Research On Key Technologies Of Online Monitoring Of SO₂ And NO_x In Ultra-low Emission Flue Gas

Air pollution is one of the most serious environmental pollution and threatens for human survival.The flue gas emitted by coal-fired power plants and cement plants is one of the main sources of air pollution.As our country pays more attention to air pollution,it has implemented ultra-low emission standards for stationary sources.This standard requires that the SO₂、NO and NO₂ in the flue gas emitted by coal-fired boilers is below 100 mg/m³.For gas-fueled boilers,SO₂ is below 35 mg/m³,NO and NO₂ are below 50 mg/m³.The implementation of ultra-low emission standards brings the following challenges to the online pollutant gas analyzer:（1）Extraction of ultra-low concentration gas Differential absorbance;（2）Separation of multi-component Differential absorbance;（3）Ultra-low concentration retrieval algorithm.（4）The influence of temperature and humidity on the measurement results.For a long time,our country’s pollutant gas online analyzer for ultra-low concentration has mainly relied on imports.To break this dilemma,with the support of the national key research and development program of China.We have carried out the research of ultra-low concentration SO₂ and NO_x（NO and NO₂）multi-component gas accurate monitoring and manual sampling equipment.The goal of the research is to develop an ultra-low concentration SO₂ and NO_x online analyzer.Based on the problems above,the main contents of the research are as follows:（1）For the ultra-low concentration of SO₂ and NO_x,the absorption band selection was optimized,and the absorption band of SO₂ and NO is selected to be 202～226nm,and NO₂ absorption band was selected to be 335～365nm.The shortcomings of using polynomial fitting and Savitzky-Golay filter to separate slow-change（Rayleigh scattering,Mie scattering）in the absorption spectrum was analyzed,and decided to use multi-scale wavelet transform to extract the fuzzy components in the absorption spectrum,reconstructed the fuzzy components to obtain the slow-change.By this way,the over fitting of polynomial fitting and Savitzky-Golay filter was overcame.（2）A method to separate the Differential absorbance of NO and SO₂ from mixed gas was proposed.The Differential absorbance of SO₂,NO,and NO₂ in 202～226nm was analyzed in the frequency domain.It was found that the frequency of Differential absorbance for NO is a complex and the bandwidth covers the frequency bands of SO₂and NO₂.It is difficult to decompose the Differential absorbance of NO from the mixed gas by the use of frequency domain filter.Therefore,an improved algorithm of adaptive interference cancellation filtering technology was used to solve this problem.The experimental results show that the adaptive interference cancellation filter can accurately separate the Differential absorbance of NO from the mixed gas.The selection of absorption band for SO₂ in the 202～226nm was optimized and at last206～212nm was selected.Because the Differential absorbance of NO in this band is almost a horizontal line,it can be considered that there is no absorption.This band is the most sensitive band of SO₂ in the ultraviolet band,which is suitable for monitoring ultra-low concentration SO₂.Differential absorbance of SO₂,NO and NO₂ at 206～216nm in the frequency domain was analyzed,it was found that their frequency components have almost no overlap.Therefore,frequency filtering technology is used to extract the Differential absorbance of SO₂ in the mixed gas.Measurement results indicate that this method can accurately decompose the Differential absorbance of SO₂ in the mixed gas.（3）Large measurement error is the shortcoming of the traditional least squares method for detecting ultra-low concentration gas,in this work,an improved concentration inversion algorithm-Differential absorbance line-by-line integration method was proposed and the principle of was introduced.By measuring the Differential absorbance integral value of a series of different concentrations,training the mathematical model of the concentration and the integral value.Cross-validating was implemented to the model to ensure that the concentration retrieval model has strong generalization ability,under-fitting or over-fitting was avoided.Finally,SO₂,NO and NO₂ multi-component gas analyzers was developed,nonlinear error of SO₂,NO and NO₂ were±0.4%FS,±1.06%FS,and±0.7%FS,respectively.The detection limits are0.02 ppm,0.063 ppm,0.11 ppm,24 hours.The zero drift were±0.7%FS,±0.6%FS,and±0.5%FS.Instrument performance indicators meet the requirements of ultra-low emission standards for monitoring instruments.（4）The influence of the temperature drift of the spectrometer and the gas temperature and humidity on the measurement results were analyzed,a temperature and humidity compensation device was designed.A temperature control system was designed for spectrometer,the effect of temperature drift can be ignored.The temperature and humidity compensation model was established.The measured data show that the temperature and humidity compensation method is effective.The SO₂ and NO_x analyzer developed by Chongqing University has been tested in the thermal power plant of Dalian Development Zone for 15 days.The results show that the analyzer can effectively measure SO₂ and NO_x in ultra-low emission flue gas.