Technical Market Support » Metallurgical Coal
Three coking coals and three PCI coals were selected to study the effect of coking and heating coals on the minerals present. The coals were selected to cover a wide range of mineral types. A number of techniques were used to characterise the mineral matter, including:
- Powder X-ray diffraction (XRD) directly on the coal / coke
- XRD on low temperature and high temperature ash
- Computer controlled scanning electron microscopy (CCSEM)
- Optical microscopy
Mineral Matter in Coal and Coke
Even for coke, evidence was seen for significant changes in the minerals caused by high temperature (815C) ashing and it is recommended that low temperature methods, such as RF ashing be used.
XRD based techniques were good at identifying the crystalline structures present but problematic where glassy, or species of poorly defined structure were involved, such as the identification of some of the complex species formed during the heating of the cokes.
SEM based techniques, such as CCSEM were not as effective identifying the actual mineral species, and in the case of the instrument used in this study the lack of information on low atomic weight elements limited the ability to identify reduced species in the heated cokes, e.g., SiC. However, the ternary diagrams generated by the technique do not require individual identification of mineral species and proved very useful for studying transformations occurring during the various processing steps. The CCSEM was successfully applied to cokes and relatively large coal pieces. Some further work is required to better define the sampling requirements for these relatively large materials. The CCSEM also identified included and excluded minerals in the coals. These were compared with qualitative optical microscopy analysis and a reasonable agreement was found. The CCSEM could also be used with very small sample masses, which is an advantage in PCI and drop tube combustion experiments.
Behaviour of Mineral Matter in Coking Coals During Heating (Coking and Subsequent Heating of Coke)
When there were high levels of non-aluminosilicates present in the original coal a complex mixture of species can form during heating to high temperatures. These cannot easily be characterised into recognised mineral species, but were readily studied using the ternary diagrams from the CCSEM results.
During coking the kaolinite could dehydrate to metakaolin or through to mullite. At this stage it is not clear what causes one type of behaviour over the other. CCSEM and optical microscopy were used to look at the distribution of the minerals within the coal matrix and could not identify significant differences. Coal A formed metakaolin during coking and the kaolinite was present as both finely disseminated material and clay lenses. Coal B formed mullite and contained a similar distribution of the clays.
Coal C formed mullite and the clays were present mainly as lenses. Coal A did contain a lot less kaolinite than the others. Even when the clays decompose to mullite during coking the excess silica does not appear to be mobile. The CCSEM ternary diagrams show that the average composition of the mineral grain does not change significantly from the original coal. However when the cokes were heated the change in grain composition could be seen as a loss of silica in the ternary diagrams.
The illite surprisingly survived the coking procedure. Heating bulk illite causes decomposition to spinel at around 1000C. Higher temperatures caused the decomposition of the illite (to potentially a number of species) and the release of the alkali metals contained in the clay. These were lost from the coke during the heating.
The CCSEM technique also has the capability to examine changes in mineral particle size during heating. The results indicated there was little change in the mineral particle size during coking. There was some evidence for decrease in the particle size during heating of cokes to high temperatures. This was associated with the loss of some inorganic species from the coke. However, with the relatively large size of coking coal particles studied difficulties were encountered. Modifications to sampling techniques and number of particles examined need to be made to improve the confidence in particle size results.
The percentage loss of silicon during heating of cokes was approximately the same for all three coking coal samples. This is despite one sample containing predominately quartz and the others predominately clay minerals. Fume collected in a cold trap on the outlet case from the experiment confirmed the vaporisation of silicon (mainly) and small amounts of sodium, potassium, magnesium and sulfur.
Behaviour of Mineral Matter During PCI Combustion
Coal burnout during combustion in a pilot scale PCI experiment was determined for three coals under several different conditions. Because of the relatively large uncertainties involved only qualitative conclusions were made, as follows:
- The effect of coal volatile matter and blast temperature on the amount of coal burnt was small. There was a marginal increase in burnout with increasing blast temperature and increasing volatile matter.
- Burnout shows a small increase with increasing O/C ratio of the blast.
Silicon vaporisation was also estimated for the three coals under several different sets of conditions. Evidence for both loss of silicon and gain of silicon from recondensation were observed in the data. Coal F consistently showed lower silicon loss than the other two coals. The minerals in Coal F, as assessed by CCSEM, were largely excluded and not intimately mixed with the carbon material. The quartz in Coal F was smaller in size than in the other two coals.
The CCSEM also indicated that the mineral particles after combustion were slightly larger than those in the feed coal. Fusion and coalescence of adjacent mineral species appears to be occurring. This is confirmed by the complex chemistry of individual mineral grains seen in the ternary diagrams of the PCI chars compared with the chemistry of the mineral grains in the feed coals.