Sie sind auf Seite 1von 74

This article was downloaded by: [Aston University]

On: 21 January 2014, At: 02:54


Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Mineral Processing and Extractive Metallurgy Review:


An International Journal
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/gmpr20

Pressure Acid Leaching of Nickel Laterites: A Review


a a
B. I. WHITTINGTON & D. MUIR*
a
CSIRO Division of Minerals , A.J. Parker CRC for Hydrometallurgy , P.O. Box 90, Bentley,
WA, 6102, Australia
Published online: 26 Mar 2008.

To cite this article: B. I. WHITTINGTON & D. MUIR* (2000) Pressure Acid Leaching of Nickel Laterites: A Review, Mineral
Processing and Extractive Metallurgy Review: An International Journal, 21:6, 527-599, DOI: 10.1080/08827500008914177

To link to this article: http://dx.doi.org/10.1080/08827500008914177

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
Mio. Pro. Err. MI,. Rev.. Vol. 21, pp. 527-600 Q 2000 OPA (Ovcmm Pubiirhcrs Asswiadon) N.V.
Reprints available directly from the publisher Published by l i a n y under
Photocopying permitted by licenw only thc Gordon and Breach Science
Publirhcrs imprint.
Prinsd in Malaysia.

Pressure Acid Leaching of Nickel


Laterites: A Review
B. I.WHITTINGTON and D.MUIR*
Downloaded by [Aston University] at 02:54 21 January 2014

CSIRO Division of Minerals, A.J. Parker CRC for Hydrometallurgy,


PO. Box 90, Bentley,WA 6102, Australia

(Received 4 March 1999; In final form 4 January 2000)

A review of the literature over the past 30 years on the processing of nickel laterites by
high temperature acid leaching has been carried out to provide a better understanding of
the mineralogy, leaching process chemistry and effect of operating conditions on nickel
recovery, residue properties and scaling. Particular attention is paid to the leaching
experience of the commercial Moa Bay plant and to the recently reported testwork
and flowsheets associated with the three Western Australia lalerite plants that will be
operating in 1999.
It is shown that laterites can vary significantly in their mineralogy according to loca-
tion, climate and depth, and that the main host minerals for nickel and cobalt can be
either goethite (iron bxide) or nontronite (clay) or manganese oxides. The mechanism of
-
leachine involves acid dissolution of the host mineral lattice followed bv, hvdrolvsis
, ,~
and precipitation (transformation) of a variety of insoluble oxides and sulphates of iron,
aluminium and silica under the high temperature conditions. Optimum leaching condi-
tions and final liquor composition varies according to the ore mineralogy. More funda-
mental studies have demonstrated that the rate of leaching and character of the residue
is dependent upon the level of Mg, Mn and C r in the ore, the Eh of the slurry and salinity
of the process water.
A number of studies are reviewed on the chemistry and precipitation of iron,
aluminium, magnesium and silica to understand how the process conditions affect the
solubility of the species and the nature of the scale which they form. Early work at Moa
Bay indicates that the incorporation of chromium into alunite scale also affects the
incorporation of silica and nickel and the settling of the residues. Various types of scale
have been identified during- different stages
- of leaching - and .possible means of minimisina-
scale are discussed.
The clay-rich nickel laterites in Western Australia differ from Moa Bay laterite in
mineraloav -. and have com~arativelv -
- high silica and low chromium content. Since no
commercial plant has previously processed such ores or used saline process water, there
is little published in this area. It is therefore recommended that further research be car-
ried out on understanding the process chemistry and species equilibrium from various
ore types under autoclave conditions.

*Corresponding author. e-mail: David.Muir@minerals.csiro.au


528 B. 1. WHITTlNGTON A N D D. MUIR

Keywords: Nickel laterite; Pressure acid leaching; Digestion; Goethite; Nontronite;


Alunite; Jarosite; Hematite; Sulphuric acid

1. INTRODUCTION

Nickel is used in the Western world for the manufacture of stainless


steel (65% total production in 1996), for the manufacture of alloys
and for electroplating (Barkas, 1998). Currently, the majority of nickel
is refined from sulphide ores. However, there is a trend towards the
refining of nickel laterite ores, with a particular interest in the pressure
Downloaded by [Aston University] at 02:54 21 January 2014

acid leaching process. Previous reviews of the pressure acid leaching


process have examined the technology of the laterite processing (e.g.
Anthony and Flett, 1997). In this review, we examine the technology
and chemistry of relevance to the pressure acid leaching of nickel
from laterite ores, with a particular emphasis on the new Australian
laterite processing plants. The hydrometallurgy of cobalt, a valuable
product from the pressure leaching of laterites, is also discussed.
After considering the mineralogy (Section 1) this review briefly
discusses the mechanism of laterite formation, the effect of environ-
mental conditions on the laterite mineralogy and the features of some
commercial nickel-laterite ores (Section 2). A brief outline of the vari-
ous methods used to refine the nickel from laterite ores (pyrometallur-
gical, hydrometallurgical and a combined process) is presented in
Section 3. A technological overview of the pressure acid leaching
process, and a discussion of those factors influencing the operation
of the pressure acid leaching process, are presented in Section 4. The
chemistry of relevance to the pressure acid leaching process is dis-
cussed in Section 5. It is recognised that caution is required when
comparing data from different authors using different ores or when
comparing batch versus continuous data.
The chemical formulae of a number of the minerals of interest to
the nickel mining industry, and used throughout this review, are pre-
sented in Table I.

1.1. Mlneralogy of t h e Important P h a s e s


The following is a brief description of the crystallographic struc-
tures of the economically important minerals present in laterite ores
ACID LEACHING OF NICKEL LATERITES 529

TABLE I Minerals of interest lo the nickel refining industry


alunite I X = N a . HxO) asbolan birnessite
(Co, N ~ )-I . ( M ~ ~ + O ~ ) ~Mn7013.5H20
-~
( 0 % - 2 y + ~ .nH20
chlorite cryptomelane fayalite
(Mg, Fe2+, K2Mns016 Fe2Si04
AISi3010(OH)8
Forsterite garnierite
Mg2Si04 (Ni, Mg)3SizOdOHh
jarosite ( X = Na, H3O) kaolin
XFe3(SO&(OH)6 A12Si20s(OH),
nickel olivine
Downloaded by [Aston University] at 02:54 21 January 2014

Ni2Si0,
pimelite
Ni3Si4Ol0(OH)2

(goethite, clays and manganese minerals) and those of importance in


the pressure acid leaching.
"Goethite (a-FeOOH) consists of double chains of Fe-0-OH
octahedra extending along the crystallographic z axis. These octahedra
are bound to neighbouring double chains by Fe-0-Fe and H
bonds" (Schwertmann and Taylor, 1989).
Phyllosilicates (clays) consist of alternating "octahedral" and
"tetrahedral" sheets, where the co-ordination refers to that at the cen-
tral (non-oxygen) atoms. The central atoms in the octahedral sheet con-
sist mainly of A13+ and/or ~ e " while those in the tetrahedral sheet
consist mainly of si4+. Substitution of A13+ or ~ e for ~ si4+
+ in the
tetrahedral sheet, o r of ~ i ' + ,~ e ' + o r M ~ ' + for A13+ in the octa-
hedral sheet, may result in a negatively charged lattice which re-
quires neutralisation by incorporation of cations between the sheets.
Clays are grouped according to the ordering of the octahedral and
tetrahedral sheets. Thus, a 1 : 1 group clay comprises alternating
tetrahedral and' octahedral layers while a 2 : 1 group clay contains
alternating tetrahedral-octahedral-tetrahedral layers. A third group,
the 2 : 2 structure, has a 2 : 1 grouping which also contains a separate
charged octahedral hydroxide layer.
The serpentine group, which includes the antigorite, lizardite and
chrysotile minerals, conceptually arises by substitution of 3 M ~ ' + for
2 A I ~ +in the I : 1 kaolin structure. Nickel can adopt the role of
530 B. I. WHlTI'lNGTON AND D. MUlR

magnesium as shown by the existence of garnierite, a naturally occur-


ring nickel serpentine (Deer et al., 1992).
Smectites are 2 : 1 clays in which cationic substitution may occur
in the octahedral or tetrahedral layers and generate a lattice charge.
Adsorption of water at the cations present between the layers accounts
for the increase in basal spacing upon hydration which is character-
istic of smectites. Nontronite is an iron-rich smectite found in some
nickel laterite ores. The reported analyses indicate a mineral with
compositional differences (Tab. 11).
Downloaded by [Aston University] at 02:54 21 January 2014

TABLE I1 Chemical formulae reported for serpentine and nontronite


Che~i~ica/jormulnc Authors

Serpentine
Mg3 [Si2 Od(OHh Deer et a/. (1992)
M66 [Sid OIOI(OH)~ Read (1970)
Mg5.79 Fe0.2 Nio.01W3.8 Alo.205 Od(OH)4 Hayward (1998) and reler-
ences therein
Mg3 P i 2 Osl(OH)d Golightly (1979)
M ~ S SbI . 1 5 A h 1 4 Si4 0!0(0H)8 antigorite Newman and Brown (1987)
Mg3 Si2 O,(OH), antigor~te Loughnan (1969)
Nontronite
(normalised to a unit cell containing OIO(OH)~; reported in the order: interlayer cations;
octahedral cations: tetrahedral cations)
Nao.6 (Ni. F44 (A4 S ~ ) & ~ O ( O H ) ~ ( H Z O ) ~ Lawrence (1996)
M.v (Al1.06l.'e2.7~M&.26)(Si7.~&~0.70)020(oH)~ Stucki (1988) and references
therein
Na,,, Ca0.1 M&.~(F~~,sN~o.~M&.~)(S~~.~A~O,~)O~O(OH) Tindall and Muir (1996)
M . ( A ~ o . ~ ~ F ~ x M M ~ ~ . o ~ ) ( S ~ ~ . O O A ~ I . W ) OStucki ~ O ( O(1988)
H ) ~ and references
therein
Fed Sia020(OH)4 Golightly (1979)
(j Ca, W o . 7 Fe4 (S~~.~A~U.~)O~O(OH)~.~H~O Deer et a/. (1992)
(4 Ca, Na, Mg)o7 Fed %02o(OH)4.nH20 Hayward (1998) and refer-
ences therein
Mom F e 4 . (Si,.,,
~ Alo,67)020(OH)4( M = 112 Ca, Na) Loughnan (1969)
Nao.67 Fe4.0 (Si7.31 A1067)020(OH)4 Kirk-Othmer, Vol. 6, p. 392.
Mrr Fe4.0 ( S ~ R - ~ , A I ~ J O ~ ( O(M= H ) ~112 Ca, Na) Allen and Hajek (1989)
M., (F~~.P~M&.W)(S~~.~~AII.OSF~O.II)O~O(OH)~ Stucki (1988) and references
therein
M., ( F ~ ~ . ~ o MI )&( S. I~ ~ . M AI F~~OO. I. S ~ ) O ~ O ( O H ) ~ Stucki (1988) and references
therein
M . 0 Fe4.o ( S ~ ~ . ~ A ~ O . I F ~ I . I ) O ~ O ( O H ) ~ Newman and Brown (1987)
M.v (F~~.wM&.Is)(S~~.~~A~O.MF~I.I~)~~O(OH)~ Stucki (1988) and references
therein
M V( F ~ ~ . O ~ M & . I O ) ( S ~ ~ . ~ I A ~ O . O R F ~ IStucki . ~ I ) ~(1988)
~ O ( ~and
H)~ references
therein
MX( F ~ ~ . W M & , ~ I ) ( S ~ ~ . ~ I A ~ U , I ~ F ~ I Stucki
.~S) (1988)
~ ~ Oand
( O Hreferences
)~
therein
ACID LEACHING OF NICKEL LATERITES 53 1

Lithiophorite consists of sheets containing linked M n 0 6 octa-


hedra alternating with sheets containing (Li,AI)(OH)6 octahedra
(lithiophorite ideal formula [Mn:'Mn2'012]*- [AI~L~~(oH),,]~+).
Aluminium and lithium are regarded as essential constituents of
lithiophorite, although the composition can vary from the ideal.
For example, Tindall and Muir (1996) report lithiophorite in a West
Australian laterite has the approximate formula [Fe0.4A13.8Si0.2
Li2(OH)12] ,2] and indicate lithiophorite is an
important host mineral for cobalt.
Cryptomelane (ideal formula ~ 2 ~ n ? ~ n * + 0 1comprises
6) corner-
Downloaded by [Aston University] at 02:54 21 January 2014

sharing M n 0 6 octahedra. These combine to give a three dimensional


framework with tunnels extending along the z-axis. The tunnels in
cryptomelane contain Kf,of which a minimum occupation is neces-
sary to prevent the structure collapsing (Taylor, 1987).
Jarosites, which may be present in the post-digestion leach residues,
are iron containing members of the larger alunite mineral family A B ,
(S04)2(OH)6 where A = H30, Ag, Na, K, NH4, Pb or Rb and B =
A I ~ o+r ~ e ' + .The A atom is co-ordinated to six oxygen atoms and
six (OH) ions while the B atom is co-ordinated to four O H ions and
two oxygen atoms from sulphate ions. (Palache et al., 1951; Das
et al., 1996).

2. LATERITE FORMATION

2.1. Laterite Mineralogy


Nickel laterite deposits are derived from olivine-rich igneous rocks
(Eqs. (1)-(5)). These minerals dissolve over time, the elements in
solution move and new minerals precipitate in another location. This
process requires high localised rainfalls, such as in an equatorial
humid climate where the profile is constantly wet and actively leached
(8- 17 m laterite reportedly forms every million years), or in a tropi-
cal wet-dry climate where the profile is sporadically or seasonally
leached (Golightly, 1979).
The mineral solubility grossly defines the depth zoning of a com-
plete laterite profile. In particular, the more soluble minerals (e.g.,
serpentine, talc) are present at the base of the profile while the
insoluble minerals (e.g., goethite) are concentrated at the surface
(Golightly, 1979). The different regions in a lateritic ore are classified
532 B. I. WHIlTINGTON AND D.MUIR

as follows: (i) upper overburden (Ni < 1.0%), (ii) limonite layer
(SO2x 6%; MgO x 3%; Ni x 1.4%; Co x 0.15% and Fe > 40%),
(iii) possibly a transition layer (1.5 to 2.4% Ni), (iv) saprolite layer
(SO2= 38%; M g 0 x 25%; Ni x 2.4%; C o x 0.05% and Fe < 15%)
and (v) lower base rock layer (Reid, 1996).
The environment in which the lateritisation process occurs in-
fluences the laterite mineralogy. In areas where circulation of water is
restricted (e.g., tropical wet-dry climates), smectite (e.g. nontronite,
Eq. (2); modified from that in Golightly, 1979) or quartz (Eqs. (4),
(5); modified from that in Golightly, 1979) form a "nontronite zone"
Downloaded by [Aston University] at 02:54 21 January 2014

between the limonite and saprolite layers (Krause et at., 1998 and re-
ferences therein; Golightly, 1979).

4 MgzSiO4 + 10 Hf + Mg3Si4010(OH), +5 M +
~ ~4 H
" 20
Forsterite Saponite

4 Fe2Si04 + 2 0 2 + 4 H20 + + 6 FeO(0H)


Fe2Si40,0(OH)~
Fayalite Nontronite Goethite

+
4 Ni2Si04 10 Hf
Nickel Olivine
-+ +5
Ni3Si40ro(OH)2
Pimeli te
+4 H 2 0

(Mg, Ni),Si04 + 2 H+ Si02 + 2 (Mg, ~ i ) +~ H" 2 0


+

2 FezSi04 + 0 2 + 2 H z 0 2 Si02 + 4 FeO(0H)


+

Nontronite zones can be found in ores from the Biankouma district


of the Ivory Coast; the Greenvale and Rockhampton districts of
Australia; Nicaro, Cuba; Barro Alto and other lateritic bodies in
Goias, Brazil (Golightly, 1981). Magnesite (MgC03), which is
common in Australian deposits, is also indicative of reasonably arid
formation conditions (Burger, 1996a). Manganese oxides and hydro-
xides may also be present in the transition zone (Burger, 1996b).

2.2. Upgrading of Nickel Within the Laterite


Commercially viable nickel laterites generally do not contain discrete
nickel minerals. Instead ionic nickel is incorporated within, or
ACID LEACHING OF NICKEL LATERITES 533

adsorbed by, a number of secondary oxides and silicate minerals


(Burger, 1996a).
Ion exchange occurs for minerals with highly variable chemical
compositions, such as serpentines, smectite clays, garnierite, chlorite,
vermiculite and manganese oxyhydroxides. These reactions are im-
portant since they allow for upgrading of the nickel and cobalt in
the ore. The exchange of nickel in soil water into magnesium ser-
pentine (Eq. (6)) is of particular importance because, at equilib-
rium, ~ i ' + is more stable in serpentine and M ~ more
~ +stable in the
soil water (Golightly, 1979). This exchange can result in serpentines
Downloaded by [Aston University] at 02:54 21 January 2014

("garnierite") with nickel contents exceeding 25% (Brindley and


Hang, 1973; Burger, 1996a).

The nontronite lattice also has substantial capacity for cation ex-
change and nontronites may contain in excess of 2% nickel (Burger,
1996a). The Murrin Murrin nontronite contains 0-4% nickel (Camuti
and Riel, 1996) and reportedly has the highest nickel grade in this
deposit (Monti and Fazakerley, 1996).
~ i * +or c o 3 + (ionic radii 0.69 or 0.525A respectively) can
substitute for ~ e j (ionic
+ radius 0.645A) in goethite although Burger
(1996a) reports continued oxidation of goethite results in a more ord-
ered goethite lattice and consequently a lower nickel content. Nickel
incorporation into iron hydroxides has been suggested to occur at
defect sites in goethite, or in ferric hydroxide platelets (Golightly,
1981, and references therein). Evidence by Georgiou and Papangelakis
(1998) and Schellmann (1978) suggests incorporation of Ni, Al and
Si in limonite ore occur within the goethite lattice, rather than by
adsorption onto the goethite surface. Various researchers have char-
acterised Ni and/or Co substituted synthetic goethites (Cornell,
1991; Cornell and Giovanoli, 1989; Gasser er al., 1996; Gerth, 1990
and Kumar er al., 1990). Hematite, which is found in some older
limonite zones (Golightly, 1979), is less able to incorporate nickel than
goethite.
Many of the manganese oxides and hydroxides are non-stoichio-
metric alid are able to incorporate a wide range of cations, including
cobalt, nickel and copper, into their crystal structures (Canterford,
534 B. I . WHITTINGTON AND D.M U l R

1985). A correspondence between manganese and cobalt has been


reported for various laterites (e.g. Chandra et al., 1983) and may result
from the presence of manganese/cobalt minerals (e.g. asbolan) or
exchange of cobalt into the structure (e.g, lithiophorite, Section 1 .I).
The incorporation of nickel o r cobalt into synthetic manganese
compounds has been studied by Taylor (1987 and references therein);
Manceau et a/. (1997); Kumar et al. (1989) and McKenzie (1967, 1970).
The arrangement of the minerals within a particle would be expected
to influence the ore properties. For example, an ore in which one min-
eral coats another could have quite different properties from an ore
Downloaded by [Aston University] at 02:54 21 January 2014

comprising a homogeneous mixture of the two minerals. Experimen-


tal evidence, presented by Briceno and Osseo-Asare (1995), suggests
the two limonite or limonite/garnierite ores they examined comprise
iron oxide with an outer silicate mineral coating. In particular, the
point of zero charge (PZC) values of these ores (at pH zz 2) are lower
than expected from the significant iron content of the ore (PZC go-
ethite occurs at pH 7.4-7.7 while PZC silica occurs a t pH 1.5-3.0).

