Rougher Flotation Circuit Research Articles

Prediction and Optimisation of Copper Recovery in the Rougher Flotation Circuit

In this work, the prediction and optimisation of copper flotation has been conducted in the rougher flotation circuit. The copper-recovery prediction involved the application of support vector machine (SVM), Gaussian process regression (GPR), multi-layer perceptron artificial neural network (ANN), linear regression (LR), and random forest (RF) algorithms on 15 rougher flotation variables at the BHP Olympic Dam. The predictive models’ performance was assessed using linear correlation (r), root mean square error (RMSE), mean absolute percentage error (MAPE), and variance accounted for (VAF). A simulated annealing (SA) optimisation algorithm, particle swarm optimisation (PSO) algorithm, surrogate optimisation (SO) algorithm, and genetic algorithm (GA) were investigated, using the GPR predictive function, to determine the optimal operating condition for maximising copper recovery. The predictive function of the best-performing model was extracted and used in optimising the flotation circuit. The results showed that the GPR model developed with the matern 3/2 kernel function makes the most precise copper-recovery prediction as compared to the other investigated predictive models, obtaining r values > 0.96, RMSE values < 0.42, MAPE values < 0.25%, and VAF values > 94%. A hypothetical optimisation solution assessment showed that SA provides the best set of solutions for the maximisation of rougher copper recovery, obtaining a throughput of 638.02 t/h and a total net gain percentage of 14%–15.5% over the other optimisation algorithms with a maximum copper recovery of 94.76%. The operational benefits of implementing these algorithms have been highlighted.

Comparison of Fuzzy and Neural Network Computing Techniques for Performance Prediction of an Industrial Copper Flotation Circuit

This paper presents the development and validation of five different soft computing methods for flotation performance prediction: (1) two models based on fuzzy logic (Mamdani and Takagi-Sugeno fuzzy inference system) and (2) three models based on artificial neural networks. Copper content in the ore feed, collector dosage in the rougher and the scavenger flotation circuits, slurry pH in the rougher flotation circuit and frother consumption were selected as input parameters to estimate the copper grade and recovery of final concentrate, as well as the copper content in the final tailings of the flotation plant. The training and evaluation of the proposed models were performed on the basis of real process data collected by the multiannual monitoring of industrial flotation plant operation in “Veliki Krivelj Mine”. The results showed that the proposed soft computing-based models well describe the behavior of the industrial flotation plant in a wide range of circumstances. Among the proposed algorithms, artificial neural networks gave the most accurate predictions for the final copper concentrate grade and recovery (R2 = 0.98 and R2 = 0.99, respectively) and copper content in final tailings (R2 = 0.87). At some points, fuzzy logic models are almost equally efficient, but artificial neural networks had lower values for all error functions.

Predicting the recovery and grade of a rougher flotation circuit from liberation data

A flotation model has been validated against the survey data obtained from a copper rougher flotation circuit with a circulating load. The model was developed using the hydrophobic force as a kinetic parameter, which made it possible to use the contact angles of the free and composite particles as input parameters so that the model can predict both recoveries and grades. The model also recognizes the role of bubble coarsening in determining the froth phase recoveries and grades. The model predictions are in good agreement with the survey data. The size-by-size recovery curve as predicted from the model shows an optimum particle size range that is consistent with most of the operating data reported in the literature. The model suggests that the upper particle size limit may be determined by the instability of the froth phase rather than the turbulence in the pulp phase of a flotation cell.

Flotation of Calcium Minerals with Combination of Reagents of Different Molecular Structure

The influence of non-ionic compounds in combination with sodium oleate on the contrast behavior of flotation response of calcium minerals is investigated theoretically and experimentally. The flotation and adsorption analyses are carried out with monomineral fractions of calcite and scheelite, while the flotation process analysis is conducted with an ore material screened into size grades of -44 and -15 urn. The process conditions of the highest difference in adsorption of oleate and contrast in extraction of scheelite and calcite are found. The conditions of the maximum depression of calcite in the rougher flotation circuit are determined.

