Effect of Operational Parameters on Microgranular Adsorptive Filtration (μGAF)

Abstract

Low-pressure membrane filtration has been used as a separation technique in water treatment applications for decades and is effective at removing particulates from water. However, it has limited ability to reject soluble materials such as natural organic matter (NOM). NOM can not only cause fouling by clogging the membrane pores or by accumulating on the membrane surface, but can also react with disinfectants to form harmful disinfection by-products (DBPs) and can interfere with other water treatment processes. This study investigated microgranular adsorptive filtration (μGAF), a process that integrates granular media filtration and packed bed adsorption by passing water through a layer of pre-deposited adsorbent on a filter media. With proper choice of adsorbent, μGAF can serve as a pre-treatment process prior to membrane filtration, as the pre-deposited adsorbent layer captures the contaminants on or within the layer and thereby prevents contaminants from reaching membrane. Heated aluminum oxide particles (HAOPs) were used as the main adsorbent in this study. Operational factors affecting the process performance, including flux, applied pressure and solution chemistry (pH, ionic strength and concentration of divalent cations) were investigated. The μGAF process achieved better NOM removal and fouling control with increasing flux, and applied pressure had no effect on the μGAF process. Low pH promotes NOM adsorption onto HAOPs, resulting in better fouling control. Ionic strength had a limited effect on the ability of HAOPs to collect foulants. Adding extra divalent cations to the feed solution reduced fouling slightly, possibly due to the agglomeration of polysaccharide macromolecules promoted by the divalent cations. Particulate and colloidal materials were the key foulants in the μGAF units, and soluble materials such as NOM dominated membrane fouling. Fouling of the μGAF unit could be modeled by the classic blocking models developed by Hermia, with the intermediate fouling scenario having the best fit to the experimental data (R2 ≧ 0.95). Lastly, larger bench scale μGAF/membrane processes were carried out with various types of meshes, membrane materials and feed water quality. In these tests, μGAF achieved good NOM removal and fouling control in all experiments, thereby supporting its potential in full-scale water treatment systems.