The principal objective is to define an optimal control strategy for a flotation column.
Research was carried out in order to define an optimal control strategy for flotation columns. The work consisted of trials on several columns of different sizes, installed in an industrial unit (the Sotiel Concentrator of Almagrera in Spain), and involved the polymetallic sulphide ore from Sotiel.
The tests carried out on the columns involved scanning of numerous operating conditions (air flowrate, feed flowrate, wash water flowrate, froth depth, addition of frother and/or collector and so on).
For each test, the products of the laboratory, pilot and industrial columns were sampled, after a running time long enough to ensure that column equilibrium had been reached (on the order of 0.5 h of stable operation). These products were dried and analysed. Samples of the column feed were also collected from time to time, although this stream was continuously analysed and displayed by the COURIER 300.
With this information a material balance of the columns can be established for every test in order to allow, in the phase of compilation of the results, and analysis of the relationship between operating conditions and column performance.
During the research it appeared necessary to measure and control the pH of the feed of the columns because it varied widely. This parameter may have a significant influence on the metallurgical performance of the column and on froth behaviour.
Important information on the control of the columns was collected during the work resulting in some of the laboratory column operating conditions being changed.
The research has also confirmed that the dependence between the performance of a flotation column and the characteristics of its feed is very marked and that all the feed variations cannot be compensated at column scale (important variations in grade or flow rate for example). Further work is still necessary in order to develop and evaluate more effective control strategies by broadening the simulator to model and simulate complex circuits that include flotation columns.
The work will relate principally to the polymetallic sulphide ore from Sotiel (Spain) but also to that from Moinho (Portugal) and from Chessy (France). It will consist of trials on several columns of different sizes, installed in an industrial unit, and of phases of modelling and of study by simulation.
The programme comprises 8 phases.
Phase 1 will involve instrumentation of the industrial and pilot columns. The instrumentation will mainly consist of:
sensors such as flowmeters (air and water for the bubble generator, wash water, slurry), density meters (slurry), pressure sensors (air and water in the bubble generator, slurry), level meters (height of the water froth interface);
small regulation loop apparatus in order to keep some parameters stable (flowrates, level), proportional, integral and differential(PID) controllers, variphimotors;
and a computer based data acquisition system.
Phase 2 will involve installation of one electrodischarge machining(EDM) pilot column and the BRGM laboratory column in the Sotiel plant in parallel with an industrial column. This will include:
transport and assembly of the columns and of corresponding instrumentation;
link with the online industrial X-ray fluorescence analyzer;
and realization of an equipment to split the industrial slurry flow in order to continuously feed the columns with product having exactly the same characteristics (grade, size distribution, solid concentration).
Phase 3 will involve a measuring campaign in the Sotiel plant, ALMAGRERA, and in the Aljustrel pilot plant. This will include:
collection of the information provided by the sensors;
sampling (feeds, tailings, concentrates);
and analysis of the samples (chemical, particle size distribution, mineralogical analysis).
Phase 4 will involve a compilation of the results. This will include:
modelling the feed to the column (characterization of the fluctuations using a variographic approach);
and dynamic modelling of the operation of the column. The model, which will interlink all the measurable parameters, will have a firm phenomenological base confirmed as far as possible by specific small scale trials, with the intervention if necessary of the relationship resulting from statistical treatment of the data.
Phase 5 will be a comparison between simulations of the various control strategies involving dynamic simulation on the basis of the feed model, with noise of the performance of a column as a function of the sensors assumed to be available, and of the control strategy.
Phase 6 will be improving the control strategy in the Sotiel plant.
Phase 7 will be comparisons between operation at the laboratory pilot plant and industrial scale as observed at the Sotiel plant. involving study of the problems of extrapolation from laboratory to industrial scale. If possible, the development of a procedure for small scale experimentation with a view to industrial evaluation.
Phase 8 will be utilization of the static model. The model in static form will be integrated into the simulator of the Moinho plant. Application of the model to the optimization of flowsheets.
Funding SchemeCSC - Cost-sharing contracts