Incorporating grade uncertainty and stockpiling in stochastic open stope production scheduling optimization

Traditional geostatistical methods such as ordinary kriging often generate biased and overly smoothed grade estimates. Simulation‑based techniques overcome this limitation by producing multiple orebody realisations, allowing improved representation of grade uncertainty in mine planning. This study extends an existing MILP model to a Stochastic Mixed Integer Linear Programming (SMILP) framework to maximise net present value (NPV) under uncertainty. Two case studies with six scenarios were analysed. Results show that the SMILP model incorporating stockpile management achieved the highest NPV, increasing NPV by 4% and 0.28% compared to corresponding MILP models, demonstrating the practical benefits of accounting for grade uncertainty.