Coagulation/flocculation process for dye removal using water treatment residuals: modelling through artificial neural networks

Abstract

The potential of aluminum-based water treatment residuals (WTR) discharged from water treatment plants was evaluated as a coagulant for color removal from a disperse dye solution. The effects of WTR dose, initial dye concentration, and initial pH on color removal were studied. The results showed higher color removal at lower pH values. Maximum color removals of 88, 87, and 76% were obtained for initial dye concentrations of 25, 50, and 75 mg/L, respectively, at pH 3.0 with a WTR dose of 3,000 mg/L. Different artificial neural networks (ANN) were developed for predicting the color removal. The performance of the models was found to be very good, with correlation coefficient (R²) values greater than 0.90. The results showed that simulation employing ANN incorporates non-linear behavior of the system, and the model-predicted and observed values of color removals were in close agreement with each other. The study thus indicates that reusing water treatment sludge as a coagulant for color removal would be an attractive option.

Keywords:

• Artificial neural networks
• Coagulation
• Color removal
• Disperse dye
• Water treatment residuals