Full-scale flotation performance for coal is always difficult to predict, and laboratory bench-scale testing is prone to significant operator error. While fine float and sink testing may provide an indication of a potential yield-ash curve, it too is subject to major errors.
Several forms of testing are now available to better predict flotation performance.
The ‘ultimate yield-ash curve’ for a particular feed can be determined using coal grain analysis (CGA – similar to mineral liberation analysis), in conjunction with laser particle size analysis and basic ash-by-size data. CGA data only characterise particles larger than one micrometre, and some flotation feeds contain significant amounts smaller than one micron. Thus the laser particle size analysis, in conjunction with sieve size analysis and ash-by-size may be used to back-calculate the characteristics of the minus one micron particles.
The ultimate yield-ash curve will not generally translate to actual full-scale flotation practice, so separate flotation testing is also required to translate the ultimate data to practically-achievable data. A range of specialised laboratory test types are available which are significant improvements on the AS/ISO standard methods. Basacon can assist clients to properly design and interpret sampling and analytical programs.
When CGA data are applied in combination with full-scale flotation sampling data, accurate performance models based upon the flotation feed CGA may be generated.