Simulation of Coupled Geomechanics and Multiphase Flow in Naturally Fractured Reservoirs
Timur Garipov, Stanford University, Energy Resources Engineering
Wednesday, October 18, 2017 at 10:30 AM
- Building 3, Rambo Auditorium
- Jack Norbeck
We present a Thermo-Hydro-Mechanical (THM) simulation framework for coupled fluid and heat flow with geomechanical deformation in naturally fractured three-dimensional subsurface formations.
The multi-component, multiphase flow and transport processes of both mass and heat are modeled at the macroscopic scale. The mechanical behavior of the natural fractures is modeled as a contact problem between two computational planes. The system of nonlinear conservation equations and associated constitutive relations is solved using finite-volume discretization schemes for the flow and transport of mass and heat, and a Galerkin finite-element method is used for the geomechanics. We employ fully unstructured grids, whereby a Discrete Fracture Model (DFM) is used to represent the natural fractures as lower dimensional objects. The mechanics and the flow (mass and energy) problems share the same conformal DFM-based unstructured grid. The THM modeling capabilities have been integrated into the Automatic Differentiation based General Purpose Research Simulator (AD-GPRS) developed by the reservoir simulation research group (SUPRI-B) at Stanford.
We demonstrate the simulation framework for a cold-water injection problem into a warm three-dimensional heterogeneous fractured carbonate reservoir. The computational domain reflects main geological features of the reservoir and contains three regions: the first region represents the oil and gas reservoir, the second region is a salt dome, and the third region represents the bedrock. The reservoir part contains explicitly defined fractures of various lengths and heights. Primarily, we focus on the simulation of shear deformation of natural fractures and we study the changes of the fracture network transmissivity as a function of the waterflooding operations.
We discuss the open challenges associated with the DFM-based modeling approach and focus on issues related to the contact problem, the particular THM coupling strategy, and the nonlinear solver.