Accurate prediction of the viscosity behavior of crude oil emulsion is essential for the design, selection and routine operation in many petroleum applications. The aim of this work was conducted to predict viscosity using feed forward artificial neural network FFANN. The factors are studied include the effect of mixing time, mixing speed, emulsifying temperature and shear rate. Experimental laboratory data are used to develop the viscosity correlations. The results shown that the predicted model has a good compatibility with experiments obtained in R= 0.99992 and best validation performance 143.3434 and high correlation coefficient RC= 0.98. The results also show that the achieved model is better than the conventional ANN in the prediction of viscosity with improving the overall percentage of 39.
