Improving reconciliation and grade control by statistical and geostatistical analysis / Улучшение сопоставления и контроля оценок с помощью статистического и геостатистического анализа

Accurate recoverable resource estimation and grade control procedures are the foundation of successful mining ventures. Long, medium and short term planning in a mining operation are all dependent upon precise estimations. For example, poor estimation may result in the long term in a pit being incorrectly optimised, in the medium term cashflow forecasts may be disastrously inaccurate, and in the short term the allocation of ore and waste material by grade control may be erroneous. Many open pit gold mines in Australia suffer from a lack of reconciliation between exploration and in-pit estimates and actual head grades delivered to the mill. This is often despite very good grade control sampling and geological control. Often correction factors are applied to the resource model and grade control estimates in order to approximate reconciliation with production figures. Some of the main reasons for erroneous estimation are the failure to account for one or many of the following:
— the statistical distribution of the sample data,
— the spatial continuity of the sample data,
— the regression-effect and the volume-variance relationship,
— the degree of mining selectivity (ie., dilution) that is practised, and
— the selection of an appropriate estimation method to calculate recoverable tonnes and grade. <...>