Flotation monitoring using fundamental dynamic models: Investigating the effect of particle size on attachment

Froth flotation is one of the most common and widely used separation processes in the mineral processing industry; however, its complexity means that complete understanding of the process has not yet been achieved. A dynamic fundamental model is needed to enhance the understanding of the process and to improve process control. It is challenging to capture all the phenomena involved; hence, we propose a multiscale modeling approach with a framework ranging from macroscopic (equipment scale) to microscopic (attachment, detachment) models along with distribution models for ore particle sizes. Real-time froth properties are obtained using a vision-based sensing system. These measurements can be used for developing a soft sensor for grade using statistical methods such as PCR (principal component regression) and PLS (partial least squares) regression. The fundamental model is updated in real-time using extended Kalman filtering. Attachment dynamics are studied for different particle size distributions. The modeling and soft sensing is tested on batch flotation cell for the pure mineral galena with different particle size distributions using Xanthate and MIBC as collector and frother, respectively. The model, online measurements and soft sensor are useful in developing monitoring methods for froth flotation.