Coupled Geologic Simulations using MOOSE: Understanding Ore Deposits and Aiding Mineral Exploration

This project used the open-source code MOOSE to simulate geological processes involved in the formation of mineral deposits, with the aim of improving understanding of mineral systems and providing information that can be used to aid mineral exploration.
Person

Principal investigator

Heather Sheldon heather.sheldon@csiro.au
Magnifying glass

Area of science

Geology, Mineral Exploration
CPU

Systems used

Magnus and Visualisation
Computer

Applications used

MOOSE, Paraview
Partner Institution: CSIRO| Project Code: pawsey0127

The Challenge

Mineral exploration is an expensive, high-risk activity, and is becoming more so in Australia as most un-discovered mineral deposits are likely to be located “under cover”; that is, under layers of weathered rock and sediment that make it difficult to detect the deposit at the Earth’s surface. Conventional mineral exploration techniques, such as geophysical surveys and drilling, are expensive. Thus, exploration companies need tools and information to enable them to identify the best areas in which to focus their efforts. Typically this involves geologists developing a hypothesis to explain the occurrence of known mineral deposits in a given area, then applying that hypothesis to a new area to predict where mineralisation is likely to occur. But what if there are several competing hypotheses, or a proposed hypothesis turns out to be wrong?

The Solution

This project provides an approach for testing geological hypotheses at low cost. Our approach involves simulating geological processes involved in mineralisation (e.g. faulting, fluid flow and heat transport) over geological time. This is achieved by expressing the processes as a series of coupled equations, which are solved on a computer using the Finite Element Method. We use the open-source code MOOSE to solve the equations, and Paraview to visualise the results

The Outcome

Pawsey’s Magnus supercomputer enables us to run large 3D simulations very rapidly, which in turn enables us to explore a large parameter space – that is, we can vary parameters in the models in order to identify the optimal conditions required for mineralisation.
We also use Zeus to visualise the simulation results remotely, avoiding the need to download results prior to visualisation.

List of Publications

Sheldon, Heather A, and Peter M Schaubs. 2018. “Investigating Controls on Mineralisation in the Batten Fault Zone Using Numerical Models.” In Annual Geoscience Exploration Seminar (AGES) Proceedings, 67–71. Northern Territory Geological Survey.

Sheldon, HA, and PM Schaubs. 2018. “Fluid Flow Drivers for Sediment-Hosted Pb-Zn-Ag Mineralisation at McArthur River, Northern Territory, Australia.” In Australian Geoscience Council Convention.

Figure 1. Geological model representing part of the McArthur Basin in the Northern Territory. Top: Model geometry. Colours represent different rock types. Star indicates location of known mineralisation relative to the Emu Fault, a major structure in the area. Bottom: Temperature contours in 2 geological units, illustrating effect of varying parameter values on thermal convection in this model.

Figure 2. Effect of deformation on convective fluid flow in the Emu Fault. Extending the model in a north-south direction results in downward flow that overrides convection. This happens almost instantaneously relative to geological timescales. Magnus enabled us to run simulations involving coupled fluid flow, deformation and heat transport within a few hours, contrasting with several days to run such simulations on a desktop machine.
Figure 3. 80 C temperature isosurface coloured by fluid flow rate, demonstrating the 3D complexity of this system. Previous studies on desktop computers were limited to 2D, or were only able to investigate very few scenarios due to the time required for 3D simulations.