A stochastic cut-off grade optimization model to incorporate uncertainty for improved project value

Cut-off grade is a decision-making criterion often used for determining the quantities of material (ore and waste) to be mined, ore processed, and saleable product. It therefore directly affects the cash flows from a mining operation and the net present value (NPV) of a mining project. A series of different cut-off grades that are applied over the life of mine (LOM) of an operation defines a cut-off grade policy. Due to the complexity of the calculation process, previous work on cut-off grade calculation has mostly focused on deterministic approaches. However, deterministic approaches fail to capture the uncertainty inherent in input parameters such as commodity price and grade-tonnage distribution. This paper presents a stochastic cut-off grade optimization model that extends Lane's deterministic theory for calculating optimal cut-off grades over the LOM. The model, code-named 'NPVMining', uses realistic grade-tonnage realizations and commodity price distribution to account for uncertainty. NPVMining was applied to a gold mine case study and produced an NPV ranging between 7% and 186% higher than NPVs from deterministic approaches, thus demonstrating improved project value from using stochastic optimization approaches.