SOLVING THE MINE SEQUENCING PROBLEM INCLUDING GEOMETALLURGICAL VARIABLES USING A SIMULATED ANNEALING METAHEURISTIC *J. L. V. Mariz1, L. E. Alvarez Paredes2, D. B. Mazzinghy3, R. L. Peroni4 1Mining Engineering Department, Universidade Federal do Rio Grande do Sul, Brazil (*Presenting author: jorge_valenca@hotmail.com) 2 Mining Engineering Department, Universidad Santo Tomás, Chile 3 Mining Engineering Department, Universidade Federal de Minas Gerais, Brazil 4 Mining Engineering Department, Universidade Federal do Rio Grande do Sul, Brazil ABSTRACT In the context of mine planning, the primary problem to be solved is the mine sequencing problem, which determines whether and when a mining block will be extracted, as well as its destination. Despite its importance, addressing this problem using an exact approach on a complete mineral deposit as a monolithic mathematical model is still impractical, as the number of constraints grows exponentially as more blocks are considered in the optimisation. However, although the mine sequencing problem aims to maximise the economic return of the operation, geometallurgical variables such as recovery and specific energy have historically been neglected, impacting the project net present value. This study proposes the incorporation of geometallurgical variables into the mine sequencing problem modelled as a direct block sequencing (DBS) approach, solved using a simulated annealing metaheuristic. Numerical experiments were conducted on the Marvin dataset considering four scenarios: i) without geometallurgical variables; ii) including variable recovery; iii) including variable specific energy; iv) including both variable recovery and specific energy. The results showed the importance of addressing geometallurgical variables within the mine planning scope, resulting in more robust plans and avoiding unforeseen losses. KEYWORDS Open-Pit Mining, Mine Sequencing, Direct Block Scheduling, Geometallurgy, Metaheuristic, Simulated Annealing. 1. INTRODUCTION The established method for representing a mineral deposit is to discretise it into a block model whose dimensions are usually associated with geological exploration data. Based on the interpretation of drill core samples, these blocks are then associated with information from geology (lithologies) and chemical analyses (grades of the main substances) through estimation techniques, the most established of which is ordinary kriging (Rossi and Deutsch, 2014). The decision encompassing which of these blocks to
RkJQdWJsaXNoZXIy MTM0Mzk2