Nickel laterite deposits geological overview, resources

and exploitation
M. Elias

Mick Elias Associates, CSA Australia Pty Ltd, PO Box 139, Burswood WA 6100, Australia

In: Giant Ore Deposits: Characteristics, genesis and exploration, eds DR Cooke and J
Pongratz. CODES Special Publication 4, Centre for Ore Deposit Research, University of
Tasmania, pp 205-220.

This paper reviews the nature and genesis of nickel laterite mineralisation, and describes
the relationship between deposit characteristics (both geological and non-geological) and
the successful development of lateritic deposits as commercial nickel producers. The
importance of nickel laterites lies in their huge resource base, which could potentially
provide a much greater share of global nickel production than their current level compared
to nickel from sulphides.

Most of the worlds terrestrial nickel resources are hosted in nickel laterites, the products of
intense weathering of ultramafic rocks at the surface of the earth in humid climatic
conditions. The process of lateritisation involves the breakdown of primary minerals and
release of their chemical components into groundwater, the leaching of mobile components,
the residual concentration of immobile or insoluble components, and the formation of new
minerals which are stable in the weathering environment. The combined effects of these
processes is to produce a vertical succession of horizons of differing chemistry and
mineralogy (the laterite profile), the overall structure of which is governed by the
differential mobility of the elements in the weathering zone. The detailed structure of the
profile varies greatly, and in any one place is the result of the dynamic interplay of climatic
and geological factors such as topography, drainage, tectonics, structure and parent rock
lithology. Nickel can be enriched to ore grade in parts of the profile by being incorporated
into the structure of the newly formed stable minerals or into the alteration products of
primary minerals.

Exploitation of nickel laterites provides about 40% of the worlds production of nickel.
Three process routes are used commercially, each of which is suited to only part of the
laterite profile. Hydrometallurgical processes of sulphuric acid leaching and reduction
roast-ammonia leaching are used to extract nickel and cobalt from the upper, low-
magnesium part of the profile, and smelting is used for the high-magnesium silicates lower
in the profile. The economics of nickel laterite processing are strongly dependent on grade
and composition of ore feed, economies of scale, location, availability of low-cost energy
and well-developed infrastructure. Historically, nickel laterite projects have proven
difficult to develop and reach their nameplate capacity, but the enormous surface resources
of lateritic nickel provide compelling incentive to overcome engineering challenges
inherent in their successful treatment. The outlook is for a greater proportion of nickel
production in the future to come from lateritic sources.
Introduction

Laterites are the residual products of chemical weathering of rocks at the surface of the earth, in
which various original or primary minerals unstable in the presence of water, dissolve or break
down and new minerals are formed that are more stable to the environment. Laterites are
important as hosts to economic ore deposits, as the chemical interactions which together make up
the lateritisation process can in certain cases be very efficient in concentrating some elements.
Well-known examples of important lateritic ore deposits are aluminous bauxite and enriched iron
ore deposits, but lesser known examples include lateritic gold deposits (e.g., Boddington in
Western Australia) (Evans, 1993).

Nickel laterites are the product of lateritisation of Mg-rich or ultramafic rocks which have
primary Ni contents of 0.2-0.4% (Golightly, 1981). Such rocks are generally dunites,
harzburgites and peridotites occurring in ophiolite complexes, and to a lesser extent komatiites
and layered mafic-ultramafic intrusive rocks in cratonic platform settings (Brand et al, 1998).
Lateritisation processes result in the concentration by factors of 3 to 30 times the nickel and
cobalt contents of the parent rock. The processes, and the character of the resulting laterite, are
controlled on regional and local scales by the dynamic interplay of factors such as climate,
topography, tectonics, primary rock type and structure.

CUBA o
PHILIPPINES 22 N

INDONESIA

o
NEW CALEDONIA 22 S

AUSTRALIA

LATERITES SULPHIDES

Figure 1: Global distribution of sulphide and laterite nickel deposits

Most lateritic nickel resources occur within a band about 22 degrees of latitude either side of the
equator (Fig. 1) and the giant, and in some cases highest grade, deposits are concentrated in
tectonically active plate collision zones (eg Indonesia, the Philippines and New Caledonia) where
extensive obducted ophiolite sheets are exposed to aggressive chemical weathering in tropical
conditions of high rainfall and warm temperatures, and there is the greatest opportunity for
supergene enrichment. Resources in cratonic settings can be large but tend to be lower in grade
(e.g. Murrin Murrin in Western Australia). Cratonic shield deposits in West Africa (Nahon et al,
1982) and Brazil (Schobbenhaus, 1986) are within the equatorial zone, but those in the Balkans
(Greece, Albania and former Yugoslavia) (Valeton et al, 1987) and the Yilgarn craton in Western
Australia occur at higher latitudes. The latter two examples are fossil deposits, currently
situated in temperate or arid climates quite different from the warm, humid conditions under
which they formed.

Nickel laterites play an important part in the global nickel industry and currently account for
around 40% of the total nickel production of about 1 million tonnes. About 70% of all
continental or terrestrial nickel resources are contained in laterites. Production of nickel from
lateritic sources as a proportion of total (sulphide plus laterite) nickel production has remained
fairly constant over the last ten years (Fig. 2), but is expected to grow with time as easily-won
sulphide resources are depleted. The main barriers to more rapid growth in lateritic nickel
production are the high capital cost of processing facilities, high energy requirements in the
pyrometallurgical process routes, and the technical challenges of making hydrometallurgical
processing more efficient.

This paper is divided into two sections. Part 1 describes the processes by which lateritic deposits
enriched in nickel are developed over ultramafic rocks, the environmental factors controlling the
processes, and the nature of the lateritic profile formed as a result of these processes. Part 2
discusses the production of nickel from laterites, the extraction processes used on a commercial
scale, the structure of the nickel laterite industry and describes the factors that characterise
commercially successful operations.

Sulphide
800 Laterite 50.0%

700 % Laterite 48.0%

46.0%
600
44.0%
tonnes Ni ('000)

% Lateritic Ni

500 42.0%
400 40.0%

300 38.0%
36.0%
200
34.0%
100 32.0%
0 30.0%
1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001

Figure 2: World nickel production by ore type, 1988-2001.

Part 1 Geology of nickel laterite deposits
Geology of laterite profiles on ultramafic rocks

The process referred to as lateritisation is essentially chemical weathering taking place in

seasonally humid climates over long periods of time in conditions of relative tectonic stability,
allowing the formation of a thick regolith with distinctive characteristics (Trescases, in Butt and
Zeegers, 1992). Table 1 lists the main effects of chemical weathering of rocks in general, and
how these processes are manifested in the weathering of ultramafic rocks. In summary, the
process of lateritisation involves the breakdown of primary minerals and release of some of their
chemical components into groundwater, the leaching of mobile components, the residual
concentration of immobile or insoluble components, and the formation of new minerals which are
stable in the weathering environment. The net effect of the mineral transformations and the
differential mobility of elements involved produces a stratified or layered mantle of weathered
material overlying the parent rock from which it was formed, which is generally referred to as the
laterite profile.

