The Problem:
In 2018, PanAust’s Ban Houayxai gold-silver operation in northern Laos faced a challenging issue: infrequent episodes of very poor gold recovery, often falling below 50%.
The operation was dealing with low head grades and high gold content in its tailings, with a significant portion of the gold locked in sulphide and silicate minerals. To enhance recovery and optimise processes for varying ore feeds, the mine needed a solution to predict these recovery challenges. Traditional geological studies had failed to identify the problematic areas, leading the mine to seek an innovative approach.
The Solution:
PanAust turned to PETRA and its MAXTAGeomet application, a groundbreaking orebody learning powered by AI for mine value-chain optimisation. MAXTAGeomet utilises operational data to predict plant performance, making it an ideal choice for the Ban Houayxai project. The digital twin model required geological data for training, which was sourced from two years of ore analysis between March 2016 and March 2018. This dataset included fresh, unweathered ore associated with poor gold recovery, as well as gold measurements from 1,211 composite tailings samples from the processing plant.
The Implementation:
The PETRA team integrated this data into the MAXTA software, creating a predictive model that could be applied to the mine’s block model for historical reconciliation analysis and future predictions.
Results:
Within just eight weeks, MAXTAGeomet successfully identified conditions leading to high levels of locked gold and poor recovery. These conditions included:
- High zinc grades.
- High acid-neutralising capacity (carbonate veins).
- Low copper and lead grades.
- High silver grades.
- Fresh rocks.
A graph comparing predicted versus actual tailings gold grades confirmed the model’s accuracy. MAXTAGeomet did not only identify problematic ore for gold recovery but also provided direct estimates of tailings gold grades and the probability of high tails grades for each block.
This approach quantified mining risk, supplying valuable data for cost improvement studies, analyses, and simulations for various scenarios. It also facilitated prioritisation of resource estimation for maximising value and reducing gold recovery risk over the life of mine.
Following the successful Ban Houayxai project, MAXTA found applications across iron-ore, copper-gold, and gold operations worldwide. Its capabilities expanded beyond geometallurgy, encompassing product quality, comminution energy consumption, and crusher, beneficiation, milling throughput maximisation. The deployments included:
- Drill and blast simulation and optimisation to maximise crusher throughput and availability within simultaneous product quality and load and haul constraints.
- Copper and gold recovery optimisations for real-time operational decision support.
- Drill and blast simulation and optimisation for maximising SAG mill throughput.
- Gold tailings grade prediction six hours in advance.
MAXTA orebody learning deployments demonstrated their ability to generate value quickly, thanks in part to PETRA’s investment in value chain integration via application programming interfaces (APIs). These APIs facilitated seamless interoperability with existing systems, enabling real-time MAXTA recovery optimisation and deployment into operations.
Integration with Maptek Vulcan 3D mine planning further streamlined the deployment process, allowing geometallurgy models to be operational within minutes.
PETRA CEO, Dr. Penny Stewart, emphasised the importance of leveraging existing systems and platforms for the successful adoption of machine learning and optimisation solutions, ensuring a smooth transition to more efficient mining operations.
For more information please contact us on or reach out via our contact page.
Or, visit our MAXTATM case studies page VISIT NOW
