Optimization of CO2 contents and energy saving on sweetening gas processing plant using Particle Swarm Optimization (PSO) algorithm

Natural gas is the energy source that is commonly used today. The high of energy requirement triggers the natural gas processing industry to increase its production by opening new fields. Natural gas is classified as sweet gas if CO₂ and H₂S contents are less or equal to 2% and 4 ppm, respectively. Therefore in this study, the optimization have been performed and focused at amine regenerator because of had a great impact to CO₂ capture or gas sweetening process. Amine column regeneration was modeled using neural network MultiLayer Perceptron (MLP) with Nonlinear Auto Regressive with eXternal input (NARX) structure and learning algorithm using Levenberg-Marquardt. In this system, CO₂ removal ratio is influenced by operating variables such as the concentration of feed column regeneration, temperature condenser, and temperature at bottom regenerator, which is in-turn determined by reboiler duty in the amine regenerator. The process has been optimized using Particle Swarm Optimization (PSO) for obtaining CO₂ mole fraction in the lean amine and the energy load in reboiler and condenser are minimized.