Mineral processing plants are inherently complex, with wide-ranging resource-intensive operations. Mines are looking to achieve significant process objectives to lower energy and plant costs while making better use of water and energy resources.
At the same time, they are tasked with meeting regulatory requirements and dealing with workforce challenges. To address these complexities and to offer solutions, Rockwell Automation recently hosted a webinar that helped attendees understand how to meet these process objectives.
Juliano de Goes Arantes, Pavilion Account Executive at Rockwell Automation, hosted the webinar on 20 May, supported by Steffen Zendler, Rockwell Automation’s Heavy Industry Strategy and Marketing Manager for EMEA. The pair discussed the main control challenges faced in mineral processing operations.
De Goes Arantes said that more than 75% of mineral processing plants are still using basic control strategies. An audience poll showed similar results in Africa, with 56% saying they use basic control, followed by 28% who use manual control.
“While basic control provides adequate control in terms of plant safety, it rarely achieves optimal control in quality, nor does it operate most economically,” said de Goes Arantes.
The webinar focused on Pavilion8 Model Predictive Control (MPC) from Rockwell Automation, a tool that reduces process variability and enhances stability over and above what is currently possible with more traditional control schemes. Making use of MPC, processing plants can leverage their control systems to optimise their operations.
De Goes Arantes explained:
“MPC uses a model of the minerals processing operation to predict how the process output variables will respond to changes in the process input variables and disturbances. MPC algorithms make use of machine learning, where the engine learns and updates the mathematical model using data, while the MPC then uses it for control.”
It is a simple and powerful technology which integrates into the current control system to optimise mineral processes while addressing the objectives and complexities of the plant.
“At its core, MPC uses supervised machine learning technology to assess current and predicted operational data. It compares the data to desired results, and then computes and updates the process online setpoint targets,” said De Goes Arantes.
MPC reduces variability, helps achieve plant stability, manages the process without constraints, and operates closer to specifications and performance limits while maintaining safety margins. Ultimately, the tool delivers mining customers with increased throughput, lower reagents consumption, better recovery, optimum water and energy usage, and improved process stability.