Mineral Concentration Plants Design Using Rigorous Models

摘要

The design of a mineral concentration plant plays an important role, since the operation of the plant depends on the structure of the process (Mendez et al., 2009). The concentration plant is formed by several concentration stages interconnected within a particular arrangement. Currently, the selection of the flotation circuit is based on the experience of the designer, and is complemented by laboratory tests and simulations. In the literature there are many studies that design concentration plants using mathematical programming. However, these strategies use a non-convex MINLP model that is difficult to solve. For that reason, usually simple models and systems with fewer species are used. This paper presents a methodology for the design of a mineral concentration circuit based in two steps: 1) Identify the set of optimal structures using discrete values of stage recoveries and solving several mixed integer linear programming (MILP) problems; 2) Determine the optimal design for each of the structures obtained in the previous step, using a Mixed Integer Nonlinear Programming model (MINLP) and a rigorous model for the recovery at each concentration stage. This work is based on the assumption that there are little optimal structures within a given space of species recoveries at each stage. A case study is used to validate the proposal using a system with several species.