Fe MgO Ni Co
Ferricrete
>50% <0.5% <0.6% <0.1%

Limonite
40-50% 0.5-5% 0.8-1.5% 0.1-0.2%

0.02-
Saprolite 10-25% 15-35% 1.5-3%
0.1%

Saprock
5% 35-45% 0.3% 0.01%
Bedrock

Figure 3: Schematic laterite profile developed on ultramafic rock in a tropical climate (Fe oxide-dominant
limonite zone), showing indicative chemical compositions in wt%. See Figure 4 for indicative thicknesses of
units.
The process is dynamic and gradual, and the gross stratification that one sees in the laterite
profile (Fig. 3) is essentially a snapshot of lateritisation in progress. The lowest layer reflects
early stages of weathering of the bedrock, and each layer further up represents a transformation
of that lying immediately below it, displaying progressively advancing stages of the process. In
the lowermost part of the profile (saprock), weathering takes place at contacts between minerals
and at fracture boundaries and there is abundant fresh rock and little alteration product. Further
up the profile, the proportion of surviving primary minerals decreases, and more strongly
fractured zones are completely altered eventually leaving detached boulders of intact bedrock
floating in a mixture of primary and alteration minerals in which the primary rock fabric is
preserved (saprolite). Higher layers consist totally of alteration minerals, and are marked by the
eventual loss of primary fabric. The zone above the saprolite is termed the pedolith in the
technical literature (Butt and Zeegers, 1992), but the term is seldom used in practice. The
pedolith is more often referred to as the limonite zone, the latter term being derived from the
dominant mineralogy (goethite and hematite) in this zone in oxide laterites.

Table 1: Main processes of chemical weathering and their effects in ultramafic rocks (after Butt and
Zeegers, 1992, p. 10)

General processes Effects in ultramafic rocks

1. Leaching of mobile constituents: Breakdown of olivine, pyroxene,
alkalis, alkaline earths serpentine and leaching of Mg, Ni, Mn,
Co
2. Formation of stable secondary Goethite formation, smectite formation,
minerals: Fe and Al oxides, clays adsorption of Ni from solution
3. Partial leaching of less mobile Leaching of silica in rainforest and moist
components: silica, alumina, Ti savanna climates
4. Mobilisation and partial Precipitation of Mn oxides and
reprecipitation of redox-controlled adsorption of Ni and Co from solution
constituents: Fe, Mn
5. Retention and residual Residual chromite concentration
concentration of resistant minerals:
zircon, chromite, quartz

Factors influencing weathering and profile development

The processes and conditions that govern and control lateritisation of ultramafic rocks are
numerous and varied on all scales, and consequently the nature of the profile varies in detail from
place to place in thickness, chemical and mineralogical composition, and the relative
development of individual profile zones. The main factors influencing the efficiency and extent
of chemical weathering, and consequently the nature of the profile, are:

Climate: Rainfall determines the amount of water passing through the soil, which
influences the intensity of leaching and removal of soluble components. In
addition to the amount, the effectiveness of rainfall (the extent to which
water is allowed to pass down through the profile rather than running off)
is important. Higher mean soil temperature (which is close to mean
 surface air temperature) increases the kinetics of weathering processes
(Butt and Zeegers, 1992).
Topography: Relief and slope geometry influence drainage, the extent to which water
passes into the soil, and water table level.
Drainage: Drainage affects the net water budget available for leaching from the
whole landscape.
Tectonics: Tectonic uplift increases erosion of the top of the profile, increases
topographic relief and lowers the water table. Tectonic stability allows
planation of the landscape, slowing groundwater movement.
Parent rock type: Mineralogy determines susceptibility of rocks to weathering and the
elements available for recombination as new minerals.
Structure: Faults and shears provide discrete zones of bedrock permeability; jointing
and cleavage improve pervasive alteration potential.

Clearly many of these climatic and geological factors are closely interrelated, and the
characteristics of a profile at any one place can best be described as due to the combined effect of
all the individual factors acting over time, rather than being dominated by any single factor.

The thickness of the laterite profile is determined by the balance between the rates of chemical
weathering at the base of the profile and physical removal of the top of the profile by erosion.
The rate of chemical weathering varies from 10 to 50 metres per million years, is generally
proportional to the quantity of water percolating through the profile, and is 2-3 times faster in
ultramafic rocks than sialic rocks (Nahon, 1986). Trescases (1975) has estimated that the rate of
downward movement in the base of weathering in the highlands of southern New Caledonia to be
from 125 to 140 metres per million years, but one-tenth this rate in the plateaux and terraces.
Rapid erosion rates restrict profile development to thicknesses of 50 to 100 metres, and their age
is therefore probably less than 1 million years.

Estimated rates of chemical weathering in stable cratonic settings where relief is subdued and
rainfall somewhat lower (e.g. Brazil and West Africa) are also in the range of 10 metres per
million years (Golightly, 1981). In these areas, tectonic stability and low erosion rates allow the
development of profiles over 100 metres thick, but these are older than those in tectonically
active settings.

Laterite profile types

Despite the complexity and interplay of controls, there are a number of broad features of the
laterite profile that are common to most examples, and it is possible to describe the range of
laterite types formed over ultramafic rocks in terms of three main categories on the basis of the
dominant mineralogy developed in the profile:

Oxide laterites: comprise largely Fe hydroxides and oxides in the upper part of the
profile, overlying altered or fresh bedrock;
Clay laterites: comprise largely smectitic clays in the upper part of the profile, and
Silicate laterites: comprise hydrated Mg-Ni silicates (serpentine, garnierite)
occurring deeper in the profile, which may be overlain by oxide
laterites.
Oxide laterites: These are the most common final products of lateritisation of ultramafic rocks.
In the presence of water, primary rock-forming minerals (mainly olivine and/or serpentine,
orthopyroxene and less commonly clinopyroxene) break down by hydrolysis releasing their
constituents as ions in aqueous solution. Olivine is the most unstable mineral and is the first to
be weathered; in humid tropical environments its Mg2+ is totally leached and lost to groundwater,
and Si is largely leached and removed. Fe2+ is also released but is oxidised and precipitated as
ferric hydroxide, initially amorphous or poorly crystalline but progressively recrystallising to
goethite which forms pseudomorphs after olivine. Orthopyroxene and serpentine hydrolyse after
olivine, also releasing Mg, Si and being replaced by goethitic pseudomorphs. Initially, while co-
existing ferro-magnesium minerals remain unweathered and support the rock fabric, the
transformation is isovolumetric and primary rock textures are preserved, but as the extent of
destruction of primary minerals increases, relict primary textures are lost by collapse and
compaction of the fabric resulting in a textureless massive goethite. The mineralogical
transformation involving loss of Mg and residual concentration of Fe results in the obvious and
familiar chemical trend in laterites of Mg decreasing upwards and Fe increasing upwards through
the laterite profile (Fig. 3).

