Development of a Dense Medium Separation Process for Fine Coal Based on the Use of a Teetered Bed Separator

The agreed ACARP objectives of this study were:-

• To develop a cheaper alternative to dense-medium cyclone technology for the treatment of –0.5 mm coal
• To evaluate alternative to magnetite as a potential dense-medium
• To provide data on the fundamental nature of the dense-medium separation process.

This study was based upon the dense-medium separation that occurs in a Teetered Bed Separator (TBS).The TBS uses raw coal feed, in the size range –2.00 mm x 0.125 mm, as an autogenous medium. The unit has the advantage over other devices, operating on this size range, of providing a controllable low density cut point at high separation efficiency.Such a separator is ideally suited to this study because it provides a well defined hydrodynamic regime.

It is clear from this study that the TBS system provides a realistic alternative to dense-medium cyclone technology. The study showed that such units operate best at high pulp density and low upward current flow rate. An important feature of these units was that suspension density was primarily a function of this upward water flow rate rather than of the density of the medium itself. This characteristic is quite different to that observed with dense-medium cyclones.

The behaviour of the unit in an non-autogenous mode was also investigated using media such as magnetite, sand, barite and TBS rejects. It was found that the potential exists for significant efficiency improvements by recycling a portion of TBS rejects. The work presented suggests that reductions in Ep of the order of 20% can be achieved by adopting this strategy but at the expense of an increase in D(50). Whether this increase in D(50) is a problem depends on the washability characteristics of the coal in question and the required product ash. However, even at cut points as high as 1.7 these units still perform better than other separators on this size range [see Table.]

The theoretical analysis contained in this report suggests that it might be possible to achieve the is improved separation without a significant change in D(50) by judicious selection of the TBS recycle fraction characteristics.

Because of the low Ep values encountered and the uncertainties in their precise estimation, an alternative method of measuring process efficiency is presented in which the criterion of efficiency is measured by the variation of D(50) with particle size.

In this scenario, the D(50) is related to particle size by the relation

D(50) = KS^-N

Where k is a constant for a given separation and S is the particle size. A measure of the separation efficiency is given by the value of N. As N decreases to zero , the separation becomes independent of particle size and hence represents perfect efficiency.

It is the settling velocity rather than the particle size that determines the suitability of material for recyling through the TBS unit. However, in practice this classification is likely to be carried out with classifying cyclones so that particle size is likely to remain a dominant factor in the selection of recycle material.

On the theoretical front, a new slip velocity model is presented which was very successful in predicting process parameters such as D(50) from fundamental medium properties. The agreement achieved with experimental observations on both real and model systems was excellent demonstrating the achievement of an improved understanding of the dense medium separation process.

Obviously, there is a need for further work to capitalise on this recycle effect by better defining the relationship between recycle particle size and feed particle size. It is also necessary to define the most appropriate circuit for achieving optimum benefit. These matters can be best achieved by a collaborative approach between plant operators, suppliers and research groups and will hopefully result in the delineation of the optimum beneficiation route for the treatment of fine coal.