Abstract

The present study summarizes reports on optimal flotation circuits design over the period from 1989 to present. A review is made on the resolution characteristics of each study, leading to a classification of the approaches into four different groups, in which conclusions are based on either (A) mathematical solutions without binary variable, (B) mathematical solutions with binary variable, (C) heuristic solutions, or (D) genetic algorithms. These groups aim at a common scheme of resolution, varying in both flotation and milling models employed, characteristics of their superstructures, and functions which each study proposes to optimize, as well as the effects of the application of each strategy on the form of the solution of the circuit (selection of stages, selection of equipment, and dimensions of equipment). Within these either implicit or explicit multi-objective optimization problems there are challenges strongly related to the obtaining of better flotation and milling/classification models, as well as the treatment of uncertainty related to important parameters (kinetics, composition, external economic factors) and changes in previously established configurations using methods in which the designer's criteria influence the application of models and restrictions on the problem. Also, the incorporation of environmental issues needs more attention. © 2008 Elsevier B.V. All rights reserved.