Kinetic analysis and prediction modeling by advanced machine learning of pyrolysis of dairy cattle manure from conventional and organic systems

Alessandro Cardarelli, Michele Ciambella, Pietro Fornai, Marco Marconi, Diego Pennino, Luca Tortora, Marco Barbanera,
Kinetic analysis and prediction modeling by advanced machine learning of pyrolysis of dairy cattle manure from conventional and organic systems,
Biomass and Bioenergy,
Volume 202,
2025,
108247,
ISSN 0961-9534,
https://doi.org/10.1016/j.biombioe.2025.108247.
(https://www.sciencedirect.com/science/article/pii/S0961953425006580)
Abstract: This study investigates the pyrolysis characteristics and kinetics of dairy cattle manure from organic (OSDM) and conventional (CSDM) systems through a comprehensive analysis of thermogravimetric (TG) data combined with advanced machine learning techniques. The results showed that the different diets and manure management systems influence pyrolysis thermal decomposition and kinetics. Also, CSDM releases volatile components more efficiently, suggesting a more favorable thermal decomposition process. Also, kinetic analysis revealed that the average activation energy of CSDM (203 kJ/mol) was slightly higher than that of OSDM (185 kJ/mol), suggesting that the pyrolysis of CSDM requires more energy to initiate the reactions. Moreover, a novel machine learning approach was based on training models on the empirical TG and conversion degree data, allowing for the forecasting of weight loss profiles under varying thermal conditions. For both the manure samples, the predicted TG curves with a heating rate of 25 and 50 °C/min were found to be in good agreement with the experimental ones. While the characterization of pyrolysis properties and kinetics offers vital information for the design and enhancement of reactors, the application of machine learning techniques minimizes the need for extensive empirical data collection, providing a powerful tool to accelerate the optimization of pyrolysis processes.
Keywords: Kinetics; Machine learning; Manure; Pyrolysis; Thermogravimetry; Waste management