monitoring. From the environmental perspective, the generation of waste leads to increased land use and potential damage to the watersheds due to the leaching of hazardous compounds. The same factors pose danger to the local populations, which are affected if regulatory guidelines are not followed or adequately enforced. Furthermore, low selectivity during the extraction stage increases the amount of uneconomic material being fed into the comminution and processing stages, increasing the energy demand of the operation, resulting in elevated GHG emissions. Stochastic methods are well-suited to the waste management problem, due to more accurate modelling of uncertain variables such as geochemical composition, decreasing the risk of not meeting environmental performance targets. One of the first stochastic methods taking waste management into account was introduced by Spleit and Dimitrakopoulos (2017), who proposed a stochastic mixed integer programming (SMIP) formulation designed to optimize the long-term schedule in the LabMag iron ore deposit in Labrador, Canada. As a notable addition to standard consideration in previous SMIP formulations, this work focused on avoiding stockpiling whenever possible and maximizing space in the mined-out pit, ensuring adequate destination for in-pit tailings disposal. This is done by including constraints that prioritize exposing the floor of the ultimate pit as early as possible, providing adequate space for waste storage. Moreover, the stochastic schedule minimizes the amount of waste generated by avoiding a barren shale layer and decreases transport requirements due to optimized pit shape. The geological uncertainty has been accounted for using 10 orebody realizations, with the UPL being developed by a nested Lechrs-Grossman algorithm. While achieving a significant reduction in waste generated, the stochastic schedule outperformed the deterministic solution, providing 16.9% higher NPV across the 10 year planning period. This work has been extended by Rimélé et al. by providing a two-stage SIP formulation solved using sliding time window heuristics, which aimed to simultaneously optimize the extraction sequence and destination policy with in-pit waste disposal (Adrien Rimélé, Dimitrakopoulos, and Gamache 2018). In this work, researchers address waste management by defining strips within the block model, which, after being extracted, are defined as storage units. Specific constraints are introduced to avoid the in-pit disposal from interfering with the extraction sequence, and the study manages to successfully deposit over 80 millions m3 of tailings within the original pit. Both of the studies acknowledge the fact that the studied deposit is particularly well-suited for in-pit disposal, and propose extension of the method to suit more orebody geometries. Saliba and Dimitrakopoulos (2020) address the topic of waste management from a different perspective. Instead of optimizing the destination policy, their simultaneous stochastic optimization solution starts by identifying the fraction of the material that will need to be disposed of in a tailings storage facility (TSF) and, given the fact that this facility is the main bottleneck for the operation, strategically delays its extraction to maximize the use of the existing TSF . The focus on acid mine drainage (AMD) generating waste is also present in Levinson and Dimitrakopoulos (2020), where the sulphur and carbon content of the ore blocks is simulated on top of standard grade considerations. These variables act as a proxy for acid generation potential, with sulphidic material accounted for as potentially acidgenerating (PAG) and blocks higher in carbonates as non-acid-generating. This approach, coupled with a dynamic CoG policy, allows for a more precise management of waste
RkJQdWJsaXNoZXIy MTM0Mzk2