close

Вход

Забыли?

вход по аккаунту

?

Separation and Mineralogical Analysis of Bayer Red Mud.

код для вставкиСкачать
Dev. Chem. Eng. Mineral Process., I0(5/6), pp. 475-489, 2002.
Separation and Mineralogical Analysis of
Bayer Red Mud
T. Picaro*, B. Pei, and A.R. Kane
Queensland Alumina Limited, Parsons Point, Gladstone,
Queensland 4680, Australia
and M.R. Thornber and A.B. Fletcher
CSIRO Division of Minerals, Conlon St., Waterford 6152,
WesternAustralia
This paper describes preliminary work to assess the potential for recovering valuable
components from Queensland Alumina Limited (QAL) red mud by physical
beneficiation. Upgrading of Weipa bauxites residue was trialed using a 50 mm
hydrocyclone and a Mozley Multigravity Separator (MGS) to recover desilication
product (DSP) and Ti02$-actions. The mineralogy of the original red m u h and the
various sub-$-actions produced was studied by chemical assay, selective leaching,
XIZL)and SEM techniques.
DSP is present in QAL red mud in twoforms: (0 as balls in the sue range from
0.8 to I 0 pm that appear to be liberated; and (ii) as finely grained DSP intermixed
with the other minerals. The hydrocycloning tests produced a DSP rich (39.5 wt%)
fraction at 89% recoveryfrom a feed containing 31.9 wt.% DSP.This represents an
upgrade (enrichment) of 1.24 times. These results suggest that the greatest DSP
enrichment will be achieved by targeting the 0.8 to 10 pm DSP balls.
The richest Ti02fraction produced variedfiom 20 to 36 wt. % Ti02fi.om a feed
value of 7 wt.% at a recovery of I. 5%. Mineralogical examination showed that most
of the TiOz in this fiaction was as liberatedparticles in the size range 8 to 20 pm.
Most of the Ti02particles in the other fiactions were not fully liberated from mainly
iron minerals.
The present work indicates that it is possible tofurther upgrade DSP fractions by
investigatingliner particle separation techniques, e.g. smaller, 25 mm and I0 mm,
hydrocyclones.
* Authorfor correspondence.
4 75
T. Picaro, B. Pei, A.R. Kane, M.R. Thornber andA.B. Fletcher
Introduction
In the Bayer process sodium hydroxide is used to extract alumina fiom bauxite using
high temperatures and pressures. During the alumina extraction step some of the
sodium hydroxide may be consumed by reaction with kaolin and quartz contained in
the bauxite, to produce sodium-aluminum-silicate desilication product (DSP). Each
tonne of silica that reacts during digestion consumes approximately 1.2 tonnes of
caustic soda and accounts for as much as 20% of the total alumina production costs
for processing high-silica bauxites [ 11. The DSP remains with the process residue, red
mud, which is eventually disposed of in large mud ponds. Environmental issues
associated with residue disposal are also of great concern.
Economically and environmentally speaking, there is growing interest in
processing and utilising the red mud by-product of the Bayer process. Literature
reviews [2-4] indicate that there are generally two approaches: one is to use red mud
as a “material” in its totality, such as a raw material for building products, etc. The
other is chemical-metallurgicaltreatment to recover and use the valuable components
in red mud. Elements such as NayAl, Ti and Fe are of particular interest. A number
of processes have been studied and proposed in the literature to extract caustic soda,
alumina, Ti and Fe minerals from red mud. Typical examples are hydrothermal
recovery of caustic soda using lime [S-81, recovery of TiO, by acid leaching [9],
sintering of red mud to recover caustic soda and alumina [lo], or smelting of red mud
to recover Al, Ti and Fe [ 113. New processes are also under investigation to more
efficiently and economically recover caustic soda and alumina fiom red mud [ 121.
