Determination of the froth height as a prerequisite of a model-based flotation control

Flotation represents one of the most important separation processes in the mineral processing industry. By exploiting differences of the surface properties, it enables an effective way for the separation of valuable from gangue material. Through the addition of air to the material pulp it is possible to selectively attach hydrophobic particles to air bubbles and to transport them to the upper end of a flotation cell. The froth phase on top of the pulp is crucial for the generation of a high quality concentrate, as it serves as a cleaning region to remove unselectively entrained particles from the product stream. Therefore, the detailed monitoring and control of this froth phase is a crucial requirement for an efficient and stable process. Flotation represents a complex process, where multiple input variables influence several sub-processes and output parameters. Such a multi-input multi-output problem cannot be efficiently controlled by conventional PIDs.
A model-based predictive controller (MPC) is required to reproduce the complex relations between the various in- and outputs to calculate an optimum set of parameters. For the successful implementation of MPCs the calculation time represents a critical factor. The required complexity of the model and therefore the calculation time can be reduced by measuring relevant process parameters. Such characteristic variables are the density of the pulp, the froth and the pulp height. Nevertheless these information need to be available at all time and as accurate as possible, so the complexity of the MPC is reduced permanently since no further failsafe methods are required.
This contribution focuses on a combination of a radar level meter and a hydrostatic pressure sensor to form a measurement system, being capable of measuring the three mentioned values at once. It is implemented into a pilot flotation plant to investigate its applicability for the concept of a “reduced” MPC. One scope of the investigations is the response of the radar signal to the typical flotation fluid, forming a three-phase systems (solid, liquid, gaseous), and consequently the robustness to disturbances and the reasonable sample rate of the sensor system. Additionally, the pilot plant is equipped with optical sensors to assess the accuracy.