The slow kinetics of wettability alteration toward a more water-wetting state by low-salinity waterflooding (LSWF) in oil-brine-rock (OBR) systems is conjectured to be pertinent to the electrokinetic phenomena in the thin brine film. We hypothesize that the nanoscale physicochemical heterogeneities such as surface roughness and surface charge heterogeneity at the rock/brine interface control further the dynamics of electrodiffusion and electrostatic disjoining pressure (Πel), thus the time-scale and the magnitude of the low salinity effect (LSE). In this regard, film-scale computational fluid dynamics (CFD) simulations were performed. The coupled Poisson-Nernst-Planck (PNP) equations were solved numerically in a thin water film confined between a solid surface and oil, both negatively charged. The solid surface is representative of sandstone (quartz/kaolinite) with patchwise physicochemical heterogeneity. The electrical properties of the oil are representative of a crude-oil sample. The OBR system was initially under chemical equilibrium with high salinity (HS) brine, then was exposed to low salinity (LS) brine. The time-scale of reaching chemical equilibrium under LS, and the spatio-temporal evolution of electric potential were investigated. We find that surface roughness (introduced by quartz patches on quartz surface) increases the diffusion time up to 3-fold due to increased tortuosity. However the effect of surface roughness and surface charge heterogeneity (introduced by kaolinite patches on quartz surface) on the effective diffusion coefficient (Deff) is minor. While surface roughness and surface charge heterogeneity affect the disjoining pressure (Πel) significantly, the influence of surface roughness on Πel is more pronounced under HS than LS condition. In contrast, the effect of surface charge heterogeneity is more appreciable under LS than HS. Our findings imply that the LS effect can be enhanced in rough, heterogeneously charged systems like clayey sandstone, although its magnitude depends on the charge density of the roughness and its variation with salinity. We introduce two scaling factors, namely the effective diffusion coefficient (Deff) and the retardation coefficient (ω), to upscale the nanoscale results to pore-scale and beyond.