HOME

CHAPTER 12
Beneficiation –
Concentration

Sponsored by:
SPONSOR PROFILE

Mineral Technologies is a recognised leader in mineral mineral processing operations, and the establishment of
processing solutions worldwide. significant new equipment installations worldwide.
Throughout our 70+ year history we have delivered a In addition to our collaborative approach, we proactively
comprehensive range of integrated equipment and services research and develop new equipment designs, manufacturing
that cost-effectively transform orebodies into high-grade materials and technologies to continually improve throughput,
mineral products for customers. grade and recovery, equipment durability, overall safety and
environmental outcomes that deliver cost-effective solutions
Today, our commitment continues with the ongoing release of for our customers.
an expanding range of minerals testing, plant design and mineral
separation equipment solutions across the industrial minerals, Underpinning our design expertise is our extensive operational
iron ore, coal and metals market sectors. knowledge gained from regular access to processing plants,
combined with our broad range of proprietary equipment
Our core capabilities include: and service offerings. This unique combination enables our
engineering teams to go beyond delivering standard engineering
•• mineral testing and process design
outcomes.
•• process plant design
Our Australian manufacturing facility houses the world’s largest
•• process optimisation
spiral manufacturing facility and produces over 20 000 starts
•• design, manufacture and supply of mineral processing annually. In 2010/11, we manufactured HC33 and WW6 spirals
equipment for Arcelor Mittal’s Mont Wright mining operations in Canada to
•• equipment commissioning and training. deliver the largest single order for spirals in our history.
A key element of our ongoing commitment is close collaboration Our head office, based in Australia, supports a network of
with customers in the development of innovative equipment offices located in South Africa, India and North and South
and services to meet specific mineral processing needs. This America. Our teams operate within an international engineering
has resulted in long-standing customer relationships, often and supply strategy that enables us to successfully work with
developed during the upgrading and expansion of existing customers anywhere in the world.
CHAPTER CONTENTS

Gravity concentration 267

Advantages 267
Suitability 267
Mineralogy 268
Test work 268
Flow sheet design 270
Jigs 271
Spiral concentrators 273
Shaking tables 274
Centrifugal concentrators 276
Sluices and strakes 279
Magnetic and electrostatic separation 280
Feed characterisation 280
Equipment description 280
Equipment selection 280
Cost estimation 283
Worked examples 283
Flotation 286
Data requirements 286
Jameson cell operation 287
Mechanical cell selection 288
Cost estimation 289
Worked examples 289
Lifetime costs of cells 291
Summary 291
Ore sorting 291
Sorting machines 291
Sorting mechanics 292
Cost estimation 293
Acknowledgements 293
References and further reading 293
CONTRIBUTORS

First edition text by: M Millard, R Barbagallo, N Dawson, B Ruggiero, P Baily and R Dudgeon

Revised and updated by:

Damian Connelly FAusIMM(CP), Director / Principal Consulting Engineer, Mineral Engineering Technical
Services Pty Ltd
Alexander (Sandy) Lewis-Gray, Technical Director, Gekko Systems
Denis Yan MAusIMM, Consulting Metallurgist, Mineral Engineering Technical Services Pty Ltd
Beneficiation – Concentration

This chapter presents four concentration methods in equipment specification and control system options.
ore beneficiation: gravity concentration, magnetic and Suppliers should be contacted for current, accurate
electrostatic separation, froth flotation and ore sorting prices.
techniques.
Advantages
GRAVITY CONCENTRATION Faster cash flow, higher overall recovery and lower
Significant progress has been made in the area of gravity cost-per-tonne are all significant advantages for high
separation in the last 20 years. Both test procedures and specific gravity minerals and especially for gold.
modelling are now at the stage where recoveries and Gravity gold recovery can:
grades can be predicted along with throughput data for •• improve carbon-in-pulp (CIP) leach kinetics
most commonly used equipment. Modelling is simple •• recover coarse gold that would otherwise be slow
and gives the engineer or metallurgist an accurate to leach
understanding of the outcome of an installation under •• reduce cyanide consumption
varying scenarios. •• reduce gold-in-circuit lock up
Some gravity concentration technologies have been •• reduce the CIP feed grade.
omitted from this section due to obsolescence and/or
It is environmentally friendly, as no reagents are
where no accurate means of predicting performance
required.
can be made. The reader is referred to the previous
edition of this handbook for this information. Suitability
Information is provided here on the capacities Whether or not gravity separation may be applicable to
and characteristics of a number of items of gravity a particular resource can be indicated by calculating the
equipment. This should be sufficient to permit this concentration criteria (CC) as defined by Taggart (1945):
equipment to be integrated in a preliminary flow
sheet and allow a mass balance to be calculated. The CC = (DH - DF) / (DL -DF)
capacities given for equipment are highly dependent
on the feed being treated and are only quantified by where:
detailed test work. DH specific gravity of the heavy mineral
Pricing of the equipment is relatively simple, as for DL specific gravity of the light mineral
most gravity equipment the machine is not custom- DF specific gravity of the fluid medium
designed to the task. Rather, multiple units are
Table 12.1 indicates the heavy mineral specific gravity
combined in a given treatment stage.
corresponding to the concentration criteria for a gangue
The most suitable flow sheet, and the equipment to specific gravity (sg) of 2.65 in water.
be used in it, will be determined by the particle size
Modern gravity concentration equipment has
to achieve efficient separation and the mineralogy of
considerably reduced the particle sizes corresponding
the valuable and gangue constituents. The variety
of gravity circuits is such that it is not possible to
comprehensively cover all available options. A few TABLE 12.1
general comments are made regarding flow sheet Heavy mineral specific gravity for a gangue of 2.65 specific
design. It is recommended that the reader retain the gravity in water.
services of someone with relevant gravity experience
CC DH Separation size Separation size
and seek information on theoretical and operating flow
(Taggart, 1945) (modern)
sheets given in the ‘References and further reading’
section of this chapter. 2.5 5.1 ‘To finest sands’
The prices given are in 2011 Australian dollars and 1.75 3.9 >150 �m >50 �m
include the electric motors and drives required to 1.5 3.5 >1.5 mm >250 �m
operate the equipment. Equipment prices are fixed and
prices change according to the cost of raw materials, 1.25 3.1 ‘At gravel sizes’ >1.0 mm

Cost Estimation Handbook 267

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

to the concentration criteria given by Taggart (1945). recovery by some 30 per cent because of the way the
These modern values should be viewed in the context test work was undertaken or interpreted. If too little
that in commercial practice the concentration criteria of the recirculating load is allowed to be bled to the
(CC) can be improved through the use of heavy media primary recovery circuit, or the primary unit is failing
separation techniques for particles over 500 μm (Wills, to perform because feed is too coarse, too dense or
1989). Advances in the new range of centrifugal gravity producing flakes, recoveries will not be achieved.
devices such as the in-line spinner; Knelson centre Proper sample selection, planned laboratory test
discharge (KC-CD) and extended duty (KC-XD) batch work and experienced interpretation of the test work
concentrators and continuous variable discharge (CVD) form the basis of process selection. As a minimum,
semi-continuous concentrators; Falcon semi-batch (SB), test plans should include gravity test work on samples
continuous (C) and fine (F) continuous concentrators; selected spatially throughout the orebody to determine
the multi-gravity separator and the Kelsey centrifugal the range of recoveries encountered. Test work should
jig have also improved CC. All of these machines can target samples widely dispersed throughout the
increase recovery efficiency at lower concentration oxide, transition and primary zones. Samples need to
criterion for finer particles. be obtained with greater definition for ores such as
laterite, coffee rock, pisolite and saprolite.
Mineralogy
Composites are not recommended, as variability of
Mineralogy determines the amenability to gravity recovery is hidden. Rather, geometallurgical modelling
processing. The following characteristics of the techniques are recommended as they can directly reduce
minerals present are particularly important: the risks associated with meeting production targets.
•• composition The samples should have designated coordinates so
•• degree of liberation the recovery can be assigned to a location in the block
•• density differential model and be used in the mine schedule. These data
will be used in conjunction with the variability data and
•• hydrophobicity
support the recovery predictions used for a particular
•• particle shape orebody. Geometallurgy can be used to identify:
•• particle size. •• concentration of deleterious elements
Characterising the feed is the first basic step •• drillability
in determining the most applicable separating
•• fragmentation
equipment and developing the optimum flow sheet.
•• grindability
Characterisation techniques include:
•• hardness
•• optical microscopy of polished sections – useful but
limited •• metallurgical recovery
•• QEMSCAN – new technology replacing optical •• mineral liberation
microscopy •• mineral species and mineral grade
•• heavy liquid separation – useful especially for gold •• mining recovery
because the major problem with gold is locating •• reagent consumption
sufficient numbers of particles confident that these •• smelter-enabling characteristics.
particles represent gold occurrence Electronic data can be used to evaluate various
•• scanning electron microscopy and gold analyses scenarios based on comminution and ore recovery.
combined with diagnostic leaching – useful in Hence, geometallurgy can identify the likelihood of
understanding the nature and occurrence of the gold an unwanted event and the consequences, thereby
present; for example, the size varies from colloidal permitting risk mitigation to be put in place.
through to nugget gold, metallic gold is common,
gold occurs as alloys with other metals and within Test work
sulfides. Interpretive techniques have advanced in the area
of particle liberation in ores to the point where the
Sample representivity minerals’ natural grain sizes and construction can be
Recovery based on non-representative samples is a relatively simply reviewed in a piece of drill core. There
major issue. At Mt McClure, recovery was miscalculated are several technologies such as scanning electron
so a gravity circuit was installed immediately. The microscopy coupled with mineral liberation analysis
presence of nugget gold at Bronzewing and Granites (MLA) and QEMSCAN to achieve this. Also, three-
Gold made it extremely difficult to predict likely dimensional (3D) tomography can give a 3D view
gravity gold recovery using small test samples. The of the ore with grain boundaries clearly defined and
presence of sulfides (galena) and tramp iron can cause discrete particles described and characterised. Particle
problems when cleaning up gravity concentrates for shape plays a very large role in gravity separation and
tabling. In a number of cases, gravity recoverable this can be explored by 3D tomography and taken into
gold (GRG) analysis overestimated the gravity gold account in modelling.

