Resource estimation tools

Uniform Conditioning is an increasingly popular technique in the Maptek Vulcan resource modelling toolbox.

In resource modelling, when data is sparse and widely spaced with respect to the selective mining unit (SMU), there is a constant tension between creating a resource model that produces accurate, well informed local estimates and a model that is good for predicting global resources and producing reliable grade-tonnage values.

Commonly the big picture is sacrificed for improved local estimates.

Uniform Conditioning is an estimation technique that has gained popularity in recent years.

Uniform Conditioning was originally developed to address the inability of traditional estimation methods, such as kriging, to produce reliable grade profiles that describe the global distribution of a deposit. However, in order to produce valid results large panels were required, and there was no indication of where the actual SMU grades were within those panels.

Recent developments within Maptek Vulcan grade estimation have addressed some of the earlier shortcomings of Uniform Conditioning. Alternatives to localise and rank the SMUs in a panel now make the grade assignment process straightforward and flexible.

Uniform Conditioning models allow users to validate the amount of recoverable reserves, providing an invaluable tool either at the prospecting stage or during production. When using a cutoff grade to evaluate resources and if the data is sparse, this method will mitigate the risk of overestimating or underestimating the grade and amount of ore.

Estimation method

Uniform Conditioning is available as another estimation method in Vulcan, alongside inverse distance, and simple, ordinary and indicator kriging. The process of setting up a uniform conditioning estimation is virtually the same as setting up a kriging plan. The user simply specifies the panel definitions and a localisation strategy.

Vulcan allows both regular and irregular panel definitions. For example, a panel could be defined using a set of 3D solids or by a block model variable. Simple regular panels are also allowed.

This provides great flexibility and helps in accommodating the panel shapes to the orebody and domain shape that is under study.

The corrected panel scale distributions can be assigned to the SMU scale blocks using one of four different localisation strategies. The user can then choose and adapt a method that best fits a particular mineralisation.

Uniform Conditioning in Vulcan has all the advantages and power of Vulcan grade estimation, including multi-threaded execution, batch mode and options to output additional variables that are part of the estimation process.

The latest Uniform Conditioning is available to GeoStatModeller users with the Vulcan 10.0.2 updates.


Grade estimation using ordinary kriging compared with models using different localisation strategies

Odinary kriging

Rank by SMU estimate

Random

Minimised entropy