Optimizing biogas production using artificial neural network

Abstract

The object of this study is the operating parameters of the anaerobic digestion unit. The study aims to increase the potential of biogas production. The task to select the optimal parameters of the working process of anaerobic digestion has been solved. A model of cumulative biogas and methane output in the process of anaerobic waste digestion has been constructed, which is conceptualized using the method of artificial neural network. The model is built on the basis of 11 process-related variables, such as hydraulic retention time, pH, operating temperature, etc. The plant parameters, leading to the highest volume of biogas production, were selected. It was determined that the optimal amount of biogas can be brought to 90 %, which exceeds the maximum value obtained from factory records by 12.6 % to 700 m3/t. Working conditions that led to optimal methane production were defined as the temperature of 39 °C, the total solids of 4.5 %, the organic percentage of 97.8 %, and pH 8.0. It was found that biogas production is the highest at temperature within the thermophilic range while the local maximum is achieved in the mesophilic temperature range. The model built serves to determine the optimal operating parameters for maximum biogas production and could be used to optimize biogas production productivity using limited experimental data. The model also makes it possible to predict the performance of anaerobic digestion under untested conditions. It is possible to practically use the developed biogas production model when monitoring the operation of the anaerobic digestion unit, to increase the efficiency of the process, and when adjusting the working conditions of the methane tank