This paper presents a learning curve model for forecast the operational mining cost. One of the great advantages of this model is the possibility to include the technical and economic changes in the cost performance that are incorporated by the time series. The learning model projections presents a good level of accuracy for cost data from iron ore. The comparison between actual and forecast data costs through the Stanford-B model indicates that the output values are within an acceptable range of 10% confidence interval.
mining cost; learning curves; iron ore and mining planning
