A 3D multi-mechanistic model for predicting shale gas permeability

Abstract

Shale gas reservoir consists of multiple media, including organic matter, inorganic matrix, and natural fracture. The flow of methane in shale system exhibits multiple mechanisms, such as Knudsen diffusion, surface diffusion, adsorption, and phase behavior. However, previous studies don't take these factors into account simultaneously to predict shale gas permeability. Here, we develop a 3D multi–mechanistic model to overcome this drawback. Using a classification algorithm, we first divide two pore size distributions (PSDs) of organic and inorganic pores. Then we measure the gas permeabilities of different pore spaces, respectively. Regarding organic pores, we consider Knudsen diffusion, surface diffusion, and adsorption. For inorganic pores, we only take Knudsen diffusion into account. Both models consider real gas effect and phase behavior. In natural fractures, we use Poiseuille law to obtain fracture permeability. After that, on the basis of embedded discrete fracture model (EDFM), we incorporate the effect of three pore types (natural fracture, organic and inorganic pores) to measure shale gas permeability. To verify the reliability of our model, we compare the result with the data from documented studies. Through sensitivity analysis, we can draw several conclusions as follows. Natural fracture has the greatest impact on permeability, followed by inorganic pore and organic pore. Owing to the horizontally oriented natural fractures, the permeability in horizontal direction is typically higher than that in vertical direction. As the inorganic pore size increases, permeability increases. While increasing organic pore size, the overall permeability will first drop and then pick up. The difference of the overall permeability caused by phase behavior and real gas effect is minor. Our proposed model can be used as an effective and comprehensive tool to predict shale gas permeability.