Metallurgical Sample Selection: A Simple Step that has the Potential to Sink your Project

Many metallurgical programs commonly suffer from samples that are not representative of the “rock recipes” that will eventually see the inside of a processing facility. Unfortunately, this often introduces a bias that is transmitted through all the detailed technical studies all the way to the construction of a flawed processing plant. The most common blunders include selection that is limited to the richest ore type, to a single host rock, to rocks excluded from the mine plan, or the lack of consideration for diluting rock types.

Discussion will focus on the appropriate scale of characterization including the introduction of the concept of time, mine schedules and plant scale into the selection process. The lack of consideration given to appropriate sample selection directly impacts the forecasts for hardness/recovery, flowsheet design and ultimately profitability of mining projects.

Mining BIG Data: the Future of Exploration Targeting

The application Machine Learning enhances our capability to harness exploration data to establish vectors to ore. However, careful consideration of the inputs and outputs by human geologists is required to ensure that the model does not merely predict what is already known or produce spurious results.

Resource Over-Estimation: Post-Mortem and a Path Forward

Several high-profile mineral exploration projects have recently suffered major setbacks following significant reductions in Resource and Reserve estimates. This talk explores the root causes for these, as well as some possible solutions.