Calibration of the Anaerobic Digestion Model No. 1 for anaerobic digestion of organic fraction of municipal solid waste under mesophilic condition. (August 2020)

• Record Type: Journal Article
• Authors: Fatolahi, Zahra; Arab, Golnaz; Razaviarani, Vahid
• Is Part Of: Biomass and bioenergy. Volume 139(2020)
• Publication Date: 2020-08
• Subjects: ADM1 -- Model calibration -- OFMSW -- Shannon entropy and genetic algorithms; Biomass energy -- Periodicals; Biomass -- Periodicals; Energy-Generating Resources -- Periodicals; Bioénergie -- Périodiques; 333.9539
• DOI: 10.1016/j.biombioe.2020.105661
• Languages: English
• ISSNs: 0961-9534

Abstract

Abstract: Parameters calibration of the Anaerobic Digestion Model No.1 for various substrates at different operating conditions is an essential key to widely increase the practical application of this model. In this study, the Anaerobic Digestion Model No.1 incorporated in the Matlab/Simulink platform was calibrated to simulate a mesophilic lab-scale reactor fed with the organic fraction of municipal solid waste operated at five different organic loading rates with a constant solid retention time of 25 days. The Shannon entropy was applied as a sensitivity index to identify important parameters as five indicators: biogas flow, methane flow, pH, effluent COD and ammonia. The sensitivity analysis revealed that the parameters of biogas flow indicator, with entropy higher than 0.2 dB, could cover all other indicators' parameters. The 13 important kinetic parameters including maximum specific uptake rate of sugar, propionate, acetate and C4+ (valeric and butyric acids) were estimated using the genetic algorithms. The calibrated parameters were validated with the last 50 days of operation, in which the overloading situation occurred. While the non-calibrated model failed to predict the biogas flow at the overloading situation, the calibrated ADM1 reasonably simulated the decreasing trend of the produced biogas flow. Therefore, the Shannon entropy performed well in recognition of sensitive parameters of the complicated nonlinear ADM1. Nevertheless, there are differences between the simulated and experimental results of pH and effluent COD, which resulted in the decline of accuracy for other products of biogas and methane flow. The cumulative biogas production and average composition of biogas flow were predicted well with rAE of 0.04 and 0.18, respectively. Highlights: ADM1 was calibrated for digestion of OFMSW at various organic loading rates. Model calibration used Shannon Entropy as the sensitivity index (SI). Genetic Algorithms (GAs) was used to optimize the important parameters. The calibrated model predicted well the performance of AD bench-scale.