Performance evaluation of cottonseed oil methyl esters produced using CaO and prediction with an artificial neural network

Abstract

ABSTRACTThis study deals with the analysis of performance evaluation of refined cottonseed biodiesel produced with calcium oxide and using an artificial neural network (ANN) technique to predict the performance. A two-cylinder, four-stroke diesel engine fuelled with the standard diesel, biodiesel and their blends, operated at different engine speeds, was used. The experimental results revealed that blends of refined cottonseed oil methyl ester with diesel fuel gave better engine performance and improved emissions. Comparing the results with conventional diesel fuel, B20 gave similar brake power, brake specific fuel consumption and brake thermal efficiency, and lower carbon (II) oxide and hydrocarbon, with the noticeable presence of nitrogen oxide (NOx) emission. The ANN model predicted the engine performance with correlation coefficients (R) of 0.97433, 0.99142 and 0.97889 for the engine brake power, brake specific fuel consumption and brake thermal efficiency, respectively. The mean square error between the desired outputs as measured and simulated by the model was 0.0001. Therefore the biodiesel produced from refined cottonseed oil performed better when blended with a small quantity of petrol diesel, and ANN proved to be a desirable prediction method in the evaluation of engine parameters.