Bioconversion of Locally Made Cassava Wastewater for Bio-hydrogen Production and Its Statistical Analysis: A Case of Response Surface Methodology (RSM) and Artificial Neural Network (ANN)

Abstract

This study is centered on the production of bio-hydrogen from cassava wastewater and its statistical analysis. Analysis of the sample was carried out by determining the physiochemical properties of the cassava wastewater for its suitability for industrial application. A broth medium was prepared, substrate preparation and inoculum pretreatment were carried out and the medium was cultivated following standard method. To optimize the process condition, response surface methodology (RSM) and artificial neural network (ANN) were engaged. An experimental design was carried out using RSM, three variable factors such as fermentation time (X1) (days), pH effect (X2) and substrate concentration (X3) (mg/L) were considered and 17 experimental runs were created. Results showed the physicochemical properties of wastewater had an initial pH of 5.58 (low acidity), total diffuse solid (TDS) of 3.93 mg/l, chemical oxygen demand (COD) of 0.25 mg/l and biochemical oxygen demand (BOD) of 0.16 mg/l. The statistical analysis by RSM predicted bio-hydrogen yield (HY) of 4.011 ml at X1 = -1, X2 = -1 and X3 = -0.043 variable conditions, this was validated in triplicate experiments, and the average HY was 3.98 ml. Similarly, ANN statistical software predicted HY of 4.221 ml at X1 = - 1, X2 = -1 and X3 = -0.032 variable conditions, this was also validated in triplicate experiments, and the average HY was 4.002 ml. The coefficient of determination (R2) and R-Sq. (adj.) for RSM (99.98% and 99.96%) and ANN (99.993% and 99.986%) indicate that the model fitted well for the acceptable representation of the relationship among the variables under consideration. The results of this experiment established that the use of both RSM and ANN with appropriate experimental design can give the optimum yield of bio-hydrogen, even though, ANN predict better than RSM in terms of yield of bio-hydrogen.