Challenge

Human decision making and reactive response to changing process conditions & slow feedback from KPI data result in operators often not knowing how each of their flotation cells/columns is performing & contributing to the overall recoveries.

The problem is that most Operators do not have live feedback on whether their air flow rates and reagent dosages are producing the right cell or bank performance in terms of Metal recovery, concentrate grade and mass pull. They often only get feedback on this when they take samples for lab analysis, after which they adjust their control variable setpoints.

This uncertainty means that flotation circuits will operate at wrong set points for a long time, limiting the circuit's performance and driving up OPEX($/t).

Our Solution

The intelliSense.io solution is a predictive model based on first principles modelling and flotation kinetics using a simulation type approach. The kinetic model uses lab analysis data which is fitted to the kinetic model on the flotation circuit. The digital process model is deployed as live outputs as virtual sensors. This provides a quantitative link between feed- and control variables, and final flotation performance (Grade & recovery), including real-time grade & recovery feedback per cell/bank.

Value

This live feedback on flotation performance allows Operators/Metallurgists to:

  • Make immediate parameter setpoint corrections, instead of waiting for non-real-time(Offline) lab sample results to inform the suitable plant configuration.

  • Run at the most stable air flows and reagent dosages, leading to increased circuit performance on recovery, concentrate grade and mass pull.

  • Minimize metal grade tailings.

Did this answer your question?