Evaluation and prediction of slow pyrolysis products derived from coals of different rank

When coal is heated slowly (<10 °C/min), as in the Lurgi Fixed Bed Dry Bottom (FBDB) gasification process, products formed in the pyrolysis region of the gasifier include gas liquor, condensable tar, oil and non-condensable gases. Knowledge of the temperature profile together with the coal decomposition behaviour is of great importance when designing a fixed bed gasification plant, and although the characterisation of tar and the composition thereof has been reported extensively, there is limited literature available on the prediction of tar composition formed during slow pyrolysis. The focus of this study was to investigate temperature and coal rank effects on pyrolysis product yield, and to predict (using FLASHCHAIN®) the char, tar, water and gas yields when heated at slow heating rates for coals of different rank (ranging from lignite B to bituminous C). A modified Fischer Assay setup was used to investigate pyrolysis at temperatures higher than that of the ISO 647 standard, i.e. 520, 720 and 920 °C and the tar quality was determined via SEC-UV (size exclusion chromatography-ultraviolet), GC-MS (gas chromatography-mass spectrometry), SimDis (simulated distillation) and ultimate analysis. Only the char yield was found to be rank dependent and the average molecular weight of the coal derived tars (212-415 Da) compared well with previous studies. The rank dependence based on the composition of the evolved volatiles (tar and gas), showed a linear relationship with elemental oxygen and carbon contents of the derived tar, as well as for the oxygen containing gases (CO and CO2). FLASHCHAIN® was able to provide a relatively accurate prediction of the char yield, poorer predictions of the tar and water yields and no correlation with the gas yield. The simulated results on tar composition showed poor promise. Statistical regression was also applied in order to determine correlations between coal properties and the pyrolysis product yields and composition. It was found that the mineral elements (Na2O, MgO, CaO, TiO2 and Fe2O3) have strong correlations with tar yield, thus implying that catalytic effects of the mineral matter appear to play a significant role in the formation and decomposition of coal derived tar, which is a limitation in all pyrolysis predictive models.