Table Optimization of Operating Conditions for a Superwetting Ultrafiltration ZrO2 Ceramic Membrane

Abstract

This paper represents an experimental and modeling study on the performance of a novel superwetting zirconium dioxide ultrafiltration membrane. The effects of different operating conditions were tested using an oily emulsion representing water produced from the Bakken oil reservoir (Canada). The Taguchi experimental methodology was used to predict the optimal operating conditions, and the results were validated experimentally. At optimal conditions, the ceramic membrane showed excellent performance for oil, total organic carbon (TOC), and turbidity rejections of 98%, 99%, and 99%, respectively. The highest flux achieved was 1,007 L/h·m2 at a transmembrane pressure (TMP) of 0.12 MPa (1.2 bar). pH was found to have the greatest contribution to oil rejection. After chemical cleaning, a high flux recovery of more than 97% was obtained. Hermia’s cake formation model had the closest correlation to the experimental data. Finally, an artificial neural network (ANN) model was used to fit the experimental data. The best fit was reached using the Bayesian regularization training algorithm.