Mine planning models have proved to be very effective in supporting decisions on sequencing the extraction of material in copper mines. These models have been developed for CODELCO, the Chilean state copper mine and used successfully. Here, we wish to develop a corporate model, including all mines of CODELCO. The original models are extremely large MIPs. In order to run a global model, the original models need to be reduced significantly. We develop an approach to aggregate the models. The aggregation is done both on the original data of the mine as well as on the MIP original models. The aggregation is based on clustering analysis. Promising results were obtained with data of a large underground mine. © 2008.