A joint inversion approach to characterize subsurface fracture networks based on geophysical and hydrological data

Fracture networks form critical pathways for fluid flow and transport of chemical species. Their characterization is crucial for applications involving storage and extraction of fluids in the subsurface such as unconventional oil and gas, CO2 sequestration, geothermal, nuclear waste storage, etc. Due to the extreme heterogeneity, anisotropy and uncertainties in the subsurface, characterizing fracture networks is a challenge. Just geophysical data (e.g., microseismic information) is insufficient to properly characterize the fracture networks, leading to a lot of uncertainty. One needs to incorporate multiple data streams such as geophysical, flow and tracer data, in order to constrain the fracture networks. In this work, we present a joint inversion framework to characterize subsurface fracture networks using multiple data streams. We first estimate the stochastics of the fracture orientations through a combination of focal mechanisms and clustering analysis of microseismic events. Flow and tracer observation datasets are then used to constrain the fracture lengths/size. The outcome of the proposed methodology is a discrete fracture network (DFN) that models the fracture network in the subsurface as a network of two-dimensional planar fractures in three-dimensional space. The DFN can then be used to predict flow, reactive transport or the state-of-stress in the subsurface.