Petrogenesis of Cu-Ni Sulphide Ores From O'okiep

499-514 Maier web et.
al_105-108 van Dijk copy 2013/06/18 3:25 PM Page 499
W.D. MAIER, M.A.G. ANDREOLI, D.I. GROVES AND S.-J. BARNES
499
PETROGENESIS OF CU-NI SULPHIDE ORES FROM OOKIEP

AND KLIPRAND, NAMAQUALAND, SOUTH AFRICA: CONSTRAINTS
FROM CHALCOPHILE METAL CONTENTS
W.D. MAIER
Department of Geosciences, University of Oulu 90014, Finland
e-mail: [email protected]
M.A.G. ANDREOLI
South African Nuclear Energy Corporation, South Africa; School of Geosciences, University
of the Witwatersrand, Johannesburg 2000, South Africa
D.I. GROVES
Centre for Exploration Targeting, University of Western Australia, Crawley 6009, Australia
S.-J. BARNES
Department of Sciences de la Terre, Universit du Qubec Chicoutimi, Chicoutimi G7H 2B1, Canada
2012 December Geological Society of South Africa
ABSTRACT
The petrogenesis of sulphide ores in the Ookiep district has remained controversial. Based mainly on the concentration of
chalcophile metals (PGE, Cu, Ni), it is proposed that the sulphides segregated from a basaltic magma generated during melting
of sub-continental lithospheric mantle. Sulphide saturation of the magma was delayed due to relatively high f O2 until crustal
contamination occurred during the advanced stages of differentiation. The immiscible sulphide melt was entrained and fractionated
in dynamic magma conduits. Sulphides enriched in monosulphide solid solution (mss) component precipitated at depth in the
Kliprand area of southern Namaqualand to form the Hondekloof deposits, whereas the Ookiep ores crystallised at shallower
levels from highly fractionated residual sulphide liquids enriched in intermediate solid solution (iss). Sulphides of intermediate
composition occur at Ezelsfontein. In the context of this model, the Ookiep intrusions could represent the proximal magmatic
members of an IOCG suite of deposits, raising the prospect for additional IOCG deposits elsewhere in southern Africa. The model
also predicts an enhanced potential at Ookiep for undiscovered Ni sulphide ores at depth.
Introduction
Since the middle of the 19th Century and until the recent
closure of the Carolusberg mine, the sulphide ores in
the Springbok area of Namaqualand, also known as the
Ookiep Copper District (Figure 1), have been a major
source of copper for South Africa (cf. McIver et al., 1983;
Lombaard et al., 1986; Schoch and Conradie, 1990;
Cawthorn and Meyer, 1993; Clifford et al., 1995; 2012;
Van Zwieten et al., 1996; 2004; Robb et al., 1999;
Duchesne et al., 2007). The mineralisation is hosted by
the orthopyroxenitic and, to a lesser extent, noritic and
dioritic portions of composite mafic-ultramafic pipeand dyke-like bodies of the Koperberg Suite. One of the
main reasons for the profusion of research papers
written on the Koperberg Suite is the atypical nature of
the ores compared to most magmatic sulphide ores from
elsewhere, namely their unusually high Cu/Ni and Cu/S
ratios, and the abundance of bornite among the
sulphide minerals. This has led to controversy on
the origin of the Koperberg suite and its ores, with some
authors calling for a crustal progenitor (Clifford et al.,
1995; 2004; Maier, 2000; Duchesne et al., 2007), whereas
most others invoke contamination of mantle magmas by

crustal anatectic melts (McIver et al., 1983; Cawthorn
and Meyer, 1993; Van Zwieten et al., 1996; Robb et al.,
1999).
In the present study, aimed at providing some
further constraints on the genesis of the Koperberg suite,
noble metal data are presented from several deposits in
the Springbok area, including Cu-rich ores from
Carolusberg, East Ookiep, West Ookiep, Rietberg,
Spektakel, Hoits, Homeep, Jubilee, and Nigramoep
(Figure 1). New geochemical data are also presented
from more Ni-rich ores at Ezelsfontein, also in the
Springbok area, and from the apparently co-genetic
Hondekloof deposit, situated near the village of
Kliprand, ca. 120 km southeast of the Ookiep district
(Figure 1; Taylor 1990; Andreoli and Moore 1991;
Hamman et al., 1995; Andreoli et al., 2006).
The Koperberg Suite
Geographic distribution and emplacement age
For more than a century the predominantly mafic bodies
of the Koperberg Suite were thought to occur only in the
SOUTH AFRICAN JOURNAL OF GEOLOGY, 2012, VOLUME 115.4 PAGE 499-514

doi:10.2113/gssajg.115.4. 499
499-514 Maier web et.al_105-108 van Dijk copy 2013/06/18 3:25 PM Page 500
PETROGENESIS CU-NI SULPHIDE ORES, OOKIEP AND KLIPRAND, NAMAQUALAND, SOUTH AFRICA
18 00
Coperberg W
Hoits
17 45
Spektakel
Steenkampskraal
iep
O'Ok
uts
Valp
ies/
r
a
G
50
Gariep
C
30 S
km
0
tic
n
tla
30 S
nko
Stei
pf
Springbok
Vaalputs
Kliprand
5 km
Aggeneys
veld
ters
Rich
ys
ene
Agg
Nigramoep
Ezelsfontein
Springbok
East Okiep
West Okiep
Homeep E
Jubilee
Rietberg
e
Skelmfontein thrust zon
Nama Group
Rietberg Granite
Concordia Granite
Granite Gneiss
Khurisberg Subgroup
Copper mines
Steinkopf
an
e
Oc
17 E
29 S
18 E
Namibia
19 E
Homeep W
29 30
500
Figure 1. (a) Geological sketch of western Namaqualand showing localities mentioned in the text (modified after Waters, 1989; Andreoli
et al., 2006). Tectonic terranes are shown in variable shade, double dotted lines indicate clusters of Koperberg suite and related rocks, and
solid lines show metamorphic facies boundaries (A: greenschist, B: amphibolite, C: low-T granulite, D: high-T granulite), short broken line
near coastline: staurolite zone (Pan African overprint) (b) Geology and locality map of the Ookiep area.
SOUTH AFRICAN JOURNAL OF GEOLOGY
501
sample/primitive mantle
1000
Hondekloof
100
10
mineralised gabbronorite
barren gabbronorite
Cs Rb Ba Th U Ta Nb K La Ce
Sr P Nd Sm Zr Hf Eu Ti Tb Y Yb Lu
sample/primitive mantle
1000
Hypersthenites
Norites
Leuconorites
Diorites
Glimmerite
100
10
O'okiep
1
Cs Rb Ba Th U Ta Nb K La Ce
Sr P Nd Sm Zr Hf Eu Ti Tb Y Yb Lu
Figure 2. Mantle normalised multi-element diagram of incompatible trace elements, for rocks from (a) Hondekloof, and
(b) Ookiep (data from Duchesne et al., 2007). Normalization values from Sun and McDonough (1989).
Ookiep Copper District, a relatively restricted area

in northern Namaqualand where they intrude
the orthogneisses, granitoids and granulites of the
1200 to 1000 Ma Namaqualand Metamorphic Complex
(McIver et al., 1983; Lombaard et al., 1986; Cawthorn
and Meyer 1993; Van Zwieten et al., 1996; Clifford et al.,
1995; 2004; 2012). The emplacement age has
consistently been difficult to resolve from the age of
regional metamorphism, with recent work establishing a
range between ~1040 and ~1020 Ma, indistinguishable
from a lower granulite facies metamorphic peak dated
at ~1030 Ma (P ~5 kbars, T ~750 to 800C; Robb et al.,
1999; Clifford et al., 1995; 2004 and references therein).
Since the late 1980s, detailed mapping in central and
southern Namaqualand brought to light other clusters
of intrusive rocks with many of the lithological,
geochemical, emplacement age, and structural features
distinctive of the Koperberg Suite near Springbok
(Figure 1; Andreoli et al., 1986; 1994; 2006; Hamman

et al., 1996; Read et al., 2002). This southern region
experienced upper granulite facies conditions
characterized by the stability of such rare minerals like
osumilite, as well as spinel + quartz (P >5 kbars, T ~860
to 1000C; Mouri et al., 2003; Robb et al., 1999).
These articles also note some significant differences
between clusters, namely in their Sm-Nd model ages, in
the relative proportions of rock types, and their
metallogenic signature. Copper was the main resource in
the Ookiep District, replaced respectively by magnetite/
ilmenite near Vaalputs, monazite + copper at
Steenkampskraal, and nickel near Kliprand.
Field relations and lithologies
The Koperberg Suite in the Springbok area is
represented by more than 1700 irregular and
discontinuous sheet-, plug-, pipe- and dyke-like intrusive
502
Table 1. Compositional data from Okiep district and Hondekloof.

