2D and 3D Finite Element Restorations of Geological Structures with Sliding Contact Along Faults

Guiton, M.L.E. and Zammali, C.
Institut Français du Pétrole, Rueil-Malmaison, France

One of the characteristics of structural geology, when compared to classical engineering sciences, is that the data which are useful for the construction of a sedimentary basin or a reservoir concern the present day configuration only, and are partially available over the full spatial domain. For this reason, the reconstruction of the evolution of a geological structure needs first to restore an initial configuration for the geometry of the interlayer horizons and of the known major faults, before the deformation accommodated either by folding or sliding along fault surfaces. The aim of this restoration is first to provide a quality control on the structural model, implying it to be a component of an inverse procedure which updates the model until some geological plausibility criteria are satisfied. The second aim is to provide the starting point of forward simulations dedicated to the modeling of the development of microstructures controlling the fluid flow properties.
It has long been proposed to compute the restoration through mass balancing of cross-sections with 2D geometrical algorithms, assuming a model of deformation a priori like, for instance, flexural slip in forelands (Suppe, 1983) or simple shear in extensive context (White et al, 1986). Later similar methods have proposed to restore an interlayer triangulated or parameterized surface (Rouby et al, 2000). A common limitation to these methods comes from the lack of generality of the models of deformation, in particular when considering structures with obvious 3D deformation. Recent advances have proposed to alleviate these limitations by using a finite element model with elasticity, to distribute the deformation in volumes. They, however, remain limited to 2D modeling (Maerten & Maerten, 2006), or do not account for the sliding along the faults (Moretti et al, 2006).
In this presentation, we present an evolution of the work of Moretti et al (2006) by accounting for the sliding along the faults in 3D. To the difference of the explicit relaxation method of Maerten & Maerten (2006), the method is here fully implicit, ensuring more stability of the solution with respect to the discretization of the model both in time and space, and allowing for using the same framework for direct modeling with non linear constitutive laws. The sliding condition along the faults prevents both the material penetration and the formation of gap. The advantage is to stabilize the algorithm with respect to unilateral contact conditions, and to facilitate the settings of boundary conditions which defines the solution space out of rigid body movements.
We first discuss the application of this method to a 2D foreland cross-section of a stratified system with sliding along multiple faults and interlayer horizons. The ability of the method to detect an a priori error introduced in the structural model is tested. A second application of the method is the 3D restoration of a Zagros anticline in relation to a variable throw along a thrust. Thanks to the non smoothness of the fault, it is possible to constrain the strike slip component of displacement.
Several perspectives can be envisioned for this work. The restoration method should be linked with geological rules to account for sedimentary or tectonic compaction components of the deformation. The influence of tectonic settings on the lateral boundary conditions imposed during the restoration remains also to be explored. A main advantage of the method is to provide a unique framework for both the restoration and direct basin modeling dedicated to the simulation of the thermodynamic and fluid flow history. The combination of backward restoration and direct modeling should enable the development of an inverse procedure to update the structural model.

AAPG Search and Discover Article #90066©2007 AAPG Hedberg Conference, The Hague, The Netherlands

AAPG Search and Discover Article #90066©2007 AAPG Hedberg Conference, The Hague, The Netherlands