Posted in | News | Copper

Hyperspectral Imaging to Revolutionize Heap Leach Pad Monitoring, Enhancing Mining Efficiency

In a recent article published in Remote Sensing, researchers from the United States of America employed hyperspectral imaging (HSI) to map heap leach pads, detecting features such as lixiviant ponds, clay minerals, precipitation, and irrigation pipes. This technology is based on light interaction with objects, where photons are absorbed, transmitted, emitted, or reflected, influenced by the object's grain size, atomic composition, and chemical bonds.

copper, heap leaching

Image Credit: Jose Luis Stephens/Shutterstock.com

Background

Leaching is a metallurgical process in which a specific solvent permeates a pile of stacked ore to dissolve valuable elements. This method is favored for its cost-effectiveness and straightforward technology, particularly in copper mining, accounting for 21% of copper extraction globally.

Mining companies traditionally monitor and model leach pads over their lifespan using methods to gather information such as clay content, P80, raffinate distribution, agglomeration effectiveness, and locations of plugged pipes. While these traditional methods are simple and inexpensive, they often yield discontinuous and inefficient data, underscoring the need for more advanced approaches.

The Current Study

The spectral library used in this study comprised two primary components. The first component consisted of spectral signatures from kaolinite, muscovite, biotite, chlorite, epidote, hornblende, enstatite, water with swelling clay, and montmorillonite. The second component included ground truth samples collected during fieldwork, representing irrigation pipes and precipitates found on the leach pad.

To ensure consistency within the reference spectra, the irrigation pipes and precipitates were scanned using an ASD TerraSpec analyzer at the University of Arizona, Tucson. X-Ray Diffraction (XRD) analysis identified the precipitates as hexahydrite (MgSO4·6H2O) and pickeringite (MgAl2(SO4)4·22H2O).

Hyperspectral imagery was captured covering a wavelength range from 400 to 2500 nm, comprising 270 bands for VNIR and 170 bands for SWIR. Principal Component Analysis (PCA) was employed to reduce data dimensionality while preserving spectral information, helping to identify the minimum number of endmembers needed. The HySime method was also used to estimate virtual dimensionality by selecting eigenvectors that best represent the signal subspace, optimizing for mean square error between signal projection error and noise projection.

The N-FINDR algorithm was used to identify endmembers within the hyperspectral images. N-FINDR iteratively evaluates different spectra to maximize the volume of the endmember matrix. The algorithm selects the spectra that produce the highest volume, representing the endmembers.

The abundance maps were generated using fully constrained least squares (FCLS) linear unmixing. The linear regression model represents a pixel as a combination of endmembers with associated abundances, adjusted by a noise term. The FCLS approach adds a Lagrange multiplier to enforce the abundance sum constraint, ensuring that the calculated abundances accurately represent the material distribution in the study area.

Results and Discussion

The analysis indicated that at least five endmembers are necessary due to significant variance captured by the first few components. HySime initially suggested eight endmembers before smoothing and only one afterward, the latter being unrealistic. Therefore, 10 endmembers were selected for N-FINDR analysis, considering the diverse SWIR-active mineral groups and other features on the leach pad.

N-FINDR identified 10 endmembers with generally low reflectance due to wet conditions. EM1, likely water, showed high abundance, suggesting a mixture of water and background. EM2 was identified as irrigation pipes with montmorillonite. EM3 and EM4 represented soaked mineral mixtures, EM5 resembled EM2 but with stronger plastic features, and EM6 was a mixture of chlorite and sulfate precipitate. EM7 combined pipes and kaolinite dust, while EM8, like EM2 and EM5, was affected by noise. EM9 and EM10 were wet precipitates with muscovite and clinochlore, respectively.

PCA provided significant insights: the first component highlighted mineral-related arc patterns from ore loading, and the third showed irrigation pipes. Virtual dimensionality methods performed poorly, with discrepancies between pre- and post-smoothing results. Smoothing reduced variance and suggested fewer endmembers than expected due to moisture's homogenizing effect on the ores.

Some abundance maps were useful despite data quality issues like noise and water content affecting reflectance. EM1's moisture distribution map could help adjust irrigation rates and detect low permeability areas. EM2, EM5, and EM8 highlighted irrigation pipes, aiding in maintenance.

Using ground truth libraries (red pathway) for unmixing provided more reliable results than endmember extraction. This method better captured horizontal pipelines, mineral distributions, and montmorillonite near ponds. It reflected an abundance of materials consistent with mining processes, although differences from surveys were noted due to sulfuric acid effects and spectral changes from ore processing.

Conclusion

This study revealed hyperspectral imaging can directly measure clay content and moisture distribution. Yet, N-FINDR yielded unrealistic water amounts and misclassified minerals. Conversely, using a reference library provided more realistic results but faced challenges in water content estimation and mineral misclassification.

Both techniques offer continuous map data for monitoring, but FCLS with a reference library proved more suitable than N-FINDR for accurate analysis in noisy datasets. Future research should refine methodologies to optimize mining operations further.

Source:

He, J., Riley, D.N., et al. (2024). Is Endmember Extraction a Critical Step in the Analysis of Hyperspectral Images in Mining Environments? Remote Sensing, 16, 2137. https://doi.org/10.3390/rs16122137

Dr. Noopur Jain

Written by

Dr. Noopur Jain

Dr. Noopur Jain is an accomplished Scientific Writer based in the city of New Delhi, India. With a Ph.D. in Materials Science, she brings a depth of knowledge and experience in electron microscopy, catalysis, and soft materials. Her scientific publishing record is a testament to her dedication and expertise in the field. Additionally, she has hands-on experience in the field of chemical formulations, microscopy technique development and statistical analysis.    

Citations

Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Jain, Noopur. (2024, June 19). Hyperspectral Imaging to Revolutionize Heap Leach Pad Monitoring, Enhancing Mining Efficiency. AZoMining. Retrieved on July 25, 2024 from https://www.azomining.com/News.aspx?newsID=17979.

  • MLA

    Jain, Noopur. "Hyperspectral Imaging to Revolutionize Heap Leach Pad Monitoring, Enhancing Mining Efficiency". AZoMining. 25 July 2024. <https://www.azomining.com/News.aspx?newsID=17979>.

  • Chicago

    Jain, Noopur. "Hyperspectral Imaging to Revolutionize Heap Leach Pad Monitoring, Enhancing Mining Efficiency". AZoMining. https://www.azomining.com/News.aspx?newsID=17979. (accessed July 25, 2024).

  • Harvard

    Jain, Noopur. 2024. Hyperspectral Imaging to Revolutionize Heap Leach Pad Monitoring, Enhancing Mining Efficiency. AZoMining, viewed 25 July 2024, https://www.azomining.com/News.aspx?newsID=17979.

Tell Us What You Think

Do you have a review, update or anything you would like to add to this news story?

Leave your feedback
Your comment type
Submit

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.