2.3. Worldwide R e s e ~ e S
Information regarding various laterite ore-bodies (i.e. location, chemi-
cal composition and selected mineralogy) is presented in Table 111.
A number of references examine the mineralogy of specific ore-
bodies in a more detailed manner than that presented in Table 111.
With regard to Western Australian resources, approximately 80% of
the laterite resources - including Bulong and Murrin Murrin - contain
the nickel mainly in silicate minerals (e.g. silica-smectite deposits such
as nontronite). Cobalt mainly associates with lithiophorite at these de-
posits (Tindall and Muir, 1997). At the remainder of the deposits -
including the Cawse region - the nickel is present in iron and man-
ganese oxides (e.g. goethite type deposits; Brand et a/., 1996). Goethite
is also the primary nickel host for Moa Bay laterite while serpentine
(e.g. garnierite) and talc have the highest nickel grade in tropical clay
laterites (e.g. New Caledonia, Tindall and Muir, 1997; Krause et al.,
1997).
Other nickel-laterite ore-bodies to have been characterised include
the Syerston deposit (Blight and Muir, 1996), the Marlborough de-
posit 45km NW of Rockhampton, Queensland (Griffen, 1998), the
TABLE 111 Treatment conditions and chemical composition of nickel laterite ores (including the main phases containing the element) which may be
used by laterite processing plants
Downloaded by [Aston University] at 02:54 21 January 2014

Bulong Ca~w Murrin Murrin Ravensrhorpe Marlborough Calliope Syerston Moo Bay
30 km E of 50 km N W of 60 km E of 170km W of 40 km N W Gladstone, 350 km N W
Kalgoorlie, Kalgoorlie, Leonora, Esperance, Rockhampron Qld, using New Sydney,
Location WA WA WA WA -Old. Caledonian ore NS W Cuba
Ourpur (t/a) Ni 9000 8700 45000 22,600 18740 20650
Proiected Co . 650' 20001 3000' 9902 1890' (combined4
Operating cost 1.44 0.77 0.78 0.77 1.29' - 1.40
Proiected ( $ ~ ~ l l b ) ~
Digestion
conditions
Kyle (1996) Kyle (1996) Kyle (1996) Kyle (1996) Kyle (1996)

Digest T ("C) 250 Information Information Amax type


Digest time 1.5 currently currently combined
(hrs) not not pressure/
acidlore 400 available available atmospheric
(kgltonne) leach
free acid (g/L)
Fe in solution
k/L)
Recoveries:
Ni (%)
C o (%l
Downloaded by [Aston University] at 02:54 21 January 2014
ACID LEACHING OF NICKEL LATERITES 537

Bulong deposit (Burger, 1996c), the Cawse deposit (Brand et al., 1996;
Hellsten and Lewis, 1996), the Murrin Murrin deposit (Monti and
Fazakerley, 1996; Camuti and Riel, 1996), the Ramu deposit
(Anderson, 1997), the Weda Bay deposit on Halmahera Island, East
Indonesia (Baillie and Cocks, 1998), the New Caledonian deposits
(Troly er al., 1979), the Amazonian deposits (Costa, 1997) and the
Moa Bay deposit in Cuba (Chaves et al., 1968; Sobol, 1968). Cerpa
er al. (1996) have characterised Moa Bay ores, using powder X-ray
diffraction, infra-red spectroscopy, TEM micrographs, elemental ana-
lyses, and compared the ore properties with those of "pure" iron-
Downloaded by [Aston University] at 02:54 21 January 2014

substituted serpentine and goethite.

3. PROCESSING O F LATERITE O R E S

3.1. Pyrometallurgical R o u t e s
Silica and magnesia-rich saprolite ores can be processed pyrometallur-
gically by either minimising the Fe/Ni ratio and refining a ferronickel
product or adding sulphur to form a nickel/iron sulphide matte which
can be refined further. The ferronickel process involves calcining
and pre-reducing the ore at 850- 1000°C in the presence of coal, then
smelting at 1500- 1600°C (Taylor, 1997b) to separate the nickelliron-
containing phase from the silica-magnesia slag. The sulphide matte
process is similar to the ferronickel process, except a sulphur source is
added to form a nickelliron sulphide matte (Taylor, 1997b).
The Caron process, developed in the 1920's, is a combined pyro-
metallurgical/hydrometallurgical process for lateritic ores. The nickel
and cobalt are reduced to their metallic forms by heating the ore a t
T > 700°C in a reducing atmosphere then the calcine is cooled to
less than 200°C by indirect water cooling to avoid oxidation of nickel
and cobalt (Taylor, 1997b). An ammonia-ammonium carbonate
solution then selectively leaches these materials. Boiling removes the
ammonia and precipitates basic nickel carbonate which, upon calci-
nation at 1200°C, produces nickel oxide.

3.2. Acid Leaching


The nickel and cobalt bound within the goethite or clays can be
released from the solid matrix by acid leaching. The soluble iron
538 B. 1. WHIITINGTON AND D. MUIR

species subsequently hydrolyse to basic iron sulphates which, under


the conditions present in the pressure leach (temperatures greater
than 20OoC), react to hematite and regenerate acid (Hazen and Chou,
1997). Soluble aluminium and chromium (111) sulphates also hydro-
lyse to the basic sulphate salts. However, in these instances there is
no further transition and the acid bound with these salts is not re-
leased (Hazen and Chou, 1997).
Atmospheric leaching of laterite ores can extract significant nickel
and cobalt. For example, Taylor and Cairns (1997) report test work
for the Bulong project, conducted at 90-95'C/3 hours, results in
Downloaded by [Aston University] at 02:54 21 January 2014

93 and 91% nickel and cobalt recoveries respectively. However,


the significant leaching of nearly all the iron from the solids results
in a high acid consumption and requires a high initial acid addition
of I kg/kg ore (the iron sulphates d o not hydrolyse a t this low
temperature).

3.3. Relative Efficiencies of t h e Various


Refining P r o c e s s e s
The pyrometallurgical process for refining saprolite ores and the acid
leach process both recover more nickel from the ore than the Caron
ammonia leach process (90% for matte smelting, 93% for ferronickel
smelting, 94-96% for the Amax acid leach, 75-80% for the Caron
process; Duyvesteyn er al., 1979; Reid, 1996). The acid leach process
is additionally favoured over the Caron process due to the currently
high price of cobalt, the high cobalt content of limonite ore, the high
cobalt recovery ( > 90%), lower energy requirements and lower capi-
tal costs (Reid, 1996). However acid leaching is not very econo-
mic a t processing ores containing acid consuming compounds, such
as magnesia, since acid consumption is the highest cost component in
the acid leach process (Reid, 1996). An additional point worth noting
is that the effluent from the acid leach process is more highly con-
taminated than that from the ammonia leach process, necessitating
a greater "clean-up" before disposal (Reid, 1996).
From these results we can conclude the saprolite layer, which has
a high nickel, magnesium and silica content, is best processed by
pyrometallurgical processes (Monhemius, 1987). Both the pyrometal-
lurgical and hydrometallurgical processes can accept ore from the
ACID LEACHING OF NICKEL LATERITES 539

transition body between the upper limonite layer and the lower
garnierite (saprolite) layer. The goethite-rich limonite ore has a rela-
tively low nickel content but a high iron and cobalt content and is
mainly processed by hydrometallurgical methods (Monhemius, 1987).

4. TECHNOLOGY O F THE PRESSURE ACID LEACHING


O F NICKEL LATERITES

In this section, we discuss the technology of the acid leaching process


Downloaded by [Aston University] at 02:54 21 January 2014

and present empirical experimental results for factors influencing


the efficiency of this process. The chemistry of the processes occurring
during the acid leach, which may be used to explain these empirical
results, is discussed in Section 5.

4.1. High-pressure Acid Leach "Flowsheets"


Where possible, the ore is beneficiated to improve the nickel and
cobalt concentration and improve the autoclave throughput. Conven-
tional preconcentration techniques are generally unsuccessful for
laterite ores since most of the nickel is dispersed in the iron oxides or
clays rather than occurring as a discrete mineral (Kyle, 1996; Hazen
and Chou 1997). However, it is reportedly possible to "concentrate"
the nickel in Cawse (Mason et al., 1997) or Ravensthorpe (Hamilton,
1998) ores by removal of coarse, low-grade, silica.
The slurried ore is thickened prior to feeding to the autoclaves.
Typically, thickening is to approximately 40-45% solids although the
hydrophilic nature of the nontronite present in Australian ores may
limit the slurry concentration to 25-30%. The feed slurry thicken-
ing, and in particular the solids density, significantly influence the
autoclave throughput and are therefore of some importance. The
rheology of laterite ore slurries, and methods to improve the slurry
rheology, have been examined by Avotins et al. (1979); Avramidis
and Turian (1991); Bhattacharya et al. (1998) and Lussiez et al.
(1977). Cerpa el al. (1996) report a correlation between the concentra-
tion of goethite in Moa Bay ore and the slurry properties.
Acid leaching occurs in titanium-lined autoclaves (Australian
projects) o r acid brick and lead-lined autoclaves (Moa Bay project)
540 B. I. WHITTINCTON AND D.M U l R

(Kyle, 1996) a t temperatures of approximately 250°C. The use of tita-


nium is of particular importance to the Australian processing plants
since this can resist attack from any chloride present in hypersaline
process water. After digestion, the slurry is flashed to approximately
100°C and the solids separated from the liquids in a counter-current
decantation (CCD) thickener well. The nickel and cobalt present in the
liquor are then precipitated as either the sulphides (using H2S, Fig. la),
carbonates o r hydroxides (using, for example, MgO, Fig. I b) o r ex-
tracted directly from the acidic leach liquor using solvent extraction
(Fig. 2a). The nickel sulphide or nickel hydroxide intermediates are
Downloaded by [Aston University] at 02:54 21 January 2014

then dissolved and further refined by solvent extraction/electrowinning


(Fig. 2b). The operating parameters of the various processing plants
(actual or expected) are discussed below and presented in Table 111.
The post-digestion liquor can be treated by a number of methods.
The sulphide precipitation process displays excellent selectivity for
cobalt and nickel, but not other impurities (Motteram et al., 1996).
In particular, the nickel, cobalt and zinc precipitate as the sulphides
while the magnesium, manganese, iron and aluminium remain in solu-
tion. The hydroxide precipitation has the disadvantage that copper,
zinc and some magnesium and manganese also precipitate with the
nickel and cobalt (Kyle, 1996). However, the nickel/cobalt hydro-
xide ammoniacal ammonium carbonate leaching is reportedly simpler
than the oxidative leaching required by the sulphide precipitation, is
selective for nickel, cobalt, copper and zinc and rejects iron, manga-
nese o r magnesium (Kyle, 1996; Mason et al., 1997).
One of the critical operating parameters of the pressure acid leach
process is the acid (sulphur) consumption. Magnesium, which is found
in saprolite ores, increases the acid consumption and hence reduces
the economic viability of the pressure acid leach for processing
magnesium-rich saprolite ores. The Amax process, a variation on
the pressure acid leach process, allows the processing of both limonite
and saprolite ores (to 15% magnesium; Taylor, 1997a) with minimal
acid-consumption penalty. A much simplified flowsheet for the Amax
process is presented in Figure 3. The coarse, high-magnesium content
saprolite ore is calcined to "activate" the magnesium and then fed to
the atmospheric leach circuit where it is leached in acidic leach liquor
recycled from the pressure acid leach circuit (this liquor still contains
in excess of 50 g/L H2SO4). The solids are separated from the resultant
Downloaded by [Aston University] at 02:54 21 January 2014
water

-
*?+ coarse
rejects
NilCo hydroxides
Downloaded by [Aston University] at 02:54 21 January 2014

pressure leach wash water residue


steam

Flash tanks
phosphate -1 manganese removal

.
CCD thlckenlng
$ water wash water solids
r )tailings
nickel
cathode
limestone product

cobalt
wash water dewatering sulphide
product

solution bleed

alkal
cobalt
'
i NH, recovery

NH, :t leach

nlckel cathode

FIGURE 2 Flowsheet for (a) the direct solvent extraction o f nickel/cobalt the refining o f mixed nickeljcobalt hydroxides (from Taylor,
ACID LEACHING OF NICKEL LATERITES

ore feed

water-
4
Feed preparation

acid
steam

~~
Downloaded by [Aston University] at 02:54 21 January 2014

tailings

I preleaching

resldue

solution
barren
solution
"9 H,S precipitation to recvcle

P
mixed sulph~d or waste
refining recycle

products

FIGURE 3 Amax process Rowsheet for refining Mg-poor limonite and Mg-rich
saprolite ores (from Taylor, 1997a).

slurry and subjected, with the fine, low magnesium content (limonite)
ore, to a pressure acid leach at 270°C. The "preleaching" of the high-
magnesium content ore has the advantage that it consumes residual
acid present in the original pressure acid leach liquor and also lowers
the magnesium concentration of the solids prior to the second pressure
acid leach (Krause er al., 1998 and references therein; Duyvesteyn
et al., 1979). This process has been tested by Calliope for refining New
Caledonian ores and is discussed in Section 4.1.4.
A modification to the Amax process - the Sulzer regenerative acid
leach (Sural) process - crystallises magnesium sulphate present in the
544 B. 1. WHITTINGTON AND D. MUIR

liquor after neutralisation with the high-magnesium garnierite. This


is then decomposed to MgO and SO2; the MgO being processed
for subsequent sale as a refractory while SO2 is converted to sulphuric
acid (Monhemius, 1987).

4.1.1. Production via Nickel Sulphide intermediates:


Murrin Murrin and Moa Bay
The following description of the Murrin Murrin processing plant is
modified from that presented by Motteram et al. (1997). Slurried ore -
Downloaded by [Aston University] at 02:54 21 January 2014

prepared from borewater containing approximately 600ppm chlor-


ides - is fed to three stages of steam injection preheaters. The slurry is
then injected into one of four autoclaves where it is digested for 90
minutes at 25S°C. The slurry is then flashed through three flash tanks
and the solid separated from the liquor using seven stages of CCD
washing. The liquor is neutralised to 1 g/L acid by addition of calcrete
and the gypsum removed and sent to the CCD circuit for washing.
Nickel/cobalt sulphides are precipitated from the neutralised solution
at 90°C by adding HIS at 105 kPa. Chlorides are removed from the
mixed sulphides by washing with water obtained from a reverse osmo-
sis plant and the sulphides dissolved using oxygen under pressure. The
resultant leach solution is sent to solvent extraction for removal of
the cobalt and impurities (e.g. zinc, manganese) then cobalt metal
produced by hydrogen reduction/briquetting. Nickel metal is produced
from the nickel-rich raffinate by hydrogen reduction/briquetting.
The Cuban Moa Bay processing plant, currently (early 1998) the
only operating pressure acid leach processing plant, has a similar
overall process flowsheet to that from the Murrin Murrin processing
plant. Hazen and Chou (1997) and Chalkley et al. (1996) discuss the
operation of the Moa Bay processing plant.' The ore is acid leached
to give a post-digestion liquor with a nickel concentration of 5.1 -7.0
g/L and a free acid concentration of between 20-30g/L (Sobol,
1969; Hazen and Chou, 1997). Seven CCD stages are used at Moa Bay
with a CCD ovedow concentration of 3.5g/L nickel and 15g/L
H2SO4 (Hazen and Chou, 1997).

'The operating conditions presented by Hazcn and Chou (1997) are slightly different
to those reported by Kyle (1996).
ACID LEACHING OF NICKEL LATERITES 545

The Moa Bay processing plant adds hydrogen sulphide to the raw
liquor prior to neutralisation of free acid to reduce Fe(II1) to Fe(I1)
and Cr(V1) to Cr(II1) and precipitate copper as copper sulphide
(Chalkley et al., 1996). Subsequent neutralisation of the free acid
present in the post-digestion liquor to pH 2.2-2.4 is accomplished by
addition of coral (Kyle, 1996). Sulphide precipitation at Moa Bay uses
H2S at 120°C and 1000 kPa and produces a nickel-cobalt sulphide
concentrate. Addition of recycled sulphide seed material promotes
the kinetics, promotes particle growth and reduces reactor scaling
(Motterham er a!., 1996; Matos, 1997).
Downloaded by [Aston University] at 02:54 21 January 2014

4.1.2. Production via Nickel Hydroxide


Intermediate: Cawse
The Cawse processing plant will reportedly upgrade the ore by an
initial screening prior to processing, then subject the ore to a pressure
acid leach (Kyle, 1996). The Cawse autoclave allows control of slurry
Eh by addition of either air or sulphur. The cooled slurry is neutralised
to pH 3.5 with limestone and any Fe(I1) oxidised to Fe(II1) with air.
The liquor is separated from the solids (CCD), neutralised with lime-
stone to pH 6.0 and any residual iron oxidised to Fe(II1) with air.
This precipitate is recycled to recover any nickel and cobalt co-pre-
cipitated with the iron during this stage. The mixed nickel/cobalt
hydroxides are then precipitated with MgO and redissolved in
amrnonia/an~moniumcarbonate. Oxidation of the Co(I1) ammine to
Co(II1) ammine, which is required for efficient separation of the nickel
from cobalt, occurs readily at this stage. The nickel is then separated
from the cobalt by solvent extraction and the nickel metal produced
by electrowinning. The cobalt-containing liquor is treated with am-
monium sulphide to a cobalt sulphide product (Kyle and Furfaro,
1997; Krause et al., 1998 and references therein).

4.1.3. Production of Nickel and Cobalt by Direct Solvent


Extraction of Acid Llquoc Bulong
The following is a simplified description of the Bulong nickel laterite
processing plant operation from a presentation by Hayward (1998).
Ore is crushed and slurried then thickened to a density suitable for
546 B. I. WHllTINGTON AND D. MUlR

leaching. The slurry is heated (one of four slurry preheaters) then


leached in a six-compartment horizontal autoclave vessel. The slurry
is flashed (four slurry flash vessels) and the cooled slurry processed
through a seven stage CCD thickener circuit, where it is washed and
separated into pregnant liquor and leach residue tailings slurry. The
leach residue tailings slurry is neutralised with limestone and lime
then pumped to the leach residue tailings dam. The pregnant liquor
solution, containing the nickel and cobalt, is partially neutralised
with limestone to pH 4.0 (Kyle, 1996) and the iron, aluminium and
chromium precipitated as the hydroxides. The neutralised pregnant
Downloaded by [Aston University] at 02:54 21 January 2014

liquor is then clarified, passed through sand filters to remove any


remaining solids, and then to the solvent extraction stage.
Cobalt, manganese and zinc are extracted from the liquor, then
nickel extracted from the resultant aqueous phase using a second
solvent extraction stage where it is recovered as the metal by elec-
trowinning. Cobalt is precipitated as cobalt sulphide from the loaded
strip 'liquor using sodium sulphide. The filter cake is releached by
pressure oxidation in an autoclave at 140°C, the solution purified of
zinc and copper by solvent extraction and ion exchange and the cobalt
metal obtained by electrowinning.

