Outcrops as Analog of Fractured Reservoirs: Capture Explicit Geometries, Derive Statistics and Model Behaviour

N. J. Hardebol and G. Bertotti
Delft University of Technology, Dept. of Geotechnology, Delft, Netherlands

Abstract

Predicting the mechanical response of specific geologic complex structures from the field has increasing societal and industry interest. For instance, what are the risks of fault re-activation from the production or injection fluids or gas in (abandoned) reservoirs and how can these risks be mitigated? Or, how can small-scale fractures be mechanically extended to enhance reservoir properties while the activation of bigger faults (seismic hazards) is avoided? Although these are practical concerns in regard to subsurface reservoirs, addressing the failure potential of fractures or faults in complex structural settings is a recurring research theme surrounded with numerous fundamental questions. Outcrops serve as learning ground which offer hand-on access to fault-fracture network geometries and which form ideal input for numerical modelling of fluid flow behaviour or for mechanical simulations of re-activation of a pre-existing deformation structures.

Our capabilities to capture geologic structures from the field have become much more quantitative with the help of digital acquisition hardware like tablet or handheld devices and tools like LiDAR-scan and photogrammetry. On the computational side, major advance has been made in discretizing realistic geometries with refined and hybrid meshes and mesh adaption to account for complex deformation structures (e.g. Paluszny and Matthäi, 2009). Despite these great advances in digital acquisition of geologic structures and the meshing and numerical modeling of complex geometries, no functional workflow exists that captures the complexity of field derived structures as direct input for numerical experiments. In lack of a fitting modeling workflow, complex structural models can only be built with arduous effort and is typically avoided.

The conventional approach is to conceptualize a deformation structure into simplified geometries before using them as input in numerical experiments. While conceptual models helped us establishing fundamental principles, they cannot predict the deformation response of actual structures under stress. Addressing the failure potential of fractures or faults in such specific and complex structural settings is a recurring research theme. Explicit description of a geologic structure can be done with a discrete geometric representation. Geologic software like Paradigm's gOcad/SKUA, Baker Hudges JewelSuite™ or MV Move Suite provide necessary tools to build coherent 3D geologic geometry models, but require geometric field descriptions that make up the structure to start with. The building of models from field data requires a tool for collecting discrete geologic features immediately in the field.

Geographic Information Systems (GIS) provide key functionality for digitizing geometries and assigning attribute information and its user interface is increasingly adapted to tablet PCs and use in the field. On the other hand, the design of advanced geometry models requires Computer-Aided Design software, which typically lack adaptation for use in the field. The geologic structural model design software have several tools in common with true CAD software and have the added advantage that functionality supports specific geologic feature types such as faults and horizons and which acknowledge geologic rules. However, also these software are not adapted for use in the field and do not (yet) support the entire 'life-cycle' of a structural description from field acquisition to numerical modeling.

AAPG Search and Discovery Article #120140© 2014 AAPG Hedberg Conference 3D Structural Geologic Interpretation: Earth, Mind and Machine, June 23-27, 2013, Reno, Nevada