Modeling of the interaction of a hydraulic fracture with a pre-existing three-dimensional fracture network using a fully-coupled finite element-based poroelastic model
Abstract
The propagation of hydraulic fractures is strongly influenced by pre-existing geological inhomogeneities. Linkage of natural fractures through the growth of hydraulically-induced fractures may significantly change the effective permeability of the rock mass, with implications for hydrocarbon production and environmental impact.
We model the growth of a hydraulic fracture in a domain containing many natural fractures using the Imperial College Geomechanics Toolkit (ICGT). ICGT is a finite element-based, monolithically coupled thermo-hydro-mechanical code, which can model the growth of non-planar fractures in geological media. Fractures are represented as surfaces in a three-dimensional domain, and they grow in response to local changes in the stress field along their tips. The computational model comprises three sub-models that solve for stress, displacement, and fluid pressure throughout the volumetric domain that contains fractures embedded in a linear elastic rock matrix. Fluid leakoff between the fracture and matrix is computed explicitly, via Darcy's law. The mesh automatically adapts to changes in the geometry as the fractures grow. Fracture growth is therefore not restricted to pre-existing paths, but is instead determined by mixed mode failure criteria, incorporating the three fracture tip stress intensity factors (modes I, II, and III). We demonstrate the process of fracture activation, growth, and linkage in domains containing pre-existing natural fracture networks. Fracture growth and the ensuing changes to network permeability are observed to occur before intersection between the hydraulic fracture and the fracture network occurs. The effect of hydro-mechanical fracture growth on rock mass effective permeability is quantified for different scenarios, including various initial fracture intensities, and pre-existing fracture orientations. Results highlight the significant influence that the pre-existing fracture network exerts on the final sweep of injected fluids, and the need to accurately incorporate the mechanical process of growth into discrete fracture network modelling.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.H34A..08T
- Keywords:
-
- 1805 Computational hydrology;
- HYDROLOGY;
- 1822 Geomechanics;
- HYDROLOGY;
- 1829 Groundwater hydrology;
- HYDROLOGY;
- 1832 Groundwater transport;
- HYDROLOGY