Depending on the amount of reactive silica (kaolin) in the original bauxite, DSP
as the major soda-containing component may make up anywhere from 5 to 40 wt.%
of the red mud solids. Separating the DSP fiom red mud or DSP beneficiation would
reduce the amount of inert material to be treated in prospective caustic soda recovery
processes. This would not only lower capital costs but also operating and maintenance
costs, especially for those processes which are energy intensive [lo, 121. In addition,
if the residue from the DSP upgrading step could be further processed to upgrade Fe
andor Ti minerals, then this could significantly reduce the overall down-stream
processing costs of the red mud and would also lead to a significant reduction in the
mud load for disposal.
Among the various mineral separation techniques, hydrocyclones are the most
simple and efficient devices used in the minerals industry and coupling these to the
newer technology of the Mozley Multi Gravity Separator (MGS) is likely to provide
the optimum system for fine particle processing. The MGS consists of a tapered drum
which is rotated along a horizontal axis, at the same time as a shaking action is
imparted along the horizontal axis. Feed slurry and wash water are introduced into
the drum and a series of scraper blades are used to push the concentrate material
towards the discharge. The unit combines the principles of thin-film separation (as
used in shaking tables) with the higher separation forces imparted by the drum
rotation.
The main aim of the work presented here was to investigate if hydrocyclones are
capable of producing significant beneficiation of DSP from Bayer red mud produced
at the Queensland Alumina Limited (QAL) refinery. In addition, tests were also
carried out to assess the potential for enrichment of iron and titanium minerals from
the red mud by using the MGS.
4 76
Separation and Mineralogical Analysis of Bqver Red Mud
Mineralogical studies of various separated &actions were carried out using a
number of selective leaching treatments. XRD and SEM techniques were used to
reveal the mineralogical correlation between DSP and iron minerals, as well as
between titanium minerals and iron minerals in red mud. The separation efficiency is
discussed in conjunction with the mineralogy of the red mud. Further test work has
been identified and additional processing techniques suggested in order to improve
the separation efficiency.
Experimental Procedures
(i)
Physical Separation
The red mud slurry with approximately 500 g/L or 39 wt.% solids content was
collected from the last washer underflow outlet of the QAL mud washing circuit. The
mud was diluted with water to 25 wt.% solids and thoroughly mixed and passed
through a 600 pm sieve to remove any large fragments of scale and mud that could
block the hydrocyclone.
The sieved slurry was processed through the Mozley C700 hydrocyclone test rig
using the C124 hydrocyclone (50 mm) with an 8 mm vortex finder under conditions
that gave a nominal 5 pm, dso size distribution. The pressure to the cyclone was
maintained at just below 350 kPa to give the recommended cone-shaped spray issuing
from the vortex at the bottom. Figure 1 shows details of the test.
The test rig was operated in the “desliming mode”, i.e. successive samples of the
“slimes” reporting to the overflow were collected for further analyses. The underflow
was also collected and re-run through the hydrocyclone. The first sample collected
was the initial third of the volume from the hydrocyclone overflow, the second
sample collected was the next third of the volume from the overflow, and the coarse
fraction was left as the last fraction remaining in the test rig hopper. Initially the mud
slurry (diluted to 25 wt.% solids) was sampled as soon as it was placed in the test rig
without allowing time for the mud to be sheared by the mono pump used to pump the
slurry through the hydrocyclone. The first lot of three samples (QDlA, QD2A,
QD3A) were each sub-sampled and then recombined and pumped round the circuit
for 30 minutes so that shearing of the mono-pump had sufficient time to break up
aggregated particles. QD2A and QD3A were each sub-sampled and then recombined
and pumped round the circuit for 30 minutes so that shearing of the mono-pump had
sufficient time to break up aggregated particles.
The slurry was then sub-sampled in the same way as before and giving three
more samples (QDlB, QD2B, QD3B). These slurry samples were then each passed
through the MGS to give a tail and a heavykoarse fraction. The MGS was set to its
fastest rotation and jigging rate, a horizontal tilt. The slurry and water flow were kept
to the minimum possible. These conditions should optimise the proportion reporting
to the coarseheavy fraction.
477
i? Picaro, B. Pei, A.R. Kane, M.R. Thornber andA.B. Fletcher
Figure 1. Flow diagramfor the test.
(ii) Sample Analysh
Samples of the solids from the various fractions collected from the hydrocyclone were
prepared by settling, filtration, washing with deionised water, and drying at 110OC.