268 Cost Estimation Handbook

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

Heavy liquid separation (HLS) is an excellent on rolls-crushed core samples to a size determined
predictor of the potential for gravity concentration. by the mineralogy, normally 1.5 to 6 mm. The heavy
Typically, HLS using tetrabromoethane provides yield, liquid test work indicates the size at which the
recovery and quality of the concentrate. valuable mineral becomes liberated. When combined
For gold, the typical GRG test work is to grind a 25 with mineralogical examination of the products, it
- 100 kg sample through three successively finer sizes gives a first indication of the primary grind and the
(P80 of 850 μm to 75 μm) and between each step passing regrind size at which a high recovery and/or a saleable
the tailings reground sample through a batch laboratory concentrate grade is achieved. Plotting these results
as a liberation curve is a valuable tool in determining
scale Knelson concentrator to recover gold liberated
whether gravity techniques are applicable. However,
(Laplante, 2000). A single P80 75 μm GRG can also be
heavy liquid analysis will not in itself indicate whether
undertaken, but this is not as accurate and does not
gravity separation is practical.
provide any liberation data. Note that GRG predicted
by this method will not reflect typical plant recovery, as Hard rock gold ores
100 per cent of the GRG is recovered from the sample When a gold ore is to be treated solely by gravity the
but only a portion of the feed is typically processed comments in the preceding paragraph apply. However,
in an industrial centrifugal concentrator. Hence, only more frequently gravity concentration is considered
a portion of the total GRG is recovered. As with other to prevent a build-up of coarse gold in the grinding
GRG test work, over-prediction must be allowed for. circuit, reduce downstream operating costs and ensure
Intensive leaching of the concentrates in either a slow-leaching coarse gold particles do not report to the
Consep Acacia or Gecko in-line reactor is typically tails of the leach circuit.
carried out and very high gold recoveries are usually The performance of a batch centrifugal concentrator
obtained. (BCC) in the milling circuit can be predicted using a
Interpreting gravity recovery results requires a skilled combination of GRG test work results and modelling
person who can relate the results to plant practice. by either the manufacturers or by a model produced by
Consultation with the vendors is also very useful to the AJ Parker Centre in Perth, Western Australia.
ensure the correct conclusions are drawn and current The GRG content of the ore can be determined in a
pricing information is obtained. test developed by Laplante (2000) where 25 to 100 kg
The test work required to approximate a flow sheet of ore sample is passed through a laboratory scale BCC
varies considerably with the type of ore processed, as and the concentrate and tailings analysed for gold
follows. content by size. The test can be carried out in three
stages (most comprehensive) or as a single-stage test.
Gold alluvials The critical information required for design purposes
Metallurgical testing and ore grade determination is the concentrate mass as a percentage of the overall
form part of the same exercise. Due to the nugget mass. The concentrate is assayed by size to indicate the
effect, the size of sample that must be taken to achieve gold deportment.
adequate sampling accuracy extends to several tonnes.
The GRG results can be input into various models to
Bulk sampling equipment is mandatory. The only
give an indication of gold recovery at varying machine
matter requiring metallurgical resolution is whether
sizes and percentage of recirculating load treated
exceptionally coarse or fine gold is present, as this
to determine the expected full-scale performance.
would escape a gravity roughing stage and require
Typically the full-scale performance will be between 60
separate treatment. To detect abnormally coarse gold,
and 80 per cent of the GRG result.
some test rigs are equipped with a metal detector
on the trommel oversize belt. For fine gold, reliable The performance of the low-gravity separators such as
assays can be obtained on samples of the fine fraction jigs and spirals is machine-specific. The manufacturers
of the tailings due to the reduced nugget effect and should be consulted to ensure the correct procedures are
concentration of the gold in this fraction by screening. followed and data are generated to enable equipment
sizing.
Other alluvials
Tin and tantalite alluvials and beach sands are usually Coal
sampled with equipment such as Calweld or Banka The response of a coal deposit to gravity separation
drills. Material from these samples can be parted and is more easily quantified than other applications due
subjected to the usual mineralogical analysis, which to the ability to categorise the washability by carrying
indicates the size distribution of the valuable mineral out heavy liquid tests at small specific gravity (sg)
and more significant gangue components. increments over the entire range of the coal and ash
components of the feed. Provision of this information
Hard rock ores to an equipment supplier or reference to the treatment
The minimum information that is required is detailed flow sheet for a similar coal permits a probable error of
mineralogical examination and heavy liquid test work separation or Ecart probable (Ep) value to be predicted

Cost Estimation Handbook 269

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

for specific equipment. Ep can then be used in modelling •• A single separator cannot treat a broad particle size
software to predict yield and product quality. range. Many items of gravity equipment like jigs
rely on differential free or hindered settling rates
Piloting between gangue and valuable mineral particles.
Piloting is rarely undertaken for greenfields gold project Above a certain particle size spread, the coarsest
circuits. It has occurred for chromite, metallic copper light or low-density particles will have a greater
or in additional specific applications where there settling rate than the finest heavy mineral particles.
was uncertainty. Piloting is common for brownfields •• Greatest separation efficiency is obtained by using
projects or where retrofitting of equipment is being a large number of narrow size ranges. However,
considered. For gold, batch testing using a Knelson capital cost and operational considerations impose a
or Falcon concentrator may be done. Gekko may practical limit that depends on the scale of operation
undertake testing using the in-line pressure jig (IPJ) and mineralogy of the specific deposit.
where very coarse gold is present. •• As a general rule-of-thumb, a factor of two between
Flow sheet design the finest and coarsest sizes in a given fraction is
an upper limit for tin, tantalite and tungsten hard
It is only intended to discuss gravity circuit flow sheet rock concentrators at sizes less than 500 µm. At
design in very general terms due to the great variety subscreen sizes these classifications are usually
of applications and the uniqueness of each resource. achieved in single or multiple spigot hydrosizers,
The guiding principles of all mineral dressing circuits cyclones or teeter columns. In the case of multiple
apply: spigot hydrosizers, the desliming operation can be
•• remove the valuable mineral as soon as it is liberated carried out in the same piece of equipment.
at the coarsest possible size •• It is evident from the above that material classified
•• reject barren or tailings-grade material as soon as it by screens is effectively treated on any type of gravity
is generated and at the coarsest possible size. device while products from hydraulic classification
The most complex gravity circuits are associated with processes are most successfully treated on flowing
hard rock deposits but the design principles involved film equipment and centrifugal concentrators.
for hard rock apply to any other gravity separation •• When gold is the valuable mineral, the size range
operation, as follows: that can be treated is greatly increased because of
•• For gravity processes, feed preparation is critical the higher specific gravity differential from the
for efficient operation. The comminution circuit gangue.
should be designed to minimise fines generation •• When there is a degree of natural classification of
particularly when the valuable mineral tends to form the resource, as in some alluvials, the number of
slimes, as with cassiterite. This usually involves ball size fractions to be treated is much reduced.
milling with a high circulating load if fine grinding •• Gravity equipment often has an optimal feed
is required. solids concentration. Typically, significant volumes
•• High recirculating loads of gold into the mill should of good-quality water are required. Therefore,
be avoided, as flat flakes can be produced and these the overall water management scheme for the
are not as amenable to recovery. concentrator requires considerable attention during
•• Ideally the ball mill circuit should be closed with detailed engineering. Although water rejection from
screens rather than hydrocyclones, which tend a gravity circuit is often economically achieved by
to concentrate heavy minerals in the underflow, cyclones, in complex flow sheets a certain amount
resulting in overgrinding. The exception to of buffer capacity in density tanks or thickeners
this is gold, where the concentration in the makes the plant easier to control.
underflow benefits the performance of centrifugal •• Poor quality water may necessitate reverse osmosis
concentrators. It is reasonable to use screen plants or vacuum distillation units using waste
apertures down to 0.15 mm. For example, Derrick heat from powerplants for centrifugal concentrator
screens give undersize streams with an 80 per cent fluidising water.
product passing size (P80) less than 100 µm. Given •• Apart from alluvial operations, few concentrators
a lower limit for many gravity operations of around rely solely on gravity separation. They are often
50 µm, it is nearly always possible to use screens for accompanied by flotation, magnetic separation,
primary classification. leaching or other metallurgical processes.
•• Feed must be classified into different size fractions •• One area where design concepts are changing is in
for efficient gravity separation and to remove the treatment of the finest size fractions (5 to 50 µm)
slimes. Wills (1989) noted that the presence of slimes from hard rock deposits. These fractions are typically
increases the viscosity of the slurry, which reduces treated using flotation. The flotation concept is now
the sharpness of separation. Therefore, it is usual to being challenged by a new generation of higher
remove particles less than 10 µm from the feed. capacity gravity devices capable of separations at

270 Cost Estimation Handbook

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

these fine sizes, such as the multi-gravity separator, According to Wills (1989), good separations of a
the Knelson concentrator and the Kelsey centrifugal fairly narrow specific gravity range (eg fluorite sg 3.2
jig. These machines use high apparent gravitational from quartz sg 2.7) are achieved using jigs if the feed
field techniques and appear on the verge of being is fairly closely sized, such as 3 - 10 mm. When the
accepted as fully developed production machines. specific gravity difference is larger, good concentration
•• The introduction of centrifugal concentrators was is possible over a wider size range. According to
the single largest positive change in gravity flow Campbell (1991) the typical effective operating range
sheets. The use of up to 200 g force results in higher is from 100 μm to 16 mm and that the treatment of
recovery of gold at the finer sizes. This was followed +16 mm material is carried out in a separate circuit to a
by the change from a bleed of the mill discharge to top size of over 30 mm.
a bleed of the cyclone underflow. The latest trend Square mineral jigs come in sizes from 300 mm square
has been to treat cyclone feed rather than cyclone to 1200 mm square. According to Campbell (1991), they
underflow because of higher gold recovery. have a capacity of approximately 10 m3/h/m2 whether
•• The vast majority of centrifugal concentrators today the cells are arranged as a single cell or two cells in series
are used in recirculating loads in gold grinding with a hutch water requirement of 22 m3/h/m2 for each
circuits for the recovery of free gold. Partial cell. Jig feed is normally in the range 30 - 70 per cent
recoveries down to 10 µm are observed while solids and it is frequent practice to put a dewatering
recoveries in the size range 50 to 1000 µm is where cyclone on the jig feed. It must be stressed that there
these units excel. are significant variations from this feed rate criterion
•• As machines became larger and able to process depending on the material being treated.
a greater percentage of feed, there was a move It is usual on a clean placer feed to use two jig cells
towards automatic discharge that allowed the in series for each concentration stage. For fine feeds
concentrate to flow by gravity to a hopper in the there is a reduction in jig efficiency, which may require
goldroom, thereby restricting access and theft. a third cell.
•• New developments are frequently trialled in Heavy clays resulting in a high slimes content cause
applications that have failed to be satisfactorily a large reduction in separation efficiency, requiring
resolved by conventional techniques. A limited additional cells and reduced throughput to obtain
amount of information is available on the satisfactory recoveries.
application of these new technologies to standard Single cells are seldom used; an exception is for the
duties. Pricing data on this new technology is also treatment of +16 mm material in deposits with very
preliminary and subject to significant change as coarse gold.
manufacturing procedures become established. For
In addition to the treatment of alluvials, jigs are
these reasons it is recommended that consideration
used to treat coal and hard rock ores where they find
of new technologies in the prefeasibility assessment
application at coarse sizes in the region of 700 μm to
of a resource should be made through the suppliers
12 mm (hard rock) and 400 mm (coal). Reichert cones
directly.
and spirals, although they will handle material up to
Jigs 3 mm, are most efficient below 1 mm so that jigs can be
used to recover values liberated at coarser sizes.
Jigging is one of the oldest methods of gravity
concentration but is still widely used today. This section The costs and power consumptions of different
presents mineral jigs and jigs for other applications. configurations of square jigs are shown in Table 12.2.
In very approximate terms, jigs recover around
Mineral jigs ten per cent of the feed weight to a ten per cent solids
Currently jigs are predominantly used for processing concentrate (hard rock). However, this will be highly
alluvials of all types where they can perform both variable with the jig operating settings, ragging
roughing and cleaning duties. The most frequent material used and ore processed. For cleaning stages
circuit configuration encountered in practice is for the weight recovery is up to 20 per cent.
the concentrate from a two or three-cell primary jig For larger scale alluvial operations circular jigs
to be cleaned in a two-cell secondary jig. For gold are used. The cells in these jigs are configured as
applications the secondary jig concentrate can feed a the segments of a circle with central feed point and
centrifugal concentrator or Knudsen concentrator. circumferential tailings collection. The largest of these
For some gemstone applications, product is also jigs has an area of about 42 m2 comprising 12 cells
recovered from the top of the primary jig screens. and treats 150 - 300 m3/h. This represents a slightly
If significant valuable mineral occurs below the jig’s lower throughput per square metre than a square
recovery size limit then an additional circuit is required jig. Note that the segmented cells of a circular jig are
to treat the fines. This circuit typically comprises spirals approximately equivalent in performance to two
or, in the case of gold, Knelson concentrators. square jig cells in series.