4.1.4. Production of Nickel and Cobalt Using Amax


Process: Calliope
Calliope tested an Amax-type pressure acid leach for refining New
Caledonian mixed limonite/transition ore to nickel and cobalt sul-
phides (Faris er al., 1997). The process is similar to that described in
Section 4.1, with the high-magnesium ore consisting of + 10 mesh
fraction and the limonite ore being the finer -10 mesh fraction.
Optimum nickel and cobalt extractions of 94 and 96% were obtained
during trials. Chromate in the solution is reduced to the chromic state
in a separate reduction step.

4.2. Effect of Acid-Leaching Process Conditions


on Processing Plant Operation
4.2.1. Ore Grind
Chou el al. (1977) report no effect on the rate nor extent of nickel
extraction upon grinding Moa Bay limonite. However, grinding
ACID LEACHING OF NICKEL LATERITES 547

increases the post-digestion liquor nickel : impurity ratio (i.e. Ni/


(A1 + Fe)), which suggests the finer ore particles enhance the
precipitation rate for the aluminium and iron salts.
Baillie and Cocks (1998) report grinding Weda Bay ore improves the
cobalt extraction kinetics during the subsequent high-pressure acid
leach.

4.2.2. Acid Concentration or Acid: Ore Ratio


Testwork from a number of researchers indicates it is possible to
Downloaded by [Aston University] at 02:54 21 January 2014

achieve high nickel and cobalt extractions during the pressure acid
leach process (Tab. IV).
One of the main requirements for high nickel and cobalt extrac-
tions is sufficient residual acidity in the liquor. Krause el al. (1998)
report the acid addition required for high Ni (95%) and Co ( > 90%)
extractions depends mainly on the concentration of Mg, Al and
FeC03 siderite in the ore. For leach temperatures between 240 and
270°C, Eq. (7) is reportedly valid

TABLE IV Nickel and cobalt extractions from the various ores examined during
testwork conducted for the various processing plants
acid time residual exfractions
addition digestion free acid (%)
Ore T("C) (kglkg ore) (hrs) (g/L) Ni Co Reference
Bulong 245-250 0.2-0.6 2 30-40 > 95 > 95 Taylor and
Cairns (1997)
nontronite ore 235-250 0.40 3 - 97 99 Tindall and
Muir (1997)
Cawse (talc or 250 0.30-0.37' - - > 95 > 95 Mason et al.
limo&) (1997)
Murrin 250-260 0.40 1 - 1.5 40 95 93 Motteram
Murrin er a / . 11 996)
Marlborouah- 250 - I - > 94 > 94 Grhen
(1 998)
Ramu mixed 250 0.27 I - > 96 > 96 Mason and
Hawker
limonitel (1998)
saprolite
' acid consumption.
548 B. I. WHlWlNGTON AND D.MUlR

where wt % &So4 is the mass of "pure" sulphuric acid required on a


dry ore basis and wt % Mg, Al, etc., are the elemental concentrations
in the ore (Krause et al., 1998).
Mason and Hawker (1998) report a residual free acid concentration
of 30g/L is optimal for efficient leaching of Ramu ore at 250°C (50-60
minute leach time, 0.2-0.3 kg acidlkg ore) while Kyle and Corrans
(1998) report sufficient dissolution of Ni and Co requires residual free
acid concentrations of between 30 and SOg/L H2SO4. Use of lower
acid concentrations, or acidlore ratios, reduces both the initial rate
and the ultimate extractions of nickel and cobalt (Georgiou, 1995).
Downloaded by [Aston University] at 02:54 21 January 2014

Chou et al. (1977) additionally report a drop in the Moa Bay acidlore
ratio from 0.30 to 0.27 requires a three fold increase in the leach time
to achieve 90% nickel extraction at 250°C and a 25% solids loading.
The use of higher acid concentrations also improves the product
settling (Krause et al., 1998) but increases the residual free acid con-
centration (Mason and Hawker, 1998) and increases the concentra-
tions of Fe, A1 and Cr in the liquor (Krause et al., 1998; Georgiou,
1995). The latter results from the highly acid leach liquor inhibiting
the hydrolysis reactions.
Berezowsky (1997) reports a strong interdependence between the
leach temperature and acid additions (i.e. acidlore ratios) - higher di-
gestion temperatures generally enable the use of lower acid additions
as more Fe and Al is hydrolysed.
Further evidence for the effect of acid concentration on the nickel or
cobalt extractions can be found in the papers by Chalkley et al. (1996)
and by Chalkley and Toirac (1997), examining the operation of the
Moa Bay processing plant. Here, a strong correlation exists between
free acid concentration (i.e. acid addition) and nickel or cobalt
extractions (the correlation graphs use daily data, which suggests no
other major changes to the processing plant operating conditions).
At the Moa Bay plant free acid in the post-digestion liquor, mag-
nesium dissolution and soluble aluminium/alunite scale formation
respectively account for 22, 15 and 50% of the total acid consump-
tion (Tab. V and also in Berezowsky, 1997). The remaining 13% acid
consumption is used in the dissolution of nickel, manganese, chro-
mium and iron (Tab. V and also in Hazen and Chou, 1997).
Acid consumption, and hence the required acid addition, varies
with ore type and linearly increases with ore MgO content. Serpentine
ACID LEACHING OF NICKEL LATERITES 549

TABLE V Average composition of Moa Bay latcritic ore and sulphuric acid balance
during leaching (Sobol, 1967)
Acid conrumption
Liquor Tailings
% dry Extraction to Acid & m a d Acid to toiling (kgjIO0 kg (kg1100 kg
Component basu solution (%) (kgjkg metal) (kgjkg metal) ore) ore)
Ni 1.35 95 1.67 2.140
Co 0.13 95 1.67 0.206
Fe (goethite) 47.5 0.35 2.62 0.435
Al (gibbsite) 4.0 15 5.49 3.290
Mn 0.76 40 1.78 0.543
Mg 1.O 90 4.03 3.640
Cr 2.0 5 - -
Downloaded by [Aston University] at 02:54 21 January 2014

Cu 0.024 100 1.54 0.037


Zn 0.040 100 1S O 0.060
Total combined acid 10.351
Free acid in liquor (28 g/L) 5.310
+
Total (liquor tailings) 23.84 kg/
100 ke ore

(Hazen and Chou, 1997) or saprolite ores (Baillie and Cocks, 1998)
require higher acidlore ratios than other ore types to obtain similar
high levels of nickel and cobalt extraction.

4.2.3. Solids Concentration


Krause et al. (1998) report the concentration at which the ore is fed
to digestion does not significantly influence the Ni and Co extractions
for solids loadings between 25-40%. However, it affects the dissolu-
tion of impurity elements.
Chou et al. (1977) report the initial rate of nickel extraction during
pressure acid leaching at 250°C increases with an increase in the solids
concentration from 10 to 33% solids, but slightly decreases with an
increase in the solids concentration from 33 to 45% solids.

4.2.4. Temperature
Chou et al. (1977) indicate a 250°C digestion temperature results in
a rapid extraction of nickel from a Moa Bay ore. A similar ultimate
extraction, but a higher initial rate, is obtained by increasing the di-
gestion temperature to 275°C. In contrast, leaching at 300°C lowers
550 B. I. WHlmlNGTON AND D. MUlR

the nickel yield. Leaching tests conducted by Da Silva (1992) also


indicate increasing the digestion temperature (from 220 to 270°C) can
reduce the nickel extraction. Precipitation of nickel-magnesium
sulphates (Mg, Ni)S04. H 2 0 is thermodynamically possible for leach
temperatures in excess of 270°C and formation of these materials
could account for any decrease in nickel extraction at these leach tem-
peratures. However, (Mg,Ni)S04. H 2 0 is more soluble at the lower
temperatures used for measuring nickel extraction and dissolution
upon cooling could give a misleading value of its solubility at high
temperature. Chou et al. (1977) report the nickel extraction has an
Downloaded by [Aston University] at 02:54 21 January 2014

apparent activation energy of 126 kJ/mole for temperatures between


225 and 275°C.
Georgiou and Papangelakis (1998) report the rate of nickel ex-
traction upon pressure acid leaching an Indonesian limonite ore
(230 to 270°C; 30% solids charge and 0.2acid/ore ratio) increases
with temperature. However, the cobalt extraction kinetics are general-
ly unaffected by temperature. In particular, approximately 80% of the
cobalt is rapidly extracted suggesting cobalt is present in a readily-
leached manganese phase (e.g. asbolan).
Leaching testwork conducted for the Syerston project indicates an
increase in nickel and cobalt extractions, but a decrease in iron dis-
solution, with increasing temperature (100-230°C) (Blight and Muir,
1996).
Chou el al. (1977) indicate the leach temperature, acidlore ratio
and leach time have a significant effect on the sulphate content of
the leach residue. In particular sulphur losses (i.e. solids sulphate
content) are greatest for high leach temperatures and high acidlore
ratios. This is important, since sulphur is a major operating ex-
pense in the acid leach process. The variation in sulphur content of
the leach residue with digestion time suggests initial formation of a
"transient" hydrolysis product with a high sulphur content such as
basic aluminium or iron sulphates. These decompose as the original
ore reacts, and the free acid concentration changes, to stable hydro-
lysis product(s) with a lower sulphur content (e.g. alunite and
hematite).
The digestion temperature also influences the [W/([Fe + All) ratio
in the liquor, where M = N i or Co (Georgiou and Papangelakis, 1998).
In particular, increasing the temperature from 230 to 270°C increases
the ratio obtained after 1 hour digestion. This arises from the increase
ACID LEACHING OF NICKEL LATERITES 551

in Fe and Al precipitation rate with increasing temperature and the


"inverse solubility" of the hematite and alunite solids.

4.2.5. Added Salts


The variation in liquor concentration of various elements during the
pressure acid leaching of an Indonesian limonite ore at 240°C is shown
in Figure 4. Of particular interest is the beneficial effect of added
salts (sea-water) at reducing the aluminium concentration in the
post-digestion process liquor (Fig. 4b). This is suggested to arise from
Downloaded by [Aston University] at 02:54 21 January 2014

formation of natroalunite, which is reportedly less soluble than the


hydronioalunite forming in the tap water experiments (Krause et al.,
1998). The concentration of iron is not significantly affected by the
addition of sea water (Krause et al., 1998) or salt (200 kgltonne, Blight
and Muir, 1996).
Leaching testwork conducted for the Syerston project indicates an
increase in nickel and cobalt extractions in the presence of salt (200 kg/
tonne) (Blight and Muir, 1996) while Kyle and Furfaro (1997) report
chloride present in saline water has a beneficial effect on the dissolu-
tion of nickel and cobalt from Cawse ore. The use of saline process
water is also claimed to enhance the leaching kinetics and recoveries,
and the settling and thickening characteristics, of Bulong and Cawse
project ores, but not the smectite ore used at Murrin Murrin (Kyle,
1996). Given that Bulong ore also contains smectite (nontronite), the
difference in the properties of the Bulong and Murrin Murrin ores is
surprising.
Das et al. (1997) report addition of 15g/L Na2S04 relative to addi-
tion of 15g/L (NH4)2S04 slightly increases the nickel and cobalt
extractions from laterite ores or chromite overburden, for pressure
leaches conducted at 250"C, 10% solids, 36.8 g/L H2SO4,10% W/W arn-
moniojarosite. Both ore-bodies contain goethite, and some hematite.
Das et al. (1995), in their 2' full factorial study on nickel, cobalt
and iron extraction, report nickel, cobalt and iron extractions decrease
with increasing (NH4)2S04 concentration.

4.2.6. Air Addition


In well weathered laterites, such as limonite, iron is in the form of
~ e ' + .However serpentine may contain some of the iron as ~e~~ and,
Downloaded by [Aston University] at 02:54 21 January 2014
ACID LEACHING OF NICKEL LATERITES 553

in the absence of other oxidising minerals, may result in a leach liquor


containing several g/L ~ e ~ While + . this is acceptable for sulphide
precipitation of nickel and cobalt from the leach liquor, it will cause
difficulties if the nickel refining process proceeds via an intermediate
nickel/cobalt hydroxide or if the post-digestion acid leach liquor
undergoes direct solvent extraction (as at the Bulong processing plant)
(Hazen and Chou, 1997).
Oxidation minimises the ferrous ion concentration. Vracar and
Cerovic (1997), who examine the rate of FeS04 oxidation under con-
ditions different to those in laterite liquors, report the oxidation rate
Downloaded by [Aston University] at 02:54 21 January 2014

increases with temperature (50-20O0C, IOg/L Fe(II), no H2S04, O2


partial pressure 1013 kPa) and oxygen partial pressure (203 - 1013 kPa)
but decreases with increasing initial Fe(I1) concentration (10-50g/L
Fe(I1)) or increasing acid concentration (0 or 50g/L H2S04, 140 or
200°C).
The Fe(I1) oxidation has an apparent activation energy of 51 kJ/
mol, which Vracar and Cerovic (1997) report indicates a diffusion
controlled reaction. Thus, faster reaction rates would be expected by
increasing the autoclave stirring speed above 400 rpm (not experimen-
tally verified).
Neutralisation test work conducted for the Cawse project indicates
control of oxidation in the pressure leach is important (Mason et al.,
1997). In particular iron present as Fe(III), but not the more soluble
Fe(II), is.almost completely precipitated upon subsequent neutralisa-
tion by limestone to pH 3.5. Residual Fe(I1) still present in solution
is removed in a subsequent stage by aeration at higher pH. Any nic-
kel and cobalt which are precipitated at this stage are recycled for
re-dissolution.
Baillie and Cocks (1998) report the effect of an overpressure of air
present during the high-pressure acid leaching of Weda Bay ore at
250°C. While air does not influence the nickel and cobalt extraction
kinetics, it reduces the total concentration of iron in solution (from 7.7
to 1.3 g/L) and consequently reduces the acid demand from 0.45 to
0.42 kg/kg. This negligible effect of air on the nickel and cobalt ex-
traction kinetics apparently contradicts the results from Tindall and
Muir (1998), who report lower nickel extraction rates for ores con-
taining oxidising materials such as M n 0 2 (Section 5.3.4). However,
air has a lower Eh than MnOz.
554 B. I. WHI'ITINGTON AND D.MUlR

Mason and Hawker (1998) also report addition of air after


autoclave digestion of Ramu ore (i.e. during autoclave digest "let-
down") reduces the liquor ~ e concentration
~ + from 2.5g/L to less
than 0.1 g/L.

4.2.7. Effect of Digestion Parameters


on Settling Properties
Tindall and Muir (1997) report the post-digestion slurries from
nontronite ore are more poorly settling than those from a limonite ore
(31 -39% solids underflow versus 55% solids underflow for a similarly
Downloaded by [Aston University] at 02:54 21 January 2014

treated limonite ore). However, increasing digestion temperature


(235-25O0C), increasing acidlore ratio (0.2-0.4) and decreasing stir-
ring speed (410- 130rpm) increase the settling rate and underflow
density of the slurry from digestion of a nontronite ore (Tindall and
Muir, 1997).
Tindall and Muir (1997) report a two stage process -digestion at
15O0C/45minutes then digestion at 250°C - improves the leach residue
settling for nontronite ore. The initial (low temperature) digestion
reportedly allows polymerisation of silica in solution while the sub-
sequent (high temperature) digestion allows transformation of iron .
oxide and subsequent coating of the silica particles (Tindall and
Muir report leach residues with high settling rates have high PZC
values and consist of predominately an iron oxide surface while ore
residues which have a silica coating have a lower PZC value and hence
a lower settling rate).
Briceno and Osseo-Asare (1995) report the PZC value of the post-
digestion solid, obtained by digesting a limonite/garnierite ore at
275"C, is lower than expected for a product with a high hematite
content (PZC observed at pH x 2 while the hematite PZC reportedly
occurs at pH x8.5). This suggests dissolution of the silicate mineral
coating the original ore particles and subsequent deposition of silica
on the iron oxides (Briceno and Osseo-Asare, 1995). By contrast, the
PZC value of a limonite ore after digestion at 250°C (pH 3.8) is lower
than that of the original ore (PZC at pH 2), suggesting redeposition
of silica is not as pronounced under these conditions and/or the pre-
sence of aluminium sulphate shifts the mobility curve to higher pH
values (Briceno and Osseo-Asare, 1995). As previously mentioned,
ACID LEACHING OF NICKEL LATERITES 555

the PZC values influence the settling properties of the post-digestion


slurries.
Chou et al. (1977) report the settling of post-digestion slurry de-
pends on the agitation occurring during pressure leaching. In parti-
cular, high shear2 results in a slurry with poor settling properties (long
time to settle and low compaction of the solids). In this instance, fine
particles probably inhibit the settling (although particle size dis-
tributions were not determined during these tests) and the use of low
shear probably allows particle aggregation.
Krause et al. (1998) report the settling characteristics of the post-
Downloaded by [Aston University] at 02:54 21 January 2014

digestion slurry are expected to depend on the degree of conversion of


goethite. Since most of the nickel in limonite ores is contained within
goethite, the goethite conversion can be approximated by the nickel
extraction. The results of Krause et al. (1998) indicate an increase in
settling rate with increasing Ni extraction.
High leach residue settling rates are generally obtained with high
leach temperatures, high acid additions (high acidlore ratios) and
longer leach retention (Duyvesteyn er al., 1979; Kyle and Corrans,
1998) - factors which would all be expected to increase the goethite
conversion. The settling properties also usually improve with de-
creasing Mg ore-feed content (Krause el al., 1998).
Tindall and Muir (1997) report formation of jarosite is undesirable
relative to hematite due to the resultant increase in residue mass (due
to incorporation of sulphate, hydroxide and sodium) and increase in
underflow volume.
Moa Bay ore contains approximately 2% chromium and the
chromium chemistry can influence the operation of the Moa Bay pro-
cessing plant (in contrast, nontronite ore may contain only ~ 0 . 3 %
Cr). Chromate ion (Cr0;-), which forms from oxidation of c r 3 + with
MnOz, is incorporated into alunite with silicic acid (Section 5.4.1).
This mechanism reportedly results in a cleaning of the liquor of silica
and prevents gel formation. However addition of sulphur at Moa Bay,
to control the Eh, prevents formation of 00:-and the Si-Cr con-
taining alunite species. Under these conditions, the presence of "free"

In these tests, low shear indicates mixing with a perforated paddle at 50rpm while
high shear indicates mixing with a twin turbine at 600rpm.
556 B. I. WHIITINGTON AND D.MUIR

silica gel reportedly halves the thickener performance (Tindall and


Muir, 1997 and references therein).
Free acidity also influences the silica chemistry. In particular, the
reduction in free acidity on passing from the autoclave discharge to
the wash-plant CCD system (from 26-30g/L to 10- 14g/I) may
hydrolyse the silicic acid (H4Si04) to colloidal Si02. x H 2 0 and hinder
the solid/liquid separation process (Sobol, 1969).

