On the joint multi point simulation of discrete and continuous geometallurgical parameters


On the joint multi point simulation of discrete and continuous geometallurgical parameters

van den Boogaart, K. G.; Tolosana-Delgado, R.; Lehmann, M.; Mueller, U.

Geometallurgical parameters are descriptions of the mineralogy and microstructure of the ore determining its mineralogical and microstructural characteristics. From a conditional geostatistical simulation of such properties, a processing model could compute recovery, equipment usage, processing costs, and thus the monetary value for mining and processing the block with certain processing parameters. This can be used for optimizing mining sequences or finding optimal processing parameters by solving the corresponding stochastic optimization problem.
The approach requires two properties of the simulation not provided by established geostatistical techniques:
1) Many relevant geometallurgical parameters are from non-Euclidean statistical scales like (mineral) compositions, (grain size) distribution, (grain) geometry, (stratigraphic type) categorical, etc., which might produce impossible values when simulated with standard geostatistical techniques.
2) Due to the nonlinearity of processing, the whole conditional distribution of the geometallurgical parameters is relevant and not only its mean and variance. The geostatistical simulation needs to reproduce the joint conditional distributions of all the geometallurgical parameters.
We have developed a multi-point conditional geostatistical simulation technique, which allows for jointly simulating dependent spatial variables from various sample spaces. The technique combines an MPS-type infill simulation with a new form of distributional regression to estimate conditional distributions of arbitrary scales from different information sources, including training images, training models and observed data. The distributional regression is based on a generalization of logistic regression and has some relation to both BME-type geostatistics and high order cumulants.
The method ensures simulated data to reside within the constrained sample space and honour the characteristics of the joint distribution to be reproduced. The computational effort is substantial, but affordable for a useful application with standard problems: from processing-aware block value prediction and block processing optimization as we show in the test application to a completely defined simulated model situation with a complex processing model.

Keywords: geostatistical simulation; non Euclidean scales; geometallurgy

  • Lecture (Conference)
    SMP 2014, Ore Body modelling and strategic mine planning, 24.-25.11.2014, Perth, Australia
  • Contribution to proceedings
    SMP 2014, Ore Body modelling and strategic mine planning, 24.-25.11.2014, Perth, Australia
    Orebody Modelling and Strategic Mine Planning, SMP 2014, Integrated mineral investment and supply chain optimisation, Charlton Victoria, Australia: Australasian Institute of Mining and Metallurgy, 987-1-925100-19-8, 379-388

Permalink: https://www.hzdr.de/publications/Publ-20226
Publ.-Id: 20226