Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Hydrocyclone ISWT

Download as ppt, pdf, or txt
Download as ppt, pdf, or txt
You are on page 1of 69
At a glance
Powered by AI
The key takeaways are the history of hydrocyclone development from 1891 to 1945 and their main industrial applications and uses such as clarification, dewatering, washing, and separation.

The history of hydrocyclone development started in 1891 with the first patent by Bretney. The first major industrial application was in a US phosphate plant in 1914. Later applications included heavy medium introduction for coal washing in 1939 and dense medium cyclones for coal in 1945.

Hydrocyclones are used because they are extremely versatile, simple and cheap to install with little maintenance required. They also have low space requirements and high throughput.

HYDROCYCLONE

History of Development
First patent of cyclone- 1891 (Bretney)
First application in an American phosphate
plant in 1914
Major Industrial Applications 1935
Heavy medium introduced for coal Washing
1939 (Dutch State Mines)
Dense Medium Cyclone for Coal 1945
(Driessen)

Why to use hydrocyclones???

1. Extremely versatile in application clarification, dewatering,


washing, separation of two immiscible liquids, degassing etc.
2. Simple, cheap, easy installation, little maintenance and support
structures, easy operation.
3. Less space requirement and high throughput.

Principles of Operation

Basic Features of a Hydrocyclone

SEPARATION THEORIES

1. SIMPLE FUNDAMENTAL THEORIES


o EQUILIBRIUM ORBIT THEORY
o RESIDENCE TIME THEORY

EQUILIBRIUM ORBIT THEORY


Proposed BY - Driessen (1951) and Criner (1950)
According to this concept particles of a given size attain an equilibrium
radial orbit position in the cyclone where terminal-settling velocity is
equal to the radial velocity of the liquid.
Therefore, the finer particles will find equilibrium on small radius where
the flow is moving upward while the coarser will stay on large radius
and finish in the downward flow.

Limitations
Takes no account of the residence time of the particles
All particles may not be able to attain equilibrium orbits within their
residence time.
No consideration of turbulence.
However, gives reasonable prediction of cyclone performance at
low feed solids concentration.

RESIDENCE TIME THEORY


Proposed by, Rietma (1961)
1. Considers whether a particle will reach the cyclone wall in the residence
time available.
2. Assumes homogeneous distribution of all particles across the feed inlet.
3. This theory doesnt take into account the radial flow, effect of hindered
settling at higher solids concentrations and no influence of turbulence.
4. Many other versions ( Trawinski, 1969; Kutepov et al., 1977 and 1978) of
this theory are also available.
5. Despite its different approach and assumptions, often leads to correlations
that are often in very similar forms as those from the equilibrium orbit theory.

CROWDING THEORY

Fahlstrom (1960), proposed that the cut size was primarily a


function of the capacity of the underflow orifice or spigot
opening and of the particle size analysis of the feed.
Anyone involved in cyclone testing would have observed the
strong influence of underflow orifice on the cyclone cut size.
But, the development of analytical expression to describe
crowding theory is difficult.

FLUID FLOW MODEL BASED ON NAVIER-STOKES


EQUATION
The importance of the NavierStokes equation for the simulation of
fluid flow in cyclone was realized from the outset of research in
1950s.
Over the years, a large number of studies on cyclones have made
use of this complex, nonlinear, momentum balance equation.
In cyclone studies, the NavierStokes equations have been
expressed in primitive form as well as in terms of stream function,
or stream function and vorticity.
In various studies, the flow has been considered as both turbulent
and laminar flow at high Reynolds number.
The flow has been treated as both Newtonian and non-Newtonian,
and the solution methods have varied between purely analytical and
fully computational techniques.

TYPES OF FLUID FLOW MODEL BASED ON NAVIERSTOKES EQUATION


ANALYTICAL FLOW MODEL, NUMERICAL FLOW MODEL,
TURBULENT TWO PHASE FLOW MODEL
AND MULTIPHASE
MODELING OF THE PARTICLE DISTRIBUTION
ANALYTICAL FLOW MODEL
The analytical flow models are based on mathematical solutions of the
basic flow mechanics.
Bloor and Ingham (1973, 1975a, 1977, 1987) assumed the fluid to enter
the cyclone with uniform momentum and used the form of vorticity
distribution.
However, the momentum cannot be distributed uniformly in the
entering fluid.
Useful analytical models towards comprehensive description of
particle separation mechanism in a hydrocyclone are yet to be
developed.

