Optimising biogas from food waste using a neural network model

Abstract

This study has been carried out to investigate the production of biogas by anaerobic digestion of solid-phase kitchen food waste using an artificial neural network. The network was used to model and optimise biogas production using mixed substrates of food waste with cow dung. The substrate mix percentage, plant pH level, digestion period and digester temperature were used as input parameters for the model, with biogas yield as the output. Food waste and cow dung were mixed at different compositions to a total mass of 2 kg and placed in 21 miniature digesters. The input and output parameters from the digesters were then considered in the model. The highest biogas performance level of 375 ml/g volatile solids on the 25th day of digestion was achieved by a substrate profile of 80% food waste and 20% cow dung at a temperature range of 30-40°C. On the basis of these results, kitchen food waste is shown to be highly biodegradable and an effective source of biogas.