Selecting Flotation Cells: How Many and What Size?
Lower ore grades have driven the industry to larger throughputs and thus larger equipment over
the last 20 years. Flotation cell sizes of up to 200m3 are now available with 150m3 cell
installations being common.
The question for many circuit designers is, do the old "rules of thumb" still apply when selecting
the number and size of cells required? The answer is part yes and part no.
Rule 1. "I need at least 8 cells in a rougher/scavenger train to prevent short-circuiting"
The theory of short-circuiting is not changed by the size of the cell and thus some argue that
this rule must always apply. In practice, it appears that factors other than short-circuiting are
more critical to the performance of a float circuit where 5 or more cells are employed. Issues
such as scale-up of residence time from laboratory or pilot plant tests, froth handling, number of
control stages in the circuit, stream recycling, reagent selection and addition and stability of feed
all have a measurable impact on circuit performance. The difference in performance between 5
cells and 8 cells with equivalent residence time cannot be detected in most cases due to noise
from other flotation factors. When consideration is given to capital and installation cost, plant
footprint and operability the economic solution is generally less than 8 cells.
Recommendation: Use at least 5 cells for rougher/scavenger applications.
Rule 2. "The larger the cell, the less cells in each bank"
This rule is relatively new and relates purely to the control of cell banks, in particular level
control. When cells were 3 or 8m3 and literally hundreds of cells were used, it was not
uncommon to see up to 6 cells in a bank. As the size of cells has increased, the number of cells
in each bank has reduced. This occurs for two reasons; firstly each bank is a single control stage
with its own level, thus as the cells get bigger and the number of cells used reduce, it is
necessary to reduce the number of cells in each bank to provide sufficient control points in the
circuit. Second, and perhaps more importantly, control of large banks of large cells is extremely
difficult. Thus it is now common to see 100 or 150m3 cells installed as a series of "unit" reactors.
Recommendation: "Less is Best" Maximum number of cells per bank (Refer to table below )
Cells Size OK150TC OK100TC OK50TC OK38 OK16 OK8 OK3
No. 2 2 2 3 4 5 6
Rule 3. "Froth lip loading should not exceed 1.5 tonnes per linear metre per hour "
This is definitely a rule that has been superseded by the changing design of flotation cells.
The new thinking has expanded to also consider the froth-carrying rate and emphasises the
importance of selecting flotation cells with a "froth-handling strategy" in mind. In many cases,
the use of fewer, larger cells has been of benefit as it has allowed the designer to have less froth
surface area and thus more stable froth when treating low-grade ores. These large cells can also
create a trap in situations where the throughput rate is high and the grade is also high. In this
situation, it is possible to have too little froth area, making froth removal the rate-controlling
step in the flotation process. An extreme example of this occurs where the largest available cells
are not suitable due to froth area limitations, yet may satisfy the old lip length rule. Thus
selection of the number of cells and the froth crowder/launder combination is a critical
component of circuit design and consideration of the froth carry rate and froth lip loading is
imperative.
Recommendation: Standard circuit design should use froth carry rates in the ranges below.
Always check the 1.5t/m/hr lip loading is not exceeded.
Duty Rougher Scavenger Cleaner
Carry Rate (t/m2/hr) 0.8-1.5 0.3-0.8 1.0-2.0
Rule 4. "I need at least 4 control stages in the flotation cell train to provide circuit flexibility"
A control stage is defined as a cell or cells which operate with a common level control. Many
people use the term "cell bank" to define this group of cells. Each control stage can be
considered a "unit" reactor from which a tail and concentrate can be extracted for subsequent
processing within the plant. Whilst it is possible to take intermediate concentrates from a bank
of cells, it is difficult to control the grade of this concentrate without impacting the remaining
cells in the bank as they are all linked to the same level control. Linking cells to the same level
control also creates a relationship between the air addition rates for each cell. To operate a
smooth circuit, it is best to balance the air evenly amongst a bank of cells. This is especially true
for larger cells where a difference in air rate between two cells can easily create a difference in
pulp level that is greater than the target froth depth creating a "pulping" situation in one of the
cells. A rougher/scavenger circuit with 4 stages allows the plant operator the capacity to
simultaneously operate the first stage to create high grade concentrate, the second stage to
create intermediate concentrate, the third stage to create low grade concentrate and the final
stage to control the tail grade and make a concentrate for regrind. The high, intermediate and
low-grade concentrates can be directed into the cleaner circuit as appropriate for the ore in
question.
Recommendation: Use at least 4 stages or banks when designing a rougher/scavenger circuit.
The discussion above indicates that cell selection and circuit design has become a more complex
issue with the advent of larger cells. Large surface areas and fewer units make designing for
successful froth handling and circuit control critical. The issue of froth handling must be
considered during the initial test work. Circuit designers need information on froth removal rate
versus flotation time and observations of froth stability if they are to select the correct launder
options.
The four "new" rules of thumb:
1. Use at least 5 cells for rougher/scavenger applications.
2. Less is Best. Check the maximum number of cells per bank.
3. Check froth carry rate and lip loading meet design guidelines.
4. Use at least 4 stages or banks when designing a rougher scavenger circuit.
These notes were contributed by Peter Bourke, Manager, Flotation Cells, for Outokumpu
Technology Pty Ltd. He has more than a decade's experience at the leading edge of the
development of flotation cells, having been involved successively in the installation of
Outokumpu's first 100, 150 and 200 cubic metre float cells