NUMERICAL FLOW MODEL


Numerical studies of hydrocyclones effectively started in the 1980s to
predict the flow field with considerably more accuracy and beyond the
capabilities of any possible analytical solution.
The first successful numerical simulation of the flow in hydrocyclones
was described in a series of papers by Pericleous and Rhodes (1984,
1986 and 1987). These authors used the commercial flow solver
PHOENICS.
The advent of powerful computers, as well as developments in
numerical techniques, has made it possible to think of a complete
numerical solution of NavierStokes equation either in its primitive form
or the stream function and vorticity approach.
However, CFD modeling of cyclone separators is a non-trivial
computational challenge. The swirling turbulence, the presence of solids
and the air core in the central zone make the flow in hydrocyclones
complex.

The problem is made worse because the detailed experimental


information needed to validate CFD models is difficult to obtain. CFD
models should be validated by slurry velocity and local concentration
of solids measurements in the flow field.
Single-phase flow laser Doppler anemometry (LDA) has been used to
obtain this information. However, the presence of particulate minerals
renders cyclone feed slurries opaque and LDA cannot be used unless
experimental slurry of glass beads with a fluid of the same refractive
index is prepared. Such systems exist, but the fluid is invariably toxic
and corrosive.
Gamma ray tomography is being used to obtain air core information
in a DMC at the JKMRC University of Queensland, Australia, but the
technique requires considerable care because of the radiation hazard.
He et al (1999) demonstrated that full three-dimensional modeling is
essential in order to accurately model the cyclone flow-field, as a
result of its inherent axial asymmetry. Clearly, the two-dimensional
approach cannot capture axial flow-field asymmetries.

TURBULENT TWO PHASE FLOW MODEL


Turbulence models for prediction of very high swirl flow contain empirical
constants and are still being developed. The computational cost of such
simulations is also very high. Therefore, predictions of laminar flows have been
explored to a greater extent for this research. Such an analysis can give a reliable
insight into the behavior of high viscosity slurry flow.
Most authors note that the standard k-e turbulence model is unsuitable for the
highly swirling flows that exist in cyclone separators. This has been dealt with in a
number of ways.

Hsieh (1988) used a modified mixing length model for the turbulence with a
different mixing length for each component of the momentum equation.
(Dampening problem ??)
Others have modified the dissipation equation in the standard k-e model with a
swirl correction. This modified k-e model is implemented in Fluent as an
adjustable constant in the RNG (Re- Normalized Group) turbulence model.
However, it was observed that the swirl constant needed to be increased to match
experimental behavior, but a point came where further increase made the
simulations unstable.
The application of RSM and LES to flows with particles is an active research
area.

MULTIPHASE MODELING OF THE PARTICLE DISTRIBUTION


Flows in cyclone separators are multiphase, as indeed are most flows in
mineral processing applications.
The flow in a hydrocyclone consists of solid medium that are dispersed
throughout the water. In addition, there is the air core. So this is a three-phase
flow problem.
Multiphase flows can be solved by a number of CFD techniques. These
include the Eulerian-Eulerian multiphase approach, the Eulerian-Lagrangian
approach, and the direct numerical approach.
Very few authors have attempted to solve the full multiphase flow in cyclone
with a Eulerian multiphase flow simulation. However, the models rely heavily
on many assumptions, which ultimately make the predictions differ vastly from
the actual plant data.
The Eulerian- Lagrangian approach provides a direct description of the
particulate flow by tracking the motion of individual particles. This approach is
only useful for dilute suspensions as it doesnt involve consideration of
collisions between particles and change in slurry viscosity.
The third type of approach for simulation of the motion of both the fluid and
the solid particles is termed direct numerical simulation which is not yet
explored in cyclones.