Off-Policy Q-Learning: Set-Point Design for Optimizing Dual-Rate Rougher Flotation Operational Processes

Rougher flotation, composed of unit processes operating at a fast time scale and economic performance measurements known as operational indices measured at a slower time scale, is very basic and the first concentration stage for flotation plants. Optimizing operational process for rougher flotation circuits is extremely important due to high economic profit arising from the optimality of operational indices. This paper presents a novel off-policy Q-learning method to learn the optimal solution to rougher flotation operational processes without the knowledge of dynamics of unit processes and operational indices. To this end, first, the optimal operational control for dual-rate rougher flotation processes is formulated. Second, H∞ tracking control problem is developed to optimally prescribe the set-points for the rougher flotation processes. Then, a zero-sum game off-policy Q-learning algorithm is proposed to find the optimal set-points by using measured data. Finally, simulation experiments are employed to show the effectiveness of the proposed method.

Adapting a Phenomenological Model of a Rougher Flotation Circuit to Industrial Historical Operating Data Base

The use of simulators is a powerful tool to train plant operators and to be also incorporated in the development and test of supervisory control strategies. However, the phenomenological models describing the process are relatively complex, characterized by nonlinear relationships and whose parameters are depending on many local factors, such as, the plant configuration, the individual characteristics of the equipment, the availability of on-line measurements and the characteristics of the feed, among others.In this work, the previously developed phenomenological model is adapted to the particular characteristics of the rougher circuit of an industrial flotation plant, considering its particular layout and the available information to feed the simulator. The rougher circuit consists of three lines of 8 mechanical cells processing a feed of 4000 tons/h.The new model predictions were tested for a family of feed characteristics, including variation of mineralogical species under different operating conditions. Since some variables are commonly unmeasured during the operation, additional data were incorporated to improve the model predictability. The use of the simulator is illustrated in several examples, as well a discussion of the model prediction limitations due to some particularities found in the historical operating data.

Control del Nivel de Pulpa en un Circuito de Flotación Utilizando una Estrategia de Control Predictivo

This work presents the design and implementation results of a model predictive control strategy used to control the pulp level of a Rougher flotation circuit in a mine located in the third region of Chile, which is composed by five flotation banks. The strategy considers a state space representation to model the pulp level of each bank (with a multiple input - single output model), which is obtained by a system identification procedure and uses a Kalman filter as a state estimator. To solve the optimization problem that calculates the control law, a genetic algorithm based optimization tool is used. Experimental data is used to show the results of the proposed control strategy.

Correlation between the top of froth grade and the operational variables in rougher flotation circuits

Cu grades at the top of the froth (TOF) were compared to the froth depth (hF) profiles and froth discharge velocity profiles along rougher flotation circuits. Measurements were performed in two self-aspirated flotation banks, one consisting of nine 130m3 cells (1-2-2-2-2 arrangement), and the other one of six 250m3 cells (1-1-1-1-1-1 arrangement). Two behaviors for the TOF grade were observed: (i) for hF⩽10cm, a decrease in the TOF grade along the bank was observed because either the decrease in mineral liberation or a potential recovery of slow floating gangue, (ii) for hF⩾10cm, an increase in the TOF grades with the froth depth was observed, mainly due to an increase in froth selectivity. In addition, an inverse relationship between TOF grades and froth discharge velocities was obtained in most cases. However, in the first rougher cell, both higher froth discharge velocities (7–14cm/s) and higher TOF grades (20–28%Cu) were observed. The froth in the first cell is typically loaded, which favors the froth stability as well as the concentrate discharge velocity.The TOF grades of Mo as a function of the operating variables showed the same dependency as those observed with the TOF grade of Cu. This result indicated that Cu and Mo minerals had similar flotation rates in the rougher operation, which was in good agreement with the comparable ranges of Cu and Mo rougher recoveries.