Rocktype
Ba
Sr
Cr
Ni
Cu
Ru
Rh
Pd
Ag
Os
Ir
Pt
Au
96.0
173.0
Cu-rich ores
Carolusberg
cw1905
Orthopyroxenite
59
24
14 3534
791
358
48796
1.27
93.0
1.7
cw2662
Norite
108
471
48
692
241
280
18508
0.88
44.0
0.5
cw271
Norite
117
381
64 1125
331
445
23904
0.82
2.0
0.6
29.0
0.2
13.2
50.5
cw276
Norite
106
292
85 1413
382
508
31386
0.80
2.4
1.0
44.8
0.3
19.8
60.6
cw293
Norite
197
687
48
25.0
<0.05
74.4
cw508
Orthopyroxenite
20
10
0.8
285.0
0.1
20.9
595
171
281
13407
0.36
14 3210
767
256
55470
1.78
76.3
East Okiep
e02
Leuconorite
992
699
23
34
172
216
10225
1.25
2.5
e03
Leuconorite
867
746
17
45
100
227
9990
1.12
3.5
e06
Leuconorite
242
629
35
38
83
248
12264
1.26
3.4
e08
Leuconorite
256
650
31
29
83
210
13294
1.47
5.9
0.1
5.4
<0.05
7.0
<0.05
36.9
0.5
5.6
0.1
21.4
9.0
23.2
27.5
Jubilee
j102
Pyroxene diorite
j186
Mica diorite
177
476
42 1568
457
185
22600
0.81
3.8
1192
787
13
38
333
100
0.00
4.8
j23
Pyroxene diorite
367
349
70
909
233
157
5300
0.32
4.3
j28
Mica diorite
725
698
58
31
16
176
500
0.01
5.4
j53
Pyroxene diorite
264
765
10
488
215
294
12800
0.46
j64
Pyroxene diorite
199
673
11
21
58
275
600
0.11
w02
Leuconorite
293
626
50
324
396
711
41663
1.12
w04
Leuconorite
251
667
37
436
434
572
42628
1.40
6.7
w06
Leuconorite
252
644
54
371
209
863
9140
0.30
2.1
1.1
0.6
0.2
34.3
55.2
12.6
16.3
1.6
0.5
20.5
2.8
<0.05
3.1
0.5
8.0
0.1
10.1
9.4
2.9
0.3
<0.05
1.3
0.5
1.0
0.1
0.3
West Okiep
4.9
177.8
1.6
2.5
<0.05
157.4
6.4
21.4
Hoits
U-190-275
Leuconorite
455
126
60
191
224
597
24960
2.56
3.3
10.0
0.1
36.0
U-II-230
Leuconorite
1811
365
43
178
174
623
20750
1.17
4.0
0.5
6.0
0.1
24.0
0.8
22.0
Spektakel
NW 103 124
Norite
32
156
25
396
347
438
39820
3.66
2.0
NW 103 103
Norite
102
130
20
163
108
337
9261
0.87
1.7
1.1
5.0
0.2
13.0
1.5
<0.05
3.0
1.9
Rietberg
R-1
Mica diorite
233
85
41 1165
490
673
64830
3.18
23.9
R-2 shaft
Mica diorite
1045
161
145 5297
582
1098
42230
1.49
5.0
0.9
34.4
29.0
0.2
57.0
151.9
37.0
11.0
0.3
14.0
92.5
Homeep east u-86 Leuconorite
1494
211
135 1196
164
632
13900
0.76
4.5
0.5
33.0
0.4
57.2
72
374
21 118700
7.30
33.1
0.4
331.0
92.0
0.2
686.0
233.2
67280
5.73
9.2
63.0
4.0
0.1
62.0
143.8
251.8
Homeep
Homeep west 2
Cu-ore
65 1175
Homeep west 1
Leuconorite
167
172
D1
Semi-massive Cu ore
165
370
17 5040
512
1578 142400
6.52
17.0
22.0 106.0
0.1
115.0
286 51.8
Semi-massive Cu ore
67
316
12 1088
362
948 160900
8.52
20.8
30.0
0.2
32.0
8 164.3
Leuconorite
1865
500
17
172
260
11950
0.50
3.0
9.0
8 177
Norite
170
463
11
827
224
352
11180
0.91
4.0
KNU286 48
Semi-massive Cu ore
235
232
8 1335
564
1474
92550
4.23
10.0
576 3579 1375
1517
Nigramoep
73
0.6
28.0
33.0
40.0
0.5
7.0
5.7
0.1
12.3
0.1
33.9
0.2
30.8
Intermediate ores
Ezelsfontein
i-26
Plag-mica-ap-sulf rock
33
105
170 1516 1882
935
12700 15.20
1.9
1.8
16.0
0.7
82.4
I-35
Plag-mica-ap-sulf rock 111
95
143 1876 1219
1169
2172 11.10
0.0
0.8
30.0
0.6
15.5
Ni rich ores
Hondekloof
14-19
gabbronorite
263
20
180 9885
186
2793 15.00
2.3
0.9
8.0
3.0
14-30
semi massive Ni-Cu ore 22
138
15
229 17450
141
1085 24.00
0.0
2.5
35.0
6.0
14-31.7
gabbronorite
184
19
988 3104
255
2.9
0.0
1.0
284
544
3.69
0.5
0.3
3.0
20.8
0.9
20.6
0.1
4.9
bodies comprising, in order of decreasing abundance,

quartz-bearing anorthosite, various types of diorite and
norite, and orthopyroxenite. Rocks of more alkaline
affinity,
such
as
glimmerite
and
syenite
(often refererred as shonkinite in the literature)
are less abundant (Conradie and Schoch 1986).
The mineralisation in the Springbok area consists
essentially of copper sulphides. Although mining has
now come to an end, the sulphide mineralization
produced and in reserve between 1939 and 1986
amounted to 94 million tons averaging 1.75% Cu
(Lombard et al., 1986). The mineralization is mainly
hosted by the more mafic members of the suite, namely
orthopyroxenite, norite, and, to a lesser extent, biotitebearing norite and diorite. Most economic bodies are
60 to 100 m wide and taper out at depth (Lombaard
et al., 1986). Within the pipes, the more leucocratic
phases (anorthosite-biotite diorite) are generally
concentrated at higher levels and are interpreted to have
intruded first, whereas the orthopyroxenites are more
commonly located in the deeper levels and are
interpreted to have intruded last (McIver et al., 1983;
Van Zwieten et al., 1996; Clifford et al., 1995, and
references therein). The richest ore tends to occur in the
lower portions of the orthopyroxenites, such as at
Carolusberg and Nigramoep mines (Potgieter 1996).
The intrusions are generally associated with so-called
steep structures. These are narrow, D4 deformation
zones along which the regional sub-horizontal planar
fabric established during the D2 regional deformation
has been upturned into an antiformal cusp-like shape
(Watkeys 1996; Kisters et al., 1996), possibly formed in a
transpressional/dilatational regime (Lombaard and
Schreuder 1978, Watkeys 1996). In addition, occasional
copper-bearing, high-Th-U glimmerite intrusions with
Koperberg Suite affinities were observed well to the
north of the Springbok copper district, in the Steinkopf
area (Andreoli et al., 2006).
Other mafic rocks relevant to our study that occurr in
the Springbok area include hornblende-twopyroxenebiotite granulites of the Oorkraal Suite
(formerly known as the Nuwefontein Suite: Hamman
et al., 1996; C DeBeers, personal communication 2000)
which are foliated and deformed by a pre-D3 tectonic
event. They occur as raft-like inclusions/sills in the
~1200 Ma Nababeep orthogneiss and yield an
emplacement age of 1168 9 Ma by U-Pb SHRIMP on
zircons (Lombaard et al., 1996; Robb et al., 1999;
K. Beukes, personal communication 1994). These mafic
granulites are compositionally distinct from the
more rare, hornblende-free gabbroic rocks referred
to as jotunite (Clifford et al., 2004; Duchesne et al.,
2007). The latter were dated at 1035 13 Ma and were
proposed to represent a parental chilled melt from
which the rocks and ores of the coeval (see above)
Koperberg Suite were derived by fractionation.
The geology of the Hondekloof deposits near
Kliprand (Figure 1) was investigated by Taylor (1990),
Andreoli and Moore (1991), and Hamman et al. (1996)
503
from whom the following description has been taken.