Ni and Co behave differently to the major elements. Nearly all of the original Ni and Co in the
ultramafic bedrock occurs in solid solution in olivine and olivine-derived serpentine. As these
minerals break down, the released Ni and Co ions have a chemical affinity for the newly-formed
poorly-crystalline Fe hydroxides and are incorporated and concentrated into their structure by a
combination of adsorption and replacement of Fe3+ (Gerth, 1990). Contents of 1.5% Ni and 0.1%
Co are seen in massive goethite developed from olivine containing 0.3% Ni and 0.02% Co. Ni
and Co are also incorporated strongly into Mn oxides (asbolanes) where these are precipitated by
redox reactions as veins and surface coatings on minerals and in fractures. Mn oxide minerals
containing up to 12% Ni and 8.5% Co are known (Elias et al, 1981).

Mineralogical trends in the pedolith reflect the gradual transformation of goethite to hematite.
The first-formed Fe hydroxides resulting from the breakdown of ultramafic minerals are
amorphous or poorly crystalline. Their crystallinity improves with time to well-structured
goethite with a characteristic yellow-brown colour, which is progressively replaced by red-brown
hematite is the goethite dehydrolises. The colour change is reflected in the commonly-used
terminology of yellow limonite and red limonite for the lower and upper parts of the
limonite zone respectively. The transformation of goethite to hematite is accompanied by a
loss of Ni, as hematite cannot accommodate in its lattice the Ni formerly contained in the
goethite. At the very top of the profile, a nodular fabric develops in the red limonite, which
develops further to an indurated crust as the nodules coalesce and harden. The crust is known as
ferricrete, iron crust, or by the French term cuirasse.

Oxide-dominated deposits can pass directly from totally altered goethitic saprolite down into
fresh bedrock over a distance of only a few centimetres. More generally there is an interval
between goethite saprolite and the fresh bedrock interface, as shown in Figure 3, which
comprises a mixture of goethite, oxidised ultramafic material and fresh ultramafic rock boulders.

Important examples of oxide deposits as described above are Moa Bay and Pinares (Cuba)
(Linchenat and Shirokova, 1964), and Goro and Prony (southern New Caledonia) (Golightly,
1981).
A particular variety of oxide deposit that is formed over dunite bedrock is composed of goethite
and minor clay with abundant free chalcedonic silica in forms ranging from fine-grained particles
to coarse veins and discontinuous lenses and masses. Examples of this silica-oxide laterite can be
found in association with clay laterites developed over peridotites, and it appears the lack of Al in
the dunite precursor precluded development of secondary clay. Cawse and Ravensthorpe in
Western Australia are examples of silica-oxide laterites (Brand et al, 1998).

Clay laterites: In less severe conditions of weathering, e.g., cooler or drier climates, silica is not
leached as it is in humid tropical environments, and instead combines with Fe and a small amount
of Al to form a zone where the smectite clay nontronite predominates, in place of Fe oxides.
Nontronite plays a similar role to goethite in fixing Ni ions within its lattice where they substitute
for Fe2+ and are fixed in inter-layer positions (Brand et al. 1998). Nontronite clays typically
contain 1.0-1.5% Ni in mineralised laterite. Silica in excess of that required to form nontronite
can be deposited as opaline or chalcedonic nodules in the clay. Clay laterite profiles are also
preferentially developed where there is restricted groundwater movement such as in broad areas
with low topographic relief (Golightly, 1981). The clay horizon may be overlain by a thin zone
of more Fe-rich oxide material which is generally low in Ni, and is underlain by partially
weathered saprolite containing serpentine and nontronite. Clay-dominated nickel laterite profiles
are extensively developed in Australia; e.g, Murrin Murrin (Monti and Fazakerley , 1996),
Bulong (Elias et al, 1981) and Marlborough (Golightly 1981) and in Brazil.

SILICATE CLAY OXIDE

(eg New Caledonia) (eg Murrin Murrin) (eg Moa Bay)
DEPTH (m)
0 Iron cap
Cuirasse Colluvium Limonite
Red Ferruginous overburden
limonite zone
Yellow Limonite
limonite Smectite
zone
Earthy Saprolite
20 ore

Ore with
boulders Saprolite
(Serpentine,
chlorite, Bedrock
Rocky
ore smectite )

40 Bedrock
Bedrock

Figure 4: Schematic comparison of principal laterite profile types

Silicate laterites: In situations where there is slow continuous tectonic uplift and the water table
is kept low in the profile, weathering over long periods can result in the development of a thick
saprolite zone, which may be overlain by a thin limonite zone depending on the intensity of
erosion at the top of the profile (Golightly, 1981). Silicate laterites are characterised by an
absolute enrichment or concentration of Ni in the saprolite zone which comprises altered primary
minerals such as secondary serpentine, and neoformed goethite, smectite clays and garnierite
(garnierite is a term for mixed-structure hydrous Ni-Mg silicates of low crystallinity with
affinities to serpentine, talc and chlorite; see Pelletier, 1996). Much of the nickel is derived from
that released by the recrystallisation of goethite to hematite further up in the profile. Nickel is
reprecipitated within the saprolite by substituting for Mg in secondary serpentines (which can
contain up to 5% Ni) and in garnierite which can grade over 20% Ni (Pelletier, 1996). Average
content of Ni in silicate laterites is typically 2.0-3.0%. Examples of silicate profiles are the
economically important laterites on the massifs of New Caledonia (Golightly, 1981), which
contain a large proportion of the worlds lateritic nickel resources. Silicate laterites are also the
ore source of most of the nickel currently produced from laterites.

A schematic comparison of the three profile types is shown in Figure 4.

Regional and local geological setting of nickel laterite deposits

On a global basis, nickel laterite deposits are found in two tectonic settings (Brand et al, 1998):

Accretionary terrains these are tectonically active zones often associated with oceanic or
continental plate boundaries and collision zones. Thrust faulting has obducted slabs of upper
mantle peridotites and associated rocks forming ophiolite complexes with extensive areas of
exposure at surface. Tectonic processes (i.e., uplift) play a large part in influencing the type of
nickel laterite deposits formed. Ages of host ultramafics and lateritisation range from Cretaceous
through to the late Tertiary. These terrains are typical of active and inactive island arc settings of
Indonesia, the Philippines and New Caledonia.

