The metallurgical performance of a flotation machine is largely defined by phenomena occurring in the froth zone. The water content in the froth affects recovery by influencing froth stability and mobility and, at the same time, reduces grade by mechanical entrainment of gangue particles in the overflow water. Efficient operation requires a compromise between the water carried by bubbles from the collection zone and that which overflows. It is believed that the most suitable operating strategy could be based on the measurement of froth water content, as a strong correlation with water overflow is anticipated. This work reports the testing results of an in situ electrical conductivity sensor continuously measuring the froth zone water content in a laboratory-scale flotation column. The test program included simultaneous measurement of froth conductivity and water overflow rates for changes in gas flow rate and frother concentration. The results show a stronger dependence of the measured top-of-froth water content on frother concentration than on the gas flow rate. A relatively linear trend was shown between top-of-froth water content and water overflow rate for a given air rate and frother.

The recovery of fine hydrophilic particles in flotation is related to the recovery of water. Water is carried by bubbles into and through the froth. The amount transported (entrained) depends on bubble size, gas rate and the subject of this paper, frother type. To isolate the effect of frother type from that of bubble size, gas holdup is used as the correlating variable. Measurements were made using a bubble column (no solids) operated continuously with the overflow recycled. Using an automatically controlled 7 cm foam depth, the overflow rate of water is used as a measure of 'water carrying rate'. It is shown that the common frothers can be grouped into four families based on the water carrying rate-gas holdup relationship.

Minerals flotation is a widely used process to produce base metal concentrates through the selective capture of particles on the surface of bubbles. The process performance depends on the size distribution of the bubble population generated by gas dispersion, which is characterized by three variables: superficial gas velocity, gas holdup, and bubble size. A literature review revealed that current instrumentation cannot provide reliable on-line measurement of these variables except for gas velocity. There are some promising alternatives for gas holdup, but bubble size measurement will continue to be unavailable. The use of a model that integrates the three gas dispersion variables makes possible the calculation of one of the variables when knowing the values of the other two. Modelling bubble flow has been pursued using two approaches: determining the bubble terminal velocity reduction by the presence of other bubbles and regarding the bubble swarm as a packed bed through which a fluid is allowed to flow. Models based on these approaches (drift-flux and Molerus, respectively) were found in the literature and used to assess their prediction ability. Gas holdup predictions for known values of gas velocity and bubble size showed similar trends for both models; Molerus results were always higher than those obtained with the drift-flux model, and the difference between both predictions increased with bubble size and gas velocity. A relationship between bubble surface area flux and gas holdup was explored (a single line was expected); Molerus values showed a noticeable effect of bubble size, while drift-flux results showed a minor effect of bubble size only for gas holdups below 10%. Model accuracy was established using a data set collected to characterize frother roles in flotation for six commercial frothers, which included values of the three gas dispersion variables measured simultaneously and reported at the same conditions. The results indicate that bubble size predictions obtained from Molerus model are closer to the measured values than those obtained from the drift-flux model. Drift-flux predictions systematically underestimated the measured bubble size, with relative errors between 10 and 30%, while Molerus predictions showed values around the measured size with relative errors not larger than 15% (and in most cases below 10%). The accuracy of the Molerus predictions is acceptable for applications in the control of individual cells and flotation circuits operation. Further testing to assess the performance of Molerus model with data collected in lab and industrial mechanical cells and columns, where conditions in the test volume may not be stable or homogeneous, is recommended.