A modified approach for modeling two-phase flowback from multi-fractured horizontal shale gas wells

Historically, high-frequency fluid production and flowing pressures (hourly or greater) have been gathered on nearly every multi-fractured horizontal well (MFHW), although this data has rarely been used by industry in a quantitative manner to characterize hydraulic fracture or reservoir parameters. Recently several authors have recognized the potential to extract key properties from this early-time data. This work will expand on the analytical flowback analysis model presented by Clarkson and Williams Kovacs (2013a) and modified by Williams-Kovacs and Clarkson (2013a). These works presented a data driven pseudo-analytical modeling approach for quantitatively analyzing two-phase flowback to estimate key frac properties including fracture conductivity and half-length. In these early attempts to model flowback from shale gas wells, multi-phase depletion from the fracture network was assumed to be the primary flow-regime. More recently, three flow-regimes have been observed in flowback data depending on data frequency: 1) transient flow of frac fluid within the fracture network prior to breakthrough of formation fluids (rarely seen in shale gas); 2) Single-phase depletion of water within the fracture network (also rarely seen); and 3) coupled formation and fracture flow following breakthrough of formation fluid. In this work, flow-regimes 1-3 are modeled rigorously. Flow-regime 3 is the focus of this study and will be modeled using a coupling of transient linear flow of gas from the matrix to the fractures with multi-phase depletion within the fracture network (conceptually more realistic than previous attempts). Further, dynamic fracture porosity and permeability are incorporated to better represent the physics of the problem and a modified material balance equation (MBE) is developed to account for additional drive mechanisms (fracture closure in addition to desorption and gas expansion). Finally a modified pseudo-pressure and pseudo-time are applied to before-breakthrough (BBT) single-phase rate transient analysis (RTA) and after-breakthrough (ABT) multi-phase RTA (conducted after analytical simulation using pressure and saturation dependent outputs from simulation) to improve both parameter estimates (BBT) and flow-regime confirmation (ABT). The new model is compared to the previous model using a field case study. (C) 2016 Elsevier B.V. All rights reserved.