Cratonic terrains laterites are developed on komatiites and the ultramafic phases of layered
mafic complexes of any age from Archaean through the Palaeozoic. Relative tectonic stability
allows peneplanation and the laterites are developed in moderate to subdued relief. Restricted
drainage often results in smectite formation instead of oxides. Tectonic stability allows
continuous lateritisation over extended time periods giving rise to deep weathering, and extension
of laterite formation into cooler or less humid climate zones. Examples include the nickel
laterites and bauxites of the Yilgarn craton, Western Australia, and parts of Brazil, West Africa
and the Urals in Russia/Ukraine.

On a local or deposit scale, laterites may be classified according to their topographic setting into
plateau, slope and terrace deposits (Troly et al, 1979).

Plateau deposits are affected by active drainage process but less by erosion, and hence tend to
show complete profile development and form a thicker saprolite zone. The Thio Plateau and
Koniambo deposits in New Caledonia are good examples of plateau deposits
Slope deposits are affected more by erosion and the oxide zone is poorly developed or absent.
Silicate laterite development can also be thinner than plateau deposits as the groundwater
movement has a greater lateral component, i.e. downslope. However, the increased lateral
groundwater flow can cause higher Ni grades to be developed.

Terrace deposits are relics of earlier peneplains or erosion surfaces and indicate a temporary stop
of tectonic uplift. They tend to show development of complete profiles and of thick saprolite
zones. Terraces can include the products of erosion of laterite from surrounding plateaux
(particularly oxide material which is readily eroded) and multiple cycles of lateritisation are
possible with the possibility of increased grades. Good examples of terrace deposits are found in
southern New Caledonia, where the Goro deposit is the best known.

Global resources of nickel in laterites

Total global resources of nickel contained in laterites are about 160 million tonnes (Fig. 5).
When classified into limonitic (combined oxide and clay) and silicate-dominant deposits
depending on the predominant profile zone (or in some cases the focus of a development project),
limonitic or low-Mg deposits host about 70% of the total lateritic nickel resource. Because of
their lower average grade, ore tonnages in limonitic deposits are closer to 80% of total tonnes.

When viewed on a regional basis (Fig. 6), the importance of just three countries, and in particular
of New Caledonia, becomes very clear New Caledonia, Indonesia and Australia together
account for some 60% of the total resources of lateritic nickel, and New Caledonia, despite its
small size, lays claim to 26% of the total. Nickel grade and proportion of resources in limonite
vary between the three, with New Caledonia showing higher grades in both limonite and silicate
ores. Its proportion of nickel in limonites is close to the global average of 70%, but these are
nearly entirely in the small southern region (Massif du Sud). Australia, by contrast, has almost
no saprolitic resources, and the grade of its limonitic laterites is considerably lower than other
countries.

Silicate
Limonite 30%
70% Total 160.7 Mt Ni
2712 Mt ore
9569 Mt ore 1.79% Ni
1.17% Ni 48.5 Mt Ni
112.2 Mt Ni

Figure 5: Global lateritic nickel resources (by contained nickel)

The range of individual deposit sizes and Ni grades for both silicate and limonitic (oxide and
clay) laterites is seen in Figure 7. Limonitic laterites have a maximum deposit grade of 1.6% Ni
and large tonnages, whereas silicate laterite deposits are smaller (mostly less than 50 Mt) and
grades of 2.4-2.6% Ni predominate. A number of outstanding deposits in terms of size and grade
are named in Figure 7.

Americas Other
Balkans
New Caledonia
Africa
2.49%

Brazil
1.38% 1.38% Average Ni grade

silicate

Philippines 1.88% limonite

1.25%
1.36% Indonesia
Cuba 0.86%

Australia

Figure 6: Lateritic nickel resources by region based on contained Ni tonnes, divided into limonite (oxide
plus clay) and silicate deposits.

3.0
2.8
Konia mbo
2.6
2.4
2.2
Ni grade (%)

2.0
1.8 Soroa ko
Bahodopi Goro
1.6
1.4
Sam pala
1.2
Murrin
1.0
0.8
0.6
0 50 100 150 200 250 300 350 400
Tonnes (millions)
Silicate Limonite

Figure 7: Tonnage-grade plot of individual laterite deposits. Limonite deposits include oxide and clay
laterites. See Table 4 for further information on named deposits.
Part 2 Exploitation of nickel laterite resources
From the earliest nickel production in 1875 until the end of the 19 th century when the vast
sulphide nickel deposits of Sudbury came into production, laterites from New Caledonia were
essentially the sole source of nickel in the world. Despite the fact that around 70% of global
nickel resources are hosted by laterite, production of nickel from lateritic sources has lagged
behind that from sulphides. Currently about 40% of nickel production comes from treatment of
laterite ores, and this proportion has been in the range of 30-40% for the last decade (Fig. 2).
Before the late 1950s lateritic nickel production was about 10% of total nickel, but strong
demand for nickel at that time driven largely by US requirements for the Korean War effort saw a
number of laterite operations developed which quickly doubled the supply from laterites to 20%
of total nickel produced.

The reasons why the production of lateritic nickel is not at a level commensurate with its resource
position relative to sulphide nickel are essentially economic, and are summarised in Table 2. The
average weighted cost of production (fully allocated costs net of by-product credits) of nickel
currently produced from laterites is greater than that for sulphide nickel by some 20%. The
mining cost component of lateritic nickel production costs is considerably less than for sulphides,
which are mostly produced from underground mines. Processing of nickel laterites is
considerably more expensive than for sulphides, and more than offsets the mining cost
advantage.

Table 2: Summary of issues relevant to the economics of nickel sulphide and nickel laterite deposits.

Sulphides Laterites
Treatment costs lower than for laterites - Nickel in stable oxide mineralogy which
nickel in metastable minerals, which requires energy and/or aggressive chemical
contain latent energy in the form of S attack to disassociate.
Nickel-bearing minerals can be easily Only rare opportunities to separate and
separated from barren (gangue) material at concentrate nickel-bearing minerals, and
the mine site. Only the concentrate (higher must apply treatment processes to all the
grade, lower volume) undergoes further ore. Requires handling large quantities of
treatment. material with low nickel concentrations,
and large volumes of consumables which
must be transported to the plant.
Ore and concentrate have generally High short-range variability of mineralogy
consistent characteristics within a deposit and chemistry, requires careful planning,
easy to manage metallurgy. mining and blending to produce consistent,
predictable feed for the treatment plant.
High recovery of nickel from ore possible Variable composition through profile
with only one process route restricts recovery of nickel from only part
of the system with one process route.
High value in by-products (eg. Cu, PGE) Few opportunities for by-product credits
helps to offset treatment costs of Ni (only Co).
Moderate capital cost per unit Ni capacity Very high capital cost of processing
for processing facilities. facilities more so if the energy source has
to be part of the capital cost.
Processing of nickel laterites

Methods currently in use on a commercial scale to extract nickel from laterites comprise three
main processing routes:

Smelting to produce ferro-nickel or matte,

Caron process (reduction roast ammoniacal leach), and
High pressure acid leaching (HPAL) using sulphuric acid

Table 3 shows nickel production from laterites in 2001 by each of these process routes.

