Geomechanical modeling of three-dimensional geologic structures is helpful for understanding fault related stress perturbations. The choice of which modeling technique to use depends on the available input data and physics that are to be solved, as well as a range of computational considerations such as run time and solution accuracy. For scenarios where fault and fold geometry and offsets are well defined from field or subsurface seismic data, elastic dislocation (ED) and finite element method (FEM) techniques are popular. Both approaches start from a common three-dimensional structural framework model and both produce meaningful stress perturbation results. However, geomechanical modelers must decide when to use a simpler and faster surface-based elastic dislocation model versus when a more robust volumetric finite element model is required. To explore this question, this study compares the ED and FEM approaches for solving a seismic scale (~10km) fault perturbation problem under a static linear elastic assumption. An airtight structural framework for normal faults in a relay ramp setting is built, then in one scenario the faults and horizons are exported for direct use in an ED solution. In a second scenario, a volumetrically reconciled tetrahedral mesh is created and solved using FEM with a surface contact algorithm. Solution accuracy and resolution, run time, and modeling constraints are compared to determine practical guidance for when ED or FEM models are more effective.
Fault related stress perturbations are common in many reservoirs and affect natural fracture orientation and intensity, fault linkage, fluid flow, and rock properties. In order to geomechanically model perturbations (e.g., Pollard and Fletcher, 2005) a structural model must first be created using available geologic information (e.g., Ferrill et al, 2004; Maerten et al., 2006; Hennings et al., 2012; Horne et al., 2020). Field, seismic, and well data are used to interpret faults and horizons, which form the basis to create the three-dimensional structural framework (Figure 1). The fault and horizon intersections are reconciled to constrain the fault geometry and offsets as defined by hanging wall and footwall cutoffs and fault polygons which define the gaps in a faulted horizon (Figure 2). With the fault offsets constrained the surfaces may be re-meshed with triangles or converted to a rectangular grid at a resolution sufficient to capture the geometry of the geologic structures throughout the extent of the model. In some cases, the individual surfaces in the structural framework model can be directly exported for geomechanical simulation. For example, one common geomechanical modeling approach uses the boundary element method to solve for the elastic dislocation (ED) of faults embedded in a half space (Comninou and Dunders, 1975; Okada, 1992; Thomas, 1993; Jonsson, 2002; Maerton et al., 2005; Meade, 2007; Nikkhoo and Walter, 2015). This approach assumes uniform elastic properties and solves for quasi-static stress perturbations surrounding displaced faults computed on an observation grid, typically a stratigraphic horizon or cross section line. The ED method requires low geometric modeling overhead, does not need perfectly reconciled fault and horizon meshes, and the computations run relatively quickly.