The practice of the sampling theory of broken ores

"In previous publications by the author, significant advances and generalizations of the theory of sampling of broken ores, prompted and confirmed by experimental results, have been presented. This paper shows how to implement these theoretical results in practice. The models to be used are reviewed, with a description of the various sampling parameters involved. Then the methodology for the experimental assessment of those parameters, including segregation factors, is described in simple and practical terms, along with the use of handy sampling protocol charts. Finally, the particularly important case of processing data from a sample-tree experiment is detailed, and examples of applications covering all situations and types of experiments are provided.IntroductionThe reliability of a sample of crushed material , i.e. the expected (average) deviation between its true grade and the true grade of the lot it comes from, mostly depends upon six factors:• the size (mass) of the sample (and to a minor extent, for small lots, the size of the lot);• the size of the rock particles (degree of comminution)• the nature of the mineralization at microscopic scale (i.e., within even the smallest particles);• the geological nature of the mineralization and mineral of interest;• the grade of the lot in the mineral of interest; and• in some cases, the state of segregation of the lot.In practice, one should first disregard segregation problems, and assess sample variability in the segregation-free case. Once the numerical parameters governing this variability have been assessed, then segregation factors can be assessed in turn. In the case of laboratory standard sample preparation procedures, segregation can usually be disregarded altogether."