model of the deposit is developed using geomodelling techniques, such as geostatistics, to facilitate the planning process by providing information on the shape, orientation, and size of the orebody. Subsequently, based on a set of economic and technical factors, such as commodity prices, feasible recoveries, and operational costs, the orebody is divided into sections, with a monetary value assigned to each of the blocks, resulting in an economic block model (Fathollahzadeh et al. 2021). Depending, for example, on the depth of the deposit, its size, and grades, the extraction method is selected. Open-pit mining is typically more suited to relatively shallow, large deposits, while underground extraction techniques are more applicable to deposits located at greater depths, with more complex orebody shapes. Given the fact that stochastic mine planning and scheduling methods are more commonly applied in open-pit operations, these types of techniques are detailed discussed in the subsequent sections. Mine planning process can be divided into distinct levels, ranging from strategic, long-term planning to short-term scheduling, which takes place after the initial prefeasibility study. In open-pit mines, the ultimate pit limit (UPL) definition stage (i.e., the spatial extent of the mine) is first defined. All of the extraction-related activities are encapsulated within that region. To simplify the mining process, the long-term plan is then divided into distinct stages, called pushbacks. Each pushback is then executed using the short-term schedules. A visual schematic of mine planning processes is presented in Figure 1. Figure 1 – Schematic representation of an open-pit mine planning process One of the fundamental methods widely adopted by the industry is the LerchsGrossman algorithm, developed in 1965, used to define the ultimate pit limit based on graph theory (Lerchs and Grossman 1965). The technique relies on the definition of the economic value of the constituent blocks comprising the orebody model, which are then classified as economically feasible to extract or to be discarded as waste. Given the reliance of the technique on geological knowledge of the mineralogy of the deposit, the underlying geostatistical model used to spatially predict the composition of the deposit from the borehole data is fundamental to the economic success of the mining operation. The main drawback of the traditional planning approach is its initial assumption of perfect knowledge of the deposit based on the conducted modelling stage. In reality, popular estimation techniques, such as ordinary kriging, produce overly smooth models that fail to capture the in-situ variability of the deposit (Smith and Dimitrakopoulos 1999). The lack of acknowledgement of uncertain variables can be detrimental in other areas as well. For example, developing a rigid cut-off grade (CoG) policy that fails to address fluctuations in
RkJQdWJsaXNoZXIy MTM0Mzk2