5. CHEMISTRY OF THE PRESSURE ACID LEACHING


Downloaded by [Aston University] at 02:54 21 January 2014

OF NICKEL LATERITES

5.1. Sulphuric Acid Chemistry


The free acid concentration in a liquor can be expressed by Eq. (8):

where [SO:-],,,,, is the total sulphate in solution at the end of the


reaction and [SO:-]bo,,, is sulphate stoichiometrically related to
dissolved metal salts. However, the extent of second dissociation of
sulphuric acid in the forward reaction in HSO; ++ Hf +SO:- de-
creases with increasing temperature so that for temperatures in ex-
cess of 150°C, sulphuric acid behaves as a monoprotic acid (Krause
ef al., 1998). An amount of H2SO4 in excess of that expected from
full dissociation is therefore required to maintain a specific hydrogen
ion activity at temperature. Additionally, added metal sulphates act
as a proton sink, further decreasing the hydrogen ion activity. For
example, MgS04 dissociates and the SO:- reacts with H + to HSO;
(Krause el al., 1998 and references therein). The effect of added salts
on the mineralogy of the post-digestion iron phases is discussed further
in Section 5.3.1.

5.2. CobaltlNlckellManganese Chemistry


Eh/pH diagrams indicating the stability field for Co-S-H20 or Ni-
S - H 2 0 salts at 250°C are presented by Da Silva (1992). These
diagrams indicate residual acidities in excess of 18 g/L or 1 g/L are
required if the cobalt or nickel are not to precipitate as the respective
ACID LEACHING OF NICKEL LATERITES 557

monohydrate "MeS04.H20" salts at 250°C. This result, and the


observation that sulphuric acid behaves as a monoprotic acid at high
temperatures, but not at low temperatures (Section 5.1), provide a
theoretical basis for the empirical observation that moderately high
residual acidities are required for effective nickel and cobalt leaching
(Section 4.2.2).
Manganese dioxide minerals are usually rich in cobalt, but are not
completely dissolved unless a reducing agent is present (e.g. Eq. (9)).
This can have a negative impact on cobalt recovery (Tindall and Muir,
1996; Section 5.3.4). For example, Tindall and Muir (1996) report
Downloaded by [Aston University] at 02:54 21 January 2014

the poor extraction of cobalt upon pressure acid leaching a limonite


ore, with a high (14%) M n 0 2 content, arises due to the presence of
unleached lithiophorite. In contrast, high cobalt and manganese
recoveries are obtained from nontronite ores, since reduction of the
Mn(IV) in lithiophorite with Fe(I1) present in the clays allows dis-
solution of the Mn(I1) and cobalt (Tindall and Muir, 1997). Addi-
tion of other reducing agents may generate a liquor with a high
ferrous ion concentration, which would create problems for processes
involving an intermediate hydroxide or direct solvent extraction of
the acid leach liquor (Hazen and Chou, 1997).
Typically, about 50-60% of the manganese present in a laterite ore
solubilises as sulphate (Taylor, 1997a). Results presented by Krause
et al. (1998), for the leaching of New Caledonian ore, an Indonesian
ore (Fig. 4) or Moa Bay ore (Tab. V), indicate manganese extrac-
tions of 72-90%, 50% or 40% respectively. Motteram el al. (1996)
report 91% of the manganese is extracted from a Murrin Murrin
ore while Tindall and Muir (1997) report 88- 100% manganese dis-
solution upon digestion of a nontronite ore.
By analogy with deep sea manganese nodules, some of the cobalt
and nickel present in the laterite manganese minerals are in the tri-
valent state (see Sobol, 1968). In these instances, leaching of the late-
rite ore may occur by Eqs. (10a) and (I I) (Canterford, 1985; where [O],
indicates a species oxidised by the higher valent Mn, Co or Ni).
558 B. I. WHITTINGTON AND D. MUlR

These equations are probably simplifications, especially since research-


ers indicate dissolution of M n 0 2 is known to involve the formation
of Mn(II), Mn(II1) and Mn(1V) species (Canterford, 1985).
Georgiou and Papangelakis (1998) report the pressure acid leaching
of nickel from an Indonesian limonite ore, conducted in a batch
Downloaded by [Aston University] at 02:54 21 January 2014

autoclave at 230- 270°C, 10- 30% solids loading, 0.15 -0.35 acidlore
ratios and 400-650 rpm agitation speeds, can be accurately modelled
if the assumption is made that diffusion of sulphuric acid through
the particle pores is the rate controlling step. The diffusivity values -
calculated from the experimental leach data - are very low and further
support the assumption of pore diffusion control for nickel laterite
leaching (Georgiou and Papangelakis, 1998). Small pores, 20- 30 nm
at the entrance but decreasing with depth to less than 2-3 nm dia-
meter, have been observed in atomic force microscope images of a
synthetic goethite (Fischer el al., 1996).

5.3. Iron Chemistry


Lateritic ores contain significant quantities of iron-containing minerals
(e.g. goethite, nontronite) which may also contain nickel and cobalt.
Acid leaching of these minerals should therefore be required to lib-
erate the nickel and cobalt. Dissolution of theiron requires consump-
tion of significant quantities of acid (Eqs. (12a,b)) although the
resultant precipitation of iron may, if the digestion is conducted at a
high temperature, regenerate acid (Eqs. (12c,d), (13a,b)). Formation
of jarosite entails a greater loss of acid from the liquor than formation
of hematite (cf. Eqs. (12) and (13)).

2 FeOOH + 3 H2S04 --t Fe2(S04), + 4 H20 (12 4


ACID LEACHING OF NICKEL LATERITES 559

Hayward (1998) suggests the additional leach reactions (12b) and


Downloaded by [Aston University] at 02:54 21 January 2014

(13b) occur during digestion of Bulong ore (Eq. (13b) differs from
(13a) in that more explicit account is made of the fate of any chloride
from saline waters). Hayward (1998) assumes dissolution of goethite
and nontronite (Eqs. (12a) and (12b)) occurs within the first few
minutes of leaching and that the resultant iron sulphate hydrolyses to
hematite and natrojarosite (Eqs. (12d) and (13b)).
The iron chemistry can also influence the effectiveness of the post-
digestion operation. For example, we have already seen the degree of
iron hydrolysis may influence the filterability and settling properties of
the post-digestion slurry (Section 4.2.7).

5.3.1. Mineralogy of the Post-digestion Iron Phases


Dissolution of iron present in the nickel laterites yields soluble ferric
sulphate, which subsequently hydrolyses to hematite or basic ferric
sulphates. The mineralogy of the final products depends on the liquor
acidity and digestion temperature (Hazen and Chou, 1997 and refer-
ences therein).
Tozawa and Sasaki (1986) report ferric sulphate hydrolyses in acidic
media to Fe203 and/or Fe(OH)S04 upon digestion at temperatures1
times in excess of 17OoC/30 hours. The product composition, and
hence the sulphur content of the solid, depends on the free acid in the
liquor. At the breakpoint acid concentration (69glL H2SO4at 200"C,
see Tab. V1) a slight increase in free acid results in the formation of
Fe(OH)S04 while a decrease results in the formation of Fe203 (these
phases contain 19.1 and 0 wt % S respectively).
Table VI illustrates addition of ZnS04, MgS04, CuS04 or Na2S04
alter this breakpoint acid concentration. Addition of sodium
560 B. I. WHI'ITINGTON A N D D. MUIR

TABLE V1 Effect of additives on the "break-point" acid concentration (above this con-
centration a sulphur- "rich product forms, below this concentration a sulphur- "poor"
product forms). Digestion times are in excess of 30 hours. From Tozawa and Sasaki
(1986)
Break-poinr:/ree-acid concenrrarion Sulphur concentration in solid
( a l L HSOJ (wt%)'
< breakpoint > breakpoinr
Additives k l L ) 185°C 200°C acid conc. acid. conc.
none 55 69 = 1.2 18.8
Downloaded by [Aston University] at 02:54 21 January 2014

'Weight % S: Fe20, (0%): Fe(OH)S04 (19.1%); natrojarasite N~F~J(SO.)~(OH)~


(13.2%).

significantly decreases the stability field of hematite and additionally


changes the iron mineralogy from that observed with the other
salts. In particular natrojarosite, and not Fe(OH)S04, forms for acid
concentrations above the breakpoint value (Tozawa and Sasaki,
1986).
Tindall and Muir (1997) also report formation of jarosite in post-
digestion liquors containing added Na2S04 (digestion at 250°C,
acidlore 0.4). The jarosite: hematite ratio increases with increasing
Na2S04 addition to the 21000ppm upper limit examined. Addition
of sodium as NaCl slightly reduces the jarosite: hematite ratio rela-
tive to addition of Na2S04 (1 IOOOppm Na in both). In contrast, no
jarosite is observed by X-ray diffraction if the digestion is conducted
at the higher temperature of 263°C.
Other researchers obtain similar results to those presented in Table
VI. Papangelakis et a/. (1994a) report addition of Zn (97glkg) to the
leach solution increases the dominance of the FeS04HS04 species for
temperatures between 150 and 200°C. Thermodynamic calculations
indicate addition of ZnSO4 significantly decreases the free acid con-
centration, which drives the reaction presented in Eq. (14) (Section
5.3.2) to the right and decreases the solubility of iron in the liquor
(Papangelakis et al., 1994a).
ACID LEACHING OF NICKEL LATERITES 56 1

5.3.2. Mechanism of Iron Hydrolysis


Papangelakis et 01. (1994a) conduct thermodynamic modelling of
hematite formation in acid systems. Their model includes iron hy-
drolysis and interaction with complexing anions such as OH- and
SO:-. The effect of H2SO4 concentration on the iron solubility,
which is calculated using reported thermodynamic parameters, shows
a similar trend to results obtained by other researchers at 150-
200°C (i.e. the liquor iron concentration increases with increasing
free acid concentration and decreasing temperature). Interestingly, ex-
Downloaded by [Aston University] at 02:54 21 January 2014

cluding the FeS04HS04 species from the model results in a poor


predictive model, suggesting this previously postulated species is a
true species under the conditions examined in this paper. A similar
conclusion is made for the mixed hydroxide/sulphate species (e.g.
Fe(OH)S04).
Papangelakis er al. (1994a) use their thermodynamic model to
calculate the distribution of the iron species present at the various
temperatures (150, 170, 185 and 200°C) and acid concentrations
(20- 100g/kgH2S04).FeS04HS04 and F ~ H S O are~the most domi-
nant species at all temperatures and acid concentrations examined.
The dominance of the FeS04HS04 species is suggested to explain
why basic ferric sulphate forms at high acid concentrations.
Papangelakis et al. (1994a) report the liquor concentrations of
HSO;, SO:- and H + are similar at 185 or 200°C and suggest the
hematite precipitation is mainly controlled by the way the equilibrium
constant for Eq. (14) varies with temperature.

Tozawa and Sasaki (1986) conduct a thermodynamic analysis of


the iron-sulphur-water system and report the establishment of the
bisulphate-sulphate buffer reduces the free acid concentration, which
increases the apparent acid breakpoint level (hematite forms in liquors
with acid concentrations below this breakpoint value). Refer to the
above discussion on how the sulphate-bisulphate system influences
the hydrogen ion activity at temperature (Krause et al., 1998;
Section 5.1).
562 B. I . WHITTINGTON A N D D. MUlR

5.3.3. Effect of Reaction Conditions on Iron


Concentration in the Liquor
The variation in liquor iron concentration during the digestion of a
goethite-rich limonite ore at 250-270°C is presented in Figure 5.
These results, from Papangelakis et a/. (1996), indicate the iron
concentration increases at short digestion times then decreases to ah
equilibrium value, consistent with a goethite dissolution-iron precipita-
tion mechanism. The maximum concentration ofiron in solution ( x 1 g/
L) accounts for less than 2% of the total, suggesting hematite forms
Downloaded by [Aston University] at 02:54 21 January 2014

very rapidly - probably through a homogeneous nucleation mecha-


nism because of the very high supersaturation. Thus, goethite dissolu-
tion appears to be the rate determining step. Basic iron sulphate is not
observed by X R D or TEM as an intermediate under the conditions
used in these leaching experiments. Results presented by Krause et a/.
(1998) for the leaching of New Caledonian ore o r an Indonesian ore
(Fig. 4). o r for the leaching of Moa Bay ore (Tab. V), similarly indi-
cate low iron extractions (0.2-0.5%, 0.8% and 0.4% respectively).
Tindall and Muir (1996) report addition of hematite seed increases
the goethite-hematite transformation rate, probably by removing the

Leaching time (minutes)


FIGURE 5 Concentration of dissolved iron during leaching of a goethite-rich limonite
ore at 22% solids loading. Initial iron charge approximately 140g/L (from Papangelakis
cr a/., 1996).
ACID LEACHING OF NICKEL LATERITES 563

requirement for nuclei formation prior to precipitation. This result


indicates experimental conditions exist in which goethite dissolution is
not the sole contributor to the observed rate.
The results in Figure 6 indicate the "equilibrium" iron concentration
is, for a given acid concentration, higher for slurries at lower tempera-
tures (250°C versus 270°C) or for less dense slurries. The latter suggests
"secondary nucleation on crystal fragments or on growing hematite
particles play a significant role in hematite precipitation" (Papangelakis
et al., 1996). The equilibrium iron concentration also increases with
increasing acid concentration (Fig. 6). Umetsu et al. (1977) also report
Downloaded by [Aston University] at 02:54 21 January 2014

an increase in the equilibrium iron concentration with increasing acid-


ity at 200°C according to Log [Fe]= 3.9 log [H2S04]-7.00 (concentra-
tions in g/L, equilibrium between ferric oxide/Fe(III)/H2S04).
Acharya rt al. (1992) report increasing the digestion temperature
from 170 to 210°C or decreasing the sulphuric acid concentration in
the liquor from 161 to 140g/L (190°C digestion) both decrease the
concentration of iron present in the liquor after digesting zinc leach
residue. Blight and Muir (1996) also report increasing the digestion
temperature (130-230°C) reduces the concentration of iron in the
post-digestion slurry

Free &SO4 concentration (g/L)


FIGURE 6 Effect of free sulphuric acid concentration, temperature and slurry density
on the "equilibrium" iron concentrations in the leach liquor (from Papangelakis er a/.,
564 8. I. WHlTTlNGTON A N D D. MUlR

Results presented by Blight and Muir (1996) indicate use of saline


process water (200 kg/tonne salt) for digestion does not significantly
affect the concentration of iron in solution. The minimal effect of salts
on the post-digestion slurry iron concentration is also reported by
Krause el al. (1998) and presented in Figure 4. Natrojarosite is
expected to form in saline waters of a high sodium concentration
(Eq. (13)) (Kyle, 1996).
Sobol (1969) reports the presence of N?+, Co2+ or ~ n in ~syn- +
thetic leach liquor decreases the liquor concentrations of AI'+, ~ e ' +
or Cr3+ at 230-250°C. Ni, Co and Mn can substitute into alunite
Downloaded by [Aston University] at 02:54 21 January 2014

(Section 5.4.1) suggesting the Ni, Co or Mn alunite/jarosite species


have a lower equilibrium solubility than "pure" alunite/jarosite.
Acharya el al. (1992) also report the concentration of iron present
in the liquor, after digestion of 150g/L zinc leach residue in 140g/L
sulphuric acid at 170°C, follows the trend Na+ < NH: < K' < no
additives. Examination of the stability field reported by Acharya et al.
(1992) suggests the acid concentration and digestion temperature used
in this study would precipitate jarosite and that the different iron con-
centrations reflect the different jarosite solubilities (i.e. Na-jarosite
is expected to be less soluble than K-jarosite). However, the authors
did not characterise the solids. Haigh (1967) reports a substantially
different trend where the solubility of iron (111) in the presence
of Na2S04 > Na2C03 > NH40H > K2SO4 after high-temperature
(160-200°C) sulphuric acid leaching of zinc residues (i.e. K-jarosite
is less soluble than Na-jarosite).
Results by Umetsu er al. (1977) also suggest the resultant iron
species influences the equilibrium Fe(II1) concentration (hydrolysis of
a Fe2(S04), solution at 200"C, no acid added - acid forms during
hydrolysis). In particular, since the experiments conducted with no
additives or 71 g/L Na2S04 both have similar calculated final acidities
(Tab. VII), we suggest Na-jarosite is less soluble than Fe(OH)S04
under the conditions used in this experiment.3 However, we cannot
comment on the relative solubility of hematite due to the difference in
calculated final acidities (Tab. VII).

'Tozawa and Sasaki (1986) indicate both Na2S04and H2S04concentrations influence


Na-jarosite solubility at 200°C.
ACID LEACHING O F NICKEL LATERITES 565

TABLE VII Effect of additives on the mineralogy of the final phases forming upon
hydrolysis of Fe2(S04)3 at 200°C for 12- 18 hours. From Umetsu e t a / . (1977)
Sulphur lnirial Fe(l1 I) Resultant solid Fe(I1I) equilibrium Calcularedfnal
species added conc. (g/L) phase conc. (g/L) concentrarion'
none sz 25 Fe(OH)S04 1.9 34 g/L H2S04
+
(= 80%) Fe203
71 e1L Na,SO. sz 20 Na iarosite 0.4 34 elL H,SOd.