Due to the requirements of each process and the wide range in chemical and mineralogical
composition in the laterite profile (see Fig. 3), each of these processes is only suited to a part of
the profile.

The pyrometallurgical smelting route is the oldest and most widely used process. The process
route treats the more nickel-rich silicate fraction of the profile and produces ferro-nickel (a
reduced Fe-Ni alloy which can be directly used for stainless steel production) and sulphide matte
(which can join a conventional sulphide treatment route) in an electric furnace. The technology
is relatively simple, well-tried and reliable, and has the advantage that the composition of the ore
makes it essentially self-fluxing. The process has high Ni recoveries (90%), but there is no Co
recovered when ferro-nickel is produced. The process economics are most heavily dependent on
the cost of power and the successful projects (e.g. Sorowako) have their own hydroelectric power
plants.

The Caron process is applicable to oxide laterites (limonite) with tolerance for some silicate
laterite, but excessive silica and Mg leads to decreased Ni recoveries. The process involves
drying and roasting in a reducing atmosphere, followed by low-pressure ammonia leaching. Ni
and Co are recovered by solvent extraction, and further refined to product stage by calcination
and reduction. Recoveries of Ni are around 80% and of Co only about 40-50%. Ore drying and
reduction roasting consume considerable energy, and process economics are heavily dependent
on fuel prices. Caron plants were developed before the oil crises of the 1970s, and struggled to
remain viable after fuel prices rose. None has been built since then, nor is there likely to be more
built in the future.

The HPAL process was first used commercially in 1959 at Moa Bay (Cuba), which remained the
only operating HPAL plant until the three Western Australian plants (Bulong, Cawse and Murrin
Murrin) were developed in the late 1990s. The HPAL process involves leaching ore with
sulphuric acid in an autoclave at about 250C and extracting Ni and Co from the leach liquor by
various methods such as sulphide precipitation using H2S, or solvent extraction and
electrowinning. HPAL is used to treat predominantly the oxide fraction of the laterite and has
high recoveries of both Ni and Co (over 92% in the leach stage). The process economics are
largely dependent on the cost of sulphur and the conversion of the sulphur to sulphuric acid
(sulphur-burning acid plants can generate much of the energy requirements for a HPAL plant).
As the consumption of sulphuric acid is determined mainly by the Mg level in the ore, the latter
becomes a critical factor in managing mining and blending.
Table 3: Lateritic nickel production for 2001 by process route. Ores come from local deposits except where
plants are marked with an asterisk, indicating production is from imported ores.

Tonnes % of total
Plant Company Country
Ni production

Smelting Ferro-nickel

Doniambo SLN New Caledonia 45,912

Hachinohe Pamco* Japan 35,700
Hyuga Sumitomo* Japan 19,800
Oheyama Nippon Yakin* Japan 13,600
Falcondo Falcondo Domin. Republic 21,662
Cerro Matoso BHPBilliton Colombia 38,500
Larymna Larco Greece 18,600
Morro/Codemin Anglo-American Brazil 6,000
Loma de Niquel Anglo-American Venezuela 11,600
Pomalaa Aneka Tambang Indonesia 10,300
Various (4-5) Various Russia / Ukraine 21,600
S/T 243,274 51.2%

Smelting Matte

Doniambo SLN New Caledonia 13,061

Sorowako PT Inco Indonesia 62,600
S/T 75,661 15.9%

Caron Process (Ammonia leach)

Yabulu Queensland Ni* Australia 28,500

Nicaro Cuban Govt Cuba 16,400
Punta Gorda Cuban Govt Cuba 26,900
Tocantins Votorantim Brazil 16,700
S/T 88,500 18.6%

HPAL (Acid leach)

Moa Bay Sherritt/Cuban G. Cuba 29,226

Bulong Preston Res. Australia 6,262
Cawse OM Group Australia 7,200
Murrin Murrin Anaconda Nickel Australia 24,991
S/T 67,679 14.2%

Total Laterite Production 475,114 100%

Total Sulphide Production 674,886
Total Nickel Production 1,150,000
The HPAL process can also be applied with high recoveries to clay laterites, but the presence of
silica in the ore from both nontronite and serpentine can create slurry handling problems in the
autoclave and subsequent steps; these require increased operating costs to overcome, and result in
reduced efficiencies such as through-put rates in various parts of the plant. Pure oxide ores with
low silica such as at Moa Bay are most efficient in the HPAL circuit.

Ore beneficiation

Certain types of lateritic ores can be beneficiated before being fed to the process plant.
Beneficiation is the process whereby a low-grade component of the mineralisation is separated
from the rest and rejected, leaving a component with a higher grade to be treated. This is
analogous to making a concentrate from a sulphide ore, but the concentration factor is much
smaller for laterite ores. In the beneficiation process, some of the nickel is lost to the reject
component but this is outweighed by the improved economics which result from the higher feed
grade.

The practice has been applied for many years to silicate ores, where coarser boulders and
fragments of hard, less-altered rock have much lower Ni content than the matrix of softer altered
material in which they occur. Projects where optimisation of this type of beneficiation is integral
to their economic viability include Sorowako (Indonesia) and Kopto (New Caledonia), although
the practice of screening out coarser lumps and boulders is carried out at all laterite mines. At
Kopto, a grade increase of 25% is achieved between run-of-mine ore and product finally shipped
to the smelter.

100 2
90 1.8
80 1.6
70 1.4
Ni grade %

60 1.2
%

50 1
40 0.8
% Mass passing
30 0.6
20 % Ni recovery 0.4
10 Ni grade % 0.2
0 0
0 0.5 1 1.5 2 2.5
Screen size (mm)

Figure 8: Beneficiation of silica-oxide ore from Western Australia. Plot shows parameters of material passing
screen sizes shown on the x-axis. Vertical dashed line is at 212 microns. % mass passing is the weight of the
undersize fraction at the relevant screen size as a percentage of the total bulk sample weight, and Ni grade % is the
grade of the undersize fraction. % Ni recovery is the proportion of total nickel in the bulk sample that is reported
in the undersize fraction.
Beneficiation with more effective results is possible with silica-oxide laterite. The silica
component is essentially devoid of nickel and easily separated from the associated goethite by
simple screening after crushing. Figure 8 shows an example of how screening a crushed laterite
of bulk grade 1.04% Ni at 212 microns can produce a fine fraction comprising 53% of the
original mass, containing 81% of the original nickel and with a grade of 1.44% Ni. This form of
beneficiation is carried out at the Cawse operation in Western Australia.

