Abstract: The heavy fuel oils (HFOs) and crude oils are the main oil types in the marine oil spill accidents in China. It is usually a challenge to distinguish the HFOs from crude oils due to the highly similar physicochemical characteristics. In this paper, the distributions of phenanthrene (Phe), anthracene (Ant), methyl-phenanthrene (MP) and methyl-anthracene (MA) in hundreds of HFOs and crude oils samples which were collected from all over the world were characterized. Nine new diagnostic indexes, such as Ant/(Ant+Phe) and other eight diagnostic ratios based on the MP isomers and MA, were developed for effective distinguishing HFOs from crude oils. The histogram with normal fit plots, the double ratio plots and Bayes discriminant analysis (BDA) method were employed to illustrate the effectiveness of the new diagnostic indexes. BDA model based on nine new diagnostic indexes demonstrated high precision with discriminant ratio which lay between 93.92% and 99.32%.

        Key words:　HFOs and crude oils, phenanthrenes, anthracenes, diagnostic indexs, Bayes discriminant analysis