Predictive analysis of energy consumption in minining for making decisions

This research is centered on how predictive analysis can support decision making on energetic consumption in the great copper mining in Chile. In this article it is analyzed the data base of energetic consumption in the Grinding A0, A1 and A2, in Codelco, Chuquicamata, in the period 2007–2014. This study uses the Box-Jenkins method for predicting energetic consumption in the Grinding. Before this analysis, it is achieved to predict the behavior of the time series of energetic consumption under stationary conditions, through polynomial delay equations, Seasonal Auto Regressive Integrated Moving Average (SARIMA). As a result, it is obtained that most of the consumptions lines in the Grinding A0, A1, and A2, present some type of trend and stationary. The SARIMA model was able to adapt to the behavior of the variables of the energetic consumption of the copper productive process. It is possible to predict the energetic consumption as critical information in the decision making in the Grinding A0, A1 and A2.