The Kliprand area falls within the spinel+quartz upper
granulite facies domain (D, Figure 1; Waters 1991) and
thus represents a deeper crustal level than the Springbok
area. In this area there are large bodies of megacrystic
charnockite and charnockitic orthogneiss, mafic twopyroxene granulites of the Oorkraal Suite (Table 1; cf.
De Beer, 2000) and an easterly-trending belt of
predominantly supracrustal rocks (Albat 1984; Moore
1989). The latter represents a zone of monoclinal
steepening and local shearing. It consists of calcsilicate
rocks, metapelites, biotite-garnet gneisses, quartzite and
ferruginous rocks belonging to the de Beer Kamiesberg
Subgroup of the Bushmanland Group (de Beer 2000).
Scattered within the belt are lenticular to irregularlyshaped bodies/sheets of melanocratic to mesocratic
biotite granulites/gneisses of the pre-to early syntectonic
Oorkraal Suite (Table 1; cf. de Beer, 2000) and less
deformed lenses/sheets/outcrops of mafic biotitehornblende-metagabbro, orthopyroxenite, gabbronorite,
leuconorite, anorthosite, diorite and glimmerite with the
characteristics of the Koperberg Suite (Table 2).
The nickel deposits occur as small lenticular gossans
along the basal contacts between poorly-banded,
granulitic gabbronorite and banded migmatitic biotitegarnet paragneiss. Many lenses occupy fold hinges and
are strung within a 1200 m wide, 3.5 km long belt of
steepened foliation. Hamman et al. (1996) reported
gradational contact relationships between sulphide-rich
and sulphide-poor mafic lithologies (Table 2) and
interpreted these bodies to be syn- to late tectonic
and co-genetic with the Koperberg Suite. In their model,
the sulphides were mobilized and concentrated during
tectonism (accompanied by pervasive shearing) to sites
of reduced strain represented by the fold hinges.
This resulted in thick sulphide-enriched crests with
long and thin attenuated, often disrupted fold limbs.
The lateral and vertical discontinuity of the sulphide
bodies, and possibly also that of the gabbronorite
host, is attributed to this deformational event
(J. Hattingh, personal communication 2008). Shallow
drilling in the early 1980s defined in situ ore
resources totalling 2 Mt at 0.88% Ni and 0.2% Cu.
More recently, however, renewed drilling has
significantly extended the reserves ( J. Hattingh, personal
communication).
Mineralogy and Petrology
Most of the basic bodies in the Springbok/Ookiep area
are relatively unaltered and are essentially composed of
plagioclase (An32-74) and orthopyroxene (En58-68). In most
bodies, orthopyroxene is significantly more primitive
than plagioclase (van Zwieten et al., 1996). Phlogopite,
quartz, magnetite, spinel (hercynite) and sulphides are
additional common minerals, whereas clinopyroxene is
rare (McIver et al., 1983; Conradie and Schoch 1986;
Brandriss and Cawthorn 1996). Apatite and zircon are
abundant trace phases and may be locally highly
enriched (e.g. at Ezelsfontein). Allanite or monazite are
504
Table 2. Koperberg Suite meta-gabbronorites.

Column
N=
1
SiO2
20.41
TiO2
1.35
Al2O3
7.46
Fe2O3 tot 53.31
MnO
0.05
MgO
4.87
CaO
1.75
Na2O
0.71
K2O
1.6
P2O5
0.16
LOI
9.02
Total
100.69
Rb
Sr
Ba
Y
Zr
Nb
La
Ce
Nd
Sm
Eu
Tb
Yb
Th
U
Co
Ni
Cu
V
Cr
Zn
Col.
Col.
Col.
Col.
Col.
Col.
Col.
Col.
Col.
Col.
1
2
3
4
5
6
7
8
9
10
2
ore
1
25.08
1.6
6.3
43.89
0.13
4.98
3.02
0.02
0.51
0.4
12.85
98.78
159
124
285
13
58
10
85.1
186
375
336
21
75
11
21.8
11.3
16.2
4.8
0.8
1.5
38.4
2.3
1016
16966
582
104
487
0.6
1.9
1.9
0.8
947
17195
2720
49
140
158
267
3
1
36.05
1.3
9.11
35.96
0.08
3.91
3.51
0.66
1.04
0.38
6.91
98.91
4
low grade ore
1
39.68
1.5
6.57
32.7
0.26
10.83
3.24
0.54
0.95
0.32
2.84
99.43
63
37
465
375
205
21.2
47.0
20.4
42.3
4.9
1.4
0.7
4.4
1.1
0.7
1.9
0.7
12432
3.2
0.8
320
6520
149
318
676
HKF 14-46
HKF 14-23.6
HKF 14 B
HKF 14 C
HKF 15-11
HKF 15-23.6
NAM 543
NAM 900
HKF 26-15
HKF 26-19
Samples 35, 36 (Duchesne et al., 2007)
NAM 902
also occasionally present in cupriferous biotite norite

ore (Andreoli et al., 2006). The rocks have highly
equilibrated textures, indicative of metamorphic
recrystallisation at granulite facies conditions. Together
with the age data summarized above this suggests
intrusion of the Koperberg Suite shortly before or during
peak metamorphic conditions (Robb et al., 1999; Clifford
et al., 2004; Andreoli et al., 2006), unlike many other
magmatic Ni-Cu deposits on Earth that formed in
5
1
43.12
5.24
14.57
12.62
0.09
7.93
5.26
1.38
4.59
1.98
1.98
97.82
392
201
730
61
226
54
57
153
109
6
7
Non mineralized
1
5
49.88
53.58
2.38
2.97
15.5
15.72
11.17
11.2
0.11
0.15
7.96
5.06
5.59
7.02
1.93
2.28
1.62
0.95
0.8
0.72
2.1
0.52
99.04
100.17
135
319
828
34
247
16
48.85
102.3
1
55.43
3.26
14.29
9.86
0.12
2.47
6.66
3.23
0.47
1.17
3.44
101.32
10
jotunite
1
2
56.01
52.86
1.14
2.40
13.76
13.85
7.9
13.98
0.08
0.20
5.87
4.12
5.31
8.24
1.8
2.70
2.77
1.49
1.58
0.73
2.1
98.23
100.57
4
5
45
101
42
398
133
0.5
1.8
2.0
3.2
7.2
1.5
34
92
18
179
609
26
128
42
196
173
66
147
172
17
330
170
110
156
47
197
840
71
418
22
57
124
65
13
3
3
6
1
1
321
41
165
38
72
242
257
142
172
118
143
ET 1
56
502
875
42
305
13
548
230
61
1814
21
59
136
85
173
2211
5604
38
717
13
279
520
257
5
6
75
15
ET 11
extensional, rift related settings (Barnes and Lightfoot,

2006).
Opaque minerals in the mineralized rocks comprise
approximately 50% magnetite (Cawthorn and Meyer
1993), unusually high for magmatic ores where
magnetite generally constitutes between 5 and 10%
of the total opaque component (Naldrett 1989).
In most bodies, the sulphide assemblage is dominated
by bornite and chalcopyrite (Carolusberg-type ore, e.g.
at Carolusberg, Rietberg, and Homeep), but in some

