Estimating Hydraulic Fracture Geometry by Analyzing the Pressure Interference Between Fractured Horizontal Wells

Pressure interference measurements in fractured horizontal wells have been used to characterize hydraulic fractures (Kampfer and Dawson, 2016; Roussel and Agrawal, 2017). Past work has modeled this interference using static reservoir gridblocks as a proxy for hydraulic fractures. In this paper, we show that to accurately interpret the pressure response observed in a fractured monitor well, one needs to explicitly model the fractures and their propagation as a compliant discontinuity in the reservoir. A fully-coupled 3-D geomechanical reservoir model which models fractures explicitly as open and compliant channels has been developed to simulate pressure interference between hydraulic fractures in a multi-well pad. Using this model, we simulate dynamic fracture propagation at the treatment well while monitoring pressure at the monitor well. The pressure response inside the monitor well fracture is calculated accurately by explicitly modeling the monitor well fracture as a compliant discontinuity in the reservoir rock. We study the impact of mechanical stress interference between the fractures. The model is then used to simulate and analyze the treatment pressure response observed in a pair of wells in the Permian Basin. Simulation results indicate that hydraulic fracture propagation towards the monitor well results in changes in stress on the monitor fracture. Closure and opening of the monitor fracture is manifested directly as an increase/decrease in pressure in the monitor well fracture. We show that this pressure change in the monitor well is observed primarily because of the elastic effect of mechanically squeezing the monitor fracture by the dynamically propagating hydraulic fracture (not by direct hydraulic communication). As such it is essential to model the compliance of the fractures as has been done in this study. This monitor well pressure response is systematically analyzed to estimate fracture geometry for field data obtained from a Permian Basin well pad. Our representation of the propagating hydraulic fracture and the monitoring well fracture as compliant discontinuities in the reservoir is for the first time shown to be essential to capture the pressure response observed in the field. Previous models have simplified the problem by representing the fracture as static reservoir grid-blocks, and such models are clearly inadequate. Our model captures the impact of a propagating hydraulic fracture on the pressure response observed in a fractured monitor well much more accurately. Such pressure interference analysis can provide operators with a semi-quantitative estimate of hydraulic fracture geometry, relatively inexpensively.