Evaluation of Methane Yield Prediction Models for Anaerobic Digestion of Agro-Industrial Biomass

The increasing generation of agro-industrial residues reinforces the importance of anaerobic digestion (AD) for renewable energy production and sustainable waste management. However, practical implementation remains limited by the lack of reliable predictive tools capable of estimating methane production prior to laboratory- or pilot-scale trials. Accurate methane prediction models are essential to support substrate selection, process optimization, cost reduction, and techno-economic assessment. Despite their relevance, existing theoretical models frequently show inconsistent performance when applied to complex or lignocellulosic feedstocks. Thus, this study systematically evaluated 13 mathematical models for predicting methane production from 10 agro-industrial residues, comparing theoretical estimates with experimentally determined biochemical methane potential (BMP) values under wet and dry AD conditions. Elementary and biochemical models performed poorly for lignocellulosic substrates and dry AD, with substantial deviations from the measured BMP values. Physicochemical approaches demonstrated greater reliability, particularly for pretreated substrates, where the McCarty model showed the closest agreement with the experimental data. Among the biodegradability correction factors tested, xBD No. 14 consistently improved the predictive accuracy. Overall, methane prediction performance was strongly influenced by substrate recalcitrance, pretreatment, and AD configuration, highlighting the need for more robust models for dry and lignocellulosic systems.