Probability-Assigned Constrained Kriging For Precious Metal Reserve Modeling

The production data of many mines have not reconciled geostatistical estimations based on exploration and development drilling. The discrepancies in recovered metals are sometimes as large as 30%. A number of factors are responsible for such unpleasant outcomes. A major cause, however, is the lack of geological controls in the implementation of geostatistical techniques. Geological modeling establishes geological constraints and mineralization/ore envelopes. Geological constraints are usually established through drill hole interpretation. Mineralization and ore envelopes have been delineated manually based on assay grades. The issues associated with manual methods are: •Time requirements . •Inconsistent dilution. •Uncontrollable implementation variance. Although the time issue may not be a big concern for some companies, the other two have been responsible for many undesired estimation errors. A new geostatistical approach described in this study is designed on the basis of indicator and ordinary kriging. The central component of the method is the procedure for automatic delineation of mineralization and ore envelopes. The method, referred to as probability assigned constrained kriging (PACK), is demonstrated on a Carlin-type gold deposit. The estimation results are compared to blast-hole based production panels, showing that the estimated tonnage is about 2.1 % higher than the actual production, and the average grade is about 2.4% lower.