Title: The use of artificial neural networks to predict furfural degradation in aqueous solution by advanced oxidation processes
Authors: Sinan J. Mohammed; Yasmen A. Mustafa; Ghaidaa Majeed; Raid R. Omran
Addresses: Department of Economics of Oil and Gas, University of Imam Jaafar Al-Sadiq, Baghdad, Iraq ' Department of Economics of Oil and Gas, University of Imam Jaafar Al-Sadiq, Baghdad, Iraq ' Department of Civil Engineering, University of Technology, Baghdad, Iraq ' Al Dora Refinery, Ministry of Oil, Baghdad, Iraq
Abstract: In this study, the wastewater polluted with furfural was treated by advanced oxidation processes. Both batch and continuous systems were used. Different variables in batch experiments, Fe+2, H2O2, pH, furfural concentration and the relation with the mineralisation of furfural were examined. The results indicate that a 30 mg/L concentration of Fe+2, a 1,300 mg/L concentration of H2O2, a pH of 3, and an irradiation time of 60 min at 30°C, were required to complete the mineralisation of 300 mg/L of furfural. In the continuous system, different flow rates were used. The results show that at a furfural concentration of 300 mg/L, a flow rate of 20 mL/min, and an irradiation time of 60 min, only a 64% mineralisation of furfural is achieved. The study examined the implementation of artificial neural networks (ANNs) for the prediction of furfural degradation in aqueous solution. A correlation coefficient of 0.97-0.99 was obtained between experimental and predicted output values.
Keywords: advanced oxidation process; AOP; Fenton process; artificial neuron network; furfural; photo-Fenton.
International Journal of Environment and Waste Management, 2022 Vol.29 No.2, pp.111 - 132
Received: 15 Aug 2019
Accepted: 22 Jan 2020
Published online: 01 Mar 2022 *