Fracture Characterization through Multi-Physics Joint Inversion

Stefan Finsterle1, John Kearney Edmiston2 and Yingqi Zhang2, (1)Earth Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, CA, United States, (2)Lawrence Berkeley National Laboratory, Berkeley, CA, United States
Abstract:
Natural and man-made fractures tend to significantly impact the behavior of a subsurface system – with both desirable and undesirable consequences. Thus, the description, characterization, and prediction of fractured systems requires careful conceptualization and a defensible modeling approach that is tailored to the objectives of a specific application. We review some of these approaches and the related data needs, and discuss the use of multi-physics joint inversion techniques to identify and characterize the relevant features of the fracture system. In particular, we demonstrate the potential use of a non-isothermal, multiphase flow simulator coupled to a thermo-poro-elastic model for the calculation of observable deformations during injection-production operations. This model is integrated into a joint inversion framework for the estimation of geometrical, hydrogeological, rockmechanical, thermal, and statistical parameters representing the fractured porous medium.