The tools and methods used for mineral resource modelling in the mining industry have been tried and tested over many years, with many projects and in a multitude of commodities around the world.
Among the tools in use today, we can date the kriging methods back to the
Witwatersrand deposits in South Africa and the early days of Danie Krige, Herbert
Sichel and Georges Matheron in the 1950s and 1960s. Even more impressive, inverse distance methods can be dated as far back as the 14th century!
More recently, the discovery and exploitation of world-class deposits gives rise to the search for the next big find. Shallow, easy to access deposits appear to have all been discovered. The remaining ones present significant challenges: they may be marginal deposits, occur at great depths and/or present other access constraints. Some would say that most challenges can be solved if you throw enough money at the problem, but how much is too much? What is the risk of getting it wrong? If wrong, how far off the mark are we?
These questions are not new to the resources sector. We have always known
that the mineral resource models that are built are a snapshot in time, reflecting
our best assessment of the deposit. We have also known that there is uncertainty in this model, uncertainty in the geology interpretation, uncertainty in the sample data, uncertainty in the predicted grades, and ultimately, uncertainty in the ‘optimal’ pit or stopes that may form the basis for a mine design and schedule.
The idea of quantifying this uncertainty and using it to manage risk is also not
new. Conditional simulation was posed as a potential solution to this problem in the early 1990s. Over the last three decades, we have seen the rise of geological simulations, grade simulations, and the merging of these sources of uncertainty.
In the last 15 years, the focus subtly shifted to the use of these simulated models to determine the optimal pit or for underground stope optimisation. Some have even gone so far as to build schedules and cash flow models based on these uncertainty models, to assess the uncertainty in the cash flow over the life of a mine. This is the present state of innovation in mineral resource modelling.
Today’s mineral resource models often rely on technology that can be traced
back 70 years, or even 700 years. Tomorrow’s models are going to be based on technologies that took seed over 30 years ago. The future of mineral resource estimation requires us to acknowledge that uncertainty exists. It demands for us address it, quantify it, and to use it to make more responsible risk management decisions.
