Application of stochastic mine planning in an open-pit gold project
We are thrilled to have Marcelo Godoy, Chief Technology Officer at AngloGold Ashanti, as a keynote speaker for the Mineral Resource Estimation Conference (MREC) 2025. With over 25 years of experience in the mining industry, Marcelo brings invaluable expertise in resource modelling, mine planning, and risk management. In his keynote address, Marcelo will share insights on stochastic mine planning, highlighting how it can address geological uncertainties and improve resource classification.
Marcelo Godoy is the Chief Technology Officer at AngloGold Ashanti.
Could you provide a brief overview of your keynote address?
Investment in mining projects, like most business ventures, is highly susceptible to risk and uncertainty. Effectively identifying, assessing, and managing these risks is crucial for making strategic investments and optimising operational performance. Geological uncertainty is a primary factor contributing to unmet project expectations. To mitigate this, mining companies often require projects to follow standardised resource confidence profiles during various development phases. This typically involves assessing a distribution of measured, indicated, and inferred resources throughout the life of mine. However, these resource categories are generally defined based on drill spacing without adequately considering the associated uncertainty in resource estimates. This paper presents a case study on the application of stochastic mine planning in an open-pit gold project. Unlike traditional deterministic mine planning, which relies on fixed inputs and assumptions, stochastic mine planning uses multiple simulated models of the orebody to account for inherent geological uncertainty. The results highlight the advantages of using such approach over traditional resource classification in supporting investment decisions.
What key takeaways would you like attendees to gain from your talk, and how do you believe these insights can be applied to their work in resource estimation?
- The stochastic optimisation of conditional simulated models provides more realistic production forecast and effectively manages resource risks.
- Annual forecasts derived from stochastic scheduling can support resource classification.
- Stochastic optimisation is the ultimate reserve estimator. It considers resource confidence derived from equiprobable models of the deposit, which can incorporate sampling errors. In addition, it goes further than commonly used evaluation methodologies by using probabilistic parameters to optimise key resource to reserve modifying factors.
- Resource Development Projects are complex technical constructs and take a long time to become operations. While we cannot control commodity prices, taxes, exchange rates, etc., we can progressively narrow the gap in estimated and actual tonnages and grades, recoveries, costs and productivity. That is where the stochastic mine planning is most valuable.
How does your talk address current challenges in mineral resource estimation today?
The use of conditional simulations combined stochastic mine planning address one of the fundamental problems of dealing with resource estimation uncertainty in production planning, which is not addressed by conventional methods. This case study also provides an example of how conditional simulations can help identifying issues with resource estimates.
Why do you think events like the Mineral Resource Estimation Conference are important for advancing best practices and innovation in the industry?
Because resource estimation is hard. The knowledge exchange that takes place at events like this is invaluable, it supports professional development and provides insights for practitioners to improve their resource estimation workflows. MREC 2025 is the type of event that allows attendants to learn about the latest developments and innovations applied to real world resource estimation challenges.