deposits, such as East Ookiep, Narrap, and Ezelsfontein,
pyrrhotite is abundant, minor pentlandite is present,
whereas bornite is absent (Narrap-type ore). At several
deposits, including Hoits, Nigramoep, Jan Coetzee, and
Spektakel, younger (600 to 500 Ma) metamorphic
and meteoric fluids have remobilized Cu to form
sulphides such as chalcocite and covellite, as well as
abundant sericite, epidote and chlorite (Hoits-type ore,
Clifford et al., 2012). In all deposits, the sulphides are
mainly disseminated, but may locally occur as massive
pods and as veinlets and veins, reflecting remobilization.
All rock types in the Ookiep district have a strong
crustal signature ( Nd around -9; Sri 0.7061 to 0.7272,
34S -4 to +4, and high values; von Gehlen et al., 1990;
Boer et al., 1994; Clifford et al., 1995; Van Zwieten et al.,
1996; Brandriss and Cawthorn 1996). The frequent, at
times even extraordinary enrichment of members of the
Koperberg Suite in U, Th and the REE (hosted by
accessory minerals such as zircon, monazite, allanite,
etc.) well above world averages for mafic rocks and
granulites, is also notable (Andreoli et al., 1994; 2006,
and references therein).
At Hondekloof, the host rock is mainly a mediumgrained gabbronorite consisting, on average, of
plagioclase (64%), orthopyroxene (14%), clinopyroxene
(11%), and phlogopitic mica (7%) with minor quartz
and accessory apatite and opaque minerals.
Where clinopyroxene is absent the rocks hold less
phlogopite and have a noritic composition; and where
hypersthene is less prominent, dioritic (sulphide-bearing)
compositions predominate. The plagioclase has An45-55
and orthopyroxene has En53-62. Chadacrysts of plagioclase,
clinopyroxene, phlogopitic mica, apatite and opaque
minerals occur in some of the hypersthene grains.
The clinopyroxene is colourless to pale green and
has a salite composition. It poikiloblastically encloses
most of the other minerals suggesting simultaneous
growth during recrystallization (J. Hattingh, personal
communication). The concentration of phlogopitic mica
is extremely variable, from an average of 7 up to
90 modal per cent in the glimmeritic rocks. Surprisingly,
monazite and fluorspar were found as accessories in low
grade dioritic glimmeritic ore rocks. In contrast to the
Cu-rich nature of the Ookiep sulphide assemblages,
the Hondekloof sulphides consist largely of pyrrhotite,
containing pentlandite flames and blebby chalcopyrite.
The 34S of the sulphides is -2.2 to -3.6, overlapping
with the sulphur isotopic signature of sulphides at
Ookiep.
The available data suggest that the felsic members of
the Koperberg Suite were derived from protoliths that
were isotopically heterogeneous at the regional scale.
The intrusions in the Ookiep district have Nd
significantly lower (range: -8 to -10; TCHUR ~1800 Ma)
than the monazite-bearing intrusive at Steenkampskraal
(ca. -1.8 to -1.3; TCHUR ~1050 to 1160 Ma; Andreoli et al.,
2006)
505
New results
In the present study we determined the platinum-group
elements (PGE), Cu, Ni, S, and some selected other
elements in 28 samples from the Ookiep copper district
and from Hondekloof. The data are presented in
Table 1, together with data from ten samples from
Carolusberg and East Ookiep presented previously
(Maier, 2000). All analyses were carried out at the
University of Quebec at Chicoutimi, Canada. The PGE
and Au were determined by Ni-sulphide fire assay
followed by instrumental neutron activation analysis
(INAA), using the method described in Maier and Barnes
(1999). Nickel and Cu were determined by Atomic
Absorption (AA) spectroscopy, S by LECO titration, the
remaining elements were determined by INAA using
the method of Bdard and Barnes (2002). Copper,
Ni and S contents in the samples from Carolusberg, East
and West Ookiep, and Jubilee are given in Cawthorn
and Meyer (1993).
Bulk PGE contents in the samples are highly
variable, from below detection limit to about 1 ppm.
The bulk of the PGE component consists of Pt and Pd.
Iridium contents are generally below 2 ppb. The lowest
PGE contents occur in the Ni-rich ores from Hondekloof
and Ezelsfontein, and in some Cu-rich ores, notably
from East Ookiep and West Ookiep. The most PGEenriched ores occur at Homeep, Carolusberg West, and
Nigramoep. The PGE tenors of the sulphides are highly
variable: at Carolusberg and Homeep the Pt+Pd contents
in 100% sulphide reach approximately 1 to 4 ppm,
but at most of the other localities the tenors are much
lower, notably in the Ni rich ores from Ezelsfontein and
Hondekloof.
Compared to most other magmatic sulphide ores, the
Au contents are relatively high, exceeding PGE contents
in most samples. The average Au content of all samples
is 69 ppb, compared to 29 ppb Pd and 31 ppb Pt.
Similarly high Au/PGE ratios were recorded at Caraiba,
Brazil (Maier and Barnes, 1999) and in some of the
Sudbury mines (e.g. Cu-rich sulphide stringers at Little
Stobie; Hoffman et al., 1979).
The metals correlate with S in individual deposits,
but the total sample population shows unsystematic
scatter in binary metal/S plots (Figure 3). The least
scatter is found in Cu/S and Ni/S.
The mantle-normalised PGE patterns are strongly
fractionated (Pd/Ir up to 100) with negative Ir anomalies
and a progressive increase in concentration from Ir to
Cu (Figure 4). The lowest Pd/Ir occurs in the Ni-rich
ores, at Ezelsfontein (average 37) and Hondekloof
(average 27). The Pd/Ir ratios in the Cu-rich ores are
higher than those of most other sulphides hosted by
differentiated gabbroic rocks. The closest analogies are
again the Caraiba norites-pyroxenites in Brazil (Pd/Ir 72;
Maier and Barnes 1996, 1999) and Cu-rich ores at
Sudbury (Naldrett et al., 1982).
Some new lithophile whole rock-data from the
Hondekloof gabbronorites in the Kliprand area are given
in Table 2. The bulk major-, minor and trace element
506
100000
10000
Ni (ppm)
100000
Cu (ppm)
1000000
10000
1000
1000
100
Carolusberg
East Okiep
West Okiep
Spektakel
Jubilee
Rietberg
Homeep
Hoits
Nigramoep
Eezelsfontein
Hondekloof
100
.01
.1
10
10
100
10000
.01
.1
10
100
1000
1000
Pd (ppb)
Pt (ppb)
100
100
10
10
.1
10
100
10
.1
10
100
1000
100
Ir (ppb)
Au (ppb)
1
10
.1
1
.01
.01
.1
10
100
.1
.01
S (wt%)
.1
1
S (wt%)
10
100
Figure 3. Binary variation diagrams vs S of (i) Cu, (ii) Ni, (iii) Pt, (iv) Pd, (v) Ir, and (vi) Au. Note that most of the data correlate positively
with S, but individual deposits define distinct fields.
composition of the rocks bears some similarities to the

Koperberg Suite Table 2), but also to the maficintermediate granulites and gneisses of the Oorkraal
Suite (Table 3) and the two-pyroxene granulite (jotunite)
of the Springbok area (cf. Table 2; Hamman et al.,
1996; Duchesne et al., 2007). The rocks have distinct
crustal signatures, with strong, but variable relative
enrichment in LIL and HFS elements, and negative
Nb-Ta, Sr, and Ce anomalies (Figure 2). In both the
Hondekloof gabbronorites and the Koperberg Suite in
the Springbok area the mineralised samples tend to

contain lower concentrations of incompatible trace
elements than the unmineralised samples. This trend
appears nevertheless reversed in the case of the
monazite-chalcopyrite ore at Steenkampskraal (Andreoli
et al., 1994), certain monazite/allanite-Cu sulphides ores
in the Ookiep District (Andreoli et al., 2006) and of
the monazite-fluorspar-mica-sulphide bearing diorite
(low grade ore) at Hondekloof (cf. Column 4,
Table 2).
Discussion
Unbundling the nomenclature knot.
Even a cursory appraisal of the literature on the Ookiep
copper district reveals that the origin of these mineral
deposits remains highly controversial. Because most
publications focused on the Springbok area, we contend
that at least some of the issues may be cleared by
considering the wider geographic perspective of centralsouthern Namaqualand. For instance, our observations,
both at Okiep and at Vaalputs and Kliprand, severely
constrain Duchesne et al.s (2007) model which derives
the Koperberg Suite via the fractionation of jotunite,
because field exposures of such rocks appear to be
much rarer than suggested by these authors. Duchesne
507
et al. (op. cit.) equated an occurrence of 1035 15 Ma