Analyses of these solids were done by X-ray fluorescence analysis (XRF) for major
and minor elements, and by X-ray diffiaction (XRD) peak height ratios compared to
those of standard mineral mixtures to give mineral proportions. The DSP was
removed fiom the solids by acid leaching (O.05M H2S04) and inductively coupled
4 78
Separation and Mineralogical Analysis of Bayer Red Mud
plasma emission spectroscopy (ICP) and atomic absorption (AA) assay of the leachate
were used to determine the DSP content, as well as any other major elements leached.
The Malvern laser difiaction technique was used to measure particle size
distribution. The absolute density was determined using helium pycnometry
according to ASTM D4892. Particle size distribution was measured on an original
solid sample of red mud collected from the hydrocyclone, and also measured on the
residue after the DSP had been acid leached from the original solid. The particle size
distribution of the dissolved fraction of the DSP was estimated from the difference in
size distribution of the solids before and after leaching.
The data were normalised where possible to give a comparative quantitative
mineralogy of the fractions, and fiom the mass balances the degree of beneficiation
was calculated
(iii) Mineralogical Examination
The red mud samples were leached in solutions of bicarbonate and tri-sodium citrate
with addition of dithionite to dissolve all of the iron components. The solids
remaining from this CCD leaching process were dried at 110°C and weighed and the
weight % leached was calculated. The CCD leach process was adapted from Smith
and Mitchell [ 131. This solid was then acid leached to give the wt.% DSP and then
the solid was examined by SEM and XRD. Assays of the leach liquid were related
back to the weight of solution so that the percentage weight of each element leached
was calculated.
Mineralogical examination of the QDlB sample and the leached materials
(QDlB-CCD and QDlB-CCD-ACID) was made using a JEOL JSM 5800LV
scanning electron microscope (SEM) using the Oxford Link energy dispersive system
for elemental analysis, in order to determine the nature of the DSP minerals. This
information could then be linked in with the bulk mineralogy of the samples.
In preparing the SEM samples, appropriate slurry samples were dried onto SEM
mounts and carbon coated if they were to be studied in high vacuum mode. Samples
of dried washed solids were glued to the sample mounts by spreading the powders
onto graphite tape and these were also carbon coated if needed.
Results and Discussion
Figure 1 gives the sample identification numbers, and the calculated weights and
percentages of solids in each sample fraction. Discrepancies in the mass balance are
due mainly to spillages, seepage of mud through the seals on the deslimer mono
pump, and accuracy limitations in the weight calculations. Tables 1, 2 and 3 give a
summary of the acid leach analysis, XRF results, and particle size distribution and
density, respectively for all of the sample fiactions.
4 79
I: Picaro, B. Pei, A.R. Kane, M R. Thornber and A.B. Fletcher
Table 1. Acid leach results; weight % of oxides leached by weak acid solution.
Table 2. XRF analysis of solid samplej-actionsfor the test.
Table 3. Particle size and density*
480
Separation and Mineralogical Analysis of Bayer Red Mud
I
Particle Size
Figure 2 shows the particle size changes that occur when the original mud solids
minus the >600 p fiaction were leached with weak acid, Figure 2 shows that there
is a bimodal size distribution for the original material with the fines mainly centred
around 3 pm, and the coarse being above 100 pm (the >600 pm particles had been
previously screened-off). Figure 2 shows that the acid leach has preferentially
removed particles from 0.8 to 10 p in size. Since only the DSP component in red
mud is soluble in acid, this indicates that the majority of the DSP is concentrated in
this size range. The acid leach data in Table 1 gives the most precise measurement of
the DSP content in each fraction. The DSP content has been calculated from the
leached silica based on the noselite formula, N~(AlSi04)6.Na2S04.Table 1 also
shows that there is some DSP in the coarser fractions such as 4600 and QMG3H, and
a preliminary examination of these materials by the SEM showed that the DSP was
incorporated in the large grains that were most likely fragments of scale.
