Combustion Optimization Of Large Coal-fired Boilers

Nowadays, along with rapid growth of energy consumption in our country, NOx emission is greatly increased. During the Eleventh Five-Year Plan, regional acid rain deterioration is becoming worse due to fast increasing of NOx emission, which partially offsets our great cost aiming at SO2emission reduction. Therefore, NOx is ranked as control subject of atmospheric impurity during the Twelfth Five-Year Plan. Commonly speaking, NOx mainly comes from power station and motor vehicle, of which the former one emits great volume and relatively concentrated, NOx becomes the first target for state control. So development of economy applicable and low NOx pulverized coal combustion technology is the key factor to continue to keep advantage of coal-fired boiler.Firstly this thesis analyses formation mechanism of NOx; relationship between NOx emission concentration and burning coal variety, boiler load, boiler operating parameter; of which, coal variety is the biggest influence to NOx emission concentration, and in the boiler operating parameter, oxygen content inside boiler is the most obvious influence to NOx emission concentration. Boiler operating condition and parameter can also influence NOx emission concentration, and timely adjustment of boiler burning condition and operating parameter is able to lower NOx emission concentration by10%～20%; this thesis focuses on main control methods of NOx from the aspect of flue gas treatment and low NOx combustion technology.Coal variety and operating parameter greatly influence boiler NOx emission and carbon content in fly ash, but it’s hard to show their relationship by mathematical expression. This thesis, using the600MW boiler of a power plant in Shanxi as research object, focuses on developing a prediction model of NOx emission volume and carbon content of fly ash by quoting BP-adaboost. This model, employing boiler operating parameter that influencing target value as input variable, NOx emission volume and carbon content of fly ash as output variable, uses data downloaded by boiler DCS for network training, and tests predictive ability by test sample. Model developed in this thesis can be used for actual application if predictive error of NOx and carbon content in fly ash was below5%. Then NOx emission volume can be gotten from this model, and further boiler efficiency can be calculated.Objective function of combustion optimization is established with comprehensive consideration of NOx emission and boiler efficiency. It combines prediction model of NOx emission and carbon content in fly ash with genetic algorithm to search optimum proposal under various optimization objectives. Optimization objective is:separately optimize NOx emission, then optimize boiler efficiency in the case of NOx emission at650mg/m3, and to integrated optimize boiler efficiency and NOx emission. When using genetic algorithm to optimize boiler efficiency and NOx emission, NOx can be reduced by24.8%, that’s from678.12mg/m3to509.93mg/m3, and boiler efficiency is reduced by0.22%from92.43%to92.22%.The last, fundamental principles of PSO algorithm is also presented in detail in this thesis. PSO algorithm similar with genetic algorithm, is an optimization tool based on iteration. PSO is easier than genetic algorithm in operation, such as no cross, variation operation and less parameter. What’s more, PSO algorithm is more flexible in terminate iteration than that of genetic algorithm. Therefore, this thesis, by combing PSO algorithm with NOx emission prediction model, adjusts boiler operating parameter to separately optimize NOx, and integrated optimize NOx and boiler efficiency to gain optimum results. Compared with original working condition, NOx emission can be reduced by28.3%, meanwhile boiler efficiency is reduced by0.05%under integrated optimization. From the viewpoint of optimization result, PSO algorithm is superior to genetic algorithm in less parameter, short computing time, and is applicable to online optimization.