(Clifford et al., 2004) unmetamorphosed jotunite to the
extensive sheets of Oorkraal Suite mafic (two-pyroxene,
hornblende granulites found in the same area
(cf. Lombaard et al., 1986). However, these rocks [that is
the two-pyroxene granulites in the Springbok area] are
older, yield a 1168 9 Ma emplacement age, and a 1063
16 Ma metamorphic age (Robb et al., 1999) and have
a sub-alkaline, high-Mg tholeiitic basalt composition
(cf. Table 1; Reith and Meisel, 2001). A comparable
confusion was seemingly repeated in the Kliprand area,
where Hamman et al. (1996) attributed to the
Nuwefontein Suite both the regionally extensive sheets
of two-pyroxene granulites and a string of less
metals/primitive mantle
10000
1000
Carolusberg west
West Okiep
East Okiep
100
10
1
.1
.01
Ni
Os
Ir
Ru
Rh
Pt
Pd
Au
Cu
Ni
Os
Ir
Ru
Rh
Pt
Pd
Au
Cu
Ni
Os
Ir
Ru
Rh
Pt
Pd
Au
Cu
10000
1000
Hoits
Spektakel
Rietberg
100
10
1
.1
.01
Ni
Os
Ir
Ru
Rh
Pt
Pd
Au
Ni
Cu
Os
Ir
Ru
Rh
Pt
Pd
Au
Cu
Ni
Os
Ir
Ru
Rh
Pt
Pd
Au
Cu
Pt
Pd
Au
Cu
10000
1000
Homeep
Nigramoep
Jubilee
100
10
1
.1
.01
Ni
Os
Ir
Ru
Rh
Pt
Pd
Au
Cu
Ni
Os
Ir
Rh
Pt
Eezelsfontein
1000
100
100
10
10
.1
.1
.01
Pd
Au
Cu
Ni
Os
Ir
Ru
Rh
Rh
Pt
Pd
Au
Cu
10000
10000
1000
Ru
Ni
Os
Ir
Ru
Rh
Pt
Pd
Au
Cu
.01
Hondekloof
Ni
Os
Ir
Ru
Figure 4. Mantle-normalised metal plots for the Ookiep and Hondekloof ores. Primitive mantle normalization values are from Barnes
and Maier (1999).
508
Table 3. Whole-rock compositions of selected metagabbro and mafic to intermediate granulites of the Oorkraal suite
Column
N
SiO2
TiO2
Al2O3
Fe2O3 tot
MnO
MgO
CaO
Na2O
K2O
P2O5
LOI
Total
Rb
Sr
Ba
Y
Zr
Nb
Co
Ni
Cu
V
Cr
Zn
Metagabbro
1
2
1
2
44.28
47.03
1.21
1.12
9.27
11.27
13.31
12.60
0.18
0.18
22.42
18.98
5.05
6.42
1.6
1.48
1.09
0.66
0.32
0.17
1.41
0.34
100.14
100.23
Nuwefontein
mafic granulite
3
7
52.04
2.55
15.56
11.45
0.14
6.28
6.99
1.65
1.62
0.7
0.71
99.69
meta-dior
4
1
53.09
2.73
15.37
9.77
0.12
3.72
5.81
3.05
4.09
1.02
0.54
99.31
5
2
59.46
2.26
14.85
9.13
0.07
2.55
4.78
3.28
2.27
0.86
1.07
100.55
Vaalputs
biotite granulite
6
7
58.83
1.83
15.05
8.56
0.13
2.56
4.62
3.63
2.89
0.73
0.50
99.33
35
446
839
23
175
5
90
549
37
147
791
105
18
186
270
25
117
3
74
433
33
183
2035
93
74
723
1402
39
317
9
33
93
23
199
196
125
136
333
977
56
804
25
15
22
8
114
155
157
227
475.0
1572.0
53.0
705.0
18.0
16
13
21
98
39
131
K/Rb
258
304
182
139
106
Col.
Col.
Col.
Col.
Col.
Col.
Col.
Col.
ET 23
opx+cpx+oli+plg+bio
ET 51
NAM 541
hor+opx+cpx+plg+oli>bio
ET 10
ET 13
ET 46
ET 47
NAM 533
NAM 539A
NAM 542A
KAPKAP
KAPKAP 2
R 20
PL 20
NA 278-2
NA 278-3
NAM 222
Jotunite samples 35, 36 (Duchesne et al., 2007)
ET 48
NAM 538
1
2
3
4
5
6
7
8
deformed Koperberg Suite bodies structurally

confined to an easterly-trending monoclinal, sheared
structure (Taylor, 1991; Andreoli and Moore, 1991).
Our new data, complemented by mapping by Taylor
(1991), the Council for Geoscience (C. de Beer, personal
communication, 2008) and, most recently, by
prospecting activities for nickel (Johann Hattingh,
personal communication 2008) correlate the barren twopyroxene granulites of the Nuwefontein Suite to the
Oorkraal Suite, and the nickeliferous gabbronorites and
related rocks (anorthosite, diorite, glimmerite) to the
younger Koperberg Suite. On this basis we propose to
discontinue the use of the term Nuwefontein Suite.
Magmatic lineage of the mineralised rocks.
Previous workers suggested that the Koperberg
sulphides crystallized from sulphide melt that segregated
from an alkaline basaltic or anorthositic magma,
O'okiep
2 pyrox gran
7
8
3
2
47.08
52.86
1.52
2.40
14.22
13.85
13.17
13.98
0.19
0.20
0.41
4.12
9.08
8.24
1.75
2.70
0.83
1.49
0.27
0.73
1.69
99.22
100.57
31
203
226
35
138
7
51
166
91
297
479
109
47
197
840
71
418
22
321
41
165
38
72
143
263
NAM 539
R9
DH1-703
DH1-856.8
NA-278 samples: Reith and Mesel, 2001
in response to contamination with peraluminous

granitic country rocks (McIver et al., 1983; Schoch and
Conradie 1990; Boer et al., 1994; Van Zwieten et al.,
1996). Probably the main argument is the unusual
combination of differentiated silicate phases, enrichment
in incompatible elements (including LILE and HFSE),
and crustal Nd isotope signatures on the one hand, with
elevated Cr and Ni contents on the other hand.
An anorthositic - jotunitic lineage might be consistent
with the rare dykes of jotunite in the Ookiep Copper
district (Clifford et al., 2004; Duchesne et al., 2007) and
the observed enrichment in magnetite, apatite, relatively
sodic plagioclase, and ferrous orthopyroxene in the
Koperberg Suite. The hyper-enrichment of certain
sulphide-bearing members of the same suite in Large Ion
Lithophile (LIL) and High Field Strength (HFS) elements,
as exemplified by the ~5 x 105 tonnes of monazitechalcopyrite concentrate once mined from the
Steenkampskraal thorium mine (Andreoli et al., 1994;

2006; Schoch and Conradie, 1990) would have to be
explained by a highly trace element-enriched
contaminant, such as a small-degree melt of the country
rocks, or a sub-continental lithospheric mantle (SCLM)
component (see below).
Evidence for an alkaline signature is ambiguous.
This is because the compositional characteristics of
alkaline magmas resemble those of crustally
contaminated magmas in many regards (i.e. fractionated
incompatible element patterns, volatile enrichment,
negative Nd, high initial Sr isotope ratios, etc.) which in
turn reflects the derivation of many alkaline magmas
from refractory SCLM refertilised by crustally-derived
fluids and melts (Francis and Ludden, 1995). Partial
melts of the SCLM can be alkaline if derived by small
degree of melting (e.g. kimberlites and many alkali
basalts) or tholeiitic if derived by larger degrees of
melting (e.g. Bushveld magmas; Harmer and Sharpe,
1985; Richardson and Shirey, 2009). Therefore, all that
can be said at present on the petrogenesis of the
parental magmas to the Ookiep and Hondekloof
deposits is that they are derived from a relatively
differentiated mafic magma of basaltic lineage,
containing a large HFSE/lithophile element component,
either derived from metasomatised mantle and/or via
509
contamination of asthenospheric mantle melt by crust

(cf Andreoli et al., 2006).
Stumpfl et al. (1976) also envisaged a broadly
basaltic lineage for the deposits, but in most other
aspects their model is radically different from that of
other authors in that they suggested anatexis of a lower
crustal mafic protolith and ascent and intrusion of felsic
partial melt entraining restitic material. This model
would be consistent with the high-grade metamorphic
setting of both the Curaca Valley and Ookiep ores.
In the case of a gabbroic to dioritic protolith, the restitic
phases may consist mainly of pyroxenes (Osborn, 1978).
Crustal anatexis of a dioritic protolith would probably
yield dry melts with silica activities equal to unity, and
such melts would not be able to dissolve much P2O5, Zr,
or Fe3+, resulting in refractory behaviour of magnetite,
apatite, and zircon. Segregation of these residual
phases during ascent and intrusion may yield
the unusual ultramafic assemblages predominantly
observed in the lower portions of pipes in the Ookiep
district, namely orthopyroxene-magnetite (+apatite)
(Maier, 2000).
Models of ore formation
Past workers have modelled the Ookiep ores by various
processes, none of which is considered entirely
10 7
m
ss
10 6
Ni/Pd
silicate melt
10 5
sulfide melt R=500
sulfide melt R=5000
10 4
10 2
10 4
iss
10 3
10 5
10 6
10 7
10 8
10 9
1010
Cu/Ir
Figure 5. Plot of Ni/Pd vs Cu/Ir. Ookiep and Hondekloof ores can theoretically be modeled by a multistage process, involving sulphide
segregation from differentiated magma of picritic lineage (150 ppm Cu, 40 ppm Ni, 0.35 ppb Pd, 0.0025 ppb Ir), followed by
sulphide fractionation and separation of mss (F=0.9, D of 0.22 for Cu, 2 for Ni, 0.15 for Pd and 3 for Ir) and iss (F=0.2) components.
However, ores with Cu/Ni ratios as high as observed at Ookiep have not yet been observed elsewhere. See text for discussion, and Figure 1
for symbols.
510
Hondekloof
(1)
Eezelsfontein
Okiep
(2)
Hondekloof
Eezelsfontein
(3)
Hondekloof
Basalt
Initial sulphide melt
Mss cumulate 1
Mss cumulate 2
Fractionated iss sulfide melt 1
Fractionated iss sulfide melt 2
Figure 6. Schematic model of ore formation at Ookiep and Hondekloof. Basaltic, relatively Cu-rich and PGE-poor magma was generated
through low-degree partial melting in the sub-continental lithospheric mantle. The magma differentiated during ascent through the crust.
Sulphide saturation was triggered by contamination. Continued ascent in dynamic magma conduits led to extensive fractionation of the
sulphide liquid. (1) Relatively Ni-rich cumulate ores formed by mss fractionation in deep intrusions such as Hondekloof. (2) Residual sulphide
liquid was entrained and re-deposited at Eezelsfontein. Renewed mss fractionation produced relatively Ni-rich cumulates and Cu-rich
residual liquid. (3) Continued magma flux led to entrainment of highly fractionated sulphide melt to shallowest intrusions at Ookiep.
See text for explanatio
satisfactory to explain all the characteristics of the

