Diagnostic and Predictive Geometallurgy

Geometallurgy is an intersection of?geology and metallurgy aimed at optimizing mining operations. However, the term is often used so broadly that it sometimes seems to include the entirity of geology and metallurgy.? At the recent IMPC 2024 conference there were three amazing sessions dedicated to geometallurgy, but while watching these talks it occurred to me that "geometallurgy" lumps together two fundamentally different approaches: diagnostic geometallurgy and predictive geometallurgy. This can be confusing to those new to the field, making geometallurgy less approachable.

The Ambiguity of Metallurgy

Metallurgy is an ambiguous term that can cover both extractive metallurgy—the processes used to extract metals from ores—and secondary metallurgy, which involves refining metals and making alloys. When we use broad terms like "metallurgy" without specifying the context, we risk causing confusion.

At an extreme, if we were to combine structural geology and secondary metallurgy, we might end up with "structural steel" as a geometallurgical field!

Diagnostic Geometallurgy

Diagnostic geometallurgy focuses on understanding the mineralogical and microtextural characteristics of ore deposits. It involves detailed analyses similar to those performed by a mineralogist—examining the composition, structure, and relationships of minerals at a microscopic level. This approach helps identify the factors that influence mineralogy, which in turn affect the processing behavior of the ore. Essentially, it aims to understand the 'why' behind the ore's characteristics.

Techniques such as liberation analysis, mineral species identification, and surface chemistry studies are employed in diagnostic geometallurgy. These techniques are expensive per sample and are typically used only when there is an unexplained issue in the metallurgy. Many metallurgical processes, especially flotation, depend heavily on mineral textures that cannot be observed by geologists when logging core visually. While this information may be of limited use to geologists in terms of finding additional ore, it is invaluable to metallurgists for diagnosing and addressing issues in metallurgy, offering potential solutions or at least explaining suboptimal results. Similar to medical diagnostics, a problem should exist before heavily investing in this type of analysis.

Predictive Geometallurgy

Predictive geometallurgy, on the other hand, focuses on a macro-level understanding. It integrates geological and metallurgical data into block models used for mine planning, helping to forecast the performance of different sections of an ore body during extraction and processing. This approach aims to predict key outcomes, such as recovery rates, processing costs, and potential operational challenges. Essentially, it helps forecast 'what will happen' in large-scale mining operations.

For predictive models to work effectively, predictors must already be populated within the block model so that the model can be generalized across the entire deposit. These predictors should be reliable, interpretable, quantifiable, and spatially continuous. If such relationships cannot be established with the existing data, or if the variability is too high to be acceptable, then a diagnostic approach must be employed instead.

Integrating Diagnostic Insights into Predictive Models

Effective geometallurgy hinges on the integration of diagnostic findings into predictive frameworks. By identifying the root cause of metallurgical issues, we can often discover proxies—derived from geological logging or geochemical assays—that can be used to populate the block model with the ore characteristics determined during the diagnostic phase.

Example One

High iron sphalerite, known as marmatite, cannot achieve the same zinc concentrate grade as blende, which is pure ZnS. One could spend months in the metallurgical laboratory testing various flotation reagent regimes without successfully improving the zinc concentrate grade. Diagnostic laser ablation mineralogy might map a few hundred grains to identify the issue, but how do we expand this knowledge to a resource-wide model?

The next step could be to expand QemSCAN modal analysis categories. Instead of categorizing all sphalerite into a single bin, the software could divide it into two: sphalerite and marmatite. This allows us to interpret a few dozen samples and determine the proportion of each type. However, a few dozen samples are still insufficient to populate a block model. At this stage, geometallurgists can review the classified samples and search for proxies within the ore body knowledge that could be used to populate the block model. Reviewing the ratios of Zn, Fe, and S is an effective starting point. Overlaying these ratios on major geological events helps guide domain selection and assign concentrate grade expectations accordingly.

Example Two

A skarn-porphyry deposit contains abundant copper, primarily present as chalcopyrite, which floats cleanly. However, the skarn also contains sphalerite, which typically would not float under a selective copper flotation regime. In this project, about 30% of the zinc still floats, regardless of the use of zinc depressants. Diagnostic mineralogy revealed that approximately 30% of the zinc grains are affected by chalcopyrite disease. This means that regardless of the reagents used or the fineness of the grind, it is unlikely to reduce the 30% zinc recovery.

Generally, a 30% zinc recovery is manageable, and blending can keep concentrate sale terms acceptable on paper. However, blending tests with the supergene composite showed a failure, with over 90% zinc recovery—worse than the individual tests. Further diagnostic mineralogy revealed that the supergene ore contained chalcocite, which, when ground with skarn ore, released copper that activated the sphalerite, causing it to float during copper flotation. As a result, a base zinc recovery of 30% can be applied in the block model, but the mining plan must be adjusted to increase zinc recovery to 90% whenever supergene material is simultaneously fed to the mill along with skarn ore.

In these two examples, we see that detailed advanced mineralogy was used on a small number of samples as a diagnosis. Combined geological and metallurgical explanations—or geometallurgical explanations—were determined, and these diagnoses led to practical predictive geometallurgical algorithms that could be used in mine planning.

Tools

Cancha geometallurgy software is a powerful tool for both diagnostic and predictive geometallurgy, allowing for a deeper understanding of ore body characteristics and their implications on metallurgical performance predictions. For diagnostic geometallurgy, Cancha facilitates the analysis of mineralogical data, enabling users to identify factors that impact metallurgical performance. By incorporating data from liberation analysis, users can diagnose issues, such as mineral textures that impact recovery.

In predictive geometallurgy, Cancha integrates geological and metallurgical data into domain and regression models that inform mine planning. Users can link diagnostic findings to proxies—such as geochemical assays or mineralogical indicators—that can be used to populate block models with key ore characteristics.

Conclusion

Find the reason in diagnostic geometallurgy, and then generalize the solution in predictive geometallurgy.

Post Script on Process Mineralogy

While process mineralogy is indeed a vital component of diagnostic geometallurgy, equating the two overlooks the broader, collaborative nature of the discipline. Diagnostic geometallurgy is a multidisciplinary effort that extends beyond mineralogical analysis. It involves the collaboration of a diverse team—including geologists, mineralogists, metallurgists, and engineers—who work together to integrate mineralogical data with geological context, material properties, and processing performance. This collective approach allows for a comprehensive diagnosis of how an ore's inherent characteristics influence its behavior during processing. By combining insights from various experts, diagnostic geometallurgy provides a deeper understanding and more effective solutions than process mineralogy could achieve alone. So, while process mineralogy is a key player, it's the team collaboration that truly defines and enriches diagnostic geometallurgy.

#CanchaCanHelp