Response Surface Methodology for Modeling Bisphenol A Removal Using Ultrafiltration Membrane System

In this work, the effects of various operating parameters (pressure, pH, BPA concentration, and filtration time) toward bisphenol A (BPA) removal via ultrafiltration (UF) membrane system were investigated using response surface methodology (RSM). Historical data design of RSM was used to obtain the interaction between variables and response as well as optimizing the process. The analysis of variance (ANOVA) showed that the third-order polynomial model was significant in which pH and filtration time were identified as significant terms that influence BPA removal. The 3D response surface plots revealed the two-factor interaction between independent and dependent variables. The optimization process of the model predicted optimum conditions of 99.61% BPA removal at 1 bar, pH 6.7, 10 mu g/L BPA concentration, and 10-min filtration time. The predicted optimum conditions for BPA removal were consistent with the obtained experimental values, indicating reliable application of historical data design RSM for modeling BPA removal in UF membrane system.