deposits:
i. Based on the unusual sulphide assemblage in
the Koperberg Suite, particularly the abundance of
bornite, Cawthorn and Meyer (1993), Boer et al.
(1994), and Van Zwieten et al. (1996) proposed that
the ores formed by metamorphic oxidation and
desulphidization of a primary magmatic sulphide
assemblage derived from mantle basalt, dominated
by pyrrhotite and chalcopyrite, to an assemblage
consisting of chalcopyrite, bornite, and Ti-free
magnetite:
5CuFeS2 + 5FeS + 6H2O + [8O2] = Cu5FeS4 + 3Fe3O4 + 6H2S + 5SO2
cp
po
bn
mag
i. The model would provide an explanation for the

abundance of bornite and magnetite in the most Curich ores. Similarly Cu-rich ores in other metamorphic
terrains (e.g. at Caraiba) also are characterised
by abundant bornite and magnetite (Maier and
Barnes, 1999). However, the model provides no
explanation for the extremely high Cu/Ni ratios of
the Ookiep ores, and the low Cu/Ni at Hondekloof.
Therefore, an additional or alternative mechanism
is required.
ii. The high Cu/Ni ratios and the abundance of mica in
the mineralized rocks of the Koperberg Suite could
suggest in situ hydrothermal or metamorphic
remobilization or introduction of Cu to the mafic
rocks via an aqueous fluid phase, but this model
is considered unlikely as the mineralization is

exclusively associated with the mafic-ultramafic
rocks within a high-T, low-aH2O granulite facies
environment (Waters, 1986; Clifford et al., 1994;
Andreoli et al., 2006).
iii. The ascending basic magma may have assimilated
Cu from deep-seated Cu-bearing country rocks, for
example porphyry copper deposits within the
Eburnian-aged basement rocks to the north of
the Ookiep district (Minnitt, 1986). Copper
porphyries can contain significant Au and, in some
cases, also PGE (Economou-Eliopoulos, 2005).
The Ookiep and Hondekloof rocks have distinct
crustal isotope signatures consistent with significant
crustal contamination (van Zwieten et al., 1996;
Brandriss and Cawthorn, 1996). However, the model
is unsatisfactory as presently it cannot be directly
constrained.
iv. Based on the common high-grade metamorphic
setting of the Caraiba and Ookiep ores, Maier and
Barnes (1999) and Maier (2000) suggested that the
Cu-rich sulphides formed by partial melting of lower
crustal rocks containing Cu-rich proto sulphides
(see also Stumpfl et al., 1976; Clifford et al., 1995;
2004; 2012; Duchesne et al., 2007), followed by
entrainment and fractionation of sulphide melt in
slurries of anatectic melt and restite. The model is
consistent with the crustal signature of the silicate and
sulphide minerals, the disequilibrium compositions of,
e.g., plagioclase and orthopyroxene, and potentially
the low PGE contents and high Au/PGE ratios of the
sulphides. However, the model is more difficult to
apply to the relatively Ni-rich Hondekloof ores as any
sulphide melts that formed during partial melting of
crust might be expected to equilibrate with residual
mafic components, resulting in compatible behaviour
of Ni. Hondekloof would thus have to represent a
relatively more restitic component, inconsistent with
the composition of the silicates.
v. The Cu- and Ni-rich ores could have formed through
extreme fractionation of a magmatic sulphide liquid.
The main challenge for this model is to explain the
relatively Ni-rich Hondekloof ores. The mineral
chemistry of the Hondekloof biotite gabbros and
norites is as evolved as that of the Ookiep intrusions,
consistent with a differentiated basaltic parent melt
in both districts, but Ni contents and Ni/Cu ratios of
the Hondekloof sulphides are much higher than
at Ookiep. Compositional ranges of potential
sulphide liquids can be evaluated by simulating
fractional crystallization of basaltic-picrite magma
using thermodynamic modelling software such as
PELE. Assuming a picritic starting composition
(e.g., olivine-phyric basalt from Cape Smith;
Barnes and Picard, 1995) and lower- to mid-crustal
pressure, residual basalt crystallizes pyroxene and
plagioclase of the composition recorded at
Hondekloof after approximately 50% fractional
crystallization (in detail, the modelling is dependent
511
on H2O content and f O2, but generalized trends are

sufficient for the present purpose). At this stage of
solidification, the magma would have 3 to 4% MgO,
45 ppm Ni, and several hundred ppm Cu as long as
the magma was sufficiently oxidized so that largescale sulphide saturation was not achieved.
Assuming R factors between 500 and 5000, the
sulphide melt has about 5 to 7.5% Cu, 1 to 1.3% Ni,
and Cu/Ni ratios of at least 5, an order of magnitude
higher than Hondekloof sulphides. To produce
ores with 1.5 to 2% Ni and Cu/Ni 0.2 to 0.5, as at
Hondekloof, the crystallizing mss needs to
fractionate from the residual sulphide melt after as
little as 10% crystallization. The Cu-rich Ookiep ores
may theoretically be produced by fractionation of iss
after advanced crystallization of the sulphide liquid
(F = 0.2) (Figure 5). A concern with this model is
that Cu/Ni ratios as high as at Ookiep have not yet
been recorded in magmatic ores elsewhere.
Another problem is that the metal-rich nature of the
Ookiep sulphides requires the magma to remain
undersaturated in sulphide melt during advanced
differentiation. This is difficult to reconcile with the
low PGE contents of the ores. One could argue that
the magma merely attained temporary sulphide
saturation. This resulted in segregation of only a
small amount of sulphide melt that failed to deplete
Cu in the magma, whereas the highly chalcophile
PGE were effectively stripped from the magma.
However, due to the homogenous PGE depletion of
all deposits in the district, this model seems
implausible.
In view of the above considerations, we presently
favour a composite model incorporating new ideas
as well as elements of some of the above models, as
discussed below.
Genetic links to the IOCG group of deposits.
The Ookiep deposits bear certain similarities to the
Caraiba copper deposits of Brazil (Oliveira and Tarney,
1995; Maier and Barnes, 1996, 1999). In a global context,
the Ookiep, Caraiba and Phalaborwa copper deposits
may be rare magmatic end-members of the IOCG
class of deposits (Andreoli et al., 2006, Groves and
Vielreicher, 2001). A related deposit may be the
occurrence at Otjisuzubin in the Damara belt of
Namibia, where anomalous copper is associated with
apatite in an alkaline complex comprising alkali
pyroxenite, svite, mafic pegmatoids and syenite
(Gunthorpe and Burger, 1986). Another unique deposit,
i.e. the Vergenoeg pipe of fayalite-magnetite-fluorite
probably represents a Cu-poor variant sited in the centre
of the Kaapvaal Craton (Goff et al. 2004). This array of
unique deposits suggests that the SCLM under southern
Africa was anomalously fertile and raises the possibility
of magmatic-hydrothermal IOCG deposits in the
southern African terranes. One such deposit, previously
unrecognised as a variant of IOCG, may be the Messina
copper deposit in high-grade metamorphic rocks of the
512
Central Limpopo belt. At Messina, anomalously S-poor