' ~ h sto~ch~omclry
c olthc Fc(lll) rcaclmg H2SOngcncralcd 13 assumed to correspond to that In Eqs
(12c d) (Fc2Od. Eq 0 3 a ) ( N ~ ~ c , ( S O , J ~ ( O H
Eq) ~(12c)
) (Fc(0H)SO.) Add~t~onally,
slnu thc
subhur conlcnt In the Precwmie 1ncre3seswllh ml,al Fc(lll) conccntranon (Umcrsu er 0 1 , 1977),
Downloaded by [Aston University] at 02:54 21 January 2014

thiphasc mmeralogy riporkd here may not be that cnpated by comparison with Table VI

5.3.4. Studies Examining the Leaching of Goethites


and Clays
A number of studies examine the effect of various factors on the
goethite to hematite transformation rate. Results from Tindall and
Muir (1996) indicate Na2S04or MgS04 enhance the rate of goethite-
hematite transformation while addition of A12(S04)3 or Cr2(S04)3
completely halt the transformation. The inhibition of the transforma-
tion by the latter two salts suggests the aluminium or chromium ions
adsorb onto the surface, thereby halting reaction.
The effect of Eh (ix. presence of oxidising or reducing agents
in the slurry) has a dramatic influence on the goethite to hematite
transformation rate. In particular, addition of Fe(I1) results in a rapid
and complete goethite transformation within 45 minutes at 250°C
while the transformation is incomplete 90 minutes after M n 0 2
addition (Tindall and Muir, 1996). Tests conducted using a limonite
ore, with a high (14%) M n 0 2 content, confirm the deleterious effect
of M n 0 2 on goethite-hematite transformation and hence the nickel
recovery (Tindall and Muir, 1996). Analysis of the leach residue
from this test indicates 60% of the nickel losses are associated with
unleached lithiophorite, the remainder associated with unleached
goethite. Cobalt losses are associated with unleached lithiophorite.
More recent work by Tindall and Muir (1998) confirms the effect
of slurry Eh, changed by adding Mn02, on the goethite to hematite
transformation rate (transformation rate decreases with increasing
Eh). The authors suggest at low Eh the reaction is catalysed by ~ e ~ + ,
566 B. I. WHITTINCTON AND D. MUlR

which facilitates electron transfer to the oxide lattice and promotes


breakdown of the oxide lattice by dilute acid. Under these conditions,
an increase in acid concentration increases com~etitionbetween the
~ e and ~ protons
+ for adsorption on the solid and hence reduces the
reaction rate. At higher Eh the reaction rate is governed by acid at-
tack and the rate increases with increasing acid concentration. If the
reaction is dissolution controlled (e.g. as reported by Papangelakis
ei at., 1996) Tindall and Muir suggest transformation of any inter-
mediate sulphate occurs at a higher rate than it is generated. Under
this condition, the intermediate will not be observed in the reaction
Downloaded by [Aston University] at 02:54 21 January 2014

solids. However, "if the reaction is Eh controlled, intermediates will


occur when no acid is available, as reported by Chou ei at. (1977),
and the transient sulphate will convert to hematite as the reaction
continues and acid is generated". Chou et at. (1977) also report a
leaching mechanism involving formation of hydrolysis products
with a high sulphur content which subsequently hydrolyse to products
with a lower sulphur content (e.g. hematite) (Section 4.2.4).
Results by Sobol (1968) indicate the beneficial effect of sulphur
addition on the nickel and cobalt leach kinetics (Moa Bay ore digested
at 230-240°C in a laboratory autoclave, acidlore 19.5 or 22.5%).
The majority of nickel and cobalt in the Moa Bay ore is present in
asbolan, maghemite and goethite.
Addition of elemental sulphur during leaching also minimises the
formation of and increases the c r 3 + concentration of the leach
liquor. As a result, this minimises both the ~i0:-and ~ i ' +/ c o 2 +
incorporation into the alunite (Sobol, 1969; see Section 5.4.1). The
latter improves the overall nickel and cobalt extractions. However
the decrease the silica-containing solids present in the post-digestion
slurry, and the decrease in free acid concentration in going from the
leaching to CCD, both result in formation of silicic acid gel in the
CCD tanks with consequent settling problems (Sobol, 1969).
Papangelakis el al. (1996) study the 250°C leaching of goethite pre-
sent in limonite ore using transmission electron microscopy (TEM).
Their results indicate the fine-grained goethite needles dissolve con-
tinuously during leaching and that the hematite forms by ex-siru
precipitation in solution and in the goethite cavities. The hematite
forms rapidly and does not show much additional growth during
leaching. The slurry loading affects the size distribution of the hematite
ACID LEACHING OF NICKEL LATERITES 567

particles formed. At I % solids loading the spherical hematite particles


are of 0.1 to 0.25pm diameter. Increasing the slurry loading to 22%
solids results in a less uniform particle size. Fine particles are also
found which probably result from secondary nucleation caused by
particle contact and attrition (Papangelakis et al., 1996). The presence
of fine particles would be expected to affect the settling properties of
the post-digestion slurry.
Georgiou and Papangelakis (1998) report no change in the mor-
phology of a n Indonesian limonite ore upon heating to 270°C prior to
acid addition (30% solids loading). Upon acid addition, the needle-
Downloaded by [Aston University] at 02:54 21 January 2014

like goethite aggregates dissolve continuously and hematite precipi-


tates. These hematite particles d o not grow observably during leaching
(average particle size 0.1 to 0.3 pm) and, consistent with results from
Papangelakis et al. (1996), are not uniform in shape and size. The
alunite particles observed after 30 minutes leaching a t 270°C are of a
larger size than the hematite particles. Additionally, the magnesium sili-
cate observed after 5 minutes leaching contains no nickel while, after
30 minutes leaching, all of the nickel and most of the cobalt are leached
from the manganese phases (Georgiou and Papangelakis, 1998).
Briceno and Osseo-Asare (1995) report an increase in the tempera-
ture a t which the limonite or limonite/garnierite ores are digested in-
creases the mean particle size of the resultant leach residue (from = 1.5
to -4.5 pm). In particular, there is a pronounced increase in mean
particle size for temperatures above 175°C. Briceno and Osseo-Asare
(1995) suggest this is the temperature a t which aluminium starts to
hydrolyse and the increase in particle size may result from deposition
of the hydrolysis products onto the particle surfaces. Other researchers
(e.g., Papangelakis et al., 1996) report minimal growth in the hematite
particles with digestion time, suggesting growth of other species is
responsible for this increase in particle size. Interestingly, Georgiou
and Papangelakis (1998) report alunite, formed upon digestion of a
limonite ore, has a particle size larger than that of the hematite.
A number of authors examine the sulphuric acid leaching of
synthetic goethites. The results from these studies indicate caution is
needed in applying results obtained with synthetic goethites to natural
ore samples. For example, Schellmann (1978) reports a synthetic nic-
kel goethite displays different leaching characteristics than a natural
sample, since half of the nickel is readily leached in dilute acid while
568 B. I . WHI'ITINGTON AND D. MUlR

the other half - the nickel incorporated in the goethite lattice - is


leached only in stronger acid. Tindall and Muir (1996) also report a
rapid initial release of nickel from an amorphous phase occurs upon
leaching a synthetic goethite with sulphuric acid a t 250°C. However,
a good correlation is observed between the transformation of the
synthetic nickel-goethite and the nickel extraction after this initial
nickel release, suggesting this nickel is dispersed throughout the
goethite. Negligible transformation of goethite occurs a t 220°C.
Kumar et al. (1990) report the leaching behaviour of synthetic
goethites in sulphuric acid at 50°C depends on the mode of goethite
Downloaded by [Aston University] at 02:54 21 January 2014

preparation (coprecipitation o r sorption of the metal ions onto a


goethite) and the metal ion (Co, Ni or Cu). In particular, for goethites
prepared by coprecipitation, Ni leaches more rapidly than C o which,
in turn, leaches more rapidly than Cu. For those goethites prepared
by sorption, much higher extractions are observed and the extraction
follows the trend Co > Cu > Ni. Such a result indicates the location
of the metal ions in the synthetic goethite depends strongly on the
incorporation conditions.
Das er al. (1997), in their study on the leaching of chromite
overburden a t 260°C and 36.8g/L H2S04, report the presence of
goethite in the leach residues after 3 hours reaction - despite approxi-
mately 98% of the nickel being extracted. This suggests the presence
of a pure, crystalline goethite and a more reactive nickel-containing
goethite. Interestingly, Sen et al. (1984) indicate goethite present
within a laterite ore contains varying concentrations of nickel (i.e.
nickel rich and nickel poor goethites are present).
Georgiou er a/. (1994) examine the phases forming during acid
leaching of a goethite-containing laterite a t initial pulp density I % ,
initial sulphuric acid concentration 15 g/L, 250°C digestionJ60 minutes
and 400 rpm agitation. SEM examination of the particles after 4 minu-
tes digestion indicates the appearance of "flower-like" structures.
X R D indicates only goethite and hematite are present a t this time.
Basic ferric sulphate is not observed by XRD.
In contrast to the previous studies on goethite dissolution, there are
significantly fewer studies examining the dissolution of clays. Tindall
and Muir (1997) report a digestion temperature of 235°C is sufficient
to decompose nontronite and extract most of the nickel while
Motteram et al. (1997) report high nickel extractions upon digesting
ACID LEACHING OF NICKEL LATERITES 569

Murrin Murrin nontronite ore a t 250 or 260°C. Indeed, Tindall and


Muir (1997) report nontronite decomposes within 45 minutes of
leaching a t 150°C while decomposition of goethite requires tempera-
tures of a t least 220°C. X R D and electron microprobe analysis of
residues from pressure acid leaching Bulong nontronite ore indicates
the presence of hematite, natrojarosite, alunite, amorphous silica and
meta-alugen [A12(S04)3.H 2 0 ] (Hayward, 1998).
Briceno and Osseo-Asare (1995), using X R D analysis, report the
talc present in a limonite/garnierite ore completely reacts within 15
minutes pressure acid leaching at 225°C. By contrast, some goethite
Downloaded by [Aston University] at 02:54 21 January 2014

still remains after 60 minutes a t this temperature (Briceno and Osseo-


Asare, 1995).

5.4. Aluminium Chemistry


We have previously mentioned formation of aluminium-sulphur
containing alunite scale contributes most to acid consumption at the
Moa Bay processing plant (Section 4.2.2). The idealised equations for
dissolution of aluminium and subsequent precipitation of alunite scale
are presented in Eqs. (15a-c). Blakey and Papangelakis (1993) report
any gibbsite present in the ore reacts to boehmite upon digestion and
it is this species which subsequently reacts with sulphuric acid. In the
presence of cations, such as sodium derived from the borefield water,
natroalunite precipitates (Eq. (15d)). Hayward (1998) suggests the
additional leach reactions (1 5e- 15g) occur during digestion of Bulong
ore (Eq. (15e) differs from (l5d) in that more explicit account is made
of the fate of any chloride from saline waters).

AI(OH)3 + AlOOH + H 2 0 & AlOOH + 3 HC + A?+ + 2 H z 0


(1 5 4
570 B. I. WHllTlNGTON A N D D. MUlR
Downloaded by [Aston University] at 02:54 21 January 2014

Dissolution of aluminium may result in a substantial amount of the


total aluminium being present in solution prior to precipitation (e.g.
results in Figs. 4 and 7, results in Krause et al., 1998, and leaching of
Moa Bay ore in Tab. V, indicate approximately 46%, 52-69%,
6-35% and 15% of the total aluminium may be present in solution).
Precipitation of alunite or basic aluminium sulphates regenerates acid
(Eqs. (15b,c)) (Taylor, 1997a).

Digestion time (minutes)


FIGURE 7 Liquor concentration of dissolved aluminium during preparation of hydro-
nioalunite. Initial sulphuric acid concentration approximately 49g/kg, initial alumi-
nium charge approximately 8 g Allkg (from Blakey and Papdngelakis, 1993).
ACID LEACHING OF NICKEL LATERITES 57 1

The composition of the basic aluminium sulphates, and the conse-


quent acid consumption, depend on the digestion temperature. The
hydronioalunite, which forms at temperatures below 250°C (Eq.
(ISb)), results in a greater acid regeneration/lower acid consumption
than the AI(OH)S04 phase which forms at temperatures above 280°C
(Eq. (1 Sc)). Chou et a[. (1977) also suggest formation of AI(OH)S04
in their reaction system. Formation of AI(OH)S04 can also occur at
digestion temperatures lower than 280°C if the liquor is highly acid
(Georgiou and Papangelakis, 1998 and references therein). Queneau
and Weir (1986) report "efficient formation of alunite by thermal
Downloaded by [Aston University] at 02:54 21 January 2014

hydrolysis, under conditions that provide sufficient free acid to main-


tain nickel and cobalt solubilities, requires temperatures signifi-
cantly above 200°C, e.g. 240°C". It is the temperature of this
requirement that influences the digestion temperature, since hydrolysis
of the iron to hematite reportedly occurs at 200°C.
Blakey and Papangelakis (1993) prepared hydronioalunite from
gibbsite and sulphuric acid. Boehmite is present in the solids forming
in 24.58 H2S04/kg sulphuric acid - but not 49 or 73.5 g H2S04/kg -
at 230 or 270°C (Blakey and Papangelakis, 1993). The incomplete
reaction to hydronioalunite in 24.5g H2S04/kg sulphuric acid is
suggested to arise from the presence of HSO; versus H + and SO:- at
these temperatures (Section 5.1). SEM examination of this hydro-
nioalunite product indicates an iron-rich, aluminium-poor core and
suggests the hematite added for corrosion protection of the titanium
autoclave may act as a seed for alunite crystallisation (Blakey and
Papangelakis, 1993). By contrast, any unreacted boehmite remains
as separate crystals and therefore does not act as a seed for alunite
precipitation (Blakey and Papangelakis, 1996). The difference in
size distributions of the feed and product additionally indicates a
dissolution/precipitation mechanism (Blakey and Papangelakis, 1996).
The initial acid concentration can also influence the pH profile dur-
ing the course of a batch alunite preparation reaction. At an initial
acid concentration of 24.5 g HzSOJkg the pH increases for the first 2
hours of reaction, suggesting dissolution of boehmite (an acid consum-
ing reaction) as the rate determining step. However at the higher ini-
tial concentration of49 g H2S0Jkg the pH decreases during the first two
hours of reaction, suggesting precipitation of alunite (an acid generat-
ing reaction) controls the reaction rate (Papangelakis er al., 1994b).
572 B. I. WHITTINCTON A N D D. MUlR

Papangelakis et a/. (1994b) report the hydronioalunite equilibrium


aluminium concentration increases with increasing acidity (decreasing
pH) o r decreasing temperature (230-270°C) (Fig. 8). The presence of
added salts (e.g. seawater, Fig. 4, Section 4.2.5) can also significantly
reduce the concentration of aluminium in the process liquor. This
suggests natroalunite has a lower solubility than the hydronioalunite.
Georgiou and Papangelakis (1998) examine the pressure acid
leaching of an Indonesian limonite ore (230 to 270°C; 30% solids
charge and 0.2 acidlore ratio). The aluminium concentration profiles
in Figure 9 are significantly influenced by the digestion temperature
Downloaded by [Aston University] at 02:54 21 January 2014

and differ in some aspects from those in Figure 7. The results in Fig-
ure 9 suggest a n increase in boehmite dissolution rate with increasing
temperature (gibbsite reacts initially to boehmite under these condi-
tions). Precipitation of aluminium from solution does not occur as
rapidly as precipitation of iron a t 230 or 250°C.
Blakey and Papangelakis (1996) developed a thermodynamic model
to calculate the solubility of hydronioalunite, and the solution speci-
ation, under a range of conditions. Of particular interest is the pre-
dicted abundance of aluminium-hydroxyl complexes and the absence

Liquor pH (measured at 25°C)


FIGURE 8 Equilibrium aluminium concentration as a function of temperature and
solution pH (pH measured at 25'C) (from Papangelakis er a/., 1994b).
ACID LEACHING O F NICKEL LATERITES
Downloaded by [Aston University] at 02:54 21 January 2014

Digestion time (minutes)


FIGURE 9 Liquor composition during the leaching of an Indonesian limonite ore, at
230, 250 or 270°C, 30% w/w ore loading and 0.2 acidlore addition. Liquor concen-
trations are those obtained at room temperature after the slurry was filtered at tem-
perature. Original charges of approximately 200g/L Fe and 8g/L Al (from results
presented by Georgiou and Papangelakis, 1998).

of aluminium-sulphate complexes. At low sulphate concentrations, the


dihydroxyl cation is the predominant aluminium species present.
However, at higher sulphate concentrations the A I ~ ~ O ~ ( O H )species
:;
reportedly predominates (Blakey and Papangelakis, 1996).
Blakey and Papangelakis (1996) additionally indicate the usual ex-
perimental method of measuring solution pH (i.e. filtering the slurry
at temperature, cooling the liquor to room temperature then meas-
uring the pH) may not give an accurate estimation of the liquor
acidity at temperature.
Baghalha and Papangelakis (1998a) extend upon the binary model
of Blakey and Papangelakis (1996) and predict the aluminium
solubility at 250°C in H2SO4 solutions containing MgS04 and
MgS04-like salts ("pseudo-MgS04"). The model predicts a sharp
drop in free proton ( H f ) molality at 250°C with increasing con-
centration of pseudo-MgS04 due to the higher levels of sulphate
ion combining to form HSO;. The industrial implication is that ores
containing high concentrations of Mg (e.g. saprolite) show higher pH
values at the same free acid concentration to that from leaching other
574 B. I . WHI'ITINGTON AND D. M U l R

ores (e.g. limonite) which could lower the equilibrium solubility of


AI(I11) and Fe(l1l).
Baghalha and Papangelakis (1998b) model the AI/H2S04 and
Mg/H2S04 systems using the ion-association-interaction approach.
They make no assumptions regarding the Al or Mg species present
in solution, but calculate the equilibrium solubilities using any
possible species so as to obtain the best fit with the high-tempera-
ture, experimental, equilibrium solubilities. They report MgS04 as
completely dissociated in solution in the Mg/H2S04 system a t
235, 250 or 270°C. A I ~ ( S O ~ )is; reported as the main aluminium
Downloaded by [Aston University] at 02:54 21 January 2014

species in solution in the A I / H ~ S Osystem


~ a t 250°C. This revises
the A I ~ ~ O ~ ( O H ) : ;species previously reported by Blakey and
Papangelakis (1996) as the major aluminium species in solution
at 230. 250 o r 270°C.

5.4.1. Substitution in Alunite

Sobol (1967, 1968) reports hydronioalunite, forming during the Moa


Bay leaching, can incorporate Ni, Co o r Mn, possibly as (Ni, Co,
Mn)A16(S04)4(OH)12. Sobol (1967, 1969) reports incorporation of
these metal ions in alunite correlates with a lower equilibrium alumi-
nium concentration in the pregnant liquor at 230-285°C and that the
alunite metal incorporation generally increases with increasing tem-
perature. Additionally, the metal ion content in Moa Bay leach tailings
increases with increasing metal ion liquor concentration, suggesting
greater incorporation into the alunite. The free acid concentration
also influences incorporation into alunite. In particular, increasing the
free acid concentration decreases the alunite nickel concentration but
increases the alunite cobalt concentration (Sobol, 1969).
These previous results, other work conducted by Sobol (1969) and
the known chemistry of the alunite mineral family (Das e / al., 1996)
all suggest substitution of various ions into the alunite structure can
occur according to Figure 10 and as outlined below:
(i) c r 3 + or ~ e ' +may substitute for A13+,
(ii) GO:-, but not s~o:-, may substitute for SO:- (the geo-
metry and energetic reasons make incorporation of at
this site unfavourable),
ACID LEACHING OF NICKEL LATERITES 575

(iii) K + , N a + , NH:, Ni2+, c o Z + , ~ e ~ Mn2+


+ , may substitute for
hydronium ions,
(iv) SO:- may substitute for OH-.
Sobol (1969) calculates alunite forming at Moa Bay has the molar
composition 0.01 Ni2+, 0.001 c o 2 + , 0.03 Mn2+, 0.95 H ~ O + 2.87
,
0.13 ~ e ~ 1.79
+ ,SO:-, 0.21 c~o:-, 0.44 and 5.60 OH-.
However, the composition varies with the alunite formation condi-
tions. In particular the high rate of alunite formation in Reactor A,
reportedly resulting from the high liquor supersaturation, disfavours
Downloaded by [Aston University] at 02:54 21 January 2014

incorporation of other ions into the alunite structure (alunite scale

H30+

H,O'

t
SO4- Fe3+ c1.0,~- Ni2'
Ti044- CP' Cd'
Mn2+,Fez+
Mg2'
K+
NH,'

FIGURE 10 Schematic of an alunite molecule indicating possible locations for sub-


stitution by other ions. From Sobol(1969).
576 8. I. WHITTINCTON AND D. M U l R

from Reactor A reportedly contains less impurities than alunite scale


from other Moa Bay reactors).
Sob01 (1969) reports a linear relationship between ~ r 0 : - and ~i0:-
incorporation into alunite obtained from the Moa Bay processing
plant, the results suggesting one ~ r 0 : - is incorporated into alunite
for every two s~o:-. Up to 30% of the SO:- and 30% of the OH-
may be substituted by ~ r 0 : - and respectively while up to 26%
of the A I ~ +
may be substituted by ~ e (Sobol,~ + 1969).
Caution is required if the technique used by Sobol is used to
calculate the composition of alunite from other leach residues. In
Downloaded by [Aston University] at 02:54 21 January 2014

particular, since silica in the Moa Bay leach residues is generally


present only in alunite (Sobol, 1969), the calculation of alunite Si
content is easier than if Si were also present in other phases. In these
instances, calculation of an accurate alunite Si content requires a full
accounting of the silica mineralogy (e.g. Perdikis, 1996 reports scale
present on the titanium laboratory reactor internals, after an ex-
perimental programme, contains hematite (50-70% total), alunite
(15-30%), boehmite and quartz (2-4%). Caution is also required
when calculating the concentration of other elements reportedly in-
corporated into alunite.
Czerny and Whittington (1999) use powder X-ray diffraction and
scanning electron microscopyjenergy dispersive spectroscopy (SEMI
EDS) to fully characterise the jarosite/alunite scales prepared during
the pilot plant leaching of laterite ores (results presented in'section
5.8). Despite the high clay content of the ore, minimal Si was present
in the scales from the W.A. laterite - similar to scale from tropical
laterite.

