Rob Lander and Linda Bonnell from Geocosm gave an invited presentation at the EAGE in Copenhagen regarding the state-of-the-art in reservoir quality prediction methods on June 11, 2018.
Sandstone reservoir quality prediction has moved from a reliance on simple empirical trends (e.g., porosity vs. depth) to include process-oriented forward models. Process-oriented models are more accurate than empirical trends because they simulate compaction and the kinetics geochemical reactions while accounting for composition, texture, and burial history. Additionally, they incorporate models of other rock properties such as permeability and seismic velocities. Although process-oriented models have had considerable success, they do not depict the geometry of sandstones at the pore and grain scale. “Digital sedimentary petrology” models, now under development, consider the 3D geometrical evolution of solids and pores in response to composition, texture, and burial history. These models simulate diagenetic processes with greater rigor compared to existing process-oriented models and provide input for “digital rock physics” simulations that until now have relied upon scanned physical samples. The hope is that coupling these simulation methods will make it possible to accurately predict properties of interest in geomechanics, structural deformation, seismic interpretation, petrophysical analysis, and reservoir simulation for areas away from well control.