THE DIMENSIONLESS GROUP MODEL


The advantage of this approach is that it maintains a link to mainstream fluid
mechanics.
Reynolds number (Re; effectively the ratio of inertia forces to the viscous force),
Froude number (Fr; the ratio of inertia over gravity) and Euler number (Eu; the
ratio of pressure over inertia), have been used in many models of hydrocyclones.
Nageshwararao Model and Firth model are worth mentioning.
Nageshwararao (1978) considered the following factors to describe cyclone
performance:
The Euler number
The dimensionless cut size
Recovery of water to underflow, Rf
Volumetric recovery of feed slurry to underflow, Rv
This model is presently being used in the JKSIMMET software for hydrocyclone
performance optimization. However, this model doesnt consider feed material
characteristic (size and density distributions) as independent variables.

Firth model (2003) is based on dimensionless analysis


but it considers the cyclone as a slurry subdivision
process using fluid dynamic means. The issues are:
The pressure/flow relationship which provides the energy
for the separation process.
The volumetric flow rate to the oversize flow stream
(Qs).
The volumetric proportion of solids reporting through the
spigot (s).
The modeling methodology is based on multi-linear
regression analysis of selected variables.
Although the approach appears to be practical but yet to
be validated with plant data.

THE REGRESSION MODEL


This is the group of the so-called mathematical models, which are in fact
based almost entirely on regression analysis of test data.
Lynch and Rao and Plitt models are the two general-purpose
hydrocyclone models most widely used for industrial scale simulation
studies of comminution and classification circuits.
In commercial simulation software the Plitt model (MODSIM) has been
included since the early 1980s and have been used with very few changes
since then.
Both the models are purely empirical and dimensionally inconsistent.
Plitt Model
D
34.4 u
Do
S

d 50 c

3.31

0.71

0.54

. Du Do

0.46

0.6

Di Do

1.21

exp 0.08Cv

h 0.38 Q 0.45 p l

0.5

2 0.36

P 0.24 .D1.11

50.5 Dc
Du

. L l

. exp 0.54c

It is necessary to change the


constants to fit the predicted
results to the experimental
data, means rely heavily on
new experimental data.

MECHANISM OF SEPARATION

Classical Theory
Basic Flow Profile

Areas of Application

Areas of application can be conveniently


divided into the different phase separations
as follows:
Solid from liquid
Solid from solid
Liquid from liquid
Gas from liquid

Solid from liquid

Usually referred to as a cyclone thickener


Most usefully applied in the 5 to 200 size range
Larger particles than 200 settling tanks / stationary or vibrating
screens are attractive
Smaller particles than 5 can be handled provided s- f is high
enough (settling velocity!)
Two extreme cases clarification and dewatering
Clarification separation of solid from liquid
Dewatering separation of liquid from solid
In both the cases the extent of separation depends on the feed size
distribution, the feed slurry characteristics, the design and operating
variables.
A cyclone which is long in relation to its diameter is preferred.
The cyclone in combination with other methods of separation is
generally used as a thickener.

Solid from solid


Separation is based on the differences between size,
density or shapes.
Separation based on size is termed as classification.
Low efficiency of particles below the cut point and high
efficiency above it are the desired objectives.
Removal of fines from coarse particles is termed as
desliming and is of wide interest in the treatment of
mined products.
Removal of coarse particles from fines is termed as
degritting removal of oversize foreign matter from
many chemical products is an example.

Solid from solid - continued

Cyclones used for particle separation based on density differences


usually termed as washers coal washing is an example.

Separation based on the differences in particle shape is also


possible separation of broken palm kernels from shells is an
example.
Important Comment

However, in many situations the particles differ in size distribution,


density distribution and shape distribution. Actually, the separation
occurs based mainly on the differences in particle mass and hydrodynamically on Reynolds number differences. Irregular particle
shape is very difficult to define quantitatively and its orientation in a
fluid medium is also difficult to predict.

Liquid from liquid


Example separation of oil and water.
Limitations
1.
Not possible to obtain complete separation in one stage.
2.
Difficult to identify an optimum flow rate for maximum separating
efficiency as due to the shearing action emulsification of the two
phases may occur.

