Study On The Integrated Optimization Of Coal Blending And Boiler Operation Of Power Plants

Coal blending is an important technique to produce coals of different properties. And the operation of boilers often have to be adjusted to adapt to different coal properties. In order to promote the optimal allocation and rational utilization of coal resourses, the integrated optimization of coal blending and boiler operation is proposed in this paper.The combustion characterists of coal were correlated to coal quality using two nonlinear models which were built by support vector regression machine. Combusiton kinetic parameters were predicted from coal quality. The predicted results of activation energy and frequency factor showed the mean squared errors of2.571and0.565, respectively. Thermogravimetric curves defined by the predicted kinetic parameters charaterized the whole combustion process successfully. In addition, the nonlinear model of the relationship between ash fusibility and coal quality was studied using support vector machine. Due to the additivity of coal quality parameters, the prediction models of combustion characteristics and ash fusibility could be applied for not only single coal, but blended coal.The effects of coal fineness and surface properties on combustion characteristics were studied by non-thermalgravimetric experiments and nitrogen adsorption method. The comprehensive combustion performance improved with coal fineness decreasing, or with the average pore diameter increasing. The changes of the combustion reaction activity with the specific surface area showed a decrease first and then increased.Based on the real working conditions of a330MW unit, the integrated optimization of coal blending and boiler operation was carried out using differential evolution algorithm. The boiler efficiency was increased and the emission concentration of NOX was controlled below650mg/m3.