BIF-hosted deposit unit differentiation using multivariate Gaussian processes on measure while drilling data

ABSTRACT

Measure while drilling (MWD) data collected from production holes can provide information on the location of stratigraphic units in banded iron formation-hosted iron ore deposits. Stratigraphic modelling in these deposits is typically based on data from exploration holes, and adding more densely spaced production data can potentially increase model detail at the bench scale. Previous MWD classification methods struggle to differentiate between neighbouring ore units. In this paper, multivariate Gaussian Processes (GPs) were applied to locate the contact between two iron ore units in the Dales Gorge Member in the Brockman Iron Ore Formation. Production MWD points were then labelled based on the GP output. By altering parameters of the labelling process, 24.4–49.4% of the test data were labelled, with accuracies from 81.4 to 86.8%. Classifications from the same hole were compared to ensure MWD label consistency. The results demonstrate that the proposed method can improve geological unit classification from MWD data.

KEYWORDS:

• Hamersley region
• machine learning
• modelling
• iron ore
• stratigraphy

View correction statement:

Correction

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work has been supported by the Australian Centre for Field Robotics and the Rio Tinto Centre for Mine Automation.