Session
Rock Mass Characterization and Data Uncertainty I, Verbena Room
Abstract Summary
Porphyry geological systems are typically challenging to quantify through the geotechnical characterization process. The scale of many modern porphyry deposits, and associated rock mass heterogeneity requires a large characterization dataset from which the geotechnical domaining can be done. The presented case study demonstrates a rigorous workflow supported by Machine Learning (ML) to geotechnically characterize a copper porphyry deposit located in Central Asia. The operation comprises two adjoining open pits which will extend to approximately 500 m depth over a 2500 m by 2200 m footprint.
Automated core photograph logging using ML was used to supplement the available geological, structural, alteration and geotechnical data sources. The ML workflow geotechnically classified 44,500 m of core based on degree of fracturing, using existing exploration, infill and geotechnical drill core images. This provided a consistently classified geotechnical dataset that targeted all areas of the deposit at an approximate drill hole spacing of 150 m, which provides much better spatial resolution compared to most conventional geotechnical drilling datasets. The reliability of the ML dataset was verified against 22,000 m of manual photo-logging and available geotechnical logging. When compared against the lithological, structural and alteration models of the deposit, the ML dataset was used to define zones of distinct rock mass characteristics, and supplement the other more conventional lab testing and geotechnical logging datasets.
This approach to automated characterization of core images provides the potential to efficiently collect a large and reliable dataset using existing drilling data. This can provide geotechnical practitioners with a far denser and more consistent dataset to aid in domaining and design. Importantly, the results can support the reduction of geotechnical risk through the interim pit slopes and for ramp strategies. The example also highlights the importance of the geological team responsibility to collect high quality standardized photos for future use.
Authors
Presenter
Click here to view other presentations at this event.