5.5. Magnesium Chemistry


Magnesium generally occurs as a variety of minerals in laterites.
Typically, 50-60% of the magnesium dissolves in the pressure acid
leach to form magnesium sulphate although this figure is variable. For
example, magnesium extractions of 44%, 94% and 90% are reported
for leaching an Indonesian ore (Krause er al., 1998; presented here in
Fig. 4), Murrin Murrin ore (Motteram er al., 1996) or Moa Bay ore
(Sobol, 1967; and presented in Tab. V ) respectively. Steemson (1996)
reports not all of the magnesium present in magnesium silicates is
ACID LEACHING OF NICKEL LATERITES 577

acid leachable, although this is not necessarily bad since magnesium


extractions in excess of 70% reportedly "break the backbone of the
magnesium silicate molecule" and deluge the circuit with soluble
silicon (Queneau and Berthold, 1986 and references therdin). Addi-
tionally, since magnesium sulphate does not hydrolyse further, it
quantitatively consumes acid (Taylor, 1997a).
Dissolution of clays within laterites could yield soluble silica ac-
cording to Eq. (16) (Taylor, 1997a; Hayward, 1998).
Downloaded by [Aston University] at 02:54 21 January 2014

An Eh/pH diagram, indicating the stability fields for various Mg-


S - H 2 0 salts at 250°C, is presented by Da Silva (1992). M ~ is~ +
soluble at 250°C for pH < 2 while M g S 0 4 . H 2 0 precipitates at
2 < pH < 4 (Da Silva, 1992). MgS04.6H20 is the stable phase for
temperatures less than 60-SOT, with the monohydrate being stable
at higher temperatures (Queneau et al., 1984 and references therein).

5.6. Chromium Chemistry


Chromium, present in the leach liquors as Cr(III), is oxidised to cr0:-
(Eqs. (9b), (lob), Section 5.2), which may then be incorporated into
alunite (Section 5.4.1).
Sobol (1969) reports one is incorporated into alunite for
every two (Section 5.4.1), suggesting cleaning the liquor of silica
requires sufficient cr0:- in the liquor. Addition of elemental sulphur
reduces the liquor ~ r 0 : - concentration (Eq. (17); Sobol, 1968) and
would affect incorporation of silica into the alunite. Addition of sul-
phur to the Moa Bay leach liquor is observed to result in formation
of silica gel in the CCD tanks (Sobol, 1969 and previously described
in Section 4.2.7).

CrOi- has previously been shown to retard the polymerisation of


silicic acid although the conditions (1.OM SO2, 2S°C, pH 0.5-2.5,
C r : Si molar ratios 0- 1.5; Iler, 1952) are different from those in the
processing plant CCD tanks. Up to two moles Si02 reportedly
combine with one mole CrOj- at low chromium concentrations
578 B. I. WHITTINGTON AND D. MUIR

(Iler, 1952) and so the retarding effect of CrOf could also explain
the observed gelation of silicic acid upon addition of sulphur (Section
4.2.7). Calculations using published Moa Bay ore and leach residue
compositions (Appendix A) suggest formation of Si04-containing
alunite, or other silica-containing solids, reduces the silica in solution
at the Moa Bay plant to 60% of the value expected from the silica
concentration in the original ore. Further work is required to deter-
mine if halting formation of silica-containing alunite accounts for
the observed silica gelation.
Sobol (1969) also reports a linear relationship between the CrOi-
Downloaded by [Aston University] at 02:54 21 January 2014

and Ni(I1) or Mn(I1) incorporation into alunite from the Moa Bay
processing plant, with the alunite Ni(1l) or Mn(I1) contents increas-
ing with increasing CrOi- incorporation. Thus, we would additionally
expect addition of sulphur during digestion to minimise the alunite
nickel content and hence improve the overall nickel extraction.
Sobol (1969) reports the various chromium-containing phases ob-
served during high temperature acid leaching in the Cr203-SO4-
H 2 0 , Cr203- Fe203-SO4- H 2 0 , Cr203-A1203 - SO4- H 2 0 and
Cr203-AIZO3- Fe203- Sod-H20 systems.

5.7. Silicon Chemistry


Soluble silica (Si(OH)4(,,,) can result from dissolution of quartz or
amorphous silica or from dissolution of magnesiuin silicates (Queneau
et al., 1983b) (Eq. (16)).
The equilibrium solubility of monosilicic acid in water increases
with temperature (0-300°C) but is relatively insensitive to the pH
when in acid solution (Queneau et al., 1983b). A knowledge of the
equilibrium solubility is important, since it determines the super-
saturation '3-1" at a given liquor concentration (S-1 = ( C - CJC,,
where C is the solute concentration and C, is the equilibrium or
saturated solubility). Supersaturation is one of the factors influencing
the rates of precipitation and/or scaling (Section 5.8).
Whether the monosilicic acid reacts to a gel or a precipitate depends
on the silica supersaturation, the availability of silica surface upon
which fresh silica may be deposited (i.e. seed particles), temperature,
pH and ionic strength. These factors also control the rate of gela-
tion (Queneau and Berthold, 1986). The rate of polymerisation of
ACID LEACHING OF NICKEL LATERITES 579

monosilicic acid increases with increasing supersaturation, tempera-


ture and addition of active seed. The induction period which is often
present prior to the onset of polymerisation can be eliminated by
seeding, use of high liquor supersaturation and/or high temperature
(Queneau and Berthold, 1986).

5.7.1. Silica Scaling


Queneau er al. (1983b) summarise the mechanism of silica scale
formation. "Rapid silica scaling reportedly occurs in sufficiently
Downloaded by [Aston University] at 02:54 21 January 2014

supersaturated solutions by polymerisation of soluble silica to col-


loidal particles (50-100A in size) as the stream cools from 200 to
140°C. In sufficient concentration, these particles aggregate into open
clusters of microgel which collide with walls and, once in contact,
become cemented to the wall (or the surface of the scale) by further
deposition of silica from solution around the points of contact. Thus,
the rate of increase of deposit is much faster than the rate at which
the silica molecules come out of solution. Hematite does not provide a
suitable surface for silica deposition at this low pH - as a general rule,
silicic acid is not adsorbed on the surface of an oxide if the pH is much
lower than that at which the metal hydroxide dissolves in acid".
Testwork for the Amax pressure leach process indicates silica
scaling is not a problem during single-stage flashing of the autoclave
leach liquor from 240-280°C, but is a problem when three-stage
flashing is utilised (Queneau et al., 1983b). During the first stage of this
three-stage flashing (to 200°C), the low liquor supersaturation, low
pH and the short residence time result in minimal deposition of silica.
However the liquor supersaturation during the second stage of let-
down (to 140°C) results in rapid nucleation of amorphous silica as
colloidal particles (Queneau et al., 1983b).
Iler el al. (1983) report it is possible to minimise such scaling by
recycling the slurry between adjacent flash pots. This allows dissolved
silica to precipitate onto the solids present in the slurry rather than
onto the flash-pot walls. Control of pH can also influence the degree
of scaling since the rate of silica scaling is maximum at the maximum
silica polymerisation rate. Thus, one effective method for minimising
silica scaling is to change the liquor pH to 1.7, where the rate of
nucleation and deposition are at a minimum (Queneau et al., 1983b).
580 6. I. WHITTINGTON AND D. MUlR

5.8. Scaling a n d Associated Chemistry


It is generally recognised that increasing the liquor supersaturation
increases the rate of precipitation from a solution and hence the rate
of scaling. Increased supersaturation usually (but not always) occurs
with a decrease in temperature (e.g. at the autoclave walls). Increas-
ing the kinetics of the precipitation or subsequent growth, which oc-
cur with increasing the temperature, also increase the rate of scaling.
The following summary of an article by Roach and Cornell (1985)
indicates the different mechanisms for scale formation in the Bayer
Downloaded by [Aston University] at 02:54 21 January 2014

process. Despite the significant differences in the process chemistry


relative to the nickel laterite acid leaching, these comments are rele-
vant to the high temperature acid leaching.
Scale may occur as growth scales or settled scales. Growth scales
form by nucleation of the supersaturated phase on a surface and it is
the subsequent growth on these nuclei which forms a highly crystalline
material. The two main factors influencing the rate of nucleation are
the degree of supersaturation (previously discussed) and the surface
(formation of nuclei is easier on surfaces of similar crystal structure or
chemistry to the supersaturated phase).
Settled scales occur when slurry particles, which have settled in
regions of low fluid velocity, are cemented by the supersaturated liq-
uor. These scales are frequently porous and soft, reflecting their more
rapid formation relative to growth scales (up to metreslweek versus
mm/week respectively in alumina refineries; Roach and Cornell, 1985).
Sobol(1967) calculates that approximately 30% of the original Moa
Bay ore subsequently precipitates during the high pressure leaching.
Of this, Sobol (1967) reports 99.9% deposits as a thin film on the
surface of the ore particles which are being leached while only 0.1%
deposits on the reactor walls as scale. This contrasts with work by
Georgiou and Papangelakis (1998) who report "goethite particles dis-
solve continuously while iron reprecipitates as dense hematite parti-
cles in solution by ex-siru precipitation".
Scaling is a problem in the Moa Bay acid-leach autoclaves, where
up to 20cm scale builds up in the digestion tanks and may result in
5 days descale downtime per month. Various researchers have charac-
terised Moa Bay scale. Chaves er al. (1968) report a significant alumi-
nium concentration in the scale. Mineralogical analysis indicates the
ACID LEACHING OF NICKEL LATERITES 58 1

presence of hematite, boehmite, aluminochromite [Fe(Cr,Al),04],


basic iron sulphate, hydronioalunite and hydroniojarosite (Chaves
et al., 1968). Plotting the concentration of phases present in the scale
on a phase diagram indicates the scale composition tends to fall in
certain regions: (i) hematite - hydronioalunite: the ratio of these two
phases depends on which of the four autoclaves is examined and varies
from 1:3 in the first autoclave to 1:1.8 in the second autoclave. "The
largest amounts of hydronioalunite form when the pulp mixes initial-
ly with sulphuric acid or where the reagents are mixed with particular
intensity. This indicates a high rate of reaction for gibbsite to hydro-
Downloaded by [Aston University] at 02:54 21 January 2014

nioalunite" (Chaves et al., 1968). (ii) basic iron sulphate - hydronioalu-


nite. Pure basic iron sulphate is found in the first autoclave at the
point of sulphuric acid addition (i.e. on the acid addition pipe). (iii)
hematite-boehmite. These scales are found where the acid concen-
tration is lowest, generally in the third and fourth autoclaves. (iv)
hematite-hydroniojarosite mixtures are found in scale forming in
the first autoclave opposite the acid inlet (Chaves et al., 1968).
At the point of acid injection, large yellow balls of (Fe0),S04 and
Fe(OH)S04 also form (Queneau and Weir, 1986 and references
therein). Silica also precipitates with the iron as (FeS04)4.Si04 near
the acid addition point, since the high acidity results in high con-
centrations of soluble iron and the dissolution/re-precipitation of
reactive silica minerals (Eq. (18)) (Sobol, 1969).

"Further downstream from the acid injection point, first the


(Fe0),S04 and then the Fe(OH)S04 in the scale gradually gave way
to Fe203, the stable ferric oxide in a low acid, 240°C environment.
Aluminium in the scale increases from 0.5% at the point of acid
injection to 18% further downstream" (Queneau and Weir, 1986 and
references therein).
Queneau et al. (1984) report the presence of magnesium sulphate
scale, of the formula Mg(Ni)S04.6 H 2 0 and Mg(Ni)S04. H 2 0 , in the
autoclave compartments of a high-pressure acid leach vessel used to
pilot the Amax process. The Mg:Ni ratio in these scales ranges
from I : 1 to 3: 1 and reflects the concentration in the liquor from
which the scale deposited. However, the scale mineralogy varies with
582 El. I. WHITTINGTON AND D. MUIR

ore-type - intergrown crystals of alunite and hematite form when


processing ores with low magnesium contents while MgS04. H 2 0
and highly dispersed particles of hematite form when processing
magnesium-rich ores (Queneau et al., 1984). Scale taken from a lev-
el control pot, located immediately after the autoclave, comprises
alunite-hematite and MgS04. H 2 0 .
Results from Queneau et al. (1984) indicate precipitation of mag-
nesium sulphate scale occurs rapidly since scale is observed in those
autoclave compartments where the magnesium-rich ore is added,
but not in other compartments. Precipitation of magnesium sulphate
Downloaded by [Aston University] at 02:54 21 January 2014

scale is highest where the acid concentration is low (magnesium


solubility increases with acid concentration, see Fig. 14).
Krause et al. (1998) report scale forming during continuous leaching
of limonite ore consists of a thin, iron-rich layer near the autoclave
surface, with subsequent deposition of hydronioalunite. Hematite fills
the voids between the alunite dendrites (Krause et al., 1998).
Researchers have examined the formation of scale under conditions
of relevance to the Australian processing plants. Pilot testing indicates
scaling will be a problem using Murrin Murrin conditions (ca. 100mm
scale/year, Kyle, 1996). Motteram er al. (1996) report the autoclave
used for the continuous Murrin Murrin leaching testwork contains a
water insoluble scale corresponding to approximately 0.25 wt% of
the total ore treated. XRD analysis indicates the presence of alunite,
hematite and basic ferric sulphate while the A1 : Fe ratio suggests pre-
ferential deposition of alunite from the liquor relative to hematite.
Testwork conducted for the Bulong project indicates the majority of
scale forms in the first compartment of a four compartment pilot-scale
autoclave (i.e. where the acid is added; Taylor and Cairns, 1997).
Czerny and Whittington (1999) also examined those scales forming
during the pilot plant leaching of Western Australian (Bulong) ores
at 250°C using hypersaline process water. Natrojarosite/alunite (ap-
proximate formula Na(A11.04Fe1.96)(S0~)2(OH)6) formed in the first
autoclave with minor amounts of hematite and amorphous silica.
The thickness of scale forming on the turbine blades (x 170mrn/year)
was greater than that forming on the overflow pipe, and significantly
greater than that forming in the latter autoclaves. Silica was slow to
deposit and was mainly found in scales forming in the last autoclave
compartment.
ACID LEACHING OF NICKEL LATERITES 583

By contrast scaling is reportedly not a problem during testing for the


Cawse project, which probably reflects the lower aluminium content of
this ore (Kyle, 1996). Scaling is not expected to be a problem for the
Syerston project (which also uses a low aluminium-content ore) (Kyle,
1996) or for the Marlborough project (Griffen, 1998). The absence of
scale formation during tests with Marlborough ore is suggested to
result from the low sodium content of the ore and the process water
preventing formation of natrojarosite (Griffen, 1998). However, since
non-sodium containing phases (e.g. alunite, hematite) can also form
scale this is only a partial explanation. Mason and Hawker (1998)
Downloaded by [Aston University] at 02:54 21 January 2014

estimate scale formation in autoclaves processing Ramu ore at 80mm/


year. The scale forming in the "leading compartments" of a multi-
compartment autoclave, where the acid is added, consists of hematite,
aluminium sulphate hydrate and alunite. Czerny and Whittington
(1999) report formation of a scale containing hydronium alunite
(approximate formula (Na0.17 H@o.dAlz.s6 Feo.44)(S04)2(OH)d and
hematite, at a rate of =90mm/year, during leaching of a tropical
laterite at 250°C in potable process water.
Detailed examination of the scales from leaching of a Western
Australian (Bulong) ore at 250°C revealed the composition of the scale
forming in the first autoclave depended on the exact liquor com-
position at the time (Czerny and Whittington, 1999). Hence the
composition of the natrojarosite/alunite at the centre of the scale was
different to that freshly deposited at the final solid/liquid interface
(Fig. 11). Microscale variations also occurred in the natrojarosite com-
position, with evidence of banding (shown in Fig. l l b ) where the
molar Al : Fe ratio varies from 1.72 : 1.28 to 1.20 : 1.80. It appears that
Fe and Al crystallise into the alunite/jarosite lattice in turn as the
supersaturation of Fe and Al in solution builds up in cycles.
The scale from leaching a tropical laterite also contains composi-
tional variations. However, in this instance, they reflect changes in the
hematite and hydronium alunite concentrations rather than varia-
tions in the alunite composition (Fig. 12).
Papangelakis et ai. (1994b) quantify the scale forming during alunite
synthesis at pH > 0.7 by use of a titanium coupon suspended in the
reactor. The results indicate significant scale formation occurs at high
acidities (low pH values) and a digestion temperature of 250°C rather
than 240 or 270°C (Fig. 13a). Papangelakis et ai. (1994b) suggest the
Downloaded by [Aston University] at 02:54 21 January 2014
Downloaded by [Aston University] at 02:54 21 January 2014
B. I. WHITTINGTON AND D. MUIR

a)
25-
-m i250%
20-
.27@C
15-
3 .....&-;,.,::...PC::..24Bc
rx 10- .I ,,"
, ,
,' ,*
5- ,a 9,'
.mi
'=
Downloaded by [Aston University] at 02:54 21 January 2014

04
0.0 0.1 0.2 0.3 0.4
Final acidity (M)

Scale gmwth
20-
at 27vc -I:-#.-+
15-

10-

5-
Scale gmwt
at 2WC
7 *
:
..
.J
J
- :k-
..
".'B

0s
0.0 0.1 O.? 0.3 0.4
Final aad~ty(m0ldL)

FIGURE 13 Effect of liquor acidity on alunite scale production "R", the grams scale
deposited on titanium coupon/m2 coupon. Scale deposited (a) from a pure aluminium-
sulphate-water system after a four hour period (from Papangelakis el a/., 1994b),
(b) during three consecutive limonite batch leaching experiments at 250 or 270°C and
22% solids loading (from Krause et a/.,1998).

digestion time/600rpm agitation/0.86 final solution pH. The scale


appears to consist mainly of hematite with traces of hydronioalunite.
Approximately 2.7g/m2 of scale initially forms on the coupon.
Subsequent growth of scale occurs at a much faster rate - exposure
of this "scaled" coupon to a further two digestions results in
ACID LEACHING OF NICKEL LATERITES 587

17.8gscale/m2. This result suggests a predominance of surface nu-


cleation on the covered coupon. By contrast, the increase in scale
thickness during these subsequent exposures is less than the original
deposition, suggesting subsequent growth fills the voids present in the
original scale (Georgiou et ai., 1994).
The importance of secondary nucleation on the formation of iron-
containing solids is also illustrated by results from Perdikis (1996),
who reports a reduction in scaling (from 18.5 to 5.sg/m2 titanium
coupon, 270"C, final acidity 0.23 M) with a reduction in solids loading
from 22 to 11%. The smaller number of solid-solid collisions in the
Downloaded by [Aston University] at 02:54 21 January 2014

autoclave, at the lower solids concentration, reportedly reduces the


number of new nuclei generated.

