Prediction of the size distribution of tarballs using population balance along with image analysis technique

Intensive petroleum activities in the Caspian Sea have caused oil pollution in this region. Trapped tarballs on the South coast of the Caspian Sea are one of the evidence of the oil pollution. In this study, Poloxamine copolymeric surfactant is used to agglomerate tarballs in a system containing n-hexane as a solvent, methanol as anti-solvent and the Caspian Sea tarballs. Image analysis technique is used to find the size distribution and evolution of tarball particles. The average number of tarball aggregate size increases with time and reaches a maximum value, and then declines so that it eventually reaches to a steady state and almost constant size. The progress of tarball aggregate size distribution is studied and the steady-state size distribution of the tarballs is predicted by two algorithms called back propagation artificial neural network (ANN-BP) and the Imperialist Competitive Algorithm (ICA). The purpose of this study is to predict the cumulative size distribution of tarballs in the presence of Poloxamine copolymer. The error analysis shows good agreement between the predicted values for the steady-state aggregate size distribution by the neural network combined with ICA (ANN-ICA) and real data.

کلیدواژه ها:

tarball size distribution ، tarball aggregation ، artificial neural network ، Imperialist Competitive Algorithm ، image analysis technique