Tracking the Carbonisation Performance of Vitrinite Macerals

Models predicting coke quality or coking behaviour typically include vitrinite content.  Vitrinite is a broad group comprising several macerals however, and by examining the coking behaviour of key vitrinite macerals, this project highlights important differences in the behaviour of different vitrinite macerals during carbonisation.  An improved understanding of this behaviour paves the way for more sophisticated coke quality prediction models and enhanced blending and processing strategies.

This project examined coals from the Moranbah, Rangal and Illawarra coal measures, with random vitrinite reflectances (Ro) ranging from 0.81-1.18%.  For higher rank coals, vitrinite is separated into three maceral sub-groups viz. telovitrinite, detrovitrinite and gelovitrinite.  These maceral sub-groups are further divided into different macerals.  Telovitrinite consists of telinite and collotelinite (equivalent to telocollinite).  Detrovitrinite consists of collodetrinite (equivalent to desmocollinite) and vitrodetrinite.  For the coals in this study, collotelinite and collodetrinite are the dominant macerals within the vitrinite group, and as such, were the focus of this project.

The experimental approach involved the preparation and microscopic examination of coal blocks both before and after carbonisation.  Vitrinite concentrates were prepared alongside “whole coal” samples to characterise the viscosity (using rheometry) and molecular structure (using ¹³C NMR).

The vitrinite concentrates were rich in collotelinite and these concentrates generated significant fluidity, characterised by a lower minimum viscosity than the whole coals.  While the presence of inertinite may be the cause of the higher minimum viscosity for the whole coals, it is also thought that the presence of collodetrinite may be a contributing factor.  The lower rank Rangal coals did not generate significant fluidity and this is thought to be due to collodetrinite.

These “before and after carbonisation” tests showed that the collotelinite bands in all coals undergo much greater expansion when carbonised than the collodetrinite bands.  Expansion was greater for the lower rank coals (random vitrinite reflectance <1.0%) than the higher rank coals.

For the lower rank coals, the collotelinite expansion observed for the Rangal and Illawarra samples was characterised by large “balloons” and thin pore walls, in contrast to the Moranbah sample for which the pores were smaller and the pore walls were thicker.  The collodetrinite expansion in the Moranbah coal was much greater than in the Illawarra and Rangal coals, possibly due to the influence of significant liptinite associated with the collodetrinite.

For the higher rank coals (random vitrinite reflectance >1.0%), the viscosity of the collotelinite was higher and the expansion was less which gave rise to smaller pores and thicker pore walls than for the lower rank coals.  As found for the lower rank Rangal coal, the introduction of inertinite and collodetrinite to the higher rank coals increased the viscosity.

For one of the Moranbah coals, while the vitrinite concentrate and the whole coal had similar levels of vitrinite, the vitrinite concentrate was richer in collotelinite, and had a significantly lower minimum viscosity than the whole coal.  This suggests that the fluidity development and expansion behaviour for these two vitrinite maceral types is different, as seen in both the “before and after” tests and the viscosity measurements.

It is proposed that differences in behaviour of collotelinite and collodetrinite are due to differences in the volatile release behaviour, on account of differences in the microstructure of these two macerals.

This notion of component size influencing coking behaviour mirrors existing knowledge on the effect of coal particle size on coking behaviour, whereby smaller particles undergo less expansion.  It follows that including vitrinite maceral type and possibly also maceral grain size could enable enhanced coke quality prediction models and improved strategies for coal blend designs.  Greater knowledge on the volatile release behaviour of different macerals and the effect that inclusions have on expansion could also assist with determining why particular coals have either enhanced coking behaviour or poor coking behaviour.