Coupling geomechanics with flow and tracer transport in complex fracture networks




Kumar, Ashish, Ph. D.

Journal Title

Journal ISSN

Volume Title



Hydraulic fracturing in horizontal wells has enabled economic production from ultra-low permeability reservoirs. The productivity of these hydraulically fractured wells depends on the fracture dimensions, conductivity, connectivity to the wellbore, and applied drawdown pressure. Traditional numerical simulation models used to analyze the productivity of hydraulically fractured wells assume a planar bi-wing fracture that is open and connected to the wellbore. However, several core-through field studies and fracture propagation models have demonstrated that a hydraulic fracturing process can create non-planar complex fracture networks. The conductivity and connectivity of these complex fractures are highly dependent on the in-situ stress changes due to production. Hence it is critical to consider complex fractures and the impact of geomechanics in the simulation models for analyzing fractured well productivity. A finite-volume method based geomechanics coupled reservoir model was developed to simulate production from complex fracture networks. An automated meshing method was developed to create the reservoir, and fracture mesh for any given arbitrarily shaped fracture network. The reservoir-fracture network model accounts for fracture closure effects during production. The model developed in this dissertation was used to investigate the impact of drawdown strategy (choke management) on the productivity of wells producing from complex fracture networks. The competing phenomenon of higher initial production rate and faster fracture closure depending on the applied drawdown strategy was observed. Based on NPV maximization, an optimum drawdown strategy can be calculated. The model was also applied to estimate the effective permeability of the SRV (stimulated reservoir volume) to account for complex fractures in upscaled traditional reservoir simulation models. Tracer transport was implemented in the geomechanical reservoir simulation model to analyze the impact of (a) fracture geometry, (b) fracture propagation and closure effects, and (c) fracture complexity on the tracer response curves. An effective model was created to simulate tracer tests in complex fracture networks. Closure of activated natural fractures can explain the multiple peaks in the tracer response curves observed in the field tests. A neural network-based inverse modeling was performed to estimate effective connected fracture length using peak tracer concentration values, peak times, and tracer recovery from chemical tracer flowback data. Observations from the chemical tracer analysis were combined with radioactive proppant tracer and pressure interference tests to diagnose well interference for the Hydraulic Fracturing Test Site #1


LCSH Subject Headings