Application of theoretical and physical models to smelting and refining processes

The mathematical modeling of the kinetics of metal/slag/gas reactions which occur in pyrometallurgical processes is very important in providing an insight into the coupled reactions that take place between the large number of components present among the reaction phases. A generic computer code was developed (in Pascal, called SMAK), based on the coupled reaction model proposed by Robertson et al. A new approach to kinetic modeling using free energy minimization to calculate the equilibrium condition at each interface, is also described. Unit operations, representing the different phases and interfaces in this approach, were connected together using the METSIM software. The kinetic model was applied to several metallurgical processes such as counter-current refining of low-carbon ferromanganese, zinc slag fuming (SMAK) and simultaneous de-S/de-P of pig iron (METSIM). Physical modeling of a counter-current reaction launder (CCRL) for metals refining was performed to study the fluid flow and dispersion characteristics of the CCRL. This work was conducted in a 200 cm long (L) by 20 cm wide (W) channel, using tetrachloroethylene to model \u27metal\u27 and water to model to \u27slag\u27. A thermal tracer technique was used to measure the eddy thermal diffusivity (αe) and interphase heat transfer coefficient (hov). The heat and mass transfer analogy was then applied to determine the extent of longitudinal mixing (De/uL) and to estimate the rate of interphase mass transport (kov). Nitrogen was bottom injected at a rate per unit area of 12.5 cm/min, through up to 14 central bubblers placed along the length of the CCRL. The results indicate that even under conditions of significant bottom gas injection, it should be possible to maintain acceptably low longitudinal mixing (De \u3c 20 cm2/s and De/uL \u3c 0.1) in the CCRL. Interphase mass-transfer coefficients of the order of 0.02 cm/s and higher, are predicted from the measured interphase heat transfer coefficients using the mass and heat transfer analogy. Adequate performance is likely to be possible for an industrial CCRL, based on extrapolation of these results to full-scale conditions --Abstract, page v