5.8.1. Methods to Minimise Scaling


The following summarises methods used in Bayer refineries to alleviate
scaling which are of interest to the nickel laterite acid leaching plants
(from Roach and Cornell, 1985).
Incomplete cleaning may leave sufficient scale to act as nuclei for
further growth. Therefore chemical cleaning, which removes potential
growth sites, is usually more effective than manual cleaning. The sur-
face smoothness also influences scale growth and electropolished tubes
or various plastic coatings are effective at minimising scaling. Addition
of abrasive particles to the liquor has also been used although care
is obviously required.
Formation of growth scales can be stopped, or their form altered
(e.g. to a less massive scale growth), by poisoning the growth surface
with chemical additives. In contrast, the most effective way to prevent
settled scale formation is to ensure sufficient agitation.
Chalkley er oi. (1996) claim improved acid-digestion reactor agi-
tation reduces the rate of scale formation. However, attempts to
minimise scale formation in the Moa Bay processing plant, by in-
creasing the diameter of the steam injection line and increasing the
quantity of steam used for agitation, were unsuccessful (Chalkley and
Toirac, 1997). Sobol (1967) also reports poor mixing of ore slurry and
acid at the acid injection point favours scale formation. Experimen-
tal tests indicate no scale formation in a "well mixed" autoclave and
scaling in an autoclave with no agitation (Sobol, 1967).
588 B. I. WHITTINGTON AND D.MUIR

Addition of sulphuric acid to the digestor in stages ("staged acid


addition") minimises scale formation during the leach by avoiding
localised high acid concentrations, but could adversely affect the metal
extraction kinetics (Hazen and Chou, 1997). In contrast, Fekete et al.
(1978) report staged acid addition has no adverse effect on the nic-
kel extractions, although Krause (1996) indicates increased retention
times may be required. Fekete et 01. (1978) and Duyvesteyn et 01.
(1979) report addition of sulphuric acid to the various compartments
of the pressure acid leaching autoclave, and vigorous agitation, also
minimise scale formation.
Downloaded by [Aston University] at 02:54 21 January 2014

Removal of scales can be problematic. For example, sulphuric acid


only partially dissolves the iron and aluminium sulphate scales, while
NaOH solution a t 130 o r 190°C is reportedly more effective (Queneau
et al., 1983a). Lussiez and Jha (1983) patented a method where the
autoclave is flushed with 20-l00g/L H2SO4 a t 150-250°C directly
after the leaching regime. Krause et 01. (1998 and references therein)
report flushing with 50g/L H2SO4 containing Cr(V1) or Fe(II1) cor-
rosion inhibitors a t 200°C. Another option is to shut down the auto-
clave system and physically remove the internal scale (e.g. Moa Bay)
although we have previously mentioned scale remaining after incom-
plete cleaning acts as nucleation sites for further growth.
Krause (1996) reports a digestion flowsheet in which a primary
autoclave - containing acid and slurry inlet ports - feeds into a
second multi-compartment autoclave. Since the majority of scale
forms in the primary autoclave, in which the acid and slurry are
combined, having a "spare" primary autoclave allows descaling of the
original primary autoclave with minimal disruption to production
(there should be minimal scale formation in the multi-compartment
autoclave).
MgS04 is more soluble in water than the "refractory" iron and
aluminium compounds and any method which promotes its formation
instead of the iron and aluminium scales should simplify removal
of the resultant scale. Queneau et al. (1983a) report preferential
formation of MgS04 scale is achieved using a two stage (Amax)
addition process, in which ore and acid are added to the initial com-
partment(~)of the high pressure acid leach autoclave while residue
from the atmospheric acid leaching of a n Mg rich ore is added to
the latter pressure acid leach autoclave compartment(s). The MgS04
ACID LEACHING OF NICKEL LATERITES 589

scale, which forms in those compartments where the ore/atmospheric


leach residue are added, "inhibits" formation of aluminium-con-
taining scales and minimises formation of hematite scale. Cooling the
autoclave from 270 to 180°C significantly increases the magnesium
solubility (Fig. 14) and liberates the hematite scale, which is bound
with the magnesium sulphate (Queneau et al., 1984; Siedel, 1965;
Marshall and Slusher, 1965).
Adsorption of an anionic surfactant onto the reactor surface, which
is positively-charged in the acidic environments of interest to the acid-
leaching process, has been examined by Papangelakis et al. (1994b) as
Downloaded by [Aston University] at 02:54 21 January 2014

a method of minimising scale formation. Initial results (Papangelakis


et al., 1994b) indicate addition of the surfactant - even at a concen-
tration of 0.004g/L - successfully minimises formation of alunite
scale from relatively dilute solutions. More recently, Krause et al.
(1998) extended upon this work to examine the effect of anionic sur-
factants on scale forming during a 270°C batch leaching of limonite
ore (22% solids, final acidity 21.6g/L). Their results indicate addition
of an anionic surfactant (0.4 g/L) reduces scale formation by only 37 %
and indicates the effect of the surfactant depends upon the system
and may not be so effective in plant conditions with higher percent

FIGURE 14 Dependence of the solubility of MgS04 on the liquor temperature and


sulphuric acid concentration (using data from Queneau et al., 1984; Siedel, 1965;
Marshall and Slusher, 1965).
590 B. I. WHIlTINGTON AND D. MUlR

solids. Surfactants are reported to be even less effective at higher


acidities and ineffective at lower acidities.

6. CONCLUSIONS

This review has examined the chemistry and technology of nickel-


laterite acid leaching.
The nickel and cobalt present in nickel laterites are not usually
present as discrete minerals, but as cations substituted within man-
Downloaded by [Aston University] at 02:54 21 January 2014

ganese oxides, goethite (e.g. Cawse, Syerston orebodies) and/or clays


(e.g. nontronite in the Bulong and Murrin Murrin orebodies). Be-
cause of this, it is difficult to upgrade the ore by beneficiation,
although the presence of a coarse, low-grade, silica material in the
Cawse and Ravensthorpe ore reportedly allows upgrading of these
deposits.
There exist a number of methods for refining nickel laterite ores,
and it is the ore mineralogy which often determines the most economic
process.
A pyrometallurgical process is used to treat magnesium-rich nickel
laterites containing at least 2% Ni. The ferronickel process involves
calcining and prereducing the ore then smelting to separate the
nickelliron-containing phase from the silica-magnesia slag. The sul-
phide matte process is similar to the ferronickel process, except a
sulphur source is added to form a nickel/iron sulphide matte.
A combined pyrometallurgical/hydrometallurgical process - the
Caron process - has been used to treat ores containing a wider range
of magnesium. The ore is initially roasted and reduced to ironlnickel
metal which is then leached with an ammoniacal ammonium carbo-
nate solution. High energy costs and poor cobalt recovery are the
main concerns.
A hydrometallurgical process - high temperature acid leaching - is
the method of choice for the nickel laterite processing plants being
commissioned in Western Australia. This process generally involves
dissolution of goethite (or other nickel/cobalt containing phases) to
release matrix-bound nickel and cobalt and form soluble iron sulphates
at temperatures around 250°C. Subsequent hydrolysis and precipita-
tion of solid iron oxides and basic sulphates minimises the liquor iron
ACID LEACHING OF NICKEL LATERITES 59 1

concentration and may, depending on the precipitate mineralogy,


regenerate acid. Formation of jarosite (NaFe3(S04)2(0H)6)o r basic
iron sulphate (Fe(OH)S04) entails a penalty with regards overall acid
(sulphur) consumption.
Soluble aluminium leached from clays similarly hydrolyse to the
basic sulphate salts although, since there is no further transition, the
acid bound with these solids is not released. Acid (sulphur) con-
sumption is the greatest operating cost in the high pressure acid
leach process and it is apparent that the digestion temperature and
the ore mineralogy/chemistry strongly influence the economics of the
Downloaded by [Aston University] at 02:54 21 January 2014

digestion process. Since the thermal hydrolysis occurs at a higher


temperature than required for hydrolysis of the soluble iron to
hematite (240°C versus 200°C), it is the aluminium hydrolysis reaction
which has the greatest influence on the choice of digestion tempera-
ture. The presence of magnesium compounds within the ore similarly
increase the acid consumption although, in the Amax process,
magnesium rich ore has been utilised to neutralise excess acid in
the post-digestion liquor.
Increasing the acid concentration increases the extraction of nickel,
cobalt and iron in a high-pressure acid leach. However, the use of too
high an acid concentration inhibits the hydrolysis of the iron and
aluminium salts which, in turn, increases the acid consumption. The
use of high acid concentrations also results in significant quantities
of acid remaining in the post-digestion liquor which must be subse-
quently neutralised.
The presence of soluble sulphates (e.g. MgS04) can influence the
mineralogy of the iron phases forming in a high temperature digestion
(T>185°C). In particular, addition of these salts increases the
acid concentration required for the formation of Fe(OH)S04 versus
hematite. Digestions conducted at high acid concentrations in the
saline process waters used at certain Australian processing plants re-
sult in the formation of jarosite. Digestion temperature and time also
influence the degree of hydrolysis of the iron and aluminium phases
forming during digestion, and hence the resultant mineralogy.
Researchers have investigated the effect of various salts on the kine-
tics of the goethite to hematite transformation during a high-tempera-
ture digestion. Digestions conducted in the presence of Na2S04 o r
MgS04 proceed with a higher goethite-hematite transformation
rate than digestions conducted in the absence of these salts or in
the presence of A12(S04)3 or Cr2(S04)3. Conducting the digestion
under conditions of low Eh additionally increase the goethite-
hematite transformation rate. At low Eh, the acid concentration
slows the goethite-hematite transformation while at high Eh the
rate increases with increasing acid concentration.
Experimental test results indicate substitution can occur in alunite
by: c r 3 + or ~ e for~ A13+;
+ ~ r 0 : - for SO:-; ~ i ' + ,c o 2 + ,Fe2+, ~ n ' +
for H ~ O +and for OH-. Results from Moa Bay leach product
suggest one ~ r 0 : - is incorporated into alunite for every two SO:-.
Downloaded by [Aston University] at 02:54 21 January 2014

Up to 30% of the SO:- and 30% of the OH- may be substituted by


~ r 0 : - and respectively while a maximum of 26% of the
may be substituted by Fe3+. In contrast, up to 65% of the A13+ may
be substituted by ~ e in~the+natrojarosite scale formed from 250°C
pilot plant leaching of a Bulong ore.
Addition of elemental sulphur prior to leaching improves the nickel
and cobalt extraction kinetics and additionally minimises the liquor
concentration of c~o:-.A reduction in the alunite concentra-
tion reduces the concentration of ~ i ' +and c o 2 + in alunite, which
improves the overall nickel and cobalt extractions. Addition of sul-
phur also reduces si0:- incorporated in the alunite. The silicon may
then form silica gel in the CCD tanks and generate difficulties in the
decantation process. Digestion temperature and time also influence the
post-digestion settling since partially hydrolysed slurries gener-
ally have much worse settling characteristics than fully hydrolysed
slurries. High acid additions, high leach temperatures and longer
leach retention times generally result in high post-digestion slurry
settling rates.
Scaling is a problem at the Moa Bay processing plant. Hematite,
basic iron sulphate, hydroniojarosite, hydronioalunite, Mg(Ni)-
SO4.H 2 0 and Mg(Ni)S04. 6 H 2 0 have all been reported at various
locations within the autoclaves. Scaling is also expected to be a prob-
. lem for autoclaves processing Murrin Murrin and Bulong ore ("high"
aluminium contents), but not those processing Cawse or Syerston
ore (low aluminium contents).
Of the experimental acid-leach parameters, localised acid concen-
tration has a significant influence on the mineralogy and amount of
the scale. Hydronioalunite and basic iron sulphate both reportedly
ACID LEACHING O F NICKEL LATERITES

form a t o r near the acid injection point in the autoclave. In contrast,


boehmite and hematite form in positions of low acidity.
Deliberate formation of M g S 0 4 . H 2 0 scale minimises formation
of the more-refractory aluminium-containing scales and coats the
autoclave walls with a material which is readily removed by a washing
with water a t a reduced temperature (the solubility of M g S 0 4 . H 2 0
increases with decreasing temperature). The hematite scale is bound
within the MgSO, matrix and is readily removed during this washing.
Downloaded by [Aston University] at 02:54 21 January 2014

References
Acharya, S., Anand, S. and Das, R. P. (1992) Iron rejection through jarosite pre-
cipitation during acid pressure leaching of zinc leach residue, Hydrometallurgy,
31, 101 - 110.
Allen, B. L. and Hajek, B. F. (1989) Mineral occurrence in soilenvironments, Soil Science
Society of America, Dixon, J. B. and Weed, S. B. Eds., pp. 199-278.
Anderson, E. (1997) Ramu nickel project - post pre-feasibility. ALTA 1997 Nickel/
Cobalt Pressure Leaching and Hydrometallurgy Forum ALTA Metalurgical Services
(Melbourne).
Anthony, M. T. and Flett, D. S. (1997) Nickel processing technology: a review, Minerals
Industry International, pp. 26-42.
Avotins, P. V., Ahlschlager, S. S. and Wicker, G. R. (1979) The rheology and handling
characteristics of laterite slurries, International Laterite Symposium, New Orleans,
Eds. Evans, D. J. I., Shoemaker, R. S. and Veltman, H., pp. 610-635.
Avramidis. K. S. and Turian, R. M. (1991) Yield stress of laterite suspensions, Journalof
Colloid and Interface Science, 143, 54-68.
Baghalha, M. and Papangelakis, V. G. (1998a) Pressure acid leaching of laterites at
250°C: a solution chemical model and its application, Metallurgical and Materials
Trans. B, 288, 945-952.
Baghalha, M. and Papangelakis, V. G. (1998b) The ion-association-interaction approach
as applied to aqueous H2S04-Al2(S04)3-MgS04 solutions at 250°C, Metallurgical
and Materials Trans. B, 288, 1021- 1030.
Baillie, M. G. and Cocks, G. C. (1998) Weda Bay laterite project, Indonesia, ALTA 1998
NickellCohalt Pressure Leachina- and Hvdrometallurav
~ervices(Melbourne).
-
-. Forutn ALTA Metalureical
Barkus, J. P., The outlook for nickel, Australian Nickel Forum Nickel'98, IIR Con-
ferences. Perth. Western Australia. 9 and 10 Februaw. ,, 1998.
Ber&owsky,R. M.', Laterite - new life of limonites?, Minerals Industrv International.
~ a n u a ; ~1997,
, pp. 48-55.
- Bhattacharya, I. N., Panda, D. and Bandopadhyay, P. (1998) Rheological behaviour
of nickel laterite sus~ensions. International Journal o f Mineral Processine. -. 53..
251-263.
Blakey, B. C. and Papangelakis, V. G. (1993) On the synthesis and solubility of
hydronium alunite. The Minerals, Metals and Materials Society (Warrendale),
E.waction and Processina- Division Conference. San Francisco. USA. Warren. G. W.
Ed., pp. 3- 19.
Blakey, B. C. and Papangelakis, V. G. (1996) A study of solid-aqueous equilibria by the
speciation approach in hydronium alunite-sulphuric acid-water system at high
temperatures, Metallurgical and Materials Transactions B, 278, 555- 566.
594 B. I. WHITTINGTON AND D. MUlR

Blight, D: F. and Muir, I. G. (1996) The Syerston polymetallic project, A L T A 1996


Nickel/Cobalr Pressure Leaching and Hydron~erallurgy Forum A L T A Metalurgical
Services (Melbourne).
Brand, N. W., Butt, C. R. M. and Hellsten, K. J . (1996) Structural and litho-
logical controls in the formation of the Cawse nickel laterite deposits, Western
Australia - implications for supergene ore formation and exploration in deeply
weathered terrains, Nickel '96: Mineral ro Marker, Aus. I . M . M . (Melbourne),
pp. 185-190.
Briceno, A. and Osseo-Asare, K. (1995) Particulates in hydrometallurgy Part I .
Characterisation of laterite acid leach residues, Merallurgical and Materials Trans-
actions, 26B, 1 123- 1 131.
Brindley, G . W. and Hang, P. T. (1973) The nature of garnierites I. Structure, chemical
compositions and colour characteristics, Clays and Clay Minerals, 21, 27-40.
Burger. P. A. (1996a) Origins and characteristics of lateritic nickel deposits, Nickel 96:
Mineral to Marker, Aus. I . M . M . (Melbourne), pp. 179- 183.
Downloaded by [Aston University] at 02:54 21 January 2014

Burger, P. A. (1996b) Ni/Co laterite deposits: geology, evaluation and mining, A L T A


