Sampling in the evaluation of ore deposits

Sampling is an error generating process and these errors should be reduced to a minimum if an accurate ore reserve estimation is to be made from the sample values. Error in sampling can arise from the sampling procedure as well as where and how each sample is taken from the deposit . Sampling procedure involves sample collection, sample reduction and analysis, and the error from each of these three stages has an equal influence on the total error of the process. Error due to sampling procedure should be identified and eliminated at an early stage in the evaluation programme. An ore deposit should be subdivided into sampling strata along geological boundaries, and once these boundaries have been established they should be adhered to for the evaluation programme. The sampling of each stratum depends on the small-scale structures in which the grade is distributed, and this distribution in relation to sample size controls sample variance, sample bias and the volume of influence of each sample. Cluster sampling can be used where an impractically large sample is necessary to reduce sample variance or increase the volume of influence of samples. Sample bias can be reduced by composing a large number of small samples . Sampling patterns should be designed with reference to the volumes of influence of samples, and in favourable geology, geostatistical or statistical techniques can be used to predict the precision of an ore reserve estimation 1n terms of the number of samples taken. Different are deposits have different sampling characteristics and problems which can be directly related to the geology of the mineralization. If geology is disregarded when sampling an are deposit, an evaluation programme cannot claim to give an accurate estimate of the ore reserves