Biogas production using zirconium and zinc-based nanocatalysts and evaluation using a predictive modeling approach

Abstract

Anaerobic digestion (AD), a method of converting waste into energy, is commonly used in processing various organic wastes. It has been studied and recognized for its effectiveness. This study aimed to quantify the biogas yield from the catalytic co-digestion of rumen contents, and distilled water blended cow dung. This was achieved by fabricating biodigesters for the digestion of the contents. The study was carried out using nine identical digesters. For the biodigester without the catalyst (NC), that is control, the cumulative volume of gas produced during the study was 13,320 mL for 1:3. When 5 % w/w ZrO₂, ZnO was added to the mixtures, the volume of gas increased drastically to 36,537 mL, and 21,944 mL respectively. The experimental dataset obtained after 33 days of the study was used in building the machine learning models. The best-performing model achieved during the training had a correlation coefficient between 0.9795 and 1 for the control, ZnO, and ZrO₂ catalytic loading, and the test correlation coefficient of the test datasets was between 0.9782 and 1. However, the Multilayer perceptron (MLP) model performed best in both the training and testing throughout the whole study having a Pearson correlation coefficient of 1. However, the study relied on a small test dataset of 11 entries. This study has opened possibilities to utilize anaerobic co-digestion technology not only for biogas generation but also to employ machine learning modeling for modeling and understanding anaerobic digestion from cow dung and rumen contents. Furthermore, it contributes to the sustainable development goals by offering an alternative energy source.