Spontaneous imbibition is a capillary dominated displacement process where a non-wetting fluid is displaced from a porous medium by the inflow of a more-wetting fluid. Spontaneous imbibition strongly impacts waterflood oil recovery in fractured reservoirs and is therefore widely studied, often using core scale experiments for predictions. Decades of core scale experiments have concluded that spontaneous imbibition occurs by a uniformly shaped saturation front and that the rate of imbibition scales with square root of time. We use emerging imaging techniques to study local flow patterns and present new experimental results where spontaneous imbibition deviates from this behavior.
The imbibition rate during early stages of spontaneous imbibition (the onset period) was sometimes observed to deviate from the square root of time behavior. The impact of the onset period on the imbibition process is, however, not well understood. In this work, the development of displacement fronts were visualized during the onset period, using twodimensional paperboard models and core plugs imaged using Positron Emission Tomography (PET-CT). The new experimental results provided insight on the dynamics during the initial spontaneous imbibition period. Controlled two-dimensional paperboard experiments demonstrated that restricted wetting phase flow through the surface exposed to water caused irregular saturation fronts and deviation from the square root of time behavior during the onset period. Local restriction of the wetting phase flow was observed during spontaneous imbibition in sandstone core plugs as a result of non-uniform wetting preference. The presence of nonuniform wetting resulted in unpredictable spontaneous imbibition behavior, with induction time (delayed imbibition start) and highly irregular fronts. Without imaging, the development of irregular saturation fronts cannot be observed locally; hence the effect cannot be accounted for, and the development of spontaneous imbibition in the core erroneously interpreted as a corescale wettability effect. This underlines the undeniable need for a homogenous wettability preference through the porous medium when performing laboratory spontaneous imbibition measurements. Our observations of non-uniform wetting preference will affect Darcy-scale wettability measurements, scaling and modeling. We argue that great care must be taken when preparing core plugs for spontaneous imbibition, to avoid experimental artifacts.