In the some flotation models "Froth Recovery" is introduced as a mechanism whereby mineral particles, which were recovered from the bulk slurry into the froth phase are then not recovered into the concentrate launder (they fall back through the froth phase back into the bulk slurry).
This is sometimes implemented as a multiplier (or modifier) of the recovery equation.
This method of modelling has some problems. First - it can not be reliably measured by practical means, and second there is often no attempt to relate it to anything practical. It introduces a whole new level of confusion on what the word "recovery" means that I am sure politicians would be interested in.
This method cannot be used for operations guidance or for design specification. It is a fudge factor (something which is hard to measure and justify which makes a big difference to the results) which is often used to just knock a pretty bad model into some sort of semblance of fitting some data. "Fines Entrainment" is another method one that is often used badly - more on that later.
So - is there a need for and a practical way of modelling a modification of the rate of recovery of particles to the concentrate (please self administer the "Men In Black Memory Flash Thing" and clear any confused thoughts about recovery being any vague internal thing and get back to the concept of recovery is what ends up in a product stream). Get out your copy of Taggart and open it anywhere and read a few pages. Engineers have used many words such as bypassed material, efficiency, back mixing, channelling and many other concepts to describe and quantify internal inefficiencies or capacity limits before - in many unit operations such as SX, leaching, filtration, grinding etc.
Yes, there is often a need to be able to model the limited capacity of a flotation machine to transport particles from the bulk slurry through the froth phase and to the concentrate launder. There may be a limit to the "Froth Carrying Capacity" and there may be a "Lip Length Limitation". These are both real, practical, measurable (or directly related to something easily measured) parameters that directly relate to machine design and specification, and which can provide operations guidance and plant improvement.
A good flotation model should have froth carrying capacity and lip length capacity as inputs, for each cell that has different configurations, and this combined with the rate by component by size (and maybe liberation extent if you want to go that far as well) should then predict the measured performance (REAL recovery).
If the grade changes (a new pit coming on line?) is the flotation circuit design still appropriate? Do you need more (or less) air? Are the cell configurations appropriate? Would a double launder design be better (maybe you can make one, times are tough). Or is a shift of the froth crowder required (or maybe some tests can be done on one cell to investigate this).
METSIM enables modelling of flotation with any level of layered mechanisms and with capacity limits using machine characteristic dimensions. These can be implemented as User Defined Functions, or expressions in the CtB, or expressions in the recovery by component and size functions. And you can do some "visual programming" using generic unit ops (splitters, SUB's etc.) to customise a model (and others can see from your model diagrams more easily what you have done).
METSIM provides all the practical methods of defining component recoveries in simplified and rigorous forms. You can implement any described model and you are encouraged (by the ease of implementation, testing, and calibration using FBC's) to test alternate methodologies.