Distributed Compositional And Temperature Nature Of Melts In Submerged And Open Arc Furnaces In High Carbon Ferrochrome Production

The development of feedforward semi-empirical control models for furnaces and the evaluation of mass balances are based on plant data. The models can only be as good as the data and it is therefore required to know the inherent uncertainty associated with the plant data. This uncertainty stems mainly from the fact that the slag, alloy or dust, have a distributed nature that depends on the degree of mixedness. It is commonly accepted that melts from open bath furnaces are well-mixed, and submerged arc furnaces less so. The impact of this assumption is significant, as a reasonably homogeneous melt could be represented by one sample only. As so much of metallurgical control depends on the assay information of the alloy and slag melts from furnaces, it was decided to characterise the possible variation in melt chemistry and temperature. Sampling campaigns and continuous temperature measurement of the melts were performed for different alloy and slag taps at a major ferrochrome producer. Both a submerged arc furnace and open arc furnace are evaluated in this way, by way of comparison. The slags of both open arc and submerged arc furnaces were found to be of similar degrees of homogeneity, with much spatial variance in the reducible iron and chromium oxides. However, it was found that the alloy melts deriving from open arc furnaces were significantly less homogeneous than the alloy derived from submerged arc furnace alloy production. The compositional variance could be related to the furnace operating conditions and it affects both metallurgical accounting and process control. In all cases it was found that the variance in melt solute components are larger than can be ascribed to analysis instrument error. It will be shown how this predetermined variance in melt composition could be included in process modelling and control.