Gas from liquid


The use of a cyclone for this purpose is perfectly feasible. However, it
requires departure from conventional cyclone design due to the
necessity for non-axial removal of the liquid or destruction of the air core
at the point of an axial exit so as to avoid re-entrainment of gas with
discharging liquid.
We, the mineral processing people are not interested to know
much more on the above two areas of applications of
hydrocyclones.

Performance Evaluation
The commonest method of representing cyclone efficiency is by a partition curve.

Ideal
Real

d50

The sharpness of the cut depends on the slope of the central section of the
partition curve; the closer to vertical is the slope, the higher is the efficiency.
Imperfection (I) = (d75 d25)/2d50

Many mathematical models include the term corrected d50 taken from the
corrected classification curve.
Assumption is that the solids of all sizes are entrained in the coarse
product liquid by short-circuiting in direct proportion to the fraction of feed
water reporting to the underflow.
Example: Suppose the feed contains 16 t/h of material of a certain size, and 12
t/h report to the underflow, then the percentage of this size reporting to the
underflow, and plotted on the normal partition curve, is 75%.
However, if, say, 25% of the feed water reports to the underflow, then 25% of the
feed material will short-circuit with it; therefore, 4 t/h of the size fraction will report
to the underflow due to short-circuit and only 8 t/h will report to the underflow due
to classification.
The corrected recovery of the size fraction is thus
100 x (12-4)/(16-4) = 67%.

Uncorrected and corrected classification curves

Reduced efficiency curve


Lynch (1977) suggests that this curve can be used to derive the actual
performance as it is independent of diameter, outlet dimensions or
operating conditions. (May not be always true).

Design variables

Cyclone Diameter
Aperture diameters
Body dimensions (length of cylindrical section, cone angle)
Feed inlet geometry
Interior surface finish
Materials of construction

Cyclone Diameter
The smaller is the diameter the smaller the d50, that is, the smaller is the
size of particle which can be separated.
Although larger particles can be separated in a large diameter cyclone
theoretically but the difficulties in handling large particles in suspension and the
severe abrasion within a cyclone that would result.
It is unusual to find cyclones in use with a diameter greater than 24
inch though diameters up to 48 inch are known.
At the other extreme a limit is also met due to difficulties in the
construction and manifolding together of many very small diameter
cyclones to give capacities of practical interest.
The practical limit appears to be around 10 mm.
The relationship between flow rate and cyclone diameter at a fixed
pressure drop is Q Dcz (z varies between 1.8 to 2.0)
The relationship between d50 and cyclone diameter at a fixed pressure
drop is d50 Dcx (x varies between 0.41 to 0.5).

Aperture Diameters
The principal design variables which control cyclone performance are the
three aperture sizes; feed, Di; overflow, Do; and underflow, Du.
Di = Dc/6 to Dc/7 generally used for the design purposes but there is no
real evidence of optimum diameters.
Do > Di is the general perception. However, designs vary from Do = Di to 1.6
Di.
Plant cyclones are usually equipped with a variable Du and for this reason
the size of this aperture need not play an important part in design
considerations.
Actually, Du should always be matched to the application. Normally D u
varies in the range Dc/10 to Dc/5.
Vortex finder length is also an important design parameter and normally
varies in the range of Dc/2 to Dc/3.
Vortex finder wall thickness should not exceed one-tenth of the internal
diameter.

Body Dimensions
It has been conventional to construct cyclones with a cylindrical and
a conical section.
Small diameter cyclones are also made without a cylindrical section.
Lengths of cylindrical sections in practice ranges generally from 2/3
Dc to 2 Dc.
Cone angles generally vary from 90 to 200.
The desirable cone angle or length is one of the least well defined
aspects of design.
Increase in cone angle has the effect of increasing the air core
diameter and consequently decreasing the volume of underflow for a
fixed underflow diameter.

Feed Inlet Geometry


Feed inlets can vary in shape, type and position and again there is an absence of
useful data.

Standard tangential entry

Vaned entry

Involute entry

Venturi entry

Size varies from 0.12 to 0.28 Dc.

Feed Inlet Geometry contd.

