Relevant Degree Programs
Affiliations to Research Centres, Institutes & Clusters
Graduate Student Supervision
Master's Student Supervision (2010 - 2020)
Mineral exploration is the necessary first step of any mining project. Mineral prospectivity analysis is a cost and time efficient exercise with for goals to delineate area of high prospectivity or to rank targets. The worldwide tendency in mineral exploration efforts is to focus on brownfield areas where large amount of data is available. Hence, data-driven methods for mineral prospectivity modeling (MPM) is preferred. In this research, two aspects of data-driven MPM are explored to examine the influence of them on the mineral prospectivity map using two different machine learning algorithm (random forest (RF) and support vector machine (SVM)). This research aims at demonstrating that RF algorithm is the best method for MPM in a variety of case scenarios involving different training datasets and input features.This study primary target is epithermal Au deposit in the area of the Iskut property owned by Seabridge Gold Inc. Different training dataset were created using same 18 deposit locations but different set of non-deposit locations: selected non-prospective known mineral occurrences (KMO) for Au deposits and 10 sets of random point pattern. Predictor maps were generated from publicly available and privately-owned geospatial data based on a conceptual exploration model using a mineral system approach and were separated into two input datasets: one exclusively included public data while the other included data from the private and public domain. Data-driven mineral potential models using RF and SVM (using three different kernel function) were compared based on sensitivity to parameter configuration, accuracy and performance. It was found that the accuracy of a model increases when the number of predictor maps increases. This research also showed that non-prospective KMO based uniquely on distance and commodity can introduce a bias. On one hand, almost all the SVM models are overfitting, most likely because of the scarce training dataset. Moreover, they are sensitive to outliers in the training data and require long computational time. On the other hand, the RF is easy to parameter, transparent, less sensitive to outliers and performs well. Hence, RF is the method to opt for in data scarce region over SVM.
No abstract available.
Managing the payload is key to running safe, efficient and profitable mining operations. Despite the relative simplicity of shovel-truck operations, they are currently not achieving optimum productivity. By overloading trucks, shovel operators can significantly affect the operational efficiency of a mobile fleet and, consequently, reduce the overall profitability of a mining project. Uncertainty around the real volume of the mineral payload hauled from the mine to the processing plant impacts the planning for a smart decision-making process, which increases the risk of operating the mine below a desirable standard of efficiency and profitability. Also, inefficient payload management increases the consumption of fuel and tire, which impacts the carbon footprint by raising demand for elements harmful to the environment. Modern digital and data technologies offer the potential to factor out these drawbacks and to provide digital solutions to optimize shovel-truck performance. This Thesis proposes two approaches for estimating the volume and distribution of the truck payload. The first approach is a batch-performance machine vision system for imaging-based analysis of the payload. The system is developed and tested on a 1/14 model of a mining truck, and results are visualized in an immersive augmented reality environment. The second approach is based on the utilization of a proposed model of machine earning where data from sensors embedded in different parts of the shovel are collected and streamlined to an in-house private cloud for virtualization and processing. A TensorFlow-based ML platform was then used to find correlations between the truck payload and its components for an accurate visualization and volume computation under harsh operating conditions. Finally pros, cons and discussions around its applicability in a real operation were analyzed for both approaches. Analysis of profitability, cost and requirement of computational resources complements the encouraging results attained from both approaches regarding volume computation and visualization of the mining truck payloads.
Corrosion of ground support poses a significant safety risk to underground mines that is difficult to identify and costly to address. The corrosion of embedded ground support elements is undetectable by visual inspection. This leads to unexpected failure typically mitigated by comprehensive rehabilitation that covers any areas suspected of corrosion. Rock bolts are replaced when the visible portion of a rock bolt is judged to be excessively corroded or loss of holding capacity is confirmed by the pull-out test.This thesis investigates the influenced of the mineral electrochemical properties in a heterogeneous rock mass on ground support corrosion. Previous work focused on the uniform corrosion of ground support in wet and dry conditions by empirical experiment in a real or simulated underground mining environment and on the corrosivity classification of environment by various qualitative measures. This thesis approaches the electrochemical influence of rock mass to corrosion of rock bolt by characterization of mineral electrochemical properties, numerical simulation of the galvanic corrosion process, and validation of numerical model by submersion experiment. A procedure for determining the electrochemical properties of minerals was developed with a complementary numerical model for galvanic corrosion of ground support by minerals. Additionally, the simulated numerical corrosion model was validated with a submersion experiment.The numerical model was found to have over-estimated the galvanic corrosion of steel by cathodic mineral, but predicted cathodic protection of steel by anodic mineral. The numerical model did not account for the acid byproducts of oxidized sulphide minerals that resulted in pitting corrosion. These findings allow us to predict galvanic corrosion of steel by minerals and propose a new mechanism for initation of pitting corrosion to by pyrite oxidation.