Matrix Effect In Measurements Of Carbon Content In Coal By Laser Induced Breakdown Spectroscopy And Its Correction Methods

It’s very important to develop a kind of online instrument to measure coal properties for coal-fired power plants to improve efficiency, reduce pollution, and improve safety. Laser-induced breakdown spectroscopy(LIBS) is a promising technique for online coal analysis because of its rapidity, simple sample preparation and multi-element analysis. A key issue to realize the online coal measurement by LIBS is to achieve the accurate measurement of carbon content in coal by LIBS. However, the measurement accuracy of the carbon content in coal by LIBS is relatively low due to the strong matrix effects resulted from the complex chemical composition and structure of coal.The matrix effect in the measurements of carbon content in coal is studied by spectroscopic measurements. The content variations of volatile matter and ash in coal are the main reasons for the matrix effect. The atomic carbon emission intensity is lower in the coal that has high volatile matter content due to the lower carbon atomic number densities in the plasma. As the ash content in coal increases, the plasma temperature will be increased, which will affect the atomic carbon emission intensity.The data processing method to correct the matrix effect in the measurements of carbon content in coal is developed. A spectrum standardization method is established for the data preprocessing by utilizing the molecular emissions of C2 and CN to compensate for the reduction of atomic carbon emission intensity in the coal that has high volatile matter content. In order to improve the measurement precision and accuracy, this method also utilizes the spectral information to further correct the variations of the compensated atomic carbon intensity caused by the fluctuations of the plasma parameters, including the carbon atomic number density, plasma temperature and electron number density. Then, a spectrum standardization based partial least square(PLS) calibration model is established to further improve the measurement accuracy by utilizing the spectrum standardization method to extract the main information of the carbon content in coal and utilizing the whole spectrum information to correct the residuals with PLS algorithm.The spatial confinement method is utilized to further improve the performance of the spectrum standardization based PLS model. Firstly, the spatial and temporal distributions of spectral emissions of plasmas generated without and with cylindrical cavities are studied by spatially and temporally resolved spectroscopic measurements. From the measurements, it is proved that the mechanism of the spatial confinement is the interaction of the reflected shock wave and the vapor plume in the plasma. The measurements also show that when the cavity diameter is suitable and the cavity height is close to the height of the vapor plume, the spatial confinement effect of the cylindrical cavity is optimal. Then, the cylindrical cavity of moderate size is used in the measurements of carbon content in coal. It is shown that the measurement accuracy and precision of the carbon content in coal by LIBS can be further improved by the combination of the cavity confinement method with the spectrum standardization based PLS model.