The pre-concentration of gold using a Reflux Classifier, M. le Roux, J.H. van Vreden, Q.P. Campbell, and M. Ndabeni

It is not uncommon for gold slurry ponds or screen oversize dumps to contain several different types of carbon-rich contaminants. These contaminants act as robbers of dissolved gold which has a detrimental effect during the downstream carbon-in-leach (CIL) or carbon-in-pulp (CIP) processes. With a renewed interest in the reprocessing of contaminated gold waste ponds and dumps, preconcentration of the feed to either the CIL or CIP process becomes important to maximise profits from these resources. An added advantage in removing these resources, which were once thought of as being uneconomical, is the availability of land that can further be rehabilitated for community development projects. For gravity separation, an example of suitable technology is the Reflux Classifier (RC), which can be used for the separation of materials of different densities and can be used as a pre-concentration step. In this project, a laboratory-scale Reflux Classifier was used to pre-concentrate a gold-bearing sample obtained from a waste dump situated to the east of Johannesburg, South Africa. The material had an average gold grade of 1.9 g/ton and a carbon content of 10%. This material was first fed to a semi-batch Reflux Classifier that was used to fractionate it into several different density intervals by varying the fluidisation water feed rate, followed by continuous runs at fixed water flow rates. The results showed that a cumulative gold recovery of 60%, carbon recovery of 85%, and mass pull of 41% was possible for a fluidisation water feed rate of 15 L/min, while a rate of 20 L/min resulted in a gold recovery, carbon recovery and mass pull of 83%, 92% and 79% respectively, which translated into a concentrate gold grade above the set target of 3 g/ton. Continuous operation at 20 L/min yielded an underflow stream with 84% gold recovery at a grade of 5.4 g/ton. Keywords: Reflux Classification, gold recovery