eCite Digital Repository

Machine-assisted ore deposit mass balance evaluations in 3D


Hood, SB and Cracknell, MJ and Gazley, M and Reading, A, Machine-assisted ore deposit mass balance evaluations in 3D, Resources for Future Generations (RFG2018), 16-21 June 2018, Vancouver, BC (2018) [Non Refereed Conference Paper]


Many options exist for the representation of geochemical anomalies around ore deposits, and the majority of these involve different approaches to calculating, and representing, the upper and lower portions of element frequency distributions from compositional data. So-called positive and negative anomalies are taken to represent chemical deviation from local background values and are typically interpreted at metalliferous hydrothermal ore deposits as representing mass that has been added or depleted during metasomatism linked to metal deposition.

This contribution considers the representation of geochemical anomalism in 3D space by calculating the mass added or depleted during alteration of host rocks using Grantís Isocon (1985) formulas. While conventional mass balance work is traditionally done on a sample vs. sample basis, our methodology uses groups of protolith rocks and groups of altered rocks with comparable geological histories, leveraging an adapted approach developed by Ague and van Haren (1995) which includes a measure of confidence for results. Identifying sample pairs for altered and unaltered equivalents is challenging because the correlations between multivariate data may not be intuitive, and large amounts of data are laborious to interpret manually. Machine learning offers potential solutions to problems involving large numbers of data, many variables or features, and complicated relationships between variables. Furthermore, traditional mass balance approaches do not involve representing uncertainty related to the heterogeneity of lithological units.

Our methodology involves linking protolith rocks to their equivalent altered counterparts from large exploration and mining drill hole databases using machine learning algorithms. Interpreting the results gives insight into original protolith geometry, and adds confidence measures to derived data. By evaluating the geometry of our derived data around mineralisation zones, we seek to better understand the chemical processes at work during ore genesis.

Item Details

Item Type:Non Refereed Conference Paper
Keywords:geology, ore deposits, lithology, Minto, alteration
Research Division:Earth Sciences
Research Group:Geology
Research Field:Geology not elsewhere classified
Objective Division:Mineral Resources (Excl. Energy Resources)
Objective Group:Mineral exploration
Objective Field:Copper ore exploration
UTAS Author:Hood, SB (Mr Shawn Hood)
UTAS Author:Cracknell, MJ (Dr Matthew Cracknell)
UTAS Author:Reading, A (Professor Anya Reading)
ID Code:129240
Year Published:2018
Deposited By:Earth Sciences
Deposited On:2018-11-18
Last Modified:2018-11-23

Repository Staff Only: item control page