Gold Sniffer Measurements and Analysis of the Tailings From a Floatation Process at a Gold Mine

Jim Kendall, Jack Cole, Cynthia Charlebois, and Chris Brown


The Gold Sniffer is a portable device that analyzes high resolution digital images of the surface of an unpolished mineral sample, such as drill core or a grab sample cut by a diamond saw, to detect gold particles as small as 1.6 microns as shown in Figure 1. It does this by determining which pixels have the colour of the gold alloy in a particular deposit, and then using a spatial algorithm to detect the edges of the gold particles. it uses the fractional coverage of gold in the imaged area to estimate the local gold grade. Averaging the local gold grades for typically 10 to 30 measurements provides an estimate of the gold grade of the sample. It requires 90 seconds to perform a measurement. The Gold Sniffer detects minerals in the same fashion as gold, by seeking the colour of a mineral as it is found in a particular deposit and using a spatial algorithm to detect the edges of the mineral particles. The Gold Sniffer, shown in Figure 1, consists of a digital camera, macro lens, a light source that illuminates uniformly at micron dimensions across the visible spectrum, a stage to position the measurement sample, and software that processes the digital image to identify gold and mineral particles, and display the results.

Jim Figure 1


The Gold Sniffer was used to measure tailings from a flotation process that concentrates the ore at a gold mine by extracting the sulphide minerals pyrite, chalcopyrite, gelena and sphalerite. The tailings were mixed with epoxy and formed into polished samples with 30 mm diameter. the Gold Sniffer can measure either polished or unpolished mineral samples. since each Gold Sniffer measurement requires an area of 3mm x 4.5mm, it was possible to take 24 measurements from each of the four polished samples. The Gold Sniffer measured the locations and sizes of particles of gold, pyrite, chalcopyrite, gelena and sphalerite as small as 1.6 microns. An example of a Gold Sniffer measurement is shown in Figure 2 with the locations of the gold particles indicated by the red crosses.

Jim Figure 2

The fractional coverage of each picture was used to calculate the local concentration for each mineral. Averaging the 24 local measurements provided an estimate of the gold grade and the concentration of each mineral for each sample (see Table 1).

Jim Table 1

By observing the mineral particles around each gold particle it was found that 97% of the gold particles were either on, or within 5 microns of a chalcopyrite particle. The remaining 3% of the gold particles were associated with pyrite particles. Based on the close association of gold and chalcopyrite, the gold grade of the tailings was plotted as a function of the chalcopyrite concentration of the tailings, and a straight line was fit to the data, as shown in Figure 3.

Jim Figure 3


The least squares fit line in Figure 3 shows that if more chalcopyrite is extracted from the ore by targeting this mineral in the floatation process, then more gold would be extracted. This could be achieved by optimizing the floatation process for chalcopyrite, rather than using a process that is non-specific that seeks to extract all sulphide minerals such as chalcopyrite, pyrite, sphalerite and galena, as is done at present. The slope of the least squares fit line of 719.72 ppm, provides a guide as to the gain in gold recovery that can be achieved if more chalcopyrite is extracted from the ore in the floatation process. The least squares slope specifies that for each reduction of 0.0138943% in the chalcopyrite concentration in the tailings, there is a 0.1 ppm reduction in the gold tailing grade. Note that 0.1 ppm/ 0.0138943 = 719.72 ppm. The average of the chalcopyrite tailings grade data is 0.121%. Dividing 0.0138943% by 0.121% provides the result that for each 11.5% reduction in the amount of chalcopyrite in the tailings, there is a 0.1 ppm reduction in the gold tailings grade.

A 0.1 ppm reduction in gold tailings grade would increase gold recovery by 0.1 ppm, or in other words 0.1 grams of gold per tonne of ore. If it is assumed that this increase in gold recovery in uniform for the 450,000 tonnes of ore that the mine produces per year, then this would produce an additional 1447 ounces of gold per year. Note that a Troy ounce is equal to 31.1035 grams. However, the gold and sulphide content of the ore is likely variable, so it is important to repeat the Gold Sniffer analysis described in this paper on tailings that result from processing ore from different parts of the mine. It would also be valuable information to adjust the conditions of the floatation process to focus the process on extracting chalcopyrite and use the Gold Sniffer to measure the gold grade in the tailings. If the gold tailings grade is lower when the floatation process is focused on extracting chalcopyriterather than extracting sulphide minerals in general, particularly if this could be used to consistently increase gold recovery. Therefore, this study of floatation process tailings using the Gold Sniffer has defined a set of experiments that, if the results are similar to those presented in this paper, could result in a significant increase in gold recovery.