Metal accounting is the estimation of (saleable) metal in a mine and subsequent process streams over a defined time period. Comparisons of estimates, from different sources over a specific time period, are called reconciliation. Metal accounting research is ‘themed’ into measurement network design and accounting for multiple ores sources. This research was initiated in 2004, and has involved a number of industrial test campaigns on mining operations.
Measurement Network Design: Metal accounting relies on measurements obtained from a network of samples and sensors in the plant. Metal accounting can only take place if a suitable measurement network, or system of samples and sensors, is in place. The choice of this measurement network impacts on the precision to which variables of interest, such as the final metal produced, can be estimated. This research aims to develop a methodology for designing metallurgical accounting measurement networks by mathematical programming and heuristics.
Accounting For Multiple Ores Sources: When ores are combined in blends it is commonly assumed that the metallurgical performance of the blend is additive i.e. equivalent to the weighted average sum of that for pure ores in the blend. However, there is a perception based on past experience that poor quality ores may control the metallurgical behaviour of ore blends. Here, the poor quality ore produces non-linear or ‘antagonistic’ effects, which weight the metallurgical behaviour of the blend towards that of the poor quality ore. This would impact significantly on areas such as production accounting where ores are considered additive in blends. The objective of this research area is to investigate the effects of multiple ore sources/blends on process accounting in terms of the impact on performance measures such as final recovery and grade.