| Summary: | In a coal mining environment it is often difficult to relate the coal wash table data and analysis
obtained from exploration drill core with the run of mine material being processed in beneficiation
plants. This also affects the predicted yields, budgets and the resource and reserve determinations.
The aim of this study is to determine whether a technique can be developed whereby a more
accurate relationship between the exploration results used for the modelling, resource and reserve
calculations and budgeting can be reconciled with the beneficiation results of products extracted
from the resource. In a multi-seam and multi-product environment the problem of obtaining
products at various different specifications for differing markets is exacerbated. Most product
specifications are linked to the beneficiation environment, thus the need to determine relationships
between these properties and the final products is tantamount to success.
Fortunately one of the less obscure properties is the ash content of the coal. To this a host of other
specifications with regard to other impurities, such as sulphur or phosphorous content, may be
added, depending on the application of the end product.
The physical and mechanical properties of coal however are the most important in the initial stages
of the beneficiation process. Here density and particle size distribution play a major role, since, no
matter which process is used in recovering various products, density separation is the underlying
mechanism utilized whether it is by jig, drum, cyclones or flotation.
Coal plies and particles have different relative densities, determined by maceral composition, rank,
mineral content and porosity. The physical parameters of coal are mostly controlled by moisture,
volatile content (in terms of aliphatic or aromatic hydrocarbons), ash and carbon content which
inadvertently have an effect on the overall density of the raw coal. Since the majority of products
required are initially specified at specific ash content, the probable relationship of ash to density
would be critical in solving the greater majority of associated problems.
The sedimentary depositional environment of the coal deposit would have contributed to the
distribution of finely dispersed mineral matter within the coal plies. Changes within the
environment, largely due to climatic events and differing flow regimes, could influence the amount
of associated mineral material in these plies resulting in closely associated but differing densities
within the raw material.
At Grootegeluk mine two major stratigraphical units, representative of two distinctly different
sedimentary environments are being exploited. The first unit is the Volksrust Formation which
consists mainly of intercalated bright coal (vitrinite rich) and shale in a deltaic environment, where
the coal formation is considered to be autochthonous. The second unit, the Vryheid Formation,
consists of relatively well defined dull coal seams inter-bedded with siltstone and sandstone beds in
a fluvial environment considered to be allochthonous.
Based upon extensive studies of borehole analyses and sink and float tests, the derivation of a
technique applied to the relationship of coals’ physical properties has been developed which will
enhance the understanding of its behaviour in the beneficiation environment. This has provided
significantly better correlation between coal washing characteristics determined at fixed densities
within the analytical realm and actual beneficiation results. If adopted, this technique would
enhance the credibility of resource and reserve estimations, product yield predictions and actual
reconciliation of product to source, post mortem. In so doing, the results would have considerable
impact on marketing, economics and life-of-mine expectancy. Already the AAD methodology applied
to the grade control function in a beta-testing phase has resulted in the closest monetary value to
actual production income versus the predictions provided by the geological model and the mine
planning budgetary values.
|
|---|
