Research on the Prediction Model of CO2 Emission Based on Different Operating Conditions for 100MW Boiler

摘要

Coal-fired power plant may integrate into the carbon trading market more actively with the lowering of carbon emission cost, understanding the situation of CO2 emissions from coal-fired power plants and rational allocation of carbon emission quotas. Based on one 100MW unit of a power plant in Guangdong, prediction models of the relationship between coal-fired power plant medium parameters, operation parameters and the total amount of CO_2 emission were established using support vector machine algorithm. They were then applied to study the amount of CO2 emission when the coal quality and air conditions are different, respectively under the 30%, 50% and 70% operation load. The results indicate good prediction capability of the established models, whose outcome bias ranges within 10%. In addition, by comparatively analyzing the optimum and worst coal quality and air condition based on the prediction results under different operating conditions, it indicates that selecting and optimizing coal quality and operation parameters are significant.