Optimization of the Haile Gold Mine Grinding and Flotation Line Using Model Predictive Control

"This paper describes the application of the BrainWave Model Predictive Controller (MPC) and Advanced Control Expert (ACE) Advanced Supervisory Control System on the SAG Mill, Ball Mill, and Rougher Flotation cells for a new gold processing line at the Oceana Gold - Haile Gold mine in Kershaw, South Carolina. For this project, ANDRITZ supplied the control system design and programming, operator-training simulator and advanced control. The advanced control was tested against the IDEAS™ dynamic simulator, and commissioned during the startup of the processing plant to help achieve target production as quickly as possible. Operators were able to run the equipment in a consistent and optimized manner from an early stage of the plant operation, increasing efficiency and profitability. MPC is able to reduce process variability beyond the best performance that can be obtained with conventional Proportional Integral Derivative (PID) control methods. MPC is able to optimize the control of processes that exhibit an integrating type response in combination with transport delays or variable interaction, which are characteristic of many gold processes including SAG mill weight control and Flotation cell froth level control. The experience and benefits of applying MPC during the startup of a new gold processing plant are presented in this paper.INTRODUCTION Mineral processing operations present many challenges for automatic process control due to variations in unmeasured ore properties, material transport delays, and non-linear response characteristics. Advanced process control technologies such as expert systems and multi-variable model predictive controllers are often used to improve the control of these processes to increase productivity and recovery efficiency. Typically these systems are applied after the mine has been in operation for a period of time. For this project, a complete dynamic process simulator was developed using the IDEAS platform for the purpose of training operators before the plant was constructed. Operators were able to become familiar with the process and the control system in order to become proficient in the use of the controls before the plant was ready to start. They were able to practice abnormal condition management including startup and shutdown of the process using the simulator and develop complete Standard Operating Procedures (SOPs) for these situations. The simulator was also used to test the control system functionality and operator Human Machine Interface (HMI) configuration to reduce errors and improve the performance of the control system during plant start-up."