Elucidation of the Effects of Minerals on Coke Behaviour in the Blast Furnace

The objectives of this project were to determine whether minerals in coke have an effect on hot coke strength, and if so, elucidate the mechanism(s) by which this occurs under blast furnace conditions. A secondary objective was to demonstrate the use of the Environmental Scanning Electron Microscope (ESEM) for the study of coal and coke. The basic working hypothesis was that coking coals with minerals containing alkali, alkali earth and transition elements such as Na, K, Ca, Mg, Fe and Mn suffered detrimental effects during coking that resulted in cokes that were more reactive and lower in hot coke strength than equivalent coking coals which had low concentrations of these elements. This hypothesis was confirmed by coking two coals that had sufficiently similar petrographic composition and rank but different ash chemistry and mineralogy.

Attempts were made to simulate the reaction of coke with blast furnace gas at high temperatures in the ESEM. Operational problems showed that the instrument is not yet sufficiently developed to simulate the higher temperature blast furnace environments. However a number of possibly important observations were made, especially the formation of mineral phase globules of glass on the coke surfaces above 1217°C. It was also observed that reaction between coke and gas occurred mainly on the interpore wall material and not the pore linings. This implies that coke is consumed from the outside in and may explain why coke retrieved from the raceway of blast furnaces is a smaller version of the charge coke.

SIROQUANTTM XRD was used extensively on coals, their cokes and their coke CRI residues, and clearly showed the mineralogy of each was different. The difference in mineralogy between coke and coke CRI residue is especially dramatic and shows that there is a lot of reaction occurring within the mineral phases of coke pieces during this test. This implies similar vigorous reaction of the mineral phase within coke inside the blast furnace.

Calculations were performed with the thermodynamic computer package F*A*C*T in order to determine whether gases evolved by minerals could affect coke porosity. The results showed that gas evolved by minerals in coal can be of the order of several litres per gram. Perhaps surprisingly, the main contributors were the clays ? in this case kaolinite and illite. Not surprisingly, the carbonates were also prolific gas producers but in some cases the gas was given off during the semicoke phase. It is shown however that mineral-derived gas is negligible when compared to the volume of volatile matter produced during the plastic phase and is an unlikely cause of higher porosity and therefore higher reactivity. It is possible that gases given off after resolidification by calcite for example, could increase porosity due to micro fissuring.