A New Optimum Pit Limit Design Algorithm

Although many ultimate pit limit design algorithms have been developed for the past one- quarter century, only the well known Lerchs-Grossmann's (LG) algorithm is capable of producing a truly optimum ultimate pit limit today. This paper presents a graph theory oriented algorithm for optimum ultimate pit limit design. This algorithm produces a true optimal solution and maximizes the total undiscounted net profit for a given 3-D block mine model. The algorithm performs much better than the well known LG algorithm, in terms of the reduction in computation time and computer memory requirements. More importantly the new algorithm is easier to understand and implement than the LG algorithm. Specifically, this paper discusses: 1) the computer representation of arcs and the rules of arc generations which dictate the formation of trees in the graph, 2) the mechanism of joint support and re-allocation, 3) the properties, the general procedures of the algorithm, and the suggested specific steps to implement the algorithm, 4) comparison between the new algorithm with LG algorithm, and 5) some experience in implementation of the algorithm and two case study results.