Wireframe-Free Geological Modelling - An Oxymoron or a Value Proposition?

Efficiency demands that mine geologists devote the most time to tasks where they can have the biggest effect on production. One of the most time-consuming tasks is geological interpretation of the orebody and turning this into a block model for the planning engineers. The current best practice for mines that require a geologically constrained resource model involves capturing the interpretation of geologists as three-dimensional (3D) wireframes. However, there is significant overhead in the production of valid wireframes from sectional and plan interpretations. This overhead consists of the mundane tasks of digitising polylines, then building 3D wireframes and validating them. While software can simplify this task, the current process is not the best use of a geologist’s time. Under time pressure exerted by planning engineers and the time required for modelling from polylines, geologists are forced to roll-up the presence of other elements or impurities into the interpretation and encode them in the polylines, even though it is often desirable to consider these other factors separately. While it might seem unlikely that mine geologists can produce constrained geological models without the intermediary of wireframes, the technology to accomplish this is getting closer. In this paper, the results of a trial at the Savage River magnetite mine, which is owned and operated by Grange Resources Ltd, are reported. The trial was for a ‘direct-to-block’ implicit geological modelling method. The focus was on creating a geological attributed block model from drill hole geological log data; resource estimates were not determined. The categories modelled were air, waste, low-grade ore, medium-grade ore, and high-grade ore. The block model outputs either resulted in a proportional block model (categories expressed as percentages within equivolumetric blocks) or a sub-blocked model (parent block and one sub-block), which is the preferred block model output at the Savage River mine. When compared to the traditional best practice modelling method, which takes 43 work days, the wireframe-free modelling method took only 26 hours from initial modelling, using both exploration and grade control data. An update using appended drill hole data is achievable within one day. These results represent time savings of more than one order of magnitude compared to the traditional workflow, and, as such, these figures merit serious consideration for the use of the direct-to-block modelling approach at other operations. With these time savings, geological and estimation models can be kept updated constantly; geologists will not need to set aside two months for quarterly or half-yearly modelling of the drill hole data for resource review. Such dramatic time savings will allow for the better use of highly skilled geologists, who are in short supply in the current market, and will also impact directly on downstream mining engineering workflows.