1997 Nickel/Cobalr Lnrerire Projccr Developmenr Seminar A L T A Metalurgical
Services (Melbourne).
Burger, P. A. (1996~)The Bulong, Western Australia, Ni/Co laterite deposits - a case
history, Nickel '96: Mineral ro Marker, Aus. I . M . M . (Melbourne), pp. 37-41.
Camuti, K. S. and Riel, R. G. (1996) Mineralogy of the Murrin Murrin nickel laterites,
Nickel '96: Mineral ro Marker, Aus. I . M . M . (Melbourne), pp. 209-210.
Canterford, J. H. (1985) Sulphuric acid leaching of cobalt bearing manganese wad,
Hydrlron~eraNurgy,14, 35-46.
Cerpa. A., Garcia-Gonzalez, M. T., Tartaj, P.. Requena, J., Garcell, L. R. and Serna,
C. J. (1996) Rheological properties of concentrated lateritic suspensions, Progr.
Colloid Polyn~.Sci., 100, 266-270.
Chalkley. M. E., Balan, R., Kranz, H. U. and Sanchez, R. (1996) The acid pressure leach
process for nickel cobalt laterite: a review of operations at Moa Nickel, S. A,, A L T A
1996 Nickel/Cobalr Pressure Leaching and HvdromcraNurgy Forum A L T A Metal-
urgical Services (Melbourne).
Chalkley, M. E. and Toirac, I. L. (1997) The acid pressure leach process for nickel and
cobalt laterite. Part I: Review of operations at Moa, In: Proceedings of rhe Nickel-
Cobnlr 97 Inrerna~ionalSynlposiun~,Sudbury, Ontario, Canada, CIM (Montreal)
Cooper, W. C. and Mihaylov, I. Eds., pp. 341-353.
Chandra, D., Ruud, C. 0. and Siemens, R. E. (1983) Characterisation of lateritic nickel
ores by electron-optical and X-ray techniques, U S Deparrmenr of the Inferior Report
of lnvesrigarions 8835, 12 pp.
Chavcs, R. A,, Karelin. V. V. and Sobolev, B. P. (1968) side reactions during sulphuric
acid process of extracting nickel and cobalt from Cuban laterites, Tsvernye Merally,
pp. 66-70 (English edition).
Chou, E. C., Queneau, P. B. and Rickard, R. S. (1977) Sulphuric acid pressure leaching
of nickelifcrous limonites, MeruIIurgical Transacrious B, 88, 547-554.
Cornell. R. M. (1991) Simultaneous incorporation of Mn, Ni and Co in the goethite
(a-FeOOH) structure, Clay Minerals, 26, 427-430.
Cornell, R. M. and Giovanoli, R. (1989) Etrect of cobalt on the formation of crystalline
iron oxides from ferrihydrite in alkaline media. Clays and Clay Minerals, 37,
65 - 70.
Costa, M. L. (1997) Lateritisation as a major process of ore deposit formation in the
Amazon region, E.rplor. Mining Geol., 6(1), 79- 104.
Czerny, C. and Whittington, B. (1999) Scale formation in the pressure acid leach process,
A L T A 1999 Nickcl/Cobalr Pressure Leaching and Hydromerallurgy Forum A L T A
Metalurgical Services (Melbourne).
Da Silva, F. T. (1992) A thermodynamic approach for the sulphuric acid pressure
leaching of nickeliferous laterites, Minerals Engineering, S(9). 1061- 1068.
ACID LEACHING O F NICKEL LATERITES

Das, G. K., Acharya, S., Anand, S. and Das. R. P. (1995) Acid pressure leaching of
nickel-containing chromite overburden in the presence of additives, Hydromeral-
lurgy, 39, 117- 128.
Das, G. K., Anand, S., Acharya, S. and Das, R. P. (1997) Characterisation and acid
pressure leaching of various nickel-bearing chromite overburden samples, Hydro-
merallurg.~,44, 97 - l l I.
Das, G. K., Acharya, S., Anand, S. and Das, R. P. (1996) Jarosites: a review, Mineral
Processing and Exrracrive Melallurgy Review, 16, 185-210.
Deer, W. A., Howie, R. A. and Zussman. J. (1992) An Inrroducrion ro the Rock Forming
Minerals, Longman.
Duyvesteyn, W. P. C., Wicker, G. R. and Doane, R. E. (1979) An omnivorous process
for laterite deposits, Inrernarional Larerire Symposium, Evans, D. J . I., Shoemaker,
R. S. and Veltman, H. Eds., American Institute of Mining, Metalurgical and
Petroleum Engineers, pp. 553-570.
Faris, M. D., Collins, M. J., Becker, G. S., Matheson, P. J. and Lennard, G. A. (1997)
Downloaded by [Aston University] at 02:54 21 January 2014

The Calliope project: pressure acid leaching of nickel laterile ores from New
Caledonia, In: Proceedings of the Nickel-Cobalr 97 Inrernarional Sj~mposiurn,
Sudbury, Ontario. Canada, CIM (Montreal) Cooper, W. C. and Mihaylov, I. Eds.,
pp. 409-424.
Fekete,S. O., Wicker, G. R., Duyvesteyn, W. P. C. and Shieh, D. F. (1978)Acid leaching
of nickeliferous oxide ores with minimised scaling. U S Paten: 4098870. 8 pp.
Fischer, L., Zur Muhlen, E., Brummer, G. W. and Niehus, H. (1996) Atomic force
microscopy (AFM) investigations of the surface topography of a multidomain
porous goethite, European Journal of Soil Science, 47, 329-334.
Gasser, U. G., Jeanroy, E., Mustin, C., Barres, O., Nuesch, R., Berthelin, J. and
Herbillon, A. J. (1996) Properties of synthetic goethites with Co for Fe substitution,
Clay Minerds, 31, 465-476.
Georgiou, D.. Perdikis, P. and Papangelakis, V. G. (1994) Investigating the behaviour of
iron solid products during direct acid leaching of limonitic laterites, Separarion
Processes: Heavy Merals, Ions and Minerals, Misra, M. Ed., The Minerals, Metals
and Materials Society (Warrendale), pp. 275-286.
Georgiou, D. (1995) Kinetics of nickel dissolution during sulphuric acid pressure
leaching of a limonitic laterite, M.Sc. Thesis, University of Toronto, 128 pp.
Georgiou, D. and Papangelakis, V. G. (1998) Sulphuric acid pressure leaching of a
limonitic laterite: chemistry and kinetics, Hydronierallurgy, 49, 23046.
Gerth, J. (1990) Unit cell dimensions of pure and trace metal-associated goethites,
Geochim. er Cosmochim. Acra, 54, 363-371.
Golightly. J. P. (1979) Nickeliferous laterites: a general description, Internarionol
Larerire Symposium, New Orleans, Ed. Evans, D. J. I., Shoemaker, R. S. and
Veltman, H., American Institute of Mining, Mevalurgical and Petroleum Engineers,
pp. 3-23.
Golightly, J. P. (1981) Nickeliferous laterite deposits, Economic Geology, pp. 710-735.
GriRen. A. (1998) The Marlborough laterites project, ALTA 1998 Nickel/Cobalr
Pressure Leaching and Hydromerallurgy Foruni ALTA Metalurgical Services
(Melbourne).
Haigh, C. J., The hydrolysis of iron in acid solutions, Proc. Ausr. Insr. Min. Mer.,
September, 1967, pp. 49-56.
Hamilton. J., Ravensthorpe: Australia's next nickel laterite mine?, Australias Mining
Monrhly, May, 1998, pp. 28-29.
Hayward, N., Treatment of nickel laterites - Bulong nickel project - process design, AJ
Parker CRC/AMF "Hydromerallurgy - Current Practice" Short Course, 22-24
June, 1998, Perth, Western Australia.
Hazen, N. and Chou, E. (1997) Development of process design criteria for HPAL nickel
laterite projects, ALTA 1997 Nickel/Cobalt Pressure Leaching and Hydromerollurg)~
Forum ALTA Metalurgical Services (Melbourne).
Downloaded by [Aston University] at 02:54 21 January 2014
ACID LEACHING O F NICKEL LATERITES 597

Mason, P. G., Groutsch, J. V., Mayze, R. S. and White, D. (1997) Process development
and plant design for the Cawse nickel project, ALTA 1997 Nickel/Cobalt Pressure
Leaching and Hydrometallurgy Forum ALTA Metalurgical Services (Melbourne).
Mason, P. and Hawker, M. (1998) Ramu nickel process piloting. ALTA 1998 Nickell
Cobalt Pressure Leaching and HydrometaNurgy Forum ALTA Metalurgical Services
(Melbourne).
Matos, R. R. (1997) Industrial experience with the Ni/Co sulphide precipitation process,
In: Proceedings of the Nickel- Cobalt 97 International Symposium, Sudbury, Ontario,
Canada, Cooper, W. C. and Mihaylov, I. Eds., CIM (Montreal), pp. 371 -378.
Monhemius, A. J. (1987) Treatment of laterite ores of nickel lo produce ferronickcl,
matte or precipitated sulphide, Critical Reports on Applied Chemistry Volume 17
Extractive Metallurgy of Nickel, Burkin, A. R. Ed. Wiley, New York, pp. 51 -75.
Monti, R., and Fazakerley, V. W. (1996) The Murrin Murrin nickel cobalt project,
Nickel '96: Mineral to Market, Aus. I.M.M. (Melbourne), pp. 191- 195.
Moueram. G.. Ryan, M., Berezowsky. R and Raudsepp. R . , ~ ~ u r r M i nu r r ~ nn~ckel/
Downloaded by [Aston University] at 02:54 21 January 2014

cobalt oroiect oroiect dcvelonment overview. ALTA 1996 NickellCobalt Pressure


Leachi& and ~ ~ d r > m e t a l l uForum
r ~ ALTA ~ e t a l u r ~ i cservice;
al (Melbourne).
Motteram, G., Ryan, M. and Weizenbach, R. (1997) Application of the pressure acid
leach process to Western Australian nickel/cobalt laterites, In: Proceedings of the
Nickel-Cobalt 97 International Symposium, Sudbury, Ontario, Canada, Cooper,
W. C. and Mihaylov, I. Eds., CIM (Montreal), pp. 391-407.
Newman, A. C. D. and Brown, G. (1987) The chemical constitution of clays, In:
Chemistry of Clays and Clay Minerals, Newman, A. C . D., Ed., pp. 2- 128.
Palache. C., Berman. H. and Frondel. C. (1951) The Svstem o f Mineralonv. ". Wilev and
Sons, New York, pp. 555-563.
Papangelakis, V. G., Blakey, B. C. and Liao. H. (1994a) Hematite solubility in sulphate
orocess solutions. Hvdrometallur~v'94.
-, . IMM.. SCI.,Chaoman. and Hall. London.
UK, pp. 159- 175. '
Papangelakis, \'. G., Blakey, B. C , and Kambossos, J. (1994b) Behaviour of aluminium
during direct acid leaching of limonitic laterites, Proc. Int. Symp. Impurity Control
and Disposal in Hvdrometalluraical Processes. Harris, B. and Krause. E. Ed.. CIM
( ~ o n t r k a l ) pp.
, 3i5-325. -
Papangelakis, V. G., Georgiou, D. and Rubisov, D. H. (1996) Control of iron during the
sulohuric acid oressure leachinn of limonitic laterites. Proc. Second Int. Svmo. Iron
l Hydromelal1urgy, Iron Control and Disposal, Dutrizac, J . E. and ~ a r r i s ,
~ o n t r o in
G . B., CIM (Montreal) pp. 263-274.
Perdikis, P. (1996) Scale formation in a batch reactor under pressure acid leaching of
limonitic laterite, M.Sc. Thesis, University of Toronto, Canada, 125 pp.
Queneau, P. B., Doane, R. E., Berggren, M. H. and Cooperrider, M. W. (1983a)
Controlling scale composition during acid pressure leaching of laterite and gar-
nierite ore, U S Patent 4, 415, 542. 12 pp.
Queneau, P. B., Berggren, M. H., Cooperrider, M. W. and Doane, R. E. (1983b) Control
of silica deposition during pressure let-down of acidic leach slurries, In: Hydro-
metallurgy Research, Development and Plant Practice, Osseo-Asare, K . and
Miller, J. D. Eds., The Metallurgical Society of AIME.
Queneau. P. B., Doane, R. E., Cooperrider, M. W., Berggren, M. H. and Rey, P. (1984)
Control of autoclave scaling during acid pressure leaching of nickeliferous laterite
ore, Metallurgical Trans. B, 15B,433-440.
Queneau, P. 8. and Berthold, C. E. (1986) Silica in hydrometallurgy: an overview,
Canadian MelaNurgical Quarterly, 25, 201 -209.
Queneau, P. B. and Weir, D. R. (1986) Control of iron during hydrometallurgi-
cal processing of nickeliferous lalerite ores, In: Iron Control in Hydrametallurgy,
Dutrizac, J . E. and Monhemius, A. J. Eds., Ellis Honvood Limited, Chichester,
UK, pp. 76- 105.
598 B. I. W H l l T l N G T O N AND D. MUlR

Read, H. H. (1970) Rurley's elenrenrs ofnrineralogy, 26th edition, Thomas Murby and
Co, London.
Reid, J. G. (1996) Laterite ores - nickel and cobalt resources for the future, Nickel'96:
Mineral 10 Marker, Aus. I.M.M. (Melbourne), pp. 1 1 - 16.
Roach, G. I. D. and Cornell, J. B. (1985) Scaling in Bayer plants, Chemeca 85,
pp. 217-222.
Schellmann, W. (1978) Behaviour of nickel, cobalt and chromium in ferruginous lateritic
nickel ores, Bulletirr du Bureau Rcchcrckes Geoloiques er Minieres Section 11, 3.
275-282.
Schwertmann, U. and Taylor, R. M. (1989) Iron oxides. In: Minerals in Soil
Environnro~rs,Soil Science Society of America, Dixon, J. B. and Weed, S. B. Eds.,
pp. 379-438.
Scn, R., Yarar, B., Spottiswood, D. J. and Schowengerdt, F. D. (1984) Application of
quantitative mineralogy to the concentration of a laterite ore, Proc. Second Inr.
Congress Applied Mineralogy in rhe Minerals Industry, Park, W. C., Hausen, D. M.
Downloaded by [Aston University] at 02:54 21 January 2014

and Hagani, R. D. Eds., pp. 441-466.


Siedel, A. (1965) Soluhiliries: inorganic and meral-organic conrpounds, Fourth Edition,
Linke, W. L., revisor, American Chemical Society, Washington, D. C., 2, 524-526.
Sobol, S. 1. (1967) Formation and importance of the crust during leaching of lateritic ore
at Moa Bay (Oriente, Cuba), Original in Revista Technologica (Cuba).
Sobol, S. 1. (1968) Mineral composition of the Moa Bay laterites and its influence on
the leaching process by sulphuric acid in autoclaves, Nuesrra Indusrria, Revisla
Technologica, 6, 3 -24.
Sobol, S. 1. (1969) Physicoshemical studies of the composition and conditions of
formation of the crust (scale) in the reactors of the Moa Bay leaching plant, Reeisra
Technologica, 7, 3.
Stecmson, M. L. (1996) Planning and execution of a metallurgical test program
for laterite pressure leaching, A L T A 1996 Nickel/Cobalr Pressure Leaching and
Hydrometallurgy Forum A L T A Metalurgical Services (Melbourne).
Stucki, J . W. (1988) Structural iron in smectites, In: Iron in Soils and Clay Minerals,
Stucki, J . W., Goodman, B. A. and Schwertermann, U. Eds., Reidel Publishing
Company, Boston, pp. 625-675.
Taylor, A. (1996) Engineering issues and economics of pressure leaching of laterites,
A L T A 1996 Nickel/Cobalt Pressure Leaching and Hydrometallurgy Forum A L T A
Metalurgical Services (Melbourne).
Taylor, A. T. (1997a) Pressure acid leaching, A L T A 1997 Nickel/Cohalr Laterire Project
Developmenr Seminar A L T A Metalurgical Services (Melbourne).
Taylor, A. (1997b) Process selection, A L T A 1997 Nickel/Cobalr Laterire Project
Developntenr Seminar A L T A Metalurgical Services (Melbourne).
Taylor, A. and Cairns, D. (1997) Technical development of the Bulong lalerite treatment
project, A L T A 1997 Nickel/Cobalt Pressure Leaching and Hydromerallurgy Forum
A L T A Metalurgical Services (Melbourne).
Taylor, R. M. (1987) Non-silicate oxides and hydroxides, In: Chemisrry of Clays and
Clay Minerals, Mi~~eraological Society Monograph No. 6, Newman, A. C. D. Ed.,
Longman Scientific and Technical, pp. 129-201.
Terrell. J. and Dellar, G., Nickel Provides, Prospect, December-February, 1998, W A
Dcpr. Res. Developmenl, pp. 6- 18.
Tindall, G . P. and Muir, D. M. (1996) Transformation of iron oxide in nickel laterite
processing. Proc. Second. Inr. Symp. on Iron Conrrol 61 Hydromerallurgy, Iron
Control and Disposal, Dutrizac, J . E. and Harris, G. B. Eds., CIM (Montreal),
pp. 249-262.
Tindall. G . P. and Muir, D. M. (1997) Settling rates of nickel-laterite clay-rich residues
from high-temperature acid leaching, In: Proceedings of rhe Nickel-Cobalt 97
Inrenrorional Symposium, Sudbury, Ontario, Canada, Cooper, W. C. and Mihaylov,
I. Eds., CIM (Montreal), pp. 459-470.
ACID LEACHING OF NICKEL LATERITES 599

Tindall, G . P. and Muir, D. M. (1998) Effect of Eh on the rate and mechanism of the
transformation of goethite into hematite in a high temperature acid leach process,
Hydromelullurgy, 47, 377-381.
Toutwa, K. and Sasaki, K. (1986) Effect of coexisting sulphates on precipitation of ferric
oxide from ferric sulphate solutions at elevated temperatures, Iron Conrrol in
HvdromeraNurnv. Dutrizac. J . E. and Monhemius, A. J. Eds., Ellis Horwood,
~hichester,UK, pp. 454-476.
Troly, G., Esterle, M., Pelletier, B. and Reibell, W. (1979) Nickel deposi~sin New
Caledonia: some factors influencing their formation. Inrernatronal Laterite
Symposium, New Orleans, Eds. ~ v a n i ,D. J. I., shoemaker, R. S. and Veltman,
H., American Institute of Mining, Metalurgical and Petroleum Engineers.
pp. 85- 119.
Umetsu, Y., Tozawa, K. and Sasaki, K.4. (1977) The hydrolysis of ferric sulphate
solutions at elevated temperatures, The Merulliirgical Society o / C I M , pp. I I I - 1 17.
Vracar, R. 2.and Cerovic, K. P. (1997) Kinetics of oxidation of Fe(l1) ions by gaseous
Downloaded by [Aston University] at 02:54 21 January 2014

oxygen at high temperatures in an autoclave, Hydromerallurgy, 44, 113- 124.

APPENDIX

Calculation of Silica Distribution at the Moa Bay


Refinery
The following calculations are conducted to determine the distribution
of silicon in the liquor and solids (acid-leach residues). From this infor-
mation, it may be possible to comment on the mechanism by which
Cr(V1) "inhibits" silica gel formation (i.e. incorporation of
into alunite along with or formation of a silicon~chromium
"complex").
Sobol(1967, Tab. I1 of his paper) reports the Moa Bay leach tailings
composition for 1964-65. The typical leach tailing composition
0.060% Ni, 0.009% Co, 50.2% Fe, 1.80% Cr, 4.19% AI, 0.47% Mn,
9.14% SO4, 3.53% Si02 (2.18 Si04:A1, 92% Al as alunite) is chosen
for these calculations.
Sobol (1967) also presents the composition of Moa Bay ore
(1.35%Ni, 0.13% Co, 47.5% Fe, 4.O%AI, 0.76% Mn, 1.0% Mg,
2.0% Cr) while the Si02 concentration can be obtained from Table
111 (8.3% SO2).
Sobol(1967) additionally reports that 40-50000 tonnes of Moa Bay
ore generates 12- 14000 tonnes of leach residue.
From this data, we can calculate the distribution of the various
elements between the solid (tailings) and liquid (acid leach liquor)
phases. However, the results presented in Table VIII suggest the leach

Das könnte Ihnen auch gefallen