Study On Optimizing Coal Blending Of The Power Plant Unit Boiler

The contradiction of imbalance between regional distribution of coal resources and location of major consumer in China and the lack of coal transportation capacity leads to the fact that the coal fired in power plants faces a variety of coal types which is different with the coal designed the indexes such as calorific values、volatiles、sulfurs、ashes、ash-fusions also have big differences in coals from different places. In fact, properties of a single coal rarely fulfills the demands of different consumers. Blending coal is the technology which is based on the coals with varies qualities.By making up the deficiencies mutually, it can satisfy the stable operation of the boiler, reduce pollutant emission,improve the efficiency of the power plants. So far, the power plants do not have the measuring method in the elemental analysis of the coal. Author makes use of artificial nerve working as a all-purpose approximation to solve this problem. In the paper, RBF Neural Network and BP Neural Network are all selected as tools for modeling. The coal’s industrial analysis and elemental analysis are respectively set as the neural network model’s input and output. The relative error of the verifying results are so small that two models can apply to engineering practice.Thermodynamic calculation for boilers plays an important role in the models of coal blending. Based on Visual C++ 6.0, the author develop a general software of the thermodynamic calculation for boilers. A 600MW supercritical once-through boiler is selected as the research object. The author sets the boiler as the thermal calculation model. The verifying results indicate application effect of this model is good. Based on the boiler thermodynamic calculation, boiler coal applicability are analyzed.The conclusion shows that different coals show different suitability to the boiler under the same power load.The paper analyzes the existing models about coal blending and raises a coal blending model optimizated by genetic algorithm which based on thermal calculation. Considering the operationing restrict of the main steam temperature, reheat steam temperature, blower, induced draft fan and the coal mill, the model optimizes the objective function for the lowest total cost of the mixed coal.This model overcomes the deficiency of the existing power coal blending mathematics model.The example of a boiler of 600 MW power coal blending calculation shows the effectiveness of the proposed method.The author also has carried on the preliminary analysis and design of power coal blending system. The advantages and disadvantages of a style mixed befor grinded between another style grinded in their own coal mill and mixed in the boiler are analyzed. At the same time,two different coal blending control systems are designed.