Bench-scale simulation of flotation plant performance

"A study has been done of the scale-up of laboratory bench-scale flotation to continuous plant circuit performance. It was found that the impellor speed and the air flow rate influenced the results and in this case extreme values of these operating variables were required in order to reproduce the plant rougher results in the laboratory. Despite the experimental uncertainty introduced by these extreme values, it was demonstrated that the plant performance could be predicted and experimentally reproduced with bench-scale tests.IntroductionThe mineral processing industry relies heavily on batch flotation testwork to design flowsheets, evaluate reagent s and compare various types of ore. The batch tests are generally regarded as being directly comparable within themselves, but quite often they relate poorly or not at all to continuous circuits. Obviously this dilemma generates considerable uncertainty with wide-ranging ramifications. One of the manifestations of this unfortunate situation is the commonly expressed opinion that ""it may work that way in the laboratory, but it will never work that way in the plant"".In a previous publication, a method was described whereby the results of batch bench-scale tests could be used to simulate a locked cycle flotation test series. The work to be described here is a continuation of that investigation with an extension to the simulation of an actual plant circuit. It would have been preferable to progress from laboratory batch data to a continuous pilot plant and then finally to a full-scale plant operation. However, because that possibility did not present itself, an attempt will be made to show that the performance of an existing plant circuit can be successfully simulated from batch data. The success of the procedure can be judged by the agreement among the plant results, a locked cycle test and the predicted results based on batch tests."