Based on the Minimum Gibbs Free Enthalpy algorithm (FEM) and a model of reaction zones located around electrode tips in the submerged arc furnace, an analysis of the raw material chemical composition influence on the ferrosilicon smelting process was carried out. A model of the ferrosilicon process in the submerged arc furnace is a system of two closed isothermal reactors: an upper one with a lower temperature T1, and a lower one with a higher temperature T2. Between the reactors and the environment as well as between the reactors inside the system, a periodical exchange of mass occurs at the moments when the equilibrium state is reached. Based on the model, sets of calculations were performed: for a Fe-Si-O-C system and, subsequently, for a more complex Fe-Si-O-C-Al-Ca-Mg-Ti system. For both systems, the energy and mass balance values were calculated and the effects of raw material contaminants on the efficiency of the ferrosilicon melting process were determined.

No abstract

The paper presents an analysis of the possibilities of using the enthalpy of post-reaction gases from the ferrosilicon smelting process to produce electricity. Ferrosilicon smelting in the submerged arc furnace is one of the most energy-consuming electro thermal processes. Post-reaction gases, generated during the silica reduction process with carbon, contain significant amounts of energy. In the past the issue of energy recovery from the ferrosilicon process has been repeatedly taken, but for the Polish ferroalloy industry it is still valid. In order to determine the amount of energy possible for recovery calculations based on mass and energy balances has been carried out and determined the stream enthalpy of post processing gas. For preparing the balance sheets has been used operational data from the 20 MVA furnace and the overall reaction of the silica reduction process. It was assumed that the reduction process occurs at a temperature of 1650°C, and the temperature of leaving post-reaction gas, whose main ingredients are oxides CO and SiO, is 750°C.

The classical models to optimize the batch of electric arc furnace require precise and predetermined numerical values. In order to determine the scrap steel properties (eg. chemical composition), we are not able to define these values precisely and accurately. This situation is due to the heterogeneity and specificity of this material, which is related to its origin and way of obtaining. Therefore the usefulness of classical models to optimize the batch of electric arc furnace is limited. The fuzzy numbers with trapezoidal membership function were applied as a model in order to describe the technological process of steelmaking, which is neither accurate nor of ambiguous nature of the parameters. The advantage of trapezoidal numbers is a simple algorithm of arithmetic operations as well as easy and intuitive interpretation. The structure of optimization model with fuzzy parameters is similar to the classical linear optimization model of the mixtures composition. The optimization model, that is formulated on the fuzzy ranges arithmetic principles, allows to ensure that the quality parameters of the batch mix for the electric arc furnace are as precise as the ones described by electro-steel works, even though the quality parameters of scrap steel as well as some of the parameters of steel melting technological process are described ambiguously.

In order to determine reasons of unsatisfactory production output regarding one of the 12 MVA furnaces, a comparative analysis with a furnace of higher power that showed a markedly better production output was performed. For comparison of ferrosilicon furnaces with different geometrical parameters and transformer powers, the theory of physical similarity was applied. Geometrical, electrical and thermal parameters of the reaction zones are included in the comparative analysis. For furnaces with different geometrical parameters, it is important to ensure the same temperature conditions of the reaction zones. Due to diverse mechanisms of heat generation, different criteria for determination of thermal and electrical similarity for the upper and lower reaction zones were assumed contrary to other publications. The parameter c 3 (Westly) was assumed the similarity criterion for the upper furnace zones where heat is generated as a result of resistive heating while the parameter J 1 (Jaccard) was assumed the similarity criterion for the lower furnace zones where heat is generated due to arc radiation.