Biogas and biomethane production via anaerobic digestion: Experimental analysis and modeling using hybrid ANN-GA optimization

The increasing demand for renewable energy and effective waste management has intensified interest in anaerobic co-digestion (AD) as a sustainable method for converting organic residues into biogas and biomethane. However, optimizing the AD process is challenging due to the complex interplay of operational parameters and microbial activity. This study investigates the co-digestion of potato peel waste (PPW) and chicken manure (CM), two abundant agro-industrial wastes, through batch-scale experiments under varying pH (5.0–8.5), temperature (40-65 °C), total solids (10-45%), and PPW/CM mixing ratios over a 20-day period. The highest biogas and biomethane yields (14,256 mL and 143.71 mL/gVS, respectively) were obtained at pH 7.0, 55 °C, and 25% TS with a 70:30 PPW/CM ratio. To optimize process performance, an artificial neural network (ANN) model was trained using the experimental data, achieving high predictive accuracy (R² = 0.962 for biogas and 0.974 for biomethane). The model was coupled with a genetic algorithm (ANN-GA) to identify global optimal conditions, predicted as pH 8.29, temperature 52.12 °C, TS 22.6%, and PPW/CM ratio of 0.555. Experimental validation under these conditions showed minimal prediction errors (1.5% for biogas, 1.72% for biomethane), confirming the model reliability. This study highlights the effectiveness of integrating experimental and ANN-GA modeling approaches for optimizing anaerobic co-digestion and enhancing bioenergy recovery from organic waste.