Predictive Expert Control System of a Hybrid Pilot Rougher Flotation Circuit

Abstract: In the last decades heuristic approaches based on logic rules have been one favored tool to control flotation units and plants on top of distributed control systems. The lack of fundamental knowledge on flotation processes, expressed in reliable models, and programs to assure the quality of the information, contained in available measurements, have limited the developing of robust predictive control strategies. In this paper we explore the idea of using simplified models jointly with measured disturbances to modify the set points of froth depth and air flow rate controllers. Since these target predictions are only approximate, a conventional feedback expert system, based on logic rules, continues the task of calculating new set points of the distributed controllers, in order to decrease the gap between actual and desirable metallurgical targets. This combined actions permitted a fast response of the process to change the operation of the circuit when a measured disturbance change is detected. This control system was implemented and experimentally evaluated in a pilot rougher flotation circuit, with three pneumatic cells, with froth depth and air flow rate controllers in each cell. Since the circuit is operated for the air-water system, a phenomenological model, with parameters estimated from industrial data, is fed on-line with real operating variables and virtual values, characterizing the feed (grades, solids%, particle size distribution), in order to predict the metallurgical targets (recovery and concentrate grade). Several cases are discussed showing the benefits of this control system and the possible improvements to work on.

Application of D-optimal design for optimizing copper-molybdenum sulphides flotation

Froth flotation is widely used for concentration of base metal sulphide minerals in complex ores. One of the major challenges faced by flotation of these ores is selection of the type of flotation reagents. In this study, the D-optimal experimental design method was applied to determine the optimum conditions for flotation of copper and molybdenum in the rougher flotation circuit of the Sungun copper concentrator plant. The investigated parameters included types and dosages of collectors and frothers, diesel dosage and feed size distribution. The main effects on copper and molybdenum recoveries and grades were evaluated. Results of optimization showed that the highest possible grade and recovery were obtained for Z11 as a primary collector (20 g/Mg), R407 as a first promoter collector (20 g/Mg), X231 as a second promoter collector (7 g/Mg), A65 (15 g/Mg) and Pine oil as frothers (5 g/Mg), 20 g/Mg of diesel dosage, and d80 of feed size was equal to 80 μm. The analysis of variance showed that the primary promoter collector was the most significant parameter affecting the recovery of Cu, while diesel dosage and d80 were the most significant parameters influencing the Mo recovery.

A study on the modified flotation parameters and selectivity index in copper flotation

ABSTRACTSelectivity index (SI) and modified flotation rate constants of copper and iron were evaluated to rougher flotation circuit of the Sarcheshmeh copper ore by a series of batch flotation tests. Kinetics model of Agar was fitted to experimental data. The results indicated that first-order kinetics model of Agar presented the time–cumulative recovery curves of flotation tests quite well for Cu and Fe. Also, the results showed that the selectivity between Cu and Fe enhanced with increasing the pH, solid content, and dosage collector. In addition, performance of collectors showed that the best selectivity between Cu and Fe was in the order of combination of Z11 and MBTs > flomin C4132 > combination of MBTs and flomin C4132 > combination of Z11, MBTs, and flomin C4132 in flotation of copper. It was also found that better results for flotation kinetics were obtained at pH 11.5, solid percentage 28, and dosage collector 40 g/t (15 g/t sodium isopropyl xanthate and 25 g/t mercaptobenzothiazole [MBT]). SI value was determined as 5.12 in optimum condition.

Effect of entrainment in bubble load measurement on froth recovery estimation at industrial scale

Froth recovery was calculated in a 130m3 mechanical cell of a rougher flotation circuit. This was done by bubble load determinations along with mass balance surveys. Valuable grade in the bubble load decreased in the −38μm due to fine particles entrained to the chamber of the device. The effect of fine particle entrainment on froth recovery was evaluated. A comparison between results from the raw bubble load data (assuming all particles were transported by true flotation) with those from corrected bubble load information (subtracting fine particle entrainment) was carried out. Entrainment occurred due to hydraulic transport in the bubble rear, which corresponds to the worst case scenario for froth recovery estimation. Results showed that the relative error was less than 0.3%, which allowed validation of the bubble load measurement as an effective methodology for froth recovery estimation at industrial scale.

On the collection of valuable minerals along rougher flotation banks

The collection of valuable minerals along rougher flotation banks was evaluated using top-of-froth measurements, which consist of sampling the froth surface. Considering non-selective froths observed in rougher flotation circuits, it was found the top-of-froth grades were similar to the bubble load grades, which allowed for a direct estimation of the collected mineral characteristics (size distribution, mineralogy, and grade). Thus, the flotation process evolution can be characterized by analyzing the collection of valuable minerals by true flotation instead of the classical mass balances approach, which strongly depends on the reliability of the tailings grade data. In addition, comparison among different valuable minerals was observed without significant sampling disturbances caused by solid entrainment.It is expected the top-of-froth measurement will be a useful tool for evaluating the effect of operating variables such as pulp conditioning, air flowrate and solids percentage on the true collection process along a flotation bank.

