Issue 5, p. 43 (2015)


Placer gold sampling—The overall measurement error using gravity concentration on particle size ranges during sample treatment

  • Stéphane Brochot  
  • François Mounié
IDM Guyane, Rémire Montjoly, Guyane France
E-mail: [email protected]
 Search for papers by this author
 Corresponding Author
Caspeo, 3 avenue Claude Guillemin BP36009, Orléans CEDEX 2, France
[email protected]
 Search for papers by this author

Placer deposits are generally characterized by low grade of free gold. This is the case in French Guiana where the main placer deposits are in the river bed. Most have already been exploited by very small mining companies using sluices. If this technology is efficient for coarse gold, it releases fine gold in the tailings. During the last years, studies have been performed on various sites and recoveries have been estimated between 40 and 60% depending on the size distribution of gold particles and of the quality of the sluice configuration. Many recent or ancient tailings are available with a non-negligible quantity of remaining gold, offering retreatment opportunities. They are generally found in the form of sand heaps with the shape of an alluvial fan originating at the sluice discharge. Due to the resulting large distribution heterogeneity it is necessary to take many samples at many strategically deployed locations. These samples have to be large enough to be representative of the local material. As gold is mainly liberated in this type of lot, traditional sample treatment with successive size reductions and sub-samplings is not efficient and can be very expensive. Another approach using sieving and gravity concentration per particle range is preferred and presented here. After presentation of the sampling and measurement protocol used, this paper focuses on estimation of the overall sampling error. Various tailing cases are presented for which retreatment decision depends on the level of confidence obtained for the estimate of the quantity of recoverable gold.




Abstract Views: