A Procedure To Audit A Sampling Protocol

Discrepancies in raw materials properties informed by both producer or consumer is a common problem in most areas of the industry. Commercial contracts in the mineral industry normally penalises the producer, if contaminants in the ore exceed established limits. In some situations, the supplier is penalised if the quality of the ore is below a contracted amount. Differences in lab analysis are due many sources of error, such as distinct practices used in two laboratories for sampling, preparation or analysis. Rigid contracts impose to the miner the use of rigorous quality control program and in these situations sampling plays a major role. Sampling theory provides the tools to analyse errors involved in sampling, preparation and analysis. This paper proposes a methodology to check the precision and accuracy of a given sampling protocol emphasising the application of Gy's sampling theory to broken ore. A procedure to measure the error on each step of the protocol is introduced and the results compared to the fundamental error measured by Gy's theory. The results show the applicability of the method and its relevance to audit sampling procedures aiming at error minimisation. The methodology is illustrated in a case study at a major coal producer in Brazil.