Rectangular opening with its long side parallel to the cyclone axis is
slightly better than a circular opening of equivalent area.
A length to breadth ratio of 2 to 1 is usually suggested.
Although tangential entry is almost standard entry type, Krebs claim
that the involute entry permits an increase of 25 per cent in capacity
for the same performance.
The Vaned and Venturi types are the most difficult to construct and
expensive as well and would have to show considerable advantages
before their adoption.
Feed inlet position should be level with the roof of the cyclone to
eliminate the dead space.

Interior Surface Finish

Cyclones are marketed with many qualities of interior surface finish from
natural rubber to glazed porcelain.
A rough interior wall could decrease efficiency through the promotion of
turbulence near the wall.
A rough wall gives a stronger downward flow and a greater proportion of
the feed passes to the underflow.
Surface roughness is of greater significance in a small dimensioned
unit.

Materials of Construction
Erosion is recognized as one of the major problems in cyclone operation and the
likely points of attack are:
1. The apex of the cone. Underflow apertures are enlarged and wear takes place
generally in the bottom third of the cone. Erosion patterns normally exhibit a spiral
form as if solids flow along a preferred path rather than over the entire cone wall.
2. The root of the vortex finder. This is a point of severe attack especially when the
feed apertures are too large and permit direct impingement of feed particles on
the vortex finder wall.
3. The wall, adjacent to feed entry. This may occur due to the direct impingement of
particles from a badly designed feed entry.
4. Localized attack in the cyclone body. This is due to the collection of an oversize
particle which is held by centrifugal force against the wall when the upward and
downward forces on the particles are in balance.
5. To combat erosion cyclones have been made from many different materials
varying from rubber, porcelain and bakelite through cast iron to titanium.
6. Polyurethane is having wider application as a wear resistant lining giving a
fivefold increase in life relative to stainless steel. Ceramic apex sections are more
generally applied in small diameter cyclones.

Operating Variables

Feed flow rate


Feed pressure or pressure drop
Solids concentration
Solids size and shape
Solids density
Liquid medium density
Liquid medium viscosity

Feed flow rate


Efficiency increases as flow rate increases in accordance with the relationship:
d50 Q-x (x varies from 0.53 to 0.64)
Similarly increase in flow rate increases pressure drop in accordance with the
relationship:
p Qy (y varies from 2.0 to 2.6)
In both cases the value for the exponent x or y is dependent on detail design of
the cyclone.
Feed pressure or pressure drop
From the above relationships it follows that
d50 px/y (x/y has values 0.25 to 0.27)
The determining factor in cyclone operation is pressure drop and not feed
pressure. The cyclone develops separational power through the use of fluid
pressure energy. There is thus a loss in pressure or a pressure drop across the
unit which is an important operating variable. Pressure drop usually used for
reference purposes is that between points immediately before feed entry and
immediately after overflow exit.

The range of pressure drop in common use is 5 to 50 psi with the lower limit
dictated by vortex stability and the upper limit largely by economics (decided by
pump availability and cost and wear rate at higher pressures).

Solids concentration

High feed solids concentration causes hindered settling, higher


concentrations of solids in the underflow and a change in pressure drop or
capacity.

To avoid hindered settling fluid to solid volume ratio of 8 : 1 is preferred.

Better check the behaviour of underflow discharge for correct feed solids
concentration.

(a) Is preferred when maximum removal of solids from the overflow stream is desired.
(b) is the correct operation.
(c) Although this is not desired but during dewatering partial rope discharge is preferred.

The effects of solids size and shape, solids density, liquid medium
density and viscosity on hydrocyclone performance can easily be
understood if we understand the fundamentals of particle separation
mechanism inside it.

Can we????

Hydrocyclones in Operation

Integral part of any industrial closed circuit grinding circuit

Various sizes, styles, and configurations of hydrocyclones available in the industry, including
hydrocyclones from 0.5 to 90 [13- to 2,286-mm] in diameter to meet specific performance
objectives. Hydrocyclones are also available in a variety of housing, liners and component
materials. Literally, hundreds of various hydrocyclone configurations are available to meet
specific requirements. Some of them are shown here.

