Modeling and optimization of salt removal in Reverse Osmosis System

This study focuses on modeling and optimizing salt removal in reverse osmosis (RO) systems, which are widely used for desalination of seawater. The main objective of the research is to increase the separation efficiency of RO membranes by optimizing key operational parameters such as pH, temperature, pressure, and salt concentration. To investigate the effects of operational parameters on salt removal, total dissolved solids (TDS) removal, and electrical conductivity (EC) reduction, a Regular Two-Level Factorial Design (RTL-FD) and Central Composite Design (CCD) were employed. The data were evaluated using statistical analysis methods, and second-order (quadratic) regression models were developed to predict TO system performance under various operational conditions. Additionally, the Artificial Neural Networks (ANN) method was used to enable more accurate prediction of system performance and to support the optimization process. In the RTL-FD experimental sets, it was observed that pH, pressure, and salt concentration had a significant effect on all dependent variables, while the effect of temperature was limited. In the CCD experimental sets, it was observed that pressure and salt concentration were particularly effective on the dependent variables, while pH and temperature had a limited effect. In the performance predictions made with ANN for RTL-FD and CCD, the correlation coefficients (R2) were calculated as 0.77 and 0.97, respectively. This result demonstrates that CCD provides both higher prediction accuracy and greater reliability in process optimization for non-linear relationships and complex interactions. The models developed using CCD show over 80% agreement with real data for dependent variables. The optimal operating conditions for salt removal in the TO system determined by CCD are defined as pH 9, pressure 40 bar, temperature 45 °C, and salt concentration 10 g/L. In conclusion, the research findings provide important scientific outputs for improving salt removal performance and system optimization in RO systems. When evaluated in line with the United Nations Sustainable Development Goals, the results obtained are expected to provide meaningful contributions to both the literature and industrial applications, particularly under Goal 6, "Clean Water and Sanitation."