Trabajo ganador del Premio Nacional de Minería en PERUMIN 35

Por: Hugo Escobar y Aldo Aliaga, Sociedad Minera Cerro Verde.

Abstract

A copper concentrator's production plans are based on the pounds produced and contained in the concentrates for their subsequent commercialization. 

The daily operation of concentrators is governed by two variables to be optimized: tonnage and recovery; both inversely related. Achieving the optimum balance between these two variables is the main challenge to maximize the pounds of copper in concentrates.

A tool that has provided good results in the concentrators of Sociedad Minera Cerro Verde has been the implementation of Artificial Intelligence (AI). The use of Machine Learning to model tonnage and recovery, and thus optimize the production of pounds of Cu. 

The recommendations obtained with the tool, the systematic process of reviewing key variables for the operation, the detection of instrument anomalies and the interaction between concentrator professionals during the review of recommendations, have served to improve and standardize operational decisions, breaking many paradigms along the way.

By implementing this tool and process, we were able to identify improvement opportunities in the plants that led to flotation circuit optimization projects, and to date, after more than a year of execution, an increase in the production of Cu pounds of up to 6.5% has been obtained, without losing concentrate quality and, in many cases, improving it.