Genetic models are rarely mentioned in the estimation of mineral resources and yet they are key to getting things right. The ability to conceive/perceive and apply the correct genetic model early in the exploration stage is crucial not only to the discovery of the deposit but also to the quality of the final deliverable product comprising Measured, Indicated and Inferred resources. It is important to always remember that tonnage and grade estimates of mineral deposits are only as reliable and accurate as the genetic model/geological base on which they stand.
Each genetic models can encompass a wide variety of deposits, e.g., the magmatic platinum group element deposits of Lac des Îsles (Canada), Bushveld Complex (South Africa), Stillwater Complex (USA) and the Great Dyke (Zimbabwe) share several characteristics in common. However, each deposit is unique and needs to be treated individually – to be understood, evaluated, classified and developed in line with its unique characteristics.
Depending on the genetic model, mineral deposits may be either homogeneous or heterogeneous. As a general rule, complex genetic models result in heterogeneous mineralization. The more heterogeneous a deposit, the more difficult it is to estimate and classify mineral resources, as many elements of the estimate become uncertain. This is illustrated in Figure 1.
General Trend of Increasing Difficulty in Estimating and Classifying Mineral Resources
Modified after Bujtor and McMahon, 1983
The regularity and predictability of a deposit depend upon, and are intimately related to, the geological system in which it evolved. The more we understand about how deposits form and what factors control their development and subsequent modification, the better chance we have of being able to evaluate and classify them and to plan their development. Often Qualified Persons (QPs) consider drill hole spacing and variography when classifying resources; that’s acceptable, but the underlying fact is that continuity is not a function of drill hole spacing or variography, but a function of geological phenomena. The ‘genetic/geological confidence limit’ must be high enough, not only to ensure an adequate basis for estimating and classifying tonnage and grade, but also to provide an essential framework for testing the validity and applicability of geostatistical studies.
Errors arising from an incorrect or inappropriate genetic/geological interpretation are orders of magnitude more serious than errors associated with grade estimation because:
- Changes in the genetic model can significantly alter an assessment of tonnage; and,
- Grade interpolation is correct only when governed by the genetic/geological characteristics, and not vice versa.
The reality is that, when the genetic model is correctly perceived and geological interpretation/domaining is correct, there may be very little overall grade difference as a consequence of using the various methods of interpolation available.
A quick survey of Chapter 12 in some Technical Reports prepared under NI 43-101 and filed on SEDAR, reveals that many QPs from independent consulting firms who perform site visits to mining projects in order to conduct data verification direct their efforts only towards QA/QC issues related to the quality of assays, survey and density measurements. They tend to forget that the genetic model/deposit type is part of the data that needs verification.
In conclusion, the importance of the role of genetic models in mineral resource estimation is enshrined in the CIM Definition Standards of May 10, 2014, as quoted below for the Measured resource category (see the underlined phrase in the quote):
Mineralization or other natural material of economic interest may be classified as a Measured Mineral Resource by the Qualified Person when the nature, quality, quantity and distribution of data are such that the tonnage and grade or quality of the mineralization can be estimated to within close limits and that variation from the estimate would not significantly affect potential economic viability of the deposit. This category requires a high level of confidence in, and understanding of, the geology and controls of the mineral deposit.
The underlined phrase provides explicit guidance as to the primary criteria to use when classifying a mineral resource into the Measured category. Geological continuity on its own is not enough because the directions of grade continuity will not always coincide with those of geological/lithological continuity, hence the need for a high level of confidence in the mineralization controls, i.e. the genetic model.
In recent years, Machine Learning/Artificial Intelligence (AI) has been developed and is being evolved to provide interpretations of data sets in search of clues to the understanding and interpretation of mineralized systems. However, genetic models developed in this manner are still far from yielding a high level of confidence required for the Measured resource category.