Flash flotation is a mature application of flotation technology to recover fast-floating liberated particles from within the grinding circuit before they can be ground further
Flash flotation is the rapid flotation of mineral particles to produce a suitable concentrate. Normally, the feed to a flash flotation cell is the cyclone underflow stream. The aim when treating cyclone underflow is to recover liberated fast-floating mineral particles before they present back to the grinding mill for further comminution. These particles would otherwise be ground finer and report to the cyclone overflow as very fine particles, where the recovery may be significantly lower due to slower flotation kinetics (Gaudin, Groh and Henderson, 1931; Morris, 1952).
A flash flotation cell is designed differently to a conventional mechanical flotation cell in that it operates at higher pulp densities and with coarser feed compared to conventional flotation cells. This requirement means that a number of unique features are present in flash flotation cells that are not noted in conventional flotation equipment (Figure 1).
The main difference between a flash flotation cell and a conventional flotation cell is the shape of the tank. A flash flotation cell is cylindrical, with a conical section at the bottom of the tank to assist in the hydraulic transport of very coarse particles (eg mill scats) out of the cell. These larger particles are present in the cyclone underflow stream but are too large for separation by direct flotation. The shape of the tank also allows a flash flotation cell to be self-draining in the event of a feed disruption or power failure, which would otherwise be a major issue when treating a coarse stream such as cyclone underflow. Flash flotation cells also generally have heavy wear liners (not shown in Figure 1 for clarity) in internal areas where coarse and abrasive particles flow. This is to improve wear resistance and reduce maintenance requirements.
Top outlet concept
Another physical feature of a flash flotation cell is that it may have an additional outlet located midway up the tank. This extra outlet (usually termed the ‘top outlet’) is in addition to the normal concentrate and tailing outlets and provides a third product from the cell. This arrangement is shown in Figure 2, with typical process data for each outlet indicated.
The material that reports to the top outlet is normally a low-density stream containing any misreported particles (ie fines) in the flash cell feed. The density of this stream is low because it contains a large proportion of the flash flotation cell feed dilution water, and the particle size distribution is fine because a flash flotation cell also acts as a classifier, hence removing any fine particles misclassified to the cyclone underflow. The top outlet was recently extensively studied by Newcombe (2016).
Mill grinding efficiency is strongly proportional to the mill feed density. As such, the mill feed density is normally kept as high as practicable. As the top outlet removes a large portion of the flash feed dilution water, the bottom outlet slurry density is maximised, which is beneficial for maintaining grinding efficiency. The grinding circuit water balance is maintained by piping the top outlet to the mill discharge hopper, where it replaces some of the cyclone feed dilution water normally added there (see Figure 3). As a flash flotation cell installed without a top outlet will have a bottom outlet (tailings) density that is significantly lower than the cyclone underflow density, the grinding mill will need to be larger to compensate for the lower feed density. As a result, having a top outlet is critical to maintaining grinding circuit efficiency for a given minimum grinding mill size.
Standard practice for the last 10-15 years has been to direct the top outlet stream to the mill discharge hopper to maintain the grinding circuit water balance. However, some operations have recently experimented by directing the top outlet to a secondary grinding duty or even to the main flotation circuit. As the top outlet slurry is low density and only contains fine particles, it often has a particle size distribution similar to the flotation circuit feed, meaning that it can potentially be directed there.
This change improves the overall capacity of the grinding circuit in two ways. Firstly, it directly removes some of the misclassified particles from the cyclones that report to the top outlet. Therefore, it should be possible to improve the grinding circuit throughput for the same product size distribution. The second improvement in directing the top outlet out of the grinding circuit is that there is no longer a restriction on the volume of flash feed dilution water. Adding more flash flotation cell feed dilution water should positively affect flotation performance.
Flash flotation cell location
A flash flotation cell is typically installed so that it treats a significant portion of the cyclone underflow, due to the numerous aforementioned advantages. However, in the 1980s and 1990s it was common practice for unit cells (originally from the 1950s) to operate in nickel flotation plants treating the mill discharge stream (Sceresini, 1982). It should be noted that these were typically installed and enjoyed some success in fairly low throughput (<240 t/h) operations (Severino, Guzman and Castillo, 2007).
Capital and operating costs
In most scenarios, a flash flotation cell treating mill discharge will require an additional large slurry pump, or larger mill foundations, compared to a flash flotation cell treating cyclone underflow. This is because when installed below a cyclone cluster and fed cyclone underflow, gravity is utilised to feed the flash flotation cell and the tail stream(s) drain by gravity back to the mill. Conversely when treating mill discharge, one of these streams (feed or tail) will likely need to be pumped, or the mill itself will need to be significantly higher, to accommodate feeding a flash flotation cell. While it is true that a more powerful pump would be required if the cyclone cluster is higher (to accommodate a flash flotation cell installed beneath), this incremental cost is outweighed by the cost of a complete additional pump if mill discharge is treated. The additional pump will also equate to higher power consumption and maintenance costs for the plant.
