Artificial neural network modelling for the removal of lead from wastewater by using adsorption process
Ayat Hussein Mahdi, Ghaidaa Majeed Jaid, Saja Mohsen Alardhi
Abstract
ABSTRACT For the current work, almond shell was used from activated carbon via the chemical activation with phosphoric acid (85%) by a weight rate of (1,1) to remove lead ion from contaminated water by a batch adsorption process. Activated carbon was characterized by XRD, EDX and SEM analysis. ANN (artificial neural network) modelling was excellent to forecast and check the removal efficiency and the impacts of operational parameters on the removal efficiency. The experimental variables for the adsorption process were utilized as input data for a neural network to train and calculate the output removal efficiency of lead. The activated carbon dosage was the most significant factor in adsorbing technique. The acquired experimentation data was fixed to the absorbing isotherms (Langmuir and Freundlich) and Freundlich model was found to be the closest to experimental data. The removal kinetic modelling used the pseudo-first-order and pseudo-second-order, the pseudo-second-order provided the best description of the kinetic demeanor. The novelty of this work is obtained by using Almond shell in a sustainable manner by predicting the most affecting factor on removal efficiency, which is supposed to be AC dosage, using ANN model.