Cost Estimation Handbook 271

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

TABLE 12.2
Capital costs for Russell jigs (AMTAS Pty Ltd, 2011).
Russell jig No of cells Nominal screen Approximate Power (kW) Lead ragging Capex
model area (m2) capacity (t/h) (A$) (A$)
LJ2 2 0.09 2 0.2 79 4712
J3 2 0.3 10 0.8 525 10 280
J5 2 0.5 15 1.1 863 14 954
J8 2 0.8 25 1.1 1380 17 768
J14 2 1.4 35 1.5 2173 29 177
J24 2 2.4 45 2.2 3726 49 934
J2/6 4 0.6 10 0.8 1035 15 710
J2/10 4 1 15 1.1 1725 28 690
J2/16 4 1.6 25 1.5 2760 37 662
J2/28 4 2.8 35 2.2 4830 47 465
J2/54 4 5.4 50 3.0 8383 75 977
J2/64 4 6.4 60 4.0 9936 83 680
J2/70 4 7.0 65 4.0 10 868 86 890

Circular or segmented jigs have similar operating 2,000

parameters to square jigs with the exception that they 1,800
are fed with a slurry containing 20 - 25 per cent solids. 1,600

Russell jig 1,400

Plant Cost, A$000

1,200
The Russell jig comes in sizes from 300 to 1450 mm and
has a capacity of 2 - 65 t/h. The cells are arranged in 1,000
two types, the two-cell and the four-cell options. The 800
cell arrangement is that the two cells are side-by-side. 600
With the four cells the arrangement of the cells is two 400
side-by-side and two cells long. The main use of the
200
four-cell jigs is to recover fine products or where a
large percentage of the product is likely to collect on 0
0 50 100 150 200 250 300
top of the screens. For ragging, lead shot is used and
230 kg/m2 of screen area is allowed. Throughput, t/h
Table 12.2 gives the capital cost of various models based FIG 12.1 - Alluvial plant costing based on Russell jigs
on the number of cells and tonnes per hour throughput. (AMTAS Pty Ltd, 2011).
It should be noted that the feed distributors, walkways,
access stairs, water manifolds, stainless steel fittings, The IPJ has been used in a variety of jigging
wedge wire screens, special paint treatment and wear- applications from free gold recovery in alluvial
resistant linings are not included in the cost. operations, to preconcentration of tin and silver from
-12 mm crushed rock. Feed densities up to 75 per cent
The cost of an alluvial plant base on the Russell jig is
solids can be easily handled.
shown in Figure 12.1.
Power consumption is typically less than 0.05 kWh/t
In-line pressure jig and water consumption is lower than conventional jigs
The IPJ combines a circular bed with a moveable sieve at 0.4 m3/t, due to the mechanical jigging action.
action, as shown in Figure 12.2. The screen is pulsed IPJ sizes and capital costs are given in Table 12.3.
vertically by a hydraulically driven shaft with the
length of the stroke and speed of the up and down Coal jigs
stroke varied to suit the application. The higher specific Coal washing jigs are typically driven by compressed
gravity particles are drawn into the concentrate hutch air. Baum, Batac and BMCH jigs have fairly similar
during the suction stroke of the bed. Particles are kept operational characteristics (refer to Figure 12.3) and
submerged in the slurry thus eliminating the loss of will be treated identically. Refer to suppliers for more
hydrophobic fine particles at the air-slurry interface. accurate information.

272 Cost Estimation Handbook

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

Feed

Deceleration 13
chamber
14 12 1. Inlet
2. Refuse outlet
Wedgewire
Ragging 11 3. Refuse collecting hopper
screen 1
10 4. Middlings outlet
5. Middlings collection hopper
6 6
Diaphragm 6. Discharge device
7 7. Clean coal overflow
8. Air chamber
2 4
9. Make-up water
8 10. Air-distributing tubes
11. Air-distributing tank
12. Waste-air collecting tank
9
13. Waste-air tubes with silencers
14. Valve control system
3 5
Tails discharge

Hutch water

Hydraulic ram Concentrate outlet

FIG 12.3 - Schematic diagram of a Batac jig.

The split of the jig feed to the different products is

entirely dependent on the nature of the coal being
FIG 12.2 - In-line pressure jig.
treated. For a typical Australian clean coal from an open
cut mine, ten to 15 per cent of the feed is discarded and
TABLE 12.3
this proportion increases to 50 - 60 per cent for a dirty
Gekko in-line pressure jig unit costs.
coal from an underground mine.
Model IPJ1000 IPJ1500 IPJ2400 The water associated with discard and middlings
Maximum feed rate (t/h) 25 50 100 streams is only surface moisture remaining after
removal of that material by elevators and the balance of
Maximum feed rate (m3/h) 50 100 200 the water added reports to the washed coal stream. The
Maximum feed particle size (mm) 25 25 25 amount of water required varies from about 2 to 4 m3/t
depending on feed characteristics and operational
Footprint area (m )
a 2
1.96 3.24 6.25
variables.
Installed power (kW) 1.5 2.2 4.0 Typical operating consumable costs for coal jigs are:
Capital cost – unit b $/(t/h) 4600 2900 2100 •• power consumption – 0.13 kWh/t/h coal
Capital cost – control/ 1000 •• air consumption – 17 m3/t/h coal at 30 kPa.
automation $/(t/h) Figure 12.4 indicates the price-capacity relationship
Installation cost c $/(t/h) 500 for coal jigs assuming a fairly clean easily washed coal
requiring five compartment jigs.
Maintenance cost $/t 1.1.7
Total capital cost $/(t/h) 3500 1,000
900
a = Plant floor space taken up. 800

b = Includes starter. 700

Jig Cost, A$000

600
c = Capital cost of surrounding structure, valves, pipes, etc.
500
400
The top size of the jig feed should be in the range of 300
25 - 150 mm. Classification of the feed is unnecessary 200
although limited washing is achieved below 0.5 mm 100
Smith (1991). 0
0 100 200 300 400 500 600 700 800
The cut point is a critical parameter in specifying coal
jigs and is between sg 1.55 and 1.90. Coal Jig Capacity, t/h

The number of jig compartments required depends FIG 12.4 - Cost of coal jigs.
on the coal type and shale loading and varies from four
or five for an easy washing coal to eight or nine for a Spiral concentrators
coal with high proportions of near gravity material and Spirals are very widely used in a large variety of
-12.5 mm shale. gravity separation applications with probably the

Cost Estimation Handbook 273

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

largest quantity being used in the beach sand industry. Also, a range of larger diameter, higher capacity spirals
Figure 12.5 shows the mechanics of separation in a spiral. with and without wash water, has been developed for
mineral sands and iron ore. These spirals can process
more than twice the amount of material compared
with a conventional unit. Such spirals offer significant
capital savings, especially for larger operations.
Spirals designed specifically for fine coal are also
available. These are larger in diameter than the mineral
spirals and can have slide cutters across the width of
the spiral to remove reject material to an inner channel
running next to the central column.
Spiral models and costs are summarised in Tables 12.4
- 12.6.

Shaking tables
Shaking tables represent some of the older types of
gravity separation equipment.
The Wilfley, Holman, Deister and James tables
come in approximately similar sizes and are grouped
together for costing purposes. The differences among
the models reflect details of the deck types and riffle
patterns. Differences in the intricacies and mounting
of the mechanisms make different styles of table more
FIG 12.5 - Gravity spiral cross-section. appropriate for certain size fractions but this does not
affect the relative costs.
Spirals are suitable for density separation in the size
range of approximately 30 - 2000 μm although they Shaking tables have been used traditionally for both
handle material up to 3 mm at reduced efficiency. In the roughing and cleaning duties although costs and
case of gold, they can be effective down to 20 μm. Recent area requirements have tended to restrict their use to
developments have seen new spiral models capable of cleaning following preconcentration on spirals or other
dealing with finer particles and very high-grade feeds. gravity equipment.
Most spiral manufacturers currently offer three The size range that is treated on tables is extremely
model ranges of mineral wash waterless spiral. These wide, ranging from around 15 μm to 2 mm for mineral
are classed as low-grade, medium-grade or high-grade applications and up to 15 mm for coal. However, the
depending on the heavy mineral concentration of the same decks and mechanisms are not applicable to the
feed. Typically, double, triple or quadruple start spirals entire range.
can be used, depending on the model.
The throughput capacity of tables varies with the
Most manufacturers can still supply wash water size of the feed and from one mineral to another. The
spirals. This configuration removes concentrate at capacity-size relationship for a full-size table is given
regular intervals down the length of the spiral using in Table 12.7. It is stressed that this is intended as an
small cutters. Although they have been demonstrated approximate guide and that considerable divergence
(Guest and Dunne, 1985) to be more efficient than wash from this is experienced in specific applications.
waterless spirals on a synthetic feed, they demand
A full-size shaking table has a deck approximately
greater operational labour to keep the wash water
2 m × 4.6 m and has a 1.5 kW motor drawing around
channels free and the cutters appropriately set.
0.6 kW. The wash water requirement will be from 1 to
Spirals are also produced with a water peel
4 m3/h for mineral applications and 2.4 m3/h per t/h
attachment that removes a portion of the water and
of feed for coal processing. The feed density should
slimes from the tailings discharge. It is particularly
be about 25 per cent solids for mineral duties and 35 -
useful in controlling the water balance when spirals are
40 per cent solids for coal.
used to treat mill circulating loads, removing organic
trash and dewatering for easier stacking of tailings. The distribution of feed to the various products is
The number of turns generally varies from three highly variable depending on the table settings and
to seven depending on the application. In addition, specific duty. Feed distribution is typically zero to
compound spirals, with two stages on the one column 15 per cent to concentrate and ten to 20 per cent to
in a ‘rougher-scavenger’ arrangement are available. middlings.
These have the additional advantage of eliminating The solids content of table products is up to
pumping and transfer between stages. 80 per cent for the concentrate, 70 per cent for middling

274 Cost Estimation Handbook

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

TABLE 12.4
Spiral performance characteristics (MG and HG series).
Spirals – MG series
Model MG4 triples MG6.3 triples MG6.2 triples
Mineral type used in: Feed material containing up to 40% heavy mineral
Maximum recommended feed rate kg/h/start 2 2 2
m3/h/start 8 8 8
Maximum feed size (mm) 2 2 2
Footprint areaa (m2) 3.3 3.3 3.3
Capital cost – unit b $/(t/h) $1900 $1800 $1660

Spirals – HG series
Model HG10i triples HG11 triples
Mineral type used in: Used for high-grade feed material generally from 20 - 90% heavy mineral
Maximum recommended feed rate kg/h/start 2 2
m /h/start
3
5.5 5
Maximum feed size (mm) 2 2
Footprint areaa (m2) 3.3 3.3
Capital cost – unit $/(t/h)
b
$2085 $1945

a = plant floor space taken up by a bank of six.

b = includes distributor, feed hoses, frames, subframes and launders.

TABLE 12.5
Spiral performance characteristics (VHG, FM and WW series).

Spirals – VHG series Spirals – FM series

Model VHG triples VHGS triples FM1 triples
Mineral type used in: Used in the final upgrading of very high-grade Used for fine feed with particles in
(+90% HM) feeds the range of 30 - 150 µm
Maximum kg/h/start 1.5 1.5 1
recommended feed rate
m3/h/start 4.8 4.8 5
Maximum feed size (mm) 2 2 0.15
Footprint area (m )
a 2
3.3 3.3 3.3
Capital cost – unit b $/(t/h) $2410 $2220 $3610
Spirals – WW series
Model WW6E doubles WW6+ doubles HC33 triples
Mineral type used in: Uses wash water addition for better grade control in specific applications (ie iron ore,
mineral sands)
Maximum kg/h/start 2 2 6
recommended feed rate
m /h/start
3
5 5 12
Maximum feed size (mm) 2 2 2
Footprint areaa (m2) 2.7 2.7 6.35
Capital cost – unit $/(t/h)
b
$2710 $2915 $825

a = Plant floor space taken up by a bank of six.

b = Includes distributor, feed hoses, frames, subframes and launders.

Cost Estimation Handbook 275

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

TABLE 12.6
Spiral performance characteristics (HC and LD series).

