View
222
Download
0
Category
Preview:
Citation preview
8/8/2019 Pages 1 57
1/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 1
1.0 I NTRODUCTI ON
Rio Tinto Zimbabwe Limited (RioZim) operates the Empress Nickel Refinery
(E.N.R) at Eiffel Flats, Kadoma, Zimbabwe [1]. It was built in 1968 to refine
nickel-copper matte from the Empress nickel mine using the Outokumpu
atmospheric leaching process. The mine and its associated concentrator and
smelter were closed down in 1982 after the economic mineral reserves had
been depleted. The refinery has been operating successfully as a toll refinery
since then, treating mainly matte from the Bamangwato Concession Limited
(B.C.L) operations in Botswana. This matte contains about 45-50% nickel, 40-
45% copper and 5-7% sulphur with the remainder made up of cobalt, iron and
arsenic. The refinery aims to maximize copper and nickel recoveries and to
produce nickel and copper cathodes that are satisfactory to the customers
requirements in terms of quality, quantity and size.
Sirosmelt slags are produced during the smelting of residues from
thickener 4 from the cementation and copper 1 leach circuit at E.N.R., the
slags being molten by-products of high temperature processes that are
primarily used to separate the metallic and nonmetallic constituents
contained in the bulk residues. Residue smelting is a process in which one
of the crucial roles slag plays, is the removal of impurities from the matte.
After a heat of the smelting process upon tapping, two different layers of
the slag and matte should ideally be formed due to differences in densities
8/8/2019 Pages 1 57
2/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 2
of the slag and matte, whereupon the matte is tapped onto the granulation
pit and the slag tapped into chill moulds. The matte being tapped should
ideally be free of any slag and the slag should also ideally be free of any
matte. However due to conditions prevailing in the furnace during any
heat, some of the matte gets entrained in the slag, while some matte is
oxidised due to the highly oxidising conditions that prevail in the furnace.
The oxidised matte reports to the slag. These occurrences are referred to
as metal-to-slag losses. It is envisaged that the total metal-to-slag losses
(i.e. total of Ni + Cu) should not exceed 1% upon assaying any sample of
Sirosmelt slag. However since the commissioning of the furnace high
metal-to-slag losses have been experienced, exceeding 1%. This has lead
to a lot of metal being lost with the slag being dumped at the slag dump at
E.N.R. To date the total of metal lost exceeds 250 tons (see appendix 1),
which is a loss in revenue exceeding U.S.$1,176,696,72. Thus against this
background the aim of this project is to design a suitable and cost effective
process to recover the Cu-Ni alloy from Sirosmelt slag at E.N.R.
The method considered for this project involves mineral processing since
it is a relatively inexpensive process with very little complications in the
process. Effective recovery of the alloy from slag involves separation of
the alloy from slag to acceptable recoveries. This can be achieved through
considering the differences in physicochemical properties between the
values and the gangue such as specific gravity, size, shape, colour, and
8/8/2019 Pages 1 57
3/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 3
electrical and magnetic properties. Such differences can then be
manipulated in the design of a suitable process for the physical separation
of the alloy from the slag. In order to determine such physicochemical
properties as mentioned above, characterization of the slag is necessary.
This involves a series of mineralogical analysis of the slag, such as
chemical analysis, sieve tests and metallographic analysis. Information
from the mineralogical analysis will then be used to design a preliminary
experimental flow sheet for laboratory scale, milling and physical
separation investigations to predict the behavior of the slag to such
techniques. Laboratory scale separation will be carried out. Results from
these tests will lead to design of a plant/pilot plant scale flow, diagram for
the milling of, and separation of the alloy from the slag.
8/8/2019 Pages 1 57
4/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 4
2. 0 LI TERATURE SURVEY
2. 1 Slag Characterizat ion
Slag characterisation involves investigations leading to an understanding of the
nature of the entrained alloy particles in the slag and their association with other
components in the slag [2]. This when accomplished can help in establishing the
slag to metal/alloy proportions, mode of occurrence of the entrained copper and
nickel and compositional variations within the slag stockpile. The dissemination
of the entrapped metal of interest within the slag matrix has a strong bearing on
the process technology to be used. Information from the study of slag
characteristics would help in predicting the degree of grinding required for
effective liberation of values from the gangue and effectiveness of the separation
methods in concentrating the Cu-Ni alloy.
The first stage in the characterization of the slag would involve chemical analysis
of the slag samples. Chemical analysis at Empress Nickel Refinery is carried out
for every slag-tap conducted after every heat of the Sirosmelt furnace. As the
slag is being tapped into chill moulds, a sample is taken and sent for analysis to
assay for different elements and other components in the slag and their relative
abundances. X-ray analysis is used in performing chemical analysis on the slag.
This is a non-destructive technique for confirming a samples elemental
composition. X-rays are bombarded on the sample, and a detector measures the
8/8/2019 Pages 1 57
5/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 5
secondary X-rays emitted from the sample. From these measurements a
computer can calculate a quantitative analysis of the sample.
Particle size analyses can also be carried out on slag samples. These are a
range of tests performed with the aim of determining the amount of crushing to
be performed on the slag in order to find an optimum degree of liberation of the
metal values from the gangue. These include sieve analysis and recovery versus
particle size analysis on the separation equipment of choice [3]. Sieve analyses
provide information on the size fractions to which the greatest proportion of metal
values report, leading to predictions of the optimum mesh of grind for effective
liberation of metal values from the gangue. A metal distribution curve will be
plotted for various size fractions to which the slag would have been ground [4].
Recovery versus particle size analysis shows the particle size for which the
greatest recovery is achieved for a particular particle separation process e.g.
gravity separation.
A slag characterization exercise would thus provide information on the degree of
comminution needed and the beneficiation methods necessary. It also shows the
nature of the final products to be obtained and what further processing is
necessary.
8/8/2019 Pages 1 57
6/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 6
2. 2 Part icle Size Analysis
Size analysis of the various products of a mill constitutes a fundamental part of
laboratory testing procedure [5]. In this project, it will be of great importance in
determining the quality of grinding and in establishing the degree of liberation of
the value metal alloy from the gangue at various particle sizes. In the separation
stage, size analysis of the products will be used to determine the optimum size of
the feed to the process for maximum efficiency and to determine the size range
at which any losses are occurring in the process, so that they may be reduced
[3].
The primary function of precision particle size analysis is to obtain quantitative
data about the size distribution of particles in the material [5]. However, the exact
size of an irregular particle cannot be measured. For a spherical particle, the size
is uniquely defined by its diameter, while for irregular particles the equivalent
diameter is often used. Recorded data from any size analysis should, where
possible, be accompanied by some remarks, which indicate the approximate
shape of the particles. Descriptions such as granular or acicular are usually
quite adequate to convey the approximate shape of the particle in question. A
short list of some of the more common methods of size analysis, together with
their effective size ranges is given in table 2.1 [5].
