Predicting the Coking Properties of Blends

Eleven Australian coals, ranging in rank from 0.74% Rv, max to 1.68% Rv, max, that are or could be used in blends were obtained. Ten blends of these coals were prepared to the same "effective" rank and fluidity values. It was demonstrated that preparing blends to a fixed effective rank and fluidity using one pair of rank and fluidity parameters did not fix other measures of rank or fluidity. The blends were characterised by standard analytical techniques and proton magnetic resonance thermal analysis (PMRTA). These blends were coked in the BHP Research Coke Oven. Blend bulk density and moisture content were fixed, since these are known to affect coke properties. The strength, abrasion resistance, reactivity and petrography of the cokes were examined.

It was found that cokes prepared to the same effective rank and fluidity had similar overall quality though the variation in some coke properties was significant.

The clearest source of variation in coke properties was due to the presence of one low volatile coal. When this coal was present in the blend there was a marked decrease in the resistance to abrasion as measured by severe abrasion tests - surface breakage rate constant (SBRC) and the ASTM hardness and stability - although no other indices were significantly affected. It is hypothesised that the coke made from this coal tends to flake on extended abrasion, owing to its relatively large domain size, producing dusts, but this form of abrasion has no effect on strength or reactivity.

Two factors that are considered responsible for variations in coke quality were found to have no significant effect on any coke property: variations in reflectance distribution - some cokes were made from blends with reflectance gaps of greater than 0.2% - and the presence of high volatile poorly-coking coals, even at additions as great as 25% wt.

There was no clear effect of fluidity or vitrinite content on any coke property (although in this study the range of values was relatively small).

However, there was some variation in some coke properties, notably in DI150/15, that was not explained, indicating that other factors, not yet identified, can significantly affect some properties of coke.