Multi-objective Optimization of the TEG Dehydration Process for BTEX Emission Mitigation Using Machine-Learning and Metaheuristic Algorithms

Natural gas processing involves the removal of acidic gases, followed by dehydration. The dehydration process is primarily carried out through absorption in triethylene glycol (TEG). During the dehydration process, volatile organic compounds (VOCs) including benzene, toluene, ethylbenzene, and isomers of xylene together known as BTEX present in natural gas are absorbed into the glycol solvent. During the thermal regeneration process of TEG, a substantial amount of BTEX and VOCs are emitted resulting in adverse environmental and health impacts, leading to the need for strict regulations. Effective mitigation of emission is essential for the natural gas industry. There are several process parameters associated with major equipment that may impact BTEX emission. Regulation of some of these parameters is found to have an adverse impact on the dehydration process as well. In this work, a multi-objective optimization is performed in order to find the optimal operating conditions related to the process parameters to mitigate BTEX emission. It is also essential to select the most important parameters from a larger set so that a reliable analysis can be performed using the reduced set of parameters. The analysis is performed using data-driven modeling using machine-learning algorithms, followed by optimization with a probabilistic technique. The least absolute shrinkage and selection operator (lasso) method is used for variable selection; support vector regression (SVR) is used for metamodeling; and a novel metaheuristic optimization algorithm, efficient ant colony optimization (EACO), is used for optimization. The surrogate or metamodel is generated with data obtained from process simulation carried out using ProMax. Using the lasso–SVR–EACO strategy, different optimized sets of operating conditions that minimize the BTEX emission for a stipulated dry gas water content, were obtained. The results show that BTEX released from the dehydration process can be mitigated by optimally choosing the TEG circulation rate, reboiler temperature, stripping gas flow rate, and absorber pressure.