EDITORIAL & OPINION — COMMENTARY — Resource estimation and geological constraints, Part 1

The orebody supports the mine; a mine cannot create an orebody. But we see, and everybody reads about, mines that start up based on assumptions of grade, geometry and continuity, only to find that those assumptions were incorrect. .TWhat goes wrong? We believe the problem is due to a lack of geological understanding and insufficiently constrained resource estimation procedures.

Approaches to resource estimation generally fall into two end members. One traditional approach is the polygonal or sectional method wherein grade is averaged over a core length, which basically assumes that the deposit can be mined on core size pieces. Another common approach is founded upon geologically unconstrained geostatistical parameters, which often results in grade being interpolated into zones that contain no grade. Both end-member approaches are inappropriate. There is only one “right way” of doing a resource estimate. The purpose of this article is to outline the approach and methodology surrounding this “right way,” which is both statistically valid and firmly grounded in geological reality.

The quality of a mineral resource estimate depends on the integrity and accuracy of its geological foundation. To construct a realistic resource estimate, it is necessary to have an understanding of the fundamental controls on mineralization and the three-dimensional geometry of the mineralized body.

Tight project schedules often place a limit on how well the geology is understood, and how much of that understanding makes it into the resource estimate. A distressing consequence is that many resource estimates are calculated without applying a well-constrained geological model. Poor communication between the geologists working the property and the geostatistitians building the resource model can leave critical geological constraints overlooked. And with friendly and affordable resource-estimation software available, many companies opt for the ease of unconstrained geostatistical methods to generate resource estimates.

Most resource estimates begin with drill and assay data represented on regularly spaced drill sections and/or level plans. These data plots are then used as the basis for a manual interpretation, usually along with some interpretation of the geology. A common approach in interpreting drill sections is to draw a series of grade envelopes, based on assay data, with the boundaries defined by an arbitrary cutoff grade. These grade envelopes, which are essentially grade contours, are drawn to enclose as much mineralization as possible, and so often are extremely irregular and bear little resemblance to the real geology.

It is a serious error to mark out envelopes that are based on grade, not only because doing it ignores the geology, but because it excludes the subset of the assay population that falls below an arbitrary cutoff. When low-grade assays lie along the same geological structure as high-grade ones, it is neither good geology nor good statistics to exclude them just because they fall outside a predetermined assay contour. Low values in a population of assays are just as important as high values, and must be included in grade interpolation. Not to do so assigns a heavier weight to high assay values and ignores low assay values, predetermining a high-grade outcome for the exercise.

If, instead, “real” geological data (fault zone trajectories, for example, or alteration assemblages) are used to constrain the geometry of mineralized zones, the mineralized bodies are linked to geological features that provide a firm basis for their construction. Grades are then extrapolated into these bodies based both on statistical and geological parameters.

— The authors are senior geologists with the mining and exploration division of Vancouver-based SRK Consulting Canada. They specialize in assessing the controls on mineralization and resource modeling for feasibility studies.