Liberation analysis is a powerful tool to gain insight into the mineralogy of an ore and how this may affect metallurgical processing. However, in some cases more detailed investigation of a particular aspect of process mineralogy is required to improve understanding of, for instance, realistic grades and recovery that can be expected.
The SEM platform by Zeiss that we use for liberation analysis allows exporting of mineral contents and size/shape descriptors for each particle that is analysed, as well as each grain contained within these particles. For other platforms such as MLA and QEMSCAN, GSL have developed image processing routines that allow extraction of similar particle-by-particle data from image grids that these platforms can produce.
The availability of detailed information for each particle enables a host of other detailed analyses. On the basis of this, we have developed a number of models that allow us that delve deeper into a specific aspect of the ore mineralogy. Examples of the services :
- Bootstrap analysis to determine sampling characteristics of a sample
- Generation of theoretical grade/yield-recovery curves for the full sample or for particular size fractions
- Prediction of process response based on mineral properties such as density or flotation response
- Stochastic simulation that relies on particle size to simulate grade/mass yield/recovery and predict the variation in these parameters that may be expected based on mineralogy
Below are a number case studies that give more detail on several of the process response modelling tools that we have.