Assessment of thermokinetic behaviour of tannery sludge in slow pyrolysis process through artificial neural network

In the leather industry, tannery sludge is produced in large volume. This study investigated the thermal degradation behavior of tannery sludge using thermogravimetric analysis (TGA). The experiments were carried out in an inert atmosphere using nitrogen gas at varied heating rates of 5, 10, 20, and 40 °C/min in the temperature range of 30-900 °C. For the kinetic parameters calculation, three different models, Friedman, Kissinger-Akahira-Sunose (KAS) and the Ozawa-Flynn-Wall (OFW), were employed. The average activation energy (E_a) obtained from Friedman, KAS, and the OFW methods were 130.9 kJ mol^-1, 143.14 kJ mol^-1, and 147.19 kJ mol^-1 respectively. Along with that, experiment of pyrolysis was accomplished in fixed bed reactor (FBR) at temperature of 400 °C. Biochar produced from FBR had a yield of about 71%. The analysis of gas chromatography-mass spectroscopy shows the different chemical compounds present in the bio-oil containing hydrocarbons (alkanes and alkenes), oxygen containing compounds (alcohols, aldehyde, ketones, esters carboxylic acids and the esters) and the nitrogen containing compounds. The kinetic assessment was complemented by distributed activation energy model (DAEM). In the pyrolysis of tannery sludge six pseudo-components were found to be involved. Furthermore, artificial neural network (ANN) was used to predict the activation energy from conversion, temperature, and the heating rate data. MLP-3-11-1 (Multilayer Perceptrons) described well the conversion behavior of tannery sludge pyrolysis.