Modeling of Nitrate Adsorption on Granular Activated Carbon (GAC) using Artificial Neural Network (ANN)

Abstract

High concentrations of N-containing compounds in drinking water cause health problems such as cyanosis among children and cancer of the alimentary canal. Therefore, removal of nitrate from water samples is of significant importance from the health and environmental point of view. In this work, the effective parameters on removal of nitrate by adsorption process, which included the amount of granular activated carbon (m), initial concentration (C0), contact time, pH and temperature (T), were investigated. The removal process was monitored using an on-line spectrophotometric analysis system. It was found that the content of adsorption followed decreasing order: m= 10>5>2>1g, C0= 20>15>25>10 ppm, pH=4>7>10>1 and T=25>35>45>55 0C. The three-layered feed forward back propagation neural network was used for modeling of nitrate adsorption on granular activated carbon. The comparison between the predicted results of the designed ANN model and the experimental data proved that modeling of nitrate adsorption process using artificial neuron network was a good and precise method to predict the extent of adsorption of nitrate on GAC under different conditions.