Engineered waterflooding modifies chemistry of injected brine to efficiently and environmentally friendly enhance oil recovery. The common practice of engineered waterflooding includes low salinity waterflooding (LSW) and carbonated waterflooding. Among these oil recovery methods, wettability alteration has been perceived as a critical physicochemical process for additional oil recovery. While extensive work has been conducted to characterize the wettability alteration, the existing theory cannot explain the conflict oil recovery between secondary mode (injecting engineered water at the very beginning of flooding) and tertiary mode (injecting engineered water after conventional waterflooding), where secondary engineered waterflooding always gives a greater incremental oil recovery than tertiary mode. To explain this recovery difference, a preferential flow channel was hypothesized to be created by secondary flooding, which likely reduces sweep efficiency of tertiary flooding. To test this hypothesis, computational fluid dynamic simulations were performed with finite volume method coupled with dynamic contact angles in OpenFOAM to represent wettability characteristics (from strongly oil-wet to strongly water-wet) at pore scale to quantify the role of pre-existing flow channel in the oil recovery at different flooding modes. The simulation results showed that secondary engineered waterflooding indeed generates a preferential flow pathway, which reduces recovery efficiency of subsequent tertiary waterflooding. Streamline analysis confirms that tertiary engineered waterflooding transports faster than secondary engineered waterflooding, implying that sweep efficiency of tertiary engineered waterflooding is lower than secondary engineered waterflooding. This work provides insights for a greater oil recovery at secondary mode than tertiary mode during engineered waterflooding at pore scale.