FLOW OF EMULSIONS THROUGH POROUS MEDIA

Abstract

Fluid transport in porous media has been an area of intense research due to its importance in a number of chemical and allied industries. Modeling flow through porous media at the engineering scale usually employs semi empirical equations which assume the medium to be a continuum. Only limited work is available on flow of emulsion (oil and water in different proportions) through porous media. Detailed experimentation is, therefore, needed to understand the single- and multi-phase (emulsion) flow phenomena through porous media. The dissertation reports on the preparation of oil–in–water (o/w) and water in oil (w/o) emulsions, and their stability and rheology. An in-depth study was carried out for flow of Newtonian and non–Newtonian (emulsion) fluids through porous media. Computational study of the single phase fluid flow through an isotropic porous media was carried out using viscous, Reynolds averaging Navier–Stokes (RANS) and large eddy simulation (LES) approaches. The mathematical and statistical approach was adopted to study the formation and stability of o/w emulsion with an integrated hybrid genetic algorithm (GA) coupled with feed-forward back-propagation artificial neural network (BPANN) and response surface methodology (RSM) based on Box–Behnken design (BBD). The hybrid GA model was found to be useful in the optimization of process parameters. The optimum condition predicted by the hybrid GA was 0.913 of emulsion stability index, ESI24, with 4.70% error for 50% (v/v) o/w emulsion, 2% (w/v) surfactant, 5691 revolutions per minute (rpm) of stirring for 5 min time. The rheological behavior of o/w and w/o emulsions with varying internal phase concentrations (10–80%) and an anionic surfactant, i.e. sodium dodecyl benzene sulfonate (SDBS) varying in concentration from 0.5 to 2 w/v %. at different temperatures (25–50 0C) was studied. A controlled stress rotational viscometer is used with shear rates ranging from 1 to 100 s1. The emulsions exhibited typical shear thinning behavior and described well by the power law relationship between shear stress and shear rate. Different viscosity models have been tested and fitted with the experimental rheological data using rigorous-linear/non-linear regression analysis. The Herschel–Bulkley model described the rheological data significantly well. It was observed that the emulsion viscosity and pseudoplasticity increased with an increase in oil concentration. The Herschel-Bulkley and power law models better represented the emulsion rheological behavior as compared to the Casson models. The rise in nonlinearity in Darcy law to describe the evolution of inertial effects in a porous medium has been investigated experimentally. The demarcation limit of the various flow regimes (pre-Darcy, Darcy, transition and non-Darcy) is highlighted, which enabled accurate determination of permeability (K) and form drag coefficient (FD) for different types iv of porous media. It is observed that the flow changes from weak inertial regime to quadratic Forchheimer regime at critical Re in between 5 to 10 and the microscopic inertial effect is the governing factor which leads to the transition from the Darcy to non-Darcy flow regimes. The detailed microscopic pore hydrodynamics in the form of flow streamlines and velocity contours is observed over the Darcy and non-Darcy flow regimes, i.e. 0.02  ReD  30. Further, emulsion (o/w) flow behaviour is investigated for different oil volume fractions (10-80% v/v) and porous media, such as packed bed with four different sizes of glass beads. The rheological properties of emulsions flowing through the porous media have been studied on the basis of shear stress ( w  ) versus shear rate (8us/Rh) relationship. It is observed that the emulsion volume fraction significantly influences the porous bed pressure drop. Pressure drop data d P dL  across the porous bed for all sets of o/w emulsions increases with an increase in emulsion volume fraction, increases with a decrease in particle diameter (glass beads) and an increase in medium porosity. An increase in shear stress and volume fraction during percolation of emulsions through the porous bed is observed. Further, it is found that the effective viscosity of the emulsion for porous bed of different particle diameter decreases with an increase in shear rate and resistance factor (RF) for different porous bed increases with an increase in emulsion volume fraction and decreases with an increase in emulsion injection rate. The pressure drop across the core holder increased with emulsion injection flow rate for all three different sand pack core holders (S-A, S-B, and S-C) studied in the present work. Pressure drop also increases with an increase in emulsion volume fraction due to the increase in emulsion viscosity. It is seen that the permeability ratio across the core holder decreases with an increase in pore volume (PV) of emulsion injected. The Permeability reduction behaviour of porous media is investigated by observing the change in pressure drop across the core holder as a function of number of pore volume of emulsion injected. The permeability reduction across the core holder decreased largely with an increase in emulsion volume fraction and also with an increase in the number of pore volume injected. The laminar flow and turbulent characteristics of flow through an isotropic porous media was studied assuming it to be a representative elementary volume (REV) and comprising an array of square or circular cylinders. Both porous media porosity () and Reynolds number based on particle diameter (Rep), were varied over a wide range. Viscous, Low Re k– turbulence, standard k– turbulence, pseudo-direct numerical simulations (DNS), and large eddy simulation (LES) approaches were used in these studies. The turbulent parameters predicted using low Re k––LB turbulent model were found to be in good agreement with the LES as compared to that with v2f model and the predictions of the pressure gradient were found to be in good agreement with the Forchheimer-extended Darcy law. The low Re–k–– LB model was found to be valid only for flows with Rep > 300.