Application of datamining method (ID3) to data analysis for ultra deep hydrodesulfurization of straight-run light gas oil—determination of effective factor of the feed properties to reaction rate of HDS

Six kinds of straight run light gas oils were tested under various conditions in high pressure fixed bed reactor. First the formula of the corrected WAT (weight average temperature) normalized to the standard conditions was made by trial and error method, using the parameters chosen by the information of literature. Next the analysis of datamining method (ID3) and correlation coefficient were applied. This method is very unique and useful to evaluate the validity of the chosen parameters. The valid parameters of feed properties were density (SpGr), nitrogen content, and the temperature of 90% distillation (T90) under our experimental conditions.