The slurry particle size distributions of mill discharge and the cyclone underflow must also be considered. In the mill discharge, there is usually a much larger particle size range (ie more fine particles) compared to the cyclone underflow (which should be essentially deslimed). Fine particles have a significantly higher surface area to volume ratio, meaning that the mill discharge will require more reagents to achieve the same percentage of collector surface coverage compared to the cyclone underflow. Higher reagent consumption equates to higher operating costs. Additionally, floating the coarser cyclone underflow stream reduces the chance of gangue entrainment (Allen, personal communication, 2015), allowing a higher grade flash concentrate.
It is also worth mentioning that a typical mill discharge mass flow is larger than the cyclone underflow mass flow. This means that the equipment required to treat the same proportion of each stream will be larger for a mill discharge installation, which equates to significantly higher capital and operating costs (larger drive, larger wear parts). For example, in the 1990s, a single 1200 t/h flash flotation cell treating cyclone underflow replaced 36 small unit cells treating mill discharge in a nickel concentrator in Western Australia. The main benefits reported by the site were significantly reduced operating costs and improved metallurgical performance.
As hydrocyclones separate by centrifugal force rather than purely based on size, higher specific gravity mineral particles of intermediate and coarse sizes will primarily report to the cyclone underflow. The primary purpose of a flash flotation cell is to remove liberated and fast-floating valuable mineral particles before they are overground to a finer particle size and then report to the cyclone overflow where they are more difficult to recover in a downstream conventional flotation circuit (Gaudin, Groh and Henderson, 1931; Morris, 1952).
By placing a flash flotation cell in the cyclone underflow, particles directly at risk of being overground are targeted for recovery. From a metallurgical standpoint, the cyclone underflow is the most efficient stream to install flash flotation capacity to have the most significant effect on reducing overgrinding of valuable mineral particles.
If treating the mill discharge, many of the fines that have been generated in the mill will be recovered by the flash flotation cell. Conversely, coarser mineral particles in the mill discharge will not be recovered as effectively in a flash flotation cell due to competition (for reagents, bubble and froth surface area) from fine particles. This means that the primary function of the flash flotation cell is diminished by operating on the mill discharge stream. Also note that a lot of the aforementioned fines in the mill discharge would be recovered in a conventional downstream flotation circuit anyway, even if the flash flotation cell was not installed.
For a flash flotation cell to operate well, it is essential to supply it with a consistent volumetric feed flow, which is achieved by using a well-designed feed box. From the author’s experience, having a poorly designed or maintained feed box is normally the main reason for a flash flotation cell performing poorly.
The feed box needs to representatively split the incoming slurry to the flash flotation cell and modulate the amount (volume and mass) of slurry being fed to the cell. Feed dilution is normally conducted after the feed box in the pipe to the cell. The feed box must also direct any remaining slurry back to the process (usually to ball mill feed if processing cyclone underflow). A simplified model of a flash cell feed box can be seen in Figure 4.
Flash flotation cells treating mill discharge
Despite the aforementioned points, there are still cases where flash flotation cells are effectively operating on mill discharge, especially in South America. These are normally installed in smaller plants with the flash flotation cell treating up to 240 t/h of mill discharge. At this size, the extra pumping requirement and operating costs to feed a flash flotation cell with mill discharge is not as significant as larger concentrators. It is worth noting that the largest currently available flash flotation cells can treat up to 2400 t/h, and flash flotation cells designed to treat 500 t/h and higher throughputs are all installed at the cyclone underflow.
Flash flotation testing and scale-up
Given that a flash flotation cell is normally fed cyclone underflow while operating in the grinding circuit, and has a measurable effect on the mineral recirculating load while at full scale, it is not straightforward to undertake laboratory testing or modelling to accurately predict full-scale flash flotation performance. A batch laboratory test will not be representative of the equilibrium full-scale plant recirculating load. Additionally, the limitations on laboratory flotation machines means that cyclone underflow material cannot be directly tested. While some authors have made progress in this area (eg Lamberg and Bernal (2009) and Newcombe, Wightman and Bradshaw (2014)), the validity of their findings is not further discussed in this article. Instead, a pragmatic approach of reviewing ore mineralogical data, together with laboratory flotation tests, is often the best way to qualitatively predict full-scale flash flotation cell performance.
When evaluating if flash flotation is worthy of consideration for a greenfield project, the first evaluation should be on the ore to be treated. For example, the main minerals (valuable and gangue) present, the respective hardness of the minerals, the degree of textural complexity and the size liberation profile are all important factors to consider. As flash flotation is the rapid flotation of minerals from within the grinding circuit, it relies on fast-floating minerals to be present with sufficient liberation to achieve the required concentrate grade. An initial ‘reality check’ on the viability of flash flotation often leads to flash flotation being identified as viable for ores with low to moderate textural complexity, with soft valuable minerals hosted within harder gangue matrices. It is no coincidence that the majority of existing flash flotation references are installed in plants processing these types of ores.
