Experimental studies and artificial neural network modeling of hydrogen sulfide removal from wastewater by calcium-modified coconut shell based activated carbon

Abstract

In this study, activated carbon -based adsorbent was prepared from eggshells and coconut shells. The effects of contact time, initial H2S concentration, and the calcium impregnated coconut shell activated carbon (Ca-CSAC) adsorption dosage on the hydrogen sulphide (H2S) removal efficiency and adsorption capacity were investigated. The batch adsorption data obtained from the experimental runs were employed to fit an artificial neural network (ANN) model. An initial optimization was performed to obtain the most suitable number of hidden neurons for training and validation of the ANN. The optimization results show that 16 hidden neurons was the most appropriate choice. The trained ANN was adequately validated and tested with coefficients of determination (R2 ) of 0.99 and 0.95, respectively. The ANN was found to be a robust tool for modeling of H2S removal efficiency by and adsorption capacity on Ca-CSAC under different process conditions.

Keywords

• artificial neural network
• adsorbent
• hydrogen sulfide
• coconut shell
• modelling