Optimization of polyethersulfone ultrafiltration hollow fiber membrane fouling performances using particle swarm optimization

Abstract

Nowadays, membrane separation has become a popular method for various industries worldwide. Membrane separations are often used in water filtration, food industries and gas separation. For more efficient use of membranes, the control of membrane fouling, which can be defined as fouling requiring reagents is of importance. Fouling of membranes is a significant issue for the efficiency of membrane filtration in wastewater treatment systems. Research on ultrafiltration membranes for water treatment is gaining more attention especially in production sectors. However, in solving the reduction of fouling condition problems, previous studies mostly used an experimentation that varied one of the independent filtration conditions and fixed the others. The common problem is the ultrafiltration process cannot be performed effectively due to non-optimum settings of the filtration conditions. Hence, in order to solve these issues, this study aims to use Particle Swarm Optimization (PSO) to optimize the polyethersulfone (PES) ultrafiltration hollow fiber membrane conditions for oily wastewater treatment to maximize fouling index. In this experiment, five variables were evaluated. They were pH and temperature of feed solution, time, transmembrane pressure and surface area of membrane. In order to minimize the number of experiments but still capable of quantifying the effect of each variables, Response Surface Methodology (RSM) of half factorial design was applied. The experimental plan was based on a combination of high levels and low levels, half factorial designs with resolution V, center points, as well as axial points. Furthermore, the regression models were generated by employing the Design Expert 6.0.5 software and they were found to be significant and valid. Then, the regression models obtained were proposed as the objective functions of PSO to determine the optimal fouling conditions. The MATLAB software was used to code and execute the PSO. Based on the results, the optimal conditions occurred at pH of 11.40, temperature of 32.5 °C, time of 28 minutes, transmembrane pressure of 2.97 bar and surface area of 0.042 m2. The membrane morphology under the influence of different ultrafiltration conditions was investigated via scanning electron microscope (SEM). As a conclusion, the fouling index during the ultrafiltration process of PES hollow fiber membrane has been optimized to reduce membrane fouling. The experimental results of this study can help to reduce the fouling of membranes, thus contributing to a more sustainable filtration system. As a future research direction, the solutions from PSO can be compared with other optimization techniques such as Genetic Algorithm (GA).