As the demand for a low-carbon and environmentally friendly future increases, so does the importance of mineral and metal commodities. The production of solar panels, wind turbines, energy storage systems and other green technologies require large quantities of minerals and rare earth metals. Natural Resources Canada noted that in 2019, Canada was a global leading producer in minerals required for green technology including graphite, nickel, cobalt, and others [1]. While mineral production continues to rise year over year, the ore grade, i.e., the concentration of a desired material, of multiple common minerals continues to decline.
To liberate valuable minerals from low ore grade deposits size reduction processes such as crushing and grinding are required; however, these processes account for over half of all energy consumption on the average mine. As mines are typically remote, fossil fuels are normally used as the main energy source, producing large amounts greenhouse gases, necessitating the need for more efficient size reduction processes. This could be accomplished by predicting how a particular orebody would break. With the surge in image sensing and computing technologies at mining sites many researchers are exploring ore texture and processability characteristics of the ore body. If distinct processability characteristics change based on ore textural feature from a 2D image, then general trends for optimal size reduction of orebodies of similar texture can developed.
This work builds on previous work by simulating ore breakage through the superimposition of a predetermined fragmentation pattern, called a mask, onto multiple ore textures. Synthetic, periodic black and white 2D ore textures were created to find a link between simple textural features such as different mineral grain shape, size, and orientation and processability characteristics. A Monte Carlo simulation was performed to generate a large quantity of realistic product particles using the Voronoi tessellations masking technique. To assess the processability of different textures, the percentage area distribution of valuable minerals of each ore texture was compared across the complete range of particle sizes. The valuable mineral percentage area distributions were analyzed for rate and shape of the distribution as particle size decreases, with noticeable differences between textures. The distributions were also parameterized using a two-beta mixture distribution model, expanding on the traditional one beta model developed by King [2,3,4]. These distributions can eventually help the mining industry make informed decisions on how much grinding and crushing will be required to liberate desired minerals from waste rock.
Identifer | oai:union.ndltd.org:uottawa.ca/oai:ruor.uottawa.ca:10393/42555 |
Date | 18 August 2021 |
Creators | Gottheil, Jeremy |
Contributors | Sowinski, Andrew |
Publisher | Université d'Ottawa / University of Ottawa |
Source Sets | Université d’Ottawa |
Language | English |
Detected Language | English |
Type | Thesis |
Format | application/pdf |
Page generated in 0.0012 seconds