Some Interesting Circuit Configurations

Two hydrocyclones in series for clarification and the dashed line shows possible
partial recycle of the overflow from the second stage to predilute the feed to the
first stage. If the recovery in the first stage is high enough the second stage
may not be required.

Actual configuration whenever predilution of overflow of the second stage is


required, for thickening and classification.

Two hydrocyclones in series with second stage underflow recycle, for


thickening and clarification.

Two hydrocyclones in series for thickening

Reclassification arrangement designed to sharpen the cut

Arrangement for thickening of dilute feeds

Multiple pass arrangement (small spigot to remove suspended particles


from the feed tank)

Hydrocyclone Selection Criterion


Any Hydrocyclone is inefficient. Coarse particles will report to overflow and
fine particles to underflow.
The nominal cut point for a cyclone is therefore defined as d50, i.e. the
size of particle that has 50% chance of reporting either to underflow or
overflow. This cut point is used in selecting the correct cyclone diameter.

An end user of cyclones normally


doesnt use the value d50. In practice
the selection is based on required size
analysis of the overflow i.e. 95 % minus
100 micron.
(K95 = 100 micron)

Example: A flotation circuit needs a 95% minus 75


micron feed. This corresponds to a nominal cut point
d50 =75 x 0,65 = 48,75 micron

Once d50 is defined the hydrocyclone diameter


can be selected from the following Table

Feed density
For efficient classification it is important that the feed density is as low
as possible (free settling particles).
10-15 % solids by volume

Good efficiency

15-30% solids by volume

Deteriorating efficiency

30 % solids by volume

Inefficient

Feed pressure will influence the cut point, higher pressure lower cut
point (look out for wear).

Select quantity of hydrocyclones


The volumetric capacity of a cyclone depends upon its diameter. A
larger cyclone will handle a larger capacity. Once the required
diameter has been defined then the number of units needed to handle
the given feed flow can be determined from the following table.

Suggested Text Books

1. D. Bradley, The Hydrocyclone, Pergamon Press, Oxford, London


2. L. Svarovsky, Hydrocyclones, Holt, Rinehart and Winston Ltd.,
London
3. Barry A. Wills, Mineral Processing Technology, Pergamon Press,
Oxford, London

Thank You

Performance Improvement of
Hydrocyclones in Closed Circuit
Grinding Operation

Problem Statement
Around 3040% of finer particles which were
expected to report through overflow comes to
underflow. These particles are generally the
heaviest particles (most of the cases the desired
minerals)
The particles are thus unnecessarily getting
overground which enhances the grinding cost.
Commonly addressed as short-circuiting in
literature

A study (Kawatra and Eisele, 2002) determined that the retention of fine particles in
the grinding circuit by the hydrocyclones resulted in a waste of 1.57 kW-hr for
every ton of ore ground.
Direct overgrinding losses: 23 million kW-hr/year

Excerpts of an Interesting Discussion on this Subject


Prof. Alban Lynch expressed an opinion on the use of hydrocyclones in closed grinding circuits: "the way
they are used now is an absolute nonsense, with circulating loads in some cases of well above 200%. The
future is high frequency screens". The disadvantages in the classifier is that their separation characteristics are
poor and they can produce recycling loads up to 400% although the particles which require regrinding may
comprise only a small fraction of this. These high loads limit the capacity of mills to grind new feed and
reduce the sharpness of the split, both of which may be costly. The problem occurs in both dry and wet
processes and the different approaches used in the cement and ore industries.
A screen behave in a totally different way. As the circulating load starts to increase, the shorter residence time
on the screen starts to make the screen passing fraction finer. The partition curve will start to change towards
finer cuts with the nominal aperture as the hinge point. The end result is that the circulating load increase fast
(How fast and when this will happen will depend on the screen dimensioning). All disturbances remain in the
grinding circuit. One has to be very clear and careful where to use screens or cyclones. Operations with high
ore grindability variations
Need to think twice before using screens! - Kari Heiskanen
Hydrocyclone and/or Heavy media cyclones separate particles based on the settling velocities and related
velocity phenomena and fluid density. That is why a small liberated grain of galena is bound to go to
underflow(which we do not really want) if it has the same forces acting as on a coarse unliberated particle of
equal settling velocity I made the observation during my talk with you; we need a different parameter based
on behaviour of particles in fluid media while analysing cyclone performance; we should not talk in terms of
screen size. - T.C. Rao, India

Information: mineng.com : http://min-eng.blogspot.co.uk/2014/08/a-brief-history-of-classification-in.html

How do we evaluate the performance of a hydrocyclone?


