Multi-objective particle swarm optimization of industrial natural gas dehydration process

Dehydration by TEG is widely used in the NGD process to prevent corrosion and plugging of the equipment, valves, and flow lines. In this process, TEG is usually lost in the system due to vaporization and carryover. Therefore, it is necessary to optimize the dehydration process to achieve the allowable water content in the dry gas, and also to minimize heat duty, and TEG losses simultaneously. Therefore, the problem has multi-objective for optimization. The process is simulated by using Aspen Hysys, and the thermodynamic model was glycol-package. Validation of the simulation results was done by comparing with the plant data. Several bi- and tri-objective optimization problems are solved; these problems involved minimization of water content, heat duty, TEG loss and maximization of dry gas flow rate. Wet gas temperature, flow rate, pressure, TEG temperature and mass flow rate, and rich glycol temperature were decision variables. MOPSO algorithm was used for obtaining the Pareto fronts. MOO of the NDG provides a range of optimal operating conditions and objective values; a suitable operating point can be selected based on the higher knowledge like profit.