700MW Thermal Power Unit Blending-coal Of Separate Pulverizers And Optimization Expert Supporting System Development

In recent years, domestic thermal power generating units begin to use blending coaldue to many reasons such as the contradiction between supply and demand of coal. Butirrational blending and operation leads to safety and economy problem of unit, may alsocauses the increasing of emission concentration of air pollutants such as SO2, nitric oxideand so on. This paper intends to solve above problems and ensure the safety, economy andenvironment of blending coal.Physical programming is widely used for design optimization. It can reflect designer’spreference. This approach is used on coal blending problem and preference functions arepresented. A new optimization model for coal blending is established, whose goal is tosecure safety, economy and environment of coal combustion in boiler. A satisfied solutionis resolved while knowing the coals and operating circumstance. An online evaluationsystem is proposed to evaluate the effect of coal blending. It relies on power plant real-timedata and can provide safety, economy and environment evaluation automatically. Theevaluation result can reflect the different effects of coal combustion between the same kindof boiler. An example of evaluation on a700MW unit is presented.Blending coals are placed on different mills and burners, which is widely used ondomestic large thermal power generating units. In this way, a blending scheme can becarried out when presnting operators which coal is loaded into which bunker. Differentcoals have different quality, so they have different ignition, burning and burnoutcharacteristics. Thus different coal-bunker schemes have different results. A case-basedreasoning (CBR) way is used to guide the coal-bunker schemes and relative operatingmethods. It’s a self-learning system. A case library is established by collecting SIS data andself-evaluation, and better mode can be learned. The case retrieval finds the similar casesclose to current case, provides a learning mode of history case. The case library is built on a 700MW unit and CBR is used to provide learning mode of coal-bunker schemes andrelative operating methods.On the basis, Blending-coal of Separate Pulverizers and Optimization ExpertSupporting System is developed. The system integrates coal blending model, evaluationmodel and CBR optimization model. It’s used to guide the work of coal blending in powerplant, which is intended to realize the safety, economy and environment of coal combustionin boiler.