Modeling chromatographic separation of produced gas in shale wells

Field evidence exists showing temporal variation in produced gas composition in shale wells. Preferential gas flow and sorption of the species in shale formations cause compositional variations in the gas produced from shale. This process is similar to gas chromatographic (GC) separation, in which the size of gas molecules and their affinity for walls cause separation. As in gas chromatography, shale gas contains molecules of different gases (methane, ethane, propane, carbon dioxide, etc.). When reservoir pressure is greater than critical sorption pressure, sorption process is negligible and the separation process is mainly due to differences in gas molecule speeds in pores. The extremely small size of the pores in shale adds different flow physics, such as Knudsen diffusion and slip flow that intensifies separation of gas components. Understanding and modeling chromatographic separation (CS) in shale can improve our knowledge and help us produce more valuable gas from gas shale. We have developed a numerical model to study temporal variations of the composition of gas produced from shale gas wells. The model is a physical transport model of single-phase multicomponent gas flow in nanoporous media. The governing equations are implemented into a one-dimensional numerical model and solved using a fully implicit solution method. A sensitivity study of the effect of different parameters such as reservoir pressure, length of the system, tortuosity, and permeability on the CS process is performed. The model results confirm strong CS process in shale. In an early stage of production, the component with the highest Knudsen diffusivity and slip coefficient is produced with a higher mole fraction than its in-situ composition. At a later time, the same component comprises a smaller mole fraction than its in-situ composition in the gas production stream. Lower Darcy permeability, a longer system, and higher reservoir pressure intensify the CS process. (C) 2013 Elsevier B.V. All rights reserved.