Test sieving is the most widely used method for particle size analysis. It covers a
very wide range of particle size; and will thus be used in this project for any
8/8/2019 Pages 1 57
7/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 7
particle size investigations. Sieve analysis is accomplished by passing a known
weight of sample material, successively through finer sieves, and weighing the
amount collected on each sieve to determine the percentage weight in each size
fraction. Sieving is carried out with wet or dry materials and the sieves are
usually agitated to expose all the particles to the openings.
Method Approximate useful range (microns)
Test Sieving 100000-10
Elutriation 40-5
Microscopy (optical) 50-0.25
Sedimentation (gravity) 40-1
Sedimentation (centrifugal) 5-0.05
Electron microscopy 1-0.005
Table 2.1: Some common methods of size analysis
In each of the standard series of sieves the apertures of consecutive sieves bear
a constant relationship to each other. It has been realized that a useful sieve
scale is one in which the ratio of the aperture widths of adjacent sieves is the
square root of two (i.e. 7KHDGYDQWDJHRIVXFKDVFDOHLVWKDWWKHDSHUWXUHareas double at each sieve, facilitating graphical presentation of results.
8/8/2019 Pages 1 57
8/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 8
There are several ways in which the results of a sieve test can be tabulated. The
methods (usually tabular) should show [3]:
1 The sieve size ranges used in the test.
2 The weight of material in each size range.
3 The weight of material in each size range expressed as a percentage of
the total weight.
4 The nominal aperture sizes of the sieves used in the test.
5 The cumulative percentage of material passing through the sieves.
6 The cumulative percentage of material retained on the sieves.
The results of a sieving test should always be plotted graphically in order to
assess their full significance. There are many different ways of recording results,
the most common being that of plotting cumulative undersize (or oversize)
against particle size. Many curves of cumulative oversize or undersize against
particle size are S-shaped as in fig. 2.1.
8/8/2019 Pages 1 57
9/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 9
2. 3 Milling Operat ions
Milling is carried out in order to prepare the slag from the slag dump, for
extraction of the valuable metal [5]. Apart from regulating the size of the slag, it is
a process of physically separating the grains of valuable metal from the gangue
minerals, to produce an enriched portion, or concentrate, containing most of the
valuable minerals, and a discard, or tailing, containing predominantly the gangue
minerals. The first stage in a milling process is comminution, whose major
objective is the liberation, or release, of the valuable mineral, and in this case,
alloy, from the associated gangue minerals by size reduction of the slag size to
the coarsest possible particle size. Comminution in its earliest stages is carried
out in order to make the freshly excavated material easy to handle by scrappers,
conveyors, and slag carriers [5]. Comminution in the mill takes place as a
sequence of crushing or grinding processes. Crushing reduces the particle size
of run-of-dump slag to such a level that grinding can be effected until the metal
alloy and the gangue are substantially produced as separate particles. Crushing
is usually a dry process, and is performed in several stages with small reduction
ratiosranging from three to six in each stage. The reduction ratio in crushing can
be defined as the ratio of maximum particle size entering to maximum size
leaving the crusher.
8/8/2019 Pages 1 57
10/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 10
2. 3. 1 Crushing Equipment
Crushing is the first mechanical stage in the process of comminution of the slag
material in which the main objective is the liberation of the valuable alloy from the
gangue [5]. The nature of machinery for a given crushing operation is influenced
by the nature of the product required and the quantity and size of material to be
handled. Lumps of slag to be fed can be as large as 50cm and these have to be
reduced in the primary crushing stage to 10-20cm in heavy-duty machines.
Crushing is normally done in open or closed-circuit depending on product size. In
open-circuit crushing, undersize material from the screen is combined with the
crusher product and is routed to the next operation. If the crusher is producing
ball-mill feed it is good practice to use closed-circuit crushing in which the
undersize from the screen is the finished product. Closed-circuit crushing has the
advantage of giving greater flexibility to the crushing plant as a whole.
Crushers are classified as primary or secondary crushers. In the primary stage
crushing of slag lumps from the slag stockpile, heavy-duty machines are used to
reduce the slag lumps to a size suitable for transport by conveyors and for
feeding the secondary crushers. Two suitable primary crushers are the jaw and
gyratory crushers, as illustrated in figs. 2.2 and 2.3 [5]. Jaw crushers range in
size up to1680mm gape by 2130mm width. This size of machine can handle slag
lumps with a maximum size of 1.22m at a crushing rate of approximately 725t/h
with a 203mm set. However, at crushing rates above 545t/h the economic
8/8/2019 Pages 1 57
11/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 11
advantage of the jaw crusher over the gyratory diminishes; and above 725t/h jaw
crushers cannot compete with gyratory crushers.
Jaw crushers may be divided into three main groups; the Blake, with a movable
jaw pivoted at the top, giving greatest movement to the smallest lumps, the
Dodge, with the movable jaw pivoted at the bottom, giving greatest movement to
the largest lumps, and the overhead eccentric, which is hinged at the top similarly
to the Blake, with the movable jaw suspended on the eccentric shaft. Jaw
crushers are applied to the primary crushing of hard materials and are usually
followed by other types of crushers. In smaller sizes they are used as single-
stage machines.
Figure 2.2 Cross-section through single-toggle jaw crusher
8/8/2019 Pages 1 57
12/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 12
The gyratory crusher consists of a cone-shaped pestle oscillating within a larger
cone-shaped mortar or bowl. The angles of the cone are such that the width of
the passage decreases towards the bottom of the working faces. The pestle
consists of a mantle, which is free to turn on its spindle. The spindle is driven
from an eccentric bearing below. Differential motion causing attrition can occur
only when pieces are caught simultaneously at the top and bottom of the
passage owing to different radii at these points. Crushing occurs through the full
cycle in a gyratory crusher, and this produces a higher crushing capacity than a
similar sized jaw crusher, which crushes only in the shuttling half of the cycle. For
this reason gyratories are often operated in parallel with a scalping grizzly
screen, provided the added cost of the screen is less than the cost of increased
crusher capacity.
Fig 2.3 A functional diagram of a gyratory crusher
8/8/2019 Pages 1 57
13/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 13
Secondary crushers are much lighter than the heavy-duty, rugged primary
machines. Since they take the primary crushed ore as feed, the maximum feed
size of slag material it can take will be ideally less than 15cm in diameter.
Secondary crushers also operate with dry feeds, and their purpose can be to
reduce the slag material to a size suitable for grinding. The bulk of secondary
crushing of metalliferous ores is performed by cone crushers, although crushing
rolls and hammer mills are used for some applications.
Another class of crushers in which comminution is by impact rather than
compression, by sharp blows applied at high speed to free-falling rock, exists.
