Metallurgical performance of gold and silver mineral processing flowsheets is typically evaluated using assay balance and production balance techniques. With assay balance the mill feed, tailings, and some intermediate streams are sampled and assayed. Gold circuit performance can then be evaluated using mass balance techniques. Unsteady plant conditions and difficulties in sampling and measuring moving streams are some problems with this method. Production balance measures two rather tangible quantities, tons milled and ounces poured, combines this with the tailings assay to calculate head grade and recovery. Changes to inprocess inventory must be monitored as the production balance is highly sensitive to this factor. The primary inventory location for flowsheets utilizing activated carbon is gold (and silver) on the carbon. Thus a production balance must include a sampling program to determine the amount of carbon in the various stages of process and the gold adsorbed on that carbon. To get an accurate estimate of the carbon inventory the sampling program must satisfy 5 conditions:
The leaching pachucas and the CIF vessels were modelled as continuous stirred-tank reactors, The flow out of each pachuca depends on the level in that and in the following pachuca, The flow from each CIF tank depends on the level in only that particular tank, The CIF vessels contain horizontal screens, which allow the pulp and carbon fines to pass downstream unhindered, but retain the coarse carbon, The coarse carbon can move downstream only if a screen overflows or tears, On the plant, and thus also in the simulator, a carbon-pulp mixture is pumped from each tank to the one above. during carbon transfers, In practice, the carbon-transfer pumps are switched on at relatively regular intervals, and are switched off when enough carbon has been transferred, In the simulation, the transfer pumps could be switched on and off as desired, so that various transfer schemes could be investigated, These included changing the frequency of transfers, and at the same time changing the magnitude of transfers to keep the amount of carbon transferred constant in the long term; changing the timing of transfers, etc
An upset carbon profile also upsets the solution profile, causing a loss of efficiency. In addition to the deliberately distorted profiles mentioned above, other disturbances to the carbon profile were also investigated, including attrition, carbon leaking through the screen, and screen blockages followed by overflows. Overflows and carbon leakage generally have only a local effect on the carbon profile, and their long-term effects on the overall efficiency are negligible. In contrast, the effect of attrition and the movement of carbon fines down the cascade was quite remarkable. An attrition rate of 83 g/t per hour would mean that about 3,7 kg of carbon are lost per hour, and yet this reduced the efficiency, as calculated by the simulator, by 2 per cent.