Hierarchical Bayesian Modeling of Biogas Yield from Agricultural Waste for Energy and Sustainability Outcomes

This study explores biogas production from agricultural waste as a sustainable energy alternative, focusing on regions rich in agro-waste. Using a Hierarchical Bayesian model, the research assesses the energy potential of pig dung, poultry droppings, cassava peels, and cow dung. Data were collected via anaerobic digestion in a 1-cubic-meter locally built digester, where biogas volume, methane content, and thermal efficiency were recorded. Cassava peels yielded the highest energy output (818.4 MJ/day), followed by poultry droppings (506.88 MJ/day), cow dung (348.48 MJ/day), and pig dung (13.64 MJ/day), demonstrating notable variability among feedstocks. The Hierarchical Bayesian Agricultural Yield Model (H-BAYM) captured overall and feedstock-specific impacts on biogas output. The global energy yield intercept was estimated at 417.53 MJ/day (SD = 59.74), with feedstock-specific coefficients ranging from 15.64 to 35.49 MJ/day. Regional effects varied from 41.02 to 63.08 MJ/day, reflecting local differences. The model's Deviance Information Criterion (DIC) of 135.7 indicated a good balance between model fit and complexity. Using Bayesian inference and Markov Chain Monte Carlo (MCMC), parameter uncertainties and interdependencies were reliably estimated. Cassava peels emerged as the most promising feedstock, and H-BAYM offers valuable insights for policymakers to plan region-specific biogas initiatives, advancing renewable energy goals in developing regions.