Artificial Neural Network and Mathematical Modelling of the Microwave Drying of Banana Peel Biomass

Fresh banana peels can be converted into bioproducts, biofuels and bioenergy. However, the high moisture content of the peel causes its deterioration and limits efficiencies of thermochemical processes for biofuels and bioenergy production. Hence, the microwave drying characteristics of banana peel was investigated. Banana peel slices were dried in a microwave oven at 200-1000 W and the effective moisture diffusivity, activation energy as well as drying energy requirement were determined. Thin layer drying mathematical and artificial neural network models were fitted to the drying data to describe the drying kinetics. The drying rate and effective moisture diffusivity increased while the energy required for drying decreased with increasing microwave power. The drying occurred mainly in the falling rate period. The effective moisture diffusivities were 1.26 x 10^-9 – 1.83 x 10^-8 m² s^-1 at 200 – 1000 W. The activation energy was 129 Wg^-1 while the total and specific energies required for the microwave drying were 0.17 - 0.37 kWh and 48.2 – 107.5 kWh/kg water removed, respectively. The Weibull mathematical model and an artificial neural network model, which had a single-hidden layer network with 3 neurons in the hidden-layer, well described the microwave drying kinetics of the banana peel biomass.