| dc.description.abstract | Five coal samples from the Witbank, Free State and Limpopo provinces in South Africa were studied to determine and understand the influence of minerals and other coal properties on the moisture adsorption and desorption behaviour. All the experiments were conducted in a climate chamber at isothermal conditions. The climate chamber controlled the relative humidity and temperature to which the coal particles were exposed during each experiment. The climate chamber was also equipped with a mass balance to record the increase (adsorption) and decrease (desorption) in mass, where a constant mass reading denoted equilibrium conditions. The coal samples were characterised in terms of proximate analysis, ultimate analysis, petrographic analysis, CO2 and N2 BET sorption analysis. The mineral characterisation of each coal was performed with XRF and QEMSCAN analysis, where the QEMSCAN analysis allowed for the quantitative evaluation of the minerals present in each of them. A constant particle size of +1mm –2mm was used to evaluate the adsorption/desorption characteristics for this investigation. The characterisation results indicated higher moisture– and oxygen contents for the lower ranked bituminous coal samples compared to the higher ranked bituminous coal sample. Adsorption results also indicated that the lower ranked coals samples adsorbed the most moisture whereas the higher ranked coal sample adsorbed the least moisture. The oxygen content is an indication of the oxygen containing functional groups present on the coal surface which facilitates moisture adsorption. It was therefore expected that the lower ranked coals would absorb more moisture than the higher ranked ones. QEMSCAN analysis revealed that the predominant mineral present in all the coals samples were the clay mineral kaolinite followed by quartz. The influence of kaolinite on the adsorption properties was investigated and no significant relationship was found. The kaolinite, however contributed more to the moisture adsorbed by the higher ranked bituminous coal in comparison to the lower ranked bituminous coals. This could most likely be attributed to the fact that the water uptake by the organic material of higher ranked coal is less than that for lower ranked coals. The amount of moisture adsorbed by the kaolinite seems to be less for lower ranked coal containing more oxygen and more for higher rank coal containing less oxygen. It can thus be said that the amount of moisture adsorbed in the different coal samples were influenced by kaolinite but to a lesser extent for the lower ranked coals. QEMSCAN analysis also displayed increased levels of calcite and pyrite present in the lower ranked coal samples and increased levels of illite and muscovite present in the higher ranked coal samples.
A positive relationship was observed when comparing the amount of moisture adsorbed and illite content for coals similar in rank. Increased levels of illite corresponded to increased levels of moisture adsorbed for the lower ranked bituminous coals. There was a significant amount of illite present in the higher ranked bituminous coal but no significant increase in the amount of moisture adsorbed was observed. Lower water adsorption surface areas were observed in comparison to CO2 surface areas. It was also found that the mineral matter present in the coal samples inhibited the CO2 adsorption surface areas. Modelling of the experimental data indicated that the monolayer adsorption capacity, estimated by the BET model, correlated very well with the surface oxygen content of each coal sample. This is an indication that moisture is first adsorbed at the surface oxygen groups. The modified BET model described the moisture adsorption mechanism very well for each coal at the relative pressure range applicable to this study. From the modified BET model the contribution of water adsorbed due to primary and secondary sites could also be estimated. Energies for the primary sites, ranging between 44 kJ/mol and 50 kJ/mol, were higher than those for the secondary sites, varying between 42 kJ/mol and 43 kJ/mol. This indicated that the water–coal interactions in the monolayer were weaker than those interactions in subsequent layers. The parameters estimated from both models correlated very well with the values presented in the literature. | en_US |