Leached
Size (micron)
Figure 2. Size distribution of original sol& before and afler the
DSP was leached
Figure 3 shows the particle size distribution changes that occur during the
hydrocyclone tests, through the deslimerhydrocyclone. Figure 3, as well as Table 3,
shows that the hydrocyclone is concentrating the fines into the overflow hctions,
QDlA and QD2A (for clarity QD2A is not shown in Figure 3 as it has very similar
size distributions to QDIA), and the underflow is accumulating the coarser particles
481
T.Picaro, B. Pei, A.R Kane, M,R Thornber andA.B. Fletcher
(QD3A). There are still some fines remaining with the underflow m i o n , QD3A, and
continued operations washing with water would be likely to remove these. Because of
the preliminary nature of these tests there was no optimisation of the test conditions.
However subsequent tests with 10 wt% solids indicated even better accumulation of
the sub 10-micron hction in the overflow.
From these particle size data it is obvious that the fine DSP is enriched into the
overflow fractions of the hydrocyclone, and it is this material that would be most
reactive for caustic recovery processes because of its large surface area.
Eb,
.
I
5
slze (rnlcrons)
Figure 3. Coinparkon of particle site distributionsfor the inital
hydrocyclone separations.
11
DSP Upgrade
The DSP values in Table 1 are valid if used comparatively, and Table 1 shows that
sample QDlA achieved an overall DSP upgrade of 1.24 if the screening and
hydrocyclone effects are combined. Samples from the hydrocyclone overflow,
QD 1A, QD 1B and Q D 2 4 achieved better than 1.20 upgrade while sample QD2B had
a lesser upgrade at 1.10. The tail sample, QMG2T, produced from putting QD2B
through the MGS did improve to a 1.18 upgrade. The “B” samples taken after
shearing in the pump for 30 minutes did not achieve as good an upgrade as the
supposedly more flocculated “A” samples of the first test. As mentioned above, later
482
Separation and Mineralogical Analysis of Bayer Red Mud
tests using more dilute solid content (10 wt%) in the feed mud suggested a better
liberation of the DSP in the overflow. Combining all overflow fiactions there was a
89.6% recovery of the available DSP.
The acid leached Na20 results shown in Table 1 give more relevant data to the
ultimate soda recovery. This data correlates with the DSP data that is calculated from
the Si02 leach results but it also shows an improved Na20 recovery in the MGS tails
samples, QMGlT, QMG2T and QMG3T, over the corresponding feed samples
QDlB, QD2B and QD3B as compared to the DSP recovery. The total Na20 data
shown fiom the XRF assay results in Table 2 agrees with the leach results and all
samples show some Na20 that is acid insoluble. This acid insoluble Na may be in
calcium silicate minerals and scale fragments as it correlates with the acid insoluble
Ca and the particle size.
111 TiO, Upgrade
The surprising result fiom this study was that the MGS was able to give a significant
upgrade of TiOz minerals by concentrating heavy minerals fiom the fine slime
fiactions produced by the hydrocyclone. Tables 1 and 2 show an enrichment of Ti02
in the form of anatase and rutile into the MGS heavy fractions, QMGlH and
QMG2H, such that there is an upgrading of 5.29 (35.7% Ti02) and 2.92 (19.7% Ti02)
respectively. These fiactions also have the highest levels of iron minerals, much of
which is magnetic. Thus a magnetic separation may give a further upgrade.
Although there were only 8 grams and 58 grams reporting to these fiactions
respectively, combining the TiOz contents of samples QMGlH and QMG2H gives a
recovery of 7.5% of the available Ti02. These results indicate how selective the MGS
can be. There is the potential for selectively upgrading titanium minerals using the
hydrocyclones and MGS. The tests carried out were not set to optimise the recovery
of the heavy fractions fiom the hydrocyclone overflows, and knowing where the
target minerals are in terms of size distribution should improve recoveries.
IV
Mineralogical Examinafion(SEM and XRD)
Mineralogical examinations (XRD and SEM) were done on the hydrocyclone
overflow sample (QDlB), the CCD leach of that sample (QDlB-CCD) and the CCD
plus acid leached sample (QDlB-CCD-ACID). Semi quantitative point assay data
were also collected. Such assay data are qualitative and cannot be used quantitatively.
