The Energy Potential of Residual Biomass Gasification Integrated with Internal Combustion Engine in Córdoba, Colombia using Artificial Neural Network Techniques

Abstract

In this study, the energy potential of waste biomass gasification integrated to internal combustion engine in Cordoba, Colombia was investigated using artificial neural network techniques. A model was trained with proximate and elemental analysis data of different biomasses and this model was used to estimate the gasification potential of the four most abundant biomasses in Cordoba. The model developed achieved an adjusted determination coefficient (R2) of 0.9293 for validation and 0.9048 for training, demonstrating high predictive accuracy. The results indicate that temperature positively influences energy generation potential, while moisture content and air-to-fuel ratio have a negative impact. Among the biomass types analyzed, cassava stands out with the highest energy potential, exceeding 9 GWh/year, followed by plantain at approximately 3 GWh/year, maize cobs below 2 GWh/ year, and rice husk with <0.5 GWh/year. These findings provide critical insights for optimizing biomass gasification processes and harnessing regional biomass resources for energy generation.