer quality. Multiple objectives are identified for maximizing the productivity of ethylene and minimizing the side product concentration to improve the quality and strength of the polymer. These multiple objectives are simultaneously optimized using the binary-coded non-dominated sorting genetic algorithm (NSGA-II) and its jumping genes (JG) variants at the operation stage. Thereafter, the multi-objective optimization is also carried out at design stage which includes more decision variables and hence is more challenging. A systematic approach of constrained-dominance principle for handling constraints is applied for the first time in the binary-coded NSGA-II-aJG and NSGA-II-JG, and its performance is compared to the penalty function approach. A three-objective optimization problem with the compression power as the third objective along with the aforementioned two objectives, is also studied. A comprehensive dynamic model for the industrial LDPE tubular reactor is successfully developed. The switching time and amount of off-specification polymer are minimized using multiple criteria with number-average molecular weight and normalized side products as the variables describing quality of the polymer, using dynamic optimization methods.