Fundamental Reasons For Different Coking Behaviour Of Coals From Different Basins - Behaviour Of Semi-inerts

Coking coal quality is based on a number of indices which characterise coke quality. While a number of standard tests exist to compare coals, some have well known flaws and are often designed to determine coal value rather than reflect real behaviour on a coke oven. In the case of coals from the Rangal Coal measures, the coking performance can exceed that expected from the standard coal qualities suggesting they are under-valued. One reason for this discrepancy is thought to lie in the variable nature of the macerals that the coals are comprised of. In particular, Rangal coals tend to have lower amounts of vitrinite but contain larger amounts inertinite that can potentially become fusible. Such thermal behaviour is complex and is poorly understood.

This project applied new techniques to characterise the fundamental thermal characteristics of a suite of coals "matched by rank" from the Rangal coal measure compared to other measures such as Moranbah, German Creek and Illawarra basins and to use this information to understand differences in the coking behaviour of the coals, particularly, the role of semi-inerts. These semi-inert maceral fractions are defined by higher reflectance placing them in the semi-fusinite range, but have the ability to become fusible (ie appear part of the continuous phase and lose definition as discrete particles). How these semi-inerts contribute to the plastic phase is not known. This investigation used the same coal suite as several previous ACARP studies that have evaluated pilot coking behaviour of Rangals and characterised fusible inertinite content using CSIRO's Coal Grain Analysis.

The majority of measurements showed a significant impact of coal rank. However, the results from this fundamental characterisation have shown that the Rangal coals in this study may be differentiated from the other coal basins by a lower overall thermoplasticity and this was expressed in a number of ways. They were characterised as being particularly low swelling, had a shorter plastic temperature region and exhibited low coke porosity. The Rangal coals began softening at higher temperatures and reached maximum swelling at lower temperatures. For this study, these two temperatures defined the plastic range and the Rangal coals had a more narrow plastic range than other coals.

The maximum swelling was highest in lower rank coals, however this trend became clearer when accounting for total fusible content, rather than vitrinite content alone. The Rangal coals fall outside this trend and accounting for the additional fusible content in the inertinite did not affect the result.

The specific heat calorimetry showed lower exothermic energy was released during the plastic phase of Rangal coals. The heat generated during plastic development is highest in the lower rank coals and (as with swelling) the Rangal coals fall below this trend.

Using the DETA technique the volatile tars could be characterised in terms of carbon and hydrogen during pyrolysis and for Rangal coals appeared to contain more molecular hydrogen. Coals of lower rank commit more carbon and hydrogen into the tar fraction. A comparison of the hydrogen to carbon ratio (Tar H/C) shows that the Rangal coals appear less aromatic and more like coals of lower rank. These indices are particularly novel and could not be obtained with other techniques.

Condensed tar analysis showed that all coals produced volatile tars with a high molecular weight distribution between 200-600amu. Lower rank coals showed an elongated distribution towards higher molecular weight which higher rank coals (and Rangal coals) did not have. The majority of tars vaporised below 400°C which is prior to the plastic region of most coals. Observable carbon peaks revolved around 200 and 300-350°C and this was considered to be based on 2 to 3 ringed systems with a distribution of different side chains (eg methyl groups, -CH3) which can affect boiling point. In this respect, the collected tars from Rangal coals showed a slightly higher amount of tar species vaporising around 350°C and this may have an impact on tar migration in a coke oven.