Coking behaviour of collotelinite and collodetrinite macerals

Authors:

Abstract
When predicting coke quality or coking behaviour, the more accurate models typically include vitrinite content and this work explores how the two most abundant vitrinite macerals behave during carbonisation. The insights gained highlight important differences in the behaviour of collotelinite and collodetrinite, paving the way for more sophisticated coke quality prediction models and enhanced blending and processing strategies. Australian bituminous coals with random vitrinite reflectance (Ro) ranging 0.81-1.18% were studied. Coal blocks were prepared with a 15 mm × 15 mm face running perpendicular to the bedding plane and polished to identify individual macerals using microscopy. These blocks were carbonised at 3°C/min to 525°C, set in resin, and repolished to examine the coking behaviour of the identified macerals “before and after.” Vitrinite concentrates (VCs) were prepared alongside “whole coal” samples to characterise viscosity using rheometry. Collotelinite underwent considerable expansion that was far greater than the expansion for collodetrinite. Extent of collotelinite expansion was greater for the lower ranked coals, characterised by large “balloons” and thin pore walls while higher ranked coals gave rise to smaller pores and thicker pore walls. Separated VCs were rich in collotelinite and generated significant fluidity, as indicated by their low minimum viscosity. For whole coals, minimum viscosities were higher due to the presence of inertinite, however, this work suggests higher viscosities are also due to an increased proportion of collodetrinite. The maceral picking process for one coal led to a situation where the VC and the whole coal had similar levels of vitrinite, however, the VC was richer in collotelinite and exhibited a significantly lower minimum viscosity. This result suggests the fluidity development and expansion behaviour for these two vitrinite macerals is different, and aligns with visual observations from the “before and after” tests. It is proposed that volatile release behaviour is different for collotelinite and collodetrinite due to differences in microstructure. More specifically, volatiles are better retained in collotelinite which contributes to its greater fluidity development and expansion behaviour. Collotelinite may be derived from single large pieces of wood that transform through a gel-like state during coalification which may enable it to have no, or very few, defects and therefore limited pathways for volatiles to leave. Instead, volatiles are trapped and contribute to both the liquid development and the expansion as bubbles nucleate in the liquid-like material and grow. There are no inclusions and defects to destabilise bubble interfaces, allowing significant expansion to occur, perhaps until the pore-wall material reaches neighbouring macerals or minerals. In the case of collodetrinite, we propose that it is comprised of individual components that are held together physically rather than chemically owing to its detrital origin. By definition, collodetrinite material is less than 10 µm in size and closely intermixed with other macerals and minerals. Accordingly, there are defects and grain boundaries between components through which volatiles may diffuse out. It is proposed that the relative ease of volatile release for different maceral types may play a crucial role in the coal’s fluidity development, and consequently its coking behaviour.