Optimizing Infill Drilling Decisions Using Multi-Armed Bandits: Application in a Long-Term, Multi-Element Stockpile

Area of Science:

• Mining Engineering
• Geostatistics
• Operations Research

Background:

• Mining operations require strategic decisions on additional drilling to optimize resource extraction.
• Geological uncertainty and complex blending requirements in stockpiles necessitate advanced planning methods.

Purpose of the Study:

• To present an optimization method for determining the necessity and optimal location of infill drilling.
• To enhance the value of mining operations by improving material type change and extraction sequencing.

Main Methods:

• Utilizing a multi-armed bandit (MAB) framework for infill drilling pattern optimization.
• Incorporating multiple conditional simulations to account for geological uncertainty.
• Defining the optimal pattern by maximizing material type changes within stockpiles.

Main Results:

• The proposed MAB method effectively optimizes infill drilling schemes.
• Demonstrated practical applicability in a long-term, multi-element gold mining stockpile.
• The approach successfully addresses difficult-to-meet blending requirements through improved material management.

Conclusions:

• The MAB framework offers an effective solution for optimizing infill drilling decisions in mining.
• Accounting for geological uncertainty is crucial for maximizing the value of mining operations.
• This method provides a practical tool for enhancing extraction sequencing and meeting processing targets.