Abstract
In mid to large-scale processing plants, even slight enhancements in process efficiency can yield significant benefits. The daily performance of a plant relies heavily on the effectiveness of the process control system and the operator's expertise. In conventional process plants, the control of the grinding circuit stands out as a challenging area with substantial potential for improvement due to nonlinear responses, a high degree of variable interaction, and material recirculation.
To address these challenges, model-based control approaches have proven to be effective. With the growing availability of data and the increasingly stringent control requirements for grinding circuits, the development of data-driven models provides an opportunity to enhance the capability and efficiency of advanced model process control (MPC) in these circuits.
This paper outlines a stepwise approach to developing and implementing an advanced MPC solution for a grinding circuit in a silver mine in Mexico. This approach helps mitigate investment and implementation risks while demonstrating the value at each step. The results of this process will be discussed in detail.