Research Of Combustion Diagnostics And Testing In Boiler Based On Radiation Spectroscopy

In China, the thermal power generating unites which are mainly based on coal combustion occupy the dominant position in power industry. Accurate and effective monitoring of flame in modern coal-fired furnace is of great significance to the stable, effective and security operation of boiler. This research is based on the flame of fuel obtained from a self-made flame detection system which has a core of fiber optic spectrometer. In this paper, a method for the advancement of spectral thermometry for flames and a novel spectroscopic method for measurement of emissivity which is compact and suitable for on-line measurement are proposed. Finally, by combining the temperature, the emissivity of flame and the flame flicker frequency, a combustion criterion of power plant boiler is proposed.The main contents of this paper and innovations are as follows:(1) The temperature of flame is a critical parameter for the combustion monitoring in boilers and the measurement methods based on CCD fiber optic spectrometer have become one of the research hotspots in the field of flame temperature measurement for their unique advantages. For the response characteristics of CCD itself, spectrometer needs to be standardized by blackbody furnace. However, if standardize temperature has too large difference from measured temperature, it will generate large errors. Moreover, when the response strength that measured by optic spectrometer has too large difference from the response strength of blackbody furnace, it will also generate large errors. In the present paper, methods have been put forward to improve the accuracy. Experiments had proved that the accuracy has been improvement a lot when use the methods. Thus the measure rang of spectral radiation thermometry, which bases on fiber optic spectrometer, has been improved a lot.(2) Methods used before to measure the emissivity of non-gray flames on-line, for the imperfect theories, have some defects. In this paper, direct to the flames of different fuels, a novel spectroscopic method based on the Planck’s Law for measurement of emissivity is introduced. The results are compared and validated with the emissivity models obtained in other research about the flame emissivity. Besides, Based on the measured emissivity, the other spectroscopic method for composition detection of blended fuels is also proposed.(3) The flame flicker frequency, as a reference of the stability of flame, is a very important criterion for flame detection and combustion diagnostics. However, because the flames signal in the industrial furnace has a wide spectrum. The signal includes noise in the whole range of spectrum, and the noise occupies the dominant status in high frequency range. So to extract the useful information in the signal effectively, the signal must be noise-reduced. In this paper, spectral signals acquired in the power plant boiler are noise-reduced with the method of Wavelet Threshold De-noising. After de-noising, signals are fast Fourier transformed to get the flame flicker frequency.(4) After the research of spectral flame temperature, emissivity and flicker frequency measurement method, in this paper, a novel combustion diagnostics program for power plant boilers is proposed based on the three parameters mentioned above. And the whole process of flame combustion is determined continuously. This is also the ultimate aim of this research topic: To establish a flame detection and combustion diagnostics system for the modern coal-fired boilers in large-scale thermal power stations and provide an accurate and effective basis and indemnification for the stable, effective and security operation of boiler and for the parameters adjustment when the loads are changed.