Clay ores are not amenable to upgrading, as Ni is uniformly distributed and there is no discrete
low-nickel fraction which can be separated, except minor secondary silica in places.

Structure of the nickel laterite industry

Table 3 shows that two-thirds of lateritic nickel is produced by smelting. Smelters range in size
and have up to 60,00070,000 t/yr Ni capacity (e.g. Sorowako, Doniambo). Most smelters draw
ore from nearby mines (often more than one) in integrated operations. The three Japanese
smelters (Table 3), however, import ore purchased from mines in New Caledonia, Indonesia and
the Philippines. In the case of Pamco and Sumitomo, the smelter owners have an equity interest
in some of the mining operations that provide their ore.

The Yabulu refinery in Townsville, Australia, using the Caron process was built to treat ore from
the Greenvale laterite deposit in Queensland, but now treats only ores imported from New
Caledonia, Indonesia and the Philippines after the reserves at Greenvale were depleted. The
Caron process plants in Cuba treat predominantly oxide ore from local mines.

The advent of new HPAL projects has been the most significant factor affecting the nickel laterite
industry in the last decade. For many years the Moa Bay operation in Cuba was the sole HPAL
producer. Over this time little was known about its performance as the plant was nationalised by
the Castro regime immediately after commissioning in 1959. When information became
available after the collapse of the Soviet economy in the 1990s, the Moa plant was found to be
operating remarkably well despite poor maintenance. This observation, supported by a number
of other factors, led to increased global industry interest in the potential of HPAL. These factors
included a significant drop over the last decade in the price of sulphur and sulphuric acid (due to
environmental pressures on smelters and resulting increased supply), technical advances in
solvent extraction and improved reagents to selectively separate Ni and Co from leach solutions,
and improved design and experience in autoclave technology from the gold industry.

On the back of the renewed interest in HPAL, some 15-20 lateritic nickel projects were proposed
globally and all were subjected to some form of study ranging from scoping to bankable
feasibility studies. Three greenfields plants were built in Western Australia: Bulong and Cawse
are essentially large demonstration plants of about 9500 t/yr Ni capacity which were intended to
be expanded after commissioning, and Murrin Murrin is a 45,000 t/yr Ni plant. All three
operations were brought on stream in 1999 and suffered severe commissioning difficulties,
engineering failures, longer than expected ramp-up times and higher costs, and all are currently
operating below capacity and are in severe economic circumstances. As a consequence of these
difficulties, most other project studies have been put on hold. An exception is Incos Goro
project in New Caledonia which is under construction and expected to start producing in 2004-5
with a capacity of 54,000 t/yr Ni.
The economics of laterite operations vary widely. To better understand the economic drivers in
the industry, it is helpful to refer to a list of 20 nickel laterite projects around the world (both
producers and non-producing projects in the exploration or development stage) selected on the
basis of the largest published resource size measured in contained Ni (Table 4), and consider
them in the light of the process descriptions given above.

The more profitable laterite operations (Sorowako and Cerro Matoso) are notable because of their
large size, and because they have cheap energy supplies. In the case of Sorowako, the operation
is supported by a fully dedicated hydroelectric scheme. The economics of SLNs operations in
New Caledonia are improved by the high ore grades (over 2.6% Ni) and available hydroelectric
energy, but some high-cost, oil-generated power is required. Falconbridges Falcondo smelter in
the Dominican Republic relies totally on oil for energy, and the smelter has been closed a number
of times when high oil prices have coincided with low Ni prices (most recently for three months
in late 2001). The Exmibal operation in Guatemala operated for only three years from 1977
before being closed and has not operated since; it barely reached half its production capacity, and
was penalised by small production capacity, low grade (1.8% Ni) and its dependence on oil-
generated power.

In contrast, a significant number of laterite operations are marginal or loss-making and rely
heavily on forms of Government support to remain in operation, thereby protecting reliability of
strategic supply, protecting the national economy, or maintaining high employment levels (these
are sometimes referred to as social producers). In the case of the Japanese smelters, the
assistance is in the form of tariff protection which offsets the high costs of ore purchase, ore
transport and energy. Greek producer Larco has poor economics due mainly due to poor ore
grade (1.1% Ni), high energy costs and high smelting temperatures of the Fe-rich ores, and has
received extensive Government subsidies.

The economic problems faced by the three Western Australian operations have resulted mainly
from capital over-runs and underestimation of operating costs, as well as lost production due to
engineering failures in pumps, pipes and pressure vessels unable to withstand the severely
corrosive conditions of hot, concentrated sulphuric acid slurries. The experience has shown that
in order to reduce technical risk, feasibility studies have to be more thorough, capital costs will be
high and projects will have to be large in order to achieve economies of scale, leaving the
development of HPAL projects in the domain only of the major mining houses. The capital and
operating cost over-runs of the two small Western Australian HPAL producers has exacerbated
their already precarious ability to repay capital.

Despite these events, most new development projects being studied or planned for construction
are HPAL plants (ten in Table 4). It is considered by the promoters of these projects and some
industry observers that the experience of the three Western Australia HPAL projects has shown
where the effort has to be made to overcome the operational difficulties, and that the fundamental
economics of HPAL are still sufficiently attractive for these projects to be progressed. Of the
development projects listed, Sipilou, Musongati and Moramanga have location disadvantages
away from the coast, and a number of projects are situated in countries that have political factors
not in their favour. The latter issue is one which becomes important when the high capital cost is
considered.
 Table 4: Summary of features of twenty large nickel laterite projects (producers, ex-producers and
development projects) selected on the basis of contained Ni metal in resources.

