The mine production planning defines a sequence of block extraction to obtain the highest NPV under a number of constraints. Mathematical Programming has become a widespread approach to optimize production planning, for open Pit mines since the 1960s. However, the previous and existing models are found to be limited in their ability to explicitly incorporate the ore grade uncertainty into the planning process. To overcome this shortcoming, this paper presents an Integer Programming (IP) model, for long-term planning of open Pit mines. This model is set up to account for grade uncertainty. The grade distribution function, in each block is used as a stochastic input, to optimize the model. The deterministic equivalent of this model is then achieved by using stochastic Programming, which is a form of nonlinear in binary variables. Because of the difficulties in solving large scale nonlinear models, the model is then approximated by a linear one. This formulation will yield schedules with high chance of achieving planned production targets, while maximizes the expectation of net present value, it simultaneously minimizes the variance in function.

An open-Pit mine production planning begins with determining the ultimate Pit limit of an open-Pit mine. The ultimate Pit limit solver selects blocks whose total economic value is maximum while meeting the slope constraints. In other words, a group of blocks that maximize a selected parameter, such as profit, metal content, or net present value, is considered in determining the ultimate Pit limit. Also, the ultimate Pit limit is designed to select the waste dump location, surface facilities, extractable reserves, and the amount of waste removal. The production planning problem in large-scale open-Pit mines is referred to as an NP-hard problem because it cannot be solved in a reasonable computational time. To solve this, various methods, including aggregation methods, have been proposed to reduce the size of the issue. In this paper, to evaluate the efficiency of the block aggregation technique based on the Pit values and computational times, at first, the heuristic Tabesh-AskariNasab aggregation algorithm was applied to the block models with 2400 and 11400 blocks. Then the ultimate Pit limit based on the original block model and reconstructed block models were determined using the linear Programming model. Comparing the results in both block models indicates that the block aggregation approach considerably decreased computational time while generating near-optimal Pit values. These results are more critical in large-scale production planning problems, exactly in open Pit mine scheduling. Furthermore, the slope of Pit walls was decreased by increasing the size of clusters, and the stripping ratio increased in both block models.

Determination a sequence of extracting ore is one of the most important problems in mine annual production scheduling. Production scheduling affects mining performance especially in a polymetallic open Pit mine with considering the imposed operational and physical constraints mandated by high levels of reliability in relation to the obtained actual results. One of the important operational constraints for optimization is the uniformity of metallic minerals grade after the blending process. This constraint directly affects the performance of the mineral processing plant. The sequence of extracting ore is usually determined by the order of pushbacks, which should be mined. Metallic minerals’ grade in each pushback is stochastic in nature that comes from some statistical Errors committed by the sampling. In such situations, decision making about the order of pushbacks for extraction as an exact defined process is not possible. Moreover, the decision-maker should maximize the total Net Present Value NPV as the main objective of mining operations by considering the high performance of mineral processing plant. To deal with such cases, this research proposes a model based on the chance-constrained one-sided goalProgramming and the obtained results from this procedure confirms the model’s reliability and correctness.