Pore-Scale-Motivated Mechanistic Interpretation and Simulation of Frequency-Dependent Vibration-Assisted Waterflooding for Heavy Oil Recovery

Summary Heavy oil reservoirs contain a substantial fraction of the world’s hydrocarbon resources; yet their development remains constrained by unfavorable mobility conditions that limit the effectiveness of conventional waterflooding. External vibration applied during waterflooding has been shown to enhance oil recovery in heavy oil sandpack systems by up to 24% relative to the nonvibration baseline; however, the simulation framework and pore-scale-motivated mechanistic interpretation governing the frequency dependence of this enhancement have not yet been quantitatively established. The present study addresses both gaps through an integrated Computer Modelling Group (CMG) reservoir simulation and analytical imbibition modeling framework, calibrated against sandpack experiments conducted under no-vibration (NV), 2 Hz, and 5 Hz conditions during the steady-state production regime. Among the candidate parameter classes evaluated, only relative permeability modification successfully reproduces the experimental production behavior across all three conditions. In contrast, both viscosity modification and capillary pressure modification fail, with the latter being structurally incapable of representing the dynamic pressure-production coupling that emerges progressively with increasing vibration frequency. Pore-scale force-balance analysis indicates that capillary desaturation, large-scale vibration-induced pressure mobilization, and turbulent-enhanced flow are the dominant mechanisms among the evaluated mechanisms in this study. A dual-population imbibition model is therefore proposed, in which the fast and slow populations are interpreted as representing distinct transport timescales. The fast oil population is consistent with Newtonian scaling, whereas the slow oil population is consistent with viscoelastic scaling associated with cycle-by-cycle meniscus progression and residual stress accumulation. Independent, unconstrained fits to the two production rate data sets yield transfer-rate coefficients consistent with the two scaling laws in the population-specific pattern predicted by the proposed interpretation, providing preliminary first-order evidence for the proposed framework rather than definitive validation. The relative permeability modifications required to achieve simulation matching provide a continuum-scale representation consistent with the proposed viscoelastic regime imbibition interpretation. This study develops a pore-scale-motivated quantitative framework linking vibration frequency, viscoelastic imbibition dynamics, and continuum-scale recovery enhancement during vibration-assisted waterflooding.