However they are extremely valuable for comparing the major elemental
compositions within the sample under examination.
1. DSP SEM observations
The CCD leach proved to be extremely useful in establishing the morphology of the
DSP present in residues. The XRD patterns showed clearly that haematite was
removed by this leach.
The CCD leached specimens gave micrographs that were dominated by the DSP
“wool balls” as shown in Figures 4a and 4b. It was observed that they were in the
size range from 0.8 to 10 pm although it was difficult to establish the amount of DSP
483
T. Picaro, B. Pei, A.R Kane, M R Thornber and A.B. Fletcher
present in the very fine material that was less than 0.5 pm. These balls could be
expected to have had all of the iron leached away from their outer surface. The point
assays of the spheres show iron content fiom about 0.5 to 2.0 atom % (Figure 4a and
Table 4), meaning that this level of iron is locked within the bdls and not available
for the CCD leaching. There is a negative correlation (-0.88) for the assay data
between the Ca and Na which indicates that some of the Na in the DSP is being
substituted by Ca. There is a positive conelation of 0.73between the calcium and the
titanium in the DSP indicating that they may be associated.
The general conclusion is that a significant proportion of the DSP exists as
liberated balls in the size range from 0.8 to 10 pm, and have potential for
concentration if a means can be devised. There may be some adhering iron oxides on
these particles (as much as 10% or more) that cannot be separated from the balls when
the solids are dispersed in the residue liquors. There is also a form of DSP that is
finely intermixed with the other minerals to form particles that cover the whole size
range from 100 pm to sub-micron particles. These probably include the scale
hgments although they may not all be scale.
Figure la. Back scattered electron tmage of a &pica1 area of the CCD &ached DSP
enriched residue sol&. m e spherical particles are DSP.
Points A and D are simhr in content, ILS are points C and F, and B and E. Point G
is a diflerent shaped cystal and appears to be a titanata
484
Separation and MineralogicalAnalysis of Bayer Red Mud
Figure 46. Another back scattered electron SEM image of a @pica1 area of the
CCD leached DSP enriched residue soli& at lower magniflcarion givlng an
indication of the abundance of DSP balk
Ti
Fe
1.52
2.16
0.42
0.47
3.20
1.19
1.24
1.56
3.15
0.40
4.54
1.84
11.74
0.47
485
T. Picaro, B.Pei, A.R.Kane, M.R.Thornber and A.B.Fletcher
2. Ti02 mineral SEM observations
Removal of the DSP by acid leach after the CCD leach gave a residue with no DSP
and little iron oxides. SEM observations show Ti02 minerals of about 8-20 p in
size. The TiOl rich grains which were gravity concentrated all appear to be well
liberated. Generally, the Ti02 grains are of three main types, those free of iron, those
with associated iron, and those intermixed with iron oxides.
Discussion of Separation Problems
The hydrocyclone and the MGS both rely on differences in settling rates to achieve a
separation of particles. Both the data from the SEM studies and the acid leaching of
DSP associated with particle sizing as shown in Figure 2 indicate that the major part
of the DSP is in the size range 0.8 to 10 pm. If this is targeted for recovery by
sedimentation processes then the calculation using Stokes Law shows that haematite
particles with a size range from 0.55 to 4 pm would be expected to have the same
sedimentation rate if density values of 2.3 and 5.1 g/cm3 are used for the DSP and
haematite, respectively. This represents a significant proportion of the haematite in
the residue. However, if the fine fraction of less than 1 pm could be rejected as well
as the fraction above 10 pm then the optimum upgrade of DSP using settling
techniques in water could be obtained.
The densities of anatase and rutile are so similar to haematite that the haematite
particles of the same size must be expected to settle at the same rate as the Ti02
minerals. The haematite minerals that are significantly smaller or larger can be
discriminated. The best upgrades of TiOz minerals comes from the MGS heavy
concentrate of the hydrocyclone overflow which contains the smaller particles and the
liberated TiOz mineral particles that are concentrated are in the 8 to 20 pm size range,
as demonstrated in the physical separation and confirmed here by SEM.
Future Study
(a)
DSP Upgrading
These results should be used to carry out tests to select the 0.8 to 10 pm DSP balls.