Pyrite recovery mechanisms in rougher flotation circuits

Bubble load and top of froth measurements per size class were carried out in a rougher flotation circuit containing sulphide copper minerals and pyrite. This circuit consists of four parallel banks of seven 130m3 mechanical cells. Results of testing the first rougher cell (approximately 69% Cu recovery), showed that 40–60% of pyrite in the finer class of the concentrate (−45μm) was recovered by entrainment. For the coarser classes, pyrite was preferentially recovered to the concentrate by true flotation or association. These results were validated by batch flotation tests, using the same feed material to the rougher flotation circuit.The bubble load and top of froth measurements permitted the evaluation of the pyrite transport mechanisms at industrial scale, providing estimations of pyrite true flotation and entrainment which are relevant for pyrite depression evaluation.

Mineralogical Characterization of Sieved and Un-Sieved Samples

Mineralogical characterization applied to mineral processing is now widespread. The first step for a mineralogi- cal characterization study is usually size fractionation. Preparation of polished sections is done on size fractions to reduce complications in making representative cross sections of particles with large size differences. A sample is commonly fractionated into five or six size intervals. The drawback of this procedure is that it makes liberation studies more expensive, because one sample actually produces five or six sub-samples that need to be studied, i.e. one from each size interval. Thus to reduce cost of liberation studies, it would be desirable to study the un-sized sample. This paper provides a comparative liberation study of a set of samples both using size fractions and using the un-sized samples. The samples studied are the feed, the concentrate and the tails of a lead rougher flotation circuit. The results consistently show significant differences between the sized and the un-sized samples. Nevertheless, the results indicate that un-corrected liberation data from un-sized samples can be used for comparative studies that involve several related samples. Thus, it is possible to improve (or further understand) a concentrator circuit by using mineralogical data from un-sized samples around such circuit.

Hybrid Dynamic Predictive Model for Rougher Flotation

Several dynamic models are able to represent the phenomenological behavior of flotation. On the other side, the interest in developing control strategies for large-scale processes has led to formulate novel methodologies which allow to consider the global performance of a plant, facilitating the design, validation and evaluation of more complex optimizing control strategies.In this work, we first develop a dynamic hybrid model for a flotation rougher circuit with one production line composed by four cell banks in series, connected by controlled valves. Each line is actuated by discrete valves, whose interaction with different bank pulp levels define several operating modes. The model was validated with data from an industrial plant.Subsequently, a hybrid prediction model is obtained by applying identification techniques to the different scenarios, valid for optimizing controller strategies. Our simulation results show that the proposed methodology is suitable for modelling the global behavior of a rougher flotation circuit, representing its nonlinear dynamic and providing a robust predictive models for several operating modes of the circuit.

Control of rougher flotation circuits aided by industrial simulator

Today, large size mechanical flotation cells of 100–300m3 are used in rougher operation in different flotation industrial plants, in Chile and worldwide. However, in spite of the advances in fundamental research and the notable growing in equipment size, with more complete instrumentation systems, there are still a lack of reliable data for industrial flotation modeling and simulation to advance in better control systems design. In this work, a procedure for modeling and simulation of rougher flotation banks, based on operating variables and parameters fitted from industrial data, is presented in condensed form. Recently, a new methodology for describing the industrial flotation, separating the collection and froth zones, has been developed. A non-linear distributed model to simulate rougher flotation circuits was developed, based on measurements of the main operating variables. The simulator was calibrated and tested using experimental data from the rougher operation at El Teniente Division, Codelco-Chile. The simulator was first used to sensitize the effect of changing the froth depth profile on target variables, and then to find out the best profile to reject common disturbances coming into the plant, considering some operating constraints. The flotation circuit simulator was then redesigned to incorporate an expert system following the proposed control algorithm. Several tests were performed by simulation to evaluate the control responses to different correlated series of input variables, such as feed grade, tonnage, particle size distribution, and solid percentage. Control alternatives based on the circuit simulator are discussed.

