EVALUATING THE EFFECT OF ACID MIXTURES AND SOLIDS LOADING ON FURFURAL PRODUCTION FROM SUGARCANE BAGASSE: OPTIMIZATION USING RESPONSE SURFACE METHODOLOGY AND ARTIFICIAL NEURAL NETWORK
Authors: *Amenaghawon N.A, Amagbewan E
A three variable Box-Behnken design (BBD) coupled with response surface methodology (RSM) and artificial neural network (ANN) was employed to optimise and evaluate the effect of solids loading and acid concentration for the production of furfural from sugarcane bagasse. The three variables studied were concentration of hydrochloric acid (0-6%w/w), concentration of sulphuric acid (2-6%w/w) and solids loading (5-15g). Statistical analysis of the results showed that all the variables had significant effect on furfural yield. A statistical model was developed by applying regression analysis to the experimental data. The model was statistically significant (p<0.05) and showed a good fit with the experimental data. Artificial neural network was also used to optimise the production of furfural with several training algorithm used and quick propagation (QP) yielding the best training to predict the furfural concentration. The RSM model predicted optimal levels of 6.00w/w%, 0.00w/w% and 15.00g for HCl, H2SO4 and solids loading respectively. Furfural produced was obtained to be 7.876g/L. The ANN model yielded optimum levels of 5.999w/w%, 0.001w/w% and 14.997g for HCl, H2SO4 and solids loading respectively. Furfural concentration produced at these optimum values was 8.679 g/L. Both models produced a good fitting with the experimental data. However, ANN proved to be a better optimization tool because of its higher R2 value (0.99225) and lower RMSE value (0.16905).
Affiliations: Department of Chemical Engineering, Faculty of Engineering, University of Benin, Benin City, Nigeria
Keywords: Furfural, ANN, Optimisation, RSM, Sugarcane Bagasse
Published date: 2017/12/29