Once flash flotation has passed initial viability scrutiny, batch-scale laboratory tests can be undertaken to provide qualitative results on the expected flash performance at full scale. Most large commercial flotation laboratories and reputable equipment vendors have procedures for these tests. The flotation tests are usually conducted on a screened slurry from a typical grind establishment test to measure the flotation recovery kinetics and concentrate grade.
A more robust evaluation for a greenfield project is by undertaking pilot plant testing. In this scenario the pilot plant has a true recirculating load in the grinding circuit that should be reflective of the full-scale plant. The beauty of this method is a pilot-scale flash flotation cell can be inserted or removed from the process to determine its effect.
The true effect of flash flotation can be ascertained by considering the cyclone overflow assay with and without the flash flotation cell. This is the most robust way to test the effect of flash flotation for a greenfield project and has been used successfully to verify the business case for inclusion of flash flotation at previous greenfield projects (Lamberg and Bernal, 2009). The downside of this approach is the cost and time associated with organising and conducting a pilot test program.
Some equipment vendors and commercial laboratories have developed in-house testing procedures to qualitatively evaluate the effects of flash flotation on an existing operating plant. The test procedure is similar to the greenfield test procedure, but involves collecting a plant sample and undertaking a ‘hot float’ in the laboratory at site. The results of the test provide a qualitative prediction of incorporating a flash flotation cell.
The utilisation of pilot-scale flash flotation cells in a brownfield environment is usually not recommended, as these units are typically very small relative to the cyclone underflow and particle sizes in the plant. This results in a high likelihood of the test unit becoming blocked. Additionally, only a small portion of the recirculating load can be treated through a pilot-sized flash flotation cell, so it is impossible to predict the effect of a full-scale unit on the overall grinding circuit recirculating load.
Selection of flash cell size
Once flotation tests have been completed and it has been decided that flash flotation is viable, the next stage is to assess the size of the cell(s) that will be appropriate for the mineral processing circuit under evaluation. This is largely determined by the circulating load in the mill and the mill water balance. To have a measurable effect on mineral overgrinding, the flash flotation cell should typically treat between 40 and 60 per cent of the cyclone underflow stream, and the cell size and configuration (single or dual outlet) selected to match the duty. The mass balance around the flash flotation cell is developed using typical figures (as seen in Figure 2) with design margins applied.
In greenfield applications where the plant designer is unsure if flash flotation will work, or there is insufficient sample or an unwillingness to test, it is recommended that a preliminary selection be undertaken based on the grinding area mass balance. Using this cell size, a suitable space can be left in the milling area between the cyclones and the ball mill to allow easy retrofitting of a flash cell, if further testing and a valid business case shows flash flotation to be viable.
Flash flotation is a mature application of flotation technology to recover fast-floating liberated particles from within the grinding circuit before they can be ground further. The top outlet allows the grinding circuit efficiency to be maximised and provides an additional opportunity to increase grinding circuit capacity if the top outlet is directed outside the grinding circuit. The suitability of flash flotation depends strongly on the ore mineralogy, and testing should be done to confirm whether flash flotation should be included in the flowsheet. Sites with flash flotation cells installed commonly report overall recovery improvements of several percentage points, plus other operational benefits.
Gaudin A M, Groh J O and Henderson, H B, 1931. Effect of particle size on flotation, Technical Publication No 414, Class B, Milling and Concentration, No 35, 23 p, The American Institute of Mining and Metallurgical Engineers.
Heath J, 2016. Flash flotation – an introduction [online], AusIMM Bulletin, June 2016. Available from: www.ausimmbulletin.com/feature/flash-flotation-an-introduction/
Lamberg P and Bernal L, 2009. Modelling and simulation of Skimair® flash flotation as a part of grinding-flotation circuit – case study Esperanza, Chile, in Proceedings 41st Annual Meeting of the Canadian Mineral Processors (Canadian Institute of Mining and Metallurgy: Ottawa).
Morris T M, 1952. Measurement and evaluation of the rate of flotation as a function of particle size, Mining Engineering, 4(8):794–798.
Newcombe B, 2016. Flotation within the grinding circuit – understanding and application of the concepts of unit cell operation, Mineral Processing and Extractive Metallurgy Review.
Newcombe B, Wightman E and Bradshaw D, 2014. A phenominalogical model for an industrial flash flotation cell, Minerals Engineering, 64:51–62.
Sceresini B J, 1982. The development and integration of a carbon-in-pulp gold circuit within the Kambalda nickel concentrator, in Proceedings Symposium on Extractive Metallurgy, pp 43–50 (The Australasian Institute of Mining and Metallurgy: Melbourne).
Severino F B, Guzman G and Castillo E C S, 2007. Flotacion en la descarga de la molienda primaria: una alternative para mejorar la performance metaluggica en la compania Minera Raura (Flotation on primary mill discharge: an alternative to improve metallurgical performance at the Raura Mine), in Proceedings XXVII Convencion Minera (Peru).