Modelling and simulation of rougher flotation circuits

At present, large mechanical flotation cells of 100 to 300m3 are used in rougher operation in different industrial flotation plants around the world. However, in spite of the advances in fundamental research and the notable growth in equipment size, there is still a lack of reliable data for industrial flotation modelling and simulation. In this work, a procedure for modelling and simulating rougher flotation banks that is based on operating variables and parameters fitted by empirical data from plant measurements is presented.Recently, a new methodology for describing industrial flotation by separating the collection and froth zones has been developed. This approach consists of using a new apparatus for direct bubble load measurement below the pulp–froth interface in industrial cells. The procedure allows independently estimating the froth recovery and the collection zone recovery. Metallurgical characterisation is developed by plant sampling, mass balance adjustment and applying a short-cut method for kinetic characterisation of flotation banks. In addition, the effect of grinding is characterised using flotation models with distributed flotation rates (for each particle size class) and a low number of parameters.From the experimental data, a non-linear distributed model to simulate rougher flotation banks was developed, which allows describing the industrial operation as a function of the main operating variables (e.g., mass flow rate, solid percentage, grinding degree, pulp level and superficial air rate). The simulator was calibrated and validated using experimental data from the rougher operation at El Teniente Division, Codelco-Chile.The new approach is a practical tool that is useful for flotation operation diagnosis, for robust design and analysis of new control and optimisation strategies and for evaluation of the effect that design and operating variables have upon the collection and froth zone performance in large flotation cells. This methodology can be applied to other flotation operations, such as cleaner and scavenger circuits that use mechanical flotation cells.

Industrial evaluation of a new flotation mechanism for large flotation cells

A new flotation mechanism was evaluated in the rougher flotation circuit at Collahuasi Concentrator Plant in Chile. For this purpose, the flotation mechanism FloatForce® was installed in one of three parallel rougher flotation lines, consisting of nine TankCell-160 (160m3) cells, while the other two lines operate with a conventional OK-mechanism. The evaluation was carried out based on a hydrodynamic and metallurgical characterization of the flotation circuits. Three overall samplings and mass balance adjustments for kinetic flotation characterization were performed. The residence time distribution RTD, of liquid and solids (per size class) was measured in the rougher lines using radioactive tracers. Thus, the effective mean residence time as well as the mixing condition in single flotation cells and along the rougher lines were evaluated. Also, measurements of axial solids segregation (per size class), axial mineral grade profiles, local superficial gas rate, bubble size distribution, bubble load and top of froth grades were performed in cells provided with both flotation mechanisms. The comparison of the rougher flotation operation with the FloatForce® and the conventional OK-mechanism is presented and discussed in this work.

Operation of Rougher Flotation Circuits Aided by Industrial Simulator

AbstractToday, large size mechanical flotation cells of 100 to 300 m3 are used in rougher operation in different flotation industrial plants, in Chile and worldwide. However, in spite of the advances in fundamental research and the notable growing in equipment size, with more complete instrumentation systems, there are still a lack of reliable data for industrial flotation modeling and simulation to advance in better control systems design. In this work, a procedure for modeling and simulation of rougher flotation banks, based on operating variables and parameters fitted from industrial data, is presented. This simulator is used to develop a set of logic rules in order to aid operating decisions.Recently, a new methodology for describing the industrial flotation, separating the collection and froth zones, has been developed. This approach consists of using new apparatus for direct bubble load measurement below the pulp-froth interface in industrial cells, for estimating the froth recovery as well as the collection zone recovery, independently. The bubble surface area flux SB (m2/min/m2) is determined by measuring the bubble size distribution and superficial gas rate. Metallurgical characterization is developed by plant sampling, adjusted mass balances and applying the short-cut method for kinetic characterization. Also, the effect of grinding is characterized using flotation models with distributed flotation rates and low number of parameters (e.g., rectangular distribution).A non linear distributed model to simulate rougher flotation circuits was developed, based on measurements of the main operating variables. The simulator was calibrated and tested using experimental data from the rougher operation at El Teniente Division, Codelco-Chile. The simulator was first used to sensitize the effect of changing the froth depth profile on target variables, and then to find out the best profile to reject common disturbances coming into the plant.