A Data-Driven Water-Soaking Model for Optimizing Shut-In Time of Shale Gas/Oil Wells Prior to Flowback of Fracturing Fluids

Abstract This paper presents a method for identifying the optimum soaking time between the cessation of pumping, and the flowback of hydraulic fracturing fluids after a hydraulic fracture stimulation job, to increase productivity of shale gas and oil wells. Multiple cracks were observed at the surfaces of cores from a shale oil reservoir under simulated water-soaking conditions. The observation proposes a hypothesis that the formation of cracks should increase well productivity. Well shut-in pressure data recorded in a watersoaking process in a shale gas reservoir were employed to derive a mathematical model to describe the process of crack propagation in shale gas/oil formations. This crack model was incorporated in a well productivity model to form an objective function for selection of the water soaking time. A field case was studied with the mathematical model to proof the hypothesis and explore factors affecting the optimum water-soaking time. Analysis of the model shows a quick increase of well productivity with water-soaking time in the beginning followed by a trend of leveling-off. The water-soaking process is mainly controlled by the number of cracks along the bedding plane. High viscosity of fracturing fluid corresponds to longer soaking time, while increasing water-shale interfacial tension reduces the optimum soaking time. The effect of different initial water saturations on optimum soaking time was found to be insignificant. If real time shut-in pressure data are used, this technique can translate the pressure data to dynamic crack propagation data and "monitor" the potential well productivity as a function of water-soaking time.