Design and Implementation of Advanced Automatic Control Strategy Based on Dynamic Models for High Capacity SAG Mill

"The Copper Concentrator Plant of Compañía Minera Doña Inés de Collahuasi (CMDIC) is composed of three grinding lines, the third one being the largest. Due to the capacity feature of the third line, specially its SAG mill, it is a priority that the advanced process control team of the Company deploys a supervisory control strategy that allows this unit to operate in an optimal way, in order to achieve the maximum productivity at minimum cost. It is in this kind of tasks where advanced control applications play a key role in obtaining the best economic profit of the process.The advanced process control team, formed by people from Informatics, Automation and Operations of CMDIC have assumed the challenge of generating the transition from a manual SAG Mill 1011 operation (40 x 22 feet, the largest copper SAG mill of the world), to a fully automatic SAG Mill using a multivariable control strategy instead of a rule based on expert control system (widely used in industrial mining process control applications).Therefore, CMDIC decided to work together with Honeywell APC Team implementing an advanced automatic control strategy for the SAG Mill 1011, using for this purpose a technology called Model Based Predictive Control (MPC), based on a variety of control methods that uses dynamical models of the process, obtained from input/output data, in order to print the process knowledge into a group of mathematical relationships that links the constraints (controlled variables) with available commands (manipulated variables) and measured disturbances.The main objective of the advanced control strategy is to ensure process government, tending to throughput stability and optimization, by the execution of coordinated control actions, proactively avoiding measured disturbances effects and the SAG mill overloading.The most important benefits of using a multivariable control strategy are: the automated operation of SAG mill 1011 (that was manually operated before the implementation of this application), a safer, more coordinated and more stable process operation, achieving through this an increase of throughput and a decrease of specific energy consumption, controlling the amount of oversize material production, power consumption, mill loading, noise and torque. In addition to the operational results, this challenge has consolidated a control team, with the participation of engineers and operators, diminishing the typical implementation time and achieving a greater use of the application in a short time."