Improved prediction of coal ash slag viscosity by thermodynamic modeling of liquid-phase composition

The calculation of the viscosity of a coal ash slag as a function of temperature is of interest both to obviate the difficult experimental measurement of viscosity and for mathematical modeling of ash behavior. Traditionally, the composition of ash prepared in the laboratory has been used as the data for calculation of the viscosity. This approach has inherent inaccuracies. The viscosity of the liquid will be dependent on its composition. For situations involving partial melting of the ash, as in viscous flow sintering, the composition of the melt phase will be that of the low-melting-point constituents and not that of the ash as a whole. For slags well above the original ashing temperature, high-temperature physical or chemical processes could also make the composition of the liquid different from the original ash. In the present work, we demonstrate that prediction of the liquid-phase composition at the temperature of interest, and subsequently using that predicted composition for calculation of viscosity, substantially improves the accuracy of prediction in the Newtonian flow range. Classification criteria based on ash composition, base-to-acid ratio, silica ratio, and lignite factor are used to select the most useful viscosity calculation model for a particular ash. For a set of 33 coal ashes, prepared from coals ranging from brown coal to bituminous in rank, the average deviation of predicted viscosities from experimental measurements were within 30% (a criterion considered to represent acceptable prediction) for 29 of the 33 cases.