This class is called impact crushers [6]. The moving parts are beaters, which
transfer some of their kinetic energy to the ore particles on contacting them. The
internal stresses created in the particles are often large enough to cause them to
shatter. These forces are increased by causing the particles to impact upon an
anvil or breaker plate. Examples of impact crushers include the hammer mill, the
impact milland the Tidco Barmac Crusher[7].
The Tidco Barmac crusher combines impact crushing, high-intensity grinding and
multi-particle pulverizing, and as such, is best suited in the tertiary crushing or
primary grinding stage, producing products in the 0.06-12mm size range. A
cross-section of the Duopactor, which can handle feeds of up to 650t/h, at a top
size of over 50mm is shown in fig. 2.4. The basic comminution principle
employed involves acceleration of particles within a special ore-lined rotor
8/8/2019 Pages 1 57
14/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 14
revolving at a high speed. Grinding will commence when rock enters the rotor,
and is thrown centrifugally, achieving exit velocities of up to 90 metres per
second. The rotor continuously discharges into a highly turbulent particle cloud
contained within the crushing chamber, where reduction occurs primarily by
rock-on-rock impact, attrition and abrasion.
Fig 2.4 Cross-section of a Barmac Duopactor Crusher (Courtesy www.teara.govt.nz)
8/8/2019 Pages 1 57
15/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 15
2. 3. 2 Grinding Mills
Grinding is normally the last stage in the process of comminution. In this stage
the particles are reduced in size by a combination of impact and abrasion, in
suspension in water. It is performed in rotating cylindrical vessels known as
tumbling mills[5]. These contain a charge of loose crushing bodies-the grinding
medium-which is free to move inside the mill, thus comminuting the slag
particles.
It is the purpose of the grinding section to exercise close control of the product
size. Thus investigations need to be carried out first to determine an optimum
mesh of grind in order to exercise close control of the product from any slag
grinding section. The optimum mesh of grind will depend on many factors,
including the extent to which the values are dispersed in the gangue, and the
subsequent separation process. The use of a grinding mill in the comminution of
slag form E.N.R. is envisaged to be on tailings from the first stage of gravity
concentration MZT1 (see fig. 3.1). Grinding is to be performed wet, being a
continuous process with material being fed at a controlled rat from storage bins
into one end of the mill and overflowing at the other end after a suitable dwell
time.
Tumbling mills are of three basic types: rod, rod and autogenous [6]. At Empress
Nickel refinery, ball mills (see fig. 2.5) are extensively used on the cementation
8/8/2019 Pages 1 57
16/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 16
leach circuit and on the siroleach circuit [1]. Thus a ball mill would be suitable for
grinding purposes on crushed slag from secondary crushers since such a mill
would be readily available and provisions for its use on slag recovery can be
made. A ball mill uses steel balls as the grinding medium. Closed circuit grinding
with high circulating loads would be preferable to open circuit grinding since the
former produces a closely sized end product and a high output per unit volume
compared with open circuit grinding. Grinding in a ball mill is effected by point
contact of balls and slag particles and given time, a desired degree of fineness
can be achieved.
Figure 2.5 A ball mill (courtesywww.mine-engineer.com)
8/8/2019 Pages 1 57
17/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 17
2. 4 Classif icat ion
Classification is a method of separating mixtures of minerals into two or more
products on the basis of the velocity with which the grains fall through a fluid
medium [5]. This fluid is usually water, and wet classification is generally applied
to mineral particles, which are considered too fine to be sorted efficiently by
screening. Since the velocity of particles in fluid medium is dependent not only on
the size, but also on the specific gravity and shape of the particles, the principles
of classification are important in mineral separations utilizing gravity
concentrators. Classifiers consist essentially of a sortingcolumn in which a fluid
is rising at a uniform rate. Particles introduced into the sorting column either sink
or rise according to whether their terminal velocities are greater or less than the
upward velocity of the fluid. The sorting column therefore separates the feed into
two products - an overflowconsisting of particles with terminal velocities less
than the velocity of the fluid and an underflowconsisting of particles with terminal
velocities less than the velocity of the fluid and an underflowor spigot productof
particles with terminal velocities greater than the rising velocity.
Many different types of classifiers have been designed and built which include
the settling cone(fig. 2.6), which is sometimes used as a dewatering unit in
small-scale operations [8]. It is often used in the aggregate industry to deslime
coarse sands products.
The rake-classifierutilises rakes actuated by an eccentric motion, which causes
them to dip into the settled material and to move it up the incline to the discharge.
Spiral Classifiers(fig. 2.7) use a continuously revolving spiral to move the sands
8/8/2019 Pages 1 57
18/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 18
up the slope. They can be operated at steeper slopes than the rake classifiers, in
which the sands tend to slip back when the rakes are removed.
Figure 2.7 Spiral classifier
The hydrocyclone(fig. 2.8) is a continuously operating classifying device that
utilises centrifugal force to accelerate the settling rate of particles. It is one of the
most important devices in the minerals industry [9], its main use in mineral
processing being as a classifier, which has proved extremely efficient at fine
separation sizes. It is widely used in closed circuit grinding operations but has
found many other uses such as de-sliming, de-gritting, and thickening. In the
recovery of alloy from E.N.R. slag, classification can be used before gravity
separation. Since gravity separators are extremely sensitive to the presence of
8/8/2019 Pages 1 57
19/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 19
slimes, which increase viscosity of the slurry and hence reduce the sharpness of
separation, and obscure visual cut-off points [5]. Settling cones preceding the
gravity device can be used to control pulp density within the circuit. For a
substantial increase in pulp density, hydrocyclones or thickeners may be used.
Figure 2.8 Hydrocyclone
8/8/2019 Pages 1 57
20/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 20
2. 5 Gravity Concent rat ion
Gravity concentration methods separate minerals of different specific gravity by
their relative movement in response to gravity and one or more other forces, the
latter often being the resistance to motion offered by a viscous fluid such as
water or air [5]. It is essential for effective separation of alloy particles from
gangue, that a marked density difference exists between the minerals and the
gangue. The concentration criteriongives the relative effectiveness of gravity
separation methods on a particular ore type.
).(.).(.
).(.).(.
mFluidMediuGSeralLighterMinGS
eralLighterMinGSeralHeavierMinGSonionCriteriConcentrat
=
When the quotient is greater than 2.5, whether positive or negative, then gravity
separation is relatively easy, the efficiency of separation decreasing as the value
of the quotient decreases. The motion of a particle in a fluid is dependent not
only on its specific gravity, but also on its size. Large particles will be affected
more than smaller ones. The efficiency of gravity processes therefore increases
with particle size, and the particles should be sufficiently coarse. In practice,
close size control of feeds to gravity processes is required in order to reduce the
size effect and make the relative motion of the particles specific gravity
dependant.
