Part 2: Experimental calibration of sampling parameters K and alpha for Gy?s formula by the sampling tree method ? Synopsis

Application of Gy?s formula is appropriate for calculation of the Fundamental Error (FE) prior to samples being collected at any stage of the minerals process, provided that sampling parameters K and a (alpha) have been determined by calibration of the ores. Two types of calibration procedure are well established in the industry, the Hetrogeneity test and the 32-piece sampling tree experiment. The sampling tree experiment uses the variance and average grade of metal analyses from four series of samples at different nominal fragment size, as the basic input data. Pierre Gy suggested the cube as an appropriate exponent of the nominal fragment size for use in his equation. More recent work suggests that this exponent is variable and characteristic of specific ores and therefore requires calibration. The method described here explains the experimental procedure of the calibration. It is then possible to calculate a sampling nomogram using the sampling parameters K and a. This allows the sampler to determine a suitable protocol for the sampling event that will ensure that the FE does not exceed a predetermined precision at any point in the sampling procedure.