Abstract
PRESENTERS:
Miguel Cuba, PhD, Senior Analyst, Mining Engineering, ArcelorMittal Mining Canada G.P.
Melanie Bolduc, P.Eng., General Manager, Technical Services and Long-term Planning, ArcelorMittal Mining Canada G.P.
ABSTRACT:
The unfolding algorithm is implemented in two steps: 1. numerical characterization of the folding process, and 2. unfolding of the geologic interpretation of the iron bands and drill-hole data to the initial state of the deposit. In the first step, the unfolding algorithm calculates scenarios of bounded three-dimensional deformation fields, using radial-basis-function estimators, conditioned to the specified guidelines. The unfolding algorithm warns the geologists if a numerical solution in parts of the ore bodies is unrealistic, based on a set of specified conditions. The numeric solution is chosen by consensus among the staff of the long-term geology department. In the second step, the chosen numeric solution is used to unfold the geologic interpretation of the deposit and the drill-hole data to the initial state of the deposit. In this state, the different economic attributes of the ore bodies are easily estimated by implementing traditional geostatistical techniques. The results are folded back again using the chosen numeric solution.
The custom-fit unfolding algorithm gave control to the long-term geology staff on the characterization of the geologic features of the deposit. The annual reconciliation of the long-term model with respect the short-term production model, in 2021, was under 2.7%. The unfolding algorithm played an instrumental role in achieving this performance.