8/8/2019 Pages 1 57
21/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 21
2.5.1 Gravity Separators
It is essential for the efficient operation of all gravity separators that the feed is
carefully prepared through grinding, successive regrinding of middlings,
desliming and screening. One of the most important aspects of gravity circuit
operations is correct water-balance within the plant. Almost all gravity
concentrators have an optimum feed pulp-density, and relatively little deviation
from this density causes a rapid decline in efficiency. Accurate pulp-density
control is therefore essential, and this is most important on the raw feed [5]. This
can be achieved through the use of settling cones and such other thickeners,
preceding the gravity device. A selective but comprehensive review of gravity
separators, which are best suited for purposes of recovering alloy form ground
slag, is given in the following sections.
2.5.1.1 Spirals
Spiral concentrators find many varied applications in mineral processing. A spiral
is composed of a helical conduit of modified semicircular cross-section. Feed
pulp of between 15-VROLGVE\ZHLJKWDQGLQWKHVL]HUDQJHPPWR PLV
introduced at the top of the spiral and, as it flows spirally downwards, the
particles stratify due to the combined effect of centrifugal force, the differential
settling rates of the particles and the effect of interstitial trickling through the
flowing particle bed. These mechanisms are complex being much influenced by
the slurry density and particle size. It is believed in some academic quarters that
8/8/2019 Pages 1 57
22/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 22
the main separation effect is due to hindered settling, with the largest densest
particles reporting preferentially to the concentrate, which forms in a band along
the inner edge of the stream (fig. 2.9) [10].
Ports for the removal of the higher specific gravity particles are located at the
lowest points in the cross-section. Wash-water added at the inner edge of the
stream, flows outwardly across the concentrate band. The width of the
concentrate band removed at the ports is controlled by adjustable splitters. The
grade of concentrate taken from descending ports progressively decreases,
tailings being discharged from the lower end of the spiral conduit. Spirals are
made with shapes of varying steepness, the angle affecting the specific gravity of
separation, but having little effect on the concentrate grade and recovery.
Figure 2.9 Cross-section of spiral stream
8/8/2019 Pages 1 57
23/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 23
2.5.1.2 Shaking Tables
The principle behind the shaking table is that when a flowing film of water flows
over a flat, inclined surface the water closest to the surface is retarded by the
friction of the water absorbed on the surface; the velocity increases towards the
water surface [5]. If mineral particles are introduced into the film, small particles
will not move as rapidly as large particles, since they will be submerged in
slower-moving portion of the film. Particles of high specific gravity move more
slowly than lighter particles, and so a lateral displacement of the material will be
produced (fig. 2.10). The flowing film effectively separates coarse light particles
from small dense particles, and this mechanism is utilised to some extent in the
shaking-table concentrator (fig. 2.11).
Figure 2.10 Action in a flowing film
8/8/2019 Pages 1 57
24/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 24
Figure 2.11 Shaking table
The shaking-table concentrator consists of a slightly inclined deck, A, onto which
feed at about 25% solids by weight, is introduced at the feed box and is
distributed along C; wash water is distributed along the balance of the feed side
from launder D. The table is vibrated longitudinally by the mechanism B, using a
slow forward stroke and a rapid return, which causes the mineral particles to
crawl along the deck parallel to the direction of the motion. The minerals are
thus subjected to two forces-that due to the table motion and that at right angles
to it due to the flowing film of water. The net effect is that the particles move
diagonally across the deck from the feed end and, since the effect of the flowing
film depends on the size and density of the particles, they will fan out on the
table, the smaller, denser particles riding highest towards the concentrate
launder, which runs along the length of the table. Fig. 2.12 shows an idealized
diagram of the distribution of table products. An adjustable splitter at the
8/8/2019 Pages 1 57
25/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 25
concentrate end is often used to separate this product into two fractions: a high-
grade concentrate and a middlings fraction.
Particle size plays a very important role in table separations; as the range of
sizes in a table feed increases, the efficiency of separation decreases. If a table
feed is made up of a wide range of particle sizes, some of these sizes will be
cleaned inefficiently. Thus since the shaking table effectively separates coarse
light from fine dense particles, it is common practice to classify the feed since
classifiers put such particles into the same product, on the basis of their equal
settling rates. This classification can be effected through use of classifiers such
as multi-spigot hydrosizers, and such other classified as described in section 2.4.
The capacity of a table varies according to size of feed particles and the
concentration criteria. Tables can handle up to 2 tonnes per hour of 1.5mm sand
and perhaps 1t/h of fine sand. On 100- PIHHGPDWHULDOVWDEOHFDSDFLWLHV
may be as low as 0.5t/h.
The quantity of water used in the feed pulp varies but for ore tables normal feed
dilution is 20-25% solids by weight. In addition to the water in the feed pulp, clear
water flows over the table for final concentrate cleaning. This varies from a few
litres to almost 100 litres/min according to the nature of the feed material.
8/8/2019 Pages 1 57
26/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 26
Tables slope from the feed to the tailings discharge end and the correct angle of
incline is obtained by means of a hand wheel. The table is slightly elevated along
the line of motion from the feed end to the concentrate end. The correct amount
of end elevation varies with feed size and is greatest for the coarsest and highest
gravity feeds. Normal end elevations in ore tabling range from a maximum of
90mm for a very heavy coarse sand to as little as 6mm for an extremely fine
feed.
2.5.1.3 The Mozley Laboratory Separator
This flowing film device (fig. 2.14), which uses orbital shear, is now used in heavy
mineral processing laboratories, and is designed to treat small, samples (100g of
ore) [11]. The separator consists of a v-shaped stainless steel tray measuring
128cm in length, 72cm in width, 91cm in height and weighing 150kg. Below the
tray is the shaking mechanism, which causes the tray to vibrate longitudinally at
amplitude of inch at a frequency of 120 to 240 rpm. The transverse cyclic
oscillation is at 60 to 120 rpm. At one end of the tray is the feed cone and
concentrate wash water pipe, while an irrigation water pipe runs along the
length of the tray with holes perforated on the pipe at equal separations, from
which jets of water impinge on the sample. At the other end of the tray is the
tailings launder for collection of tails. The tray is sloped at 1-50 to the horizontal
from the feed end. This laboratory piece of equipment has the advantage that its
V profile tray with end knock, when treating closely sized material, is capable
not only of duplicating heavy liquid analysis results, but of giving additional data
8/8/2019 Pages 1 57
27/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 27
in the higher specific gravity ranges. It is also able to predict shaking table
performance when treating hydraulically classified products. The separator is
therefore useful in carrying out release analysis and for prediction of reaction of
feed to shaking table concentrator, and the efficiencies of such processes in
gravity separation [5].
Figure 2.14 Schematic diagram of a Mozley laboratory separator
2. 5. 1. 4 Cent rif ugal Concent rators
A number of centrifugal gravity separation devices, designed to treat ultra-fine
particles, are available for the gravity separation of a mixture of ground alloy and
8/8/2019 Pages 1 57
28/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 28
gangue in slag. The Mozley Multi-Gravity Separator (MGS) shall be considered in
this section.
