(506a) Performance Decline Due to Cake Enhanced Concentration Polarization In Cross Flow Membrane Filtration: An Unsteady Electrokinetic Model and Experimental Observation

Membrane separation processes are being widely used in vast array of industries as a single treatment paradigm for separation of colloids, macromolecules, organics and ions. This is either achieved through series of progressively lower molecular cut-off membranes, or through use of nanofiltration process which can remove multiple molecular weight and size solutes. The major problem associated with membranes is fouling which increases the operating pressure, decreases the performance and requires cleaning. All of these directly increase the overall energy consumption and cost of the process. Fouling by colloids is a commonly encountered phenomenon in membrane process and considerable attention has been devoted toward mitigation of fouling. Most of the fouling studies have ignored the concept of electrokinetic transport phenomena through the interstitial space of fouling deposit to explain the mechanism of performance decline. Therefore, the study of fouling mechanism is still important to develop better understanding of the process. The application of nanofiltration membrane for produced water treatment has led us to consider and explain the membrane fouling mechanism due to colloids (silica and clays), organics, and ions. In this context, an unsteady model has been developed to predict the performance decline due to fouling of salt rejecting membranes. The model combines the unsteady colloidal deposition and cake enhanced osmotic pressure phenomenon to predict the permeate flux and observed rejection decline during cross flow membrane filtration of charged colloidal particles and electrolyte solution. The unsteady colloidal deposition is modeled based on mass balance of colloidal particles within concentration boundary layer and cake layer is represented as a swarm of non-interacting, incompressible, spherical charged particles using the Kuwabara cell model approach. The model also considers the development of streaming potential and electroosmotic back flow, due to transport of finite size ions around the charged colloidal particles of the cake layer, in light of classical Levine-Neale model of electrophoresis in concentrated dispersions. The electroosmotic back flow induces resistance to permeation in addition to the hydraulic resistance of the cake layer, referred as electroviscous resistance. The colloidal deposition and electrophoresis analysis are then combined with the standard thin-film theory or Darcy’s law of cross flow membrane filtration. The cake enhanced osmotic pressure in Darcy’s equation is modeled based on the hindered back diffusion of ions through the tortuous path the cake layer. These two phenomena: unsteady colloidal deposition and cake enhanced osmotic pressure are combined together and solved simultaneously to predict the permeate flux and observed rejection. The model considers the average porosity of the cake layer as the only fitting parameter for determining the performance of the filtration process. In order to validate the developed model some nanofiltration experiments were carried out. In these experiments, flux decline due to the fouling, salt rejection and silica mass deposition on membrane were measured for 5 hrs run. It was found out that the developed model can well predict the experimental results with correlation coefficient of higher than 0.95. The novel aspect of the model is it includes the dynamic growth and electroviscous resistance of the cake layer, and couples them with film theory to develop a process simulation model which depends on only one fitting parameter, porosity.