Traditionally, a single sample can determine the fate of a large volume of material in a mine. This decision, often involving tonnes of ore or waste, hinges on the accuracy of that one sample. This reliance on a limited data point presents a significant cost risk.
To illustrate, consider the impact of sample size on mine tonnage. During my time as Chief Geologist at the Longos Gold Mine, the open pit in Paracale, a single sample represented 7 tonnes of in-situ material. A misclassification could lead to incorrectly categorizing this material, resulting in lost revenue (if ore is classified as waste) or unnecessary processing costs (if waste is classified as ore).
Quantifying the Impact:
Let's consider the economic consequences of such a misclassification. Assuming the average gold content per tonne of ore is X grams (g) at a price of Y dollars per gram ($/g), the potential financial loss due to a misclassified 7-tonne batch would be:
- Loss from discarded ore: (X g/tonne * 7 tonnes) * ($Y/g) = Z dollars
- Additional processing cost for waste: (processing cost per tonne) * 7 tonnes = W dollars
- The total potential cost (Z + W) highlights the importance of using an appropriate sample size to minimize misclassification risks.
Moving Forward: A Data-Driven Approach
This example highlights the high stakes associated with limited sample size. Implementing statistically robust sampling techniques, such as collecting multiple samples, can significantly reduce the economic risk impact of misclassification and improve overall mine profitability.