Figure 5 outlines a possible conceptual flowsheet using a series of small
hydrocyclones to provide a gradual or stepped upgrading of DSP from the red mud.
The objective is to use combinations of the 50 mm, 25 mm, or 10 mm cyclones to
produce cuts in the red mud that contain sufficient DSP to be economically viable for
caustic soda recovery.
In Figure 5 the red mud is initially passed through a 50 mm hydrocyclone to
collect a sub 10 pm fraction in the hydrocyclone overflow. This fraction would
contain the majority of the DSP (89% as found in this work). In addition to
concentrating the DSP the amount of red mud to be treated subsequently is reduced,
e.g. in this work the overflow fiom the 50 mm hydrocyclone was 75% of the total red
mud feed. The overflow would then be passed through a smaller hydrocyclone, say a
25 mm or 10 mm hydrocyclone to remove the fine iron fraction (<1 pm) in the
overflow and leaving a significant proportion of the DSP in the 0.8-10 pm size range
486
Separation and MinerafogicafAnalysis of Bayer Red Mud
+
ovemwv enriched in
DSP
< 10 um fraction
(89% of feed DSP)
I
Red Mud
Feed
-1 , 1
ovemow to
I
2nd Stage
Hydrocyclone
(25 r n m l l O mm)
I
1st Stage
Hydrocyclone
(50 rnm)
+
underflow enriched in DSP
2-8 um fraction
(to soda recovery)
underRow to tailings dam
>I
0 urn fraction
(1 1% of feed DSP)
Figure 5. Conceptualjlowsheet of DSP upgrading circuit.
in the hydrocyclone underflow fiaction. The lowest d50cut obtainable is at 2 pm with
the 10 mm hydrocyclone. This may well remove a significant part of the very fine
iron as the overflow. However at the same time DSP losses would have to be
expected with the overflow. The result of this treatment would be to produce a
fraction which is rich in DSP and at the same time reduce the amount of inert material
which is treated in the caustic soda recovery process.
Selection of the optimum combination of hydrocyclones would require further
testing as well as economic evaluation since cuts at such small particle sizes would
require considerable pumping capacity. The throughput per hydrocyclone for a 10
mm unit can be expected to be 0.1 to 0.2 m3 per hour for example. However already
there are significant commercial applications of 50 mm hydrocyclones such as the Mt
Keith nickel mine in Western Australia where a total of 2240 Mozley 50 mm
hydrocyclones are treating a combined capacity in excess of 6000 kL/hr (1 0 M tonne
annually) to deslime a sub 10 pm fraction [14]. Further advances in fine particle
processing in the future could make the upgrading of DSP fiom red mud
commercially attractive, especially if used to feed a caustic soda recovery process
which would itself benefit fiom treating a lower flowrate of concentrated DSP.
(b) Ti02 Upgrading
The strategy for obtaining the highest recovery of TiOz minerals to an economic Ti02
grade lies in developing the best gravity-based means of collecting richer Ti02
fiactions at the same time as separating the DSP rich fiactions. It was found that a
significant proportion of the TiOz minerals are liberated and the degree of liberation
487
T. Picaro, B. Pei, A.R. Kane, M.R. Thornber andA.B. Fletcher
can be improved. There are several options available, as demonstrated here, for
improving the recovery and grade of the TiO2:
1.
Treating the fractions that contain most Ti so that the iron particles, which appear
to be fiiable, are broken down into smaller particles that can be slurried away
from the more robust Ti02grains.
2. A weak acid leach could give an advantageous liberation of Ti02 mineral grains,
and even a dithionite-citratetype of leach could be used to clean more of the iron
away.
3. The indication that the Ti02 minerals associated with the magnetic fraction are
mainly sub micron and not worth collecting, if true, could mean that removal of
the magnetic fractions from a Ti02rich gravity concentrate would give a valuable
upgrade.
Conclusions
Separation tests were carried out for red mud samples using a 50 mm hydrocyclone in
combination with the Mozley multi gravity separator (MGS). The separation gave an
upgrade up to 1.24 times (24%) producing an overflow stream with 39.5% DSP at a
recovery of 89%.
