Modeling multiphase flow in shale

The production of crude oil and natural gas from shale reservoirs has become a significant portion of the total production in the United States. Understanding the multiphase flow mechanisms is vital to the development of shale reservoirs. This dissertation addresses this issue. Firstly, both gas and liquid non-Darcy flow mechanisms in shale formations are investigated separately. For shale gas reservoirs, a sorption model consisting of adsorption and dissolution is proposed to better estimate the total sorbed gas of shale gas reservoirs at high pressure. Then the apparent permeability model is established by combining free gas rarefaction effect and the surface diffusion of adsorbed gas. For shale oil reservoirs, we carefully analyze the published studies related with liquid low velocity non-Darcy flow phenomenon, and discuss the inexistence of Threshold Pressure Gradient (TPG). Then a low velocity non-Darcy model without TPG is introduced. Both the vertical well model and multi-fractured horizontal well model are used to study the production performance of shale reservoirs considering this non-Darcy flow effect. Secondly, the multiphase flow mechanisms in shale formations are studied using pore network modeling techniques. A quasi-static pore network model is established and validated in the Bentheimer sandstone core before being applied in a shale formation. The effects of non-Darcy flow mechanisms are incorporated within this pore network model, some typical results are obtained. Moreover, based on the calculated macro-properties (capillary pressure and relative permeability curves) from the quasi-static pore network model, the analytical spontaneous imbibition model is introduced and a new solution algorithm is proposed. Flowing this workflow, the effects of wettability and absolute permeability on the spontaneous imbibition are studied. Finally, we propose a dynamic pore network model, which directly simulates the process of water imbibition in shale formations. The mathematical model and solution algorithms are presented in details. This proposed model is applied within the Barnett shale formations and a systematical sensitivity studies are conducted. Attempts are made to investigate the water imbibition mechanisms in a micro-scale point of view.