Experimental investigation of two phase flow in porous media: Effects of surfactants on immiscible displacement processes at the pore network scale

Glass micromodels are useful tools for visualization of multiphase flow at the pore and pore network level. They are especially useful for observation of immiscible displacement processes, including the displacement of water by air during remediation using air sparging, or during infiltration or drainage of non-aqueous phase liquids (NAPLs). Detailed steps in the design and construction of glass micromodels used for visual observation of the air sparging process at the pore network level are presented in this study.; The application of the micromodels is demonstrated by reporting the results of air displacement tests in the micromodel in the presence of water, and a water/surfactant solution. The tests were performed at six different airflow rates, and demonstrated that the final air saturation in the micromodel was increased approximately 40% in the presence of the anionic surfactant sodium dodecyl sulfate (SDS). The displacement process in the presence of SDS was also vastly improved as relatively uniform air distributions occurred and air channel flow was significantly reduced. Additionally, the air breakthrough time decreased and the steady-state air saturation in the micromodel increased as the air flow rate was increased, as was anticipated.; The movement of the non-wetting air through the micromodel pore network was simulated using a computer network model. The advancement of the air/aqueous interface was simulated as a discrete stepwise process in which the interface advanced through the network element with the least resistance (lowest threshold pressure) to invasion by the non-wetting fluid. According to the fluid configuration, the pressure field in the network was calculated and the pore air saturation was updated in each time step. The process was thus repeated and the model simulation was stopped once only residual water or water/surfactant remained in the model. The numerical simulation results agree well with the experimental data.