The principle of the MGS can be visualized as rolling the horizontal surface of a
conventional shaking table into a drum, then rotating it so that many times the
normal gravitational pull can be exerted on the mineral particles as they flow in
the water layer across the surface [12]. Fig. 2.15 shows a cross-section of the
pilot scale MGS.
Figure 2.15 Pilot scale MGS (Courtesy Natgroup)
The plant-scale MGS consists of two slightly tapered open-handed drums,
mounted back to back, rotating at speeds variable between 90 and 150rpm,
enabling forces of between 5 and 15g to be generated at the drum surfaces. A
sinusoidal shake with an amplitude variable between 4 and 6cps is
superimposed on the motion of the drum; the shake imparted to one drum being
balanced by the shake imparted to the other, thus balancing the whole machine.
A scraper assembly is mounted within each drum on a separate concentric shaft,
driven slightly faster than the drum but in the same direction. This scrapes the
8/8/2019 Pages 1 57
29/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 29
settled solids up the slope of the drum, during which time they are subjected to
counter-current washing before being discharged as concentrate at the open,
outer, narrow end of the drum. The lower density minerals, along with the
majority of the wash water, flow downstream to discharge as tailings via slots at
the inner end of each drum.
2. 6 Assessment of Met allurgical Ef f iciency
In the concentration of alloy in the Sirosmelt slag, the recoveryis defined as the
percentage of the total metal contained in the slag that is recovered in the
concentrate. Thus a recovery of 90% will mean that 90% of the alloy in the slag is
recovered in the concentrate and 10% lost to the tailings. Recovery is usually
expressed in terms of the valuable end product. The ratio of concentrationis the
ratio of the weight of the feed to the weight of the concentrates. It is a measure of
the efficiency of the concentration process and it is closely related to the gradeor
assayof the concentrate. The value of the ratio of concentration will generally
increase with the grade of concentrate. The grade, or assay, usually refers to the
content of the marketable end product in the material.
The enrichment ratiois the ratio of the grade of concentrate to the grade of the
feed and again is related to the efficiency of the process. Ratio of concentration
and recovery are essentially independent of each other, and in order to evaluate
a given operation it is necessary to know both. There is an approximately inverse
relationship between recovery and grade of concentrate in all concentrating
8/8/2019 Pages 1 57
30/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 30
processes. If an attempt is made to attain a very high-grade concentrate, the
tailings assays will be higher and the recovery low. If a high recovery of metal is
aimed for, there will be more gangue in the concentrate and grade of concentrate
and ratio of concentration will both decrease. It is impossible to give figures for
representative values of recoveries and ratios of concentration. The aim of milling
operations is to maintain the values of ratio of concentration and recovery as high
as possible, all factors being considered.
Since concentrate grade and recovery are metallurgical factors, the metallurgical
efficiencyof any concentration operation can be expressed by a curve showing
the recovery attainable for any value of concentrate grade. A typical recovery-
gradecurveshowing the characteristic inverse relationship between recovery
and concentrate grade is shown in fig. 2.16.
Fig. 2.16 Recovery and concentrate grade
8/8/2019 Pages 1 57
31/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 31
Concentrate grade and recovery, used simultaneously, are the most widely
accepted measures of assessing metallurgical performance; thus the two
measures will be used in this project to assess metallurgical performance of
gravity separation processes to recover alloy from slag. These two measures can
be combined into a single index defining metallurgical efficiency of the
separation:
gm RRES =..
Where S.E.=Separation efficiency
Rm = % recovery of the valuable mineral
Rg = % recovery of the gangue into the concentrate
Ff
Cc
Rm%100
=
The gangue content recovery of the concentrate = ( )%100100 mc
Where m is the percentage metal content of the valuable mineral, i.e.
Gangue content = ( ) mcm 100 ,
Rg = C x gangue content of concentrate/gangue content of feed
8/8/2019 Pages 1 57
32/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 32
Rg =( ) fmcmC 100
Therefore, ( ) ( ){ }fmcmCf
CcRR gm
=
100
100
( )
( )ffm
fcCmRR gm
=
100
8/8/2019 Pages 1 57
33/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 33
3. 0 EXPERI MENTAL I NVESTI GATI ONS
3. 1 Methodology
Firstly, a series of mineralogical analysis of the slag, such as chemical
analysis, sieve tests and metallographic analysis, will have to be conducted.
Mineralogical analysis of the slag, which involves characterization of samples
of the slag, was carried out, aimed at establishing the following:
a) Slag to metal/alloy proportions,
b) Mode of occurrence of entrapped copper and nickel,
c) Compositional variations between samples.
It was hoped that this exercise would also offer an opportunity to investigate
the degree of grinding required for effective liberation of the values from the
gangue, and effectiveness of the separation methods in concentrating copper
and nickel [13]. Information from the mineralogical analysis will then be used
to design a preliminary experimental flow sheet for laboratory scale, milling
and physical separation investigations to predict the behaviour of the slag to
such techniques. Laboratory scale separation will be carried out. Results from
these tests will lead to design of a plant/pilot plant scale flow diagram for the
milling of, and separation of the alloy from the slag.
8/8/2019 Pages 1 57
34/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 34
3. 2 Chemical Analysis
Chemical Analysis was done to establish the elements in the slag and their
relative abundances. This was accomplished through collection of statistics from
E.N.R. Sirosmelt plant assay record sheets for the period from 01/09/2002 to
25/09/2002 from the daily product accounting samples and the average assays
for different components from the slag-tap were calculated. The results are given
in 4.1.
3.3 Metallographic Analysis
Two polished sections were prepared from slag samples by being first ground to
flat surfaces, then being successively ground and polished until a mirror finish
was obtained for each surface. The sections were studied using a Zeiss reflected
light microscope and photographs were taken. Results obtained which are meant
to establish the mineralogy of the slag, are shown in 4.2.
3. 4 Part icle Size Analysis
A 4.5kg sample of the slag was crushed in a laboratory jaw crusher then ground
to passing 1mm. A sieve analysis was carried out in the size range plus 1000m
to minus 53m. The sieves used were the following: 1000m, 850m, 500m,
425m, 355m, 212m, 125m, 75m, 53m aperture sieves. Each of the 9 size
fractions was assayed for copper and nickel content and the metal distribution
was computed. The results of the metal distribution are given in section 4.3.
