ANN-GA, ANFIS-GA and Thermodynamics base modeling of crude oil removal from surface water using organic acid grafted banana pseudo stem fiber

Abstract

Banana pseudo stem fiber adsorbent (BPSF) with well-ordered framework of porosity was fabricated via esterification reaction between organic acid and the biomass. The resulting adsorbent (BPSF) was applied for the removal of crude oil from water surface. Optimization and modeling of the process parameters was done using response surface methodology (RSM), artificial neural network-genetic algorithm (ANN-GA), and adaptive neuro-fuzzy inference system-genetic algorithm (ANFIS-GA). Result shows that the optimum removal of oil occurred at oil water ratio of 0.4 g /100cm3 with 94.8% oil removal. Meanwhile, BPSF exhibited high adsorptive potential at a very low pH of 4 with 95.12% oil removal. Thermodynamic studies revealed activation energy, change in enthalpy and change in entropy of the process as (19.55, 24.68, -0.438 KJ/mols) and (53.82,31.66, -0.186 KJ/mols) indicating non-spontaneous process. Equilibrium modeling revealed that the composite material was highly matched to Langmuir isotherm model with regression coefficient > 98% with maximum adsorption capacity of 53.26 mg/g. Optimization of the process shows the optimum conditions as 100 °C, 0.4 g/100cm3, 2.1 g, 6 and 75 mins for temperature, oil concentration, adsorbent dosage, pH and time. RSM, ANFIS, and ANN models adequately predicted the oil removal with correlation coefficient > 0.97 but statistically, ANN was the best model followed by ANFIS and RSM models. Results obtained from this investigation has shown that esterified banana pseudo stem fiber composite is an efficient, economic viable and sustainable adsorbent since the properties obtained is competing favorably with commercial adsorbents.