Introduction to the Optimizer
The core purpose of the Flotation Optimization Application (the Flotation App for short) is to empower flotation operation teams to optimize their Flotation App. In other words, to best align it with its intended purpose. At the heart of the Flotation App is the Flotation Optimizer. Other features of the Flotation App, like the Virtual Sensors and the Simulator, are supplementary to the Optimizer.
The Flotation Optimizer essentially seeks to answer the core question:
What is the best way to operate a particular flotation circuit with a particular feed, to best meet its performance objectives?
We can unpack this question into different parts:
"What is the best way to operate"
The Flotation Optimizer is not a stabilising control system seeking to control a flotation circuit to entered setpoints. It is an optimization system that determines the best operating condition (based on the feed) at steady-state and provides the resulting optimal setpoints to lower-level control systems (or Operators during manual control) so that the control systems can change the process actuators (like valves and/or pump speeds) to achieve and maintain the process at these recommended setpoints. The variables for which the Flotation Optimizer provide recommended setpoints are called Control Variables. See Figure 1 below."a particular flotation circuit"
The Flotation Optimizer seeks to optimize a flotation circuit within the constraints of its current design (e.g. equipment types & -sizing), configuration (e.g. flow routing) and instrumentation (e.g. sensors, actuators & control systems)."a particular feed"
The material properties of the feed ore and the feed rate are considered to be given, outside the scope of the Flotation Optimizer. The most accurate & most recent data must be used for the Feed Variables as the Flotation Optimizer will assume that this is the feed that's entering the process & optimize the process accordingly."to best meet its performance objectives"
The Flotation Optimizer depends on the Value Driver being set, describing what the flotation circuit is trying to achieve. The Value Driver consists of two parts: setting Rewards for the Performance Variables (e.g. maximise metal recovery) and limits for the Control Variables (e.g. keep the Rougher 1 airflow rate between 300 and 600 m3/h).
Figure 1: Simplified diagram of the interaction between the Flotation Optimizer and control systems. The exact variables will depend on the specific site configuration.
Main Variables
The following list of variables are the ones typically used by the Flotation Optimizer. This is very specific per site though, as it's possible to configure the Flotation Optimizer to use what the client measures and controls - and to make a simplifying assumption or more comprehensive plan if any important measurement is lacking.
Configuration Variables
These are set once and not changed in real-time.
Flotation circuit configuration (layout & flows)
Position of different variables in the circuit (e.g. sensor location)
Flotation cell* types & dimensions
Impeller dimensions & RPM
Operational ranges of all variables
*While the word cell is used, this can also refer to a flotation column or -bank with a shared level.
Feed Variables
These should be updated as frequently as they are available. Can use forecasts or the most recent measurements.
Feed Pulp density or %solids
2 Feed particle size measurements (e.g. P50 & P80)
Feed flow rate
Feed grades of all components tracked in the site's Flotation App (g/t or %)
Performance Variables
The Performance Variables depend on the task or intended purpose of a specific flotation circuit. This can differ between commodities. The following are the typical Performance Variables that the Flotation Optimizer has targeted thus far.
Concentrate metal grade (g/t or %), typically kept above a spec minimum
Metal recovery into the concentrate (%), typically maximised
Gangue grade in the concentrate (g/t or %), typically kept below a spec maximum
Mass pull for the overall circuit, typically kept below a maximum (to protect downstream overloading)
Control Variables
This depends on the current control strategy at the site, as the Flotation Optimizer should provide setpoints that can be used by the existing control systems. The following are what the Flotation Optimizer has most often provided thus far.
Air flow rate per flotation cell
Froth depth or slurry level per flotation cell
Reagent dosage (g/t), per reagent & per dosing point
Next read: