Image analysis is more and more considered as a powerful tool for quantitative analysis of ores and concentrates. Apart from being very much dependent upon the quality of the imaging principle and the adequacy of the imaging sensor, quantitative image analysis is also the result of a long chain of material, image and pixel sampling strategies. The most reliable way to demonstrate the accuracy and the precision of modal analysis obtained from images is probably to use a blended material with known proportions of pure minerals and confront the results obtained from different analytical tools. This operation has been performed on a series of mineral blends mimicking the tailings of a flotation operation in a plant operating on massive sulfide mineralisation. The intention was to be able to better predict acid mine drainage potential from direct mineralogical observations._x000D_
Selected sulfide minerals: pyrite (Py), pyrrhotite (Po), chalcopyrite (Cp) and sphalerite (Sp) were mixed as variable amounts within gangue minerals (quartz, calcite, dolomite and siderite). Typical blends contained six to ten per cent (Py); one to three per cent (Po); 0.5 to one per cent (Cp, Sp); 15 to 40 per cent carbonates and 40 to 75 per cent quartz by mass. The 1000 gram blends were carefully sampled and prepared for X-ray diffraction, chemical analysis by ICP-AES and optical microscopy image analysis (OMIA). Results obtained for the different analytical techniques are compared and discussed revealing the limits and strength of each method in the accurate quantification of sulfide and gangue minerals.