Statistical Methods For Mineral Engineers -

You cannot measure every ton in the stockpile. You take 30 samples. What can you say about the remaining 500,000 tons?

In the processing plant, statistical methods are used to monitor efficiency and optimize recovery. Statistical Methods For Mineral Engineers

Design of Experiments (DOE): Systematically varying process parameters (e.g., pH, reagent dosage, grind size) to determine their effect on recovery. This is far more efficient than "one-factor-at-a-time" testing.

Modern practice uses weighted least squares, where each measurement is assigned a variance (from sampling and analytical error). Measurements with low variance receive small adjustments; bad actors receive large adjustments—flagging them for review. You cannot measure every ton in the stockpile

Practical output: A reconciled feed grade that is statistically more reliable than any single direct measurement. Modern practice uses weighted least squares, where each

| Tool | Mineral Engineering Application | Why Interesting | |------|--------------------------------|------------------| | Moving Average (EWMA) | Real-time smoothing of XRF assay streams | Filters out high-frequency noise to show true trend. | | Control Charts (Shewhart) | Monitoring mill power draw, density, pH | Detects special-cause variation before a spill or crusher jam. | | Linear Regression | Relating Bond Work Index to throughput | “For every 1 kWh/t increase in Wi, throughput drops 12 t/h.” | | Monte Carlo Simulation | Predicting monthly metal production given grade and recovery uncertainty | Turns “maybe 10,000 oz” into “10% chance <9,200 oz, 50% chance ~10,500 oz.” | | Taguchi Methods | Designing a flotation reagent dosage experiment with minimal tests | 8 experiments instead of 81 – finds optimum without bankrupting the lab. |