assemblages of chalcopyrite, bornite and hypogene
chalcocite and native copper, associated with hematite,
occur in large breccia pipes (Jacobsen, 1974).
The deposit occurs close to known diamondiferous
kimberlite pipes at River Ranch and Venetia, and
Jacobsen and McCarthy (1975, 1976) have suggested a
genetic association with deep alkaline intrusions.
It is suspected that the potential for IOCG deposits in
southern Africa may not have been realised because of
the anomalous nature of the deposits discovered to date,
including the Ookiep deposits discussed here.
In the context of the IOCG model, the Ookiep
magmatism would have been triggered by crustal
extension, for example in response to orogenic collapse
caused by extreme radiogenic heating at mid- to lower
crustal depths (Mouri et al., 2003; Andreoli et al., 2006).
As a result of this event, small-degree partial melting of
the SCLM produced magmas that are variably enriched
in volatiles, CO2, LILE, HFS, Au, Cu, P, but also SiO2, Cr,
MgO and other compatible elements. In contrast, PGE
contents of the magma were low, analogous to many
alkaline magmas elsewhere (Maier and Barnes, 2004).
The magma ascended rapidly into the crust due to its
high volatile content. In the oxygenated magmas S was
stable as sulphate (Mungall et al., 2006) preventing early
sulphide saturation and allowing Cu and Au to be
concentrated in the magma until the advanced stages of
differentiation (Campbell et al., 1998, and references
therein). Sulphide saturation was eventually triggered by
contamination and addition of external S, causing low
34
Nd and S values. The sulphides underwent
extensive fractionation during continued magma ascent
in dynamic magma conduits, possibly favoured by
oxidation (cf Wohlgemuth-berwasser et al., 2012).
This resulted in deposition of relatively Ni-rich ores in
deep intrusions such as Hondekloof, and progressively
more Cu-rich ores with decreasing depth of
emplacement (Figure 6). Because PGE and Au are
incompatible into mss, they are enriched in the most
Cu-rich deposits. The relatively high Au/PGE ratios of
the ores mirror those of the SCLM (Maier et al., 2012).
Conclusions
The genesis of the Namaqualand Cu-Ni ores can be
linked to other Cu-Au-P-LREE enriched systems, such as
Caraiba and the IOCG group of deposits. Key
components of the model include magma generation at
craton margins, in response to melting of metasomatised
SCLM in a back arc environment. The volatile-rich,
oxidized magmas formed through relatively small
degrees of partial melting, resulting in low PGE contents.
The magmas ascend rapidly into the crust through
trans-lithospheric structures, delaying sulphide
saturation until relatively advanced stages of
differentiation, thus causing pronounced enrichment
in Cu, and incompatible lithophile elements.
Sulphide-bearing silicate rocks at Ookiep and
Hondekloof show remarkable similarity in terms of
petrography and composition of silicate minerals,

leaving little doubt that they are co-genetic. In contrast,
the sulphide ores at the two localities are of a distinctly
different composition. At Ookiep there occur mainly
Cu-rich sulphide minerals, dominated by bornite and
chalcopyrite, with Cu/Ni ratios reaching >100.
At Hondekloof there occur Ni-rich ores dominated by
pyrrhotite with Cu/Ni ratios between 0.1 and 0.5. Ores
of intermediate composition occur at Ezelsfontein.
The range of compositions can be explained by
fractionation of magmatic sulphide melt during
magma ascent and emplacement in dynamic conduit
systems. This is particularly clear at Hondekloof because
these ores are too Ni-rich to represent unfractionated
primary magmatic sulphides segregated from their
relatively differentiated host rocks. The model has
potential exploration implications as it suggests potential
in Namaqualand for more Ni-rich ores of the
Hondekloof type. The Ni-rich ores would be expected to
occur at deeper crustal levels than the Cu-rich ores,
consistent with the emplacement of the Kliprand
intrusion in upper (T ~860oC) granulite facies rocks
relative to the Ookiep ores in lower granulite
(T ~750oC) facies.
Acknowledgements
We thank Gold Fields of South Africa, the Ookiep
Copper Company, and the Foundation for Research
Development of South Africa for financial support.
Leslie Terreblanche, Koos Beukes, and several other
staff at Nababeep mine offices guided WD Maier
during fieldwork and mine visits at Carolusberg and
Nigramoep mines. Rodger Hart (iThemba Labs
Gauteng) kindly provided neutron activation analyses.
Reviews by E Ripley and S Prevec helped to improve
the paper.
References
Andreoli, M.A.G. and Moore, J., 1991. Geology and mineralogy of the
Hondekloof nickel prospect, Kliprand, Cape Province. Atomic Energy
Corporation of South Africa Internal Report GEA-968, 32pp. (2 unbound
maps).
Andreoli, M.A.G., Smith, C.B., Watkeys, M., Moore, J.M., Ashwal, L.D. and
Hart, R.J., 1994. The geology of the Steenkampskraal monazite deposit,
South Africa: implications for REE-Th-Cu mineralization in charnockitegranulite terranes. Economic Geology, 89, 994-1016.
Andreoli M.A.G., Hart R.J., Ashwal L.D. and Coetzee H., 2006. Correlations
between U, Th content and metamorphic grade in the western
Namaqualand belt, South Africa, with implications for radioactive heating
of the crust. Journal of Petrology, 47, 1095-1118.
Barnes, S.-J. and Maier, W.D., 1999. The fractionation of Ni, Cu and the noble
metals in silicate and sulphide liquids. In: R.R. Keays, C.M. Lesher,
P.C. Lightfoot, C.E.G. Farrow (Editors), Dynamic processes in magmatic
ore deposits and their application to mineral exploration. Geological
Association of Canada, Short Course Notes, 13, 69-106.
Barnes, S.-J. and Picard, C.P., 1995. The behaviour of platinum-group
elements during partial melting, crystal fractionation, and sulphide
segregation: an example from the Cape Smith Fold Belt, northern Quebec.
Geochimica et Cosmochimica Acta, 57, 79-87.
Boer, R.H., Meyer, F.M. and Cawthorn, R.G., 1994. Stable isotopic evidence
for crustal contamination and desulfidation of the cupriferous Koperberg
Suite, Namaqualand, South Africa. Geochimica et Cosmochimica Acta,
58, 2677-2687
Brandriss, M.E. and Cawthorn, R.G., 1996 Formation of anorthosite and

leucotonalite during magma hybridization in the Koperberg Suite of
Namaqualand, South Africa. South African Journal of Geology, 99, 135-151
Campbell, I.H., Compston, D.M., Richards, J.P., Johnson, J.P. and Kent, A.J.R.,
1998. Review of the application of isotopic studies to the genesis of CuAu mineralisation at Olympic Dam and Au mineralisation at Porgera, the
Tennant Creek district and Yilgarn Craton. Austrailian Journal of Earth
Science, 45, 201-218.
Cawthorn, R.G. and Meyer, F.M., 1993. Petrochemistry of the Ookiep copper
district basic intrusive bodies, northwestern Cape Province, South Africa.
Economic Geology, 88, 590-605.
Clifford, T.N., Barton, E.S., Retief, E.A., Rex, D.C. and Fanning, C.M., 1995.
A crustal progenitor for the intrusive anorthosite-charnockite kindred of
the cupriferous Koperberg Suite, Ookiep District, Namaqualand, South
Africa; New isotope data for the country rocks and intrusives. Journal of
Petrology, 36, 231-258.
Conradie, J.A. and Schoch, A.E., 1986. Petrographical characteristics of the
Koperberg Suite, South Africa - an analogy to massif-type anorthosites?
Precambrian Research, 31, 157-188.
Duchesne, J.-C., Vander Aurewa, J., Liegeois, J.-P., Barton, E.S. and Clifford,
T.N., 2007. Geochemical constraints of the petrogenesis of the Ookiep
Koperberg Suite and granitic plutons in Namaqualand, South Africa: a
crustal source in Namaquan (Grenville) times. Precambrian Research,
153, 116-142.
Economou-Eliopoulos, M., 2005. PGE potential of porphyry deposits.
In: J.E. Mungall (Editor), Exploration for platinum-group elements,
Mineralogical Association of Canada, Short Course Series V 35.
Fraser, K.J., Hawkesworth, C.J., Erlank, A.J., Mitchell, R.H. and Scott-Smith,
B.H., 1985. Sr, Nd and Pb isotope and minor element geochemistry of
lamproites and kimberlites. Earth and Planetary Science Letters,
76, 57-70.
Gibson, R.L., Robb, L.J., Kisters, A.F.M., and Cawthorn, R.G., 1996. Regional
setting and geological evolution of the Ookiep Copper District,
Namaqualand, South Africa. South African Journal of Geology,
99, 107-120.
Goff, B.H., Weinerg, R., Groves, D.I., Vielreicher, N.M. and Fourie, P.J., 2004.
The giant Vergenoeg fluorite deposit in a magnetite-fluorite-fayalite-REE
pipe: a hydrothermally altered carbonatite-related pegmatoid? Mineralogy
and Petrology, 80, 173-199.
Groves, D.I. and Vielreicher, N.M., 2001. The Phalaborwa (Palabora)
carbonatite-hosted magnetite copper sulphide deposit, South Africa: an
end-member of the iron-oxide copper gold rare-earth element deposit
group?. Mineralium Deposita, 36, 189-194.
Gunthorpe, R.J. and Burger, A.D., 1986. Geoogy and economic evolution of
the Otjisazu alkaline igneous complex, central South West Africa/
Namibia, In: C.R. Anhaeusser S. Marke (Editors), Mineral Deposits of
Southern Africa. Geological Society of South Africa, 2255-2260.
Hamman, J.N., Rozendaal, A. and Jordaan, W., 1996. Gabbro norite hosted
Ni-Cu-(Co) sulphide mineralization in southern Namaqualand and its
relationship to the cupriferous Koperberg Suite of the Ookiep Copper
District, South Africa. South African Journal of Geology, 99, 153-167.
Hronsky, J.M.A. and Groves, D.I., 2009. Towards a Unified Model for
Magmatic-Hydrothermal Gold Metallogeny with Implications for Orogenic
Gold. Abstr., Proc 10th SGA meeting, Townsville, Australia.
Jacobsen, J.B.E., 1974. Structure of the copper ore bodies and country rocks
at Artonvilla mine, South Africa. Transactions of the Institute for Mining
and Metallurgy, 83, B60-B64.
Jacobsen, J.B.E. and McCarthy, T.S., 1976. Possible late Karoo carbonatite
and basalt intrusions at Messina, Transvaal. Transactions of the Geological
Society of South Africa, 78, 153-159.
Jacobsen, J.B.E. and McCarthy, T.S., 1976. An unusual hydrothermal copper
deposit at Messina, South Africa. Economic Geology, 71, 117-130.
Jana, D. and Walker, D., 1997. The influence of sulphur on partitioning of
siderophile elements: Geochimica et Cosmochimica Acta, 61, 5255-5277.
Kisters, A.F.M., Charlesworth, E.G., Gibson, R.L., and Anhaeusser, C.R., 1996.
The relationship between antiformal and monoclinal steep structures in
the Ookiep Copper District, Namaqualand: a case study. South African
Journal of Geology, 99, 185-195.
Lombaard, A.F. and Schreuder, F.J.G., 1978. Distribution pattern and general
geological features of steep structures, megabreccias and basic rocks
513
in the Ookiep Copper District. In: Verwoerd WJ (ed.) Mineralization in