Ti02 minerals anatase and mtile were concentrated into two heavy fractions from
the MGS with an average particle size of about 12 pm. These gave an upgrade of 5.29
and 2.92 respectively for a combined recovery of approximately 7.5% of the liberated
Ti02.
Solid sample hctions, that were enriched in DSP and T i 4 produced from red
mud were characterised. The scanning electron microscope (SEM)was used to
compare samples, prepared by leaching the DSP and/or the iron minerals, by
observing the size, morphology, mineral associations and degree of liberation of the
DSP and Ti minerals.
The DSP is present in two forms: as balls in the size range from 0.8 to IOpm, that
appear to be liberated and as finely grained DSP intermixed with the other minerals to
form particles that cover the whole size range from mm to sub-micron particles.
These probably include the scale fiagments although they may not all be scale.
If the DSP balls are targeted for recovery, combinations of the 50 mm, 25 mm,
and 10 mm hydrocyclones could be used to reject the coarse fraction above 10 micron
as well as the fine fiaction of less than 1 pm to produce cuts in the red mud solids that
contain sufficient DSP to be economically viable for caustic soda recovery. The
results of this work should be used to carry out further tests to select out the 0.8 to 10
pm ball type of DSP.
The liberated Ti02 mineral particles that were concentrated were in the 8 to 20
pm size range as confirmed here by SEM. The densities of anatase and rutile are so
similar to haematite that the haematite particles of the same size settled at the same
rate as the Ti02 minerals.
488
Separation and Mineralogical Analysis of Bayer Red Mud
Acknowledgements
The authors would like to thank the management of Queensland Alumina Limited for
their permission to publish this paper.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Baksa, G, Vallo, F., Sitkei, F., Zoldi, J., and Solymar, K. 1986. Complex causticisation: an effective
means for the reduction of NaOH losses in an alumina plant, Light Metals, pp 75-80.
Haake, G. 1988. Red mud a waste or a valuable by product. Neue Hutre, November, 33(1 I), 424429.
Thakur, R.S. and Sant B.R. 1983. Utilization of red mud: Part I: Analysis and utilization as raw
material for absorbents, building materials, catalysts, fillers, paints and pigments. Journal of
Scientifrc and Industrial Research, February, 42,87-108.
Piga, L., Pochetti, F. and Stoppa, L. 1993. Recovering metals from red mud generated during
alumina production. Journal ofMetals, November, 54-59,
Peffer, H.C. 1909. Recovery of soda in the manufacture of alumina, US Patent 938,270.
Hudson, L.K., Savage, K.I., and Stinson, J.M. 1977. Process for recovery of alumina from high-silica
ore, US Potent 4,044,095.
Cresswell, P.J. and Mlne, D.J. 1982. Australian Potent 8810Y82.
Solymar, K., and Zoldi, J. 1993. Lime in the Bayer Process Present State and Future Trends, Light
Metals.
Maitra, P.K. 1993. Recovery of titanium dioxide from red mud, UNEP Industry ond Environment,
September, 42-45.
Hrishikesan, K.G. 1977. Process for recovering soda and alumina values from red mud, U.S.Patent 4
045 537.
Fursman, O.C., Mauser, J.E., Bulter, M.O. and Stickney, W.A. 1970. Utilization of red mud residues
from alumina production. Bureau of Mines Report of Investigations. 7454.
Picaro, T. 1997. RedMud Processing, Australian Patent WO 91129992.
Smith, B.F.L. and Mitchell, B.D. 1987. Characterisation of poorly ordered minerals by selective
chemical methods, in “Handbook of Determinative Methods in Clay Mineralogy”, Wilson, M.J.
editor, Blackie, Glasgow, page 275.
HydrocycloneTechnology - 2240 Mozley Hydrocyclones Installed by AN1 at Mt Keith, WA, Mining
Maguzine London, July 1995.
-
489
Документ
Категория
Без категории
Просмотров
0
Размер файла
1 422 Кб
Теги
bayer, separating, red, analysis, mud, mineralogical
1/--страниц
Пожаловаться на содержимое документа