The design of a mine should aim to optimise the market value by maximising revenues whilst minimising costs. In open pit mining, the nature of the extraction technique translates this objective into the maximum recovery of reserves while minimising waste extraction subject to technical, economical and safety constraints. Since the introduction by Lerchs and Grossman, in 1965, of a model that is now considered a classic in the area, open pit design has reached a level of maturity and acceptance by industry. These models are being used on a regular basis by designers and planners. This paper proposes a new alternative model for open pit design. It is based on genetic algorithms (GA) and exploits a reduction of the number of decision variables of the more traditional techniques. The novel approach provides more flexibility than traditional models; it can also be extended to consider planning problems that traditional open pit design models are not capable of handling. The algorithm has been designed for two dimensional problems and tested in a simple instance. The results obtained in this simple test are encouraging and are a good supporting argument for the development of a GA model capable of handling three-dimensional (3D) instances. The model presented in this paper is just a first step towards a full GA to solve the pit design problem and it is believed that novel tools like the one presented herein could become great allies in tackling challenging problems in open pit design, an illustration of such a challenging problem is presented and discussed._x000D_
FORMAL CITATION:Saavedra-Rosas, J, 2010. A new genetic algorithm for open pit design - the two-dimensional case, in Proceedings Mine Planning and Equipment Selection (MPES) 2010, pp 95-104 (The Australasian Institute of Mining and Metallurgy: Melbourne).