metamorphic terranes. Special Publication of the Geological Society of
South Africa, 4, 269-295.
Lombaard, A.F. and the Exploration Department staff of the Ookiep Copper
Company Ltd., 1986. The copper deposits of the Ookiep Copper District.
In: C.R. Anhaeusser and S. Maske (Editors), Mineral deposits of Southern
Africa. Geological Society of South Africa, 1421-1445.
Maier, W.D., 2000. Platinum-group elements in Cu-sulphide ores at
Carolusberg and East Ookiep, Namaqualand, South Africa. Mineralium
Deposita, 35, 422-429.
Maier, W.D. and Barnes, S.-J., 1996. Unusually high concentrations of
magnetite at Caraiba and other Cu-sulphide deposits in the Curaa valley,
Bahia, Brazil. Canadian Mineralogist 34, 717-731.
Maier, W.D. and Barnes, S.-J., 1999. The origin of Cu sulphide deposits in the
Curaca Valley, Bahia, Brazil: Evidence from Cu, Ni, Se, and platinumgroup element concentrations. Economic Geology, 94, 165-183.
Maier, W.D. and Barnes, S.-J., 2004. Pt/Pd and Pd/Ir ratios in mantle-derived
magmas: a possible role for mantle metasomatism. South African Journal
of Geology, 107, 333-340.
Maier, W.D., Peltonen, P., McDonald, I., Barnes, S.J., Barnes, S.-J., Hatton, C.
and Viljoen, F., 2011. The concentration of platinum-group elements and
gold in southern African and Karelian kimberlite-hosted mantle xenoliths:
Implications for the noble metal content of the Earths mantle. Chemical
Geology, 302-303 (2012), 119-135.
Marsh, J.S., Ewart, A., Milner, S.C., Duncan, A.R. and Miller, R.M., 2001.
The Etendeka igneous province: magma types and their stratigraphic
distribution with implications for the evolution of the Parana-Etendeka
flood basalt province. Bulletin of Volcanology, 62, 464-486.
McIver, J.R., McCarthy, T.S. and Packham, B. de V., 1983. The copper-bearing
basic rocks of Namaqualand, South Africa. Mineralium Deposita,
18, 135-160.
Minnitt, R.C.A., 1986. Porphyry copper-molybdenum mineralization at Haib
River, South West Africa/Namibia. In: C.R. Anhaeusser and S. Maske
(Editors), Mineral Deposits of Southern Africa. Geological Society of South
Africa, 1567-1585.
Mungall, J.E., Hanley, J.J., Arndt, N.T. and Debecdelievre, A., 2006. Evidence
from meimechites and other low-degree mantle melts for redox controls
on mantle-crust fractionation of platinum-group elements, Proceedings of
the National Academy of Science, 103, 12695-12700.
Naldrett, A.J., 1989. Sulphide melts - crystallization temperatures, solubilities
in silicate melts, and Fe, Ni, and Cu partitioning between basaltic magmas
and olivine. In: J.A. Whitney and A.J. Naldrett (Editors), Ore deposits
associated with magmas, Reviews in Economic Geology, 4, 5-20.
Naldrett, A.J., Innes, D.G., Sowa, J. and Gorton, M.P., 1982. Compositional
variation within and between five Sudbury ore types. Economic Geology,
77, 1519-1534.
Oliveira, E.P. and Tarney, J., 1995. Genesis of the Precambrian Cu-rich
Caraiba hypersthenite-norite complex, Brazil: Mineralium Deposita,
30, 351-373.
Potgieter, J.E., 1996. Exploration in the Ookiep Copper District, Northern
Cape Province, South Africa: an overview. South African Journal of
Geology, 99, 209-220.
Read, D., Jarvis, N.V., Williams, T., Knoper, M. and Andreoli, M.A.G., 2002.
The degradation of monazite: implications for the mobility of rare-earth
and actinide elements during low-temperature alteration. European
Journal of Mineralogy, 14, 487-498.
Robb, L.J., Armstrong, R.A. and Waters, D.J., 1999. The history of granulitefacies metamorphism and crustal growth from single-zircon U-Pb
geochronology: Namaqualand, South Africa. Journal of Petrology,
40, 1747-1770.
Schoch, A.E. and Conradie, C.A., 1990. Petrochemical and mineralogical
relationships in the Koperberg Suite, Namaqualand, South Africa.
American Mineralogist, 75, 27-36.
Stumpl E.F., Clifford T.N., Burger A.J., Zyl V (1976) The copper deposits of
the O'Okiep district, South Africa: new data and concepts: Mineral
Deposita 11, 46-70
Sun, S.-S. and McDonough, W.F., 1989. Chemical and isotopic systematics of
oceanic basalts: implications for mantle composition and processes.
In: A.D. Saunders and M.J. Norry (Editors), Magmatism in the Ocean
Basins, Geological Society, Special Publication 42, 313-345.
514
Taylor, E., 1990. The geology and Ni-metallogenesis in the Namaqualaand

Movile Belt, Kliprand, North-West Cape, South Africa. Unpublished BSc.
(Honours) Project, University of the Witwatersrand, South Africa, 69pp.
Van Zwieten, A.J.M., McCarthy, T.S. and Cawthorn, R.G., 1996.
A petrogenetic model for the Koperberg Suite: evidence from the Jubilee
Mine, Namaqualand, South Africa. South African Journal of Geology,
99, 121-134.
Von Gehlen, K., Nielsen, H., Rozendaal, A. and Jensen, M.L., 1990. Sulfur
isotopes indicate a homogenous, probably mantle source for the Ookiep
copper ores, Namaqualand. Abstract Geocongress 90, Geological Society
of South Africa, Cape Town, 582-585.
Watkeys, M.K., 1996. The Klondike steep structure, Ookiep Copper District,
South Africa. South African Journal of Geology, 99, 169-183.
Woodland, S.J., Pearson, D.G. and Thirlwall, M.F., 2002. A platinum-group
element and Re-Os isotope investigation of siderophile element recycling
in subduction zones: comparison of Grenada, Lesser Antilles Arc, and the
Izu-Bonin Arc. Journal of Petrology., 43, 121-198.
Editorial handling: S. Prevec.

Petrogenesis of Cu-Ni Sulphide Ores From O'okiep

Uploaded by

Copyright:

Available Formats

Petrogenesis of Cu-Ni Sulphide Ores From O'okiep

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Petrogenesis of Cu-Ni Sulphide Ores From O'okiep

Uploaded by

Copyright:

Available Formats

What is the controversy around the origin of the Koperberg suite and its ores?

What is the controversy around the origin of the Koperberg suite and its ores?

What new data is presented in the study?

What new data is presented in the study?

499-514 Maier web et.

al_105-108 van Dijk copy 2013/06/18 3:25 PM Page 499

W.D. MAIER, M.A.G. ANDREOLI, D.I. GROVES AND S.-J. BARNES