Coal Preparation » Fine Coal
The overall objective of this project was to leverage off the significant past research on better classifying cyclones to build and demonstrate the industrial benefits of the flat-bottomed classifying cyclone technology at an appropriate scale, namely 380mm diameter cyclone. Specifically, coal-independent performance efficiencies were to be determined using:
- Size separation partition curves and related characteristic parameters α, d50 and Rf;
- RD-by-RD size partition curves and related characteristic parameters α, d50 and Rf.
These were then to be used as comparators against such curves generated by contemporary classifying cyclone technologies.
The new flat-bottomed classifying cyclone design produced pilot scale separations characterised by:
a. Very low Rf values tending to zero
b. α-values of 2.59-2.64.
- Narrow RD fraction basis:
a. Variable Rf values in the range 0.00 to 0.72.
b. Variable α- values in the range 1.21 - 6.28.
Measured α-values are considered artificially low due to feed %solids changing during testing. If better controlled, as occurs in properly operated industrial plants, sharper separations are anticipated.
A CFD/DEM model was developed to aid the proposed design of the flat-bottomed cyclone and to evaluate the working conditions on a preliminary basis. Such a model was not fully exploited in this project and improved flat-bottomed unit configurations, leading to higher capacity and α-values, are likely outcomes from further work in this area. As is extension beyond the 380mm diameter limitation.
As analytical testing was restricted to unit g environments, segregated narrow RD F/S analyses were confined to plus 38-micron particles only, which represented around only 53-55% of the feed solids, and much less (15.9-16.1%) of the overflow solids. This impacted the degree of definition that could be afforded to each narrow RD range partition curve. However, the effect on the overall partition curve determinations was much less.
It is considered there are significant further efficiency improvements to be had for the flat-bottomed cyclone design. These could possibly be realised by optimising flat-bottomed dimensional aspects, conceivably enhanced by A.I., to reduce d50c variation with particle RD. This optimisation could most efficiently be undertaken by CFD/DEM modelling followed up by practical demonstration.
The greatest source of measurement uncertainty is believed to reside in the fine coal float/sink procedure. Quantification of these uncertainties was not attempted.
The main findings of this project are that, from pilot scale testing a 380mm diameter flat-bottomed classifying cyclone, its overall performance can be characterised as exhibiting overall α values higher than around 2.6 and overall Rf values tending to zero.