Spirals – HC series
Model HC 1 quads HC1RS quads HC33 triples
Mineral type used in: These super-high-capacity spirals have been designed specifically for more economical and
compact high tonnage plants. The facility to add wash water can be added on some models
Maximum kg/h/start 6 6 6
recommended feed rate
m /h/start
3
10 10 12
Maximum feed size (mm) 2 2 2
Footprint areaa (m2) 6.35 6.35 6.35
Capital cost – unit $/(t/h)
b
$835 $1630 $760

Spirals – LD series
Model LD7 triples LD7RC triples
Mineral type used in: Fine coal beneficiation
Maximum recommended kg/h/start 3 3
feed rate
m3/h/start 12 12
Maximum feed size (mm) 3 3
Footprint area (m )
a 2
7.12 7.12
Capital cost – unit $/(t/h)
b
$1390 $1760

a = Plant floor space taken up by a bank of six.

b = Includes distributor, feed hoses, frames, subframes and launders.

and around 20 per cent for tailings. Wash water is

frequently required to mobilise the product streams.
The cost of a full-size table is around $25 000 with
variations depending on whether a standard or oversize
deck is used, the details of the table support structure
and whether or not variable-speed control is provided.
Floor area is minimised by using double- or triple-
deck tables but these do not represent an equipment
cost saving.
Half-size tables with decks of approximately 1 m ×
2.1 m are also available at a cost of $21 000. The capacity
is approximately one-third of that of a full-sized table
and wash water requirements are approximately FIG 12.6 - Gemini shaking table.
0.6 to 3  m3/h. The above costs reflect the price of
the equipment only and tables require substantial Centrifugal concentrators
foundations particularly if multiple decks are used.
Centrifugal concentrators include the Kelsey jig, the
Another type of shaking table that has found use Falcon concentrator and the Knudsen in-line spinner.
particularly in gold applications is the Gemini table
(Figure 12.6). This has a maximum feed size of 1 mm Kelsey jig
with an optimum range of 20 to 800 μm. Optimum feed The Kelsey centrifugal jig (KCJ) was invented by Chris
density is 60 per cent solids although any fluid slurry Kelsey as a means of efficiently separating particles
can be treated. For the largest size table, maximum that are too fine and/or have too small a difference
capacity is 500 kg/h with optimum capacity 450 kg/h in specific gravity to be separated efficiently by more
and the wash water requirement is 2.25 m3/h. Installed conventional (1 g) gravity separation equipment. The
power is 0.75 kW and the cost of the unit is $27 000. separating mechanisms in a KCJ are the same as for
A summary of selected shaking table models and a conventional (coarse) jig; however, the ragging and
costs is given in Table 12.7. feed particles are spun so that they experience higher

276 Cost Estimation Handbook

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

TABLE 12.7
Shaking table performance characteristics.
Holman-Wilfley Pty Ltd shaking tables
Model 800 2000 3000 7000 8000
Mineral type used in: Recovery of precious metals, copper wire, synthetic diamonds, chromite, heavy mineral sands
and gold
Maximum recommended feed rate 60 - 75 60 - 450 100 - 800 500 - 2500 200 - 2500
(kg/h)
Maximum feed size (mm) 2 2 2 2 2
Footprint area (m )
a 2
2.3 6.0 4.1 10.8 10.2
Installed power (kW) 0.37 1.5 1.5 2.2 1.5
Capital cost – unit $/(kg/h) $148 $47 $34 $12 $14

Gemini shaking tables

Model GT60 GT250 GT1000
Mineral type used in: Designed to produce a gold concentrate that can be directly smelted to bullion
Maximum recommended feed rate 30 115 450
(kg/h)
Maximum recommended feed rate 720 1500 2280
(m3/h)
Maximum feed size (mm) 1 1 1
Footprint area (m )
a 2
1.29 2.93 4.77
Installed power (kW) 0.75 0.75 0.75
Capital cost – unit $/(kg/h) $550 $220 $60

a = Plant floor space taken up.

g-forces (typically around 25 g), which improves size of 500 μm down to approximately 6 - 10 μm. Their
separation (Figure. 12.7). main applications include mineral sands, tin-tantalum-
Kelsey jigs are used to recover fine and/or low-sg tungsten, gold, base metals and iron ore-chromite. Unit
differential mineral particles from a maximum top feed rates are very much application-specific, but rates
of up to 50 t/h solids are possible using the larger of the
FEED two available KCJ models.
H20
Kelsey jig models and costs are summarised in
RAGGING
HUTCH WATER Table 12.8.

Knelson concentrator
The Knelson concentrator is essentially a high-speed
centrifuge that traps heavier particles between the ribs
of a rotating cone. The spinning motion of high-speed
centrifuge against fluidisation water causes separation
of the gold particles. The concentrator can operate in a
batch or semi-batch sequence. The main components
consist of a riffled concentrating cone, drive motor,
water chamber and fluidisation water unit, as shown in
TAILING Figure 12.8. These concentrators are universally used
CONCENTRATE for gravity gold recovery in grinding circuits with a
capacity of 300 to 1000 t/h.
The accentuated gravitational forces in the separator
means that a broader than usual size range can be
handled and the manufacturers claim that gold
FIG 12.7 - Kelsey jig cross-section. particles from 3 mm to 1 μm are recovered. The Knelson

Cost Estimation Handbook 277

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

TABLE 12.8
Kelsey jig performance characteristics.
Model J1300 J1800
Mineral type used in: Zircon, rutile, tin, tantalum, tungsten, gold and nickel. In addition, test work has
achieved positive results for chromite, iron ore, niobium, base metals
(Pb, Zn, Co, Cu) and other applications
Maximum recommended feed rate (t/h) 20 50
Maximum feed size (mm) 0.5 0.5
Footprint area jig onlya (m2) 5 13.6
Installed power (kW) 40 b
60b
Capital cost – unit $/(t/h) $70 000 $50 000

a = Plant floor space taken up.

b = Jig only.

a fluidised recovery process. Knelson CVD models

range from the CVD 6 (2 t/h) to the CVD 64 (300 t/h).
All models are fully automated. A summary of selected
Knelson CVD models and costs is given in Table 12.10
(note that CDV 6 is not shown).

Falcon concentrator
The continuous (C) or ultra-fine (UF) models of the
Falcon concentrator can be used, depending on the
application. A general design is shown in Figure 12.9.
Designed for continuous duty, these machines can
produce mass yields as high as 40 per cent. This
technology is ideal for scavenging or preconcentrating,
since no water is added during processing. Concentrates
are deslimed and partially dewatered, typically to
70 per cent solids by mass, which makes subsequent
processing easy and inexpensive.
Unit capacities are up to 100 t/h. Forces up to 300 g
FIG 12.8 - Knelson concentrator cross-section. can be produced, which allows for recovery of fine
particles. These units are fully automated with a typical
concentrator is also unusual as it is insensitive to the availability of 95 per cent.
solids content of the feed as long as the slurry is fluid Applications include recovery and upgrade of tin,
and a limiting feed volume is not exceeded. There are tantalum, tungsten, chromium, cobalt, iron, fine
two production models of the Knelson concentrator: oxidised coal and many other minerals.
the centre discharge (KC-CD) and the extended duty
(KC-XD). Selected specifications and costs are given in Knudsen bowl and in-line spinner
Table 12.9. The Knudsen bowl is a centrifugal concentrator
Knelson CVD concentrator comprising a riffled cone that rotates about a vertical
axis. The units come in a single size capable of handling
The Knelson CVD concentrator was developed
up to 5 t/h of -4 mm feed, although the optimum feed
specifically to operate in higher mass yield applications
rate is closer to 3 t/h. The power requirement is about
where the target metal or mineral is available in larger
0.37 kW giving a rotational speed of just over 100 rpm.
quantities than can be effectively recovered in the batch
Knelson. The CVD concentrator uses similar principles The machine is batch-operated. It has to be stopped
of mineral separation and recovery to that of the batch for concentrates to be washed out through a drain
machine, but allows the concentrate to be ejected from point at the base of the cone. Hence its application to
the fluidised bed continually. Pinch valves, located at production operations is limited to gold and precious
the base of the fluidised rings, are kept closed by air metals, as these produce small weights of concentrate.
pressure. By releasing the air pressure periodically, Knudsens are most frequently used in cleaner duties
concentrate can be ejected without interruption to following jigs, spirals or cones. Some operations use
production. Similar to the batch machine, the CVD uses a quantity of mercury in the concentrator to improve

278 Cost Estimation Handbook

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

TABLE 12.9
Knelson semi-continuous (batch) concentrator characteristics.

Model CD 10 CD 12 XD 20 XD 30 XD 40/ XD 48/ XD 70

QS 40 QS 48
Maximum recommended feed rate (t/h) 8.0 20.0 80.0 150 250 400 1 000
Maximum recommended feed rate (m /h) 3
10.0 27.0 109.0 205.0 340 545 1 360
Maximum feed size (mm) 6.0 6.0 6.0 6.0 6.0 6.0 6.0
Footprint areaa (m2) 0.5 0.6 1.2 2.3 4.0 6.7 13.4
Installed power (kW) 1.0 1.5 - 3.8 5.5 - 7.5 11 - 22 30 - 56 30 - 75 150 - 375
Maintenance cost ($/t) $0.075 $0.075 $0.075 $0.075 $0.075 $0.075 $0.075
Maintenance cost ($/annum) $526 $1314 $5256 $9855 $16 425 $26 280 $65 700
Total installed capital cost $/(kg/h)
b
$8875 $3918 $1451 $1060 $828 $675 $540

a = Plant floor space taken up.

b = Includes instruments, automation, maintenance structure if required, placement in plant, connection of services and ancillary feed valves.
TABLE 12.10
Knelson CVD performance characteristics.

Model CVD 20 CVD 32 CVD 42 CVD 64

Maximum feed rate (t/h) 35.0 80 120 300
Maximum feed rate (m /h) 3
75.0 170 250 636
Maximum feed size (mm) 1.7 1.7 1.7 1.7
Footprint areaa (m2) 2.3 6.7 6.7 13.3
Installed power (kW) 11.0 30.0 30.0 - 38.0 75.0 - 150.0
Maintenance cost ($/t) $0.012 $0.015 $0.010 $0.010
Maintenance cost ($/annum) $3632 $10 584 $10 548 $25 838
Total installed capital cost b $/(t/h) $6829 $4575 $3008 $1877

a = Plant floor space taken up.

b = Includes instruments, automation, maintenance structure if required, placement in plant, connection of services and ancillary feed valves.

gold recovery. A summary of selected Knudsen bowl

data and costs is given in Table 12.11.
The in-line spinner builds on the low water use of a
Knudsen bowl and includes a vortex bar situated in the
feed bowl to promote fluidisation of concentrate along
with automatic cleaning. Selected in-line spinner data
and costs are summarised in Table 12.12.

Sluices and strakes

Sluices are used commercially for treatment of alluvials.
They are labour-intensive and inefficient so are not
recommended.
Strakes or corduroy tables are also labour-intensive
but could be considered for use on the mill discharge
or the circulating load in a gold circuit. The cloth is
normally changed up to twice a shift and washed to
recover the concentrate for amalgamation. In addition
to the high labour requirement, these devices represent
high security risks.
Designs for sluices and strakes are found in Taggart
FIG 12.9 - Falcon concentrator. (1945).

Cost Estimation Handbook 279

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

TABLE 12.11
Knudsen bowl performance characteristics.
Mineral type used in: Specifically designed to recover gold from alluvial or hard rock deposits, upgrading and recovery of gold
from concentrates from other gravity separation stages, exploration and evaluation of gold deposits, and
other mineral applications for tin, tungsten and scheelite
Maximum recommended feed rate (t/h) 3000
Maximum feed size (Mm) 8
Footprint area jig onlya (m2) 0.6
Installed power (kW) 0.37
Capital cost – unit $/(t/h) $4465

a = Plant floor space taken up.

TABLE 12.12 separations can only be successful for de-dusted, dry

In-line spinner performance characteristics. feeds at elevated temperatures.

