MODELING FOR OPTIMUM CONDITIONS OF MEMBRANES FLUX FOR WASTEWATER TREATMENT IN SEASONING INDUSTRY

The seasoning product industry that uses saltwater and freshwater animals as ingredients tends to grow continuously. However, product development has an impact on the environment, such as water pollution. Therefore, this project aims to model and study the optimal flux conditions of the developed membranes for the treatment of wastewater from seasoning products using 3 factors, stirring speed (200-600 rpm), pressure (1-3 bar), and initial concentration (500-1500 mg/L) by the artificial neural network model and the quadratic polynomial regressions.

It had been found that  the neural network model outperformed than the quadratic polynomial regression model because of a higher coefficient of determination (R2=0.964). With relative importance and the decision coefficient of the input variables of the neural network model, it was found that the most important factor was pressure, followed by stirring speed and  concentration. respectively. The optimal conditions for the lowest percentage of decrease in membrane flux were a stirring speed of 238 rpm, a pressure of 1 bar, and an initial concentration of 500 mg/L, resulting in a lowered membrane flux percentage of 36.35.