Development and Calibration of a Reconciliated Mineralogy Method Based on Multitechnique Analyses: Application to Acid Mine Drainage Prediction

The field of quantitative mineralogy has evolved significantly through the development of image analysis techniques and information technology. Quantitative mineralogical characterization can now be routinely conducted for determining ore compositions and understanding the behaviour of minerals within mineral separation processes. The characterization of mine tailings is another important application of quantitative mineralogy. The estimation of acid and neutralization potential as well as the interpretation of kinetic leaching test results are often based on quantitative mineralogical knowledge (mineral abundance, texture, metal distribution amongst minerals, etc.). The objective of this paper is to determine and compare the accuracy of four mineralogical characterization techniques that are often used for environmental sample analysis: (1) X-ray diffraction, (2) normal analysis from chemistry analyses, (3) normal analysis from specific gravity measurements and (4) SEM-EDS image analysis. For this purpose, three artificial samples were assembled by mixing pure minerals in proportions often observed in mineral processing plant tailings. BILMAT, a statistical data reconciliation software, was applied to reconcile the chemical analyses with the results of other mineralogical analysis techniques and to estimate the acid and neutralization potential of actual mine tailings samples. The acid mine drainage was qualified from the acid potential and neutralization potential equilibrium as determined through quantitative mineralogical analysis. The statistically reconciled data made it possible to predict the metal deportment as a function of their bearing minerals and therefore, in the case of neutral contaminated drainage, to assess metal leaching potential.