Mineral type used in: Specifically designed to recover Feed characterisation

gold from in-line pressure jig Mineralogical analysis of the feeds to be processed is
concentrates and from other gravity
required to confirm the minerals present and to define
separation stages
the gangue associated with the valuable minerals.
Maximum recommended feed rate (kg/h) 2000 30 000 Minerals can be broadly grouped according to their
Maximum feed size (mm) 6 6 magnetic susceptibility as shown in Table 12.13. For
electrostatic separations it is important that particles
Footprint areaa (m2) 0.6 –
be suitably treated beforehand, because contaminants
Installed power (kW) 2.2 3 such as surface coatings can strongly influence the
Capital cost – unit $/(t/h) – – behaviour of particles.

a = Plant floor space taken up. Equipment description

– = Either not applicable or not available.
For convenient reference, the equipment available is
subdivided into groups corresponding to the broad
definition of minerals given in Table 12.13. Equipment
MAGNETIC AND ELECTROSTATIC SEPARATION is summarised in Table 12.14.
Magnetic and electrostatic separation techniques
exploit differences in magnetic and electrostatic Equipment selection
properties of particles in a feedstock. For magnetic While the equipment separation processes are usually
separation the property is magnetic susceptibility, based on metallurgical tests, some guidelines have
and for electrostatic separation the property is particle been derived from years of plant practice. These are
surface conductivity. summarised in this section.
To provide capital costs for magnetic and electrostatic
separation equipment, tests are conducted in order to Magnetic separators
confirm the: Some considerations for magnetic separation follow.
•• mineralogy of the feed •• If high-intensity separations are required, the
•• optimum circuit design separators must be protected from highly magnetic
material such as magnetite. Passing feed material
•• size range of the feed
through low-intensity separators separates the highly
•• optimum magnetic method (wet or dry) magnetic material, with the non-magnetic stream
•• tonnage rate to be processed. further treated on either high-intensity magnetic
These tests are conducted on sizeable samples (50 kg separators (WHIMS) for wet feed or induced roll
minimum). Equipment suppliers have test units set up magnetic separators (IRMS) for dry feed.
that enable performance to be assessed on production- •• Wet separations are only performed when either the
scale equipment. minerals to be recovered, or the gangue to be rejected,
In deriving these test data it is recognised that are strongly magnetic; refer to Table 12.13. The use
magnetic separation can be performed on either wet of WHIMS and low-intensity magnetic separators
(slurry) or dry feeds. Further, magnetic separation (LIMS) constitute primary separations and in many
equipment may have either low- or high-intensity cases either the magnetic or non-magnetic fractions
magnetic field strengths. High-tension (electrostatic) are further treated in dry circuits. If the minerals to

280 Cost Estimation Handbook

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

TABLE 12.13
Magnetic susceptibilities and electrostatic properties (Hunt, Maskowitz and Banerjee, 1995; Outotec, 2011a).

Mineral Magnetic susceptibility Field strength Electrostatic response

(Dimensionless SI units × 10-6) (Gauss, G) –
Arsenopyrite 3000 – Conductor
Biotites 1500 - 2900 10 000 - 18 000 Non-conductor
Cassiterite 1100 – Conductor
Celestite -16 - 18 – Non-conductor
Chalcopyrite 23 - 400 – Conductor
Chromite 3000 - 120 000 10 000 - 16 000 Conductor
Fayalite 5500 11 000 - 15 000 –
Franklinite 450 000 3000 - 5000 Conductor
Galena -33 – Conductor
Garnets 2700 12 000 - 19 000 Non-conductor
Goethite 1100 - 12 000 15 000 - 18 000 Non-conductor
Halite -10 - 16 – Non-conductor
Hematite 500 - 40 000 13 000 - 18 000 Conductor
Illite clay 410 – –
Ilmenite 2200 - 3 800 000 8000 - 16 000 Conductor
Iron 3 900 000 – Conductor
Jacobsite 25 000 – –
Lepidocrocite 1700 - 2900 – –
Limonite 2800 - 3100 16 000 - 20 000 Non-conductor
Maghemite 2 000 000 - 2 500 000 3000 - 5000 –
Magnetite 1 000 000 - 5 700 000 1000 Conductor
Montmorillonite clay 330 - 350 – –
Olivines 1600 11 000 - 15 000 Non-conductor
Pyrite 35 - 5000 – Conductor
Pyrrhotite Fe7S8 3 200 000 1000 - 4000 Conductor
Pyrrhotite Fe9S10 170 000 1000 - 4000 Conductor
Pyrrhotite Fe10S11 1700 – Conductor
Pyrrhotite Fe11S12 1200 – Conductor
Pyrrhotites Fe(1-x)S 460 - 1 400 000 – Conductor
Quartz -13 - 17 – Non-conductor
Serpentinite 3 - 75 000 4000 - 18 000 Non-conductor
Siderite 1300 - 11 000 10 000 - 18 000 Non-conductor
Sphalerite -31 - 750 – Conductor
Titanomaghemite 2 800 000 – –
Titanomagnetite 130 000 - 620 000 1000 - 3000 Conductor
Troilite 610 - 1700 – –
Ulvospinel 4800 – –

Notes: Field strengths refer to those strengths used on commercial equipment when making mineral separations; – = data not available.

Cost Estimation Handbook 281

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

TABLE 12.14
Magnetic equipment description.

Wet separation
Field strength Equipment term Configuration options Unit capacity Feed particle size
range (�m)
Low intensity Wet drum mag separator Single pass - -
500 - 1 500 (LIMS)
Double pass 15 - 20 t/h/m 20 - 1000
(gauss, G)
Either mags or non-mag retreat - -
High intensity mag Wet high intensity mag Single pass - -
separator separator (WHIMS)
68 mm width 15 - 25 t/h 20 - 850
1500 - 15 000 G 120 mm width rotors 30 - 50 t/h/machine -
10 000 -12 000 G SLon (brand) Single pass 5 - 150 t/h <13 000
Dry separation
Low intensity 500 - Dry drum mag separator Single pass - -
1500 G (DIMS)
Double pass 15 - 20 t/h/m 45 - 6000
Either mags or non-mags retreat - -
High intensity Rare earth (RE) drum mag Single pass 8 - 10 t/h/m 45 - 6000
1500 - 20 000 G separator (RE drum)
Double pass - -
Either mags or non-mags retreat - -
Rare earth roll mag Single pass 3 - 5 t/h/m 45 - 6000
separator (RE rolls)
Double pass - -
Either mags or non-mags retreat - -
Induced roll magnetic sep Double pass 2 - 4 t/h/machine 45 - 3000
(IRMS)
Triple pass - -
Either mags or non-mags retreat - -
Cross-belt magnetic sep Single pass 1.5 - 25 t/h/ 45 - 1000
machine
High-tension roll (HTR) 250 mm dia × 1524 mm length - -
separator
Single, double 4 - 6 t/h/machine 75 - 1000
Triple pass non-cond or cond retreat - -
Electrostatic plate 1524 mm length plates 2 - 3 t/h/machine 75 - 1000
separator (ESP)
Five pass non-cond retreat - -

be extracted as magnetics are weakly magnetic, then •• High-tension rolls are used as primary units with
dry separations are preferred. backup from electrostatic plates on roll middlings
The size range of minerals in the feed dictates to some and final cleaning stages.
extent the separation process selected: •• The feed requires elimination of fines or dust
particles (-75 μm).
•• wet separation is applicable from 1.0 mm to 20 μm
•• The grain surfaces of the minerals to be separated
•• dry separation is effective from 6.0 mm to 45 μm.
are to be free of staining and clay or slimes coatings,
Electrostatic separators which affect conductivity characteristics.
Some considerations for electrostatic separation follow. Tribostatic separators
•• The feed must be heated to temperatures of plus Tribostatic separators use the tendency for particles to
80°C to eliminate surface moisture, which affects take on an electrostatic charge when agitated together.
particle conductivity. Some will take on a positive charge and others a

282 Cost Estimation Handbook

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

negative charge. These charged particles then pass Mineral Assay (%) Magnetic response
between charged plates with the positively charged Quartz 3.0 Non-magnetic
particles attracted towards a negatively charged plate
Magnetite 10.0 Very magnetic
and the negatively charged particles attracted towards
the positively charged plate. Tribostatic separation Ilmenite 38.0 Magnetic
is only effective when the particles are dry and at Rutile 2.0 Weakly magnetic
temperatures greater than 100 - 110°C. Leucoxene 15.0 Non-magnetic
Cost estimation Zircon 14.0 Non-magnetic
The method for cost estimation is based on the Garnet 13.0 Weakly magnetic
knowledge of the unit capacity of the separation Hornblende 3.0 Weakly magnetic
equipment and estimating the number of separators Others 2.0 Non-magnetic
required at each stage of the defined circuit. The
Total 100.0
prices for each separator are obtained from equipment
suppliers (Table 12.15 shows indicative costs). The total Size analysis of the feed confirms that the fraction of
price for separation equipment is the sum of the costs +850 μm material is 2.0 per cent. Laboratory tests have
of units at each stage. shown that to maximise the extraction of magnetics
and minimise the loss of non-magnetics, the gauss
TABLE 12.15 levels required are:
Separation equipment costs – indicative only.
•• low intensity – 800 G
Magnetic separation Cost a of each (A$) •• first pass WHIMS – 8 000 G
•• second pass WHIMS – 13 000 G.
Low-intensity wet drums 125 000
(3 m length drums) – double pass The requirement is to reject the magnetic minerals
leaving the valuable rutile and zircon enriched in
Wet high intensity mag separator – 16 pile 425 000 the non-magnetics for further dry processing. The
model – narrow rotor width processing rate required is 50 t/h.
Wide rotor width 450 000 The plant circuit developed to treat this heavy mineral
Low-intensity dry drums concentrate is shown in Figure 12.10.
Rare earth drums (1 m length rolls) – 90 000 Important aspects of this circuit are:
double pass •• Prescreening – a two-stage screening circuit is
Rare earth rolls (1 m length rolls) – double 135 000 included to ensure that no +850 μm oversize enters
pass the magnetic separators. In particular the WHIMS
units are usually fitted with 2.5 mm air gap rotors;
Induced roll magnetic separators 150 000 hence the maximum particle size entering the
High tension rolls, 250 mm dia, 1524 mm length WHIMS should be one-third of the air gap, which
is 850 μm.
Double pass 40 000
•• Low-intensity wet drums – a level of ten per cent
Triple pass 50 000 magnetite in the feed requires two stages of drums
Electrostatic plates 75 000 to ensure maximum rejection ahead of the WHIMS
circuits. The first drum extracts the magnetics
a = These costs are included to provide working examples for this text. (magnetite) at 800 G with non-magnetics flowing to
Prices should be confirmed with equipment suppliers prior to definitive the second drum operating at 1500 G. Non-magnetics
estimates. from the second drum is WHIMS feed where the
magnetite content must be less than 0.5 per cent.
Once the total cost of separation equipment is
•• WHIMS – to ensure maximum rejection of magnetics,
established, then an order-of-magnitude cost of the
test work confirmed that two stages were required.
installed plant circuit is calculated using a factor
Using Reading’s equipment, a wide rotor (120 mm)
dependent on the type of circuit involved. The equipment
machine is used as the primary separator with a
cost may be assumed equal to 25 per cent of the installed
narrow motor (68 mm) as the cleaner unit.
cost if no more accurate information is available.
By reference to the costs detailed in Table 12.15, the
Worked examples costs of the separation equipment are estimated as:
Worked examples of costings for four types of magnetic Low-intensity wet drum
separation circuit are presented below. 950 dia × 3.0 mm wide $125 000
Wet magnetic separation 1 only 16 pole WHIMS -Wide Rotor $450 000
Assume that the mineralogy of the feed has been 1 only 16 Pole WHIMS -Narrow Rotor $425 000
established by testing as: Total $1 
000 
000

Cost Estimation Handbook 283

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

50 t/h HMC
HCM ex Wet Dry magnetic separation is usually part of a dry
Concentrator Plant
separation circuit that incorporates electrostatic
separation as well as magnetic separation. Rare
Feeding Hopper earth (RE) drum magnetic separators are now used
Conveyor extensively to produce ilmenite products with RE belt
and roll magnetic separators used to produce rutile and
50 t/h zircon products. Laboratory test work has confirmed
that the ilmenite can be fractionated into a number of
potential products including:
Trash Screen +2.0 mm

2.0 mm O/S •• primary ilmenite (very magnetic) – 50 - 54 per cent

TiO2
1 t/h
U/S •• secondary ilmenite (less magnetic) – 58 - 60 per cent
TiO2.
Trash Screen +0.85 mm To achieve these products, a plant circuit as shown
2.0 mm O/S in Figure 12.11 is installed, with a processing rate of
35 t/h. Important aspects of this circuit are:

Rejects
Low intensity Drums Mags Stockpile
950 ø x 3 m Width
Double Drum 5 t/h
RE Drums

44 t/h

1st Stage WHIMS Mags RE Rolls

1 x Reading 120 mm
Width Rotor - 16 Pole 19 t/h

Mags
N/Mags 25 t/h Mags
26 t/h Stockpile

2nd Stage WHIMS

N/Mags
1 x Reading 68 mm
Width Rotor - 16 Pole 7.0 t/h
18 t/h

Product Stockpile

FIG 12.10 - Wet magnet plant circuit – 50 t/h.

