Universal simulation models based on (a) validated mass transfer relations and (b) thermodynamic modeling procedures for glass melts are developed to predict the evaporation rates of volatile species from a large range of glass melt compositions. Depending on the glass composition, temperature of the surface of the melt, local composition of the atmosphere, exposure time of a melt layer to the combustion atmosphere, and local gas velocities above the glass melt surface, the evaporation rates of volatile species can be estimated. Laboratory-scale transpiration evaporation experiments have been used to study evaporation kinetics, to derive mass transfer relations, as well as to validate the sodium evaporation modeling results for sodium-silicate melts as well as for soda-lime-silicate melts. In these investigations, the molten sodium-silicate and soda-lime-silicate melts are exposed to atmospheres of flowing gases with controlled composition and gas flow rates. In a humid atmosphere for example, sodium mainly evaporates as NaOH. From the measured NaOH evaporation rates and the mass transfer relations, the NaOH vapor pressures in equilibrium with the molten glass at prevalent temperature and furnace atmosphere composition were derived. The NaOH vapor pressures are assumed to be in equilibrium with the glass melt composition at the surface of the melt. During the evaporation test, the Na2O surface composition will change due to depletion. This leads to equilibrium vapor pressures decreasing with time, reflecting the changing chemical activity at the glass melt surface. The results of evaporation tests for sodium-disilicate and soda-lime-silicate glass melts are shown. Chemical activities derived from these measurements are compared with the results of thermodynamic modeling, using a method based on a glass melt from ideal mixtures of associate (stoichiometric) species of structural compounds with known thermodynamic properties.