Design of CO₂ Thickeners Using Genetic Algorithm and Implementation in Enhanced Oil Recovery Simulations

Enhanced oil recovery using CO2-based flooding is a famous alternative to tackle both the problems of environmental pollution and enhanced fuel requirement of modern times, by CO2 sequestration with crude oil. This paper presents a framework for the molecular design of CO2 thickeners using an artificial neural network and genetic algorithm. This framework consists of two parts. The first part is a set of simulations to determine the properties of the thickening molecules. The second part is the framework to design molecules and finally a set of simulations to test the newly formed designs in different pore sizes of calcite material to understand the implementation and performance of these compounds on the ground scale of enhanced oil recovery. The results show that all of the molecules improve the viscosity as compared to systems with pure carbon dioxide and are comparable in performance to the oligomer of polyFast. It is also observed that the molecules are moderately synthesizable with an SA score between 2.5 and 5 (on a scale of 1 to 10, with 10 being the hardest), in comparison to the oligomer of polyFast with an SA score of 3.86. The model and the framework’s output and performance proved satisfactory in such cases where there is no readily available data from experiments to build data-based models for molecular design.