Applying a 25 per cent factor for equipment cost to

installed capital cost for the circuit gives:
Plant cost = $1 000 000 = $4 000 000
0.25 FIG 12.11 - Dry magnet plant circuit – 35 t/h.

Dry magnetic separation •• RE drum magnets – these units are very selective
The mineralogy of the feed has been established by and can fractionate ilmenite into various grade
assessing a conductor flow as: products by adjusting the drum speed. The drums
Mineral Assay (%) Magnetic response used are usually 400 mm diameter and can be
Quartz 1.4 Non-magnetic produced in two- or three-stage units. The first
stage is often a scalper to remove highly magnetic
Ilmenite 76.7 Magnetic
material. The magnetic strength of the RE magnets
Rutile 12.9 Non-magnetic is typically 6500 G.
Leucoxene 3.7 Weakly magnetic •• RE roll and belt magnetic separators – these
Zircon 2.0 Non-magnetic magnetic separators are stronger than the RE
Garnet 2.1 Weakly magnetic drum and are typically 7000 - 8000 G; however,
most of the additional strength is derived from
Monazite 0.1 Weakly magnetic
the tight magnetic gradient brought about by the
Others 1.1 Non-magnetic configuration of the magnet. The RE roll and belt
Total 100.0 can be used to scavenge ilmenite and reduce the iron

284 Cost Estimation Handbook

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

content of rutile and zircon products. Separation Monazite 1.0 Non-conductor

can be controlled by varying the roll speed and belt Others 4.0 Other
thickness.
Total 100.0
•• IRMS – to extract the remaining ilmenite and
Note: total proportion of the conductor material
produce low iron and zircon products, induced roll
is 68 per cent of the feed with the non-conductors
magnets operating at 14 000 - 16 000 G are used.
32 per cent.
Referring to the costs detailed in Table 12.15 the costs
of the separation equipment are estimated below. A The high-tension electrostatic circuit aims to produce:
rare earth drum (RED) and a rare earth roll (RER) are •• a conductor-enriched fraction which feeds to a dry
indicated: magnet circuit for ilmenite-rutile separation
2 × 1.5 m primary REDs 180 000 •• a non-conductor-enriched fraction which feeds to a
3 × 1.5 m RERs 405 000 zircon circuit for separation of zircon away from the
gangue minerals such as garnet.
5 only IRMS 750 000
The plant circuit, developed from test work, is shown
1 × 1.5 m RER 135 000
as Figure 12.12, with a processing rate of 50 t/h.
1 only electrostatic plate/rectifier 75 000
Total $1 
545 
000
50 TPH HMC
Applying a 25 per cent factor for equipment cost to
Primary H.T.
installed capital cost for the circuit gives: Roll Units
Plant cost = $1 545 000 = $6 180 000 8 only
N/Cond
Mid
0.25 30.0 TPH Cond
8.0
12.0 TPH
2.4 TPH
These estimates do not include the equipment to dry 98 TPH 14.4 TPH
the original heavy mineral concentrate and the costs Cond CL N/Cond CL
indicated refer to the separation circuit only. Roll Units Roll Units
6 only 2 only
Operating cost Cond
6.0
Mid
TPH
Mid N/Cond
A major component of the operating cost of a magnetic 3 TPH Cond
N/Cond
separation circuit is the power cost. As a permanent 2.0 TPH 9 TPH
magnet, RE magnets draw no power, except for the 13.0 TPH
28.0 TPH
motor drive of the roll or belt. As a result, RE magnets 1st Stage
Elect Plates
operate at a fraction of the cost of electromagnets. Cond 4 TPH
6 only
Taking into account maintenance and other costs, there
9.0 TPH N/Cond
is a large cost difference as shown in Table 12.16.
2nd Stage
3 TPH Elect Plates
TABLE 12.16 4 only
Cond
Magnetic separator operating cost for ilmenite (Outotec, 2011b). 6 TPH
N/Cond
Magnetic separator Operating cost (%)
Final Final
Cross-belt 100 Conductors Non-Conductors
35 TPH 15 TPH
Induced roll 50
Rare earth roll 15 - 25 FIG 12.12 - HT roller electrostatic plant circuit – 50 t/h.
Rare earth drum 15 - 20
Important aspects of this circuit are:
•• New generation high-tension rolls – these units
High-tension electrostatic separation incorporate an ionising wire electrode as well as a
The mineralogy of the feed (heavy mineral concentrate) semi-conductive plate electrode that improve the
separation and are used as part of an integrated
has been confirmed as:
electrostatic circuit. Each unit consists of three roll
Mineral Assay (%) Magnetic response passes with double sided machines being used.
Quartz 3.0 Non-conductor •• Electrostatic plates – these treat all the middling
llmenite 55.0 Conductor flows from the high-tension roll stages. High-
Rutile 10.0 Conductor tension middlings consist of coarse non-conductors
and fine conductors. This mixture of sized grains is
Leucoxene 3.0 Conductor
readily processed over electrostatic separators. Each
Zircon 14.0 Non-conductor unit consists of five plate passes using double sided
Garnet 10.0 Non-conductor machines.

Cost Estimation Handbook 285

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

•• Rectifiers – solid state rectifiers provide the potential This section provides the basic guidelines for
of 20 000 - 30 000 V to the HT electrode wire on the selecting equipment and estimating capital costs for
roll machines and the plate electrodes. These units these flotation processes.
convert 240 V AC supply to the 20 000 V DC for use
in the machines. Rectifiers are sized on the basis of Data requirements
4 mA required per HT roll (6 × 4 = 24 mA per roll Although economics is an important consideration,
separator) and 15 mA for each plate separator. ultimately the process design selection depends on
Referring to the equipment costs of Table 12.15, the detailed laboratory and pilot plant test work.
separation equipment costs for this circuit are: The mineralogy of the orebody needs to be well
•• primary HTR defined and compared with an existing operating mine
•• eight only 2 × 3 × 1.8 m × 270 mm diameter with similar ore. This gives an initial indication of the
machines $1 120 000 flotation process that is most likely to be successful. If
minerals have been conditioned effectively and float
•• complete with rectifiers
quickly and easily, then pneumatic flotation should be
•• conductor cleaner HTR investigated.
•• six only 2 × 3 × 1.8 m × 270 mm diameter machines However, the first step is the performance of bench-
$840 000 scale mechanical flotation tests to define the recovery
•• complete with rectifiers time curve for the valuable mineral. Concentrate
•• non-conductor cleaner HTR samples are initially collected over small time
•• two only 2 × 3 × 1.8 m × 270 mm diameter increments. This will indicate whether to select a
machines $280 000 mechanical or pneumatic machine.
•• complete with rectifiers If pneumatic flotation has been used previously for
similar ores then a small pilot plant column or Jameson
•• electrostatic plates
cell/G-Cell is run to confirm that these less-expensive
•• five only 2 × 5 stage × 1.8 m plates $375 000 processes are viable and to define the criteria needed
•• complete with rectifiers to size the full flotation circuit.
•• total $2 615 000 G-Cells are by definition pneumatic flotation cells and
Applying a 25 per cent factor for equipment cost to so have all of the advantages covered above, but differ
installed capital cost for the circuit gives: significantly in that the additional use of centrifugal
effect results in significantly better kinetics along
Plant costs = $2 615 000 = $10 460 000 with grade and recovery benefits. This automatically
0.25 translates into reduced capital and operating cost
compared to both mechanical cells and conventional
These estimates do not include the equipment to dry
pneumatic flotation cells.
and heat the heavy mineral concentrate and the costs
refer to the separation circuit only. The accelerated kinetics were evaluated by Eurus
Mineral Consultants (EMC) using SUPASIM flotation
FLOTATION modelling software from laboratory rate tests. EMC
found that eight tank cells of 19 minutes normal
Froth flotation is the most common beneficiation
residence time were equivalent to a single G-Cell. Air
process for the recovery of sulfide and oxide minerals
consumption on a total basis was also 50 per cent lower
containing copper, lead, zinc and nickel. Including
costs for concrete foundations, structural steel, flotation for the G-Cell compared to the tank cell bank.
equipment, pipework, electric instrumentation and There are large savings in capital costs if the
building costs, flotation accounts for between eight pneumatic processes or pneumatic hybrid circuits
and 18 per cent of the total project direct capital costs. are shown to give satisfactory results. Newell (1990)
An estimation of costs for different size plants is also gives good comparisons between cost for column and
outlined by Newell (1990). conventional cells. If Jameson cells are used in place
of the column, then additional savings are made by
The main froth flotation processes and equipment in
using column cells. Power requirements are much
use in Australia today are:
reduced because no air compressor is required. From
•• mechanical Newell’s (1990) paper and information supplied by
•• self-aerating cells – Wemco and Denver MIM Holdings Ltd – Marketing of Technology, the
•• forced draft – OK, Dorr Oliver and Agitair comparative cost summary in Table 12.17 is typical.
•• pneumatic Mechanical flotation
•• column with height to diameter ratio of 10:1
A rough estimate for a mechanical flotation circuit
•• Jameson cell with height to diameter ratio of 2:1 is made if the ore is similar in nature to an existing
•• Imhoflot G-Cells. operation for which the scale-up figures are defined.

286 Cost Estimation Handbook

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

TABLE 12.17 Conditioning can be done in the grinding circuit or in a

Flotation cell type cost comparison. mixed tank or in a high-intensity conditioner.
The conditioning retention time and the degree of
Mechanicala Column hybrid Jamesona mixing are also defined by test work. For high-intensity
Capital 197 100 8 conditioning, power requirements expressed as kW/m3
must be determined in the laboratory.
Operating 150 100 80
Power 276 100 60 Column or Jameson cell
The raw data required for column or Jameson cell
a = Costs relative to the central column, which is given as 100. scale-up is somewhat more extensive and difficult to
obtain than the raw data obtained from mechanical cell
Alternatively, and preferably, pilot plant tests are run laboratory tests. Finch and Dobby (1990) provide the
to define the retention time required in continuous basis for column scale-up.
operation to achieve optimum results. The test work should define the following terms to be
Kalapudas (1985) and other workers compared used for the scale-up:
results obtained from a laboratory batch test, pilot plant •• carrying capacity limit
and full-scale flotation plants. The nickel recovery and •• rate constants
preparation efficiency curve for Outokumpu Kotalahti
•• superficial gas rise velocity (Jg).
nickel-copper ore is shown in Figure 12.13.
Amenability tests are sometimes used to establish
that column or Jameson systems give better results
than mechanical cells. This involves establishing grade
recovery curves for both systems. Finch and Dobby’s
(1990) column flotation text gives some examples
where the column outperformed mechanical cells.
While residence time is of considerable importance
in mechanical cells and conventional columns, the
Jameson cell/G-Cell treats residence time somewhat
differently, preferring a design based on a number
of stages to achieve the required recovery. Increased
recovery in Jameson cells is achieved by increasing
the number of downcomers. This is due to the bubble-
particle collision-attachment-detachment mechanism,
which is different in each type of flotation cell.
FIG 12.13 - Ni recovery as a function of flotation time in the bulk
It is normal, using column cells and particularly with
flotation of Kotalahti Ni-Cu ore.
Jameson cells and G-Cells, to directly produce final
The optimum rougher circuit retention time is defined grade concentrate when the mineral is well liberated
when the recovery curve stops the steep increase or and free floating.
the separation efficiency curve starts to flatten out. To Due to the very different operating principles
achieve 88 per cent Ni recovery, the scale-up factor compared to mechanical tank cells, it is generally
for roughers is 10/4 = 2.5. For an overall recovery of preferred to size G-Cell plants from running smaller
95 per cent an additional nine minutes for the scavenger pilot plants. This is because the G-Cell cannot be
circuit is required. This gives an overall scale-up for successfully scaled down to bench scale due to the
roughers and scavengers of 19/8 = 2.4. unique aerator design. However, where this is not
possible, a method has been developed based on data
Kalapudas (1985) also found that in comparing results
from bench mechanical float tests. This method is
obtained from a four-cell bank with sizes of 0.5, 1.5, 3
described in detail by Imhof, Lotzien and Sobek (1993).
and 16 m3, the plant recovery curve was the same from
The method is based on applying a similar logic to that
all sizes. Thus, pilot plant recovery curves for 0.5 m3
used in a McCabe-Thiele diagram transposed onto a
cells can be used to directly scale-up to full-scale cells.
grade versus time curve.
Both bench-scale and pilot plant tests are needed to
define retention times and stages of flotation required Jameson cell operation
before mechanical cells can be selected. A Jameson cell is used widely as a complementary
If typical scale-up factors are known for particular technology to conventional mechanical flotation and
ore types then indicative retention times are calculated not as a replacement. Due to the ability to recover
from bench-scale tests. fast-floating, highly liberated mineral particles using
The type of conditioning required prior to froth a very short residence time, Jameson cells are most
flotation also needs to be defined by the test work. effectively used at the head of cleaning circuits and/or

Cost Estimation Handbook 287

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

straight after regrind circuits. Using the cells in these

locations means that the particles can be recovered at a
final concentrate grade, while also reducing the load on
downstream flotation units. Conventional mechanical
cells, which operate with much longer residence times,
can be used to recover the remainder of the mineral
particles that are less hydrophobic and slower floating.
An example of a flotation circuit using Jameson cells in
this configuration is shown in Figure 12.14.