Product
Contain %
Predom ion
ed Ni in % Ni Co Processing
Project Country Ownership inant Project status capacit
resourc grade grad route
ore type y
e e
(t/yr Ni)
New
Goro INCO-BRGM Oxide >5 Mt 1.56 0.18 Feas. Study HPAL
Caledonia
Matte
Sorowako Indonesia INCO-Sumitomo Silicate >5 Mt 1.80 Producer 68000
smelting
Sampala Indonesia Rio Tinto Oxide >5 Mt 1.34 0.10 Exploration HPAL
New Falconbridge- FeNi
Koniambo FeNi Silicate 3-5 Mt 2.58 0.07 Feas. Study
Caledonia SMSP smelting
Falconbridge- Oxide-
Sipilou Cote d'Ivoire 3-5 Mt 1.48 0.11 Feas. Study HPAL
SODEMI silicate
Western Anaconda-
Murrin Murrin Clay 3-5 Mt 0.99 0.06 Producer 45000 HPAL
Australia Glencore
BHP Billiton- Oxide-
Gag Island Indonesia 3-5 Mt 1.35 0.10 Feas. Study HPAL
Aneka Tambang silicate
Matte
Bahodopi Indonesia INCO Silicate 3-5 Mt 1.77 Feas. Study
smelting
SLN New
Eramet-SLN Silicate 3-5 Mt 2.40 Producer 60000 Smelting
Operations Caledonia
Weda Bay- Oxide-
Weda Bay Indonesia 2-3 Mt 1.37 0.12 Exploration HPAL
Aneka Tambang silicate
Pinares de Cuban
Cuba Oxide 2-3 Mt 1.07 0.12 Exploration HPAL
Mayari Government
FeNi
Pomalaa East Indonesia INCO Silicate 2-3 Mt 1.83 Exploration
smelting
Camaguey Cuba BHP Billiton Clay 2-3 Mt 1.30 0.05 Exploration HPAL
Oxide-
Musongati Burundi Argosy 2-3 Mt 1.31 0.08 Exploration HPAL
silicate
Moramanga Madagascar Phelps Dodge Oxide 2-3 Mt 1.11 0.10 Feas. Study HPAL
New New Caledonian
Prony Oxide 2-3 Mt 1.40 0.14 Exploration HPAL
Caledonia Government
FeNi
Euboea Island Greece Larco Oxide 2-3 Mt 1.00 Producer 20000
smelting
Matte
Exmibal Guatemala INCO Silicate 1-2 Mt 1.83 ex-producer 11300
smelting
FeNi
Cerro Matoso Colombia BHP Billiton Silicate 1-2 Mt 2.35 Producer 55000
smelting
Dominican FeNi
Falcondo Falconbridge Silicate 1-2 Mt 1.23 Producer 34000
Republic smelting

Ingredients of a successful laterite project

The experience of history has shown that new nickel laterite development projects have a very
patchy record of success. Many have suffered from construction cost overruns, unforeseen
technical difficulties and inability to reach nameplate capacity, and that the three new Western
Australian HPAL projects so far seem to be suffering a similar fate is a clear indication that
laterite projects require exceptionally high standards of engineering and technical excellence to
be successful. However, there are a number of natural attributes of nickel laterite deposits that, if
they applied in new projects, would improve their chances of successful development and
becoming profitable operations. These attributes can be described in the four categories of
quality, scale, location and infrastructure.
Ore quality

It is an old adage that good mines are made from good orebodies. Ore quality for lateritic nickel
deposits depends on factors such as:

Grade the highest possible grades of both Ni and by-products, principally Co, improve
efficiency of plant utilisation and decrease the effect of internal waste included
in the ore stream. The ability to beneficiate ores can be an advantage, although
it must be weighed up against the cost of mining more ore than is needed for the
mill.
Consistency continuity and consistency in grade and other physical and chemical
properties allows for less variability in composition of material sent to the plant.
Efficiency in the plant relies heavily on maximum control and minimum
variation in feed composition.
Ore and overburden thickness greater ore thickness and less overburden improves
stripping ratio of overburden to ore.
Mineralogy in HPAL, oxide mineralogy is preferable to clay mineralogy. Although Ni
recoveries for the two ore types are similar, in clay laterites the presence of
colloidal silica in slurries and solutions released by breakdown of the clay
causes problems with high pressure pumping and solid-liquid separation.
Higher slurry densities can be achieved with oxide mineralogy, increasing
through-put rates.
In smelting, the Si:Mg ratio in the feed is critical to controlling melt
temperatures and slag reactivity and viscosity. The ratio is strongly influenced
by mineralogy, particularly the occurrence of serpentine.
Deleterious elements in HPAL, Mg and Al are strong acid consumers, and high levels
of Al (as can be found in overburden) can cause the formation of alunite scale in
the autoclave.
Free silica occurs irregularly in places as veins and boxworks in all types of
nickel laterite. If it occurs in smelter feed, it can cause major variations in the
Si:Mg ratio, and it is therefore to be avoided.

Scale

The high capital cost of laterite plants requires large capacity treatment plants to achieve
economies of scale, and a long mine life to allow payback of capital. As a rule of thumb, a
minimum plant capacity of about 45,000 t/yr Ni is required for a viable greenfields HPAL plant
to keep unit cost of capital to a minimum (a capital cost of US$10 per pound of annual nickel
capacity is sometimes referred to as a benchmark). For a smelter, about 20-25,000 t/yr is the
minimum viable size, but this depends mostly on Ni grade and cost of energy. Resources which
show the potential to allow the definition of reserves sufficient for a mine life of 30 years are
considered necessary: this would in most instances require several hundred million tonnes of ore
for low-grade deposits, and at least 50-100 million tonnes for high-grade silicate ore.

Location

Because of the large amounts of consumables required for the operation of HPAL and Caron
plants, a coastal location for the plant is preferred. Where possible, the plant should be located
close to the minesite to minimise the transport and handling of ore. Smelting operations in some
cases are located close to energy sources or markets, with the ore being imported to the smelter.
In that case, coastal locations for the ore sources are preferable. Murrin Murrin suffers
economically from the cost of transporting 500,000 t/yr of sulphur some 800 km from the
seaboard to the minesite whereas this will not be the case for Goro, Gag Island and Weda Bay.

Infrastructure

The three main infrastructure requirements for laterite operations are water, power and access.
Water consumption of hydrometallurgical processes (Caron and HPAL) is high, but in tropical
climates water availability is often not a problem. The problem of both water availability and
quality exists in more arid locations such as inland Australia. Smelters are heavy users of power
and nearby potential sources of low-cost energy are advantageous, such as hydroelectricity or
natural gas. Access issues are important particularly in areas of rugged topography and uplifted
terrain where ore has to be transported to a plant or shipping terminal on the coast.

The need for provision of infrastructure can add greatly to the already-high capital cost pf laterite
operations. The location of the three Western Australian laterite projects close to a natural gas
pipeline has been to their advantage, and the two smaller projects, Cawse and Bulong, are located
close to the WMC Kalgoorlie Nickel Smelter from which they derive their sulphuric acid. A
coastal location alone for a plant is not necessarily an advantage by itself, unless there is a port
developed to handle materials and freight. A port can cost tens of millions of dollars to construct.

