The self-heating of coal has been and remains a significant issue for the coal industry. It has caused closure of mines, loss of life and equipment and every year costs the industry millions of dollars. Much work has been performed in an attempt to investigate the phenomenon of coal self-heating and to detect it. The earliest methods of detection relied on physical indicators such as heat, haze, sweating or smell. In fact, heatings were often first detected by smell. The major compounds that contribute to the odour of self-heating coal have not previously been definitively identified.
Gas indicators used for detecting coal self-heating were developed during the twentieth century. These include CO make, Graham’s Ratio and the CO/CO2 ratio amongst others. However, most of the practical tests used to develop these gas indicators are based on small-scale laboratory tests. These tests use a high airflow to coal mass ratio and are typically conducted on finely crushed coal. The current understanding of gas indicators in the mining industry is based on these small-scale tests. These existing tests do not allow for the complex reactions that take place in a coal pile. For example, they do not consider the effect of moisture on the gas evolution patterns as a hot spot develops and migrates.
At The University of Queensland a 2-metre column apparatus was developed to study the gas evolution pattern associated with bulk coal self-heating. This column holds approximately 50 kg of coal and is an intermediate step between small-scale and large-scale tests. Large-scale tests use tonnes of coal, and have trouble associating the gas evolution from the developing hot spot to its originating location, due to the size of the rig. The hot spot location effect can be investigated in the 2-metre column since the column is only 0.2 m diameter and gas sampling ports are located along its length.
This thesis examines the correlation between gas evolution, coal temperature and the location of the gas source with respect to hot spot location. Two different coals were tested in the 2-metre column and in a small-scale test. Small-scale test gas results for these coals were plotted against temperature. Equations for the curve of best fit for this data were determined and used to develop a model which converted the temperature profile of the column into a single temperature that represents the effective reactivity of the coal averaged over the whole column (effective temperature). The model was also used to predict the exhaust gas carbon monoxide and oxygen concentration based on the column temperature profile and a good correlation was obtained. The model was unable to account for the effect of secondary seam gas desorption, which may affect the rate of oxidation. This showed that the small-scale tests are indeed valid for modelling some aspects of larger scale self-heating events, however, the complexity of the effect of seam gas content on the rate of coal oxidation must be accounted for. Also, no evidence of secondary reactions downstream of the hot spot was observed in the 2-metre column. It was also found that the 2-metre column tests were able to provide a better estimation of what gas indicator values to use as part of Trigger Action Response Plans (TARPs) than the small-scale tests, and initial TARP values have been suggested for the MAU4 and SPCK5 coals.
The generation of hydrogen by an acid metal reaction was also documented, where the acid is simply the mildly acidic moisture liberated from the coal. It was found that significant amounts of hydrogen can be generated at low temperature by this mechanism. Historically, high levels of hydrogen have been thought to indicate an advanced self-heating event. However the bulk 2-metre column testing shows that in a real mining situation high levels of hydrogen can in fact be generated under otherwise benign conditions. Importantly, this explains spurious real world observations in coal mines where high levels of hydrogen have been encountered whilst other gas indicators have not shown signs of an advanced heating.
Anecdotal evidence has suggested that the key odour of coal self-heating can be described as a musty, oily, benzene-like (sweet) smell. In general, the odour has been attributed to the presence of ketones and aldehydes. Benzene and acetone were not detected at levels significantly above their odour thresholds in the 2-metre column tests. However, acetaldehyde was identified as the dominant constituent of the odour. Other compounds which significantly contribute to the odour are acrolein, isovaleraldehyde and propionaldehyde. As the hot spot temperature increases, so do the concentrations of the significant odour compounds. The odour of these compounds is dependent on their concentrations, therefore as the concentration increases, the apparent odour of the coal self-heating event changes and increases in strength.