The Local Mode Prediction Model Of Near Infrared And Its Application In Oil And Gas

Near infrared spectrum contains a large number of characteristic information in thehydrogen-containing groups of the organic molecules. The oil and gas molecules are mainlyconstituted by C-H group. Therefore, the near-infrared spectrum can be used for efficientqualitative and quantitative analysis of oil and gas’s composition and properties.First, the influence on the infrared absorption intensity caused by the temperature andpressure changes is studied. The quantitative analysis of oil and gas is based on the Lambert-Beer Law. However, the Lambert-Beer Law is only applicable in ideal conditions, and it willbe affected by many factors in practical applications. Two modification methods reducing theinfluence of the temperature and pressure on C-H group’s infrared absorption are given: datafitting and theoretical calculations. Data fitting, based on the method of least squares withexperimental datas, is to gain the relationship among the temperature, pressure and infraredabsorption coefficient at different peaks of C-H group. In theoretical calculation method, theinfrared absorption coefficient is regarded as the product of C-H group’s the absorption linesintensity and linear function. The Voigt function, which is chosen as the linear function, hasno accurate values and is approximately calculated with the method of the polynomialexpansion summation in this paper. Secondly, the problems of the local mode predictionmodel of C-H group’s infrared absorption spectrum and its application in the oil and gasanalysis are studied. The predictive local mode models of methyl group, methylene group andmethenyl group have been built and the model parameters have been calculated. The vibrationfrequencies of C1~C4alkanes and alkenes have been predicted with the model and makecomprehensive evaluation to the model with the results in detail. The models of methyl group and methylene group in the alkenes have been modified and gain better predictive effect.Finally, a soft measurement scheme is given for the on-line analysisi of oil and gas.