提出了一种衰减全反射红外光谱法快速测定糠醛抽出油烃族组成含量的方法, 设计了一套重油烃族组成快速分析系统。 实验收集了62个减压蜡油经糠醛精制后的抽出油样品, 采用NB/SH/T0509-2010方法测定其饱和烃、 芳烃含量, 由差减法求得胶质和沥青质总含量。 采用偏最小二乘法(PLS)分别建立了糠醛抽出油饱和烃、 芳烃和胶质+沥青质含量的定量校正模型, 其SEC分别为1.43%, 0.91%和1.61%, SEP分别为1.56%,1.24%和1.81%, 均小于NB/SH/T0509-2010方法规定的再现性误差, 模型预测精密度小于标准方法的重复性误差要求。 与现行标准方法相比, 该方法速度快, 操作简单, 可显著提高糠醛抽出油烃族组成分析效率。 该分析系统也可用于其他重油分析, 具有很好的推广应用前景。

A set of rapid analysis system for hydrocarbon composition of heavy oils was designed using attenuated total reflection FTIR spectrometer and chemometrics to determine the hydrocarbon composition of furfural extract oils. Sixty two extract oil samples were collected and their saturates and aromatics content data were determined according to the standard NB/SH/T0509-2010, then the total contents of resins plus asphaltenes were calculated by the subtraction method in the percentage of weight. Based on the partial least squares(PLS), calibration models for saturates, aromatics, and resin+asphaltene contents were established using attenuated total reflection FTIR spectroscopy, with their SEC, 1.43%, 0.91% and 1.61%, SEP, 1.56%, 1.24% and 1.81%, respectively, meeting the accuracy and repeatability required for the standard. Compared to the present standard method, the efficiency of hydrocarbon composition analysis for furfural extract oils is significantly improved by the new method which is rapid and simple. The system could also be used for other heavy oil analysis, with excellent extension and application foreground.