Evaluation of CO2 Injection in Shale Gas Reservoirs through Numerical Reservoir Simulation and Supervised Machine Learning

Abstract

CO2 geological storage is an important means to decarbonize the economy, but also expensive in field applications. CO2 injection into shale gas reservoirs can significantly increase the economic incentive by enhancing shale gas production through CO2 preferential adsorption in shales. This work presents a practical framework to evaluate and predict the CO2 adsorption process and CH4 production in shales through a combination of numerical reservoir simulation and supervised machine learning.