Application of anaerobic digestion model No. 1 for simulating fermentation of maize silage, pig manure, cattle manure and digestate in the full-scale biogas plant

Abstract

This article proposes a modification of the Anaerobic Digestion Model No. 1 (ADM1) to simulate biogas production in a full-scale agricultural biogas plant. The structure of the model was modified by dividing carbohydrates into three fractions: starch, cellulose, and hemicelluloses, and by dividing the protein fraction into rapidly and slowly degraded proteins. Lactic acid was also introduced to the model as a component of maize silage. The model was calibrated in several stages. In the first stage, model coefficients were selected based on a review of the literature to ensure that the results of the simulation were logically correct. In the following stage, the results of the sensitivity analysis were used to select parameters that significantly influenced the modeled results. The final values of the parameters were determined by analyzing the model’s accuracy within the 95% confidence interval. The variance of model coefficients, used in the analysis of confidence intervals, was determined based on the calculated optimal values. Even when biogas production suddenly decreased around day 225 of the experiment, due to a lack of fresh substrate, the model predictions were accurate. In real-world processes, a considerable decrease in biogas production is observed only after several days. The value of the Nash-Sutcliff efficiency coefficient (NSE = 0.5665) calculated for the calibration dataset demonstrated that the modified model fit the experimental data well. Among the inhibition constants, the inhibition constants for the uptake of propionic acid (KI_H2_pro = 4.6·10−8 kg COD·m−3), butyric acid, and valeric acid (KI_H2_c4 = 5.0·10−8 kg COD·m−3) indicate that the hydrogen concentration strongly affects the inhibition of acetogenesis. The hydrogen concentration should be monitored because an increase predicts process failure.