Simulation modeling of zinc concentrates roasting in the fluidized bed furnaces

The simulation roasting process of the sulphide zinc concentrates in the fluidized bed furnaces was studied. The charge humidity, oxygen content in the blast, coefficient of charge feeding into the furnace were used as independent variables. As for dependent variable in the model, the temperature of fluidized bed was chosen. All the variables on the basis of which mathematical models were realized were virtual, and the experiment was designed. The independent variables ranging in terms of impact intensity on the fluidized bed was estimated by absolute value of the regression model coefficients. The higher the coefficient (without accounting a free model element), the greater is its affect on the dependent variable temperature. It is established that the greatest influence on the temperature of the fluidized bed (according to the model in a dimensionless scale) is exerted by the oxygen content in the blast, and further, in descending order: coefficient of concentrate loading into the furnace; the coefficient of excess blast and, the least - the moisture content of the concentrate. In this case, it should be taken into account that the data can be either dimensionless or dimensional, if this is proved by other methods, for example, by the method of expert estimation. It is illustrated that to estimate the model adequacy along with Fisher-Snedecor’s statistics it is possible to use the correlation coefficient between the experimental (virtual) and the calculation (modeling) vectors. The proposed method for the development of mathematical models of sulfide concentrate roasting has the following advantages over physical methods: substantial energy savings time, significant savings in materials for design, high quality mathematical model. It is established, that for the ranking of independent variables it is more correct to use a matrix in a dimensionless scale.