Hydrodynamic effects of air sparging on hollow fiber membrane filtration in a bubble column reactor

• Main Author: Xia, Lijun
• Other Authors: Law Wing-Keung, Adrian
• Format: Theses and Dissertations
• Language: English
• Published: 2014
• Subjects: DRNTU::Engineering::Environmental engineering::Water treatment
• Online Access: https://hdl.handle.net/10356/59976
• Institution: Nanyang Technological University

Submerged membrane filtration systems have many advantages over conventional water and wastewater treatment systems including better effluent quality, smaller footprint, higher organic loading and lower surplus sludge production. However, membrane fouling remains a challenge. Air sparging is now a standard approach to reduce concentration polarization and fouling of membrane modules in membrane bioreactors (MBRs). The hydrodynamic shear stresses, bubble-induced turbulence and cross flows induced by air sparging all help alleviate the deposition of foulants onto the membrane surface. However, despite its popularity, the detailed quantitative knowledge on the effect of air sparging on membrane fouling remains lacking in the literature due to the complex hydrodynamics generated by the gas-liquid flows around the membranes. To date, there is no valid model that quantifies the relationship between the membrane fouling performance and the flow hydrodynamics induced by air sparging as far as we are aware. This study examines the relationship between the membrane filtration performance and the bubble–induced hydrodynamics around a submerged hollow fiber membrane module in a laboratory scale bubble column reactor (BCR). Membrane filtration tests were first carried out to investigate the relationship in a quantitative manner. The fouling performance was assessed under constant flux conditions with the monitoring of the trans-membrane pressure (TMP). Two axial locations were tested in order to differentiate the effect of flow regimes on the cake formation. Moreover, two aeration schemes of continuous and alternate air sparging were also studied to understand the associated flow dynamics and its influence on the membrane filtration performance. Additionally, factors such as the membrane module orientation, switching interval of alternate aeration as well as the nature of feed solutions were considered in the experiments. Due to various constraints, most measuring techniques for hydrodynamic diagnostic are not applicable in multiphase reactors. In this study, a novel dynamic pressure sensing technique was developed to characterize the hydrodynamics of the BCR with and without the submersion of the membrane module. The transient dynamic pressure fluctuations were recorded at different air flow rates, and special computational algorithms were applied to extract the hydrodynamic characteristics from the dynamic pressure signals. Finally, with the known hydrodynamic characteristics, the correlation with the measured membrane performance in the BCR was explored. In addition to experimentation, computational fluid dynamics (CFD) modelling was also performed to study the time-averaged and transient flow behaviour inside the BCR. Three-dimensional numerical simulations were conducted using the Euler-Euler approach, with both the realizable k-ε turbulence closure and Large Eddy Simulation (LES) to account for the turbulence effects. The BCR with and without the submerged membrane module were simulated respectively. Owing to the importance of the membrane configuration on the filtration efficiency, the membrane module geometry was simulated realistically as twenty solid tubes in this study to take account of the effect of individual fibers in order to improve the accuracy of the numerical predictions of the hydrodynamics around the membrane fibers. Overall, the computer-based method enabled the assessment of the fluid motions over the entire domain of the BCR, which was essential for system diagnostics and optimization. In summary, the membrane filtration experiments confirmed the existence of an optimal air flow rate for the air sparging of a membrane module. Various factors such as the flow pattern, switching interval of alternate air sparging as well as membrane module orientation were all shown to have an influence on the membrane filtration performance. The dynamic pressure sensing measurement technique demonstrated its capability on the hydrodynamic characterization of the BCR. Nonlinear parameters of the Hurst exponent (H) and largest Lyapunov exponent (LLE) can be potentially used to delineate the fluid motions inside the BCR as well as to determine the range of the optimal air flow rate for a membrane filtration system. In addition, the implementation of CFD simulations with specific construction of membrane geometry provided detailed information on both the transient and time-averaged flow behavior inside the BCR, especially the flow along and within the fiber bundle. Therefore, the integrated research of the three approaches enabled the full understanding of the hydrodynamics inside the multiphase reactor and its relationship with the membrane performance, which was rarely reported in earlier studies. The knowledge is useful for the optimization of the air sparger configuration and system operation.