Litho-Based Pore Scale Waterflooding in Carbonate Reservoirs: A Holistic Approach to Predict Displacement Efficiency

Grainstones deposited on carbonate ramps are excellent petroleum reservoir formations and are important for energy needs. Waterflooding is routinely used to augment oil recovery and many carbonate fields have long production histories. Future management of these "mature" assets requires knowledge of how oil production can be sustained and enhanced but requires understanding the pore-scale displacement processes. Despite decades of waterflooding in carbonate oilfields a plausible displacement efficiency prediction is not yet trivial. To evaluate waterflooding economics, it is crucial to know the residual oil saturation (Sor) and where oil is entrapped by capillarity in the reservoir. Microfluidic waterflooding experiments provide a means to visualize pore-scale phenomena within different carbonate minerals (calcite, dolomite, and gypsum) and petrographic textures, to estimate microscopic displacement efficiency. By using analogues of carbonate ramp reservoir-lithologies (in terms of texture, unstructured-irregular pore networks and varied mineralogical compositions) realistic evaluations of displacement efficiency were determined for different mineralogical compositions. The quantitative test results matched closely Arab formation SCAL published data. It was determined that multi-mineralic grainstones undergoing waterflood likely experience contemporaneous imbibition and drainage, giving rise to complex multiphase flow due to the existence of different states of wettability. This wettability contrast induces "capillary jumps" across wettability-boundaries at the interface between different lamina or textures. These "capillary leaps" account for increase in oil recovery as they occur but leave behind bypassed oil. Consequently poly-mineralic arrangements have a lower oil recovery compared to mono-mineralic cases. It was observed that distinct Sor are achieved at different injected pore volumes, despite sharing similar porosity & permeability, thus the relationship between Sor and porosity/permeability is weak. Thus, predicting waterflooding efficiency requires the different carbonate minerals Sor to be incorporated in dynamic simulation.