Automatic Control of Flotation Columns - Advanced Control and Supervision of Mineral Processing Plants - page 275

Tabl e 6. 1 References describing several flotation column control applications

Reference Controlled variables Control algorithm Type of column

[78] Froth depth, bias PID Industrial

[79] Froth depth Backstepping, Pilot, 2 phases

model reference
nonlinear control

[61] Grade, recovery Fuzzy logic supervising Pilot, industrial

local loops

[62] Grade, recovery Expert system Industrial

[30] Froth depth, bias Review of different methods Pilot, 2 and 3 phases

[68] Froth depth, gas hold-up MPC Laboratory

[4] Froth depth PID Industrial

[21] Froth depth, bias PID Laboratory, 2 phases

[31] Froth depth Gain-scheduled PI Pilot, 2 phases

[81] Grade Fuzzy logic Industrial

[82] Column performance through Fuzzy logic Industrial

pulp/froth zone densities
and/or grade and recovery

[33] Froth depth, bias, gas hold-up Decentralized PI Pilot, 2 phases

[34] Froth depth, bias, gas hold-up PI + MPC Pilot, 2 phases

[83] Column concentrate grade and Expert system Industrial

tailing grade of cleaner scavenger

[84] Froth depth, bias MPC Pilot, 2 phases

[69] Froth depth, gas hold-up MPC Pilot, 3 phases

[36] Froth depth, bias, gas hold-up PID Pilot, 2 phases

[85] Froth depth, gas hold-up MPC Laboratory, 3 phases

6.4.3 Application Examples

Two examples are presented, the first a multivariable control (3 × 3) for a two-phase

system and the second a multivariable control (2 2) for a three-phase system. In

both cases, the MPC was selected for its ability to handle the existing operating
constraints listed later on.
In the first example [34], regulatory PI controllers were implemented on an in-
strumented pilot flotation column (internal diameter: 5.1 cm; height: 732 cm), as
illustrated in Figure 6.11. Since the primary interest was the development of a regu-
latory control (column hydrodynamic characteristics), a two-phase system was suf-
ficient to illustrate the advantages of the proposed approach. Adequate frother con-
centration (Dowfroth 250) was used to provide proper bubbling conditions and froth
stability. The controlled secondary variables were froth depth, gas hold-up and bias
rate. Froth depth was measured using the weighted average method described at the
end of Section 6.3.1 [24], whereas gas hold-up value was estimated from measured
conductivity values using Maxwell's equation (Equation 6.10) and bias rate was es-
timated using the method proposed by Maldonado et al. [32] detailed in Section
6.3.2. The manipulated variables were tailing flow rate, wash water rate and gas rate

×