The 2017 GAC-MAC annual meeting is a special edition: Their 150th anniversary edition! And contemplating the past is an inspiration for the future of geosciences, and let's be part of it.

Following its tradition of innovation for mineral exploration techniques, IOS makes a special effort to deliver to the scientific community its last achievements, especially in geometallurgy and exploration geochemistry:

Automated optical gold grain counting: A quantum leap

Session Q51: Glacial Processes and Deposits: Advances and Application : Monday, May 15, 3:00pm

Presented by Mr. Alexandre Néron, this talk will present the recent advances in automated gold grain counting in glacial sediments. This method, widely used by the industry, remains the most efficient technique for the detection of blind gold deposit under a glacial sediment cover. However, this technology has little evolved for 40 years; however it will be shown how it can be improved with the use of modern technologies. Gold grains in deposits are minutes, and it is demonstrated that minute grains are poorly recovered by conventional methods, such as shaking table and binocular counting. The ARTGoldTM procedure, based on SEM automated gold grain detection, although outstandingly efficient, remains limited by machine availability. Thus, a similar automated procedure has been developed on a motorized petrographic microscope. The method uses optical recognition and image analysis to scan heavy mineral concentrates and detect gold grains as small as 5 micrometres in size. Almost as robust as SEM scanning, it allows to detect, image and locate the grains, which can either be classified in terms of morphology, or transferred to a SEM for EDS analysis. Being accomplished on affordable equipment, the method will allow multiplication of the processing units and to circumvent SEM bottlenecking.

Gold Deportment: Insights from Exploration Techniques

Session GE1 : Geometallurgy - From Exploration to Remediation: Monday, Mai 15, 9:30am

Presented by Mr. Réjean Girard, this talk aims to bring a fresh look at the gold deportment issue in mills, as seen from the eyes of a heavy mineral laboratory operator. Although gold abundance in mine tailing is only tens of ppb, this gold can amount to tens of millions of dollars for a mid-size mining operation. Consider that about one ton of gold is lost per year in operations like Canadian Malartic or Eleonore mines. To improve gold recovery, it is primordial that its deportment is quantified properly. However, the problem of the very rare occurrence rate of gold grains makes it almost impossible, or at least non-efficient, to use conventional QEMSCAN or MLA approaches, since they will not provide a dependable representativeness within a realistic counting time. Pre-concentration is required, which is impaired by smallness of the grains and the low abundance. An array of innovative concentration and counting methods will be presented, as routinely used in exploration. These include the ARTGoldTM fluidized bed concentrator for grains larger than 20 micrometres, a centrifuge-based heavy liquid concentration using TSL at a density up to 4.0 g/cc, as well as using conventional SEM-based methods for encapsulated or non-liberated gold grains. Examples will also be presented.

Lift-Index : A method for the recognition of complex and heterogeneous signal in soil geochemistry

Session EI1 : The Transport and Fate of Contaminants in the Natural Environment, Monday, May 15, 2:20pm

Resulting from 35 years of experience in exploration geochemistry, this talk presents the most efficient approach developed by Réjean Girard in geochemical survey interpretation protocols. Recognition of the footprint of a mineralized occurrence in the secondary environment is the cornerstone of exploration geochemistry. Deciphering this signature is typically done using statistical approaches, which are limited to cases where metals are sufficiently released by the occurrence and fixed by the environment in order to generate a contrasting signal. However, statistics fail to reveal such signatures in cases where the metal behavior is severely affected by heterogeneous physical or chemical environment. A deterministic approach, trying to deconvolute the modification of the geochemical signal depending on physicochemical conditions prevailing through the liberation, transport and accumulation environments, can remove part of the variability, improving signal to noise ratios. However, where the environment is complex or heterogeneous, deconvolution of the signal may become very fastidious, if not impossible. To circumvent these limitations, the use of the machine learning based on Bayesian statistics is proposed. The approach enables to characterize complex or heterogeneous signatures associated with a known mineral occurrence, and to detect such signature within a larger dataset outside the realm of the target occurrence. Using HyperCube® platform, a Bayesian estimator has been developed which calculates the similarity between each analyses of a survey and the expected multivariate signal associated with the known deposit. This estimator, the "Lift index", is the normalized sum of every "simple-lift" associated with an analysis, where these are the over-representation ratio of a value in the target population compared to its abundance in the overall population. Compared to conventional statistics, the method does not take account of only elevated or anomalous values, but of every range of values present in the population, as well as the rate of occurrence of non-numerical entries such as colours, etc. Examples will be presented for surveys conducted over the Abitibi clay belt, where mineralized occurrences were detected below 50 metres of overburden.

Modal, Textural and Chemical Variations in the Vanadiferous Magnetite Deposit of the Lac Doré Complex (Québec) and its Implication on the Chemical Purity of Magnetite Concentrates

Session GE1 : Geometallurgy - From Exploration to remediation (Poster): Monday, May 15

This poster shows the preliminary results of Jean-Phillipe Arguin, a doctorate student at UQAC. It discusses the mineralogical variability across the stratigraphy of the magnetite layered sequence of the Lac Doré Complex. This deposit is currently under evaluation by VanadiumCorp Resource, for the production of high purity vanadium chemicals needed as electrolytes for redox flow batteries. The study described distribution of contaminants susceptible to ends in magnetite concentrate across the stratigraphy of the complex, either as solid solution in titanomagnetite, as exsolutions or caused by non-liberated phases. Textural relations between minerals are quantified in order to predict their liberation in milling circuit. A detailed attention on ilmenite and hercynite exsolution textures, size and abundance is addressed.

Optimal Density to Fractionate Organic and Mineral particles of Québec Boreal Lake Sediments

Session EI1 : The transport and fate of contaminants in the natural environment (Poster) : Monday, May 15

This poster presents the results of Ms Mélanie Aubin master studies, which deal with metal deportment in lake bottom sediments from northern Québec. Lake bottom sediment geochemistry is a greenfield exploration technique widely used by the mineral industry for the determination of regional endowments. Surprisingly, very little is known about where and how the metals are accumulated and fixed in such sediments, and more specifically on which components of the sediment do the metal are deported. Lake bottom sediments, or gyttja, are made of various proportions of detrital silicoclastic particles, clays, precipitates such as oxides, hydroxides and carbonates, fibrous undecomposed organic remains, fully decomposed organic matter, and silica tests from radiolarians and diatoms. Each of these components can contain or scavenged metals in various proportions, and thus control the metal budget of the sediment. The presentation discusses results of a method development on organic matter gravimetric separation, as well as preliminary results on metal distribution by the use of EDS mapping with a SEM.

Photomosaic of the BSE image of gold grains recovered from 2.5 kilograms of tails from an Abitibi mine. Grains, ranging between 6.5 and 57 microns, accounts for about 15% of the 70 ppb reported in the tails.

http://www.kingstongacmac.ca/en/home/