8/8/2019 Pages 1 57
35/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 35
3. 5 Recovery versus Part icle Size Analysis
A Mozley Multi-gravity Laboratory Separator was chosen as the technology for
separation. The operation of this separator is explained in 2.4.1.3. 100g of slag
were crushed for different size fractions in the range minus 1000 to plus
53microns and collected using sieves. The sieves used were the following:
1000m, 850m, 500m, 425m, 355m, 212m, 125m, 75m, 53m aperture
sieves. The 100g sample of slag crushed to a particular size was placed on a
Mozley Multi-gravity Laboratory Separator and wetted. The cyclic motion of the
tray mobilised the slag particles enabling stratification to take place. The heavy
metal particles sank to the tray surface and were thrown upstream by the end-
knock action. The lighter (gangue) mineral particles were carried downstream by
the flow of irrigation water to discharge via the tailings launder.
Interpretation of the data was carried out by assay analysis of the separator
products. For each particle size fraction, concentrates were collected and sent for
assaying for Cu and Ni. The assay results were used to calculate the recovery
(see appendix 3) for each particle size fraction and recovery versus particle size
curves drawn (see 4.4).
3. 6 Gravit y Concent rat ion Test work on Sirosmelt Slag
Samples were collected from different parts of the slag dump in varying amounts.
The samples were crushed in a laboratory jaw crusher and ground in a cone
crusher to pass the 850m screen. The grinding product was mixed
8/8/2019 Pages 1 57
36/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 36
homogeneously by coning and quartering to obtain a 3kg sample. A Mozley
Multi-gravity Separator Table was used for gravity separation (see 2.4.1.3). A
flow sheet was first designed for the gravity separation program to be followed.
This was done in order to achieve the maximum recoveries possible taking into
account that a satisfactory recovery cannot be possible to achieve in one pass.
Thus the sample would have to be subjected to a series of gravity separation
process runs through the Mozley Gravity Separator Table. Thus a rougher-
scavenger-cleaner-recleaner type of flow sheet (fig. 3.1) was designed, taking
into account its efficiency in achieving high recoveries without compromising the
grade of the concentrate.
A 3kg sample of crushed slag was fed dry into the feed cone at one end of the
Mozley Multi-gravity Separator, (MMGS) table under a running stream of water to
form a pulp of about 25% solids. The MMGS table was turned on and
simultaneously the water through the irrigation and concentrate pipes was
turned on. The feed cone was unscrewed to let the feed flow as a pulp onto the
table. The pulp was sprayed with a stream of water from the irrigation and
concentrate wash pipes to enhance flow. The table was sloping towards the
tailings end at approximately 5o. The sample was subjected to the gravity
separation action for one minute after which the machine was switched off. The
slag ground to passing 850m and was fed to the MMGS table. The concentrate
from the rougher separation run, MZT1, was collected, dried and weighed and
was subjected to another run which is the cleaner separation run, MZT2 (after a
8/8/2019 Pages 1 57
37/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 37
small measured amount was taken for assaying purposes). The concentrate from
MZT2 was taken for the recleaner separation run, MZT4 to give separate products
of concentrate and tailings. The MZT2 tailings were collected, to be re-fed to
MZT1 together with fresh feed, while the concentrates were treated as the final
concentrate, dried and weighed and taken for assaying. The tailings from MZT1
was further ground to 80% passing 125m and subjected to a scavenger
separation run on the MMGS, MZT3 under the same conditions as described
above. The tailings from MZT3 were collected as the final tailings, dried and
weighed and taken for assaying while the concentrate was collected, to be re-fed
to MZT1 together with fresh feed. Two products were obtained for each run with
the lighter gangue minerals being collected at the tailings end while the heavier
value metal was obtained at the concentrate end. The products were decanted
and oven dried at 105oC for 2hours. A sample was extracted from each product
and taken for assaying. The results for the above test work are shown in 4.5.
8/8/2019 Pages 1 57
38/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 38
Fig 3.1: Schematic experimental flow sheet for the concentration of alloy from slag
Slag Feed
850m
MZT1
MZT2Grind to 80% passing 125 m
MZT3MZT4
concentrate tailings
concentrate
Final tailing
tailings
Finalconcentrate
concentrate
tailings
8/8/2019 Pages 1 57
39/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 39
3. 7 Spiral Concent rat ion Test work on Sirosmelt Slag
A 20kg sample of crushed Sirosmelt slag was mixed with water to form a pulp of
density 40% solids and was pumped into a spiral concentrator, at the top. As the
pulp flowed spirally downwards stratification of particles due to centrifugal force
occurred. Two separate products were obtained at the concentrates and tailings
ports. The products were dried and weighed and taken to assay for copper and
nickel. The results obtained are summarized in 4.6.
8/8/2019 Pages 1 57
40/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 40
4. 0 RESULTS AND DI SCUSSI ON
4.1 Chemical Analysis
TABLE 4.1:CHEMICAL ANALYSIS OF SAMPLES OF THE SLAG
%Ni %Cu %S %Fe %FeO %Co %CaO %SiO2 %Al2O3 %MgO %Cr2O30.97 1.12 0.10 10.06 12.06 0.36 28.80 26.70 11.74 5.70 1.84
The results above show that the average assays for Cu and Ni are above
the expected limits of a total of 1% for both, at 0.97% and 1.12%
respectively, totaling 2.09%. This high metal-slag loss situation can be
remedied through a mineral processing technique.
4. 2 Metallographic Analysis
The results of metallographic analysis of slag samples are shown in figs. 4.1a
and 4.1b.
Figs 4.1b Fig. 4.1b
8/8/2019 Pages 1 57
41/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 41
Slag samples showing different phases present in the slag. [Magnification: 50x]
After careful study and analysis of the phases shown above, the following
deductions were made:
i. Most of the copper and Nickel in the slag occurs as an alloy within matte
granules. Minor hosting phases of copper and nickel include chalcocite
(Cu2S), cuprite (Cu2O), tenorite (CuO), bunsenite (NiO) and covellite (CuS),
in order of decreasing abundance [14].
ii. Generally, Au, Pd, and Pt correlate with Cu and Ni. This indicates that the
three are associated with Cu and Ni in the alloy. This could be confirmed by
microbe analysis.
iii. Slag fragments make up the ultimate majority of the samples. The slag
particles are mainly massive silicate intercepted by carbonate and iron
oxide, mainly magnetite. Fine matte particles (< 75m) occur as rounded
globules in relatively coarse fragments of slag.
iv. Majorities of the entrained matte particles have clear margins and only rare
cases of coating were observed. The coatings are broken rims of cuprite
and/or magnetite. The rest of the magnetite occurs mainly as spherical and
rod-shaped globules and less commonly as euhedral particles in the slag.
8/8/2019 Pages 1 57
42/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 42
v. Free matte granules occur as rounded to subangular particles, most of them
free of oxidation rims. Majority range in size from 600 to 1000m, average
size 300m. Patches of cuprite were visible inside slag. Finer particles of
cuprite,
8/8/2019 Pages 1 57
43/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 43
From the results shown, it was observed that metal distribution in the various size
ranges shows that 92.77% of the copper and 91.09% of the nickel is contained in
the size fraction minus 1000 to plus 125microns. This can be seen from the
graph in fig. 4.2 where the size range is the steepest part of the graph. The
effectiveness of any separation process outside this range becomes less.