Performance curve or Tromp curve

This is normally S shaped curve and a variety of quantitative expressions have


been used to describe the shape of the curve.

The entire procedure of plotting the data in such a manner assumes that
the hydrocyclone classifies particles based on particle size only.

It takes into account the quality as well as the quantity of the products. (The
quality being explained through the separated particle size and the quantity
through the recovery of each distinct size.)

How cyclones are selected for a specific


application?

liberation size of the desired mineral and the quantum of minerals to


be processed per unit time
Based on liberation size , Cut size (d50) of hydrocyclone is chosen

Hydrocyclone manufacturers chart

What are the inherent bottlenecks?


It is commonly observed that the relatively heavier
and finer size fractions are not collected through
cyclone overflow till the size fractions become much
finer than the desired cut size.
In the process these particles get re-circulated back
to the grinding circuit which unnecessarily consume
significant amount of energy.
As a result, this inherent limitation of the
hydrocyclone significantly affects the entire process
economics as around 30-40% of the feed particles
normally follow this trend of being overground.

Previous approaches to minimize


overgrinding
Water injection cyclone(Gmax cyclone)

Three product cyclone

Draw back: Coarser cut size (d50)

Observations
The inefficiency of a hydrocyclone to separate particles at a desired
size may be the combinations of the following three factors:

Density variation of the feed particles


Collection inefficiency
Short circuiting

Our Hypothesis
Hydrocyclone separates particle based on relative settling velocity which is
not only a function of particle size but also dependent on particle density and
shape

Settling Velocity ( cm/s)

100
D1
D2
D3
D4

80
60
40
20
0
10

100

1000

Size (micron)

Equal settling velocity Particle will have same orbit of rotation and report to
the same stream

Case Study
A project Optimization of Comminution Circuit Throughput and
Product Size Distribution by Simulation and Control was taken by S.
K. Kawatra (Principal Investigator) and T. C. Eisele, Department of
Chemical Engineering of Michigan Technological University. Their
primary objective was to identify and provide a solution for inefficient
grinding circuit.

They identified two major bottlenecks


Underutilization of equipment, due to bottlenecks in the
system that prevent all parts of the circuit from operating
at full capacity.
Over grinding, due to material being retained in the circuit
after it has reached its target size, and therefore, being
ground to finer than the target size.

Particle size distribution along with respective recovery to underflow

Recovery to underflow
Size
(micr
o
meter
)
10
12
14
17
20
25
30
38
42
45
52
60
74

Quartz

Magnetite

0.05
0.05
0.05
0.07
0.09
0.11
0.25
0.4
0.6
0.8
0.9
0.92
0.98

0.11
0.115
0.12
0.32
0.48
0.8
0.85
0.92
0.94
0.97
0.98
0.99
1

Feed (50:50
mixture)
0.1
0.15
0.115
0.26
0.42
0.7
0.76
0.8
0.86
0.91
0.94
0.96
0.99

Partition curve of magnetite and quartz


mixture

Cut size (d50) of magnetite : 22


micron
Cut size (d50) of quartz :

Justification
The underflow product through the hydrocyclone had a much higher
concentration of magnetite particle ranging from 20 to 39um as shown figure
above.
40

Recovery to underflow

Magnetite
Quartz
30

20

10

0
10

100

1000

Size (Micron)

Recovery curve of hydrocyclone underflow

Quartz particle of 39 micron is having equal settling velocity of a magnetite particle

of around 22 micron. This explains why the magnetite particles within the range of 22
(close to 20 micron as observed) to 39 micron report to underflow instead of reporting
to overflow.

You might also like