Sampling error is the degree to which the concentration of an element differs from the true element concentration of the material from which the sample was collected. Gold mineralization commonly exhibits sampling errors as large as 50–100%. As a result, collection and preparation of drill samples from Au mineralization can provide significant challenges for the geoscientist, largely because of the coarse particulate nature of Au. To avoid this, geoscientists have opted to collect and prepare larger drill samples to reduce the magnitude of this ‘nugget effect’. Unfortunately, the ‘nugget effect’ can cause sampling error at any stage of sample treatment when sub-sampling takes place. Knowledge of the magnitude of error at each sub-sampling step is necessary to identify strategies to reduce overall error. This is because reduction of only the largest component of measurement error will reduce total measurement error in the most numerically efficient and effective manner. With knowledge of the various components of sampling, preparation and analysis costs, the sample treatment strategy that will most cost-effectively reduce sampling error can be identified. Regrettably, the typical absolute and relative errors that occur during sample collection and preparation of sub-samples with finer particle sizes are not generally known for many Au deposits.
Results from three Au drilling projects document the magnitude of sampling, preparation and analytical errors experienced, and range from 22 to 46%, 7 to 20%, and 1 to 13%, respectively. These results are derived from large, laboratory-blind, duplicate quality control/quality assessment (QA/QC) programmes involving sample treatment protocols that would be considered to be appropriate for coarse Au-bearing samples. These QA/QC programmes measured the sampling errors, and ensured that they were minimized on these projects.
In general, results indicate that a very large component of total measurement error is introduced during the collection of the initial sample, and that subordinate amounts of error are introduced during preparation and analysis. As a result, undertaking extraordinary efforts to reduce preparation or analytical errors does not result in a significant total measurement error reduction. In contrast, the collection of larger initial samples can result in the substantial reduction of total measurement error.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.