Particles greater than 1000micronsconstitute 49.62% of the copper and 37.45%
of the nickel respectively, while particles less than 125micronsconstitute 7.23%
of the copper and 8.91% of the nickel. Therefore to effectively liberate the
particles and still save on comminution costs the particles should be ground in
the above size range.
4. 4 Recovery versus Part icle Size Analysis
Fig 4.3: Recovery Vs Particle Size for Copper
0
10
20
30
40
50
60
70
80
90
1000 850 500 355 212 125 75 53 53-
size (microns)
Recovery(
%),Grade(%)
grade
recovery
8/8/2019 Pages 1 57
44/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 44
The performance of the separator varied with the feed size as shown on the
graph in fig. 4.3 and fig 4.4. Recoveries of copper and nickel by size show that
plus 60%recoveries occur between minus 850micronsto plus 125microns. The
effectiveness of the gravity separation process falls for particles below
125micronsdue to loss of values in fines at fine sizes. The fall at 500micronsis
probably due to loss of values with gangue due to little unlocking at that size for
metal particles entrained in the slag at sizes lower than 850microns. This shows
that in a circuit to be designed for the recovery of metal values from E.N.R. slag,
the first gravity separation run should be performed on particles ground to
Fig 4.4: Recovery vs Particle Size for Nickel
0
10
20
30
40
50
60
70
80
90
1000 850 500 355 212 125 75 53 53-
size (microns)
recovery(%),grade(%)
grade
recovery
8/8/2019 Pages 1 57
45/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 45
passing 850microns, whereupon the tailings from this run would be further
ground to passing 125micronsthen subjected to another gravity separation run.
4. 5 Gravit y Concent rat ion Test work on Sirosmelt Slag
The results are summarised in the experimental flow sheet in fig. 4.5
Generally high recoveries in the concentrate were obtained for each of the
stages above. The final separation stage MZT4 gave the highest recoveries of
Nickel and Copper in the concentrate, of 88.71% and 92.34% respectively. At
this stage it became unnecessary to continue with further separation as high
enough recoveries were obtained. A high grade of the alloy, totalling 85.62% was
also obtained. Thus it can be safely concluded that further concentration beyond
this stage in not necessary.
The concentrate of the alloy obtained from gravity separation can be recharged
together with pellets into the Sirosmelt furnace or added to the granulated matte,
and fed to the siroleach plant.
8/8/2019 Pages 1 57
46/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 46
Fig 4.5: Experimental flow sheet
Feed
wt %
% Cu, %Ni
Rec (Ni, Cu)
MZT1
MZT2
MZT4
MZT3
wt %
% Cu, %Ni
% Cu, %Ni
% Cu, %Ni% Cu, %Ni
% Cu, %Ni
% Cu, %Ni
% Cu, %Ni
wt %
wt %
wt %
wt %
wt %
wt %
Rec (Ni, Cu)
Rec (Ni, Cu)
Rec (Ni, Cu)
Rec (Ni, Cu)
Rec (Ni, Cu)
Rec (Ni, Cu)
Rec (Ni, Cu)
wt %
% Cu, %Ni
Rec (Ni, Cu)
100.00
1.12, 0.97
100.00, 100.00
concentrate tailing
Grind
80%,
125m
tailing concentrate
Final concentrate tailing
Final tailing concentrate
70.36
13.27, 2.35
48.15, 46.67
29.63
36.00, 6.02
51.85, 53.33
49.64
22.41, 5.81
47.88, 30.90
50.36
49.40, 6.23
52.12, 69.11
35.93
51.40, 34.22
88.71, 92.34
64.07
1.20, 0.61
11.29, 7.66
6.11
37.40, 32.86
55.01, 57.13
93.89
1.83, 1.75
44.99, 42.87
8/8/2019 Pages 1 57
47/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 47
4. 6 Spiral Concent rat ion Test work on Sirosmelt Slag
Table 4.4: Spiral Concentration of Sirosmelt Slag
The concentration grade is low, at 6.74% for Cu and 6.13% for Ni. However the
recoveries are high at 92.86% and 97.51% for copper and Nickel respectively.
Thus although little metal is lost to tailings, the resultant grade after spiral
concentration is low.
Sample wt wt Assay Cu Assay Ni Recovery Recoveryg % % % Cu % Ni %
Spiral Feed 20000 100 1.12 0.97 100.00 100
Spiral Conc 3086 15.43 6.74 6.13 92.86 97.51124
Spiral Tail 16914 84.57 0.0945 0.028546 7.14 2.48876
8/8/2019 Pages 1 57
48/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 48
5. 0 CONCLUSI ONS
1. The average assays for Cu and Ni are above the expected limits of a total
of 1% for both, at 0.97% and 1.12% respectively, totalling 2.09%. This
high metal-slag loss situation can be remedied through a mineral
processing technique.
2. Since metallic species in the slag occur in the size range from 1000m, grinding at 1mm can be envisaged to liberate majority of the
matte particles coarser than 200m. For particles less than 15m, very
fine grinding can unlock them.
3. The large difference in specific gravities between the Ni-Cu alloy and the
slag, which gives a correspondingly high concentration criterion, can be
manipulated in choosing a physical separation technique for concentrating
the Cu-Ni alloy. The appropriate technique chosen was gravity separation,
which utilises differences in specific gravities between materials in
separating them.
4. Metal distribution in the various size ranges shows that 92.77% of the
copper and 91.09% of the nickel is contained in the size fraction minus
1000 to plus 125microns. Therefore to effectively liberate the particles and
8/8/2019 Pages 1 57
49/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 49
still save on comminution costs the particles should be ground in the
above size range.
5. Recoveries of copper and nickel by size show that plus 60%recoveries
occur between minus 850micronsto plus 125microns. The effectiveness
of the gravity separation process falls for particles below 125micronsdue
to loss of values in fines at fine sizes.
6. The Mozley multi gravity separator table produced high recoveries for both
copper and nickel in the final concentrate, at 92.34% and 88.71%
respectively, after subjecting the slag to a rougher-scavenger-cleaner-
recleaner gravity concentration circuit. The grades for copper and nickel
were 51.40% and 34.22 for the same process. Thus it can be concluded
that shaking tables can be used in the concentration of the alloy in the
slag.
7. The spiral concentrator produced very low grades of the alloy although
recovery was high. This shows that little separation between the gangue
and the alloy was effected. Thus their use in concentrating the alloy was
less effective.
8/8/2019 Pages 1 57
50/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 50
6. 0 RECOMMENDATI ONS
6.1 In a plant production scenario milling can be effected through
crushing the slag using a jaw crusher and using a Barmac crusher
for size reduction down to passing 1mm. A complete milling circuit
would encompass all the aspects like screening and sizing.
