Groupe d’études et de recherche en analyse des décisions


All-inclusive stochastic short-term production scheduling approach of an iron-ore deposit with future multi-element ore control data


Short-term mine production scheduling optimization is developed as a single formulation where mining considerations, production constraints, uncertainty in the orebody metal quantity and quality as well as fleet availability are evaluated together to define a well-informed sequence of mining that results in high performance during a mine's operation. A stochastic integer program is developed for the above. However, it is noted that partly- informed and ultimately costly decisions can be taken in the above development because of imperfect geological knowledge and information available during the actual ore control stage, affecting the performance of short-term schedules. To address this issue, orebody uncertainty models are updated by simulated future ore control data to account for local scale material type and grade variability affecting grade control classification of materials being extracted and sent to different ore/waste destinations. This updating leads to substantially improved short-term schedules accounting for better potential classifications of ore and waste materials. In general, the updating of multi-element orebody uncertainty models is based on the correlation of exploration data and past ore control data; the updated orebody uncertainty models are then used to optimize, while accounting for uncertainties, the short-term production scheduling, leading to better performance in terms of matching ore quality targets and delivering recoverable reserves anticipated. The above is demonstrated in a case study at an iron-ore deposit.

, 28 pages