A paper presented discussing coal flotation with mechanical flotation machines and their set-up / design for the best results. The reference is as follows;

Murphy, B, Heath, J L and Dube, R, (2014) A commentary on design of mechanical flotation circuits for coal applications, in Proceedings of the Fifteenth Australian Coal Preparation Conference 2014, pp 234-246 (Australian Coal Preparation Society: Broadmedow).

In the recent past in Australia the pneumatic machines have been more favoured for coal applications however I suspect in a number of cases this was more an industry trend then people doing test work and a comprehensive comparison to get the best machine for the job. If a very high upgrade is required then perhaps pneumatic is the best way to achieve this but for most cases I think mechanical technologies (if setup correctly) can achieve the required results. 

Most of the answers to your questions can be found in this paper.

Paper - “New Coal Preparation Plant Circuitry for India Metallurgical Coal”, Saurabh S., Wang Z., Mahapatra A., & Pal S. in Vol.- I, Section (Coal Preparation), Page 374-381 at “Proceedings of the XIII International Seminar on Mineral Processing Technology – 2013 (India)”, Published by CSIR & IMME India in Dec 2013

The CPP circuits that they typically use in North America and Australia are:

Coarse Circuit - 50 x 1mm - DMC (50 x 1.4mm in Australia)

Intermediate Circuit - 1 x 0.15mm - Reflux Classifier or Spiral or WOC/Spiral combination (1.4 x 0.15mm in Reflux Classifier or Spirals in Australia)

Fine Circuit - 0.15mm x 0 - Wemco Flotation or Column Flotation (0.25mm x 0 in Wemco or Column or Jameson Cells)

As far as I know, Chinese CPP circuits are same as those in North America.

The answers to some of the questions that you may not find in that paper are given below:

•Typically 1mm x 0.25mm is processed in Reflux Classifier (RC). The advantage of using RC is that, based on the wash-ability data and SG50, RC can cut at very low SG and at very low Ep. RC along with another density based separation circuit (typically DMC) will help in circuit optimization (getting max metallurgical coal as primary product). Nowadays most of the metallurgical coal companies are upgrading their CPP intermediate circuits (1mm x 0.25mm or 0.15mm) with RCs because they realize the economics.

•The recommended maximum particle size for coal flotation is 300 micron. In Australia flotation is typically done for -0.25mm (-0.15mm in North America). If you read the paper mentioned above, I clearly explained the reasons for floating -0.25mm and NOT -0.5mm.

•For coarse separation at 1mm, screening is primarily used. For fine separation, classifying cyclones are used. But one should make sure that the desired classification cut points on these cyclones are D90s or D95s and NOT D50s.

I hope that the paper referenced above along with some explanation should be able to answer your questions. 

I am looking for the modern trends which must be supported with adequate beneficiation theories. Referring to the comments about using dense medium cyclones up to 1 mm or 1.5 mm bottom size, is that an efficient proposition. It is known that for lower fractions DMCs are less efficient. On the contrary higher top size for RC is not desirable. How to tackle this aspect!

This is not quite the "Floatation" method but I market a range of "Ultrasonic Processors" that will clean and wash Coal of any particle size. It can then be separated via a separation trough. There is no limit to how many tons you can process per hour as our Ultrasonic Troughs are modular and can be sized to your needs. Best of all is the ten year guarantee and simple operation.

I completely agree that in coal flotation, once you reach about 250-300 micron, you want to start looking at alternative technologies other than traditional flotation methods. Dennis Phillips did a significant amount of work back in the late 80's at Virginia Tech which demonstrated the dramatic drop in recovery and increase in variability when the particle size increased above about 250 microns. You can probably get Dennis's thesis through the Virginia Tech website, but ultimately he determined after running 100s of tests, that while you could get high recovery from time to time, it was impossible to maintain a high efficiency once coarse particles were introduced as coarser froths are inherently less stable and more susceptible to plant fluctuations (i.e. chemical or solids loading).

For the +300 micron fraction, you certainly want to look at teeter-bed or spiral technology depending on the quality of the coal and its wash ability. A new technology which may have a place and which combines flotation and gravity separation is the HydroFloat. This was tested in coal and the results presented at the last ACPS conference in Australia. With this device, successfully coal flotation/density separation can be carried out to up to several millimetres. At this point, it has been tested, but not installed on a full scale in coal, but has been successfully installed for treating industrial minerals.

We can certainly attest to the capacity of the HydroFloat cell to float coarse sulphides up to around 2 mm. We used a lab-scale HydroFloat cell in two PhD projects in our long-running AMIRA P260 flotation project. The low turbulence conditions provided by the fluidised bed and minimal froth results in low levels of bubble-particle detachment! Moreover, we were able to float liberated material with much lower critical hydrophobicity, i.e. less reagent. For coal, this bodes well for recovery of partially oxidised material.