6.2 The laboratory scale gravity separation equipment used is useful in
duplicating gravity separation behaviour of equipment like James
shaking table. But due to the gradual phasing out of the James
shaking table as a gravity separation device over the past years,
such equipment as spirals and the Mozley C902 Multi Gravity
Separator Drum are highly recommended for alloy from slag (AFS)
applications.
6.3 The product from gravity separation is of a high grade at high
recoveries. Thus further concentration beyond gravity separation is
not recommended considering costs of operation versus
improvement in grade.
6.4 A spiral concentrator flow sheet incorporating many spirals in a
separating circuit needs to be designed in order to maximise the
upgrading exercise, for a higher grade of alloy to be obtained from
spiral concentration.
8/8/2019 Pages 1 57
51/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 51
6.5 The product from gravity separation can be fed into the Sirosmelt
furnace together with pellets or fed to the siroleach together with
granulated matte.
6.6 The recommended plant flow diagram for the whole process of
recovering Ni-Cu alloy from slag is shown in appendix 6.
8/8/2019 Pages 1 57
52/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 52
7. 0 BI BLI OGRAPHY
1. RioZim E.N.R., Empress Nickel Refinery Operations Manual, RioZim
E.N.R., Zimbabwe, 2001, pp1-5
2. Hausen, D.M., Evaluation and Optimisation of Metallurgical
Performance, SME Inc, 1991, Chapter 17
3. Bernhardt, C., Particle Size Analysis, Chapman and Hall, London, 1994,
pp8-26
4. Anon. Test Sieving, British Standard 1796, London Press, London, 1976,
pp9-11
5. Wills B.A., Mineral Processing Technology, Sixth Edition, Pergamon
Press, Oxford, 1996, pp7-8, 116-124, 142-176
6. Lewis, F.M. et al. Comminution: A guide to size reduction system
design. Min. Eng., 1976, pp28, 54-55
7. Rodriguez, D.E. The Tidco Barmac Autogenous Crushing Mill-a circuit
design primer, Minerals Engineering, 1990, pp53-54
8. Taggart, A.F. Handbook of Mineral Dressing, Wiley, New York, (1945),
pp213-215
9. Pearse, G. Some Manufacturers of hydrocyclones, Mining Magazine,
1988, pp106
10. Mills, C., Mineral Processing Plant Design, AIMME, 1978, pp38-39
11. Wills, B.A., Laboratory simulation of shaking table performance,
Mineral magazine, 1981, pp87-89
8/8/2019 Pages 1 57
53/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 53
12. Mozley R. et al., The Mozley Mineral Separating Systems, 2000, pp2
13. Evans, A.M., An Introduction to Ore Geology, Blackwell Scientific
Publications, Oxford, 1980, pp5-6
14. Craig, J.R., Vaughan, J.D., Ore Microscopy and Petrography, John
Wiley and Sons, Canada, 1981, pp1-10
8/8/2019 Pages 1 57
54/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 54
8. 0 APPENDI CES
Appendix 1: Calculation Of Total Metal-Slag Losses
Total tonnage of slag to date= 11958.3 tons
Total amount of copper lost to slag= 0.97% of 11958.3 tons = 115.99551tons
Total amount on Nickel lost to slag= 1.12% of 11958.3 tons = 133.93296 tons
Total amount on alloy lost to slag= 250 tons
Total revenue lost due to copper lost to slag= U.S.$1,600,00 x 116 =U.S.$185,592,82
Total revenue lost due to nickel lost to slag=U.S.$7,400,00 x 134 =U.S.$991,103,90
Total revenue lost due to Ni + Cu lost to slag= U.S.$1,176,696,72
Appendix 2: Calculation of Concentration Criterion
).(.).(.
).(.).(.
mFluidMediuGSeralLighterMinGS
eralLighterMinGSeralHeavierMinGSonionCriteriConcentrat
=
8/8/2019 Pages 1 57
55/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 55
Appendix 3: Calculation of Recovery
Ff
Ccery
%100covRe
= or
Ff
Ttery
%100covRe
=
WhereC = wt % of concentrate,c = grade of concentrate,T = wt % of tailingt = grade of tailingF = wt % of Feed,F = grade of feed
Appendix 4: Copper And Nickel Distribution In The SizeFractions
Size Mass Wt Assay
Assay
Distribution
Distribution
Cumulative Cumulative
(Microns)
g (%) Cu(%)
Ni(%)
Cu (%) Ni (%) Distribution Cu(%)
DistributionNi (%)
1000 1219.5
27.05 2.11 3.44 49.62 37.45 49.62 37.45
850 292.3 6.48 0.67 1.91 3.78 4.98 53.39 42.43500 669.9 14.86 1.37 2.61 17.72 15.61 71.11 58.04425 381.1 8.45 0.97 2.25 7.16 7.65 78.27 65.69355 243.6 5.40 0.76 2.49 3.55 5.41 81.82 71.11212 657.5 14.59 0.69 2.69 8.75 15.79 90.57 86.90125 299.3 6.64 0.38 1.57 2.19 4.19 92.77 91.0975 354.1 7.86 0.50 1.13 3.40 3.57 96.16 94.6653 210 4.66 0.39 1.45 1.60 2.72 97.76 97.38-53 180.3 4.00 0.65 1.63 2.24 2.62 100.00 100.00
Total 4507.6
100.00 1.12 0.97 100.00 100.00 100.00 100
8/8/2019 Pages 1 57
56/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
D. Muchena 2003 56
Appendix 5: Effect of Particle Size on Gravity Separation
Size Concentrate
(Microns) mass wt Assay Assay Recovery
Recovery
Recovery
g % Cu % Ni % Cu % Ni % (Cu+Ni)%
1000 8.8 4.4 20.4 14.44 5.317536
18.46977
7.544094
850 17.4 8.7 39 13.32 63.42056
60.67225
62.69752
500 19.2 9.6 37.4 14.02 33 51.56782
36.59244
355 19.7 9.85 30.8 14.86 50.14545
58.78353
52.66405
212 27.4 13.7 25.2 15.14 63.81516
77.10706
68.22938
125 14.3 7.15 35.8 14.72 84.47855
67.03694
78.52565
75 8.5 4.25 34.4 14.72 36.73367
55.36283
40.85323
53 2.3 1.15 43.4 14.58 15.84444
11.56345
14.495
53- 1.7 0.85 42.2 7.15 6.97859
9
3.72852
8
6.19608
6Total 119.3 59.65
8/8/2019 Pages 1 57
57/57
Designing a process for the recovery of Nickel-Copper alloy from Sirosmelt slag at Empress NickelRefinery
Appendix 6: Suggested Alloy From Slag (AFS) Plant flow Diagram
Recommended