Mechanical cell selection

If the process design parameters indicate that
mechanical cells are required and airflow control is not
an important variable, then the self-inducing air type
flotation cells are selected. These are the cheapest form
of mechanical cells; however, they provide the fewest
control options and least flexibility to control variable
ores. A flotation tank cell is shown in Figure 12.15.
If air is to be used in the control strategy then the
forced draft type mechanical cells are preferred. In this FIG 12.15 - Flotation tank cell to short-circuiting.
case, the cost of an air blower and piping needs to be
added to the cell costs. The number of cells in a bank is chosen to minimise
For both types of cells the calculation of the cell size short-circuiting. Construction of the flotation cell tank
and number of cells in the bank is the same. From the also affects short-circuiting. Cell types with short
tonnage to be treated, the volumetric flow to flotation is dividing walls will have more short-circuiting than
calculated. The retention time required for a particular cells with full dividing walls with small openings.
process is then determined. From this information the Lindsberg (1988) developed a curve based on data
total cell volume required is calculated. from operating flotation circuits that relates the number
It is important to determine the net cell volume by of cells in a bank with the retention time in each cell.
subtracting the mechanism volume occupied in the Thus if retention time is large in each cell in a bank,
tank. Some operators also subtract an allowance for air it may be assumed that little short-circuiting will take
entrainment. However, if the design data are based on place and fewer cells are required in the bank. Circuits
pilot plant work then the calculated cell size already that fall above the line in Figure 12.16 are found to have
has air entrainment built into the retention time value. minimal short-circuiting.

FIG 12.14 - Flotation circuit incorporating Jameson cells.

288 Cost Estimation Handbook

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

FIG 12.17 - Ideal mechanical cell flotation air blower curve.

FIG 12.16 - Residence time in one cell versus number of cells
(units) in relation. to control airflow over a reasonable valve movement
range. Often air valves are too big to vary the airflow
Typically, four cells are sufficient for a particular effectively. Constant flow blowers, such as ‘roots
flotation process, providing the retention time is blowers’, should never be used in flotation circuits
greater than five minutes in each cell. The curve shown as they cannot be used to vary airflow. The most
in Figure 12.16 allows for shorter banks of cells using economical blowers are single- and double-stage
larger cells than have been used in the past. centrifugal fan type blowers.
For cleaning circuits where the tailings from a Blowers can either be positive displacement (constant
particular stage are usually reprocessed, fewer larger flow) or centrifugal. Positive displacement blowers
cells can be used. The valuable mineral that has short- require a blow-off system for excess air or a variable-
circuited on the first pass can be collected when speed drive to control blower speed.
represented to the cleaner circuit.
Cost estimation
Mechanical cells that self-induce air use more power
The relationship shown in Figure 12.18 is based on
than forced draft cells. However, when the power
budget prices for Outotec flotation cells manufactured
consumed to produce air is added, the forced draft cells
in Australia. Cell costs include epoxy lining, internal
consume similar total power to the self-inducing cells.
dart valves and standard launders.
This will depend on the flotation cell manufacturer.
For self-inducing air cells, each mechanism disperses Worked examples
up to the minimum required quantity of air. Above this Three worked examples are presented.
figure, if additional air is forced into the mechanism,
large bubbles that adversely affect the flotation process Mechanical cells
are created. In sizing the air blower there should be no Assume batch and pilot flotation curves for a particular
safety margin applied to the maximum air requirement ore are the same as the curves shown in Figure 12.13,
nominated by the cell manufacturer. then:
Another factor to consider is that varying amounts •• rougher retention time = 10 min (88 per cent
of air are required for each stage of flotation. This will recovery)
reduce the total air requirement and also limit the size
•• scavenger retention time = 9 min (95 per cent
of the blower:
recovery)
•• for roughing – 50 - 90 per cent of dispersion capacity
•• solids feed rate = 500 t/h
•• for scavenging – 80 - 100 per cent of dispersion
•• pulp flow at flotation density = 1100 m3/h
capacity
•• rougher-scavenger flotation volume = 1100 × 19/60
•• for cleaning – 40 - 80 per cent of dispersion capacity. = 348 m3.
Note that the air requirement for flotation should be If one bank of cells at 50 m3 nominal size is assumed:
at constant pressure but control is needed to vary the
airflow over the total flotation circuit. The pressure of Net volume = 50 × 0.95 = 47.5 m3
the blower must be sufficient to overcome the line and
valve pressure losses and the slurry feed head. The Since the design is based on pilot plant work, air
air blower curve profile demonstrated in Figure 12.17 entrainment has been taken into account.
shows the type of air blower characteristics that should
No. of cells in the bank = 348/47.5 = 8 cells
be sought.
Care should be taken in sizing the air control valves From Figure 12.16, the eight cells with two-minute
in the air pipework so that they can be properly used retention time are acceptable and minimal short-

Cost Estimation Handbook 289

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

FIG 12.18 - Cost per volume for Outotec flotation cells.

circuiting will occur. Larger and fewer cells are not Based on Newell (1990) the cost of flotation
possible for this circuit. equipment is 0.7 to 0.5 times the cost of the total
•• from Figure 12.18, the cost of each 50 m3 cell is flotation plant including concrete, structural steel,
$8000/m3 platework, pipework, electrics and building costs and
basic equipment including the conditioner and blower.
•• therefore, 50 × $8000 = $400 000
Then the total flotation plant costs:
•• thus, the cost for eight cells = 8 × $400 000 = $3 200 000.
The volume of concentrate treated in the cleaning = $3.96 M/0.5 = $7.92 M.
circuit will depend on the feed grade of the ore.
Jameson cell
Assume first cleaner feed = 100 m3/h
Using Figure 12.14 for this example and assuming
second stage feed = 50 m3/h
rougher pulp flow rate is 1100 m3/h, MIM Holdings
To ensure that all slow floating minerals collected in Ltd – Marketing of Technology assumed the following
the rougher cells are refloated in each cleaning stage, typical design values:
the retention time must be longer than the rougher
•• a superficial gas rise velocity Jg = 1.3 cm/s
retention time of ten minutes.
•• for a staged recovery of 65 per cent, three stages are
Say both cleaning stages 15 min required to guarantee the 95 per cent recovery
then first cleaner volume = 25 m3 •• the carrying capacity of the cell is based on the
and second cleaner volume = 12 m3 volume of air input to the cell.
Because tailings are recycled to the head of the Then it is estimated that six 4 m diameter Jameson
rougher section then a two-cell group followed by a cells with ten 250 mm downcomers are required for
three-cell group of 8 m3 cells could be used. rougher scavenger duty.
From Figure 12.18: •• Budget price is $875 000
•• the cost per unit of 8 m3 cells = approximately For cleaning stages, the Jameson cell normally
$10 000 per m3 requires one stage only and one 2 m diameter Jameson
cell with three 200 mm downcomers is recommended.
•• therefore 8 × $10 000 = $80 000 per cell
•• Budget cost is $70 000
•• total cost for five cells = $400 000
No blower or compressor is required for Jameson
•• total cell equipment = $400 000 + $3 200 000 = cells.
$3 600 000.
•• Total cost for Jameson cells $945 000.
Cost of the air blower and conditioning tank is about
ten per cent of the flotation cells cost. Imhoflot G-Cell
Total cost of flotation equipment = 1.1 × $3.6 M Using the same data as in Figure 12.16 for a pulp flow of
= $3.96 M 1100 m³/h and mechanical rougher flotation residence

290 Cost Estimation Handbook

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

time of ten minutes, it is estimated that three G-Cells of • produce final product in specialised applications
4.8 m diameter would be required for roughing duty. (eg diamond).
The budget price for the cells including In Australia, ore sorting is currently, or has been,
instrumentation is $720 000. practiced at the following locations (the commodity
Imhoflot cells can typically produce a higher-grade and type of sorting are also given):
concentrate; therefore only one-stage cleaning is • Argyle – diamonds, X-ray
necessary. For cleaning duty, two G-Cells of 2.2 m • King Island – scheelite, photometric
diameter are required. The budget price for the cleaner • Mary Kathleen – uranium, radiometric
stage is then $150 000.
• Mt Carbine – wolframite, photometric.
• No blower or compressor is required for G-Cells
The beneficiation of an ore by sorting involves
• conditioning tank cost is the same as for mechanical inspection and recognition of the worth of each
flotation at approximately $300 000 particle followed by separation into either a valuable
• total cost for G-Cells is then $1 170 000 or worthless fraction. Hand sorting is still practiced
• again using the factor of flotation equipment in places where labour is cheap, other separation
accounting for 50 per cent of the flotation plant. or preconcentration techniques are not appropriate
Total G-Cell flotation plant is: $1 170 000/0.5 = $2.34 M. or the commodity is very high value such as gold or
diamonds. Increasing cost of labour and advances in
Lifetime costs of cells crushing, grinding and alternative separation processes
Lifetime operating costs of flotation equipment are have led to the decline in hand sorting.
often not considered as thoroughly as they deserve to Mechanical detectors can detect only one property
be, bearing in mind that the initial capital of a flotation at a time and the property must be present uniformly
cell is relatively small compared to operating costs over in the valuable particles to cause separation of the
the life of the project. Typical ownership costs over a selected particles. For sorting to be economic, recovery
FIG 12.19 in Figure 12.19.
25-year mine life are represented of the valuable mineral must be high and this high
recovery requirement has limited the success of sorting
operations. The sophistication of various sensing
devices and electronic technology has increased the
Initial investment costs (6%) number of successful sorting machines available.
Energy (68%)
Maintenance (7%) Sorting machines
Reagents (19%)
For successful sorting, the valuable mineral must be:
• essentially liberated from the gangue at coarse sizes
• consistently identifiable by a detector within the
FIG 12.19 - Breakdown of a large flotation cell expenses over a residence time available in the machine.
25-year life (Rinne and Peltola, 2008).
Ore sorters are usually classified on the basis of the
type of detection system used, which include:
Considering that energy consumption is the major
cost factor over the life of a flotation cell it would make • conductivity
sense then to try to reduce the power consumption • magnetic
as much as possible. One way to accomplish this is • optical and photometric (Figure 12.20)
through the use of a variable-speed drive (VSD). • radiometric
Summary • X-ray.
While mechanical flotation cells represent a proven
conventional technology, there are large potential
overall savings in capital and operating costs and
power consumption for pneumatic flotation cells,
especially Jameson cells and G-Cells. This situation
should encourage initial expenditure on laboratory
and pilot plant test work to evaluate these alternatives.

