Towards stochastic interpretation of faulted structures using graph theory

in: 2016 RING Meeting, ASGA

Abstract

The interpretation of faults from geological and geophysical data is often challenging, due to observation gaps and to the complexity of geophysical processes in fault zones. As a result, geological interpretation of faulted structures may be ambiguous and under-constrained. Stochastic structural modeling can sample structural uncertainties and may be used in multi-scenarios studies. However, existing stochastic structural modeling methods either neglect some sources of uncertainty or only consider very sparse or specific data types. In this article, we propose a theoretical framework to stochastically interpret heterogeneous data before the construction of a 3D subsurface model. The interpretations are represented by two graphs: the first one gathers data belonging to the same geological structure (G correlation ) and is used to correlate fault observations. The second one (G branch ) details the contacts between these structures. These graphs could be updated during stochastic simulations to generate several interpretations while accounting for geometric and geologic rules.

Download / Links

BibTeX Reference

@INPROCEEDINGS{,
    author = { Godefroy, Gabriel and Bonneau, Francois and Caumon, Guillaume },
     title = { Towards stochastic interpretation of faulted structures using graph theory },
 booktitle = { 2016 RING Meeting },
      year = { 2016 },
 publisher = { ASGA },
  abstract = { The interpretation of faults from geological and geophysical data is often challenging, due to
observation gaps and to the complexity of geophysical processes in fault zones. As a result, geological
interpretation of faulted structures may be ambiguous and under-constrained. Stochastic structural
modeling can sample structural uncertainties and may be used in multi-scenarios studies. However,
existing stochastic structural modeling methods either neglect some sources of uncertainty or only
consider very sparse or specific data types.
In this article, we propose a theoretical framework to stochastically interpret heterogeneous
data before the construction of a 3D subsurface model. The interpretations are represented by
two graphs: the first one gathers data belonging to the same geological structure (G correlation )
and is used to correlate fault observations. The second one (G branch ) details the contacts between
these structures. These graphs could be updated during stochastic simulations to generate several
interpretations while accounting for geometric and geologic rules. }
}