Environmental considerations

Environmental issues that need to be considered when developing new laterite projects include
mining, processing, waste disposal and closure issues (Dalvi and Poetschke, 2000). Mining of
laterite deposits is shallow (generally less than 50 metres deep) but develops a large footprint
and therefore large areas must undergo post-mining rehabilitation. In tropical areas re-vegetation
is less of a problem than in arid areas. Processing issues relate to the disposal of tailings, effluent
and emissions to the environment. Placement of tailings and disposal of effluent can be a
problem in tropical climates due to high rainfall and low evaporation rates. Deep sea tailings
disposal is technically feasible where the coast is near and sea-floor topography is suitable, but
encounters opposition from environmental groups and is in some cases not permitted by
governments.
Summary and conclusions

Nickel laterite deposits form where olivine-rich rocks are exposed to chemical weathering in
humid climatic conditions over a sufficient time to allow the concentration of nickel in various
parts of the laterite profile. Deposits form at all scales of size and degree of nickel enrichment,
but the right combination of geological and climatic factors can allow giant deposits to develop.
In summary, these are:

- large areas of exposed olivine-rich ultramafic (especially dunite and harzburgites),

such as are found in ophiolite complexes in current or former island arc and oceanic
plate collision settings,
- warm, seasonally humid tropical climatic conditions over periods in excess of one
million years,
- tectonic processes allowing a balance between rates of erosion and downward advance
of the weathering front, and development of a topography that provides for a low
water table and free drainage of the profile, and
- jointing and fracturing in the bedrock allowing penetration of groundwater.

Commercial development of nickel laterite projects is a high risk undertaking due to the high
capital costs involved and the need for the application of the highest standards of technology and
engineering. It helps to have a giant, quality orebody, but successful projects require a
favourable combination of geological, mineralogical and mining factors, technical and
engineering factors related to the process flowsheet, infrastructure-related factors and
environmental considerations (Dalvi and Poetschke, 2000). Although currently lagging behind
sulphides as sources of nickel, laterites are well positioned to increase their production levels and
lower their costs due to their huge resource position and continuing improvements in processing
technology and engineering
.

Acknowlegements

Numerous discussions with Dr C R M Butt (CSIRO, Australia) over many years have helped to
develop the ideas expressed in this paper, and I look forward to more. Dr N Brand (now with
Anglo American Exploration, Perth) is also thanked.

Comments from reviewers have also substantially improved the manuscript.

References

BRAND, N.W., BUTT, C.R.M. and HELLSTEN, K.J., 1996. Structural and lithological controls in the
formation of the Cawse nickel laterite deposits, Western Australia - Implications for supergene ore
formation and exploration in deeply weathered terrains. In: E.J. Grimsey and I. Neuss (Editors), Nickel
'96. Conference proceedings Kalgoorlie, 27-29 November 1996, Australian Institute of Mining and
Metallurgy, Publication Series No. 6/96, pp. 185-190.

BRAND, N.W., BUTT, C.R.M. and ELIAS, M., 1998. Nickel laterites: classification and features. AGSO
Journal of Australian Geology & Geophysics, 17(4): pp. 81-88.

BUTT, C.R.M. and ZEEGERS, H., 1992, Regolith Exploration Geochemistry in Tropical and Subtropical
Terrains. Handbook of Exploration Geochemistry, Volume 4, (G.J.S Govett, Editor). Elsevier,
Amsterdam, 605pp.

DALVI, A.D. and POETSCHKE, G.M., 2000. Technical and economic factors related to base metal
projects. Paper presented at The 39th Annual Conference of The Metallurgical Society of the Canadian
Institute of Mining, Metallurgy and Petroleum (CIM), Ottawa, August 20-23, 2000.

ELIAS, M., DONALDSON, M.J. and GIORGETTA, N., 1981. Geology, mineralogy, and chemistry of
lateritic nickel-cobalt deposits near Kalgoorlie, Western Australia. Economic Geology, 76: pp. 1775-1783.

EVANS, A.M., 1993. Ore Geology and Industrial Minerals. Blackwell Scientific Publications, Oxford,
390pp.

GERTH, J., 1990. Unit-cell dimensions pf pure and trace metal-associated goethite. Geochimica et
Cosmochimica Acta, 54: pp. 363-371.

GOLIGHTLY, J.P., 1981. Nickeliferous laterite deposits. Economic Geology, 75th Anniversary Volume:
pp. 710-735.

LINCHENAT, A. and SHIROKOVA, I, 1964. Individual characteristics of the nickeliferous iron

(laterite) deposits of the northeastern part of Cuba (Pinares de Mayari, Nicaro and Moa). International
Geological Congress, 24th, Montreal 1964, pt. 14, sec. 14, pp 172-187.

MONTI, R. and FAZAKERLEY, V.W., 1996. The Murrin Murrin nickel cobalt project. In: E.J. Grimsey
and I. Neuss (Editors), Nickel '96. Conference proceedings Kalgoorlie, 27-29 November 1996, Australian
Institute of Mining and Metallurgy, Publication Series No. 6/96, pp. 191-195.

NAHON, D., 1986. Evolution of iron crusts in tropical landscapes. In: Rates of Chemical Weathering of
Rocks and Minerals. Academic Press, London, pp. 169-191

NAHON, D., PAQUET, H. and DELVIGNE, J., 1982, Lateritic weathering of ultramafic rocks and the
concentration of nickel in the western Ivory Coast. Economic Geology, 77: pp.1159-1175.

PELLETIER, B., 1996. Serpentines in nickel silicate ore from New Caledonia. In: E.J. Grimsey and I.
Neuss (Editors), Nickel '96. Conference proceedings Kalgoorlie, 27-29 November 1996, Australian
Institute of Mining and Metallurgy, Publication Series No. 6/96, pp. 197-205.

SCHOBBENHAUS, C., 1986. Geologia do niquel. Principais depositos minerals do Brasil, DNPM-
CVRD vol II, pp. 251-255.
TRESCASES, J.-J., 1975. L'volution Gochimique Supergne des Roches Ultrabasiques en Zone
Tropicale. Formation des Gisements Nicklifres de Nouvelle-Caldonie. Memoire 78. ORSTOM, Paris,
259pp. (In French).

TROLY, G., ESTERLE, M., PELLETIER, B. and REIBELL, W., 1979. Nickel deposits in New
Caledonia, some factors influencing their formation. In: International. Laterite Symposium, New Orleans
1979, Society of Mining Engineers AIME, New York, pp. 85-119.

VALETON, I, BIERMANN, M., RECHE, R. and ROSENBERG, F. 1987 Genesis of nickel laterites and
bauxites in Greece during the Jurassic and Cretaceous, and their relation to ultrabasic parent rocks. Ore
Geology Reviews. 2, pp.359-404.