DEVELOPMENT OF A MULTI-FEEDSTOCK BIODIESEL PLANT AND OPTIMIZATION OF ITS OUPUT

Abstract

The quest for increased energy security and the need to develop renewable alternatives to fossil fuel has brought about the development of renewable sources of energy. In this study, a biodiesel processing plant was designed and fabricated for producing biodiesel from used and unused oils. The developed plant consists of nine sections namely; oil storage, methanol storage, water tank, mixing tank, reactor, wash tank, separating section, glycerol storage and biodiesel storage. The plant majorly consists of three 3 kW variable speed sparkless electric motor, three 2.5 kW pumps, 220 V AC alarm, four flow switches, 220 V timer relay, 60 A contactor, 3 thermostatically controlled 300 kW electric heater, sawdust insulation, 12 ball valves, 3 thermostats, a separating funnel, shaft and a stirrer of diameter 25 mm with seven blades that rotate in the reactor. The stirrer is driven by the 3 kW electric motor. The plant was designed to have an output capacity of 200 litres per day. A stainless steel of 2.5 mm thickness was used in the fabrication of the tanks for each of the nine sections of the plant because of its excellent combination of high strength, formability and corrosion resistance. Performance evaluation of the plant was conducted using used frying oil and palm olein oil. The biodiesels produced were characterized and its properties compared in respect of the American Society for Testing and Materials (ASTM D6751) and EN 14214 limits for biodiesel. Also, emissions of some pollutants from the diesel engine were measured at 2000 rpm using the AUTOplus Gas Analyzer for diesel and blends of biodiesel while the Fourier Transform Infrared (FTIR) Spectroscopy was also used to analyse the functional groups of the biodiesel produced. The effects of five process parameters namely; reaction time, temperature, stir speed, catalyst concentration and methanol-oil ratio on the process plant yield were investigated. Optimization of the five process parameters was carried out using the central composite design model and response surface methodology. Taking the biodiesel yield as the response of the designed experiment, the data obtained were statistically analysed to obtain a suitable model for optimization of biodiesel yield as a function of the five process parameters. The model was validated through a comparison of its output with the ones obtained from the actual experiment. It was revealed that the developed model could adequately predict the biodiesel yield within the limits of the process variables investigated at 95% confidence level