A parametric fault displacement model to introduce kinematic control into modeling faults from sparse data

Abstract

Fault-related displacements impact oil and gas flow predictions at reservoir scales. In this contribution, we integrate a quantitative description of fault-related deformation directly embedded into the structural modeling workflow. Consistent fault displacements are produced by using numerical fault operators that deform horizons in accordance with theoretical isolated fault displacement models to generate kinematically consistent structural models.We compare structural modeling approaches based on such fault operators with those relying on interpolation. Several synthetic cross-sections are generated from a reference high-resolution structural model of the Santos Basin, Brazil. Models are reconstructed from this 2D synthetic sparse dataset using both methods. Their ability to produce consistent structural models is assessed by comparing reconstructed and reference models. On this example, kinematic modeling improves the quality of automatically generated models when only few and poor quality observations are available, thus reducing the time needed for structural validation.

3-D Structural geological models: Concepts, methods, and uncertainties

“…The amount of fault displacement is determined by interpolation from observation data (e.g. Mallet, 1992;Calcagno et al, 2008), by interactive techniques (Caumon et al, 2003) or using specific near-field fault models (Calcagno et al, 2008;Georgsen et al, 2012;Laurent et al, 2013;Godefroy et al, 2018). As compared to other geological surfaces which always terminate onto other surfaces, fault surfaces can have tip lines not associated to any other surface.…”

Integrating fault kinematics into implicit 3D modelling of fault networks

“…A curvilinear fault frame (gx, gy, gz) (Godefroy et al, 2018;Laurent et al, 2013) is interpolated by first modelling the geometry of the fault surface (gx). This can be achieved by any implicit interpolation method.…”

“…The third coordinate (gz) is orthogonal to both the normal to the fault surface and fault slip direction. The fault coordinates are all normalised for the fault volume so that the minimum value is -1 and maximum value 1 at the tips of the fault (Godefroy et al, 2018). A volumetric displacement is calculated from the fault frame coordinates that defines the magnitude of the displacement caused by the fault.…”

A Triangulated Irregular Network Constrained Ordinary Kriging Method for Three-Dimensional Modeling of Faulted Geological Surfaces

“…Although these methods are widely used in geological applications, an interpolated surface, which is initially constrained by all sample points and then cut and moved by faults, may no longer pass some of these controlled points. Fitting observations to get a geological surface that honors both horizontal and fault observations become difficult [32]. Optimization methods, such as numerical optimization and particle swarm optimization, can be used to move the faulted surface to honor the constraint of observations.…”

“…Optimization methods, such as numerical optimization and particle swarm optimization, can be used to move the faulted surface to honor the constraint of observations. The former finds an ''optimal'' set of parameters for trishear algorithms of fault modeling [33], [34], and the latter chooses kinematic parameters by minimizing a cost function consisting of the misfit between the data observed and the built surface model [32]. However, the complexity of the modeling workflow grows, as numerical calculation and kinematic interpretation is introduced by these methods, and, therefore, they should be implemented cautiously.…”