ORE SORTING
Ore sorting technology applications include:
• preconcentrate below cut-off grade material to
increase ore reserves
• reject coarse waste to reduce comminution costs FIG 12.20 - Colourimetric ore sorting machine.

Cost Estimation Handbook 291

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

In sorting machines, particles from 5 to 300 mm are The sensing method that can be used in a particular
analysed individually at a fixed rate of particles per situation depends on the ore to be treated. A wide range
second; hence higher tonnage rates are achievable with of mineral properties may be used in sorting. Types of
coarser particles. Any one machine handles particles sensors are presented in Table 12.18.
in approximately 2:1 or 3:1 ranges at tonnages from All these detection systems combine the various
25 to 180 t/h. For example, particles may be in the sensors with some electronic discriminating procedure.
range 10 to 25 mm at a rate of 30 t/h, or 75 to 175 mm Modern sorting machines use microprocessors
at 180 t/h. Maximum tonnage rates always depend on to interpret the sensor signal and developments
the material being presented to the sorter as well as the
in electronic hardware have led to the present
sorting precision required.
sophisticated ore sorting machines.
Sorting mechanics Ejection
A modern sorting operation has three distinct stages:
Ejection is the mechanical separation of the detected
1. singulation particles from the gangue particles of the ore. Separation
2. detection of the detected mineral particles may be achieved using
3. ejection. one of:
The overall recovery efficiency depends on the •• a solenoid plunger to push detected particles off the
successful completion of each stage. In general, the conveyor
singulation and ejection stages are rate determining, •• a solenoid-activated deflector plate to direct each
and the detector determines the separation efficiency. particle onto the proper waste or concentrate belt
Singulation •• solenoid-activated air valves to eject selected
particles while they are in free fall from the end of
Singulation is the control of the flow of feed so that
the sorting conveyor.
each particle is presented individually to the detector
for observation. Singulation techniques are classed as The two major requirements of an ejection system are
either: accurate timing and rapid recovery.
•• in-line – particles move in single file In free fall ejection systems, to cover the total range
of particle sizes from 10 to 175 mm, two separation
•• single-layer – particles move in a band of one layer
schemes are used:
in depth.
1. For 10 - 30 mm sizes, 80 air valves are used on
The single-layer method allows a much greater
presentation width of 800 mm. These valves have
capacity. However if the detection system requires
minimum open times of 3 - 4 ms, so that blast
scanning from all directions, the in-line method is
resolution is 10 mm across the presentation width,
normally used because it allows for scanning during
and 12 - 16 mm in the direction of rock motion (belt
free fall in air.
speed 4 m/s).
Detection 2. For 30 - 175 mm particles, 40 valves are used on
Detection is sensing the presence or absence of some width of 800 mm. The minimum open times of these
characteristic of the valuable mineral in the ore and larger units is approximately 7 ms, giving minimum
the electronic evaluation of the signal received. The blast length of 28 mm. This system has been shown
detection stage of an ore sorting operation comprises capable of separating 120 mm particles of sg 4.7.
a sensing device and an electronic signal processing Ultrasort model specifications for optical, radiometric
component. and electromagnetic ore sorters are given in Table 12.19.

TABLE 12.18
Ore sorter sensor types.
Sensor Mineral property Application
Optical Reflectance, fluorescence, transparency Base metals, precious metals, industrial minerals,
diamonds
Photometric Monochromatic reflection, absorption Industrial minerals, diamonds
Infra-red Reflection, absorption, heat conductivity-dissipation Base metals, industrial minerals, precious metals
Radiometric Natural or induced gamma radiation Uranium, precious metals
X-ray Atomic density-transparency, fluorescence Base metals, precious metals, coal, diamonds
Electromagnetic Conductivity Base metals

292 Cost Estimation Handbook

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

TABLE 12.19
Ultrasort model specifications.

Model ‘UFS’ Model ‘ULS’

Size range (mm) 5 - 80 40 - 300
Feed rate (t/h) Up to 80 Up to 300
Recovery (%) Up to 99 Up to 99
Grade (%) Up to 99 Up to 99
Weight (kg) 8000 30 000
Electric power type Single and three phase Single and three phase
Electric power (kW) Approximately 10 Approximately 12
Compressed air (m /t)
3
30 30
Compressed air type Blasted Blasted
Ejection system 80 (8 or 10 mm) or 120 (5 mm) ejectors 60 (14, 16 or 18 mm) ejectors

Cost estimation for their comments and review of the technical content
of that section. The author stresses that the opinions
Sorting systems generally offer lower capital and
expressed in that section are solely his own.
operating costs over alternatives for ore preconcen-
tration; for example, high frequency electromagnetic Both engineering and operating companies are in
ore sorting versus dense media separation (DMS). the debt of equipment suppliers for the service they
Operating costs for DMS can be up to $3.50/t at 200 t/h provide in giving budget prices and quotations for
compared to $1.75 - 2.50/t for ore sorting for a similar equipment for projects and studies that are frequently
capacity plant (Materials World, 2011). some distance from or may never reach fruition. The
Operating costs very much depend on location (and opportunity is seldom available to express gratitude for
associated water shortage and electricity costs). The this service and the response of suppliers to requests
dominant operating cost component in ore sorting for technical and pricing information for this text has
is compressed air at 30 - 50 m3/t. For a power cost of been excellent.
$0.04/kWh the compressed air cost is approximately Companies and individuals who provided
$0.7/t feed. Maintenance costs can be up to $0.35/t. information for ‘Gravity concentration’, along with
Ore sorting uses specialised equipment tailored to their locations and specialisations, are:
particular operations and hence generic cost information •• AMTAS Pty Ltd – Allan Russell
does not exist. The vendors should be contacted for cost
•• Bulk Materials Coal Handling) – Sydney (coal jigs)
information for particular applications. However, given
an estimate by the vendor, the capital cost factor of an •• Clyde Carruthers – Sydney (shaking tables)
ore sorting plant (ore sorter plus ancillary equipment) •• CPG Resources Mineral Technologies Pty Ltd –
is in the range of five to eight times the cost of the ore (spirals, Reichert cones, Kelsey centrifugal jigs,
sorter unit (Mular, Halbe and Barratt, 2002). Table 12.20 Knudsen bowls, shaking tables)
provides two indicative installed capital costs. •• David Mathieson and Associates – Sydney (Knelson
concentrators)
ACKNOWLEDGEMENTS
•• Denver Equipment – Sydney (coal jigs)
The author of the section on ‘Gravity concentration’
•• IHC Holland NV – Molendijk, Netherlands (mineral
thanks Graham Wylie (CPG Resources) and Rod Watts
jigs)
•• MET – Melbourne (spirals)
TABLE 12.20
Capital cost examples of two ore sorting plants. •• Peter Campbell – Gold Coast (mineral jigs).

Sorter Throughput Capex Project REFERENCES AND FURTHER READING

(t/a) (A$)
AMTAS Pty Ltd, 2011. Private communication.
Radiometric 2M $6.56 M Salamanca,
Berkley Resources, 2009. Announcement to the Australian
Spain (Berkley
Securities Exchange [online]. Available from: <http://www.
Resources, 2009) asx.com.au/asxpdf/20091202/pdf/31mhcmnb2w4tfn.pdf>
Electromagnetic 220 000 $4.3 M Jubilee, Western [Accessed: 11 January 2013].
Australia (2006) Campbell, P, 1991. Private communication.

Cost Estimation Handbook 293

 CHAPTER 12 – BENEFICIATION – CONCENTR ATION

Collier, D, Dwyer, F B, Thompson, R and Wulff, E, 1973. Materials World, 2011. Sorting out low-grade nickel ores
Ore sorters for asbestos and scheelite, in Proceedings [online]. Available from: <http://www.iom3.org/news/
International Conference on Mining and Metallurgy, London, sorting-out-low-grade-nickel-ores> [Accessed: 7 April
pp 1007-1022. 2011].
Finch, J A and Dobby, G S, 1990. Column Flotation, first edition Miller, V R, Nash, R and Schwaneke, A E, 1974. Detecting
(Pergamon Press: Oxford). and sorting disseminated native copper ores, US Bureau
French, R R, 1968. Beneficiation of low-grade gypsum by of Mines, report of investigations RI7904.
electronic color sorting, Transactions Society Mining Mular, A, Halbe, D and Barratt, D, 2002. Mineral Processing
Engineers, 241(Sept):331-334. Plant Design, Practice and Control Proceedings (Society for
Goode, J R, 1975. Recent coarse ore sorting advances, Canadian Mining, Metallurgy, and Exploration: New York).
Mining Journal, June, pp 49-50. Newell, A, 1990. Column flotation versus conventional
Goodman, R H and O’Keefe, K F, 1980. Tin ore treatment flotation, a comparison of costs, in Proceedings Mining
at Renison Ltd, Renison, Tas, in Mining and Metallurgical Industry Capital and Operating Cost Estimation Conference
(The Australasian Institute of Mining and Metallurgy:
Practices in Australasia: The Sir Maurice Mawby Memorial
Melbourne).
Volume (The Australasian Institute of Mining and
Metallurgy: Melbourne). Newman, P C and Whelm, P F, 1953. Photometric separation
of ores in lump form, in Recent Developments in Mineral
Guest, R N and Dunne, R C, 1985. An evaluation of gravity
Dressing, pp 359-381 (The Australasian Institute of Mining
separators by use of a synthetic ore, Journal of the South and Metallurgy: Melbourne).
African Institute of Mining and Metallurgy, June.
Private correspondence with flotation machine
Hunt, H, Maskowitz, B and Banerjee, S, 1995. Magnetic manufacturers.
properties of rocks and minerals, in Rock Physics and Phase
Rinne, A and Peltola, A, 2008. On lifetime costs of flotation
Relations (ed: T Ahrens), (AGU: Washington DC).
operations, Minerals Engineering, 21:846-850.
Imhof, R M, Lotzien, R and Sobek, S, 1993. Pneumatic
Robinson, R I, 1980. Tin-copper ore treatment at Cleveland
flotation: A reliable procedure for a correct plant layout,
Tin Ltd, Luina, Tas, in Mining and Metallurgical Practices
in Proceedings XVIII International Mineral Processing in Australasia: The Sir Maurice Mawby Memorial Volume
Congress, pp 971-978 (The Australasian Institute of Mining (The Australasian Institute of Mining and Metallurgy:
and Metallurgy: Melbourne). Melbourne).
Jenkinson, D E, Gough, P B, King, H G and Daykin, K W, Schapper, M A, 1977. Beneficiation at large particle size using
1974. Coal sorting by X-ray transmission, in Proceedings photometric sorting techniques, Australian Mining, April,
Tenth International Mineral Processing Congress, London, pp 44-53.
pp 1023-1041.
Smith, C, 1991. Private communication.
Jubilee Mining, 2006. Mined products made responsibly Taggart, A F, 1945. Handbook of Mineral Dressing, tenth
[online]. Available from: <http://www.infomine.com/ printing 1967, p 11-02 (John Wiley & Sons: New York).
publications/docs/InternationalMining/Chadwick2006j.
Wills, B A, 1985. Mineral Processing Technology, third edition,
pdf> [Accessed: 15 November 2011).
pp 522-530 (Pergamon Press: Oxford).
Kalapudas, R, 1985. A study on scaling up of laboratory batch
Wills, B A, 1989. Minerals Processing Technology, fourth edition
flotation data to industrial size flotation, in Proceedings XV
(Pergamon Press: Oxford).
International Mineral Processing Congress, France.
Woodcock, J T (ed), 1980. Mining and Metallurgical Practices
Kelly, E G and Spottiswood, D J, 1982. Introduction to Mineral
in Australasia: The Sir Maurice Mawby Memorial Volume,
Processing, pp 237-242 (John Wiley and Sons: New York). contains descriptions of hard rock and alluvial tin,
Laplante, A R, 2000. A standardized test to determine gravity tungsten, tantalum, beach sands and coal circuits
recoverable gold, McGill University. (The Australasian Institute of Mining and Metallurgy:
Lindsberg, R, 1988. Column flotation of various process Melbourne).
stages, SME Annual Meeting, Phoenix, Arizona.

294 Cost Estimation Handbook