Fundamental studies on the chemical aspects of freeze linings

Abstract

The application of freeze lining technologies has been shown to be highly effective in extending the working lives of pyrometallurgical furnaces operating under aggressive process conditions and corrosive liquid melts. Freeze linings are obtained by cooling the outer furnace walls, resulting in the formation of solid protective deposits on the inner walls of the furnace linings. Of particular interest to process operators are the freeze lining thickness and the heat loss through the furnace walls, both of which are directly related to the bath-freeze lining interface temperature. In most of the predictions for thermal steady state conditions, the assumption has been made that the temperature at the freeze lining/liquid interface is equal to the liquidus temperature of the bulk slag, and that the primary phase forms a dense sealing layer at this deposit/liquid interface. As a result of recent research on these systems, there is now extensive experimental evidence to demonstrate that the interface temperature can be below the liquidus and that the primary phase is not necessarily present at the deposit/liquid interface at thermal steady state conditions. These observations clearly indicate that other factors, in addition to thermal parameters of the system, need to be taken into account in the design of freeze lining systems. In the present study, heat and mass transfer and elementary reaction steps, and the associated thermal, physical, and chemical factors that can influence freeze lining behaviour are identified. From experiments undertaken using CaCl2-H2O solutions, accurate measurements have been made of deposit growth and microstructural changes taking place in the deposit over time were directly observed. These experiments have demonstrated the importance of nucleation rate of the primary phase and the thermal history of the freeze lining. High-temperature experiments in the ‘Cu2O’-‘Fe2O3’-MgO-SiO2 system in equilibrium with metallic copper have clearly demonstrated that at thermal steady state, bath/freeze lining interface temperatures are lower than the liquidus temperatures, and the possibility of operating processes at subliquidus bath temperatures. These experiments also indicated that an increase in bath temperatures results in an increase of the steady state interface temperature. High-temperature experiments in the Al2O3-CaO-SiO2 system were designed to study the effect of slag viscosity. Differences in the characteristics of the deposit/liquid interface were observed, however, it was unclear whether the observed differences in freeze lining behaviour were caused by the changes in viscosity, or the differences in morphology of the primary phase. The spontaneous decrepitation of freeze linings containing the dicalcium silicate phase on cooling clearly illustrated iv the need for caution when designing freeze lining systems that may form compounds undergoing significant changes in volume due to polymorphic transformations. The research supports the view that deposit/liquid interface temperatures in freeze lining systems are determined by a complex combination of heat transfer, mass transfer and chemical processes, and that thermal steady state conditions of freeze linings are best described in a more general dynamic steady state framework.