2D Network Simulation of Dispersion

The phenomenon of dispersion widely exists in many fields such as groundwater systems and oil and gas extraction, and studying dispersion issues has important theoretical and practical significance. We used particle-tracking method to simulate solute transport in 2D network backbone characterized by pore connectivity and pore-size heterogeneity. Within an individual pipe, solute transport obeyed either of two different flow mechanisms without or with dispersion, namely mean flow and Taylor-Aris dispersion. The longitudinal dispersion coefficient DLM for mean flow is smaller than that value DL of Taylor-Aris dispersion. The relative difference between them decreases with velocity or Péclet number. The mixing rule at nodes has a negligible effect on longitudinal spreading, but has a significant effect on transverse spreading, especially for the nearly homogeneous media. An increase of the disorder in the network diminishes the difference between two mixing rules. The evolution of longitudinal dispersion coefficient over diffusion coefficient presents three different patterns at different velocities for the homogeneous porous media, such as monotonously increasing trend, decreasing first and then increasing trend and monotonously decreasing trend. But all are approximatively replaced by the power law DL ∝ Dmβfor a high disorder, where β decreased with increasing velocity.